LIST OF CONTRIBUTORS Peter J. Buckley
Leeds Business School, University of Leeds, UK
Steve Burt
University of Stirling, Scotland
Susan Cartwright
Manchester School of Management, UMIST, UK
John Child
The University of Birmingham, UK
David Faulkner
Said Business School, University of Oxford, UK
Pervez N. Ghauri
Manchester School of Management, UMIST, UK
Marc Goergen
Manchester School of Management, UMIST, UK
Robin Limmack
University of Stirling, Scotland
Robert Pitkethly
Said Business School, University of Oxford, UK
Fionnuala Price
Manchester School of Management, UMIST, UK
Luc Renneboog
Tilburg University, The Netherlands
David M. Schweiger
University of South Carolina, USA
Phillippe Very
EDHEC, France vii
INTRODUCTION In the first volume of this series we explained our view of the need for a forum to bring together the growing but disparate literature on mergers and acquisitions. We noted in particular the apparent gulf between the finance and management literatures despite the fact that many papers in both camps were motivated by a desire to explain the unsatisfactory outcomes that result from many mergers. Judged by the response to the first volume, the idea of bringing together academics from different disciplines researching this topic has struck something of a chord, and we continue with this general theme in this volume. We start with a number of papers that pursue the issues of integration and organisational change. In the first paper, Schweiger and Very focus upon the importance of integration in determining value creation, a matter which they observe has historically been rather neglected. They suggest issues that need to be considered a priori as well as those that must be managed post completion of the deal. Related to this general theme, in the second paper Pitkethly, Faulkner and Child examine the issue of integration, using a study of the acquisition of U.K. targets by foreign acquirers. They find substantial differences in approach by acquirers of different nationalities. Faulkner, Child and Pitkethly draw on the same dataset to examine the organisational change mechanisms adopted to achieve this integration, and find that whilst clear national differences to post acquisition change exist, there is no clear best practice that emerges for improving performance, although they identify “self confidence” on the part of the acquirer as perhaps the most important factor. On a related theme, Cartwright and Price examine the attitudinal preferences of management towards foreign M&A partners, and investigates whether these have changed in the presence of growing internationalisation. They find that attitudes have changed little and that managers still prefer to merge with or be acquired by companies from countries that are perceived to be culturally similar. Our next two papers examine acquisition success or failure from two different financial perspectives. Goergen and Renneboorg examine the issue of ix
x
INTRODUCTION
value creation in large European M&As, using announcement period (short run) returns of both acquirers and targets. They then go on to investigate the characteristics of the bid that determine share price reactions. The authors find evidence consistent with synergies being the prime motivation for bids in firms that generate positive total wealth gains. However, outside this group they note that there is evidence that managerial hubris leads to poor decision making in takeovers. Adopting a different approach to financial performance in acquisitions, Burt and Limmack look at the operating cash flow performance of acquirers and targets rather than their equity returns. Their study is of takeovers in the retail industry, and they find that there is an improvement in operating performance that appears to be associated with improved asset utilisation. They note that this finding of improved performance is not incompatible with the findings from long run “event” studies as the benefits of such improved performance could have been passed on to the target firms’ shareholders. Our final two chapters are review papers. Limmack provides a comprehensive literature review on the wider issue of corporate diversification, including the more recent literature which offers a more subtle interpretation of the so-called “diversification discount”. Guari and Buckley offer a wide ranging review of M&As from a number of perspectives and seeks to analyse the recent boom in M&A activity together with examining theoretical and societal impacts of M&A. We would like to conclude by thanking our long-suffering editorial assistant, Maureen Costelloe, for her excellent efforts in assisting us with this series. Without her administrative skills the editors would doubtless still be wondering about organising the first edition rather than looking forward to the third.
NOTE FOR CONTRIBUTORS We welcome papers on all aspects of mergers and acquisitions and in future editions intend to continue with the theme of bringing together contributions from a range of disciplines spanning management, accounting, finance and economics. In the first instance, please send manuscripts (either in Word or PDF format) to Margaret Cannon at the University of Manchester Institute of Science and Technology, e-mail:
[email protected] Cary Cooper and Alan Gregory Editors
CREATING VALUE THROUGH MERGER AND ACQUISITION INTEGRATION David M. Schweiger and Philippe Very INTRODUCTION Successfully integrating mergers and acquisitions (M&As) has become an increasingly important topic both in the academic and practitioner literature during the past decade (Cartwright & Cooper, 1996; Hubbard, 1999; Galpin & Herndon, 2000; Habeck, Kroger & Tram, 2000; Marks & Mirvis, 1998; Feldman & Spratt, 1999; Schweiger, 2002; Haspeslagh & Jemison, 1991; Schweiger & Goulet, 2000; Hitt, Harrison & Ireland, 2001). Almost everyone writing on this topic acknowledges that integration is a critical part of M&A value creation. In spite of these acknowledgements, the literature has not directly demonstrated a clear linkage between value creation and the integration process. It is the primary purpose of this chapter to do so.
VALUE CREATION AND INTEGRATION DEFINED Before demonstrating any linkages it is critical that we first define what we mean by value creation and integration. Value Creation Most companies transacting M&As rely on discounted cash flow (DCF) models to value a target company (Brunner, Eades, Harris & Higgins, 1998).
Advances in Mergers and Acquisitions, Volume 2, pages 1–26. © 2003 Published by Elsevier Science Ltd. ISBN: 0-7623-1003-0
1
2
DAVID M. SCHWEIGER AND PHILIPPE VERY
A DCF model provides an acquiring firm with reasonable estimates of what a target may be worth to it. Based on this model, economic value is created when the return on capital employed in an acquisition exceeds a target’s weighted average cost of capital (Copeland, Koller & Murrin, 1995). To develop a DCF model an acquirer typically forecasts cash flows over some period of time (e.g. five years) and a terminal value for a target, and discounts those values at the weighted average cost of capital. The final valuation depends on the weighted average cost of capital chosen, cash flows forecasted (e.g. revenue, cost, net working capital, and investment forecasts), and terminal value.1 Often, multiple estimates based, on different assumptions and scenarios, are developed. Based on the DCF methodology, it is apparent that an acquirer’s ability to accurately forecast is critical to value creation. Perhaps, the ability to realize these forecasts after a deal is closed is more important to value creation. (Csiszar & Schweiger, 1994).2 This is based on the premise that prior to closing valuation is only a hypothetical exercise, although stock prices may move based on the announcement of a deal. If realization of cash flows is critical to value creation then we must understand the important elements that influence “realization.” It is the major premise of this chapter that cash flows are realized as a result of the success of the integration process. As such, the remainder of this paper attempts to demonstrate the linkage between the nature and management of the integration process and cash flow realization and value creation. Value Creation and Pricing Although DCF is a major tool employed in valuation, companies often rely on market-based measures to determine what to pay for a particular target. Market based measures of valuation establish what other comparable companies are worth or what recent comparable acquisition transactions have traded for. Essentially, these values establish a base of what companies have sold for in the market. However, they may not accurately reflect what a particular target is worth to a specific acquirer. For example, the market price may exceed the value established by a DCF model. This may be due to an active market whereby there are multiple bidders who drive up the price (but not necessarily the value) of a target (Hitt et al., 2001). At any moment a target has an intrinsic value that is based on what a firm is worth as a stand-alone entity. This value is based on the stream of cash flows it can produce as a going concern. If an acquirer pays more than this value
Creating Value Through Merger and Acquisition Integration
3
based on market prices, value is likely to be destroyed. Unless, a buyer can utilize the acquisition to improve the cash flows of either the target, itself, or both value cannot be created. In common language, this has come to be known as the concept of “synergy.” The relationship between price, synergy and value is illustrated in Fig. 1. The figure illustrates that value can be created when the price paid for a target (Price 1) is below its stand-alone value. In this case, an acquirer must ensure that value does not leak from the target. When the price (Price 2) exceeds the stand-alone value, synergies must be captured for value to be realized. In this case, changes must be made in either the target, the acquirer, or in both firms for cash flows to be improved and value to be realized. When the price (Price 3) exceeds all synergies, there is no chance that value can be created. Recent research (Sirower, 1997) has found that when premiums are paid for acquisitions, the likelihood of value creation decreases significantly. Sirower argues that often synergies are either not present, are exaggerated by a buyer, or cannot be effectively realized through integration. In this chapter we focus on one factor, the extent to which buyers can effectively integrate targets. In the next section we will define sources of synergy that have the potential to improve cash flows (Schweiger, 2002).
Fig. 1.
Pricing, Synergy and Value Creation.
From: D. M. Schweiger (2002) M&A Integration: A Framework for Executives and Managers, New York: McGraw-Hill.
4
DAVID M. SCHWEIGER AND PHILIPPE VERY
Sources of Synergy 3 There are four basic sources of synergy: cost, revenue, market power, and intangibles. Cost synergies are often the easiest to document and capture in a merger or an acquisition. Revenue, market power and intangibles on the other hand are increasingly more difficult to achieve. Each is described below. Cost Synergies Reducing costs is one clear way to increase cash flows. Historically, it has been the most common form of synergy and often the easiest type to capture. There are two types of cost synergies: (1) Fixed cost reduction; (2) Variable cost reduction. Fixed cost reduction is often associated with economies of scope and scale and productivity. It is also associated with reducing general, administrative and sales expenses through headquarters and support function consolidation, gaining economies of scale in operations, sales force and distribution optimization and reduction of transaction costs in the supply chain. Variable cost reduction is associated with increased purchasing power and productivity. Both forms of cost reduction often come with the physical consolidation of activities between the combining companies. The synergies that can be accrued here depend very much on the nature of the cost structure of the business model being employed by an acquirer. This source of synergy is likely to be present in almost every M&A. Revenue Synergies Revenue synergies are often hoped for but rarely realized in M&As. Typically, revenue synergies are associated with cross-selling products or services through complementary (i.e. non-overlapping) sales organizations or distribution channels that serve different geographic regions, customer groups or technologies. A key assumption underlying this source is that complementary markets desire the same products and services. In addition, revenue synergies can be derived from broadening a company’s products and services to provide needed bundling or a more complete offering. Critical to this synergy is to leverage complementary capabilities without additional costs. This includes: (1) Increased sales productivity by selling more volume with the same number of or fewer sales people (this may also create cost synergies). (2) Cross-selling products through complementary sales organizations and distribution channels.
Creating Value Through Merger and Acquisition Integration
5
(3) Reducing fixed new product development costs by utilizing complementary products; (i.e. reducing the per unit cost of each product or service through increased volume). Market Power Synergies This type of synergy results from the elimination of competitors and capacity from a market. This synergy has been a critical element in many mature market consolidations where there is over capacity. It allows an acquirer to maintain or increase prices in the market thereby improving margins and cash flows. Intangible Synergies This type is the most difficult to achieve. It does not easily lend itself to quantification. Intangibles include brand name extensions and the sharing of knowledge and know how. Rarely do they accrue through the physical consolidation of activities. They rely on the ability to transfer the intangible capabilities of one organization to the other (Haspeslagh & Jemison, 1991). Negative Synergy The synergies described above focus on the creation of value. However, it is important to note that M&As are interventions in organizations that can create disruptions that can destroy the intrinsic value of a firm. We call this value leakage or negative synergy. Although it is critical to manage for positive synergies, it is also important to manage against negative synergies. Rarely in a merger or an acquisition is only one form of synergy present or sought. In a large merger clearly multiple synergies are present. In a small acquisition, only one source of synergy (e.g. pre-commercial technology) may be sought. Depending on the particular deal the number and importance of synergies possible varies. Moreover, the realization of each form of synergy cannot be assumed. The integration process must be managed with the achievement of each type in mind. Synergy and Integration Whether and how much synergy exists in a particular acquisition is likely to be a function of the strategic objectives driving a deal. It is a major premise of this chapter that different objectives create different opportunities for potential synergies. Further, the realization of synergies depends on different types and levels of integration. In the next section we define integration and show how different forms intersect to create different integration issues.
6
DAVID M. SCHWEIGER AND PHILIPPE VERY
Integration Defined The integration literature recognizes that there are a variety of ways in which people and assets can be combined and has developed frameworks to explain them (Bastien & Van de Ven, 1986; Haspeslagh & Farquar, 1987; Schweiger & Ivancevich, 1987; Shrivastava, 1986; Shanley, 1988; Buono & Bowditch, 1989; Napier, 1989; Lubatkin, Calori, Very & Veiga, 1998). For the most part, the frameworks are quite similar. Thus we rely on an approach proposed by Schweiger (2002). This approach specifically recognizes that within a particular merger or acquisition different approaches can be used based on geographical areas, product lines and functions. Schweiger defines four primary approaches to integration. Consolidation. The extent to which the separate functions and activities of both an acquirer and a target firms are physically consolidated into one. Standardization. The extent to which the separate functions and activities of both firms are standardized and formalized, but not physically consolidated (e.g. separate operations may be maintained, but the operations are made identical). This is typical when acquirers formally transfer best practices across firms. Coordination. The extent to which functions and activities of both firms are coordinated (e.g. one firm’s products are sold through the other firm’s distribution channels). Intervention. The extent to which interventions are made in an acquired firm to turnaround poor cash operating profits, regardless of any inherent sources of synergy (e.g. change management team, drop unprofitable products).
STRATEGIC OBJECTIVES, SYNERGIES AND INTEGRATION: PUTTING IT TOGETHER4 Based on the discussion above it is important to understand how the strategic objectives driving a merger or an acquisition are related to the achievement of synergies and thus the different approaches to integration. Simply put, different strategies result in different sets of potential synergies with different integration challenges. Failure to manage these challenges diminishes the realization of needed cash flows and thus value creation. Below is a series of strategic objectives and the synergies underlying them. Also discussed are some of the integration challenges associated with achieving these synergies. It is important to note that an acquirer may be seeking multiple strategies and synergies in a particular merger or acquisition. More will be said about this later.
Creating Value Through Merger and Acquisition Integration
7
Consolidate Market Within Geographic Area When this objective is employed, the basic goal is to acquire competitors in the same geographic market. Depending upon the nature of the market it may include a region within a country, a country itself, a continent, or the globe. The potential sources of synergy in this strategy thus include: • Lower variable costs of raw materials through increased purchasing power. • Lower fixed costs through elimination of redundant functions; e.g. corporate staff, information technology, sales staff. • Lower fixed costs through better utilization of fixed assets; i.e. economies of scale. • Higher prices through elimination of capacity from the market place. • Economies of scope in the sales organization through sharing of products and services developed by each organization. Market consolidation requires high levels of organizational consolidation, standardization and coordination. This means that: • Most, if not all, organizational functions and activities are consolidated and standardized. • High levels of redundancy of management and employees exist and reduction in force is likely to take place at all organizational levels. Retention of key people in non-consolidated areas, or highly competent people in general, is important. • Differences in organizational culture, identity and management practices between the acquirer and the target have to be resolved. • Differences in strategy, policies, operations, brand names, etc. have to be resolved. Extend or Add Products, Services or Technologies Typically, opportunities exist to increase competitive capabilities in the marketplace by acquiring new products and services, skills and technology, and access to complementary distribution (if products/services utilize different distribution channels). The potential sources of synergy in this strategy thus include: • Lower variable costs through increased purchasing power, where acquired products and services utilize the same basic raw materials. • Lower fixed costs through elimination of redundant functions; e.g. corporate staff, human resources and information technology.
8
DAVID M. SCHWEIGER AND PHILIPPE VERY
• Lower fixed costs through better utilization of fixed assets (i.e. economies of scale) where products share basic platforms. • Lower fixed costs through better utilization of fixed investments such as advertising and brand development; i.e. economies of scope. • Increased revenue through sharing of products and services developed by each organization. • Increased market share (i.e. volume and revenue) by providing a more competitive lineup of products and services. • Increased capabilities through acquiring new technologies. Extension requires moderate levels of organizational consolidation and standardization and high levels of coordination. This means that: • Some functions and activities, especially with respect to general and administrative overhead (e.g. MIS, accounting, human resources and sales), may be consolidated and standardized. Operations may be consolidated and standardized depending upon the extent to which new products/services can be provided within the existing operating infrastructure. • Redundancy of management and employees and reduction in force are likely to take place in those areas being consolidated. Retention of key people in consolidated as well as non-consolidated areas, or highly competent people in general, are important. This may especially be the case where the value of the target is largely tied up in intangible assets such as people: for example when acquiring R&D skills. • High levels of coordination are required across organizations/sales forces if cross selling is required. • Differences in organizational culture, identity, and management practices between the acquirer and the target have to be resolved. • Differences in strategy, policies, operations, and brand names, may have to be resolved. Enter a New Geographic Market In this case, the objective is to extend the business into geographic areas where a firm has had no presence. Typically, this objective is employed by firms who are rolling up a fragmented industry and by firms who are taking advantage of market deregulation and liberalization. In roll ups the market is highly fragmented and characterized by numerous small firms with very small market shares. Usually, one or several firms see an opportunity to grow revenues and profitability. Often roll-ups begin by consolidation within a geographic area (i.e. consolidation) and then by
Creating Value Through Merger and Acquisition Integration
9
geographic extension. Market deregulation and liberalization have influenced numerous industries. Essentially they have opened previously closed markets. The potential sources of synergy in this strategy thus include: • Lower variable costs through increased purchasing power, where acquired products and services utilize the same basic raw materials. This is now accomplished on a larger geographic scope than before. • Lower fixed costs through elimination of redundant functions; e.g. corporate staff, human resources and information technology. • Lower fixed costs through better utilization of fixed investments such as advertising and brand development over a broader geographic scope. • Increased revenue through sharing of products and services developed by each organization. • Increased sales volume through geographic expansion. • Increased competitiveness by being better able to serve customers with needs for broader geographic coverage. Market entry requires low levels of organizational consolidation where there is little geographic overlap, but may require high levels of standardization and coordination. This, however, is dependent upon the extent to which the firms can take advantage of doing things the same way across geographic markets. If each market is relatively independent due to strong pressures for localization, there may be few if any opportunities for synergies. If markets are interconnected in some fashion, synergistic opportunities increase dramatically. Certainly standard products across markets would lead to significant efficiencies in new product development and operations, as well as developing a standard for a best practice. This means that: • Very few functions and activities are consolidated. Some aspects of general and administrative overhead (e.g. MIS, accounting, human resources, sales) are consolidated and standardized. • Operations may not be consolidated, depending upon the economics of the business. They may, however, be standardized to the extent that one basic approach (e.g. best practice) for providing the product/service across all markets makes sense. • High levels of consolidation and standardization within new markets (e.g. as in market consolidation) may take place if geographic entry is the first step in a regional market concentration strategy. • Redundancy of management and employees and reduction in force is likely to take place in those areas being consolidated. Retention of key people in non-consolidated areas, or highly competent people in general, is important.
10
DAVID M. SCHWEIGER AND PHILIPPE VERY
In geographic expansion retention of people is important in non-redundant areas since acquirers neither have the internal people available to staff the new operation nor have “deep knowledge” of the local area. • High levels of coordination are required across organizations/sales forces if cross-selling is required. • Differences in organizational culture, identity, and management practices between the acquirer and the target have to be resolved, unless each geographic area is different. • Differences in strategy, policies, operations, brand names, etc. have to be resolved, depending upon the degree of standardization. Vertically Integrate In this case, the objective is to enter into either sources of supply or distribution. Such moves are made to increase value added into the business or gain control over more aspects of the business. Vertical integration requires very low levels of organizational consolidation, standardization but high levels of coordination. The potential sources of synergy in this strategy thus include: • Lower variable costs of raw materials through control over raw materials and value added retained within the new company. Accounting for the costs depends on the transfer pricing agreements established within the new company. • Lower overall costs through improved product development and manufacturing interfaces. • Lower fixed costs through elimination of redundant functions; e.g. corporate staff, information technology, sales staff. Costs associated with managing the new vertical relationship internally replace costs of managing the external relationship (e.g. purchasing function). This means that: • No functions and activities are consolidated with the exception of cash management, treasury functions, and financial statements. Moreover, the senior management of the target is likely to report into the acquirer’s structure. • Operations are neither consolidated nor standardized. • Key interrelationships are established, as the new acquisition supports existing parts of the business. • Redundancy of management and employees and reduction in force is not likely, although some key people in the target may leave or be replaced. Retention of key people is important.
Creating Value Through Merger and Acquisition Integration
11
• Differences in organizational culture and identity and management practices between the acquirer and the target are not an issue, unless there are certain aspects of the acquired firm the acquirer cannot live with. • Differences in strategy and brand names have to be resolved. Enter a New Line of Business In this case, the objective is to enter businesses where the acquirer has little or no previous experience. Typically, opportunities exist to grow revenues, add distribution, and add new products/services, acquire new technologies, acquire new management talent with different perspectives. The potential sources of synergy in this strategy thus include: • Lower fixed costs through elimination of few redundant functions; e.g. corporate staff, information technology, sales staff. • Lower cost of capital for combined firm by reducing firm risk through diversification. • Intangibles such as broader pool of available management talent and business know-how. This means that: • No functions and activities are consolidated with the exception of cash management, treasury functions, and financial statements. Typically, the acquired company operates either as a: (a) division of the acquirer, whereby the senior management of the target reports into the acquirer’s structure, or (b) wholly-owned subsidiary whereby the acquirer controls the target through board representation. • Operations are neither consolidated nor standardized because there are no sources of operational synergy (e.g. operational cost reductions or crossselling opportunities). • High levels of consolidation and standardization may take place in subsequent acquisitions in the new line of business if the initial acquisition is a platform for further market concentration. • Redundancy of management and employees and reduction in force is not likely, although some key people in the target may be leave or be replaced due to historical performance. Retention of key people is important since the acquirer rarely has its own people to transfer to the target. • Differences in organizational culture, identity, and management practices between the acquirer and the target are an issue, unless there are certain
12
DAVID M. SCHWEIGER AND PHILIPPE VERY
Table 1.
Linking Strategic Objectives and Synergies. Strategic Objective
Consolidate within a geographic area
Extend or add new products, services, or technologies
Enter a new market
Vertically integrate
Enter a new line of business
Moderate Low High Low
Low None None Low
Type of Synergy Cost Revenue Market power Intangible
High Low High Moderate
Low High Moderate Moderate
Low High Low Moderate
aspects of the acquired firm the acquirer cannot live with. Differences between the senior management teams may be an issue. • Differences in strategy and brand names have to be resolved. Table 1 summarizes the five types of strategic objectives and the synergies possible within each type. In each box is presented the level of synergistic benefits present in each strategic objective.
STRATEGIC OBJECTIVES AND MAJOR INTEGRATION ISSUES As we noted earlier, effective integration is critical for cash flows to be realized and value to be created. However, researchers suggest that integration is very challenging and have identified numerous problems that acquirers face when integrating (Cartwright & Cooper, 1996; Hubbard, 1999; Galpin & Herndon, 2000; Habeck, Kroger & Tram, 2000; Marks & Mirvis, 1998; Feldman & Spratt, 1999; Schweiger, 2002; Marks & Mirvis, 1998; Haspeslagh & Jemison, 1991; Schweiger & Goulet, 2000; Schweiger & Walsh, 1990). For example, many researchers have argued that cultural differences are a major impediment to integration. When examining culture, however, researchers have failed to examine the strategic objectives or synergies being sought. In other words, they have assumed that cultural differences are always an issue in M&As. We argue that this may not be the case, or at least the magnitude of the issue may vary by strategic objective. In this section of the chapter, we identify five issues
Creating Value Through Merger and Acquisition Integration
13
gleaned from the literature that may facilitate or hinder the realization of synergies. Then we discuss contingencies created by the strategic objectives. Five Major Integration Issues Numerous issues may be faced when integrating a target. For example, designing a common information system often raises technical difficulties such as compatibility of computer hardware, general architecture, and software languages. Although these difficulties can threaten the integration or create additional costs, we focus on the emerging research that has identified the human side of M&As as a critical causal factor in integration success. The literature on integration is rather eclectic. In their review of the human side of M&As, Schweiger and Walsh (1990) identified numerous themes, research methods and selected variables examined by researchers. In order to isolate issues, we primarily relied on three reviews of past research: Buono and Bowditch (1989), Schweiger and Walsh (1990) and Schweiger and Goulet (2000). The first two reviews focused on human issues, while the last one dealt more broadly with integration issues. Based on these reviews we identified five major issues faced when integrating. These are presented below. Each issue described hereafter has been identified at least by two of these research reviews. Individual Uncertainty and Ambiguity As Haspeslagh and Jemison (1991, p. 187) stated, “the immediate postacquisition is pregnant with expectations, questions, and reservations, among the personnel and the managers of both the acquired and acquiring organizations.” During this period some employees perceive threats while others perceive opportunities. Risberg (1999) and Larsson and Risberg (1998) make a distinction between two kinds of issues: uncertainty and ambiguity. Uncertainty occurs when employees feel a lack of information. An uncertain situation can be clarified by gathering more information. Ambiguity is characterized by the inconsistency of the information provided to the employees. More communication is itself not sufficient for resolving ambiguous situations; what prevails is the consistency and clarity of the future communication flows (Feldman, 1991). Uncertainty and ambiguity explain why employees react to a merger announcement and to the inherent changes. They are concerned about their future in the combining organization. As such, these issues contribute to the fact that, on average, acquisitions are characterized by a loss of productivity; defection of competent executives, managers and employees; absenteeism; poor morale; safety problems; and
14
DAVID M. SCHWEIGER AND PHILIPPE VERY
resistance to change during the first months of the post-acquisition period (Pritchett, 1985; Schweiger & DeNisi, 1991; Cartwright & Cooper, 1997; Marks & Mirvis, 1998; Hubbard, 1999). These in turn contribute to value leakage and an inability to realize projected cash flows and synergies. Organizational Politics M&As often lead to a change in ownership for acquired firms, which often leads to changes in their organization and management practices. Power bases are also likely to shift as authority structures change and sources of power (e.g. expertise) needed in the organization change. As these happen instability is created, as employees perceive threats or opportunities; i.e. some people will perceive that they have “gained” whereas others will perceive that they have “lost.” These conditions are ideal antecedents to organization politics – that is to say “those activities taken within organizations to acquire, develop, and use power and other resources to obtain one’s preferred outcomes” (Pfeffer, 1980, p. 7). Consequently, M&As can create an excellent context for political tactics like scapegoating, controlling information, networking or manipulating people. As Pfeffer and Salancik (1977) argued, the greater the organizational politics the greater the sub-optimization within organizations; thus, if too many people jockey for their own interests, the overall firm’s performance is likely to decline. Power and politics have rarely been the direct focus of M&A research, with two exceptions. The first is Schweiger et al. (1987) who studied executive actions for managing human resources before and after a merger. They found that one of the greatest challenges for executives was to minimize warfare among employees and to avoid “playing favorites,” especially in staffing decisions. In other words, effective managers were perceived as those who avoided or minimized political behavior. The second is research on the “theory of relative standing” which has been used to explain target top-management behaviors (e.g. Hambrick & Cannella, 1993). This theory asserts that the status an employee feels for himself in a social setting is based on how he compares his status to others in a proximate social setting. According to Hambrick and Cannella (1993, p. 736), “acquired executives are placed in a new social setting in which comparisons to acquiring executives as well as comparisons to their prior situation are inevitable and salient.” This line of research suggests that the loss of standing, and resulting loss of power and stature, can lead to the turnover of executives. When this happens there may be a loss of leadership talent needed to drive the changes required to realize synergies and cash flows.
Creating Value Through Merger and Acquisition Integration
15
Finally, political behavior during a merger can foster so much internal organizational competition that executives, managers and employees fail to attend to external competition and other important market and business issues (Haspeslagh & Jemison, 1991). Again, the net result can be unrealized synergies and cash flows as customers defect to aggressive competitors. In conclusion, political behavior can lead to the loss of key people, the demotivation of others needed to implement changes to realize synergies and cash flows. Voluntary Departure of Key People Key people are those who are necessary for value preservation (e.g. relationships with key customers) or synergy realization (e.g. important technology knowledge). Therefore their retention becomes critical to the success of a merger or an acquisition. Employee or top-management voluntary turnover is seen as a consequence of what Buono and Bowditch called “dysfunctional combination-related behaviors” (1989, p. 245), citing the example of a merger where engineers and scientists left during the integration phase. Jemison and Sitkin (1986) suggested that such turnover could potentially come from acquirer arrogance. Haspeslagh and Jemison (1991) and Very (1999) underlined that those who leave are often the most talented. The reason is that they can easily find a new job. Most of the research dealing with voluntary departure focuses on target topmanagers. For example, Walsh (1988) found that acquisitions cause increased top management turnover in comparison with ordinary conditions. In most studies, researchers do not make clear the distinction between voluntary and involuntary turnover in their empirical work, although they build their framework and interpret their results from a voluntary perspective. Reviewing past research on that theme, Risberg (1999) concluded that results do not clearly help us understand why managers leave a company. But she agreed that top management voluntary turnover is a problem many acquirers face and have to overcome in order to keep valuable skills and knowledge. Therefore, departure of valuable employees is likely to threaten the intrinsic value of the integrating firms and prevents the realization of synergies and cash flows. Loss of Customers Many stakeholders are affected by an acquisition: customers, bankers, suppliers, and competitors (Csiszar & Schweiger, 1994; Schweiger, 2002). However, research on issues related to stakeholders is very poor. The limited research tends to focus on customers. Many researchers cite loss of clients as
16
DAVID M. SCHWEIGER AND PHILIPPE VERY
a major threat characterizing acquisitions, but most of them only posit relationships, interpretations or explanations for customer defection. Hax and Majluf (1996) used the Merck-Medco merger to explain how vertical integration changed a customer-supplier relationship into competitive rivalry. Such a change in the rules of the game may benefit a merging firm’s initial competitors. Customers may also leave when they perceive, like employees, uncertainty or ambiguity about the future; or when they are concerned about whether existing contracts and agreements will be honored after a deal is closed (Csiszar & Schweiger, 1994); when the merger leads to a too much concentration of their suppliers (Very, 1999); or when they have to deal with new procedures and policies (Buono & Bowditch, 1989). Whatever the motivation for exiting, retaining and satisfying the most important customers of both firms remains necessary to sustain the firm’s historical revenues, and thus avoid value leakage. Moreover, loss of customers can also affect expected synergies like cross-selling that is aimed at enhancing revenues and thus cash flows. Cultural Resistance Schweiger and Goulet (2000), in their review of the literature, conclude that culture is a complex issue. Organizational cultural differences and clashes are identified by most researchers and practitioners as primary cause of M&A failure, both in domestic and cross-border deals. However, recent research comparing domestic and cross-border deals, suggests alternative findings (e.g. Larsson & Finkelstein, 1999; Morosini et al., 1998; Very et al., 1997; Weber et al., 1996). This research tends to show that the existence of cultural distance might not be directly associated with poor performance. Conclusions indicate that the relationship with performance is more complex than initially assumed. For example Very et al. (1997) found that the level of autonomy given to an acquired firm influences the culture-performance linkage. Some researchers introduced a cultural process, called acculturation, to explain performance (e.g. Nahavandi & Malekzadeh, 1988; Larsson, 1993; Very, Lubatkin & Calori, 1996). These researchers contend that the success of integration may depend more upon how cultural integration is managed than upon initial cultural similarities (Schweiger, 2002). Using this process perspective, Larsson (1993) connects acculturation to the reduction of conflict. In brief, questions remain about the conditions under which cultural problems and their interplay with other dimensions occur and how they influence performance. The integration process at least seems to moderate the relationship. Despite these unanswered questions, research has
Creating Value Through Merger and Acquisition Integration
17
shown that when cultural incompatibility exists, employee resistance emerges. As a consequence, target top management turnover is likely to increase and acquirers can face strong barriers for implementing their integration plan (e.g. Schweiger & Goulet, 2000). In other words, cultural resistance may have a negative impact on synergy realization and cash flows if cooperation between combining firms is not achieved. Characteristics of Integration Issues The five issues discussed above are clearly not independent, For example, uncertainty and ambiguity can lead to the departure of key people, but such departure can also emanate from unfavorable organizational politics or a recruitment opportunity offered by a competitor. Relationships among the issues would be of interest if we were studying interventions for solving those problems. Since each dimension can be either independent or linked, we considered each one separately so as to isolate the specific challenge they create. Each issue is likely to have a negative effect on value preservation and/or synergy realization, and thus cash flows. For example, individual uncertainty and ambiguity decrease the productivity of employees, diminishing the firm’s cash flows (value preservation). Moreover, when such a loss of productivity happens, the realization of synergies can be threatened: employees will not easily share their competencies as long as they are concerned about whether they will be retained. The same analysis can be made for the four other issues. However, as we will see below, the importance of a particular issue depends upon the strategy behind the acquisition.
LINKING ISSUES TO STRATEGIC OBJECTIVES Research on acquisitions has neglected incorporating integration issues into a contingency framework, with some notable exceptions (e.g. Bower, 2001; Hunt, 1990; Haspeslagh & Jemison, 1991). For example, most researchers view cultural issues as being pervasive across all types of deals. Therefore, they have generally been studied on populations of deals with little regard to the nature of those deals. In the following paragraphs, we will analyze the conditions under which the five issues described above affect integration are most salient5. The five issues focus primarily on people’s reactions to change. From the perspective of an acquired firm, the arrival of a new owner can be perceived as troublesome, depending on the nature and size of the organizational changes
18
DAVID M. SCHWEIGER AND PHILIPPE VERY
that are required. For Ansoff (1989) citing Macchiavelli, resistance to change is proportional to the degree of discontinuity in the culture and power structure introduced by the change. Resistance comprises cultural and social aspects, at both the individual and collective level. Tichy (1982) characterizes strategic change as a period favoring the redistribution of power. Consequently, the importance of issues such as organizational politics, cultural resistance and loss of customers is likely to depend upon the nature and size of changes. The integration approaches described earlier differ according to the level of changes that they induce. While coordination suggests slight changes in an organization, standardization suggests the use of common procedures and practices, and consolidation suggests physical changes due to the combining or sharing of common resources. We could deduce that, the more that integration mechanisms are based on consolidation, the greater the cultural resistance, organizational politics and loss of customers. Facing a change in ownership, people need assurances about their future.6 Argyris (1990), Pondy, Boland and Thomas (1988) have noted that any strategic shift is unavoidably accompanied by individual fears. Consequently, whatever the acquisition strategy, individual uncertainties and ambiguities should always be considered important issues. The fifth issue, departure of key-employees, is also a consequence of change: the greater the change, the greater the probability of voluntary departure. But, regardless of the nature and depth of the implemented changes, retention of key employees often remains a major issue. Key employees possess skills needed to realize synergies and cash flows. The acquisition strategy helps identify key employees needed for value to be created. For instance, when entering a new line of business, an acquirer has generally little knowledge of how to manage this new field. Therefore, an acquisition provides a way to learn from some target top-managers who constitute inimitable resources for the acquirer. When consolidating a market within a geographical area, the acquirer often seeks to increase its market share, eliminate capacity, thanks to consolidation and standardization of best practices. Consequently, those target employees that master best practices (if any at the acquired firm) need to be retained. In sum, there are generally some people that need to be retained, those who possess valuable competences according to the acquisition strategy. Consolidating a market within a geographical area requires high levels of consolidation, standardization and coordination. Therefore, organizational politics and cultural resistance are likely to emerge and slow the integration process. Changes can also threaten customers, if their relationship with a firm is modified. The adoption of best practices suggests that target employees who possess the required skills must be retained.
Creating Value Through Merger and Acquisition Integration
19
When extending products, services or technologies, synergies should be achieved primarily through coordination and moderately through standardization and consolidation. Therefore, changes in power structure and culture should essentially affect consolidated sub-units in both organizations. As the objective of an acquisition is often to extend the scope of products or services offered to customers, they should positively perceive the acquirer’s strategy. Departure of key-people is problematic if the acquirer has not successfully secured and integrated knowledge about the target products and technologies. When entering a new geographic market, the extent of synergies and the integration choices depend upon the pressures for developing a locally dedicated strategy. When entry corresponds to the implementation of a geographic concentration strategy, an acquirer could face difficulties similar to those pertaining to consolidating market within geographic area. When an acquisition is a first step in a very different market, the adaptation of the strategy, the low levels of synergy and the logical choice of coordination mechanisms militate for slight changes in the target. Consequently, as the organization remains quasi unchanged, internally and in its relation with its clients, few integration issues are likely to emerge. One main challenge remains, however: the retention of skilled top-managers from the target, since the acquirer has little or no knowledge of the new country. When vertically integrating, an acquirer looks to efficiently coordinate both firms’ activities. Therefore, there is a low risk of organizational politics and cultural resistance. When vertically integrating downstream, an acquirer instantly modifies its relationships with its customers. Customer, however, could indeed become competitors. In such cases, customers are likely to feel uncomfortable with the strategic move and could choose to change their procurement source.7 As vertical integration can be equated to entry into a new business, an acquirer needs to learn about this new field as well. Thus, the retention of key managers, again, appears to be important. When entering a new line of business, the unrelatedness with existing businesses restricts synergies to intangible ones (e.g. managerial knowledge). Therefore, few changes are likely be made to the power structure and culture of the organizations. Customers may even be different. However, as an acquirer has no previous knowledge about a new business, retention of target topmanager is likely to be important. Table 2 summarizes our propositions by delineating the integration changes, strategic objectives and issues that need to be managed. Our approach is not deterministic. The propositions essentially underline the linkages between strategic objectives, synergies, integration approaches and integration issues. Variables other than strategic objectives can influence the
Integration changes Strong change due to: • high level of consolidation • high level of standardization • high level of coordination Moderate change due to: • moderate level of consolidation • moderate level of standardization • high level of coordination Slight to moderate changes due to: • low level of consolidation • moderate level of standardization • high level of coordination Slight changes due to: • low level of consolidation • low level of standardization • high level of coordination Slight changes due to: • low level of consolidation • low level of standardization • high level of coordination
Consolidate within geographic area
Extend or add product or service
Enter new geographic market
Vertically integrate
Enter new line of business
Individual fears Organizational politics in consolidated areas Cultural resistance in consolidated areas Departure of product/service experts at the target
• • • •
• Individual fears • Departure of business experts at the target
• Individual fears • Loss of customers (when acquiring downstream) • Departure of business experts at the target
• Individual fears • Departure of target managers with local geographic or country expertise
Individual fears Organizational politics Cultural resistance Loss of customers Departure of employees mastering best practices
• • • • •
Major issues that need to be managed
Acquisition Strategy, Integration Changes and Contingent Issues to Manage.
Acquisition strategy
Table 2.
20 DAVID M. SCHWEIGER AND PHILIPPE VERY
Creating Value Through Merger and Acquisition Integration
21
design of integration plans and the type of issues faced. The friendliness or hostility of a transaction, the price premium paid, the relative size of the target company, and the prior performance of the target could certainly influence integration choices and employee reactions. Also, the industry might influence the type of issues faced by the acquirer. When the value of a firm resides in its people, such as in investment banking, retention of key employees is of particular importance, regardless of the strategic objectives. A limit to our framework is that we have restricted it to five issues. There are, however, other potentially salient issues such as the loss of suppliers, the reaction of competitors, and the emergence of a collective political or ideological resistance at the target that are important. Consequently, our framework is intended to be illustrative, not exhaustive.
CONCLUSION AND IMPLICATIONS That M&As, on average have not lived up to the financial expectations of those transacting them has become a common theme in both the academic and practitioner literatures. Many reasons have been advanced for this, ranging from lack of strategic fit of acquisition or merger partners to poor due diligence. In this chapter we demonstrated how one historically neglected element of the M&A process, integration, impacts the success of M&As. Specifically, we illustrated the complex relationship between valuation, pricing, strategic objectives, synergies and integration. We also presented a framework that demonstrated the contingent relationships among different strategic objectives, synergies and the integration issues that ensue. The ideas presented in this chapter were not only intended to help you understand the integrations issues that must be managed after a deal closes (i.e. during the integration phase of a merger or an acquisition), but were also intended to provide important insights into the issues that need to be considered a priori during the transaction and transition phases (see Schweiger, 2002). It is a major premise of this chapter that integration issues must be considered during all phases of the M&A process to ensure that valuation and pricing decisions are realistic and financial results are realized. Three issues clearly illustrate this. First, before the closing of a deal, dealmakers must factor integration issues into their valuations (e.g. DCFs). Specifically, they should examine how different issues impact the success of changes needed in the acquirer, the target, or both firms and the impact they have on subsequent cash flows and earnings.
22
DAVID M. SCHWEIGER AND PHILIPPE VERY
To the extent that changes lead to positive synergies value is created. Conversely, to the extent that they lead to value leakage (i.e. negative synergy) value is lost. Therefore, different integration scenarios should be developed, leading to a series of more realistic M&A valuations and prices. Second, before or after the closing of a deal, an acquirer can intervene (the fourth integration mechanism of Schweiger’s, 2002 framework) to prepare and implement actions aimed at dissipating or avoiding the emergence of potential threatening issues. For example, when an acquirer fears strong cultural resistances from a target it can engage interventions such as discussions, intercultural mirroring workshops, gathering employees from both companies, organizing reciprocal visits at worksites, work on future common values that will characterize the new group to mitigate the resistance. Third, many acquirers seek to learn from their past experiences. Identifying issues early in the M&A process can subsequently help identify any contextsimilar deals an acquirer has faced in the past. According to Zack (1999) context is important since the knowledge gained from dissimilar types of deals may not generalize very well. The contingency framework presented in this chapter defines one such context that may be helpful. An acquirer may then look at how it managed issues in the past and whether it did so successfully or not and why. The insights gained from context may help an acquirer develop more realistic integration plans and even choose managers who have successfully handled these issues before. These managers could then lead or take significant roles in subsequent integrations. Based on these reasons it is apparent that an acquirer could greatly benefit from associating probabilities to the occurrence of future integration issues and from measuring the potential impact these issues have on the future realization of synergies. The framework that we presented in this chapter could be helpful, even though it identified only five challenges. In terms of research, future investigations on integration should clearly differentiate acquisitions with regard to their financial and strategic objectives when trying to explain integration outcomes and performance. It is important that research differentiate among types of M&As so that similarities and differences in integration issues among deals are clearly understood when examined. We believe that this will lead to a much richer set of findings with significant practical value. Working on one type should be valuable for identifying the specific challenges and actions that need to be managed. Such an approach based on a typology would avoid the growing confusion that has characterized past research – for example under what strategic objectives is cultural resistances likely to occur.
Creating Value Through Merger and Acquisition Integration
23
Future research should also examine the complex relationship between integration and financial outcomes. For example, one could examine valuation models that were employed to price a deal and the results that were subsequently realized. Gaps could then be examined to determine why they occurred. For example, where revenue synergies (e.g. cross selling) were projected, research could determine whether they were realized and why or why not. Clearly, this would require in depth analysis with a high degree of access to an acquirer. But, without such research it may be very difficult to acquire a detailed understanding of these issues. It was the intent of this chapter to provoke thinking and to stimulate additional research in the area of M&A integration. It was also the intent to link a number of important variables that have heretofore been treated somewhat independently. Although the chapter has some limitations (e.g. the breadth of challenges considered) we hope that we have stimulated you to participate in and help extend this area of inquiry. Clearly, M&A integration research is in its infancy with additional research needed to help us understand how it impacts value creation.
NOTES 1. Although beyond the scope of this paper it is important to note that there is much debate concerning the use of terminal values and which approach is best. As, such the value of a firm can vary greatly depending upon which approach (e.g. P/E multiple, perpetuity value) is chosen. Further the weighted average cost of capital can also greatly affect valuation and can vary depending upon the percentage of debt and equity used to finance the acquisition. 2. This argument presumes that the market (in the case of publicly traded companies) recognizes the importance of return on capital and cost of capital in its trading activity. 3. Much of this section of the chapter is drawn from David M. Schweiger, “M&A Integration: A Framework for Executives and Managers,” New York: McGraw-Hill, 2002. 4. Much of this section of the chapter is drawn from David M. Schweiger, “M&A Integration: A Framework for Executives and Managers,” 2002, McGraw-Hill. 5. A salient issue is one that could threaten the value creation process. 6. We do not mean that everybody in both organizations will feel uncomfortable with the future. We mean that, in any deal, a substantial proportion of individuals will feel periods of uncertainty and ambiguity. Some people could see career opportunities as the most important direct consequence of the acquisition. These opportunities will guide their behaviors. 7. It should be noticed that the suppliers of an acquirer could react the same way when the deal corresponds to an upstream vertical integration. We focused on loss of customers because this is the one most often identified in the literature.
24
DAVID M. SCHWEIGER AND PHILIPPE VERY
REFERENCES Ansoff, H. I. (1989). Implanting strategic management. Englewood Cliffs NJ: Prentice Hall. Argyris, C. (1990). Overcoming organizational defenses: Facilitating organizational learning. Boston: Prentice Hall. Bastien, D. T., & Van de Ven, A. W. (1986). Managerial and organizational dynamics of mergers and acquisitions. SMRC Discussion Paper No. 46. University of Minnesota, Strategic Management Research Center. Bower, J. L. (2001). Not all M&As are alike – And that matters. Harvard Business Review, 79(3), 93–101. Bruner, R. F., Eades, K. M., Harris, R. S., & Higgins, R. C. (1998). Best practices in estimating the cost of capital: Survey and synthesis. Financial Practice and Education, 8, 13–28. Buono, A. F., & Bowditch, J. L. (1989). The human side of mergers and acquisitions. San Francisco, CA: Jossey-Bass. Cartwright, S., & Cooper, C. L. (1996). Managing mergers, acquisitions and strategic alliances: Integrating people and cultures. Oxford: Butterworth-Heinemann Ltd. Copeland, T., Koller, T., & Murrin, J. (1995). Valuation: Measuring and managing the value of companies. New York: Wiley. Csiszar, E. N., & Schweiger, D. M. (1994). An integrative framework for creating value through acquisition. In: H. E. Glass & B. N. Craven (Eds), Handbook of Business Strategy (pp. 93–115). New York: Warren, Gorham and Lamont. Feldman, M. S. (1991). The meanings of ambiguity: Learning from stories and metaphors. In: P. J. Frost, L. F. Moore, M. R. Louis, C. Lundberg & J. Martin (Eds), Reframing Organizational Culture. Beverly Hills, CA: Sage. Feldman, M. L., & Spratt, M. F. (1999). Five frogs on a log. New York: Harper Business. Galpin, T. J., & Herndon, M. (2000). The complete guide to mergers and acquisitions. San Francisco: Jossey Bass. Habeck, M. M., Kroger, F., & Tram, M. R. (2000). London: Financial Times/Prentice Hall. Hambrick, D. C., & Cannella, A. A. (1993). Relative Standing: A Framework for Understanding departures of acquired executives. Academy of Management Journal, 36(4), 733–762. Haspeslagh, P. C., & Farquhar, A. B. (1987). The acquisition integration process: A contingent framework. Paper presented at the meeting of the Strategic Management Society, Boston. Haspeslagh, P. C., & Jemison, D. B. (1991). Managing acquisitions: Creating value through corporate renewal. New York: The Free Press. Hax, A. C., & Majluf, N. S. (1996). The Strategy Concept and Process: A Pragmatic Approach (2nd ed.). Upper Saddle River, NJ: Prentice Hall. Hitt, M. A., Harrison, J. S., & Ireland, R. D. (2001). Mergers & acquisitions: A guide to creating value for stakeholders. New York: Oxford University Press. Hubbard, N. (1999). Acquisition strategy and implementation. London: MacMillan. Hunt, J. W. (1990). Changing pattern of acquisition behaviour in takeovers and the consequences for acquisition processes. Strategic Management Journal, 11, 69–77. Jemison, D. B., & Sitkin, S. B. (1986). Corporate acquisitions: A process perspective. Academy of Management Review, 11, 145–163. Larsson, R. (1993). Barriers to acculturation in mergers and acquisitions: Strategic human resource implications. Journal of European Business Education, 2(2), 1–18.
Creating Value Through Merger and Acquisition Integration
25
Larsson, R., & Finkelstein, S. (1999). Integrating strategic, organizational and human resource perspectives on mergers and acquisitions: A case survey of synergy realization. Organization Science, 10(1), 1–26. Lubatkin, M., Calori, R., Very, P., & Veiga, J. (1998). Managing mergers across borders: A twonation exploration of a nationally bound administrative heritage. Organization Science, 9(6), 670–684. Marks, M. L., & Mirvis, P. H. (1998). Joining forces. San Francisco: Jossey-Bass. Morosini, P., Shane, S., & Singh, H. (1998). National cultural distance and cross-border acquisition performance. Journal of International Business Studies, 29(1), 137–158. Nahavandi, A., & Malekzedah, A. R. (1988). Acculturation in mergers and acquisitions. Academy of Management Review, 13, 79–90. Napier, N. K. (1989). Mergers and acquisitions, human resource issues and outcomes: A review and suggested typology. Journal of Management Studies, 26, 271–289. Pfeffer, J. (1980). Power in organizations. Boston: Pitman. Pfeffer, J., & Salancik, G. D. (1977). Administrator Effectiveness – The Effects of Advocacy and Information on Achieving Outcomes in an Organizational Context. Human Relations, 30(7), 641–659. Pondy, L. R., Boland, J. R., & Thomas, H. (1988). Managing ambiguity and change. New York: Wiley. Pritchett, P. (1985). After the Merger: Managing the Shockwaves. New York: Dow Jones – Irwin. Risberg, A. (1999). Ambiguities Thereafter: An Interpretive Approach to Acquisitions. Lund University Press, Malmö. Schweiger, D. M (2002). M&A Integration: A Framework for Executives and Managers. New York: McGraw-Hill. Schweiger, D. M., & Goulet, P. (2000). Integrating acquisitions: An international research review. In: C. Cooper & A. Gregory (Eds), Advances in Mergers and Acquisitions (vol. I, pp. 61–91). Greenwich, CT: JAI/Elsevier Press. Schweiger, D. M., & DeNisi, A. S. (1991). Communication with employees following a merger: A longitudinal field experiment. Academy of Management Journal, 34(1), 110–135. Schweiger, D. M., & Ivancevich, J. M. (1987). The effects of mergers and acquisitions on organizations and employees: A contingency view. Paper presented at the meeting of the Strategic Management Society, Boston. Schweiger, D. M., Ivancevich, J. M., & Power, F. (1987). Executive action in managing human resources before and after being acquired. Academy of Management Executive, 1, 127–138. Schweiger, D. M., & Walsh, J. P. (1990). Mergers and acquisitions: An interdisciplinary view. In: K. Rowland & G. Ferris (Eds), Research in Personnel and Human Resources Management (vol. 8, pp. 41–107). Greenwich, CT: JAI Press Inc. Shanley, M. T. (1988). Reconciling the rock and the hard place: Management control vs. human resource accommodation in acquisition integration. Unpublished paper, Graduate School of Business, University of Chicago. Shrivastava, P. (1986). Postmerger integration. Journal of Business Strategy, 7, 65–76. Sirower, M. L. (1997). The synergy trap. New York: The Free Press. Tichy, N. M. (1982). Managing strategic change. New York: Wiley Interscience. Very, P. (1999). Ce Que Ne Disent Pas Les Chiffres. L’expansion Management Review, 93, 70–74. France.
26
DAVID M. SCHWEIGER AND PHILIPPE VERY
Very, P., Lubatkin, M., & Calori, R. (1996). A cross-national assessment of acculturative stress in recent European mergers. International Studies of Management and Organization, 26(1), 59–88. Very, P., Lubatkin, M., Calori, R., & Veiga, J. (1997). Relative Standing and the Performance of Recently Acquired European Firms. Strategic Management Journal, 18(8), 593–614. Walsh, J. P. (1988). Top management turnover following mergers and acquisitions. Strategic Management Journal, 9, 173–183. Weber, Y., Shenkar, O., & Raveh, A. (1996). National and corporate cultural fit in mergers/ acquisitions: An exploratory study. Management Science, 42(8), 1215–1227. Zack, M. E. (1999). Managing codified knowledge. Sloan Management Review, 40(4), 45–58.
INTEGRATING ACQUISITIONS Robert Pitkethly, David Faulkner and John Child INTRODUCTION During the 1990s there was significant direct investment into the U.K. by foreign firms (Child et al., 1998). Between 1986 and 1995 companies from the USA, Japan, Germany and France were the largest sources of foreign investment, accounting for between 70.8% and 80.9% of the total. In 1996, foreign acquisitions of U.K. firms exceeded the total value for all other European Union countries combined, and the U.K. was topped world-wide as a take-over target only by the USA (KPMG, 1997). An extensive search using numerous sources covering the period 1985 and 1994, identified 1,422 U.K. activities comprising new acquisitions, joint venture formations, collaborations or consortia involving foreign investment, but excluding greenfield developments or expansions of existing facilities. It found that 79% of these activities involved acquisitions.1 This chapter concentrates on the key issue of integrating acquisitions using a study of acquisitions of U.K. companies by companies from the USA, Japan, Germany and France. Four major areas are involved. These are overall level of integration, the control and communication methods adopted, and the strategic philosophy pursued by the new parent company in relation to the subsidiary.
CATEGORIES OF CHANGE Acquisition provides a potentially powerful lever for the direct application of foreign management practice insofar as 100% ownership legitimates foreign owners’ authority. Moreover, the greater investment demands of outright Advances in Mergers and Acquisitions, Volume 2, pages 27–57. Copyright © 2003 by Elsevier Science Ltd. All rights of reproduction in any form reserved. ISBN: 0-7623-1003-0
27
28
ROBERT PITKETHLY, DAVID FAULKNER AND JOHN CHILD
acquisition encourage acquirers to devote greater attention to their new subsidiaries. Acquisitions also weaken the ability of local management to resist the introduction of new practices more than do joint ventures or collaborations. Greenfield sites might offer even less resistance, but do not offer such a clear comparison of practices prevailing before and after FDI. Many factors may bear upon changes in management practice in subsidiaries following acquisition. Firstly, there may be “background changes” which would have occurred anyway even in the absence of an acquisition. These changes in management practice may come from general conditions affecting U.K. industry during the period of study, be they the influence of new management ideas or the economic cycle of boom and recession. Secondly, however there may be various “acquisition effects” which only occur following acquisition. Amongst these are changes which would have occurred anyway eventually and which may, following an acquisition, be catalysed by it and proceed faster or more effectively. New investment may be made in plant and information technology, providing opportunities for new practices to be introduced. The rationale for many acquisitions is to exploit perceived opportunities for securing a greater return from assets, and this may bring about a further impetus for change in management practice within the subsidiary. There is also likely to be a general “new broom sweeps clean” effect and some changes implemented which would never have occurred in the absence of acquisition. The acquisition effect is however characteristic of acquisitions per se, rather than reflecting any particular foreign approach to management and organisation. It comprises both changes which would have occurred anyway but which are catalysed by the acquisition, as well as changes which occur directly as a result of the acquisition but which do not differ by nationality of the acquirer. Thirdly, there is change that is specific to the nationality of the new owner, which may be called the “transfer of foreign practice effect”. This comprises the transfer of foreign management practice to domestic companies following foreign acquisition. It can also proceed by emulation within domestic companies, whether acquired or not, as Oliver and Wilkinson (1992) noted in the case of Japanese-type production methods. These considerations prompt the questions “what is being transferred” from foreign investing companies, and “what are we comparing?” (Morris & Wilkinson, 1996, p. 727.) The first concerns the characteristics of management practices that are transferred directly or through emulation. What differences in this respect might one find between companies from each of the “big four” countries investing in the U.K.? The second question concerns the difference that FDI, as opposed to acquisition by U.K. companies, makes to the management practices introduced into U.K. subsidiaries. This chapter
Integrating Acquisitions
29
addresses the first of these questions, and in particular the extent to which what is transferred is affected by the overall integration, control, communication and strategic philosophy exhibited by the new parent company. Until the 1980s many students of management assumed that there were general principles that might be applied to situations, irrespective of the culture of the companies being studied. The dominant view was that the appropriate approach to management and organisation should be determined in the light of prevailing contingencies, particularly those established by the market, technology and scale of operation. Culture was either thought to be of limited relevance (e.g. Hickson et al., 1974) or just one of several contingencies to be considered (cf. Child, 1981). In the last two decades, however, the pendulum has swung strongly in the opposite direction. Company and national culture are now seen as critically important in selecting management methods, strategies and structures (cf. Hampden-Turner & Trompenaars, 1993). A growing body of research on national management systems, and relevant national cultural differences, has led to the expectation that companies of different nationalities will introduce distinctive management practices. At the same time, markets and corporations have been globalising rapidly, and many more companies now face two distinct cultures, their own, and that of a foreign partner or parent. Work on acquisitions and their performance (Haspeslagh & Jemison, 1991; Norburn & Schoenberg, 1994) as well as on the effects of differing national management cultures on the performance of acquisitions (Very et al., 1996; Morosini & Singh, 1994) has also lead to interest in the wider implications of national and managerial culture for acquisitions and their performance. Very et al. (1996) studied differences in acculturative stress and Calori et al. (1994) differences in control mechanisms in acquisitions by U.S., French or U.K. companies of French and British firms. However, the literature which studies post acquisition integration generally avoids looking at national differences altogether or looks more at national differences between the acquired and acquiring company rather than between different nationalities of acquiring company. Thus while differences in national management are much studied, differences in national approaches to acquisition integration which this chapter addresses have been relatively neglected.
MAJOR ISSUES This chapter focuses on the effects of acquisitions in four major areas. Firstly, the level of integration of the subsidiary into the parent. Secondly, the methods and systems adopted by the parent to control its new acquisition. Thirdly, the method of communication used. And fourthly, the strategy and philosophy
30
ROBERT PITKETHLY, DAVID FAULKNER AND JOHN CHILD
the new parent adopts in relation to the subsidiary. The second, third and fourth of these factors depend crucially on the overall attitude the parent takes to integrating the new company into its overall corporate structure. Integration The importance of the extent to which an acquired company is integrated into its parent organisation rapidly becomes apparent once one talks to acquired companies. There have been a number of writers who have touched on the issue of integration. Norburn and Schoenberg (1994) identify the need for “relatively specialised integration skills different from those required within an intra-U.K. context” and identify three needs. These are integration by facilitating a transfer from owner-management to professional management, the proactive transfer of skills to overcome a lack of integration, and the need to overcome potentially conflicting national cultures. Morosini and Singh (1994), while concentrating on implementing a “national culture-compatible strategy” as a means to improving the performance of acquisitions, draw attention to the difficulties of integrating resources across both acquiring and acquired companies, something seen as detrimental to the performance of the acquisition. Datta (1991) also highlights the importance of integration and the finding that procedural integration problems are less detrimental to performance of the acquisition than cultural integration problems associated with some but not all differing management styles. Shrivastava (1985) identifies three types of integration : procedural, physical and managerial/ socio-cultural, the last of which is found by Datta to encompass the potentially important cultural differences in management style. Gall (1991) identifies integration as a key organisational issue faced by management of a new acquisition and emphasises the role of employee communication in building a positive post-acquisition climate. The overall degree of integration achieved following an acquisition, and the degree of control and communication involved in that integration process, are therefore issues of great interest. This is because an inappropriate level of integration may be detrimental to performance. Thus a tendency to over or under integrate as a result of cultural factors hindering integration or pressuring moves towards it may result in sub-optimal solutions. Haspeslagh and Jemison (1991) have proposed a set of “metaphors” to classify acquisitions into four types depending on whether their needs for organisational autonomy and for strategic interdependence are high or low. In their typology a “Holding” acquisition involves a low need for both organisational autonomy and strategic
Integrating Acquisitions
31
interdependence. “Absorption” acquisitions involve a low need for organisational autonomy but a high need for strategic interdependence. “Preservation” acquisitions have a high need for organisational autonomy but a low need for strategic interdependence. “Symbiosis” involves a high need for both organisational autonomy and strategic interdependence. However, while this categorises acquisitions, it is clear that: The usefulness of choosing an overall metaphor for an acquisition integration does not change the fact that acquisitions bring with them many positions and capabilities, the integration of which, seen in more detailed perspective, might be best served by a different approach (Haspeslagh & Jemison, 1991, p. 146).
It is thus possible in considering acquisitions to anticipate both cases where either the overall picture or one detail of it suggests one of the types of acquisition Haspeslagh and Jemison describe (Holding, Preservation, Symbiosis or Absorption) and yet also to see that in most cases there is a subtlety in the approaches adopted by some managers and a multitude of detail including the many resources, capabilities and other factors to be considered. Furthermore, it is difficult in practice sometimes to distinguish clearly between Haspeslagh and Jemison’s Holding and Preservation categories. Angwin (1998) however, gives increased focus to the two categories by distinguishing between a holding category where the acquirer tries to effect a turnaround but without any degree of integration as opposed to a preservation approach where the acquired company is also left unintegrated but in order to continue making good profits. Another way to distinguish them is to say that holding acquisitions are not concerned with preserving capabilities but with remaining as uninvolved as possible. Whichever way one looks at acquisitions, there appears to be a potential continuum in the degree of integration, or what might be called a spectrum of integration. This is illustrated in Fig. 1 below, which illustrates the acquisition by company A of company B with varying degrees of integration. This spectrum of integration ranges from acquisitions with little integration (1–2 on the scale, corresponding to Preservation and Holding) and where the parent and subsidiary remain distinguishable and their functions largely unintegrated, to those where the integration is almost total (6–7 on the scale, corresponding to Absorption) and all functions and departments of B absorbed into A. Symbiotic acquisitions could be arranged at intermediate points on this continuum corresponding to partial integration where some but not all functions and departments of B are integrated into A. While this view simplifies integration into one continuum it does of course comprise or summarise a multitude of different components.
32
ROBERT PITKETHLY, DAVID FAULKNER AND JOHN CHILD
Figure 1 also suggests the way in which the integration of the new subsidiary may vary. With a low level of integration (1–2), regular financial and other operating figures will be required for the parent to monitor the performance of the subsidiary. Some top level personnel changes may be initiated and some restrictions are likely to be imposed on capital spending. However the subsidiary will continue to operate and present itself to the market much as before acquisition. With higher levels of integration (3–5), the new parent is likely to take over and run centrally whole areas of activity. This is likely to cover strategy, and may involve finance, personnel policy and systems, procurement, product development, IT systems and possibly the whole area of branding and management of the company image. Depending on how strongly the new parent regards the reputation of the subsidiary’s name and trade marks it may or may not decide to continue using them. At these partial levels of integration the parent is likely to have recognised that it has something to learn from the acquired company. However, it will only centralise functions if it believes this is to the advantage of the corporation as a whole. The highest integration levels (6–7) correspond to total absorption into the parent’s organisation. Brand names may be retained if they are strong but, particularly in service organisations, may be discontinued after a transitional period. Control When one company acquires another it needs to exercise some control over the company that is now acting in its name and using its resources. Control is in many ways the antithesis of trust since the greater the level of trust between the companies the less the perceived need for tight control systems (Faulkner, 1998). However, control can take many forms. Control systems may be limited to control over budgets and capital expenditure. They may involve appointing staff to key positions in the subsidiary company, carrying out certain important functions like planning and personnel in the parent company, or imposing ‘need for approval’ requirements on identified decisions (Geringer & Hebert, 1989). The control system selected illustrates the degree to which the parent is willing to grant a level of autonomy to the newly acquired subsidiary, and may be crucial in terms of influencing the level of motivation of the acquired company personnel. While not discussing national culture, Goold and Campbell (1988) draw attention to three main approaches to managing subsidiaries depending on the degree of control and planning by the centre that is involved.
Fig. 1.
Spectrum of Integration.
Integrating Acquisitions 33
34
ROBERT PITKETHLY, DAVID FAULKNER AND JOHN CHILD
Control of new acquisitions is seen as a key issue by Calori, Lubatkin and Very (1994) who study the effect of culture on the process of integration. Calori et al. rely in their analysis on the control strategy dimensions of centralisation and formalisation identified by Child (1972, 1973). Their research found that French firms exercise higher formal control of the strategy and the operations and lower informal control through teamwork than American companies when they buy firms in the U.K. American firms however were found to exercise higher formal control through procedures than the British when they buy firms in France. In fact Dunning writing about an earlier wave of FDI notes that “we may perhaps say with some certainty that U.S. managerial and financial control is more likely than not to be fairly rigid for the first five years or so [after investing in a U.K. firm]” (Dunning, 1958, p. 112). Other research has shown that French decision-making is concentrated towards the top of hierarchies (Horowitz, 1978; Hickson & Pugh, 1995). Research by Maurice et al. (1980), comparing French with West German and U.K. manufacturing firms, found that in France there are usually more levels in the hierarchy. French hierarchies tend to be more top-heavy, with between 1.5 and twice as many supervisors and managers as in German firms. Calori and De Woot (1994) add to this hierarchical characterisation by noting that French companies have a far higher number of organizational levels, and a lower level of participation than German or other northern European countries. Culturally, the Germans emerge from surveys as tending to have high levels of uncertainty avoidance (Hofstede, 1991). This is associated with attaching a high value to stability, and has been taken by commentators to be a main reason why in German organisations there tends to be a strong orientation to the use and adherence to rules, and a heavy stress on control procedures (HampdenTurner & Trompenaars, 1993). Relationships exhibit a high level of formality and commitment to paper, and attention to detail is painstaking. As Hickson and Pugh (1995, p. 97) have put it, One of the most characteristic aspects of the German culture, which certainly strikes an outsider, is their way of managing uncertainty through an emphasis on planning and orderliness.
This penchant for order is, according to some writers, manifest in attention to organizational structure rather than to process (e.g. Stewart et al., 1994). Germans also tend to score highly on Hofstede’s measure of “power distance” and this cultural disposition manifests itself in the presence in most German organisations of ordered hierarchies, at least of status hierarchies.
Integrating Acquisitions
35
The issue of control is of course closely linked to that of autonomy and research by Datta and Grant (1990) suggests that autonomy should be proportional to the unrelatedness of the acquisition’s business. Abo (1994) describes the Japanese management system as very flexible with few rigid job demarcations. Workers, supervisors and managers collectively take part in the discussion of managerial and operational functions. Assembly-line workers are responsible for on-line inspection and quality. There is a use of “implicit control” based on shared corporate norms and understanding. The importance of control of acquisitions and the fact that a wide variety of approaches to it exist are thus widely acknowledged in the literature. Communication Communication is partly a matter of systems and style, but also a question of the communicator’s skill at getting his message across to his colleagues. This is an area where nationality is very important, since different cultures have different attitudes towards communication, and language barriers inevitably make communication difficult. What is undeniable is that differing nationalities have differing communication styles ranging from the indirect, often unspoken style and “implicit understanding” of Japanese culture, through the distant understatement and explicit comprehension of British culture, to the close but highly direct communication style of U.S. culture. Relative to this continuum French and German communication styles might perhaps be tangential, differing on issues of logical necessity and formality respectively. For example French organisations have more non-managerial white collar specialists in either commercial-cum-administrative or technical functions, which reflects the French tendency to separate technical, planning, administrative and supervisory tasks from execution and operational ones (Sorge, 1993). In order to make this segmented and complex operation predictable and reliable, French organisations tend to use written rules, instructions, and communications extensively. The formality of German organisations on the other hand has already been noted (Stewart et al., 1994) and this formality can be seen to extend to methods of communication too. The methods and style of communication adopted by a parent company in dealing with a foreign subsidiary are thus by the very differences which exist between the parent and subsidiary’s natural communication style bound to play a critical role in the integration of any foreign subsidiary into its parent company’s organisation.
36
ROBERT PITKETHLY, DAVID FAULKNER AND JOHN CHILD
Strategic Philosophy The parent company’s strategy or philosophy is clearly crucial to the subsidiary’s future well-being. This comprises the parent company’s management style and culture as well as whether it adopts a primarily strategic or financial orientation. It also comprises the extent to which autonomy is granted to the subsidiary, and the expectation or otherwise of immediate and short-term profits. It establishes a framework within which the new subsidiary will have to learn to work. The short-termism of U.K. management has been frequently cited, both in academic studies (Lane, 1995) and by those working in industry (Marsh, 1995). Acquisitions by French companies on the other hand tend to be strategic rather than based on short-term financial considerations. This strategic emphasis is often phrased in a quasi-military language in which the contestants (for France has a large number of contested take-over bids) adopt, by analogy, the stance of the great generals and marshals (Barsoux & Lawrence, 1990). In contrast U.S. management culture places a stronger emphasis on achieving short-term financial results (Jacobs, 1991) and many U.S. companies tend to be managed for the short-term maximisation of profits and the satisfaction of shareholders (Calori & De Woot, 1994; Lawrence, 1996). The strategic approach or philosophy adopted by a new parent company in relation to a new foreign subsidiary is thus another critical component involved in the integration of the subsidiary into the parent’s organisation.
SAMPLE AND METHOD The data used in this chapter and chapter 3 were obtained primarily through semi-structured interviews with managing directors of acquired companies. The companies investigated are profiled by nationality of acquirer in Tables 1, 2, 3 and 4 below. The essential details of the company selection process for the interviews were as follows. A list of potentially relevant examples of FDI was prepared using data from the Central Statistical Office, Reuters, Predicasts, DTI, the Journal “Acquisitions Monthly” and other sources. This covered acquisitions of U.K. companies by U.S., French, German and Japanese companies in the period 1985–1994. Interviews were held with companies selected from among those companies acquired by a U.S., French, German or Japanese company. Forty of 79 companies asked (50.6%) agreed to being interviewed with the companies being divided equally among the four nationalities. Interviewed companies were spread over the whole of the United Kingdom with the
Integrating Acquisitions
37
Table 1. Acquirer
USA 1 USA 2 USA 3 USA 4 USA 5 USA 6 USA 7 USA 8 USA 9 USA 10
Major MNC (H)* Large Nat’l retailer (L) Major transport co (H) Major MNC (H) Major Int’l consult’y (H) Large Nat’l hi-tech co (L) Major MNC (H) Major MNC (H) Large Nat’l co (H) Large Int’l co (H)
USA Acquisitions.
Acquiree
Acquiree Industry
Acquiree Condition
Small family co Perfumier 4 local transp’t cos Specialist manuf’r Nat’l IT consult’y Hi-tech defence co National FM co Small Mfg co Specialist Start-Up Small national co
Medical implants Cosmetics Courier Automobiles IT consult’y Electronics Facilities Mgt Synthetic fibres Insurance Engineering
just profitable losses losses losses profitable profitable profitable just profitable profitable just profitable
* International experience: H = High, M = Medium, L = Low; Information gleaned from the interviews.
exception of Wales and Northern Ireland. Seventy percent of companies interviewed were in the manufacturing sector and 30% in service industries. U.K. companies acquired by U.K. companies were excluded from the interview program since the cross-cultural interactions occurring in the transfer process between companies, which it was intended to study, would obviously not be present in such cases. The interview schedule covered a wide range of management practices. Background information on the story behind the acquisition was also gathered.
Table 2.
Japanese Acquisitions.
Acquirer
Acquiree
Industry
Condition
J1 J2 J3 J4 J5 J6 J7 J8 J9 J10
Small family co Ex-sub’y of MNC Old branded co Nat’l branded co City merchant bank Medium Eng co National branded co Ex Danish sub’y co Old Manuf’g co Old ex-sub of U.S.
Engineering Engineering Household goods Computers Banking Engineering Mens clothing Polymers Engineering Pharmaceuticals
losses losses losses losses losses just profitable just profitable losses just profitable just profitable
Major national co (L) Major national co (L) Major national co (L) Major MNC (M) Large domestic bank (L) MNC (H) MNC (H) Major Int’l co (M) Major MNC (H) Large Nat’l family co (L)
38
ROBERT PITKETHLY, DAVID FAULKNER AND JOHN CHILD
Table 3. French Acquisitions. Acquirer
Acquiree
Acquiree Industry
Acquiree Condition
F1 F2 F3 F4 F5 F6 F7 F8 F9 F10
2 Small Private cos City firm City firm Domestic co Small domestic co IT consultancy co Brand name co Regional co Ex-sub of U.K. plc 2 regional cos
Defence Banking Banking Engineering Marketing IT consultancy Adhesives Water Pharma Water
losses losses just profitable profitable just profitable profitable losses profitable profitable profitable
Major MNC (M) Large MN bank (H) Small Parisian bank (L) Major domestic co (L) Major MNC (H) MNC (H) MNC (H) Major MNC (H) Family co (L) Major MNC (H)
Respondents were asked open ended questions about which changes they felt had been the greatest and had the most impact on the company. The interviews involved questions under the following headings: (1) General Background to the acquisition: Reasons and consequences. (2) Background to major changes and influences: (a) Discussion of 2–3 Major Areas of Change; (b) Patterns of Influence; (c) Integration.
Table 4. German Acquisitions. Acquirer
Acquiree
Acquiree Industry
Acquiree Condition
G1 G2 G3 G4 G5 G6 G7 G8 G9 G10
Family co Small regional co City bank Specialist Manf’r Truck agency Small family co Small local co Med. National Small family co Small national co
Pharmaceutical Engineering Banking Hi-tech medical Automobiles Fire security Furniture Trailers Chemicals Construction
profitable losses just profitable just profitable losses just profitable losses profitable profitable profitable
Major MNC (H) Large national co (L) Major landesbank (H) Major MNC (H) Major MNC (M) MNC (H) Medium national co (L) Medium national co (L) Major MNC (H) Domestic co (L)
Integrating Acquisitions
39
Table 5. Interviewees. No. of Interviewees CEO/MD /General Manager Chairman Other Director Non Board Level – Manager/Director TOTAL Subsidiary Board Level Directors Appointed from Parent Appointed from Subsidiary Joined company before acquisition Average no. of years since acquisition Average no. of subsidiary employees
French German Japanese
U.S.
TOTAL
Total
5 2 1 2
8 1 1 0
8 0 2 0
8 1 1 0
29 4 5 2
73% 10% 13% 5%
10
10
10
10
40
100%
8 0 10 8 7.4 983
10 3 7 7 5.6 212
10 0 10 8 7.2 922
10 2 8 7 5.5 1213
38 5 35 30 6.4 833
95% 13% 88% 75%
(3) Impact of acquisition on Performance: (a) What have been the main benefits of the acquisition? (b) What have been the main disadvantages of the acquisition? (c) How has the acquisition contributed to profitability and growth? In each case the interviews lasted for at least an hour and in all save one case were conducted at the offices of the acquired company. The interviewees were primarily (73%) CEOs, Managing Directors or General Managers of the subsidiary company, most of whom had been with the subsidiary company since before the acquisition and almost all of whom (95%) were board level directors of the subsidiary company, (cf. Table 5 above). Some (13%) had been appointed by the new parent company from among parent company staff. In order to avoid any bias that might still result from such a mix, the interviews focused on significant events and minimised more judgemental and qualitative questions. However, all interviewees were also asked to rate the degree of integration achieved by the acquisition on the scale of 1 to 7 described above. All interviews were tape recorded.
FINDINGS Overall Findings On the 1–7 scale of integration described above, the 40 companies interviewed had an average integration level of 3.61. However within this overall figure
40
ROBERT PITKETHLY, DAVID FAULKNER AND JOHN CHILD
there were significant national differences in the level of integration. American companies were the most committed to total integration of subsidiaries with an integration level of 4.77, while Japanese and German companies tended to be the least inclined to integrate with average integration scores of only 3.05 and 2.95 respectively. French companies interviewed had an average integration level of 3.65. An ANOVA test showed that the overall differences between the average integration scores for each country were significant at the 7% level ( p = 0.072). Overall, the interviews showed that the most common form of control imposed was through financial means, i.e. approval and monitoring of subsidiary budgets, and control of capital expenditure. This form of control was generally allied to the provision of a strategic framework within which subsidiary decision-making was to be confined. Differences between national approaches in this area mirrored those found in levels of integration. The parents were generally able to reassure their subsidiaries that their investment was for the long-term. Communication was an area in which there was a clear differentiation between individual national styles. From the interviews it was apparent that American companies were professional communicators relishing the use of first names, regular meetings at all levels, notice boards with mission and vision statements on them and company newspapers. Communication between Japanese companies and their U.K. subsidiaries did not seem as easy or open in comparison. German companies on the other hand appeared to veer between the stiffly formal and the selfconsciously informal, while French companies seemed to suffer little self-doubt, communicating well amongst themselves but informing subsidiary staff only on a ‘need to know’ basis and adopting what one interviewee referred to as a generally ‘colonial attitude’. The four countries’ attitudes to the four key areas concerned with are summarised in Table 6 below. These four areas of overall integration, control, communication and overall strategic philosophy will now be reviewed in more detail in respect of each of the four countries.
COMPARISON BY ACQUIRER NATIONALITY American Companies (a) Integration American post-acquisition management tended primarily towards total absorption even where this required some time for readjustment. In the case of a diversified U.S. manufacturer’s subsidiary USA32 the new U.S. manager said:
Integrating Acquisitions
Table 6.
USA
France
41
National Contrasts in Parent-Subsidiary Relationships. Strategic Philosophy
Integration
Control
Communication
Fully Integrated
Targets/Budgets Strict Financial Control Strategic & Financial Control Budgets/Systems “Advisers” Varied: Budgets/Systems Informal Controls
Open but Formal
Short Term Financial
Need to Know/ Top Down Need to Know/ Implicit Upward Formality Downward Informality
Long Term “Imperial” Long Term Strategic Long Term Indistinct
Japan
Partially Integrated Not Integrated
Germany
Not Integrated
The people here seem to hate all things American, even the size of the note pads. Some of the personnel still have to adjust to the fact that they are now part of a very successful American-based multinational, not a little family company.
In this case the company was very much absorbed and other examples of total integration of U.K. companies by U.S. parents abounded. USA4 progressively absorbed a U.K. subsidiary strengthening both control systems and financial reporting. The MD said the new subsidiary: had been totally integrated. There was about a year of separateness, but for full integration 5 years have been needed. We have a well organised account management structure which doesn’t tolerate weak performers.
In some cases the nature of the business made integration inevitable. For example, USA5 a U.S. freight company totally integrated a U.K. company into its world-wide organisation reorganising its systems to match USA5’s global systems. Of the ten USA acquisitions interviewed five had an integration level of 5 or more, three were at the 3–4 level and only 2 at the non-integrated 1–2 level. With an average score of 4.77 the preference of U.S. parent companies was for a high level of integration. (b) Control An emphasis on financial control and shorter financial time horizons was typical of the U.S. subsidiaries interviewed. In many cases the change in financial controls was sudden and related to a lack of the controls expected in a U.S. financial environment with quarterly reporting requirements which
42
ROBERT PITKETHLY, DAVID FAULKNER AND JOHN CHILD
privately owned U.K. companies do not face. The MD of a small U.K. company USA6 acquired by a larger U.S. company said: The Americans had to justify their investment. They put in financial reporting systems much quicker and ones which were compatible with their own systems and what they have done has improved the business. It needed a financial controller rather than a part time accountant.
The need for regular financial returns by U.S. companies facing quarterly reporting constraints was common. According to the U.K. MD of USA7 this is also linked with a pragmatic style of management depending on trust but with a heavy emphasis on performance: The relationship with the U.S. CEO was one of trust. He asked ‘is this British Management delivering that which we require?’ . . . if it was he didn’t interfere . . . . He said to me ‘. . . you know what style of management we have? It’s the management of the Mafia – send us the money and we leave you alone!’ . . . it worked very well.
A requirement for consistency with the corporate profile often accompanied the financial controls. This could be tempered by a realisation that there had to be a balance between integration involving instilling big company values and trying to preserve the small company’s entrepreneurial spirit and flexibility. However, all ten U.S. acquired companies interviewed reported substantial tightening of control systems especially financial ones. (c) Communication The U.S. companies interviewed also tended to pursue informal communications in a relatively formal way. That is not to say that U.S. parent companies are always formal but as one MD of a U.S. multinational manufacturing company’s subsidiary USA8 said: On the one hand there’s a very informal style. I was at the site when the take-over took place and the CEO of USA8 came over and said “Call me Hank”, well you wouldn’t call the chairman of UK1 [the former parent Co.] “Hank”. He wouldn’t invite you to . . . . But on the other hand there’s a very high degree of toughness and insistence on conformity . . . conform – or you’re dead. With UK1 . . . a signal from on high was a signal for wide ranging debate – was this meaningful, useful and were we going to obey it! USA8 cannot cope with a stand-alone subsidiary. They have to integrate everything.
Companies acquired by U.S. parents, reported moves towards a shorter planning time horizons and employment philosophy but the toughness mentioned above also extended to employee relations as the MD of USA8 also said: I find American business culture pretty difficult. UK1 never talked about caring and valuing people but its actions showed that it did. USA8 talks a huge amount about caring and valuing people but when the going gets tough . . . it doesn’t care at all.
Integrating Acquisitions
43
(d) Strategic Philosophy The American philosophy of post-acquisition management appears to be very much a hands on approach. The MD of USA7 which had been German owned but which was bought by a U.S. company said: They expect instant returns . . . . There’s no well, we wait for three years, make sure there’s synergies and then look for a return – they want it now. ”The shock is not just financial, . . . the speed at which change is to be introduced is quite extraordinary. So no doubt about it, a much more interventionist type of policy, much more dynamic, much more forcible and much more demanding.
Despite a shortening of financial time horizons imposed by USA acquirers, the larger size of many U.S. parent companies relative to their U.K. subsidiaries meant that very significant investments were still made in the subsidiaries. The MD of a small U.K. technology based company USA9 said that the main benefit of the acquisition of his company was that: Our company was going down the tubes fast. USA9 has been very successful in turning around the business and putting in huge sums of money. One must give them credit for that which UK2 (our former owners) probably could not have done.
Another MD of the U.K. service company USA9 said: Now we are preparing five year plans . . . before we never went further forward than 12 months. But at the same time, there is pressure for quarterly results . . . . You get all kinds of absurd requests to make more profit or collect more debt each quarter . . . so we’ve noticed both a lengthening and a shortening of the time horizons.
However, while experiencing the financial demands of U.S. ownership the subsidiaries of some U.S. companies have tended to have more autonomy over capital expenditure and changes in strategy. One MD of a U.S. company’s subsidiary USA10 said that, with a fairly dynamic approach, it may be possible to drive strategy which the parent company might not think of. In the case of the U.K. company concerned this was mainly apparent through marketing and liaison initiatives which it, not its U.S. parent, had initiated. The U.S. acquirers tended to install a formal multi-year planning systems, and to adopt a global approach to business but demand good short-term financial results. The implication is that if the company does not perform in a relatively short timescale it or at the very least the MD will be divested. Japanese Companies (a) Integration Japanese companies’ attitude to integration contrasted strongly with the American one, with any change achieved incrementally by slow adaptation to the parent company’s norms.
44
ROBERT PITKETHLY, DAVID FAULKNER AND JOHN CHILD
One Japanese Bank, had acquired a U.K. financial services company in order to establish a presence in the City of London but having done so made very few changes, assuming that the U.K. management understood the business better. In another case, a U.K. pharmaceutical company was acquired inadvertently by a Japanese Pharmaceutical firm when it acquired a U.S. owned group in order to preserve a Japanese licensing arrangement. Nevertheless, it provided the U.K. firm with substantial financial support and technical assistance despite having little other involvement with the company. The new Japanese owners of a U.K. consumer products company J1 gave it steady financial support, and did not interfere in operational matters, leaving the company essentially unintegrated. A hands-off attitude which surprised the director interviewed. In the case of an U.K. electronics company (J2) acquired by a major Japanese company, which again remained unintegrated, the MD said: Historically most of their sales organisations are totally controlled from Japan . . . but not with us! We are quite an experiment . . . the numbers are not crucial so long as we are going in the right direction.
In all acquisitions there is a balance between allowing the subsidiary freedom from interference by the parent and on the other hand seeking the benefits that closer integration could bring. There is thus a balance between independence and intervention along the spectrum of integration. In the case of Japanese acquisitions the balance is often one where supportive independence dominates. Of the ten Japanese acquisitions interviewed none had been integrated at the 5 or higher level, six were at the 3–4 level and four were left at the 1–2 nonintegrated level, giving a low average of 3.05 overall. (b) Control Japanese companies not only use budgets and financial control systems to monitor their acquisitions but also other methods of keeping informed. Japanese companies uniquely introduced “advisers” and they generally permitted rather more decision-making autonomy to their new subsidiaries. The “advisers” were often managers sent abroad for two to four years to gain international experience prior to promotion back in Japan. In one subsidiary it was said that: There are no formal links, there have been some placements of Japanese personnel but they’ve been in new staff roles . . . . The few that have done it have been perceived by J8 to be up and coming managers and have come to learn about management internationally rather than take part in day to day management.
Integrating Acquisitions
45
However, in many cases these advisers had substantial expertise in some specialist area. J7’s MD said that once he realised one Japanese manager was a highly skilled engineer he was able to make very good use of his expertise in production management. The advantages of specialist technical support were also available as the MD of J6 found: We ran into a problem on a product. It required technology on adhesives which we hadn’t got experience of and they quite easily said we’ll send our adhesives man in . . . . But then they had all these experts very, very, focused experts.
In general where decisions were required from Japanese parent companies they could prove rather slow in arriving according to several interviewees. The MD of a U.K. retailing company J5 said that a problem with their Japanese parent was: Getting a clear answer, a clear ‘yes go ahead’. We keep battling away and sometimes they’ll never say no. If they said ‘no, go away’ we’d know where we stand but their culture doesn’t permit them to say no.
Another MD of a Japanese subsidiary J6 said: It was very frustrating, it would sometimes take weeks to get an answer. In reality you actually knew the answer to it. It was very annoying for the guys on the shop floor.
On the other hand where trust over a particular issue or range of issues had been established considerable independence could be expected but the trust and respect had to be earned. The MD of J7 said: I learned very quickly . . . that because of the communication problems they are only interested in figures not written words. It makes my life very much easier.” “My first presentation in Tokyo was to the effect that we needed 3–4 years to get things going. Where we stand today is where we predicted we would be . . . [4 years ago] . . . achieving what you promise is very, very important to a Japanese . . . so it’s a very happy relationship.
The ways in which Japanese acquirers stand out compared to the other national groups support some but not all of the normal characterisations of Japanese management practice. Their long-term and strategic orientation and collective orientation were clearly apparent. As the MD of a pharmaceutical company’s subsidiary J9 said: There is a feeling that we should know what we need to do and that we don’t need to go to [our parent] for counsel. But they know what is going on and have been very supportive.
Overall Japanese companies attitude to control was to believe in budgets and forecasts and expect them to be met in detail, to give considerable operational latitude but to take the big decisions in Japan, often agonisingly slowly.
46
ROBERT PITKETHLY, DAVID FAULKNER AND JOHN CHILD
(c) Communication The area of communication can be a difficult one with a Japanese acquirer for both linguistic and cultural reasons. For example in one interview with a previously family owned engineering company J1 the MD said: Because all Japanese companies’ managers speak good English, language is not a problem on a day-to-day basis. It does, however, pose a barrier in terms of control and information. Virtually all written communication to and from Japan is in Japanese and I cannot understand it. Nor is it translated for me. I therefore don’t know for certain exactly what they are saying.
Language problems can thus lead to communication barriers. However, in another Japanese acquisition (J10) the interviewee said: I think language is undoubtedly an issue at times but it’s never a significant problem. I think it can be at the more junior level . . . . When it’s just straightforward technical reports its no problem but when they are trying to get over areas of subtlety it can be difficult.
Thus communication barriers due to language could be overcome more easily where technical terminology was involved, perhaps due to the preponderance of imported foreign words or “gairaigo” in technical Japanese. (d) Strategic Philosophy The Japanese companies’ strategic philosophy was consistently long-term, even if the strategic details could often appear to be rather fuzzy and ad hoc. The MD of a U.K. pharmaceutical company J3 acquired along with its U.S. parent by a Japanese company said: I think we have benefited from the take-over from being able to address issues which prior to the take-over would have been difficult . . . through lack of funding or lack of strategic direction. Sometimes the patience for the return amazes even me. They seem quite laid back about it . . . the return on these things will stretch way out into the future.
Another U.K. company, J4, which manufactured a specialist consumer product and was bought by a major Japanese firm said that: The main benefit has to be the investment that was made. Our subsequent success has stemmed from the fact that it gave us the breathing space to move forward . . . .
The same company J4 is also a good example of the “slingshot effect” in that the subsidiary was eventually sold due to the parent’s financial difficulties in Japan and left having received substantial technical and financial support. This, when put together with their own in-house development arguably put them ahead of their former Japanese parents. With the exception of a longer-term employment philosophy, we did not find significant differences between Japanese and other acquirers in personnel
Integrating Acquisitions
47
policy changes. Virtually every company had introduced some or all of the operational practices associated with Japanese companies. It appears that Japanese acquiring companies adjust their HRM practices to suit the local U.K. context, while in the operations area most companies have gone a long way towards adopting a Japanese approach, as Oliver and Wilkinson (1992) concluded. All ten Japanese acquisitions interviewed described new owners whose major merit was their long-term philosophy, and willingness to back their purchase with financial resources. The logical incremental attitude commonly ascribed to Japanese companies meant that while detailed plans were often not in place or apparent, general objectives behind the acquisitions such as developing closer customer relations, or becoming more international were discernible. German Companies (a) Integration In a similar way to Japanese companies, German acquirers tended to avoid closely integrating their new acquisitions if at all possible. Of the ten companies interviewed none had an integration score above 6; five were in the 3–4 range and the other five were at the 1–2 non-integrated level, the average German score being 2.95. A major German safety equipment manufacturer acquired a small entrepreneurial firm (G6) in the same line of business. Unusually the entrepreneurial MD has remained with the company with the German parent saying: We don’t want you to change. Stay the way you are, because you have the ability to move in a market place more than we can.
Another U.K. company, in this case a very high tech instrument company (G4) acquired by a German competitor with a larger global presence, was also left alone by its new parent. The MD said: They have never sought to change any thing. They really have left us alone, I don’t think we have learnt much from them, but they may have learnt something about the flexibility of a small company, and what they need to do to become more flexible.
It is worth noting that where little change occurred it was often due to the subsidiary’s management gaining the trust and acceptance of the new parent’s management and convincing them that it knew its business better than they did.
48
ROBERT PITKETHLY, DAVID FAULKNER AND JOHN CHILD
(b) Control The German attitude to control revealed by the interviews is varied, ranging from Board representation and use of management accounts to putting in a MD who then controls very directly. The subsidiary managers interviewed generally felt that their German parent companies exercised relatively little influence over their acquisitions. Compared with the other firms, the German acquisitions were less likely to have parent company managers as CEOs, or as marketing and R&D directors. One successful subsidiary (G4) of a German acquirer comments: We report with management accounts on a monthly basis which we send direct to Germany . . . . Since we have never been wildly off budget, there is no major response to these. Anyway financially we outperform our parent as a whole.
Even less successful companies enjoyed a fair degree of freedom. One (G6) manager said: We are visited from Hamburg once a quarter. They get involved in the thinking as well as the numbers. But in general their aim is to keep a tight rein on the finances and not worry too much about how we achieve the results.
But there were also more muddled companies. One (G8) director said: The German company has a combined board of five people all of whom came over in the first few months to see their new acquisition . . . to be totally honest all of their views totally differed. So we produced a five-year business plan and rang them up and they said ‘that’s good, get on with it’ . . . . We did look to them for guidance in the early days, . . . but their ability to make decisions was somewhat strange to say the least.
Another German company (G9) started out very autocratically but showed that personal relationships as well as success could influence the degree of freedom a subsidiary enjoyed. The MD of the subsidiary commented that: . . . We would use G where we needed them but push them away where we didn’t . . . . we were able to wave our results as justification for the UDI that we declared. It was a quirk of the changing times within G and our success that lead them to that method of control. It was also personal. I very quickly developed a relationship of 100% trust with the chairman in both directions and he left it to me.
(c) Communication The issue of personal relationships also concerns communication. On the issue of formality two photographs belonging to the MD of a company (G10) which had been part of a U.S. led MBO subsequently sold to a German company illustrate the point. One showed the U.S. led MBO team on a beach at 10am celebrating the fact that they had just bought the company. The other showed a row of sober suited people in a wood panelled hotel reception room the day
Integrating Acquisitions
49
the contract to sell the company to a German company was signed. This the U.K. MD felt graphically illustrated the difference in formality between the two companies. However, while there may sometimes be a sober suited formality associated with German management there is also a sense, as the interviews revealed, in which German management style can be very informal. Various explanations were offered for this apparent paradox. One MD in a German owned subsidiary (G2) held that it was a generational issue: The Germans are quasi-formal and the first management team was very traditional and formal. The new management however is much younger and much less formal. We have ‘casual days’, ‘funny tie days’ and ‘sandals days’ and this sort of thing!
Other managers saw increased informality as a reaction to the recession of the early 1990s and a learning process whereby a more informal approach would preserve some of the small company characteristics that large German company might learn from in a recession. In the latter case this approach was contrasted with an earlier approach which involved management quite literally “by the book” using a rigid set of procedures. A German manager managing a U.K. subsidiary (G7) said: We Germans are still pretty formal, . . . I can use first names with the British but use titles with my German colleagues.
Despite such cultural differences German managers were seen to be closer to U.K. culture than many other nationalities as a financial services company (G3) manager observed: The culture is not that different. A bit more formal, but they are in my opinion far closer to us than any other European lot. You wouldn’t have any trouble having dinner with them.
(d) Strategic Philosophy German parent companies, as Japanese companies, had a long term view of investment decisions. However, even this could have an apparent downside by being almost too good. The MD of a manufacturing subsidiary G10 said: I think our financial director would almost have welcomed being pressed more to be selfsufficient. It was almost a failing of the Germans to be as supportive as they were. Quite frankly I think it would have done us good if they had said “look there’s no more money – you’ve got to sort yourself out.
Looking at several other acquisitions it seems that for long term financial support to be a success the subsidiary has not only got to have the support but also some idea about how to use that support and about the long term direction
50
ROBERT PITKETHLY, DAVID FAULKNER AND JOHN CHILD
of the company – whether that direction is self-generated or dictated by the parent. Just support or just direction is insufficient. This was reflected by the above MD’s sense that the U.K. company under its German owners suffered from a lack of strategic direction: it failed to have a long term strategic objective. The business objectives were not known by the board so how on earth could they be transmitted to the rest of the troops.
German management seems definitely influenced by the differing nature of the home financial markets (compared to U.S. financial markets). For example the manager of a German company’s subsidiary (G3) said what might be difficult for a U.S. company manager to say: They are long-termist. They do not have making a profit itself as a prime matter. They are not profit driven. They want to know about quarterly results but they are not dominated by them.
The remarkable aspect of these findings is that while they confirm the long term nature of the German management approach they directly contradict one of the assumptions about German management practice found in the literature. Commentators tend to stress a high level of formality, rule-orientation, orderliness and formal provisions for participation as salient characteristics of German management. These features did not distinguish the changes brought about by German investors in the U.K.; in fact quite the contrary. This may reflect, to some extent, the observation made by Stewart et al. (1994) that a German penchant for order is manifest in organisational structure rather than in process, since several of the changes which were distinctively less formal among German acquisitions concerned process rather than structure per se. Nevertheless, our findings raise questions regarding the validity of the typical German management stereotype, even if it represents an over-simplified view which is in any event subject to change. However it can be challenged both on grounds of the present contradictory findings and also in view of the fact that confirmation of other elements in the stereotype presented by the literature was absent from our comparative data. French Companies (a) Integration The French acquirers interviewed were in the middle of the integration spectrum at an average level of 3.65. Less determined to integrate acquired companies than American parent companies but more than Japanese and German companies. Four of the ten companies interviewed were at a 5 + level of integration, a further four at 3–4 and two at a non-integrated 1–2 level. The
Integrating Acquisitions
51
level of integration varies from total integration of a previously British IT consultancy company to form a French based transnational to water companies that have been largely left alone to operate substantially as before, at least in part due to regulatory restrictions. A manager in a French owned U.K. public utility company said: the general philosophy is “Local Management” although they are interested in this high up view they really leave the rest of it to local management. . . . Because we are a regulated business there’s a limit to what they can do.
(b) Control National characteristics reveal themselves much more in French companies’ way of exercising control. In general they operate very hierarchically and determine major decisions between French managers, regarding acquired company personnel with a ‘colonial’ attitude. Another example of the colonialist attitude to acquisitions is noted by Empson (1998). An executive from a French acquired aerospace company (F1) states: Things are decided informally amongst Frenchmen to the exclusion of British middle management. If contacts were plotted, pretty well all the informal links would be between the French.
The British part-time chairman of a French-acquired financial services company (F2) said: They have difficulty in understanding the British concept of a Board. So far as they are concerned a Board is a kind of registration. It is like a levee, not a discussion and decision taking body.
French companies seem to place great emphasis on the distinction between strategic decisions (to be made by French managers at HQ) and operational decisions (to be made by local managers on the spot) though financial controls remain important. As an executive from a French-acquired water company (F10) said: They appointed a French Financial Controller . . . that is one post they try and keep for themselves. On the operational side it’s all-local.
(c) Communication Communication tends to be very hierarchical and top-down with little involvement of British managers in the decision-making hierarchy. The head of a financial services company (F3) acquired by a French bank said: Communications are largely on a need to know basis. I don’t get involved in the Paris business at all although I try to network as much as I can. I don’t ferret around with things that don’t concern London.
52
ROBERT PITKETHLY, DAVID FAULKNER AND JOHN CHILD
An executive of an engineering company (F4) added: The French hierarchy is a ‘power’ hierarchy and it is the power that we find the most difficult to live with. I don’t think most of industry in the U.K., certainly a TQM-inspired participative modern ethos, has the power element that the French have built in to everything and they do have an arrogance and a dominance characteristic that one finds very difficult to live with.
(d) Strategic Philosophy French companies generally have a long-term strategic philosophy in relation to acquisitions. A manager in a major French electronics company’s subsidiary (F1) said: A strategic rather than a financial orientation exists; it has to, since the company loses money. However, it is thought to be important to stay in the market.
The MD of a high-tech manufacturing firm (F9) acquired by a French company said: The way in which the business is being approached is interesting ‘the centre and several self-sufficient satellites’ but in my opinion it was never going to be a workable philosophy because it weakens the whole thing rather than strengthens it . . . . It’s wonderful to manage a subsidiary where one is totally empowered but on the other hand how is one going to grow the business without the help of the group?
Another interviewee (F5) said: Decision-making is collective, but the French are very autocratic, but on big strategic issues only.
This suggests that the issue may depend on the size of the decision being made. A French financial services company, for example, was said to reserve larger decisions for itself but to allow its British subsidiary (F7) a lot of local autonomy. Indeed one personnel director related an incident shortly after being acquired when this became quite clear: The FD spoke about the state of the business, you can imagine the air of gloom . . . then another (U.K.) director said “What do you want us to do?” and there was a pregnant silence and the French director paused, looked straight down the table and said ‘Monsieur if we have to tell you what to do we have the wrong people’. So I thought this is good news because nobody is going to tell us what has to be done.
In other words the French management considered that it was not for them to influence the U.K. management so much as for the U.K. management to run their business. This apparently surprised the U.K. manager concerned who said he had always regarded French companies as bureaucratic and centralist. This also shows that to a certain extent it is up to a subsidiary to manage its owners as much as be managed by them.
Integrating Acquisitions
53
CONCLUSION This chapter has reported the findings of interviews with managers of a range of acquisitions of U.K. companies by foreign companies and has concentrated on four main issues. The level of integration of the subsidiary, the control methods and systems adopted by the parent, methods of communication and lastly, the strategy and philosophy of the new parent concerning the new subsidiary. All of these factors depend on and reflect the overall attitude a parent takes to integrating a company into its overall corporate structure. Past research (Calori et al., 1994) has looked at differences in control mechanisms between nationalities but more general studies of post-acquisition integration (e.g. Haspeslagh & Jemison, 1991) have tended to discuss differences between acquired and acquiring companies more than national differences between acquiring companies as in the present study. Haspeslagh and Jemison (1991) characterised acquisitions in terms of their needs for strategic interdependence and organisational autonomy. At first sight these two axes seem to be orthogonal. However with needs for autonomy and interdependence, being in some sense opposites it would seem that replacing them with a single measure, overall need for integration, would achieve the same power of distinguishing types of acquisitions more economically. This is reinforced by the fact that H&J found no instances of their holding category of acquisitions (those with a low need for both autonomy and strategic interdependence). The present research confirms though that both a need for organisational autonomy and a need for strategic interdependence are critical. As outlined above, there were several German acquired companies where the new parents wished to keep their subsidiary intact precisely in order to preserve and learn from their unique managerial character. There were also companies such as the U.S. acquired freight company where absorption was total leaving almost no trace of the subsidiary’s former independent existence. A number of factors affect the degree of integration to be pursued. Haspeslagh and Jemison (1991) single out both acquired company quality and acquisition size. The present investigation has also identified as other important factors the control methods and systems adopted by the parent, the methods of communication and the strategy and philosophy of the new parent concerning management of subsidiaries, as well as the focal variable of this study: nationality. Organisationally embedded capabilities that require organisational boundaries for their preservation are no less strategic capabilities than others that require the boundaryless conditions of absorption to maximise their benefits. It seems therefore that the overall choice to be made and thus the key dimension involved is one of the degree of integration to be adopted. The
54
ROBERT PITKETHLY, DAVID FAULKNER AND JOHN CHILD
key factor to be assessed with respect to any capability is whether that capability’s benefits following acquisition will increase or decrease with the degree of integration of the two companies. Paradoxically therefore the complexity of the issues involved in acquisitions and revealed by the present research has suggested a simplified categorisation of acquisitions using the degree of integration of subsidiaries. While Haspeslagh and Jemison (1991) characterised acquisitions in terms of their need for strategic interdependence and organisational autonomy, the new scale simply ranges from not integrated, through partially integrated to fully integrated where the subsidiary organisation is no longer distinguishable within the parent company. This does not ignore the fact that many components contribute to deciding a company’s position on the scale of integration following acquisition but neither does it concentrate on particular components to the exclusion of others. In presenting such a scale one must therefore emphasise the multi-dimensional nature of post-acquisition integration. The research showed that U.S. subsidiaries tended to be more integrated than other nationalities and German and Japanese subsidiaries less integrated than others with French subsidiaries somewhere in between. The interviews also showed that, while the degree of integration differed, the means of control used were common, albeit not applied to the same degree. Key examples were financial controls related to investment but common cost controls were also involved as well. The use of advisers and or informal controls in Japanese and German subsidiaries contrasted with the stricter financial controls of U.S. companies. The way in which parents communicated with their subsidiaries also showed distinct differences between the four nationalities studied with the main parameters being the degree of formality/informality, the degree of openness or more limited ‘need to know’ communication, and the directness of communication. As expected, Japanese companies operated more on a need to know and implicit communication style, U.S. companies were open but formal while French companies tended to a more autocratic top-down style. German companies who might be expected to be unduly formal were found to be upwardly formal in dealing with their parent but downwardly very informal in communicating with their subsidiary. These findings reflected the overall strategic philosophy adopted by the four nationalities of acquiring company. U.S. companies tended towards a short term financial approach to the three key issues while the other three countries were predominantly long-term in their approach. However they differed in that while Japanese companies had a strategic approach generally involving substantial financial support, French companies had a more “imperial”
Integrating Acquisitions
55
approach and German companies a range of indistinct though generally long term approaches. The lessons to be learned from the research findings are several. Firstly that while there are many similarities between acquisitions in general there are substantial differences in approach by acquiring companies of different nationalities. Secondly, communication is critical in building links between parents and subsidiaries but words must be backed by consistent actions if credibility is to be maintained. Linguistic and cultural problems with communication need more attention than they usually receive. Thirdly, a range of control styles is possible, ranging from the very strict financial controls favoured by many U.S. parents to the much more “laissez faire” attitude of Japanese parent companies and even some French companies. One might compare these findings with those of Goold and Campbell (1988) who drew attention to three main approaches to managing subsidiaries (Strategic Planning, Strategic Control and Financial Control) depending on the degree of control and planning by the centre that is involved. The present research shows that strategic and financial control and planning are important elements of managing an acquisition. However, it also shows that many other elements are involved and equally important to the success of the acquisition, not least the issues of the differing cultures of the acquiring and acquired companies and communication between them, as well as the overall level of integration. This again emphasises the multi-dimensional nature of postacquisition integration. The research also suggests several areas for further investigation. Investigating potential links between integration and performance is one area. For example does subsidiary profitability at the time of acquisition affect approaches to integration? This paper does not report results concerning postacquisition performance. However it is worth noting that despite the distinct national approaches to managing the control and integration of subsidiaries, no one country’s acquisitions were significantly more successful than any other. This all leads to the overall impression that it is not so much whether companies, control, communicate with or integrate new subsidiaries but the particular ways of doing so that they adopt in any given situation. Norburn and Schoenberg (1994) suggest that with an international acquisition there is a need for “relatively specialised integration skills different from those required within an intra-U.K. context” but, as this research has shown, such skills and the approaches adopted will differ from nationality to nationality. The multidimensional nature of acquisition management and the issues of integration, control and communication are therefore shown to be open to different national
56
ROBERT PITKETHLY, DAVID FAULKNER AND JOHN CHILD
approaches but none the less important for that. They remain of continuing interest as important factors in managing the acquisition process.
NOTES 1. These sources are unlikely to be comprehensive. In particular, smaller deals may go unnoticed by the British press and unrecorded by the DTI and CSO. The actual level of activity will therefore certainly be higher than that captured by the above sources. However, it is believed that the relative size of the number of activities by type, nationality and industry will reflect the overall picture, and these are the only data readily accessible to researchers. 2. All companies for reasons of confidentiality are referred to by numbered codes denoting the parent company nationality: USA for U.S., J for Japanese, G for German and F for French parent companies where n = 1 to 10. Former U.K. parent companies are denoted by U.K.
REFERENCES Abo, T. (Ed.) (1994). Hybrid factory: The Japanese production system in the United States. New York: Oxford University Press. Angwin, D. (1998). Post acquisition management of corporate take-overs in the U.K. Unpublished PhD thesis, University of Warwick. Calori, R., Lubatkin, M., & Very, P. (1994). Control mechanisms in cross-border acquisitions: An international comparison. Organisation Studies, 15(3), 361–379. Calori, R., &De Woot, P. (1994). A European management model. Hemel Hempstead: PrenticeHall. Child, J. (1972). Organization structure and strategies of control: a replication of the Aston study. Administrative Science Quarterly, 17, 163–177. Child, J. (1973). Strategies of control and organizational behaviour. Administrative Science Quarterly, 18, 1–17. Child, J. (1981). Culture, contingency and capitalism in the cross-national study of organisations. Research in Organisational Behaviour, 3, 303–356. Child, J., Faulkner, D., & Pitkethly, R. (2000). Foreign direct investment in the U.K. 1985–1994: The impact on domestic management practice. Journal of Management Studies, 37(1), 141–166. Datta, D. K (1991). Organizational fit and acquisition performance: Effects of post-acquisition integration. Strategic Management Journal, 12(4), 281–297. Datta, D. K., & Grant, J. H. M. (1990). Relationships between type of acquisition, the autonomy given to the acquired firm, and acquisition success: An empirical analysis. Journal of Management, 16(1), 29–44. Dunning, J. H. (1958). American Investment in British Manufacturing Industry. London: Allen & Unwin. Empson, L. F. (1998). Mergers between professional services firms: How the distinctive organisational characteristics influence the process of value creation. London Business School, University of London, Ph.D. Thesis. Faulkner, D. (1998). Are Trust and control opposing forces or complementary functions in alliances. EGOS Conference Paper, Maastricht, June.
Integrating Acquisitions
57
Gall, E. A. (1991). Strategies for merger success. Journal of Business Strategy, 12(2), 26–29. Geringer, J. M., & Hebert, L. (1989). Control and performance of international joint-ventures. Journal of International Business Studies, (Summer), 235–254. Goold, M., & Campbell, A. (1988). Managing the diversified corporation: the tensions facing the chief executive. Long Range Planning, 21(4), 12–24. Hickson, D. J., & Pugh, D. S. (1995). Management worldwide. London: Penguin. Hofstede, G. (1991). Cultures and organisations. Maidenhead: McGraw-Hill. Hofstede, G. (1996). Riding the waves of commerce: A test of Trompenaars’ ‘model’ of national cultural differences. International Journal of Intercultural Relations, 20, 189–198. Horowitz, J. (1978). Management control in France, Great Britain and Germany. Columbia Journal of World Business, 13, 16–22. Jacobs, M. T. (1991). Short-term America: The causes and cures of our business myopia. Boston, MA: Harvard Business School Press. KPMG (1997). Corporate finance survey. London. Lawrence, P. (1996). Management in the USA. London: Sage. Maurice, M., Sorge, A., & Warner, M. (1980). Societal differences on organizing manufacturing units: A comparison of France, West Germany and Great Britain. Organization Studies, 1, 59–86. Morris, J., & Wilkinson, B. (1996). The transfer of Japanese management to alien institutional environments. Journal of Management Studies, 32, 719–730. Morosini, P., & Singh, H. (1994). Post cross-border acquisitions: Implementing ‘national culture compatible strategies to improve performance’. European Management Journal, 12, 390–400. Norburn, D., & Schoenberg, R. (1995). European cross-border acquisition: How was it for you? Long Range Planning, 27(4), 25–34. Oliver, N., & Wilkinson, B. (1992). The Japanisation of British industry: New developments in the 1990s (2nd ed.). Oxford: Blackwell. Sorge, A. (1993). Management in France. In: D. J. Hickson (Ed.), Management in Western Europe. Berlin: De Gruyter. Stewart, R. et al. (1994). Managing in Britain and Germany. Basingstoke: Macmillan. Shrivastava, P. (1985). Post merger integration. Journal of Business Strategy, 7(1), 65–76. Very, P., Lubatkin, M., & Calori, R. (1996). A cross-national assessment of acculturative stress in recent European mergers. International Studies of Management and Organisations, 26(1), 59–86.
ORGANISATIONAL CHANGE PROCESSES IN INTERNATIONAL ACQUISITIONS David Faulkner, John Child and Robert Pitkethly INTRODUCTION Following the previous chapter’s discussion of how acquiring companies for the USA, Japan, Germany and France integrated their U.K. acquisitions with regard to the overall level of integration attempted and achieved, control, communication and strategic philosophy, this chapter discusses the organisational change mechanisms adopted in order to achieve this integration. The previous chapter gives details on the methodology adopted for the research and the criteria for the choice of sample. The codes for the companies (e.g. US04) follow those laid out in Tables 1–4 (pp. 37–38).
A LITERATURE REVIEW Corporate and national culture are now seen by many researchers as critically important in the determination of management methods, strategies and structures (cf. Hofstede, 1991; Hampden-Turner & Trompenaars, 1993). This view can be traced back to Roberts (1970) in her evaluation of the influence of culture on organisation. To Sorge (1982), culture was the dominant force to be considered in evaluating international performance. Boisot (1995) takes a similar view when analysing information theory across national boundaries. Loveridge (1998) implicitly accepts the importance of underlying culture in Advances in Mergers and Acquisitions, Volume 2, pages 59–80. Copyright © 2003 by Elsevier Science Ltd. All rights of reproduction in any form reserved. ISBN: 0-7623-1003-0
59
60
DAVID FAULKNER, JOHN CHILD AND ROBERT PITKETHLY
influencing management methods in flagging up the contention that complete convergence to a globalised best-practice may never come about. Child (1981) proposes a comparative examination of national differences in organisational practices between nations whose economic conditions are similar, in order to assess the importance of culture in determining management practices. Redding (1994) argues along similar lines. Boisot (1986) using two dimensions namely codification and diffusion of information, notes the different preferences of different cultures for information along these dimensions which he designates culture space. He claims that it explains the Anglo-Saxon preference for market transactions, and the Chinese preference for network capitalism (Boisot & Child, 1988, 1996). This form of theory can also be applied to the way in which different cultures record information and transform it from tacit to explicit (Nonaka & Takeuchi, 1995) leading to different organisational forms and practices for different cultures. Similarly Liu (1986) shows how the factorial analysis orientation of the West contrasts with the synthetic-holistic orientation of the East leading to different forms of preferred decisionmaking. A country’s institutions are shaped by its social and cultural history as well as by the logic of expediency, and they in turn shape a country’s management practices (Biggart & Guillen, 1999). The varying institutional factors of different nations are therefore also important in cross-cultural research into management practices. Mokhiber and Weissman (1999) point out that it can also work in the other direction as the management power of large multinationals can influence the evolution of a country’s institutions. The interaction of national culture and management practice is therefore by no means necessarily uni-directional, and more research is needed to develop a fullyfledged theory of the respective interactive roles of culture and organisation. Even in the absence of such a theory however, many researchers would concur that different nations with different cultures do manage companies differently, and would hypothesise that not all such differences are due to environmental contingencies. Calori, Lubatkin and Very (1994) in their research conclude that the French exercise higher formal control of strategy and operations, and lower informal control through teamwork in U.K. acquisitions, than do the Americans. They also conclude that the Americans exercise higher procedurally-based control than do the British in French acquisitions. being conscious of the influence of national administrative heritage should help reduce anticipated cultural problems in the integration process following international acquisitions . . . Firms are prone to carry their home practice with them as they move into foreign markets (Calori, Lubatkin & Very, 1994).
Organisational Change Processes in International Acquisitions
61
Perlmutter (1969) had long ago warned of the risk of ethnocentricity where the buyer does not make any adaptation to local practice, and carries on as though in its domestic market. Gates and Egelhoff (1986) concentrate on the importance of the country of origin of MNCs in their choice of control mechanisms over new acquisitions. Whitley (1992) addresses the issue from the viewpoint of the local company, and stresses the importance of local institutional and infrastructural circumstances in influencing effective consolidation of acquisitions across borders. A growing body of research on national management systems and relevant national cultural differences has led to the expectation that companies of different nationalities will introduce distinctive management practices. At the same time, markets and corporations have been globalising rapidly, and many more companies now face two distinct cultures, their own, and that of a foreign partner or parent. Work on acquisitions and their performance (Haspeslagh & Jemison, 1991; Norburn & Schoenberg, 1994) as well as on the effects of differing national management cultures on the performance of acquisitions (Very et al., 1996; Morosini & Singh, 1994) has also led to interest in the wider implications of national and managerial culture for acquisitions. There is considerable interest in the impact that the integration of the new subsidiary may have on its post-acquisition performance. Nahavandi and Malekzadeh (1988) find that the success of implementation is highly dependent upon the attitudes to the mode of acculturation held by both the acquirer and the acquired company. They see the four possible modes as integration, assimilation, separation and deculturation. Sales and Mirves (1984) see conflict, and hence probable failure of the acquisition, as most likely when firms with radically different corporate cultures come together. Kogut and Singh (1988) find that the greater the cultural distance between the two countries involved, and the higher the level of uncertainty avoidance in the potential market-entering firm, the less likely it is to adopt acquisition as a method of market entry, preferring greenfield development or joint venture. The impact of the acquirer’s national management practice on the postacquisition behaviour and performance of the acquired company is also an issue of concern. The implication of the globalisation of markets and the multiple management cultures faced by U.K. management, is that inward FDI, through the control and influence it gives to foreign management, will stimulate the adoption of management practices which contrast with past practices of the acquired U.K. firm. As Shrivastava (1986) points out there are several areas in which management practice influence might come about; e.g. in accounting and budgeting systems, in physical assets, product lines, production systems and technologies; and most importantly at a managerial and socio-cultural
62
DAVID FAULKNER, JOHN CHILD AND ROBERT PITKETHLY
level. Integration or at least influence are not always needed or in fact achieved at all these levels. It might be suggested that the managerial and socio-cultural level is the most important one, since Buono and Bowdich (1985) find in their research that where cultures collide and cause a merger to fail, the predominant observed factor is the discontent felt by the non-dominant partner based on subjective, i.e. cultural, rather than objective features. The importance of the national origin of the acquiring company in crossborder acquisitions is thus by now a well established and much studied phenomenon. However past study of the balance between direct application of foreign management practice and its adaptation to local conditions, leaves open the question of how far and in what manner FDI induces domestic changes in the management practices of acquired companies. Once the waves of foreign investment have swept over U.K. Industry, what is their effect on U.K. management practice and how are the changes involved achieved?
CHANGES PROCESSES INVOLVED IN INWARD FDI The integration process involved in ‘digesting’ the acquisitions can be depicted as in Fig. 1 below. Distinctly different approaches were taken to the task of bringing about change and improving performance in newly acquired companies. It is appreciated that other factors than the national identity of the acquirer are likely to influence post-acquisition behaviour, including the performance of the subsidiary at the time of acquisition, the prior international experience of the acquirer, the relative size of the partners, and perhaps the industry involved. It is impossible to formally control for all possible factors in a qualitative study. However, a varied range of companies in a wide range of industries was chosen
Fig. 1.
Post-Acquisition Change Processes.
Organisational Change Processes in International Acquisitions
63
for interview in order to seek to avoid bias in any particular direction regarding these other possible influencing factors.
OVERALL FINDINGS The following is the general cross-nationality picture of the acquisition process. As shown in Table 1 below, of the forty companies interviewed 30% were acceptably profitable at the time of purchase, a further 30% were at break-even, and 40% were making losses. By the time the interviews were carried out, the profitable companies had increased to 55%, the break-even ones reduced to 23%, and the remaining loss makers were only 22%. So with 78% companies in profit after acquisition compared with 60% prior to acquisition, a qualified approval could be given to the take-over process. An important issue is which of two parties initiates any post-acquisition change. In the companies interviewed the initiator was by no means dominantly the acquirer, as shown in Table 2 below. In 14 cases the acquirer led and in a further four acted as catalyst, but in 13 cases the subsidiary claimed to have
Table 1. Companies
Profitability of Interviewed Companies. USA
Japan
Germany
France
Total
Acceptably: before Profitable: after Just: before Profitable: after
4 8 3 0
0 4 4 3
4 4 3 3
4 6 2 3
12 22 12 9
Loss: before Making :after
3 2
6 3
3 3
4 1
16 9
Table 2. Change Initiator Acquirer Acquirer as catalyst Subsidiary Both Little change
Initiators of Change.
USA
Japan
Germany
France
Total
6 1 2 0 1
2 2 5 1 0
3 0 3 1 3
3 1 3 1 2
14 4 13 3 6
64
DAVID FAULKNER, JOHN CHILD AND ROBERT PITKETHLY
initiated the changes. In three companies changes were attributed to the joint efforts of the parent and subsidiary, and in the remaining six acquisitions, little change was said to have taken place. It is thus as likely that change will be initiated by the subsidiary as by the parent company following an acquisition, though as can be seen in Table 2 initiation by the parent is more frequent where U.S. parent companies are involved. As regards the method of post acquisition change, in 12 cases the subsidiary was absorbed into the parent or had a Managing Director appointed by the parent imposed upon it as shown in Table 3 below. However in seven companies the change was effected by the parent and subsidiary companies working together. In the rest of the companies interviewed, the parent company either adopted a “hands off” management style (five companies) more or less leaving the subsidiary to manage itself, or (in eight companies) allowed the subsidiary to effect changes for which the subsidiary’s management team had obtained prior approval. The degree of independence possessed by subsidiaries thus varied considerably, and subsidiaries were by no means confined in all cases to close management by their new parent companies. All acquirers provided finance, even if only in the sense of making the purchase. However, there was a wide variety in their attitude to dividends, some leaving the money in and some taking it out. As shown in Table 4, some positively improved the credit rating of the subsidiary by providing credit support for local loans. Only 18, just under half of those interviewed, provided strong technical and other support activities to bring the subsidiary up to the standard of the parent. The numbers exceed forty, since some parents contributed resources under more than one heading. In terms of monitoring progress in bringing about change 27% of the companies appointed a new MD to the subsidiary and sometimes other senior Table 3. Method of Change. Change Method
USA
Japan
Germany
France
Total
Absorption MD appointed Joint action Hands-off Action by sub Decisions by HQ Little change
5 1 1 2 0 0 1
0 1 4 2 3 0 0
0 3 0 1 2 1 3
0 2 2 0 3 1 2
5 7 7 5 8 2 6
Organisational Change Processes in International Acquisitions
Table 4.
65
Resources Introduced.
Resources introduced
USA
Japan
Germany
France
Total
Finance Technology Systems & IT support Mediocre products Asset investments Credit support
10 6 1 1 0 0
10 4 0 1 0 2
10 4 0 0 2 2
10 4 0 0 0 0
40 18 1 2 2 4
Table 5.
USA Japan Germany France Total % of Co.s
Parent Company Appointments.
MD & others
Senior Appointments
Advisors
Junior Appointments
None
Total
4 1 3 3 11 27%
0 4 2 0 6 15%
0 2 0 0 2 5%
0 4 2 1 7 17%
6 1 4 6 17 42%
4 11 7 4
NB some rows and columns add up to more then the total of 40 cases or 10 per country as some cases involved appointments in more than one category.
executives, a further 15% made senior appointments but left the existing MD in place, 17% made junior appointments, often ostensibly just for training purposes, but presumably also to monitor events in the subsidiary. An astonishingly high 42% made no appointments to the subsidiary at all, but merely tied it into a parent company appointed board resident in the parent company’s country (see Table 5).
DIFFERING NATIONAL APPROACHES TO CHANGE U.S. Companies The U.S. acquirers were the most internationally experienced of the four nationalities surveyed as shown in Table 6. Although international experience
66
DAVID FAULKNER, JOHN CHILD AND ROBERT PITKETHLY
is often important in achieving high international performance, this is only the case where the experience has led to increased know-how, and this is not always the case (Simonin, 1997). The U.S. acquirers were also the most likely to buy an already profitable company. The change process in U.S. acquisitions tends to be characterised by three main features. Change is initiated by the new parent company, it is often effected by absorption into the parent company in the manner described by Haspeslagh and Jemison (1991), and strong backing is usually given to new subsidiaries, not just in the form of finance, but also in support activities and technology. As the MD of a U.K. firm acquired by a major U.S. MNC (US5) said: After the take-over in 1990 they progressively absorbed us, and we are now totally integrated into US5 (U.K.). There were huge negatives at first. ‘The yanks are coming; they have no sense of irony.’ This took three years to settle down.
However, companies usually experienced an increase in performance following U.S. acquisition. Five out of ten U.S. parent companies absorbed their new subsidiary into the American parent’s organisation and systems in such a way that the subsidiary Table 6. Acquirer USA 1 USA 2 USA 3 USA 4 USA 5 USA 6 USA 7 USA 8 USA 9 USA 10
Acquiree Major MNC (H)* Large National retailer (L) Major transport company (H) Major MNC (H) Major International consultancy (H) Large National hi-tech company (L) Major MNC (H) Major MNC (H) Large National company (H) Large International company (H)
USA Acquisitions. Acquiree
Small family company Perfumier
Acquiree Industry
Medical implants Just profitable Cosmetics Losses
4 local transport Courier companies Specialist manufacturer Automobiles National IT consultancy IT consultancy Hi-tech defence company National FM company Small Mfg company Specialist Start-Up
Acquiree Condition
Losses Losses Profitable
Electronics
Profitable
Facilities Mgt. Synthetic fibres Insurance
Profitable Just profitable Profitable
Small national company Engineering
Just profitable
* International experience: H = High, M = Medium, L = Low; Information gleaned from the interviews.
Organisational Change Processes in International Acquisitions
67
could no longer be clearly distinguished from the parent. This involved adoption of the parent’s logo, brands and company culture and left little trace of the original company’s former independent identity. This was very much acquisition by absorption in the form identified by Haspeslagh and Jemison (1991). An example of such an absorption is that described by the general manager of US1. We are now a professionally organised part of the Health Care Division of USA1 and fully integrated into its organisation structure. I formally report to the EBC (European Business Centre) for manufacturing and to North America for business applications and R&D. There is no one person, but an integrated multi-point approach When stand alone businesses are acquired for example their R&D is not closed down but linked in to the world-wide network.
Only U.S. companies applied a total absorption approach to their acquisitions in any of the situations researched. They either absorb the subsidiary entirely, or leave it nominally independent but subject to substantial intervention by the parent. Thus U.S. parents are the main source of change in acquisitions by U.S. companies (cf. Table 2). One interesting aspect of the integrationist approach favoured by several U.S. companies was that it did not necessarily involve imposing U.S. appointed MDs on to the new subsidiaries. In four cases the U.S. parent did appoint a new MD, and took over the running of the new subsidiary. However, in six cases the parent company made no new appointments and worked through the existing management team. There was also no use of advisors, senior sub-MD appointments or junior trainee appointments i.e. they had slim and direct management lines from subsidiary to parent. The interventionist approach could, however, lead to problems as the general manager of the acquired subsidiary of US3 commented: The U.S. expats expected to put in an exact blueprint of US3 in the new country and were surprised when it didn’t work. Cultures are different. It took the Americans three or four years, a lot of money and some quite painful decisions to realise this.
One positive aspect of this interventionist approach was that such companies were also the most likely to supplement the finance provided with whatever other managerial and technological support services were needed to bring the new subsidiary up to the required standard. As the MD of the subsidiary of US4 put it: Quality was improved dramatically. US4 provided people for product design, training and management services . . . They brought money, armies of bureaucrats, and international professionalism.
However traumatic an acquisition by an American company might be to a small new subsidiary frightened of losing its identity and autonomy as it disappears
68
DAVID FAULKNER, JOHN CHILD AND ROBERT PITKETHLY
into the maw of the corporate monster, the result was in most cases a substantial improvement in its economic performance. Of the ten companies acquired, only two were not acceptably profitable at the time of interview, whereas six had been so at the time of acquisition. The distinctiveness of U.S. company attitudes to acquisition was noted by the MD of a U.K. company acquired by a U.S. (US10) company, who had also worked previously in a subsidiary of a German company: So if I was to say what is the lesser of the two evils – is it the benign neglect of the Germans and the avuncular style of management and the helter skelter dynamics of the Americans – I would go for the American style quite frankly . . .
Even if individual preferences for particular styles of management must also be allowed for in the comment, this does illustrate the difference in approach of companies from the two countries. However, the Americans are not always teachers rather than learners. As the MD of a U.K. company put it: The strategic planning group came to talk over the take-over. They’re not daft and they recognise that there’s been some learning on both sides which will help them in future to integrate smaller companies in a way which will not destroy the flexibility and culture and entrepreneurial attitude but gives them what they need in terms of embracing core US8 values, so they look at this as some form of role model for the future.
This suggests that, while U.S. companies can and do add value to their acquisitions, they may, at least sometimes, take care not to stifle the advantages which are part of being a smaller company. Japanese Companies The Japanese acquirers were the least internationally experienced of the four countries (see Table 7). Most were large domestic companies seeking to become more international. In their selections of acquisitions, they bought more loss-makers than did the other nationalities (see Table 1). Despite this, their track record at improving performance was good. However, their company attitudes towards acquisitions, and to the implementation of change following them, differed quite markedly from those of the other nationalities investigated. This is characterised by several distinctive features. The manner in which change was effected was much more by co-operative or indirect means than other countries. Consistent with this low-key approach to managing subsidiaries, most of the initiatives for change among the companies interviewed came from the subsidiaries themselves, rather than their Japanese parent companies. However, the support provided by the Japanese parent companies, particularly finance,
Organisational Change Processes in International Acquisitions
69
Table 7. Japanese Acquisitions. Acquirer J1 J2 J3 J4 J5 J6 J7 J8 J9 J10
Acquiree Major national company (L) Major national company (L) Major national company (L) Major MNC (M) Large domestic bank (L) MNC (H)
Acquiree
Acquiree Industry
Acquiree Condition
Small family company
Engineering
Losses
Ex-subsidiary of MNC
Engineering
Losses
Old branded company
Household goods Losses
National branded company City merchant bank
Computers Banking
Losses Losses
Engineering
Just profitable
Mens clothing Polymers
Just profitable Losses
Engineering Pharmaceuticals
Just profitable Just profitable
Medium engineering company MNC (H) National branded company Major international Ex Danish subsidiary company (M) company Major MNC (H) Old manufacturing company Large national Old ex-subsidiary of US family company (L)
played a large role in making implementation of such subsidiary initiatives possible. The economic condition of the companies acquired seems to have been of far less concern to Japanese companies than to companies of the other nationalities. Six out of the ten Japanese – owned companies interviewed were making losses at the time of purchase, and the other four were only close to break-even. This suggests that Japanese companies made their purchases more for strategic reasons than short-term financial ones. A further factor may stem from the difference between U.K./European accounting practice and Japanese accounting practice in that there is less pressure on Japanese parent companies to consolidate the accounts of their poorly performing foreign subsidiaries with their home accounts than for U.K./European companies, with the result that Japanese companies have another reason to take a more relaxed and longer term view of making their subsidiaries’ operations profitable (Tomohide, 1998). However despite these differences, Japanese companies do not seem to have been less successful in rendering previously unprofitable businesses profitable. Of the ten companies interviewed four had been helped into an acceptable level
70
DAVID FAULKNER, JOHN CHILD AND ROBERT PITKETHLY
of profit, three were just profitable and only three remained as loss-makers following acquisition (see Table 1). The critical difference was how that change was effected. Japanese companies intervened less and imposed their culture and systems on their subsidiaries far less than other nationalities. The MD of a U.K. company acquired by a Japanese firm (J10) said of his firm’s acquisition: When the acquisition took place J10 said ‘Well we know nothing about trading in the international market so therefore we’re not going to tamper with the organisation. You carry on with what you’ve been doing over the past 100 years’ . . . . It might have been better if there had been more integration, at least at an earlier stage. To this day there has been no integration. It’s almost as if the companies are associates.
So far as initiation of change was concerned the subsidiary was frequently the main source of change (see Table 2) In only one case was a ‘changemaker’ executive from Japan (J1) installed with the remit to shake up the new subsidiary in the manner of a troubleshooting U.S. executive, and this seems to have been a noticeably unsuccessful experiment. Generally, the method of bringing about change was either by leaving it to the local team, with the Japanese parent company merely encouraging or setting an example, or by the Japanese parent and subsidiary working closely in a joint team (see Table 3). The Japanese acquisitions also illustrated a point seen in the U.S. cases, which is that in any acquisition there is a balance to be struck between allowing a company the freedom to pursue it’s aims without interference from the parent, while also allowing it to benefit from the parent’s resources, such as finance, technology and R&D, that it could take advantage of as a more integrated part of a larger group. In the case of the Japanese companies though, this balance sometimes erred towards a lack of integration. The MD of the company acquired by J2 said: The Japanese financial director tells me what I need to know and I tell him what he needs to know and there is tremendous trust between us. If we had not been part of J2 we would not be here today, but they had absolutely no international experience and so said when they arrived ‘we’re not going to change anything. Everything will remain the same.’
Of all the resources put into their subsidiaries by Japanese parents, financial assistance was the most significant. There were several examples of this, and the help took several forms usually being in the form of guarantees or finance for investment. One was described by the MD of the U.K. subsidiary of a large Japanese company (J08): . . . the activities, when they were taken over by the Japanese were loss making and we needed secure guarantees. We needed time and our credit rating wasn’t very good and
Organisational Change Processes in International Acquisitions
71
we got letters of comfort from our parent company which were accepted by the banks here, and because of our parent company’s massive involvement with Japanese banks we also got funding through these banks in London.
Japanese parent companies used their domestic financial muscle to improve the subsidiary’s credit rating. As the MD of a Japanese owned U.K. merchant bank (J05) said: The only resource they put in was money. Two major defensive injections. To give you an idea how bad it was the bank had capital of £30 million. Loan losses were £120 million. It lost all its capital four times over! But our Japanese parent bank have a Comfort Letter to the Bank of England as do all foreign owned financial institutions, which is a huge advantage to us, as we as a small bank have a higher credit rating than Rothschilds, Hambros and so forth, because we have a big brother shareholder. But the Japanese produce no actual business for us.
In many cases the amount of financial support and relaxed attitude to shortterm performance surprised the U.K. managers as the MD of J6 said: I’ve had a lot of dealings with companies like TI who measure cash-flow sometimes daily; yearly would be an innovation for J6, but that’s from one extreme to the other. Over the years they have had a very, very long-term view.
Japanese parent companies provided substantial support services in only a few cases. They were generally much less involved and more inclined to help where possible only from afar (cf. Table 4). In terms of monitoring or change-making appointments Japanese companies had a varied approach. They were the only nationality to make use of monitoring ‘advisors’ located in the subsidiary (see Table 5). Although they only appointed one MD outright, they frequently made senior appointment while leaving the U.K. MD in position. They also often made junior training type appointments, which also had some monitoring effect. However, the only overall trend discernible in the appointments made was a reluctance to take the helm, and take direct responsibility for the performance of the new subsidiary. The concept of a team approach was much more in the ascendant, and although it might take longer, was no less successful than more direct methods of intervention. German Companies The German acquirers were very mixed in international experience, including major German based MNCs and other large German companies with little such experience, but seeking to increase their international reach (see Table 8). They were equally mixed in the economic condition of their acquisitions at the time of purchase (Table 1). However they consistently failed to turn around
72
DAVID FAULKNER, JOHN CHILD AND ROBERT PITKETHLY
Table 8. German Acquisitions. Acquirer G1 G2 G3 G4 G5 G6 G7 G8 G9 G10
Acquiree Major MNC (M) Large national company (L) Major Landesbank (H) Major MNC (H) Major MNC (M) MNC (H) Medium national company (L) Medium national company (L) Major MNC (H) Domestic company (L)
Acquiree
Acquiree Industry
Acquiree Condition
Family company Pharmaceutical Small regional company Engineering
Profitable Losses
City bank Specialist manufacturer Truck agency Small family company Small local company
Banking Hi-tech medical Automobiles Fire security Furniture
Just profitable Just profitable Losses Just profitable Losses
Medium national
Trailers
Profitable
Small family company Chemicals Small national company Construction
Profitable Profitable
their loss-maker. At the time of interview, the profitable companies were still profitable, the break-even one still breaking even, and the loss-makers still making losses. It became apparent that German acquirers sometimes lacked the confidence that their proven domestic management methods would translate effectively onto an international stage. Their actions tended therefore to be far from clearcut and purposeful. The U.K. Chief executive of the German engineering parent (G02) of a Tyneside company said: G2 are very long-term orientated, and that has been reflected in the culture. It is almost an adverse reflection as there is a casualness in attitude to losses. They don’t seem to view this as seriously as we do in the U.K. where a loss is a serious matter. They are very traditional, isolated in terms of the world market place. The German market is different from the world market. They are willing to let people find their feet. But they don’t bring a flow of benefits. They haven’t really tried to impose any particular industrial concepts or ways of working on us.
There was little consistency in the area of the initiation of change. Among the 10 cases investigated, the German company rarely took the lead (see Table 2). It might be possible to ascribe some of this apparent indecisiveness of German parent companies to their senior management and board structures, with their two-tier governance system and/or collegiate philosophy (Lane, 1989). The benign neglect G10s subsidiary suffered was also an opportunity for development with the German company’s support. However, that support could
Organisational Change Processes in International Acquisitions
73
lead to complacency about the availability of support to bail the company out of any mistakes. As the MD noted: It was a feeling that nobody actually cared much about us actually. I think our financial director would almost have welcomed being pressed more to be self- sufficient. It was almost a failing of the Germans to be as supportive as they were. . . . So it was a lack of leadership from the top and thus of strategic planning.
Obviously if the U.K. subsidiary looked up to a German board for guidance it could sometimes receive mixed advice as the MD of one U.K. firm found (G8): We did look to them for guidance in the early day bearing in mind that they were the owners but their ability to make decisions was somewhat strange to say the least . . . . The German company is just owned by three people all of whom are active in the business with a combined board of five people all of whom came over in the first few months to see their new acquisition. . . . to be totally honest all of their views totally differed.
However, whatever factors one might ascribe the indecisiveness to, what perhaps distinguished companies which overcame such confusion was whether they treated it as a disadvantage to be endured or as an opportunity. The same U.K. Managing Director for example having encountered a hydra-like German management style decided to propose and pursue his own plans: . . . So we sat down and said all right we know where we’ve been, we know what the U.K. market is, we know what to do. So we produced a five year business plan and rang them up and said look here.. would you like to see what we’ve got to say? [And they said] ‘that’s good get on with it’ and that’s exactly what we did.
Just as in the case of Japanese companies, German companies found independence and integration both to have benefits requiring a balance to be struck. As the U.K. MD of an engineering subsidiary (G06) said: The number of changes have been pretty minimal and have happened through their decisions but after consultation. Numbers employed have actually gone down slightly. The company’s position hasn’t changed much since the acquisition in either profits or sales.
A subsidiary often needs direction, and if this is not provided by the parent company, it has to be provided by the subsidiary management. In common with most other nationalities and especially Japanese parent companies, German parents could be a particularly useful and generous sources of funds for their subsidiaries, often supporting them to such an extent that without them the company would have ceased to exist (see Table 4). However, when comparing all the companies studied in this investigation, it appears that for any acquired subsidiary to be a success it not only requires financial support from its parent company, but also some idea about how that
74
DAVID FAULKNER, JOHN CHILD AND ROBERT PITKETHLY
support is going to be used, and about the longer term direction of the company. Just one or the other is insufficient. Moreover, while the ideas about how to direct the long-term future of the company can come from either the parent or the subsidiary, if they come from neither, the acquisition will merely drift. This seems to be precisely what happened in the case of several of the German acquisitions studied. The acquisitions by German companies can therefore perhaps be characterised by their generous financial but indecisive strategic support, and this is shown by the methods used to implement change. In three of the companies interviewed, a ‘changemaker’ in the form of a German Managing Director for the new subsidiary was appointed. In one case change came from Germany, where it was decided behind closed doors, but only after consultation with the local management team. In the remaining cases, the subsidiary took advantage of the new stability that came from being part of a financially powerful group to carry out the necessary changes themselves, often under the watchful eye of monitoring parent company executives. However, in three companies very little change had occurred at all (see Table 3). German companies were very varied in their response to new acquisitions in terms of new staff appointments. In three cases they appointed a new MD, in two cases they satisfied themselves merely with some senior appointments, and in two cases with some junior appointments. In four cases they made no appointments at all to the new subsidiary (see Table 5). This of course left the success of the new subsidiary very much at the mercy of the incumbent management, who could either use the opportunity to move on to greater things or drift in the state they were acquired in, kept afloat by the German parents’ financial support. A final comment from the U.K. MD of case G4: G4 put in some key personnel but left us well alone otherwise, as we are probably more sophisticated than G4 are. I don’t think we have learnt much from them, but they may have learnt something about the flexibility of a small company, and what they need to do to become more flexible themselves.
French Companies Most of the French acquirers in the sample were pretty experienced internationally (see Table 9) and had therefore developed a distinct style of international management. They tended to buy already profitable companies, or strategically important ones (see Table 1) which they were generally successful at bringing into profit. From the interviews it appeared that French companies were very self-confident in their approach to their new acquisitions. Indeed they were liable to display a very ‘colonial’ attitude in a number of cases,
Organisational Change Processes in International Acquisitions
Table 9. Acquirer F1 F2 F3 F4 F5 F6 F7 F8 F9 F10
Acquiree Major MNC (M) Large MN bank (H) Small Parisian bank (L) Major domestic company (L) Major MNC (H) MNC (H) MNC (H) Major MNC (H) Family company (L) Major MNC (H)
75
French Acquisitions. Acquiree
Acquiree Industry
Acquiree Condition
2 Small private companies Defence City firm Banking City firm Banking
Losses Losses Just profitable
Domestic company
Engineering
Profitable
Small domestic company IT consultancy company Brand name company Regional company Ex-subsidiary of UK plc 2 regional companies
Marketing IT consultancy Adhesives Water Pharmaceuticals Water
Just profitable Profitable Losses Profitable Profitable Profitable
discussing proposed courses of action only with fellow Frenchmen. As the MD of the subsidiary of F4 put it: The French hierarchy is a ‘power’ hierarchy, and it is the power that we find the most difficult to live with. They do have an arrogance and a dominance characteristic that one finds very difficult . . .
There was little consistency in how the French acquirers handled their new subsidiaries (see Table 2). Sometimes they initiated change unilaterally, and in one case they acted as the catalyst for change. However, in other cases the subsidiary initiated the change process itself. In one case the process was initiated by both sides and in two cases there was little change to report at all. French initiators tended to discuss matters with other Frenchmen, and then issue orders. But more often there was either a joint task force working on change, or one composed of only local management. Perhaps the best way of characterising the greater certainty regarding French companies’ attitude to their subsidiaries is that they were more adamant that, either the subsidiary was there to represent the parent’s business, and had to deliver the performance required, or else that the direction was a matter which required close involvement of French managers, seconded or appointed from the parent company. An example of the former attitude occurred in a French owned U.K. plastics company (F7) soon after the acquisition:
76
DAVID FAULKNER, JOHN CHILD AND ROBERT PITKETHLY The Technical director got up and . . . part way through his presentation . . . said to the table ‘What do you want us to do?’ and there was a pregnant silence and the (French Co.) director paused, looked straight down the table and said ‘Monsieur if we have to tell you what to do we have the wrong people’ . . . . Certainly the French have questioned our strategies, but they have expected us to identify them and make them happen.
In effect the director was saying that it’s not for the parent company to influence the subsidiary, so much as for the subsidiary to run its part of the parent’s business, albeit with some monitoring. The other attitude combined a more centralist but also independent approach. It was favoured by a French pharmaceutical company (F9), and summarised by the U.K. MD as: The philosophy of my boss was to see the company as a standalone company, the wording is affiliate rather than a subsidiary. That was a good initial aim but to be a bit critical, it didn’t give us the leverage of being part of a bigger group.
This again brings us back to the independence-integration dilemma facing parent companies in dealing with their subsidiaries. The same MD recognised this as a problem inherent in an approach involving independence: “The way in which the business is being approached is interesting ‘the centre and several self-sufficient satellites’ ”. As with the other nationalities, French companies provided the stability of finance, credit support and sometimes some actual managerial and technical support activities (see Table 4). Apart from the view of French managers as centralist bureaucrats looking down on local managers, the picture of French companies’ attitude to integrating their acquisitions is somewhat variable, and depends strongly on individual company culture and specific circumstances. The attitude of French companies to monitoring or bringing about change was also mixed. In three cases, they moved in by appointing a new MD and senior staff, but in six cases they made no new appointments and in one appointed only juniors (see Table 5). Thus while there is a noticeable tendency towards some elevated centralist approach, this may not be incompatible with allowing subsidiaries a fair degree of independence. As the MD of the U.K. subsidiary of F5 put it: “The French are very autocratic. My boss feels he is the boss and he decides, yet at the same time he doesn’t interfere with me.” However, along with this centralist and combined attitude to independence and integration the attitudes toward longer-term strategic investments provided a contrast with the American short-term profit expectations. As the U.K. Chairman of F2’s subsidiary put it:
Organisational Change Processes in International Acquisitions
77
The French are strategic to a fault. They were too tolerant of loss, much more than the British would have been. With this went a tendency to centralisation and bureaucracy and to fitting things into strategy.
CONCLUSIONS As Hampden-Turner and Trompenaars (1994) show, the forms of capitalism and hence management practice vary substantially by nationality of acquirer, and national paradigms are evident. There is therefore likely to be a variety of different methods, at least partially nationality influenced, of integrating new acquisitions. It is even possible, as this research suggests, that there is no single best-practice way of treating a newly acquired company and getting good performance out of it. The way in which individual acquirers behaved also depended in part not just on nationality, but also on the condition of the acquisition company at the time of purchase, and the acquirer’s company culture and hence preferred management methods. Thus, where an acquired company was in crisis, a more interventionist approach was adopted than in a company making acceptable profits. Furthermore acquirers with little international experience tended to be more tentative in their actions than major MNC acquirers. However, there were discernible differences between approaches by companies of different nationalities, irrespective of the international experience of the acquirer or the economic condition of the subsidiary. The American acquirers made the most effort to ensure that their acquisitions were profitable at the time of purchase. They then tended to absorb them into the parent company systems and to demand rapid high performance. One way or the other the U.S. absorptive management style usually achieved it, as at the time of interview only two U.S. acquisitions were still making losses. Japanese companies were less concerned to buy existing profitable companies, as their aims were more long-term than those of American companies. They tended to be gentler in their treatment of their acquisitions, often acting primarily as a catalyst, rather than as an owner. They were the only nationality to appoint ‘advisors’ to monitor events in the new subsidiary. Their results were no worse than those of American companies however. This was remarkable given their relative international inexperience. Considering the better initial condition of the U.S. acquisitions, this co-operative and catalytic post acquisition change style was generally very successful. The French style tended to be ‘centralist’. They either appointed a new French MD and took decisions after discussions in France, or they left the local team in day to day charge of operations but determined high level strategy at
78
DAVID FAULKNER, JOHN CHILD AND ROBERT PITKETHLY
headquarters. This contrasts somewhat with Calori, Lubatkin and Very’s (1994) experience that the French tended to exercise high formal control of both strategy and operations. The French approach was generally effective. German companies were less successful or certain in their methods. There was no discernible German method of change-making, as their actions varied from appointing an MD and giving orders to leaving well alone and hoping for the best. This eclectic, perhaps unfocussed and largely non-interventionist approach achieved no turnarounds in their troubled acquisitions, but maintained profit in their profitable ones. In line with Whitley’s (1992) thesis, institutional factors may have played a role in inhibiting leadership, particularly the dual board system, but this does not seem an adequate explanation for their poor performance. In a qualitative study conducted through interviews with British managers, it is possible to do no more that identify some hypotheses for future research. In our view, some of the most interesting hypotheses that might usefully be subjected to further testing are the following: (1) The national origin of the acquirer has significant impact on the management methods adopted in integrating a newly acquired company. (2) No one method of treating a new subsidiary is consistently more effective than another in achieving superior performance, but the confidence of the acquirer in adopting a particular method consistently is an important success factor. (3) The U.S. are typically ‘absorbers’, the Japanese ‘preservers’ or, symbiosisists’, the French ‘colonialists’ while the Germans lack a clear integration style. (4) U.S. acquirers seek short-term profits and hence tend to buy already profitable companies. (5) The Japanese, French and Germans take a longer-term strategic outlook, and are thus less concerned with their acquisition’s current profitability. The overall conclusion from the study is that clear national differences in approach to post-acquisition change do exist. Furthermore, it appears from our observations that there is no clear best practice for improving performance from an acquired company, but that self-confidence (otherwise described as ’Leadership’ cf. Olie, 1994) on the part of the acquirer in its preferred method of integration is perhaps the most important factor, as this transfers to the subsidiary a belief that it is in the hands of a ‘winner’ and leads to consequently enhanced motivation and ultimately performance.
Organisational Change Processes in International Acquisitions
79
REFERENCES Bartlett, C., & Ghoshal, S. (1989). Managing across borders. London: Hutchinsons. Biggart, N. W., & Guillen, M. F. (1999). Developing difference: social organization and the rise of the auto industries of South Korea, Taiwan, Spain and Argentina. American Sociological review, 64, 722–747. Boisot, M. (1986). Markets and hierarchies in cultural perspective. Organization Studies, 7, 135–158. Boisot, M. (1995). Information space: A framework for learning in organizations. Institutions and Culture. London: Routledge. Bowman, C., & Ambrosini, V. (1997). Using single respondents in strategy research. British Journal of Management, (June), 119–132. Boisot, M., & Child, J. (1988). The iron law of fiefs: bureaucratic failure and the problem of governance in the Chinese system reforms. ASQ, 33, 507–527. Boisot, M., & Child, J. (1996). From fiefs to clans: explaining China’s emerging economic order. ASQ, 41 600–628. Buono A. F., & Bowditch, J. (1985). When cultures collide: The anatomy of a merger. Human Relations, 38(5), 477–501. Child, J. (1981). Culture, contingency and capitalism in the cross-national study of organizations. Research in Organizational Behavior, 3, 303–356. Child, J., Faulkner, D. O., & Pitkethly, R. H. (1998). Foreign direct investment in the U.K. 1985–1994: The impact on domestic management practice. Journal of Management Studies, 37(1), 141–166. Calori, R., Lubatkin, M., & Very, P. (1994). Cross-border acquisitions: An international comparison. Organisation Studies, 15(3), 361–399. Gates, S. R., & Engelhoff, W. G. (1986). Centralization in headquarters-subsidiary relationships. Journal of International Business Studies, 17(2), 71–92. Hampden-Turner, C., & Trompenaars, F. (1993). The seven cultures of capitalism. London and New York: Doubleday. Haspeslagh, P. D., & Jemison, D. B. (1991). Managing acquisitions. New York: Free Press. Hickson, D. J., Hinings, C. R., McMillan, C. J., & Schwitter, J. P. (1974). The culture-free context of organization structure: A tri-national comparison. Sociology, 8, 59–80. Hofstede, G. (1991). Cultures and organizations: Software of the mind. Maidenhead: McGrawHill. Kogut, B., & Singh, H. (1988). The effect of national culture on the choice of entry mode. Journal of International Business Studies, 19(3), 411–433. KPMG (1997). Corporate finance survey. London. Lane, C. (1989). Management of labour in Europe. London: Elgar. Liu, I-M. (1986). Chinese cognition. In: M. H. Bond (Ed.), The Psychology of the Chinese People (pp. 73–105). Hong Kong: OUP. Loveridge, R. (1998). Review of “The changing European firm-limits to convergence”. Organization Studies, 19, 1049–1053. Mokhiber, R., & Weissman, R. (1999). Corporate predators: The hunt for megaprofits and the attack on democracy. Monroe, Maine: Common Courage Press. Morosini, P., & Sing, H. (1994). Post cross-border acquisitions: Implementing “national culture compatible strategies to improve performance”. European Management Journal, 12, 390–400.
80
DAVID FAULKNER, JOHN CHILD AND ROBERT PITKETHLY
Morris, J. J., & Wilkinson, B. (1996). The transfer of Japanese management to alien institutional environments. Journal of Management Studies, 32, 719–730. Nahavandi, A., & Malekzadeh, A. R. (1988). Acculturation in mergers and acquisitions. Academy of Management Review, 13(1), 79–90. Norburn, D., & Schoenberg, R. (1995). European cross-border acquisition: How was it for you? Long Range Planning, 27(4), 25–34. Nanaka, I., & Takeuchi, H. (1995). The knowledge creating company. New York: OUP. Olie, R. (1994). Shades of culture & institutions in international managers. Organisation Studies, 15(3), 381–405. Oliver, N., & Wilkinson, B. (1992). The Japanization of British industry: New Developments in the 1990s (2nd ed.). Oxford: Blackwell. Perlmutter, H. (1969). The tortuous evolution of the multinational corporation. Columbia Journal of World Business, 4, 9–18. Redding, S. G. (1994). Comparative management theory: Jungle, zoo or fossil bed? Organization Studies, 15, 323–359. Roberts, K. H. (1970). On looking at an elephant: An evaluation of cross-cultural research related to organizations. Psychological Bulletin, 74, 327–350. Sales, A., & Mirvis, P. (1984). When cultures collide: Issues of acquisition. In: J. R. Kimberly and R. E. Quinn (Eds), Managing Organizational Transition (pp. 103–133). Homewood, Illinois: Irwin. Simonin, B. L. (1997). The importance of collaborative know-how. Academy of Management Journal, 40, 1150–1174. Shrivastava, P. (1986). Postmerger integration. Journal of Business Strategy, 7(1), 65–76. Sorge, A. (1982). Cultured organization. International studies of management and organization, 12, 106–135. Tomohide, S. (1998). Personal communication. Whitley, R. (Ed.) (1992). European business systems: Firms and markets in their national contexts. London: Sage. Very, P., Lubatkin, M., & Calori, R. (1996). A cross-national assessment of acculturative stress in recent European mergers. International Studies of Management and Organisations, 26(1), 59–86.
MANAGERIAL PREFERENCES IN INTERNATIONAL MERGER AND ACQUISITION PARTNERS REVISITED: HOW ARE THEY INFLUENCED? Susan Cartwright and Fionnuala Price ABSTRACT Cross border mergers and acquisitions (M&As) are an integral part of international business. Although M&A activity is predominantly driven by a rational-economic model, cultural attitudes are likely to play a role in influencing selection decisions and management integration practices. This Chapter reports on a study to establish whether different national managerial groups (n = 480) have similar/dissimilar attitudinal preferences towards M&As with foreign partners. Comparisons are made with an earlier study.
INTRODUCTION The dollar value of completed mergers, acquisitions and divestitures worldwide in 2000 increased by almost 25% to more than $1.7 trillion, and set a record for the sixth consecutive year. A significant trend in the recent pattern of merger and acquisition (M&A) activity has been the increase in foreign acquisitions. Whilst the USA continues to be a major acquirer of overseas companies, the value of these deals during the period 1991–2000 was significantly less than Advances in Mergers and Acquisitions, Volume 2, pages 81–95. Copyright © 2003 by Elsevier Science Ltd. All rights of reproduction in any form reserved. ISBN: 0-7623-1003-0
81
82
SUSAN CARTWRIGHT AND FIONNUALA PRICE
the level of investment flowing into the U.S. in terms of foreign acquisitions of U.S. companies. In 2000 alone, over a thousand U.S. companies were acquired by overseas buyers at a value of $340 billion. In contrast, a little over 2,000 U.S. companies were bought by foreign acquirers in a ten year period between 1978–1988. The U.K. has similarly seen an increase in inward foreign direct investment, mainly from the USA, Japan, Germany and France (Child et al., 2000). In 1996, foreign acquisitions of U.K. companies exceeded the combined total value of all other European Union countries (KPMG, 1997). It would seem that like their U.S. counterparts, U.K. companies are both highly acquisitive and at the same time, extremely attractive acquisition targets. The global trend in M&A activity is further supported by the increasing number and value of cross border deals in Europe (see Table 1 below). Cross border deals represent a significant and growing aspect of global M&A activity far exceeding domestic deals in terms of average value and representing an increasing proportion of the total value of all deals done across the world – 41% in 2002; compared to 24% in 1996 (KPMG, 2000). The underperformance of M&As continues to be the focal point of much debate and attention. Estimates of M&A failure range from 80% (Marks, 1988) to 50% (Hunt, 1988; Weber, 1996; Cartwright & Cooper, 1997) thus emphasising the notion that M&As invariably do not deliver what is expected in terms of increased profitability or economies of scale. Whilst overall some sectors, e.g. banking and insurance, tend to record higher success rates than others in terms of enhanced shareholder value, the success of a few top
Table 1.
Cross Border Deals in Europe 1991–2000.
Year
No. of Deals
Value ($ billion)
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
1,097 1,030 875 1,096 1,372 1,294 1,422 1,602 2,236 2,422
24.0 36.2 28.9 38.9 52.9 50.6 87.8 146.2 312.4 526.7
Source: Mergers and Acquisitions, 2001.
Managerial Preferences in International Merger
83
performers often masks the failure of the majority (Financial Times, August 2000). According to a recent report (KPMG, 2000), 83% of recent deals failed to deliver shareholder value and an alarming 53% actually destroyed value. The report concludes that underperformance is the outcome of excessive focus on “closing the deal” at the expense of focusing on factors that will ensure its success. Hussey (1998) also cautions organisations to explore in detail the wider impact of strategies before committing to any strategic decision to acquire. According to Hussey, acquisition failure can be the outcome of any one or more of three failures of analysis: a misconceived action, failure to think through and implement actions that will enable the acquisition to deliver the intended benefits or financial over-extension. In the context of both domestic and international M&As, strategic fit is obviously an important consideration in target identification. However, the recognition of synergistic potential, it would seem is no guarantee that it will be realised, particularly if there is a lack of cultural fit. When Daimler and Chrysler merged in January 1999, the event was heralded as the biggest ever auto merger and the year’s smartest deal (Fortune, January 1999). As the partners’ markets scarcely overlapped the strategic fit was perfect. Yet, despite the synergistic potential, within two years the combined company was worth less than Daimler Benz was before the merger and the hostile relationships between the U.S. and German management groups were widely reported in the business press (The Economist, 13 December 2000). Clearly, the challenges of marrying the entrepreneurial style of a U.S. business with the conservatism of a German company proved more difficult than was expected (Schoenberg, 2000). Experiences such as this serve to emphasise the importance of partner or target selection based on additional and different criteria from traditional practice (Jemison & Sitkin, 1986), particularly the potential problems of cultural integration. As the level of international M&A activity has increased there has been a parallel growth in the management literature on the influence of national culture on M&A integration and outcomes, and to a lesser extent on the issue of partner/target selection. In a recent review of the literature on cross border acquisitions, Schoenberg (2000) observes that research studies confirm theoretical reasoning that the relative national cultures influence the eventual outcome of an international acquisition. Although the interaction of organisational cultures cannot be ignored (Larsson & Risberg, 1998), it has been argued that cross border combinations between organisations with similar corporate cultures may not be sufficient if the national cultures conflict. Consequently, it has been proposed
84
SUSAN CARTWRIGHT AND FIONNUALA PRICE
that the challenge for leadership in cross border M&As is to successfully accommodate and integrate both national and organisational cultures – i.e. to address what has been termed the “double-layered acculturation process” (Nahavandi & Malekzadeh, 1998). In the context of international business more generally, Zarkada-Fraser (2001) draws attention to the potential influence of national stereotypes on international relationships and decision-making. Many studies (Diamantopoulos et al., 1995; Tse et al., 1996; Zang, 1996) have demonstrated the extent to which consumer purchase decisions are influenced by country-of-origin (COO) or product country image (PCI) effects. The evaluative nature of stereotyping is considered to be a significant barrier to international understanding and cooperation (Klineberg, 1964). Importantly, as a frame of reference, the emotional nature of stereotypes can often override logic and lead to irrational decisions. By their very nature, cultural stereotypes are a condensation of reality in that they simplify and over-generalise the characteristics of a societal group. In the absence of detailed knowledge and direct experience of a potential merger partner or acquisition target, stereotypes offer a means of reducing the cognitive complexity of a decision. It has been suggested (Cooper & Kirkcaldy, 1995) that the selection of suitable international partners or collaborators is influenced, in part, by cultural stereotypes. This chapter reports on recent research conducted amongst senior-middle managers to investigate whether different national managerial groups have similar/dissimilar attitudinal preferences toward foreign M&A partners. In replicating an earlier study based on data collected in 1994 (Cartwright, Cooper & Jordan, 1995) it examines the extent to which attitudes may have changed in light of the continued growth in internationalisation.
ACCULTURATION AT THE NATIONAL LEVEL: DISTANCE AND ATTRACTIVENESS National or societal culture is a pervasive influence on the attitudes and behaviour of its members. The national culture in which an organisation operates will to a greater or lesser extent influence the values, strategy, behaviours and style of work organisation that companies will adopt (Hofstede, 1980). Ethnocentrism is identified by many studies as a barrier to international management (Cartwright & Cooper, 1996; Hodjetts & Luthans, 1994) and is defined as a belief that one’s way of doing things is superior to that of others. This perspective invokes cultural stereotyping and presents potential obstacles
Managerial Preferences in International Merger
85
to the effective integration of diverse national cultures in an international business context, particularly when the goal of many M&As is to assimilate the acquired or smaller partner into the dominant culture. Stereotyping can influence the way in which we interpret and classify behaviour we have observed and how subsequently we recall that behaviour, i.e. it can determine what we select to notice in the first place and hence what we later remember. Stereotypical attitudes are enduring and difficult to displace because of our tendency to focus on information and behaviours which reinforce our stereotype and discount, as exceptional, that which is inconsistent. Many years ago, Dunning (1958), in a study of U.S. investment into British manufacturing companies identified the tendency of foreign parents to impose their management principles and practices on their acquisitions – what is termed the “foreign practice effect”. Traditionally, the U.S. has preferred outright acquisition to merger or joint venturing. Jaeger (1983) also found a national culture pattern in ways in which organisations manage their cross border acquisitions. More recent evidence (Child et al., 2000) supports this foreign practice effect. In a study of over 200 U.K. acquisitions by U.S., Japanese, German, French and U.K. companies, it was found that the very process of being acquired led to significant changes in management practices, particularly towards more performance-related rewards. However, foreign acquired companies tended to experience a wider range of changes than domestically acquired firms. Interestingly, U.S. and French acquirers exerted more influence than Japanese or German acquirers. Furthermore, the French and German management approach did not conform to the accepted view of national management practice. In terms of selecting a compatible foreign merger partner or acquisition target, Larsson and Risberg (1998) note that organisations tend to prefer to invest in neighbouring territories or those with which they have the closest economic, linguistic and cultural ties. Schoenberg (2000) suggests that attractiveness in international M&As can be viewed in terms of cultural differences as well as cultural similarities. Whilst differences can lead to greater acculturation stress and integration difficulties, cultural differences do not necessarily result in negative outcomes. In a study of 129 European cross border acquisitions Schoenberg and Norburn (1998) found that only differences in cultural attitudes towards risk negatively impacted on acquisition performance. In contradiction to Child et al. (2000), other researchers (Larsson & Risberg, 1998) have argued that cultural differences at the national level do not have such a negative impact as differences at the organisational level in domestic M&As, because there is a greater awareness and appreciation of
86
SUSAN CARTWRIGHT AND FIONNUALA PRICE
national culture differences and a greater tolerance for multi-culturalism. In particular, a number of studies (Very et al., 1997; Calori, Lubatkin & Very, 1994; Cartwright & Cooper, 1993) have found irrespective of nationality if the buyer’s culture is perceived as being relatively less controlling then it is more likely to be perceived as being more attractive. The issue of compatibility of national cultures is frequently discussed and studied within the framework of Hofstede’s (1980) and Trompenaars’ (1993) classifications. The country indices for Power Distance, Uncertainty Avoidance, Individualism and Masculinity developed by Hofstede have provided a means of representing cultural distance between collaborating companies. Most research studies on M&As have tended to focus on the dimensions of Power Distance and Uncertainty Avoidance with equivocal results (Schoenberg, 2000). Very, Lubatkin and Calori (1998) incorporated all four dimensions into their study of the performance of U.K. and French mergers and found that French companies acquired by the British performed significantly poorer than domestic acquisitions. However, overall the results were not straightforward, making it difficult to generalise. The results of the survey conducted in 1994 (Cartwright, Cooper & Jordan, 1995) on which the following study is based, found that the mainly Northern European sample of managers showed stronger preferences for merging with other Northern European and American organisations. In addition, they ranked Japan, Italy and Spain amongst their least preferred partners. In terms of Hofstede’s (1980) classification, Northern European countries and the USA tend to cluster in terms of their orientation towards “individualism” as opposed to “collectivism” which is highly characteristic of both Japan and Spain. Furthermore, the preferences expressed were found to map the pattern of actual M&A activity during the previous year. Individualism pertains to societies in which ties between individuals are loose and where an individual is expected to look after his or herself and their own immediate family. Typically, the concept of “I” and short-term individual self-interest dominates over the wider and longer term implications of “we”. In contrast, collectivism pertains to societies in which individuals from birth onwards are integrated into strong cohesive groups, which throughout their lives, continue to protect them in exchange for loyalty. In a business context, this translates into very different employer-employee contracts and attitudes towards organisational relationships. In highly individualistic cultures like the USA and the U.K., this contractual relationship between employer-employee is based on supposed mutual advantage and
Managerial Preferences in International Merger
87
reciprocal exchange. Similarly, partnerships between organisations are predominantly founded on the opportunity to capitalise on a situation of immediate strategic advantage irrespective of the quality of the interpersonal relationship between the partners. In collectivist cultures like Japan, the relationship between employer-employee is more familial and developmental with a heavier moral foundation and more linear approach to career progression. Decisions about interorganisational collaborations are carefully and extensively considered with the focus as much on the trust and ongoing quality of the relationship between the parties as the potential strategic advantage. As individualism is strongly linked to the capitalist system, M&A activity is very “individualistic” in cultural orientation. This apparent tendency of managers to prefer national cultures which are to be perceived to be “more like us” suggests that M&A selection decisions may be influenced by an underlying desire to reduce cultural differences and avoid cultural distance. However, increased globalisation and an expansion in international management education may lead to a greater acceptance and appreciation of the potential value of cultural differences.
METHODOLOGY The questionnaire which formed the basis of this study was adapted from the 1994 survey and consisted of three parts: (i) Biographical/organisational information: items relating to nationality of respondent, nationality of parent company, organisational activity, size and experience of M&A. (ii) Attitudinal preference – this section invited respondents to place in rank order (1–3) their preferred/least preferred choice of foreign merger partner or acquirer and the rationale for their choice. (iii) Compatibility – this section required respondents to indicate which foreign country they considered was most/least compatible in terms of managerial style. In addition, data was collected on managerial expectations concerning their organisation’s future involvement in M&A activity over the next three years. Following negotiation with airport authorities at a major international U.K. airport, access was granted to the Executive lounges over a two week period during summer 2001. During this period questionnaires were distributed and
88
SUSAN CARTWRIGHT AND FIONNUALA PRICE
collected from business passengers awaiting flights. A total of 480 questionnaires were collected and analysed using SPSS for Windows.
RESULTS Biographical and Organisational Information Completed questionnaires were returned by 22 different nationalities. The majority of respondents were Northern European. Not surprisingly, given that the data were collected at a U.K. airport, 45% of the sample was British (n = 217). Swedish managers comprised 11% (n = 54) of the sample as did Dutch managers (n = 51), with U.S. managers (n = 24) and Irish managers (n = 24) both accounting for 5% of the sample. The majority of the sample worked for British (25%) or American organisations (24%). As the majority of the sample (66%) worked for organisations whose nationality was different from their own, the sample could be satisfactorily described as being international. Almost three-quarters (72%) worked for very large organisations with over 1,000 employees and in the main described their responsibilities as being in the area of strategy (34%) or operations (48%). Their organisations represented a broad spectrum of industries including manufacturing, pharmaceuticals, telecoms, healthcares, energy and financial services. As Table 2 illustrates, the respondents were representative of organisations that were highly active in M&As and were particularly acquisitive. Over half had acquired another organisation and one-third had merged during the last five years. Furthermore, 62% indicated that it was likely/extremely likely that they would make further acquisitions within the next three years. In the same period Table 2.
Type of Activity in Which Organisation Had Been Involved During the Last 5 Years (1996–2001).
Merged Acquired another organisation Been taken over Target of an unsuccessful bid Party to a joint venture Entered some other form of strategic alliance
No.
%
153 276 69 33 145 159
32 58 14 7 30 33
Managerial Preferences in International Merger
89
86% indicated that they expected to become involved in joint ventures/strategic alliances, while 20% expected to merge. Compared with 1994 data, the pattern indicates a further significant increase in M&A activity. Attitudinal Preferences Table 3 shows the overall and highest ranking ‘preference dimension’ choices for each of the analysed nationalities. This analysis was restricted to national groups which contained at least 10 respondents. It is notable that the total sample selected the U.S. as the most popular (28%) 1st preference, with the U.K. being a close second (25%). Forty-seven percent of German respondents chose their own nationality, a similar percentage of U.S. respondents (46%) also chose their own nationality; with 40% of U.K. respondents choosing their own nationality. All these respondents were currently working for organisations of different national parentage to their own. In 69% of cases the reasons given for their choices was perceived cultural compatibility, which was cited as being at least four times as important as market potential or management approach. Finally, an analysis was conducted to investigate the combined frequency of the 1st, 2nd and 3rd ranking. The results found that the U.S. remains the most preferred merger choice for the total sample (18%) with the U.K. again a close second (16%). However, the France (4%), German (10%) and Switzerland (5%) also entered the rankings. Table 4 shows the least preferred choices of merger partner or acquirer for the overall sample and the subset of national groups. While Japan emerged as the least preferred choice for the overall sample and the nationality subsets, it is interesting to consider the second preferences which tend to mirror the 1994 survey data. Table 3.
Total Sample USA U.K. Danish Dutch German Irish Swedish
Most Preferred Merger Partner or Acquirer. No. of Cases
1st Preference
480 24 217 19 51 15 24 54
U.S. U.S. U.S. U.S. U.S. German U.S. U.K.
90
SUSAN CARTWRIGHT AND FIONNUALA PRICE
Table 4.
Total Sample USA U.K. Danish Dutch German Irish Swedish
Least Preferred Merger Partner or Acquirer. No. of Cases
Least Preferred
2nd Least Preferred
480 24 217 19 51 15 24 54
Japan Japan Japan Japan Japan Japan Japan Japan
France France France France Italy/France France Ireland Italy/USA
A range of issues relating to incompatible culture and differences in working practices, e.g. lack of directness, slow decision-making, dominate the reasons given by respondents in choosing Japan. Further analysis was conducted to investigate the combined frequency of the 1st, 2nd and 3rd rankings. The results found that Japan remained the least preferred partner followed closely by France and Italy. Finally, the results were examined using the country indices for the four dimensions developed by Hofstede (1980). Table 5 relates to the Power Distance dimension. The data shows a uniform clustering of nationalities choosing either their own nationality (hence an identical power distance rating) or nationalities with very similar positions on the Power Distance Index. It is notable that Danish managers expressed a preference for merging or being acquired by a national
Table 5. Subset Nationality U.S. U.K. Danish Dutch German Irish Swedish
Power Distance.
Power Distance
1st Preference
Power Distance
40 35 18 38 35 28 31
U.S. U.K. U.S. U.S. German U.S. U.K.
40 35 40 40 35 40 35
Managerial Preferences in International Merger
Table 6. Subset Nationality
91
Individualism/Collectivism.
Individualism/ Collectivism Index
1st Preference
Index Score
91 89 74 80 67 70 71
U.S. U.K. U.S. U.S. German U.S. U.K.
91 89 91 89 67 89 89
U.S. U.K. Danish Dutch German Irish Swedish
culture which is associated with rather more formality and greater power distance. However, Danish culture scores very low on this dimension comparative to most other countries. It is notable that the Power Distance Index for Japan, the least preferred nationality, is 54 which is very much higher than the nationalities sampled. Table 6 presents the results in terms of the Individualism/Collectivism dimension. In the context of partner preference the results support the notion that there is a cultural attraction amongst countries which are high on individualism and an avoidance of highly collectivist cultures such as Japan which has a score of 46 on this index. Table 7 compares the results in terms of the Masculinity/Femininity dimension.
Table 7. Subset Nationality U.S. U.K. Danish Dutch German Irish Swedish
Masculinity/Femininity.
Masculine/ Feminine Index
1st Preference
Index Score
62 66 16 14 66 68 5
U.S. U.K. U.S. U.S. German U.S. U.K.
62 66 62 62 66 62 66
92
SUSAN CARTWRIGHT AND FIONNUALA PRICE
Table 8. Subset Nationality U.S. U.K. Danish Dutch German Irish Swedish
Uncertainty Avoidance.
Uncertainty/ Avoidance Index
1st Preference
Index Score
46 35 23 53 65 35 29
U.S. U.K. U.S. U.S. German U.S. U.K.
46 35 46 46 65 46 35
The countries represented in the sample differ significantly in their orientation on this dimension. Denmark, The Netherlands and Sweden are all regarded as feminine cultures which place a high value on relationships, caring and the quality of life. In contrast, the USA, U.K. and Germany are considered to be high on masculinity, although Japan is the most “macho” culture with a score of 95. Table 8 relates to the Uncertainty Avoidance dimension. Low scores are indicative of a greater propensity to tolerate uncertainty and to take risk. With the exception of Germany, all the countries represented in the sample are considered to be relatively high risk takers. In contrast, Japan has an extremely high Uncertainty Avoidance index of 92 and is risk averse preferring to plan carefully and take a longer-term perspective.
SUMMARY AND CONCLUSIONS Over the last decade, considerable effort has been expended in increasing the awareness of both researchers and practitioners to the importance of human factors in M&As. Certainly the potential importance of culture is increasingly recognised as a factor which influences M&A integration and subsequent outcomes. However, the complex interplay between national and organisational culture and its level and direction of influence is still perplexing in light of the current and still limited research evidence. Rather than focussing on the inherent characteristics of different national cultures, it may also be fruitful to focus more on their inherent attitudes and strategies towards M&A activity and their level of cultural tolerance. For example, it is feasible to hypothesise that some cultures may be more inclusive and less ethnocentric than others. It has been argued that better M&A outcomes could be achieved if more attention was paid to culture at the selection stage. In the context of
Managerial Preferences in International Merger
93
international M&As, culture may already influence partner selection decisions. The results of this study show that, given a choice, managers would choose to merge or be acquired by a foreign national culture which they perceive to be similar and hence compatible with their own, and are highly avoidant of cultural distance. The evidence of this study suggests that managerial attitudes have changed little over the last 7 years or so. The highly individualistic cultures of the USA, U.K. and Northern Europe tend to prefer to enter M&As with each other rather than with Southern European or Japanese partners. Perhaps because as highly individualistic cultures, they are perceived more likely to recognise and accept the instrumentality of a merger or acquisition and are more familiar, so to speak, with the “rules of the game”. This study is based on a limited sample of European nationalities. It would be interesting to conduct a similar study with a sample drawn predominantly from more collectivist cultures. It is impossible to determine to what extent the decisions made by the managers in this study were influenced by cultural stereotypes, management education (e.g. exposure to Hofstede’s theories) or direct experience. Without doubt, management educators and researchers have a role to play in influencing international management practice. Consequently, there is a pressing need for further research in this area. Existing studies have heavily relied upon the Hofstede (1980) model of national culture despite its acknowledged limitations (Very et al., 1998). Although not a direct replication, the later work of Trompenaars (1993) suggests that the relative position of some countries, particularly the developing economies, in respect of certain dimensions may have shifted over time. Therefore, it may be an appropriate time to conduct a large-scale study to review and possibly revise the index scores. Alternatively, it may be helpful to develop new cultural measures specifically tailored for use in M&A research.
REFERENCES Calori, R., Lubatkin, M., & Very, P. (1994). Control mechanisms in cross border acquisitions: An international comparison. Organisation Studies, 15, 361–379. Cartwright, S., & Cooper, C. L. (1997). Managing mergers, acquisitions and strategic alliances: interpreting people and cultures. Oxford: Butterworth Heinemann. Cartwright, S., Cooper, C. L., & Jordan, J. (1995). Managerial preferences in international merger and acquisition partners. Journal of Strategic Change, 4, 263–269. Child, J., Faulkner, D., & Pitkethly, R. (2000). Foreign direct investment in the U.K. 1985–1994: The impact on domestic management practice. Journal of Management Studies, 37(1), 141–167.
94
SUSAN CARTWRIGHT AND FIONNUALA PRICE
Cooper, C. L., & Kirkcaldy, B. D.(1995). Executive stereotyping between cultures: The British vs. German manager. Journal of Managerial Psychology, 10(1), 3–6. Diamantopoulos, A., Schlegelmilch, B. B., & Preez, J. P. (1995). Lessons from pan-European marketing? The role of consumer preferences in fine-tuning the product-market fit. International Marketing Review, 12(2), 38–52. Dunning, J. H. (1958). American investment in British manufacturing industry. London: Allen & Unwin. Hodgetts, R. D., & Luthans, F. (1994). International management (2nd ed.). New York: McGrawHill. Hunt, J. (1988). Managing the successful acquisition: A people question. London Business School Journal, (Summer), 2–15. Hofstede, G. (1980). Cultures consequences: International differences in work-related values. Beverly Hill CA: Sage. Hussey, D. (1998). Strategic management: Past experiences and future directions. In: D. Hussey (Ed.), The Strategic Decision Challenge. Chichester: John Wiley & Sons Ltd. Jemison, D. B., & Sitkin, S. B. (1986). Acquisitions: The process can be a problem. Harvard Business Review, 60(6), 109–121. Klineberg, O. (1964). The human dimension in international relations. New York: Holt, Rinehart & Winston. KPMG (1997). Mergers and acquisitions A global research report – unlocking shareholder value: the keys to success. Amsterdam: KPMG. KPMG (2000). Dealwatch. Amsterdam: KPMG. Larsson, R., & Risberg, A. (1998). Cultural awareness and national vs. corporate barriers to acculturation. In: M. C. Gertsen, A-M. Søderberg & J. E. Torp (Eds), Cultural Dimensions of International Mergers and Acquisitions. Berlin: Walter de Gruyter. Marks, M. L. (1988). The merger syndrome: The human side of corporate combinations. Journal of Buyouts and Acquisitions, (Jan/Feb), 18–23. Nahavandi, A., & Malekzadeh, A. R. (1998). Leadership and cultural transnational strategic alliances. In: M. C. Gertsen, A-M. Søderberg & J. E. Torp (Eds), Cultural Dimensions of International Mergers and Acquisitions. Berlin: Walter de Gruyter. Schoenberg, R., & Norburn, D. (1998). Leadership compatibility and cross border acquisition outcome. Paper presented to 18th Annual Strategic Management Society. International Conference, Orlando. Schoenberg, R. (2000). The influence of cultural compatibility within cross-border acquisitions: A review. In: C. L. Cooper & A. Gregory (2000). Advances in Mergers and Acquisitions (Vol. 1). New York: JA1. Trompenaars, F. (1993). Riding the waves of culture. London: Economist Books. Tse, A., Kwan, C., Yee, C., Wah, K., & Ming, L. (1996). The impact of country-of-origin on the behaviour of Hong Kong consumers. Journal of International Marketing and Marketing Research, 21(1), 29–44. Very, P., Lubatkin, M., Calori, R., & Veiga, J. (1997). Relative standing and the performance of recently acquired European firms. Strategic Management Journal, 18, 593–614. Very, P., Lubatkin, M., & Calori, R. (1998). A cross-national assessment of acculturative stress in recent European mergers. In: M. C. Gertsen, A-M. Søderberg & J. E. Torp (Eds), Cultural Dimensions of International Mergers and Acquisitions. Berlin: Walter de Gruyter. Weber, Y. (1996). Corporate culture fit and performance in mergers and acquisitions. Human Relations, 49(9). 1181–1202.
Managerial Preferences in International Merger
95
Zang, Y. (1996). Chinese consumers’ evaluation of foreign products: The influence of culture, product types and product presentation format. European Journal of Marketing, 30(12), 50–68. Zarkada-Fraser, A. (2001). Stereotyping in international business. In: C. L. Cooper, S. Cartwright & P. C. Early (Eds), The International Handbook of Organizational Culture and Climate. Chichester: John Wiley & Sons.
VALUE CREATION IN LARGE EUROPEAN MERGERS AND ACQUISITIONS Marc Goergen and Luc Renneboog ABSTRACT In this paper, we analyse the short-term wealth effects of large (intra)European takeover bids. We find large announcement effects of 9% for target firms, but the cumulative abnormal return that includes the price run-up over the two-week period prior to the event rises to 20%. The share price of the bidding firms reacts positively with a statistically significant announcement effect of only 0.7%. We also show that the type of takeover bid has a large impact on the short term-wealth effects of target and bidder shareholders with hostile takeovers triggering substantially larger price reactions than friendly mergers and acquisitions. When a U.K. target or bidder is involved, the abnormal returns outperform those of Continental European bids. We also find strong evidence that the means of payment in an offer has a large impact on the share price reactions. A high market-to-book ratio of the target leads to a higher bid premium but triggers a negative price reaction for the bidding firm. Bidding firms should not further diversify by acquiring target firms that do not match the bidder’s core business. We also investigate whether the predominant reason for takeovers is synergies, agency problems or managerial hubris. We find a significant positive correlation between target shareholder gains and total gains for the merged entity as well as between target gains and Advances in Mergers and Acquisitions, Volume 2, pages 97–146. Copyright © 2003 by Elsevier Science Ltd. All rights of reproduction in any form reserved. ISBN: 0-7623-1003-0
97
98
MARC GOERGEN AND LUC RENNEBOOG
bidder gains. This suggests that synergies are the prime motivation for bids and that targets and bidders tend to share the resulting wealth gains.
1. INTRODUCTION The clustering of mergers and acquisitions through time is a striking phenomenon. The first European merger wave started approximately in 1880 and ended in 1904. It was fuelled by the industrialisation started by the discovery of the steam engine. The prime incentive for these mostly horizontal mergers was the creation of monopolies. The introduction of anti-trust legislation played an important role in the second merger wave which started in 1919 and continued throughout the 1920s. The old monopolies were broken up and mergers and acquisitions were used to achieve vertical integration. The third merger wave started in the 1950s, but reached its peak only in the mid1960s. The focus turned towards diversification and the development of large conglomerates to face global markets. The development of new financial instruments and markets (e.g. the junk bond market) facilitated the financing of acquisitions and led to the fourth wave (1983–1989) which was fuelled by technological progress in biochemistry and electronics. The fifth M&A wave emerged in the 1990s (1993–2000) and went hand in hand with a long economic boom period, stock exchange development and the growth in the internet- and telecommunications industries. In 2001, the collapse of consumer confidence in these industries as well as the overcapacity in the traditional sectors caused an abrupt reduction in merger activity. Over 1993, the total dollar value paid for target firms in the U.S. and Europe doubled after 4 consecutive years of reduction in M&A activity. A sharp rise could be observed as of 1996: the total value of U.S. and European acquisitions rose to USD 1,117 million (with Europe accounting for 37%). In the following years the M&A wave gained even more strength with a value of USD 1,574 million in 1997 (35% of which was realised in Europe), USD 2,634 million in 1998 (33% in Europe), USD 3,319 million in 1999 (47% in Europe), USD 3,451 million in 2000 (43% in Europe). The year 1999 was not only a remarkable year, because the European M&A market was now almost as large as the U.S. market or because 12% of the total value was accounted for by deals in excess of USD 100 billion, but also because an exceptionally high number of hostile takeovers took place in Europe. There were 369 hostile bids in Europe in 1999 compared to only 14 in 1996, 7 in 1997, 5 in 1998 and 35 in 2000.1
Value Creation in Large European Mergers and Acquisitions
99
In this paper, we aim at investigating the wealth effects within the major mergers and acquisitions wave of the 1990s. We have chosen 1993 as the starting year for the following two reasons. First, it is the first year with a substantial increase in M&A activity since 1989. Second, the Maastricht Treaty of 1992 cleared the way for a European Monetary Union culminating in the introduction of a single currency in 1999. As most of the M&A research concentrates on the U.S. and U.K. markets and most studies concentrate on M&As in one single country, we believe that a European study will yield interesting results. The paper is organized as follows: Section 2 summarises the main findings of previous studies on mergers and acquisitions. Section 3 describes the data sources, variables and methodology. Section 4 investigates the short-term wealth effects for target and bidder firms of mergers, friendly acquisitions and hostile takeovers. A correlation analysis in Section 5 sheds some light on the type of takeover motives. Section 6 then analyses the determinants of the market price reactions to M&A announcements and Section 7 concludes.
2. THE DETERMINANTS OF BIDDER AND TARGET RETURNS IN THE LITERATURE The literature on the wealth effect of M&A announcements for target shareholders is unanimous: these shareholders receive an average premium within the 20 to 30% range over and above the pre-announcement share price. For the 1960s to 1980s, Jarrell and Poulsen (1989), Servaes (1991), Kaplan and Weisbach (1992), for instance, report average U.S. target share price returns of 29% for 1963–1986, 24% for 1972–1987 and 27% for 1971–1982, respectively. In the 1990s, the U.S. abnormal announcement returns remained at a similar level of 21% (Mulherin & Boone, 2000). In contrast, there is little consensus about the announcement wealth effects for the bidding firms. About half of the studies report small negative returns for the acquirers (see e.g. Walker, 2000; Mitchell & Stafford, 2000; Sirrower, 1994; Healy, Palepu & Ruback, 1992) whereas the other half finds zero or small positive abnormal returns (see e.g. Eckbo & Thorburn, 2000; Maquiera et al., 1998; Schwert, 1996; Loderer, 1990). Considering that the average target is much smaller than the average acquirer, the combined net economic gain at the announcement is positive, albeit very small. The M&A literature has discovered a variety of profitability drivers. First, the announcement of tender offers and hostile takeovers generates higher target as well as bidder returns than the announcement of friendly mergers or acquisitions (see e.g. Gregory, 1998; Loughran & Vijh, 1997; Lang et al.,
100
MARC GOERGEN AND LUC RENNEBOOG
1989). Second, when the bidding management owns large equity stakes, bidding firms obtain higher returns (see e.g. Healy et al., 1997; Agarwal & Mandelker, 1987). This suggests that managers are more likely to undertake value-destroying mergers, if they do not own equity in their firm. Third, allcash bids generate higher target and bidder returns than stock-for-stock acquisitions (see e.g. Yook, 2000; Franks & Harris, 1989; Franks et al., 1988; Huang & Walking, 1989). The announcement that the takeover will be paid with equity may signal to the market that the bidding managers believe that their firm’s shares are overpriced. This is in line with the fact that managers time the issues of shares to occur at the high point of the stock market cycle. Fourth, acquiring firms with excess liquidities destroy value by overbidding. Several papers show evidence that free cash flow (Jensen, 1986) is frequently used for managerial empire-building purposes (see e.g. Servaes, 1991; Lang et al., 1991). Fifth, corporate diversification strategies destroy value (Maquiera et al., 1998; Berger & Ofek, 1995). This confirms that companies should not attempt to do what investors can do better themselves, i.e. creating a diversified portfolio. Sixth, the acquisition of value-companies leads to higher bidder and target returns. Rau and Vermaelen (1998) show that the acquisition of firms with low market-to-book firms generates high returns (of about 12% on average) to the shareholders of the bidding firm whereas the acquisition of glamour firms (with high market-to-book ratios) leads to substantial negative returns.2 The main motive for mergers and acquisitions is the value creation resulting from synergies. These synergies are called operating synergies, if there are economies of scale or scope, and are called informational synergies, if the combined value of the assets of the two firms is higher than the value the stock market attributes to the assets. For example, informational synergies consist of the creation of an internal capital market: slack-rich firms with poor investment possibilities acquire slack-poor firms with outstanding growth opportunities.3 Informational synergies can also consist in minimising transaction costs or bankruptcy costs. However, Warner (1977) shows that the reduction in direct bankruptcy costs (due to less than perfectly correlated earnings of the bidder firm and the target firm) is small. In Anglo-American markets, the role of hostile takeovers as a disciplinary force to remove poorly performing management is also often emphasized. This market for corporate control seems to be more active in the U.S. (Morck et al., 1988; Bhide, 1989; Martin & McConnell, 1991) than in the U.K. (Franks, Mayer & Renneboog, 2001). In this paper, we investigate the short-term returns in large European takeovers. We also analyse whether the type of offer has an important impact on the premium paid for the target’s shares. Furthermore, we distinguish among
Value Creation in Large European Mergers and Acquisitions
101
different means of payment of the bid: all-cash offers, all-equity offers or bids combining cash, equity and loan notes. Given that the level of stock market development and the corporate governance regulation differs substantially between the U.K. and Continental Europe, we investigate whether the abnormal returns for target and bidder firms are significantly different. Industry effects and year-of-bid effects are also analysed. We examine the announcement effect of unsuccessful bids in order to see whether the market already accounts for this ex post effect at the moment of the first announcement.
3. DATA AND METHODOLOGY 3.1. Sample Selection and Data Sources Data on European acquisitions – involving both a European bidder and target – were collected from the ‘Foreign deals’ section of the monthly publication Mergers & Acquisitions Report for the period 1993–2000. This report gives the names of the firms involved in the acquisition, the value of the transaction (in USD and local currency) and the type of deal (merger, acquisition, acquisition of majority/minority control, or divestiture). Additional information, such as the means of payment in the offer, the status of the bid (hostile or friendly) and multiple bidder involvement is also frequently reported. To be included in our sample, either the bidder or the target (or both) must be listed on a European stock exchange, and the announcement date must be available. We restricted the sample to large acquisitions only, with a deal value of at least USD 100 million (equivalent to about 100 million at the current exchange rate). We also used information from the Financial Times (FT) to check the data quality from the Mergers & Acquisitions Report (hereafter the Report), and to collect missing announcement dates or other missing information. We also required that at least two articles about the mergers and acquisitions were published in the Financial Times so as to exclude non-recurring rumours. The resulting sample consists of 228 merger or acquisition announcements. We also include the cases where a bid is made for part of a firm (a divestiture). In these cases, the target share price reaction is that of the divesting firm. We adopt the distinction between mergers and acquisitions made by the FT and the Report. Both sources describe a merger as a transaction between two parties of roughly equal size, whereas in an acquisition the larger party takes over the smaller one. A takeover (attempt) is classified as hostile, if the potential target rejects the offer for whatever reason. Hostility may, among others, result from a bargaining strategy to extract a higher premium for the target shareholders (Schwert, 2000) or from the target’s directors’ viewpoint
102
MARC GOERGEN AND LUC RENNEBOOG
that the proposed plan is incompatible with the target’s strategy. We also classify an acquisition with multiple bidders as hostile and report these cases separately. Lack of share price and/or accounting information reduced the sample to 185 offer announcements. For these cases, 139 bidders and 129 targets are listed. The final sample consists of 55 mergers, 40 (friendly) acquisition bids, 40 hostile takeover bids and 21 hostile takeovers with multiple bidders. In addition, 29 of the 185 announcements refer to divestitures. Information on share prices and market indices, on the risk-free rate by country (3-month Treasury Bill rates), on risk measures and accounting information was collected from Datastream. For both target and bidder, we collected the following additional information using Datastream and the Financial Times. Industry (SIC) codes indicate the degree of corporate diversification. The market capitalization of target and bidder (MV), the market to book-value (MVBV) representing the growth potential of the target, dividend yield (DY), interest coverage (IC) representing potential financial distress, the amount of liquid assets (NC) (cash and short term loans, deposits and investments) divided by total market value, the price to cash flow (PCf), assets per share (APS; measured as net tangible assets (fixed assets less depreciation, plus longer-term investments and current assets, less current and deferred liabilities) divided by the number of shares at year-end), and the return on equity (ROE) were also collated. We calculated relative target to bidder size using the market capitalisation at least 6 months prior to the announcement. Finally, we also recorded the country in which the targets and bidders are located: our sample embraces firms from 18 European countries. 3.2. Methodology We measure the short-term wealth effects for bidding and target firms by calculating cumulative abnormal returns in an event study. The event window starts 6 months prior to the announcement date to capture the effects of rumours or insider trading. There is little consensus on when to start measuring announcement period returns, as evidenced by the great variety of practice in published work. The precision when using a short measurement period is doubtful, if there is a leakage of information before the first mention in the financial press. On the other hand there is evidence that bids follow positive movements in the acquirer’s stock price, with the danger that starting the measurement period too early will lead to an overstatement of the takeover returns.
Value Creation in Large European Mergers and Acquisitions
103
To calculate the expected returns and verify the robustness of the returns, we use 6 different measures of beta. First, we estimate the beta by running the market model over a 9-month period (195 trading days) ending 6 months prior to the event date. Second, as the beta from the first method is calculated over a period well before the event date, we estimate the beta over the 9-month period ending 1 month prior to the event date. This second method may more accurately incorporate recent changes in systematic risk, but in turn may be influenced by the event. Third, we use the Datastream beta which is corrected for mean-reversion. Fourth, we also adjust betas for mean-reversion using the Merrill Lynch method based on Blume (1979), in the following way: ai = 0.34 + *i 0.67 where ai is the beta adjusted for mean-reversion and i is the beta estimated using the market model over a 9-month period ending 6 months prior to the event. Fifth, the betas from method 1 are corrected for regression to the mean according to Vasicek’s technique using Baysian updating (Vasicek, 1973). The degree of adjustment towards the mean depends on the sampling error of beta: vi = [2i /(2*1 + 2i1] · *1 + [2*1 /(2*1 + 2i1)] · i1, where vi is the Vasicek-beta for security i, *1 is the average beta across the sample of shares estimated over a 9-month period ending 6 months prior to the event date (period 1), i1 is the beta from the market model over period 1, 2i1 is the variance of the estimate of beta for security i measured over period 1, and 2*1 is the variance of the average beta measured over period 1 (Elton & Gruber, 1995). Sixth, we calculate Dimson-betas to control for inaccurate betaestimation resulting from thin trading which biases beta downwards (Dimson, 1979; Marsh & Dimson, 1983). These betas are the sum of 5 parameter estimates of the market model in which the current level of the daily market return, as well as its first three lags and one lead are included. The model is estimated over a 9-month period ending 6 months prior to the event date.4 For all 6 estimation methods, the betas are trimmed at the 5%–95% distribution range. As none of the main results of this study are significantly influenced by the choice of beta estimation technique, we only report results based on the Dimson-betas corrected for thin trading. The abnormal returns are calculated as the difference between the actual daily returns and the expected returns obtained from the CAPM. The cumulative average abnormal returns (CAAR) are then calculated over the event period. The standard significance tests we apply are the ones from e.g. Kothari and Warner (1997). The one-day test statistic is: AR (AR)
104
MARC GOERGEN AND LUC RENNEBOOG
where
t = 41
(AR) =
1 (ARt AR )2 199 t = 240
The test statistic for CAAR is CAAR (AR)T
where T is the number of time observations. The total gain for each pair of target firm and acquiring firm is measured by: CAARTotal =
CAARTarget * MVTarget + CAARAcquirer * MVAcquirer MVTarget + MVAcquirer
where MV denotes the market value of the target’s or acquirer’s stock before the beginning of the event window (Cybo-Ottone & Murgia, 2000).
4. SHORT TERM SHAREHOLDER WEALTH EFFECTS 4.1. Target Versus Bidding Firms Figure 1a shows that the announcement of a takeover bid causes substantial positive abnormal returns for the sample of European bids. On the event day, a return of 9% is realised. Strikingly, as the cumulative abnormal returns over an event window starting one month prior to and including the event date amount to about 21% (panel A of Table 1), it seems that the takeover bid was anticipated, probably as a result of rumours or of insider trading. On average, investors owning a target company for a period starting 3 months prior to the event date (60 trading days) and selling at the end of the event day would earn a return of 24%. After about 30 trading days, the average cumulative abnormal return decreases by about 3% as a result of the fact that some takeover bids are unsuccessful or the fact that a long period to finalise the offer raises doubts about the ultimate success of the negotiations.5 The effect of the takeover announcement on the wealth of the bidders’ shareholders is very small (Fig. 1b): at the announcement, the share price only rises by 0.7% (significant at the 1% level) (panel B of Table 1). All other abnormal returns in an event window of 6 months are not statistically significant. The reason why the share prices of the 139 bidding firms hardly change in this event window compared to the target firms is that the
105
Fig. 1a.
Cumulative Abnormal Returns for Target Firms.
Value Creation in Large European Mergers and Acquisitions
106
MARC GOERGEN AND LUC RENNEBOOG
Table 1.
Cumulative Abnormal Returns for Target and Bidding Firms.
This table shows cumulative abnormal returns over several event windows for target firms and bidder firms. ***, ** and * stand for statistical significance at the 1%, 5% and 10% level, respectively. Source: own calculations. Panel A: Target firms Time Interval Event date [–1, 0] [–2, + 2] [–5, + 5] Window prior to and at event [–10, 0] [–20, 0] [–60, 0] Window around event [–30, + 30] [–60, + 60] [–90, + 90] Observations
CAAR (%)
t-value
9.01 12.96 15.92
29.53*** 26.88*** 22.25***
19.13 21.58 24.12
19.35*** 18.62*** 17.56***
23.43 21.78 21.59 129
13.91*** 9.18*** 7.44***
CAAR (%)
t-value
0.70 1.18
2.98*** 3.18***
Panel B : Bidding firms Time Interval Event day [–1, 0] [–2, + 2] Window around event day [–30, + 30] [–60, + 60] [–120, + 120] Observations
0.39 –0.48 0.41 139
0.30 –0.26 0.16
announcement effect is an amalgamation of positive and negative share price reactions depending on the type of the bid (see next section). The cumulative abnormal returns obtained by this study are close to the ones reported by Franks and Harris (1989) and Higson and Elliott (1998). Franks and Harris report CAARs of 21% for large U.K. targets and of 0% for U.K. bidders over the period 1955–1985 in the event month. Higson and Elliott find CAARs of 30% for the target shareholders in the largest bids and of 0% for the bidding shareholders over the period 1975–1990. Recent research on the wealth
107
Fig. 1b. Cumulative Abnormal Returns for Bidding Firms.
Value Creation in Large European Mergers and Acquisitions
108
MARC GOERGEN AND LUC RENNEBOOG
effects over longer time periods for cross-border acquisitions by U.K. firms reveals that such operations do not generate any gain (nor loss) for the bidders’ shareholders (Gregory & McCorriston, 2002; Conn et al., 2001) 4.2. Hostile Versus Friendly Takeover Bids We also analyse the market reactions to the different types of takeover. For the targets, we distinguish between mergers (40 cases), acquisitions (18 cases), hostile takeovers (28 cases) and hostile takeovers with multiple bidders (14 cases). For all of these types of takeovers, Fig. 2a shows a strong positive announcement effect (significant at the 1% level). As expected, hostile takeover bids trigger the largest share price reactions for the target ( + 13%) on the announcement day. These reactions are significantly higher than the ones for the other types, i.e. 9% for mergers and 6% for acquisitions (see panel A of Table 2). When a hostile bid is made, the share price of the target immediately reflects the expectation that opposition to the bid will lead to upward revisions of the offer price. Surprisingly, the announcement reaction to a takeover bid with multiple bidders is low at 7%, but Fig. 2a depicts that a large upward price movement starts already 1.5 months prior to the announcement. Figure 2a also depicts that there are large differences in the price run-ups for the different types of bids. Whereas the upward price reactions prior to the bid announcement are limited to two weeks for hostile takeovers and for friendly acquisitions, it seems that in the case of mergers, rumours or insider trading occur already 1.5 to 2 months prior to the announcement (compare the cumulative abnormal returns for the event windows [–10, + 1] and [–40, + 1] in panel A of Table 2). A hostile takeover announcement generates a CAAR of more than 25% over the 2 week-period preceding and including the announcement day. At the event date and over the 2 months prior to the first announcement of the bid, the returns to the target shareholders for hostile takeovers vastly outperform those of friendly mergers and acquisitions (panel B of Table 2). The difference in returns between merger and friendly acquisition announcements is limited to the event date and to the 2-week period prior to the announcement. For the longer symmetric event windows (6 months and longer) differences between the types of bids are no longer statistically significant. Figure 2b breaks down the CAAR for the bidder by type of bidding firm. Shareholders of bidding firms clearly react differently to announcements of mergers, acquisitions and hostile takeovers. The instantaneous price reaction on the event day is 2.2% and 2.43% for mergers and unopposed acquisitions, respectively (panel C of Table 2). However, on average, the bidder’s
Fig. 2a.
Cumulative Abnormal Returns of Target Firms by Type of Bid.
Value Creation in Large European Mergers and Acquisitions 109
Cumulative Abnormal Returns of Target and Bidding Firms by Type of Bid.
Merger
% 5.96 11.33 20.01 20.34 23.49 26.52 23.99 18
12.54*** 6.04*** 8.12*** 6.55*** 6.08***
Acquisition
19.00*** 17.24***
t-value
4.52*** 3.62** 2.68**
10.98*** 5.41***
6.34*** 7.62***
t-value
Event Window [–1, 0] [–2, + 2] [–10, 0] [–40, 0] [–30, + 30] [–60, + 60] [–90, + 90]
% 3.81 5.33 12.44 5.82 13.94 4.77 2.05
Hostile takeovers – Mergers
7.59*** 6.72*** 6.47*** 3.59*** 5.03*** 1.22 0.43
t-value difference
% 6.64 6.62 5.63 8.89 11.20 1.85 4.85
Hostile takeovers – Acquisitions
10.40*** 6.56*** 4.36*** 3.78*** 3.18** 0.37 0.80
t-value difference
Panel B : Significance of differences in target CAARs across types of takeover bids
Event day % [–1, 0] 8.80 [–2, + 2] 12.62 Window prior to and at event [–10, 0] 13.20 [–40, 0] 23.41 Window around event [–30, + 30] 20.76 [–60, + 60] 23.59 [–90, + 90] 26.79 Observations 40
Time Interval
Panel A: CAARs of target firms by type of takeover bid
% 5.63 6.67 10.59 5.55 9.67 7.84 8.44
Hostile takeovers – Multiple bidders
34.69 28.36 28.83 28
25.64 29.23
% 12.60 17.95
Hostile Takeover
8.67*** 6.49*** 4.25*** 3.51*** 2.70** 1.55 1.37
t-value difference
11.37*** 6.60*** 5.49***
10.84*** 6.79***
22.81*** 20.54***
t-value
% 2.83 1.29 –6.81 3.07 –2.73 –2.92 2.80
Mergers – Acquisitions
25.03 20.53 20.39 14
15.05 23.68
% 6.98 11.28
Multiple Bidders
5.16*** 1.48 –3.28*** 1.68 0.90 0.68 0.54
t-value difference
5.60*** 3.26*** 2.65**
4.23*** 2.87***
8.62*** 8.82***
t-value
This table shows cumulative abnormal returns over several event windows for target firms and bidder firms by type of takeover bid (merger, friendly acquisition, hostile takeover, hostile takeover with multiple bidders). ***, ** and * stand for statistical significanceat the 1%, 5% and 10% level, respectively. Source: own calculations.
Table 2.
110 MARC GOERGEN AND LUC RENNEBOOG
Merger
0.72 –1.67 –1.66 38
1.78 4.86
4.20*** 2.95*** 1.19 0.93 0.77
% 2.43 1.94
Acquisition
5.22*** 6.55***
t-value
0.27 –0.45 –0.36
2.02** 2.45***
5.06*** 2.56***
t-value
[–1, 0] [–2, + 2] [–10, 0] [–40, 0] [–30, + 30] [–60, + 60] [–90, + 90]
Event Window –4.71 –7.78 –7.16 –7.14 –4.39 –3.72 –4.24
%
Hostile takeovers – Mergers
–10.89*** –11.38*** –8.69*** –5.66*** –1.84* –1.10 –1.03
t-value difference
–4.94 –5.37 –5.11 –7.37 –2.34 –0.99 –0.51
%
Hostile takeovers – Acquisitions
–10.62*** –7.31*** –2.24*** –3.14*** –0.91 –0.27 –0.11
t-value difference
–2.43 –4.28 –1.97 –1.47 –0.66 2.28 –0.61
%
Hostile takeovers – Multiple bidders
–1.62 –0.69 –1.15 32
–3.33 –2.51
% –2.51 –3.43
Hostile Takeover
Continued.
Panel D : Significance of differences in bidder CAAR across types of takeover bids
Event day % [–1, 0] 2.20 [–2, + 2] 4.35 Window prior to and at event [–10, 0] 3.83 [–40, 0] 4.63 Window around event [–30, + 30] 2.76 [–60, + 60] 3.03 [–90, + 90] 3.09 Observations 41
Time Interval
Panel C : CAARs of bidding firms by type of takeover
Table 2.
–4.59*** –5.11*** –1.50 –1.28 –0.23 0.55 –0.12
t-value difference
–0.66 –0.20 –0.27
–10.84*** –1.56
–5.61*** –4.85***
t-value
–0.23 2.41 2.05 –0.23 2.04 4.70 4.75
%
Mergers – Acquisitions
–0.96 –2.96 –0.54 17
–1.36 –1.04
% –0.08 0.85
Multiple Bidders
0.51 3.39*** 2.23*** –0.31 0.82 1.35 1.11
t-value difference
–0.27 –0.58 –0.09
–1.59 –0.59
–0.13 0.81
t-value
Value Creation in Large European Mergers and Acquisitions 111
Fig. 2b.
Cumulative Abnormal Returns of Bidding Firms by Type of Bid.
112 MARC GOERGEN AND LUC RENNEBOOG
Value Creation in Large European Mergers and Acquisitions
113
shareholders seem to disapprove of hostile takeovers. When the bid is contested, the announcement abnormal return is –2.5%. These differences are statistically significant at the 1% level as shown in panel D of Table 2. 4.3. Comparing U.K. Bids With Continental European Offers As 85% of the companies listed on the London Stock Exchange are widely held, U.K. firms are continually auctioned and hence, an active market for corporate control exists. In contrast, in Continental Europe the number of listed firms is much lower and most listed firms (around 85%–90% for Germany and France) have concentrated ownership (for a detailed overview of ownership and control in Europe, see Barca & Becht, 2001). Consequently, hostile takeovers are largely ruled out in Continental Europe as most firms have a controlling shareholder. Not surprisingly, more than half of the sample of listed target and bidding firms is from the U.K. (77 out of 129 targets and 76 out of 139 bidders) (see panel A of the table in the appendix). For the majority of the European bids, the bidder and target are from the same country (65%). As there is a high degree of disclosure in the U.K., a liquid and well-developed equity market (McCahery & Renneboog, 2002) and a higher degree of shareholder protection (La Porta et al., 1997), higher premiums in takeover offers are expected for U.K. firms. Figure 3a confirms this conjecture: the announcement effect is substantially larger for U.K. target firms (12.3%) than for Continental European firms (6%), as shown in Table 3 (panel A). There is not much difference in terms of the price run-up prior to the announcement: in both Continental Europe and in the U.K., significant positive abnormal returns are generated 2 to 3 months prior to the announcement. U.K. target shareholders who own equity as of 2 months prior to the announcement and sell on the day of the announcement can earn (on average) a premium of more than 38%, more than double the return earned by the Continental European target shareholders (15%) over the same period (panel A of Table 3). Figure 3a also depicts that, whereas the post-announcement cumulative abnormal returns are not statistically different from zero, they are substantially negative for Continental European target firms for the 1.5 to 3 months after the announcement day. Hence, in spite of the lower bid premiums in Continental Europe, it seems that the market price reactions to the announcements are overoptimistic and that returns are subsequently corrected.6 Figure 3b and panel B of Table 3 report the returns for the shareholders of the bidding firms. Bidding shareholders in the U.K. earn more than those in Continental Europe. Over a five-day window centred around the announcement date, U.K. bidders obtain a cumulative abnormal return of 1.5% versus 0.9%
Fig. 3a.
Cumulative Abnormal Returns of U.K. and Continental European Target Firms.
114 MARC GOERGEN AND LUC RENNEBOOG
Value Creation in Large European Mergers and Acquisitions
Table 3.
115
Cumulative Abnormal Returns of Target and Bidding Firms: U.K. Versus Continental Europe.
This table shows cumulative abnormal returns over several event windows for target firms and bidder firms by location (U.K. versus Continental Europe). ***, ** and * stand for statistical significance at the 1%, 5% and 10% level, respectively. Source: own calculations. Panel A: CAARs of Target Firms: U.K. vs. Continental Europe Time Interval
U.K.
t-value
Event day % [–1, 0] 12.31 29.09*** [–2, + 2] 17.42 26.03*** Window prior to and at event [–10, 0] 20.25 15.79*** [–40, 0] 38.30 14.66*** [–60, 0] 38.14 10.05*** Window around event [–30, + 30] 29.83 12.76*** [–60, + 60] 29.32 8.91*** [–90, + 90] 28.87 7.17*** Observations 48
Continental Europe
t-value
U.K. – Continental
t-value differences
% 5.95 8.85
13.99*** 13.15***
% 6.35 8.56
14.96*** 12.75***
10.05 14.95 17.98
8.89*** 7.56*** 3.61***
10.20 23.35 20.16
8.94*** 5.64*** 4.25***
17.53 14.82 14.84 52
7.46*** 4.48*** 3.67***
12.30 14.49 14.03
5.25*** 4.39*** 3.47***
U.K. – Continental
t-value on differences
Panel B: CAARs of Bidding Firms: U.K. versus Continental Europe Time Interval
U.K.
t-value
Event window % [–1, 0] 1.04 3.41*** [–2, + 2] 1.51 3.11*** Window prior to and at event [–10, 0] 0.61 0.79 [–40, 0] 1.19 0.92 [–60, 0] –0.06 –0.08 Window around event [–30, + 30] 0.19 0.11 [–60, + 60] –1.65 –0.69 [–90 + 90] –3.12 –1.07 Observations 65
Continental Europe
t-value
% 0.40 0.90
1.19 1.69*
% 0.64 0.60
1.03 0.35 2.31
0.89 0.22 1.03
–0.42 0.84 –2.37
–0.93 0.68 –1.26
0.91 0.54 0.68 74
0.49 0.21 0.21
–1.101 –2.193 –3.806
–0.62 –0.87 –1.24
1.98* 1.17
Fig. 3b.
Cumulative Abnormal Returns of U.K. versus Continental European Bidding Firms,
116 MARC GOERGEN AND LUC RENNEBOOG
Value Creation in Large European Mergers and Acquisitions
117
for Continental European bidders. Whereas there is evidence of trading on rumours in target shares or of insider trading, this is not the case for the bidding firms. 4.4. Takeover Bids By Means of Payment The average bid value of our sample is USD 5,469 million (see panel B of the table in the appendix). The distribution of the bid value is highly skewed: the median offer is USD 575 million. The majority of bids (excluding the divestitures: see Section 4.8) are cash offers (93 out of 156 cases or 60%). Twenty-four percent of the offers are all-equity offers and the remainder consists of combinations of cash and equity (11%), of cash and loan notes (2%), of equity and loan notes (2%) and of cash, equity and loan notes (1%). Payment for smaller targets usually takes place in cash: the average value of allcash offers amounts to USD 1,489 million while that of all equity-offers is USD 14,255 million (with medians of USD 443 and 2,580 million, respectively). In 12 cases out of the 93 all-cash offers, the bidders also gave the target shareholders the opportunity to accept an all-equity offer or a combined offer (with a higher value than the cash offer).7 If managers of an acquiring firm know that their shares are worth more than the current market price, they should prefer to finance the acquisition with cash. Hence, future changes in the stock price will only benefit the shareholders of the bidding firm. Conversely, if the bidding management believes that its stock is overvalued, they should prefer paying for the acquisition with equity. Hence, asymmetric information on the bidder’s share value will have some bearing on the choice between cash or equity payments. We find strong evidence that the share price reaction of target firms is sensitive to the means of payment for the target’s shares. Cash offers trigger substantially higher abnormal returns (10% at announcement) than offers including the bidders’ equity (6.7%) and combined offers of cash and equity (5.6%) (Fig. 4a). Panel A of Table 4 shows that when the price run-up starting two weeks prior to the event day is included, cash offers trigger CAARs of almost 20% versus 14% and 12.5% for all-equity bids and combined bids, respectively. Panel B shows that, whatever the event window, the share price performance of cash-financed bids outperforms the one of other bids at the 1% significance level. Figure 4b shows an entirely different picture for bidding firms. Over both short and longer term windows, the shareholders of the acquiring firms greet equity offers more favourably (1%) than cash offers (0.4%) (panel B of Table 4).8 This implies that the choice to make an all-equity offer does not signal to the market that the bidder’s equity is overvalued. Within
Fig. 4a.
Cumulative Abnormal Returns of Target Firms by Means of Payment.
118 MARC GOERGEN AND LUC RENNEBOOG
Value Creation in Large European Mergers and Acquisitions
Table 4.
119
Cumulative Abnormal Returns of Target and Bidding Firms by Means of Payment.
This table shows cumulative abnormal returns over several event windows for target firms and bidder firms by means of payment (all-cash, all-equity or a combination of cash, equity and/or loan notes). ***, ** and * stand for statistical significance at the 1%, 5% and 10% level, respectively. Source: own calculations. Panel A: CAARs of target firms by means of payment Time Interval Event day [–1, 0] [–2, + 2] [–10, 0] [–40, 0] Window around event [–30, + 30] [–60, + 60] [–90, + 90] Observations
Cash bid
t-value
Equity bid
t-value
% 9.89 13.56 19.87 27.49
35.81*** 21.95*** 21.69*** 15.54***
% 6.65 11.38 14.13 12.23
16.07*** 12.30*** 10.30*** 4.62***
28.89 28.75 28.64 62
13.39*** 9.46*** 7.71***
13.03 12.89 14.93 25
Combined bid % 5.63 13.24 12.41 16.81
4.03*** 2.83*** 2.68***
t-value
11.68*** 12.28*** 7.76*** 5.44***
17.46 5.66 0.91 13
4.64*** 1.07 0.14
Equity offers – Combined offers
t-value difference
Panel B : Significance of differences in target CAARs across types of payment Cash offers – Equity offers Event Window [–1, 0] [–2, + 2] [–10, 0] [–40, 0] [–30, + 30] [–60, + 60] [–90, + 90]
% 3.24 2.18 5.74 15.26 15.87 15.86 13.72
t-value difference
36.10*** 10.84*** 19.25*** 26.52*** 22.61*** 16.04*** 11.34***
Cash offers – Combined offers % 4.26 0.32 7.46 10.68 11.44 23.09 27.73
t-value difference
30.81*** 1.02 16.27*** 12.07*** 10.59*** 15.19*** 14.91***
% 1.01 –1.86 1.72 –4.58 –4.43 7.23 14.02
0.20 –0.24 0.19 –0.35 –0.31 0.43 0.75
Panel C: CAARs of bidding firms by means of payment Time Interval
Cash bid
Event day % [–1, 0] 0.37 [–2, + 2] 0.90 Window prior to and at event [–10, 0] –0.16 [–40, 0] –1.18 Window around event [–30, + 30] –0.33 [–60, + 60] –1.44 [–90, + 90] –1.52 Observations 83
t-value
Equity bid
t-value
Combined bid
t-value
1.68* 1.83*
% 0.98 2.57
3.01*** 3.52***
% 0.13 0.22
0.35 0.27
–0.20 –0.84
2.60 5.15
2.40** 2.46**
–0.71 –0.20
–0.58 –0.09
–0.19 –0.59 –0.51
3.50 2.72 2.13 33
1.37 0.76 0.48
–1.42 –1.39 –4.21 23
–0.49 –0.34 –0.85
120
MARC GOERGEN AND LUC RENNEBOOG
Table 4. Continued. Panel D : Significance of differences in bidder CAAR across types of payment
Event Window [–1, 0] [–2, + 2] [–10, 0] [–40, 0] [–30, + 30] [–60, + 60] [–90, + 90]
Cash offers– Equity offers
t-value difference
% –0.61 –1.67 –2.76 –6.33 –3.83 –4.16 –3.65
–9.95*** –12.08*** –13.36*** –16.01*** –7.94*** –6.11*** –4.39***
Cash offers – Combined offers % 0.24 0.68 0.55 –0.98 1.09 –0.05 2.70
t-value difference
2.97*** 3.79*** 2.05** –1.89* 1.74* –0.05 2.49**
Equity offers – Combined offers % 0.85 2.35 3.31 5.36 4.93 4.11 6.34
t-value difference
8.93*** 11.00*** 10.43*** 8.75*** 6.59*** 3.91*** 4.93***
the sample of large take-over bids, the relatively smaller ones are all-cash bids whereas the relatively larger ones involve equity. Consequently, it may be that the market realises that for large deals the choice of means of payment is restricted. 4.5. Takeover Bids By Industry In this sub-section, we analyse whether our results are driven by particular industries. We created the following 5 industry groups based on the SIC classification: (i) (ii) (iii) (iv) (v)
energy, natural resources, waste development and utilities (7 firms); production and manufacturing (45 firms); services (21 firms); retailers, stores, pubs, hotels (11 firms); and banking and insurance (16 firms).
Panel A of Table 5 shows that, on the announcement day, bids targeting retail and manufacturing firms trigger the strongest positive price reactions, 14.4% and 10.9%, respectively. For longer time intervals of e.g. two months there are no substantial differences between the industries. Our results for banks are consistent with the findings of Cybo-Ottone and Murgia (2000), who found a significant and positive 15.3% announcement effect for European target banks. The strong decline in the prices of financial and energy target firms in the postannouncement period reflects the fact that a few of the bids were ultimately unsuccessful (Fig. 5a). However, the picture for bidding firms by industry looks different (Fig. 5b). Some industries show positive CAARs (manufacturing, retailing) whereas
Fig. 4b. Cumulative Abnormal Returns of Bidding Firms by Means of Payment.
Value Creation in Large European Mergers and Acquisitions 121
Cumulative Abnormal Returns of Target and Bidding Firms by Industry.
Energy
2.83** 2.78** 2.01*
4.57*** 3.90***
t-value
24.14 24.86 26.82 45
% 10.87 15.16
Manufac·
10.36*** 7.58*** 6.69***
25.77*** 22.73***
t-value
26.90 27.10 26.88 21
% 7.34 10.50
Services
Energy
Event day % [–1, 0] –1.91 [–2, + 2] –0.83 [–5, + 5] –1.07 Window around event [–30, + 30] 0.70 [–60, + 60] 7.64 [–90, + 90] 12.34 Observations 7
Time Interval
0.13 1.02 1.35
–1.98* –0.54 –0.47
t-value
–0.001 –1.66 –2.47 63
% 1.89 2.92 3.24
Manufac·
–0.00 –0.56 –0.69
5.00*** 4.88*** 3.65***
t-value
0.47 2.90 0.67 26
% –2.35 –2.19 –3.73
Services
Panel B : Cumulative Abnormal Returns of Bidding Firms by Industry
Event day % [–1, 0] 5.06 [–2, + 2] 6.83 Window around event [–30, + 30] 17.31 [–60, + 60] 21.30 [–90, + 90] 21.11 Observations 7
Time Interval
Panel A : Cumulative Abnormal Returns of Target Firms by Industry
0.16 0.70 0.13
–4.43*** –2.61** –3.00***
t-value
6.96*** 4.98*** 4.04***
10.48*** 9.48***
t-value
4.13 5.37 5.92 20
% 2.07 2.19 2.66
Retailers
21.26 18.22 17.88 11
% 14.35 16.87
Retailer
1.43 1.32 1.19
3.94*** 2.64** 2.16**
t-value
4.83*** 2.94** 2.36**
17.99*** 13.38***
t-value
–2.03 –8.95 –9.83 22
% 0.44 –0.15 –1.28
Bank
21.12 8.83 2.77 16
% 4.03 10.06
Bank
–0.63 –1.96* –1.76*
0.75 –0.16 –0.93
t-value
5.19*** 1.54 0.40
5.48*** 8.63***
t-value
This table shows cumulative abnormal returns over several event windows for target firms and bidder firms by industry. ***, ** and * stand for statistical significance at the 1%, 5% and 10% level, respectively. Source: own calculations.
Table 5.
122 MARC GOERGEN AND LUC RENNEBOOG
Fig. 5a.
Cumulative Abnormal Returns of Target Firms by Industry.
Value Creation in Large European Mergers and Acquisitions 123
Fig. 5b.
Cumulative Abnormal Returns of Bidding Firms by Industry.
124 MARC GOERGEN AND LUC RENNEBOOG
Value Creation in Large European Mergers and Acquisitions
125
other industries have negative announcement effects (energy, services). The latter difference is largely due to the fact that the energy and services industries count more hostile takeovers (see Section 4.2). Financial firms (banks, insurance) realise positive (but not significant) returns and negative (significant) CAARs over longer time periods (Table 5). These findings are consistent with those from Cybo-Ottone and Murgia (2000) for Europe and Frame and Lastrapes (1990) for the U.S. 4.6. Failed Versus Successful Bids In this sub-section, we address the question as to whether the markets are able to anticipate the ultimate success or failure of the merger negotiations. The merger or acquisition negotiations are assumed to be ultimately successful if the Financial Times reports acceptance of the bid by the target’s shareholders. Conversely, the takeover attempt is considered to be a failure if the bidder abandons negotiations within a 6-month period subsequent to the announcement. Out of the 185 announcements 37 failed. When focussing on the 100 announcements involving listed target firms (excluding divestitures), 27 were unsuccessful. Out of these 27 target firms, only 16 were categorised as hostile bids due to resistance by the target’s management or due to the fact that multiple firms were attempting to acquire the target. Hence, there were also 11 cases of unsuccessful mergers and (friendly) acquisitions. For both the (ultimately) failed and successful bids, we find a significant positive announcement effect for the target firms (Fig. 6a). The event day effect is significantly larger (by 5%) for the successful bids than for the failures. However, for the two-week window prior to and including the event day, there is no difference in the CAARs between failed and successful bids (panel A of Table 6). When the price run-up over a 3-month period is included, the failed bids significantly outperform the successful bids by 30% versus 23%. Whereas the abnormal returns for the targets in the successful bids are not significantly different from zero subsequent to the announcement of the bid, the abnormal returns for the failed bids nose-dive (by 7.5%) between 2 and 3 months subsequent to the event. This is a result of the collapse of the (friendly) merger negotiations or of the successful hostile opposition by the target’s management. However, it should be noticed that the cumulative abnormal return for the companies on which a failed bid was launched does not revert to the level prior to the announcement. This implies that the stock prices of these targets still contain a merger premium reflecting the possibility of another potential takeover bid in the (near) future.
Fig. 6a.
Cumulative Abnormal Returns of Target Firms by (Ex Post) Successful versus Failed Bids.
126 MARC GOERGEN AND LUC RENNEBOOG
Value Creation in Large European Mergers and Acquisitions
Table 6.
127
Cumulative Abnormal Returns of Target and Bidding Firms by (Ex Post) Successful and Failed Bids.
This table shows cumulative abnormal returns over several event windows for target firms and bidder firms by outcome of the negotiations (failure versus success). ***, ** and * stand for statistical significance at the 1%, 5% and 10% level, respectively. Source: own calculations. Panel A : CAARs of Target Firms by Ex Post Success or Failure of the Bid Time Interval
Failures
t-value
Event date % [–1, 0] 5.51 8.49*** [–2, + 2] 10.83 10.55*** Window prior to and at event [–10, 0] 16.42 9.59*** [–20, 0] 24.61 7.23*** [–60, 0] 29.87 4.31*** Window subsequent to event [ + 1, + 40] –1.89 –0.45 [ + 1, + 60] –7.53 –14.78*** Window around event [–30, + 30] 28.69 8.00*** [–60, + 60] 25.02 4.95*** [–90, + 90] 27.03 4.38*** Observations 27
Successes
t-value
Failed offers – Successful offers
t-value difference
% 10.30 13.75
27.84*** 23.51***
% –4.79 –2.92
–10.36*** –3.99***
16.30 19.74 22.41
16.74*** 14.58*** 7.31***
0.12 4.87 7.46
0.24 4.69*** 4.43***
–0.10 –2.20
–0.14 –1.54
–1.79 –5.33
–0.40 –4.53***
21.49 20.58 19.58 73
10.51*** 7.15*** 5.56***
7.20 4.44 7.46
2.82*** 1.23 1.69*
Panel B : CAARs of Bidding Firms by Ex Post Success or Failure of the Bid Time Interval
Failures
t-value
Event date % [–1, 0] –0.73 –1.45 [–2, + 2] –0.97 –1.22 Window prior to and at event [–10, 0] –0.12 –0.66 [–20, 0] –0.18 –0.68 Window subsequent to event [ + 1, + 40] –0.09 –0.44 [ + 1, + 60] –0.92 –1.24 Window around event [–30, + 30] 0.55 0.20 [–60, + 60] 1.96 0.50 [–90, + 90] –1.61 –0.34 Observations 29
Successes
t-value
Failed offers – Successful offers
t-value difference
% 1.08 1.75
4.05*** 4.16***
% –1.81 –2.72
–5.48*** –5.21***
1.00 0.91
2.21*** 1.70*
–1.12 –1.09
–2.68*** –1.97**
–1.21 2.30
–1.39 1.68**
1.02 –3.22
1.50 –2.42***
0.36 –1.13 –0.96 109
0.24 –0.54 –0.38
0.19 3.09 –0.66
0.11 1.20 –0.21
128
MARC GOERGEN AND LUC RENNEBOOG
The announcement effect for unsuccessful bidding firms is negative, but not statistically significant from zero (Table 6; Fig. 6b). This negative effect can be explained by the fact that two thirds of the subsample of failed bids consist of hostile takeovers (see Section 4.2). 4.7. Bids Made Prior and Subsequent to January 1st 1999 M&A activity during the 1990s is characterised by continuous increases in volume, in average bid value and hence in total bid value. European M&A activity grew by more than 280% over the period of 1996–1999. The year 1999 was not only remarkable in terms of the total bid value, but also in terms of the number of hostile takeovers: there were a staggering 369 hostile offers. Shelton (2000) reports evidence that bidder gains fall during merger peaks, suggesting that bidders are more aggressive, display greater tendencies to over-pay for target firms or assume more risk in pursuing takeover projects. Hence, we split the bids into two categories based on the year in which they were made: bids before 1999 and those in 1999–2000. About half of the bids were made in 1999–2000.9 Part of the difference in wealth effects between the two periods may be the result of the introduction of the Euro. However, panel A of Table 7 shows that, on the announcement day, there is little difference in terms of the price reaction for bids that took place prior to 1999 and those in 1999/2000: for both samples the abnormal return is around 9%. Still, Fig. 7a suggests that the price run-up for pre-1999 offers only started one month prior to the announcement whereas that of the bids in 1999–2000 was generated for a period starting as long as 3 months prior to the announcement. Over long windows, for instance over a 6-month symmetrical event window, the recent bids yield higher CAARs (almost 25%) than the pre-1999 ones (19%). We also investigate the difference in announcement reactions for firms bidding prior to and after 1 January 1999, but do not find any difference in abnormal returns (panel B of Table 7). 4.8. Divesting Firms The 185 announcements include 29 bids for a divestiture of a division from a listed firm. In all 29 cases the initial bid is a friendly attempt, but one case involved multiple bidders and was therefore considered to be a hostile bid. In half of the cases, bidder and target are located in the same European country and in more than one third of the cases bidder and target are U.K.-based. Ninety percent of the bids are cash financed; two offers were financed by both cash and loan notes and in one case a combination of equity and loan notes was
Fig. 6b.
Cumulative Abnormal Returns of bidding firms by Ex Post Successful and Failed Bids.
Value Creation in Large European Mergers and Acquisitions 129
130
MARC GOERGEN AND LUC RENNEBOOG
offered. The market value of the divestitures is smaller than that of the average bid for an independent firm and averages USD 480 million (median of USD 415) (see the table in the appendix). Table 8 and Fig. 8 show that the
Table 7. Cumulative Abnormal Returns by Year of Bid. This table shows cumulative abnormal returns over several event windows for target firms and bidder firms by year of bid (prior to 1999 and 1999/2000). ***, ** and * stand for statistical significance at the 1%, 5% and 10% level, respectively. Source: own calculations. Panel A : CAARs of Target Firms by Year of Bid Time Interval
pre-1999 offers
Event day % [–1, 0] 08.80 [–2, + 2] 14.02 Window prior to and at event [–10, 0] 19.82 [–40, 0] 21.78 Window around event [–30, + 30] 22.71 [–60, + 60] 18.47 [–90, + 90] 18.98 Observations 48
t-value
1999/2000 offers
t-value
1999/2000 offers – pre1999 offers
t-value differences
23.30*** 23.48***
% 09.20 11.99
18.90*** 15.58***
% 0.40 –2.03
0.92 –2.93***
17.02*** 8.52***
17.56 24.15
11.13*** 9.34***
–2.26 2.37
–1.99** 1.31
10.89*** 6.29*** 5.28***
24.10 24.83 24.00 52
8.96*** 6.56*** 5.18***
1.39 6.36 5.02
0.57 1.87* 1.21
t-value
1999–2000 offers – pre1999 offers
t-value differences
Panel B : CAARs of Bidding Firms by Year of Bid Time Interval
pre-1999 offers
Event day % [–1, 0] 0.55 [–2, + 2] 1.22 Window prior to and at event [–10, 0] 0.16 [–40, 0] 1.72 Window around event [–30, + 30] 0.39 [–60, + 60] –0.10 [–90, + 90] –0.47 Observations 74
t-value
1999–2000 offers
2.13** 2.98***
% 0.87 1.14
2.17** 1.80*
% 0.32 –0.08
0.96 –0.16
0.36 1.59
0.03 0.07
0.11 0.10
–0.13 –1.65
0.29 1.60
0.27 –0.05 –0.19
0.41 –0.91 –1.81 65
0.18 –0.29 –0.47
0.02 –0.81 –1.33
0.01 –0.31 –0.42
Fig. 7a.
Cumulative Abnormal Returns of Targets Firms: Bids Made Pre- and Post-January 1st 1999.
Value Creation in Large European Mergers and Acquisitions 131
Fig. 7b.
Cumulative Abnormal Returns of Bidding Firms: Bids Made Pre- and Post-January 1st 1999.
132 MARC GOERGEN AND LUC RENNEBOOG
Value Creation in Large European Mergers and Acquisitions
Table 8.
133
Cumulative Abnormal Returns of Divesting Firms.
This table shows cumulative abnormal returns over several event windows for divesting firms. ***, ** and * stand for statistical significance at the 1%, 5% and 10% level, respectively. Source: own calculations. Time Interval
CAAR
Event day [–1, 0] [–2, + 2] Window around event [–30, + 30] [–60, + 60] [–120, + 120] Observations
% 0.31 3.46 0.81 1.03 –10.06 29
t-value
0.51 3.60*** 0.24 0.22 –1.50
announcement of the divestitures is greeted by the market as positive news for the divesting firm as the CAAR of a 5-day window around the event day is 3.5% (significant at the 1% level). The reasons that are reported for the divesture in the announcement press statements include: return to core business (69%) and the generation of cash to pursue a focus strategy (21%). Hanson and Song (2000) find that buyers and sellers on average receive significant positive percentage returns, although significant returns only occur in the 1990–1995 period. Other recent U.S. divestiture research by Mulherin and Boone (2000) for the period 1990–1999 concludes that the combined target and bidder return at the announcement averages 3.5%, while the announcement return for corporate divestitures averages 3%. The results are consistent with the views that divestitures create value by moving assets to the buyer’s more efficient operating environment and that divestitures resolve agency problems.
5. TAKEOVER MOTIVES: SYNERGIES, AGENCY OR HUBRIS? Although most bidding firms make statements about the potential synergies from mergers and acquisitions, frequently the forecasted benefits are not obtained. This may be the result of poor synergy forecasts by the bidding management or the fact that the takeover was initiated for entirely different reasons such as managerial hubris or other agency problems. We will attempt to distinguish between these three different takeover motives by performing a correlation analysis of the target, bidder and total announcement gains.
Fig. 8.
Cumulative Abnormal Returns of Divesting Firms.
134 MARC GOERGEN AND LUC RENNEBOOG
Value Creation in Large European Mergers and Acquisitions
135
If synergies are the main motive for the merger, we assume that the management of both the target and acquirer intend to maximise shareholder value. Hence, the wealth effects of the takeover for both target and acquirer shareholders should be positive and the division of the value created should depend on the relative bargaining power of target and bidder. In addition, the wealth gains for the target shareholders should be positively correlated to those of the bidder shareholders and the total wealth effect. A second motive for a takeover may be agency related: in this case, the selfinterest of the bidder’s management is the prime reason for the offer. Managers may prefer to stimulate corporate growth rather than corporate value as their private benefits tend to be more substantial in the larger firms. For example, Conyon and Murphy (2002) show that for the U.K., corporate size (and not corporate performance) is the main determinant of the level of managerial salaries, bonuses as well as of the allotment of stock options. Hence, managers may be tempted to use free cash flow for ‘empire building’ (Jensen, 1986). Similarly, Shleifer and Vishny (1989) suggest that managers may make acquisitions such that the combined entity will depend even more on their personal expertise. Hence, they may exploit this dependency and extract value from the acquirer: both the total value of the combined entity as well as the wealth of the bidder’s shareholders will be lower. As a result, the correlations between the target’s value and the bidder’s value and between the target’s value and the total value will be negative. A third takeover motive may be the bidding management’s hubris, which hinges on the assumption that the management makes mistakes in evaluating potential targets (Roll, 1986). If there is an equal probability that managers are over- and underestimating the synergies of potential takeovers, and managers make a bid after having overestimated synergy values, they may mostly pay too much for the target. As a result, the higher the target’s gain, the lower the bidder’s gain, such that there is a wealth transfer from the bidder to the target with the total gain being zero (Berkovitch & Narayanan, 1993). Hence, the correlation between the target’s and bidder’s wealth changes is negative whereas the one between the target’s and total wealth change is zero. Panel A of Table 9 summarizes the expected signs of the correlations. In order to test these hypotheses, we select the 68 takeovers for which both target and bidder are listed. The average total gain is calculated on the event day (panel B of Table 9) and over the time window [–10, + 1] (panel C), using the market capitalizations of bidder and target as weights. Total wealth gains over these periods amount to 4% and 6%, respectively, of the combined entity. Fifty-eight percent of takeovers in this sample have positive total wealth gains. For the whole sample, we find that the correlations between the target and total
136
MARC GOERGEN AND LUC RENNEBOOG
Table 9.
Correlations Between Target, Bidder and Total Wealth Gain.
This table tests whether or not takeover bids were made for reasons of synergy, agency or hubris. Source: own calculations. Panel A : Expected sign of correlation
Synergy Agency Hubris
Expected sign Correlation Target and Total Gain
Expected sign Correlation Target and Bidder Gain
Positive Negative Zero
Positive Negative Negative
Panel B : Correlations between target, bidder and total event day gain Correlation Target and Total Gain Total sample (64 observations) Positive total gain sub-sample (42) Negative total gain sub-sample (22)
0.4545** 0.2474* 0.2359
Correlation Target and Bidder Gain 0.0617 –0.1990 –0.1267
Panel C: Correlations between target, bidder and total gain over period [–10, 0]
Total sample (64) Positive total gain sub-sample (42) Negative total gain sub-sample (22)
Correlation Target and Total Gain
Correlation Target and Bidder Gain
0.6330*** 0.5541** 0.1393
0.4155** 0.1763* –0.1640*
gain in panel B are significantly positive. When we measure the wealth gain over a longer event window, the correlation between the target and the bidder gains is also significantly positive. This suggests that, on average, the large European takeover bids in the 1990s were motivated by synergies. However, as the motives for individual firms may still be different, we also analyse the correlations for the subsamples of takeovers with positive vs. negative total wealth effects. We find that the synergy hypothesis for the firms with positive wealth effects is corroborated (see panel C). In contrast, for the companies with negative gains, we find no correlation between target and total gain and a
Value Creation in Large European Mergers and Acquisitions
137
negative correlation – only statistically significant for longer term wealth effects – between target and bidder gains. This suggests that in about a third of the firms, managerial hubris may, to a large extent, be responsible for poor decision making about takeover bids. The findings from this study are in line with those of Gupta et al. (1997) and Berkovitch and Narayanan (1993) who both find strong evidence that synergy is the prime motive for takeovers. The latter study also finds evidence that agency problems and hubris are a motive for acquisitions.
6. DETERMINANTS OF SHORT-RUN WEALTH EFFECTS FOR TARGET AND BIDDING FIRMS We regress the cumulative abnormal returns of target and bidding firms (in separate regressions) over several windows ([–1, 0] and [–10, + 1]) on variables capturing: (i) the type of takeover (merger, acquisition, hostile takeover), (ii) the means of payment (all-cash offer, all-equity offer or a combination of cash, equity or loan notes), (iii) the takeover characteristics (relative size: target/bidder), (iv) the target and bidder characteristics (net cash held by target over market value of equity, performance of target, interest coverage of target, growth potential of target (MV/BV), degree of diversification of bidder, industry of target and bidder, (v) corporate governance characteristics (U.K. target, U.K. bidder). We correct both target and bidder regressions for industry effects. The sample size is 100 for the target firms and 139 for the bidder firms. Table 10 shows that the type of takeover bid is an important determinant of the short-term wealth effects (on the event day and for the period including the price-run up) for both target and bidder firms. In comparison to merger offers, hostile bids trigger large positive abnormal returns for the target shareholders but significant, negative returns for the bidder. This follows from the fact that bidder shareholders are fearful that the management’s motives for the bid are agency problems or hubris. In contrast, target shareholders expect that opposition against the offer will lead to upward bid revisions. When the offer is cash financed, the target’s share price will increase more than when the bid consists of an all-equity offer or a combination of equity, cash and loan notes. An all-cash offer signals that the bidder’s equity is undervalued. It may also signal the bidder’s confidence in the value of the synergies as the bidder does not want to share future value creation with the target shareholders. However,
CAAR [–1, 0] Coeff· p-value 2.269** 2.141** –1.859* 2.200** 0.472 –1.289 1.724* 1.656* –1.003 0.389 –1.641 1.921* 1.846* –0.467 –0.4096 –0.3616 –0.7758
0.0643 0.0253 –0.0152 0.0964 0.0035 –0.0743 0.0006 0.0369 –0.0001 0.0497 –0.1932 0.0682 0.0189 –0.0564 –0.0463 –0.0717 –0.0337 100 0.295 0.141 0.013
Dep variable
Intercept Bid characteristics hostile acquisition friendly acquisition cash payment Bidder and target characteristics Relative size (target/bidder) Target cash reserves/ Market cap. Target market-to-book ratio Target ROE Interest coverage Bidder diversification Bidder + target : same industry U.K. target U.K. bidder Industries Energy Services Retail Financial Observations R2 Adjusted R2 Signif. of F-value –0.0724 –0.0636 –0.0738 –0.0443 100 0.352 0.215 0.006
0.0004 0.0341 –0.0004 0.0031 –0.1425 0.0561 0.0325
0.0016 0.0042
0.0849 0.0241 0.0742
0.0848
–0.897 –0.402 –0.566 –1.313
1.683* 1.528 –0.639 0.142 –1.425 1.782* 1.699*
0.529 0.211
2.363** 1.999** 2.590***
2.673***
CAAR [–10, 0] Coeff· p-value
Target firms
–0.0232 –0.0217 0.0495 0.1052 139 0.303 0.223 0.001
–0.0004 0.0082 0.0004 –0.0081 0.218 0.0463 0.0411
0.0007 0.0003
–0.0502 0.0264 –0.0342
0.0553
–0.565 –0.688 1.585 0.754
–2.421*** 0.979 0.137 –1.698* 1.555 2.014** 1.911*
0.299 0.846
–2.521*** 0.449 –1.856*
2.152**
CAAR [–1, 0] Coeff· p-value
–0.0633 –0.0452 0.0315 0.0212 139 0.341 0.246 0.001
–0.0002 0.0073 0.0007 –0.0091 0.315 0.0382 0.0445
0.0001 0.0004
–0.062 0.0074 –0.014
0.824
–0.744 –0.148 1.041 0.251
1.684* 0.265 0.463 –1.735* 0.572 1.856* 1.874*
0.189 0.677
–2.522*** 0.537 –1.251
2.428***
CAAR [–10, 0] Coeff· p-value
Bidder firms
This table shows OLS regressions of cumulative abnormal returns over different event windows for target and bidder firms.
Table 10. Determinants of Short Term Wealth Effects for Target and Bidding Firms.
138 MARC GOERGEN AND LUC RENNEBOOG
Value Creation in Large European Mergers and Acquisitions
139
for the very large takeover targets it may be difficult to raise large amounts of cash such that the bidder has to resort to an equity or combined offer. The share price reaction for bidding firms to a cash offer is (weakly) negative. This may result from the concern that management may bid too high a premium whereas when the target shareholders accept an equity offer, they share some of the risk from the acquisition. The impact of the following target and bidder firm characteristics is also investigated: the relative size of the target compared to the bidder, the cash reserves held by the target firm, the target’s market-to-book ratio, the target’s return on equity and interest coverage, the degree of the bidder’s diversification, the fact whether or not bidder and target are in the same industry and the country the bidder and target are located in. Table 10 shows that the relative size is not significant, which may be explained by the fact that this study only concentrates on large European bids. The amount of cash reserves held by the target company may have an impact on the bid and hence on the share price reaction at the announcement, because a target firm with substantial cash reserves may finance its own takeover. Table 10 shows that this is not the case in our sample. For a target firm with growth opportunities (as reflected in a high market-to-book ratio), the market expects a premium whereas Table 10 suggests that the market is anxious that the bidder will overpay for growth options. Neither the target’s performance (return on equity) nor the financial distress measure (interest coverage) has any bearing on the abnormal returns of the sample targets and bidders. In addition, the fact that the takeover is within the same industry does not influence the announcement effect of a takeover. In contrast, we find some weak evidence (at the 10% level) that the bidder’s share price reacts negatively to a takeover bid when the bidder is already diversified. The regressions also analyse whether the location of the bidder and target – U.K. versus Continental Europe – influences the announcement effect. We find strong evidence of significantly positive abnormal returns when the target firm and the bidding firm are U.K. based. This is because the U.K. has an active market for corporate control which is largely inexistent in Continental Europe.
7. CONCLUSIONS The 1990s were characterised by a large increase in European M&A activity. In 1999, the total deal volume, the average deal value and the number of hostile takeovers almost reached U.S. levels. In this study we analyse the market reactions to takeover bids in large M&A deals (185) with a value of at least USD 100 million. Our sample comprises 55 mergers, 40 friendly acquisition,
140
MARC GOERGEN AND LUC RENNEBOOG
40 hostile takeovers, 21 hostile takeovers involving multiple bidders and 29 divestitures. The short-term wealth effects are remarkably similar to those of U.S. and U.K. studies. We find large announcement effects of 9% for target firms, but the cumulative abnormal return that includes the price run-up over the two-week period prior to the event rises to 20%. The share price of the bidding firms reacts positively with a statistically significant announcement effect of only 0.7%. We also show that the type of takeover bid has a large impact on the short term-wealth effects of target and bidder shareholders. For hostile takeovers, the announcement effect for target firms is substantially higher (12.6% on day 0 and almost 30% including the price run-up) than the one for mergers and friendly acquisitions (8% on day 0 and 22% including the price run-up). Hence, the market seems to expect that opposition against a bid will lead to a revision of the offer and ultimately to a higher bid premium. This is confirmed by the share price reaction of bidding firms: a hostile takeover triggers a negative price reaction of 2.5% whereas the announcement of a merger or friendly acquisition generates a positive announcement abnormal return of 2.5%. The location of bidder and target firms also seems to have an important impact on short-term wealth effects: both U.K. bidders and targets generate significantly higher returns than their Continental European counterparts. This can be partially explained by the higher incidence of hostile takeovers in the U.K. and in the more developed U.K. takeover market. We also find strong evidence that the means of payment in an offer has a large impact on the share price reactions. All-cash offers trigger an abnormal return of almost 10% upon announcement (27.5% including price run-up) whereas all-equity bids or offers combining cash, equity and loan notes generate a return of 6% only (14% including the price run-up). Cash bids are more frequent for smaller targets, though. Surprisingly, the market reacts more positively ( + 1%) to bidding firms which resort to equity to pay for the takeover. This implies that the choice of means of payment does not act as a signal to the market of the over- or undervaluation of the bidder’s equity. Contrary to past research, the size of the target firm relative to the size of the bidder does not have an impact on target and bidder wealth effects. The reason for this may be that this study focuses on large takeover deals and that therefore the average relative size is high. There is no evidence that the past returns of target and bidder firms influence the share price reactions around the bid announcement. However, the market-to-book ratio of the target matters in terms of the bid premium. A high market-to book ratio for the target leads to a higher bid premium combined with a negative price reaction for the bidder.
Value Creation in Large European Mergers and Acquisitions
141
Finally, we find that bidding firms should not further diversify by acquiring target firms that do not match the bidder’s core business. We also investigate whether the predominant reason for takeovers is synergies, agency problems or managerial hubris. We find a significant positive correlation between target shareholder gains and total gains for the merged entity as well as between target gains and bidder gains. This suggests that synergies are the prime motivation for bids and that targets and bidders tend to share the resulting wealth gains. However, these findings are only valid for the companies generating positive wealth gains. In companies with negative total wealth gains, there is no significant correlation between target and total wealth gains whereas the correlation between target and bidder gains is significant and negative. This implies that – given that the total wealth effect is not positive – a dollar gain to the target’s shareholders coincides with a dollar loss for the bidder’s shareholders. Thus, it seems that for a third of firms, managerial hubris leads to poor decision making on takeovers.
NOTES 1. As reported in an M&A report by Morgan Stanley using Thomson Financial Securities Data, April 2001. 2. For an excellent overview of post-merger performance and of the motives for mergers and tender offers: see Agrawal and Jaffe (2000). 3. Still, the empirical evidence investigating the creation of an internal capital market shows that diversified firms do not rely significantly less on the outside capital market than non-diversified firms (Comment & Jarrell, 1995). 4. The systematic risk of all 6 estimation techniques is calculated using the all-share index for each country. For example, the betas of U.K. targets and bidders are calculated using the FT-All Share Index. 5. After the first announcement of a bid, it still takes several months before the merger or acquisition is accepted and the target firm stops trading. In only 11 out of 129 cases, the target firm is no longer traded within 40 trading days subsequent to the announcement. Respectively, 24 and 36 target firms are delisted 60 and 100 trading days subsequent to the announcement. We reduce the event window of target firms to 80 days after the event day. 6. The post-announcement correction in abnormal returns is not explained by the fact that the Continental European sample consists of more bids which fail ex post. Quite the contrary is true: there are more failed bids (related to hostile takeover attempts) in the U.K. 7. This choice between an all-cash offer or a combined offer consisting of cash and equity is given in all 12 cases at the first announcement of the bid. In contrast, in 4 cases, a cash offer was added to an initial all-equity or combined offer as a sweetener some time after the first announcement. 8. Upon the editors’ suggestion, we investigated whether the means of payment is still significant after correcting for the payment of a premium. We find that cash offers
142
MARC GOERGEN AND LUC RENNEBOOG
still have a (modestly significantly) negative impact (at the 10% level) on the share price reaction of the bidder. 9. An analysis of the takeovers by year for the period 1993–1998 does not give significant differences in abnormal returns at the announcement and over longer time windows.
ACKNOWLEDGMENTS We are grateful to Marco Becht, Rafel Crespi, Julian Franks, Carles Gispert, Alan Gregory, Rez Kabir, Colin Mayer, and Joe McCahery for stimulating comments.
REFERENCES Agrawal, A., & Mandelker, G. N. (1987). Managerial incentives and corporate investment and financing decisions. Journal of Finance, 42, 823–837. Agrawal, A., & Jaffe, J. (2000). The post-merger performance puzzle. In: A. Gregory (Ed.), Advances in Mergers and Acquisitions (Vol. 1, pp. 7–41). Elsevier Science. Barca, F., & Becht, M. (2001). The control of corporate Europe. Oxford University Press. Berger, P. G., & Ofek, E. (1995). Diversification’s effect on firm value. Journal of Financial Economics, 37, 39–65. Berkovitch, E., & Narayanan, M. P. (1993). Motives for takeovers: An empirical investigation. Journal of Financial and Quantitative Analysis, 28, 347–362. Bhide, A. (1989). The causes and consequences of hostile takeovers. Journal of Applied Corporate Finance, 2, 36–59. Blume, M. (1971). On the assessment of risk. Journal of Finance, 26, 1–10. Comment, R., & Jarrell, G. A. (1995). Corporate focus and stock returns. Journal of Financial Economics, 37, 67–87. Conn, C., Cosh, A., Guest, P., & Hughes, A. (2002). Long-run share price performance of U.K. firms engaging in cross-border acquisitions. Working paper Centre for Business Research, WP214 University of Cambridge. Conyon, M., & Murphy, K. (2002). Stock based executive compensation. In: J. McCahery, P. Moerland, T. Raaijmakers & L. Renneboog (Eds), Corporate Governance: Convergence and Diversity. Oxford University Press. Cybo-Ottone, A., & Murgia, M. (2000). Mergers and shareholder wealth in European banking. Journal of Banking & Finance, 24, 831–859. Dimson, E. (1979). Risk measurement when shares are subject to infrequent trading. Journal of Financial Economics, 7, 197–226. Dimson, E., & Marsh, P. R. (1983). The stability of U.K. risk measures and the problem of thin trading. Journal of Finance, 38, 753–783. Eckbo, E., & Thorburn, K. (2000). Gains to bidder firms revisited: domestic and foreign acquisitions in Canada. Journal of Financial and Quantitative Analysis, 35, 1–25. Elton, E., & Gruber, M. (1995). Modern portfolio theory and investment analysis (5th ed., p. 715). New York: John Wiley & Sons. Frame, W. S., & Lastrapes, W. D. (1998). Abnormal returns in the acquisition market: the case of bank holding companies, 1990–1993. Journal of Financial Services Research, 14, 145–163.
Value Creation in Large European Mergers and Acquisitions
143
Franks, J., Harris, R., & Mayer, C. (1988). Means of payment in takeover: results for the United Kingdom and the United States. In: A. Auerbach (Ed.), Corporate Takeovers: Causes and Consequences (National Bureau of Economic Research. Distributed by University of Chicago Press). Franks, J., & Harris, R. (1989). Shareholder wealth effects of corporate takeovers. Journal of Financial Economics, 23, 225–249. Franks, J., Mayer, C., & Renneboog, L. (2001). Who disciplines management of poorly performing companies? Journal of Financial Intermediation, 10, 209–248. Gregory, A. (1997). An examination of the long run performance of U.K. acquiring firms. Journal of Business Finance and Accounting, 24, 971–1002. Gregory, A., & McCorriston, S. (2002). Foreign acquisitions by U.K. limited companies: long-run performance in the U.S., Continental Europe and the rest of the world. Working paper Financial Markets Research Centre, University of Exeter. Gupta, A., LeCompte, R. L. B., & Misra, L. (1997). Acquisitions of solvent thrifts: Wealth effects and managerial motivations. Journal of Banking and Finance, 21, 1431–1450. Hanson, R. C., & Song, M. H. (2000). Managerial ownership, board structure, and the division of gains in divestitures. Journal of Corporate Finance, 6, 55–70. Healy, P., Palepu, K., & Ruback, R. (1992). Does corporate performance improve after mergers? Journal of Financial Economics, 31, 135–175. Healy, P., Palepu, K., & Ruback, R. (1997). Which takeovers are profitable: strategic or financial? Sloan Management Review, 38, 45–57. Higson, C., & Elliott, J. (1998). Post-takeover returns: The U.K. evidence. Journal of Empirical Finance, 5, 27–46. Huang, Y., & Walking, R. (1989). Target abnormal returns associated with acquisition announcements: payment, acquisition form and managerial resistance. Journal of Financial Economics, 19, 329–350. Jarrell, G., & Poulsen, A. (1989). The returns to acquiring firms in tender offers: evidence from three decades. Financial Management, 18, 12–19. Jensen, M. C. (1986). Agency costs of free cash flow, corporate finance and takeovers. American Economic Review, 76, 323–329. Kaplan, S. N., & Weisbach, M. (1992). The success of acquisitions: evidence from divestitures. Journal of Finance, 47, 107–138. Kothari, S. P., & Warner, J. B. (1997). Measuring long-horizon security price performance. Journal of Financial Economics, 43, 301–339. Lang, L. H. P., Stulz, R., & Walkling, R. A. (1989). Managerial performance, Tobin’s Q and the gains from successful tender offers. Journal of Financial Economics, 24, 137–154. Lang, L. H. P., Stulz, R., & Walkling, R. A. (1991). A test of the free cash flow hypothesis: The case of bidder returns. Journal of Financial Economics, 29, 315–335. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. (1997). Law and Finance. Journal of Political Economy, 106, 1113–1155. Loderer, M., & Martin, K. (1990). Corporate acquisitions by listed firms: the experience of a comprehensive sample. Financial Management, 19, 17–33. Loughran, T., & Vijh, A. (1997). Do long-term shareholders benefit from corporate acquisitions? Journal of Finance, 52, 1765–1790. Maquiera, C., Megginson, W., & Nail, L. (1998). Wealth creation vs. wealth redistributions in pure stock-for-stock mergers. Journal of Financial Economics, 48, 3–33. Martin, K. J., & McConnell, J. J. (1991). Corporate performance, corporate takeovers and management turnover. Journal of Finance, 46, 671–687.
144
MARC GOERGEN AND LUC RENNEBOOG
McCahery, J., Moerland, P., Raaijmakers, T., & Renneboog, L. (2002). Corporate governance regimes: convergence and diversity. Oxford University Press. Mitchell, M., & Stafford, E. (2000). Managerial decisions and long-term stock price performance. Journal of Business, 73, 287–329. Morck, R., Shleifer, A., & Vishny, R. W. (1988). Managerial ownership and market valuation: an empirical analysis. Journal of Financial Economics, 20, 293–315. Mulherin, J. H., & Boone, A. L. (2000). Comparing acquisitions and divestitures. Journal of Corporate Finance, 6, 117–139. Rau, P. R., & Vermaelen, T. (1998). Glamour, value and the post-acquisition performance of acquiring firms. Journal of Financial Economics, 49, 223–253. Roll, R. (1986). The hubris hypothesis of corporate takeovers. Journal of Business, 59, 197–216. Servaes, H. (1991). Tobin’s Q and the gains from takeovers. Journal of Finance, 46, 409–419. Shelton, L. M. (2000). Merger market dynamics: insights into behaviour of target and bidder firms. Journal of Economic Behavior & Organization, 41, 363–383. Shleifer, A., & Vishny, R. (1989). Managerial entrenchment: the case of manager-specific investments. Journal of Financial Economics, 25, 123–139. Sirrower, M. (1997). The Synergy trap: how companies lose the acquisition game. New York: The Free Press. Schwert, G. W. (2000). Hostility in takeovers : in the eyes of the beholder? Journal of Finance, 55, 2599–2640. Vasicek, O. (1973). A note on using cross-sectional information in Bayesian estimation of security betas. Journal of Finance, 28, 1233–1239. Walker, M. (2000). Corporate takeovers, strategic objectives and acquiring-firm shareholder wealth. Financial Management, 53–66. Warner, J. B. (1977). Bankruptcy costs: some evidence. Journal of Finance, 32, 337–347. Yook, K. (2000). Larger return to cash acquisitions: signalling effect or leverage effect? Working paper. Johns Hopkins University.
37 119
93 37 18 45.9% 3 3 2 12 4
All Cash Bid All Equity Bid Cash/Equity Bid % Cash in Cash + Equity Bids Cash/Loan Notes Bid Equity/Loan Notes Bid Cash/Equity/Loan Notes Bid Choice Cash or Equity Bid Equity Bid (later: also cash offer made)
Failed bid Successful bid
77 76 101
U.K. Target U.K. Bidder Bidder and Target same country
0 29
26 0 0 – 2 1 0 0 0
11 13 15
29 0 0 0 1 28
Number
Number 156 55 40 40 21 0
Divestitures
Takeovers
Table A1. Sample Distribution and Descriptive Statistics.
Total sample Mergers Acquisitions Hostile Takeovers Multiple Bidders Bid on divestiture
Panel A: Sample composition:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
APPENDIX
Value Creation in Large European Mergers and Acquisitions 145
17878 4.26 0.07 3.97 50.80 11.53 6.13
Market Capitalization (USD m) MV/BV Ncash/MV Dividend yield (%) Interest Coverage Price/Cash flow ROE (%)
Targets
Divestitures (29 cases) Value of bid (USD M) Cash Bid (USD M) Cash + Loan Notes bid USD M) Equity + Loan Notes bid (USD M)
Takeovers (156 cases) Value of bid (USD M) Cash Bid (USD M) Equity Bid (USD M) Cash + Equity bid (USD M) Cash + Loan Notes bid (USD M) Equity + Loan Notes bid USD M) Cash + Equity + Loan Notes bid (USDM) Median 415 317 292 0
0
15779
Mean 481 504 217 370
Median 15694 147 0 0 0 0
15192 8.88 0.16 3.50 32.59 36.36 7.11
Stdev.
Mean 575 1489 14255 3084 114 29881
Panel C: Descriptive statistics of Means of Payment
Mean
Panel B: Descriptive Statistics (USD million)
Stdev. 310 318 106 0
22106
Stdev. 100 3403 28196 5318 13 15363
21568 4.01 0.09 2.78 13.41 10.51 5.5
Mean
Bidders
Min 110 110 142 0
0
Min 147280 0 0 0 0 0
28038 5.20 0.14 1.59 13.66 7.71 4.83
Stdev.
Max 1239 1239 292 370
31410
Max 218 24600 147280 19895 127 42729
15033 8.13 0.08 3.19 5.51 9.41 2.11
Q25 $212.55 160.4 142 0
0
Q25 2492 0 0 0 0 0
29694 22.3 0.08 2.27 5.99 7.81 4.52
Divesting firms Mean Stdev.
Q75 682.75 681.75 292 0
423 142 370
Non-zeros median
31410
443 2580 655 114 34052
581 100 0 0 0 0
Non-zeros median Q75
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
Table A1. Continued.
146 MARC GOERGEN AND LUC RENNEBOOG
THE OPERATING PERFORMANCE OF COMPANIES INVOLVED IN ACQUISITIONS IN THE U.K. RETAILING SECTOR, 1977–1992 Steve Burt and Robin Limmack ABSTRACT Takeover activity has played an important role in the restructuring of the U.K. retail sector over the past two decades and appears likely to do so in the future. Debate about the impact of this restructuring has focused to a large extent on the impact on competition, customers and employees. At the micro level research on takeovers has generally been directed to the likely shareholder wealth effects. Studies using methodologies based on the analysis of security returns have generally concluded that while shareholders of target companies experience wealth gains from takeover activity, on average shareholders of acquiring companies experience wealth losses. If takeover activity is not even in the interests of the shareholders of companies that make acquisitions, then it may be that the potential detrimental effects on other groups should produce a policy bias against takeovers. There are, however, a number of reasons why this approach ought not to adopted without question. First, questions have been raised about the methodologies adopted in studies based on analysis of security return behaviour, and in particular of those studies which evaluate long-run behaviour. Second the studies themselves are not Advances in Mergers and Acquisitions, Volume 2, pages 147–176. Copyright © 2003 by Elsevier Science Ltd. All rights of reproduction in any form reserved. ISBN: 0-7623-1003-0
147
148
STEVE BURT AND ROBIN LIMMACK
unambiguous in their conclusions on the net wealth effects. Because of the criticisms of studies based on security price behaviour, a number of researchers have taken an alternative approach to measuring the long-run effects of particular economic events, including takeover activity. This approach involves analysis of the cash flow operating performance of companies involved. In the current paper we apply this approach to an analysis of the operating performance of a sample of U.K. companies involved in takeover activity in the retailing sector over the period 1977–1992. We also investigate potential sources of changes in operating performance, together with the effect of the takeover activity on employment in the retailing sector.
INTRODUCTION Recent news announcements confirm that the period of restructuring of the U.K. retailing sector is not yet complete. Much of the public debate on these developments has related to the benefit, or otherwise, to the customer from restructuring activities and in particular to their effect on competition. In a few cases concern has also been expressed about the effect of heightened competition on employees, especially in those areas of retailing that are able to obtain inventories from non-U.K. suppliers. Surprisingly one stakeholder group that appear to have been taken for granted in this debate are the owners of the companies involved, and in particular the shareholders of the acquiring firm. It is not clear whether this oversight has arisen because shareholders are not an obvious candidate for public support or because it is assumed that, of all groups, shareholders are those most likely to benefit from takeover activity. However the latter assumption is not necessarily correct. Nor is the majority of the population of the United Kingdom unaffected by the impact of financial decisions on the wealth of shareholders of companies involved. There is currently a fairly large body of research that casts doubt on the ability of the management of companies involved in takeover activity to generate wealth for their shareholders. Studies generally agree that the shareholders of target companies experience large, significant wealth increases from the premiums paid to their companies when acquired. However it is not clear that the latter does not simply involve a transfer of wealth from the shareholders of bidding companies rather than any net wealth gain arising from synergies. In particular those studies that have examined the long-run postoutcome share price performance of acquiring companies have generally concluded that takeovers are not in the interests of shareholders of bidding companies.1 Such a conclusion will not come as a surprise either to those who
The Operating Performance of Companies Involved in Acquisitions
149
are familiar with literature on corporate governance or with the early accounting studies on takeovers in the U.K., published in the late 1960s and early 1970s.2 One of the features of governance literature relates to the inability of various governance mechanisms to overcome the agency problem. The recent survey on corporate governance by Shleifer and Vishny (1997) begins by highlighting the level of disagreement about the effectiveness of various governance mechanisms. While the threat of takeover may be seen by some as a means by which inefficient management may be disciplined,3 takeover activity itself also appears to be one of the manifestations of the problem. One of the problematic aspects of security market based studies of the wealth effects of takeovers is that they are dependent on the choice of an appropriate model of expected share price behaviour.4 Studies of long-run security returns behaviour are to a certain degree inconsistent with the accepted efficient markets paradigm that provides the foundation for studies of economic events. The latter studies are based on the assumption that share prices react in a speedy and unbiased manner to the expected present value of the future cash flow changes arising from the event under investigation. The apparent contradiction between the results of long-run studies of post-bid share price behaviour and the efficient markets paradigm have led a number of authors to suggest that the results may be a reflection of the choice of an inappropriate returns prediction model rather than the acquisition per se. However such a conclusion then leaves the question as to what is an appropriate model? While researchers have, in general, rejected early models of security price behaviour used in takeover studies, they are divided as to the choice of an appropriate alternative. As one alternative a number of authors have suggested using a methodology which tests for changes in operating performance following the acquisition rather than focussing on the wealth effects, if any.5 This approach has already been adopted in studies of the performance of companies following seasoned equity issues (Loughran & Ritter, 1997). It has also been used to a limited extent in studies of post-outcome acquiring company performance, both in the USA (Healy et al., 1992) and in the U.K. (Manson et al., 1994). The current study adopts this approach and tests for improvements in operating performance following takeovers of companies involved in the retailing industry in the United Kingdom. Thus our focus is not on the distribution of gains following takeovers but whether there is evidence for operating improvements that would give rise to wealth gains, however distributed. This approach itself is not without its critics as accounting-based measures of performance are subject to potential manipulation, lack of riskadjustment and choice of arbitrary time frame for analysis. However the approach taken in this study uses cash flow measures that are subject to reduced
150
STEVE BURT AND ROBIN LIMMACK
opportunity for manipulation and uses a time frame for analysis beginning five years prior to the acquisition and ending five years post-acquisition. We also adopt a matched firm approach that minimises any potential bias arising from incomplete risk adjustment. The structure of the remainder of this paper is as follows: Section (2) describes the sample and data used in the current study together with the basic methodology; Section (3) describes the results of our preliminary analysis; Section (4) extends the analysis to investigate possible sources of operating improvement; finally Section (5) provides a summary and conclusions together with suggestions for further research.
2. DATA AND METHODOLOGY Data The sample of takeovers selected for analysis was identified from the population of takeovers of U.K. listed companies in the retail sector that were initiated and completed over the period 1 January 1977 to 31 December 1992. In order to obtain comparable data for the analysis it was necessary to further restrict the sample to those takeovers in which both the acquirer and the target were publicly listed. The full set of tests used in the current paper also require accounting data to be available for our sample of bidding and target companies, and their respective controls, for a period of up to twelve financial years. This period commences six years prior to the year in which the takeover was initiated and, except for the target companies, extends to a period of five financial years following the end of the financial year in which the takeover was completed. However in order to reduce survivorship problems we impose no requirement as to the minimum period for which financial data must be available beyond one year following the takeover year.6 A total of 49 acquisitions satisfied this minimum requirement, representing £8.86 billion in purchase consideration at an average size of £174 million. While our final sample is fairly heavily biased towards large acquisitions it nevertheless represents a reasonable proportion of the value of takeovers of U.K. retailing companies in the period studied, and covers a range of retail sectors. The bias towards large acquisitions will be referred to in the final part of our analysis. Methodology Although we have identified disagreements over the choice of appropriate methodologies to be used in studies involving security returns, the choice of
The Operating Performance of Companies Involved in Acquisitions
151
measurement bases for evaluating operating performance is itself not unambiguous. Earlier accounting-based studies in the U.K. (Singh, 1971; Meeks, 1977; Cosh, Hughes & Singh, 1980; Holl & Pickering, 1988) and the U.S. (Lev & Mandelker, 1972; Mueller, 1980) focused on changes in net income to total sales as a performance measure. Other studies that have used operating income to total assets or sales to measure operating performance include those by Ravenscraft and Scherer (1987) and Jarrell (1995) in the U.S. The use of earnings based on accrual accounting is not without criticism, including those relating to inconsistent accounting methods (Appleyard, 1980) and opportunities for earnings management. Erickson and Wang (1999) provide evidence that managers of acquiring firms manipulate earnings in the quarter prior to the bid for equity offerings.7 The potential problems of earnings-based measures of operating performance has led a number of authors to adopt measures based on operating cash flows. Rayburn (1986) and Bowen, Burgstahler and Daley (1986) provide evidence that cash flows can incrementally explain abnormal stock returns. Bowen, et al. (1986) found ‘that cash information is consistent with the information impounded in security prices and also has incremental explanatory power beyond that contained in accrual flows alone’ (p. 746). According to Dechow (1994, p. 5) ‘many financial analyst regard operating cash flow as a better gauge of corporate financial performance than net income, since it is less subject to distortion from differing accounting practices.’ 8 The choice of operating cash flows as our numerator is consistent with that adopted in other similar studies including those of Manson, Stark and Thomas (1994) in the U.K. and Healy, Palepu and Ruback (1992) and Clark and Ofek (1994), Anand and Singh (1997) and Harford (2000) in the USA. Our measure of operating performance is therefore pre-tax operating cash flows deflated by the opening book value of operating assets. We use the book value, rather than market values partly to avoid the problems associated with market anticipation of the gains from the takeover.9 In addition Barber and Lyon (1996) demonstrate that the use of market values, rather than book values, adds little if anything to the measure of performance.10 Operating profit is calculated before the following items: depreciation and goodwill amortisation, gains or losses on the disposal of fixed assets, interest receivable and interest payable, tax, extraordinary items, and dividends received and paid. Operating cash flow is calculated based on operating profit adjusted for non-cash changes in operating assets and liabilities. The definition of operating cash flow used in this study is similar to that used by Manson et al. (1994) but differs from that used by Healy et al. (1992). The latter authors define operating cash flows as sales minus cost of goods sold and selling and
152
STEVE BURT AND ROBIN LIMMACK
administrative expenses, plus depreciation and goodwill expenses. In their paper Healy et al. (1992) make no adjustment for current accruals and nonoperating gains. Operating assets are defined as the total of net fixed assets plus current assets, less cash and marketable securities.11 Growth in operating assets is defined as the change in book value of operating assets over the year divided by the book value of operating assets at the beginning of that year. To avoid bias caused by the additional expenses involved in the acquisition itself and the effect of timing differences for the date of consolidation within the acquisition year, financial results for that year (t = 0) are excluded from comparison with pre-bid periods. In order to provide a benchmark measure of pre-acquisition performance for each takeover we identify a pro-forma measure (Ppre i,t) for the acquirer and target combined in each of the five years prior to takeover. This measure is computed by deflating the operating cash flows for the acquirer (target) by the asset value of the acquirer (target) at the beginning of the corresponding financial year and then computing the weighted average value for acquirer and target combined using the relative asset values as weights. The measure of asset value used in this study is based on the book value of total assets, excluding cash and marketable securities. Thus for firm i in year t the measure is calculated as:
Ppreit =
OCFA AssetA
AssetA OCFT + Asset(A + T ) AssetT
AssetT Asset(A + T )
or Ppreit =
OCFA + OCFT Asset(A + T )
A and T represent the acquirer and target respectively; OCF represents operating cash flow; Asset refers to the asset value of the respective company; and t is the financial year (t = –5 . . . , –1). A matched-firm control is obtained for each target and acquirer with matching based on industry and size.12 In order to reduce problems caused by imprecise matching we adopt a procedure similar to that used by Healy et al. (1992) and weight control company returns by the relative asset values of the two sample firms at the beginning of the relevant year.13 Although Barber and Lyon (1996) demonstrate that matching based on prior performance is more appropriate,
The Operating Performance of Companies Involved in Acquisitions
153
they also show that the use of industry matching can improve the power of tests over longer time periods. Because of a limited number of potential observations in each industry group we are unable to match exactly by prior performance and industry.14 We also use a series of tests, based on regression analysis, which attempt to control for difference in performance between our sample and the control companies. Pre-acquisition control returns (Pcpre i,t) are calculated as described above for the acquirer and target. The control-adjusted operating performance (APc) is obtained by subtracting the pro-forma control company measure of operating performance from the pro-forma measures of operating performance for the respective takeover i for year t. APci,t = Ppre i,t Pcpre i,t for t = –5, . . . –1 The post acquisition measure of operating performance is calculated as the operating cash flow of the newly combined firm in each of years 1 to 5, deflated by book value of operating assets at the beginning of the relevant year.15 As explained previously we use book values in order to exclude from the denominator the capitalised value of cash flow changes arising from the takeover. In their study Healy et al. (1992) use market values and exclude the change in market value of the target and acquirer from five days before the first offer until the date when the target is delisted on the assumption that this period captures all the wealth effects of the bid. However one of the primary reasons which we state for using cash flow analysis is that the results of studies based on the analysis of security returns have suggested that there is a long-run post outcome wealth effect. It would therefore be more appropriate to exclude all post-bid wealth changes in the denominator. Such an approach would, however, produce a denominator that was a mixture of market values and book values but measured at different points in time. We choose instead to use book value as the consistent measurement base throughout the study. One of the potential problems with the use of book values as denominator relates to the impact on measures of performance of the use of fair value adjustments at the time of the acquisition. We refer to this potential problem in the discussion of results. The combined control company performance in post acquisition years is computed by weighting the performance of the individual control companies by the relative operating asset values of acquirer and target at the end of the last financial year prior to the acquisition (t 1).16 We also test the sensitivity of our results to the alternative weighting method.
154
STEVE BURT AND ROBIN LIMMACK
In addition to the above, a summary measure of operating performance for the combined acquirer and target is calculated for each acquisition based on the median control-adjusted performance for the five years prior to acquisition and a similar summary statistics is constructed for the five years post-acquisition. APcprei = Median [APci 5, . . ., APci 1] APcposti = Median [APci,1, . . ., APci,5] In order to control for relative pre-bid performance we then apply the following model: c BT c (PM post,i Ppost,i) = + (Ppre,i Ppre,i) + I
where c (PM post,i Ppost,i) = the post acquisition median annual control-adjusted operating return for ‘company’ i c (PBT pre,i Ppre,i) = the pre-acquisition median annual control-adjusted operating return for the same ‘company’
= is the measure of the abnormal control-adjusted return and is independent of pre-acquisition returns = the slope coefficient that captures any correlation in returns between pre- and post acquisition years, and represents the performance that would have been achieved independent of the acquisition.
3. PRELIMINARY RESULTS We begin our analysis by compiling data on rates of growth in operating cash flows and operating assets for the acquirer, target and control companies. In Table 1 we present preliminary data on the characteristics of the firms involved in takeovers and the respective control companies. Panel A summarises the rates of growth in cash flows for the five years prior to the bid for targets, acquirers and their respective controls. Targets appear to demonstrate higher rates of growth on average than their controls from year –5 to 4, –4 to –3, and –3 to –2 but lower rates of growth in year –2 to –1. However the results are influenced by a number of extremely high values in each of these years. Using both Analysis of Variance and Mann-Whitney tests we find no significant difference in cash flow growth rates between targets and their controls at the 5% level for any period. The rate of growth in operating cash
The Operating Performance of Companies Involved in Acquisitions
155
flows of acquirers is significantly higher (at the 10% level) than that of the control companies only from year –3 to year –2. The median annual rate of growth in cash flows for acquirers over the entire five-year pre-bid period is, however, significantly higher than that of the controls at the 1% level. As described in Panel B of Table 1, the rate of growth in operating assets is significantly higher for targets than their controls in years –5,–3 (based on mean values) and in year –1. For acquirers, there is a significantly higher rate of growth in years –4 and –2 only. Based on this data alone there is only slight evidence of acquirers buying growth. Table 1. Pre-Bid Rates of Growth in Operating Cash Flows and Operating Assets for Companies Involved in Retail Acquisitions Undertaken in the Period 1977–1992. Figures in Cells Represent Mean Followed by Median Values. Period relative to bid year (year 0)
Target
%
Control Control- Acquirer Control Control- Observations adjusted adjusted %
%
%
%
%
N
64, 16 97, 19 101, 41 93, 32
–29, 15 37, 10 9, 10 39, 4
93, 5 60, 8 97c, 25 54, 43
36 39 41 44
60, 38
22, 14
38a, 30a
44
22, 14 –51, 4 12, 14 25c, 3 8, 5 1, 2 10, 10 29c, –4 14, 10 8, –2
36 39 41 44 44
Panel A: Cash Flow Growth Rate: Year –5 to –4 Year –4 to –3 Year –3 to –2 Year –2 to –1 Median annual performance over years –5 to –1
105, 21 101, 25 4, –17 386, 26 26, 8 360, 32 38, 2 –23, 14 61, 1 –50, 9 6, 16 –57, 7
–42, 32
19, 21 –61, 5
Panel B: Operating Asset Growth Rate* Year –5 Year –4 Year –3 Year –2 Year –1 Median annual performance over years –5 to –1
34, 15 13, 13 22, 11 26, 12 56, 16
15, 15 13, 6 7, 6 14, 10 3, 5
18b, –4 0, 4 15b, 9 12, –3 53c, 16a
–29, 15 37, 10 9, 10 39, 4 22, 14
36, 16
11, 9
25, 7
21, 13
9, 10
Note: Asset growth rates are calculated using start-of-year values. Significance tests refer only to the control-adjusted returns. a Indicates significantly different from zero at 1% level of significance. b Indicates significantly different from zero at 5% level of significance. c Indicates significantly different from zero at 10% level of significance.
12, 4
44
156
STEVE BURT AND ROBIN LIMMACK
Panel A of Table 2 provides a summary of the post-bid rates of growth in operating cash flows of the newly combined firm relative to the pro-forma control. The base year against which growth is measured is the financial year ending prior to the year in which the acquisition takes place. With the exception of the period from year –1 to year + 4, there is no significant difference in the rate of growth of cash flows for the newly combined firm compared to the control firms. As shown in Panel B of Table 2, the control-adjusted median rate of growth of operating assets is actually significantly negative in all five postacquisition periods, relative to year-1. The results provide further support for the view that the acquisition activity is not simply growth driven. Rather it appears that, as that the rate of growth in operating cash flows is not significantly lower than in the pre-bid period (Table 2, Panel A), the combined company is making more efficient use of operating assets post-acquisition. We next compare the pre-bid operating performance of both the target and acquirer sample relative to their control firms. The results of this analysis are summarised in Table 3. Panel A of Table 3 provides a summary of the deflated operating cash flow returns of acquirers and control firms for years –5 to –1 prior to the bid year while Panel B of the table provides similar data for the targets and their control companies. Although there are large differences in mean performance reported between acquirers and their controls, the variation in returns across observations is large and in no year is the difference statistically significant. For target companies, the only pre-acquisition year in which the control adjusted target return is significantly different from zero is year –5. In that year both mean and median target returns are significantly lower than that of their matched firm control. In the remaining four years the performance of the target companies is not significantly different from that of their control companies. There is some tentative evidence from these results that acquirers generally achieve higher control-adjusted operating performance than their targets. However the results are not statistically significant. The lack of statistical significance in the results reported in Panel B do not allow the conclusion to be drawn that targets in retail acquisitions have a history of poor pre-bid performance. However neither of the above results of themselves rule out the possibility of economic benefits from takeover activity. For consideration of this issue we need to examine post-bid operating performance. We measure operating performance of the pro-forma combined firms for each year before the acquisition using the sum of acquirer and target operating cash flows divided by the sum of acquirer and target book value of operating assets at the beginning of that year. Control group cash flow returns are also
The Operating Performance of Companies Involved in Acquisitions
157
Table 2. Post-Bid Rates of Growth in Operating Cash Flows and Operating Book Value of Assets for Companies Involved in Acquisitions in the Retailing Industry in the U.K. Undertaken in the Period 1977 to 1992a. Period relative to bid year (Year 0)
Year –1 to Year + 1
Year –1 to Year + 2
Year –1 to Year + 3
Year –1 to Year + 4
Year –1 to Year + 5
%
%
%
%
%
65 24 51
50 39 –2
61 43 4
98 63 64b
117 81 –6
103 69 33
156 47 109
153 139 13
322 96 226c
272 156 116
–24 21 –34b
–17 45 –36b
–7 48 –42b
0 66 –69b
4 76 –77b
44 27 17
86 51 35
106 59 47
175 81 94
206 93 113
44
44
41
38
34
PANEL A: Cash Flow Growth Rate Median: Firm Control Control Adjusted Mean: Firm Control Control Adjusted
PANEL B: Operating Asset Growth Rate Median: Firm Control Control Adjusted Mean: Firm Control Control Adjusted observations (n)
a Operating cash flow is defined as operating profit before tax and extraordinary items, adjusted for depreciation and goodwill and changes in working capital (that is, changes in stocks, trade debtors and prepayments and changes in creditors and accruals). The operating book value of assets at the beginning of the year is the book value of shares plus net debt less cash and marketable securities. Before the acquisition (year –1), cash flow and asset values of the combined firm are weighted averages of the acquirer and target values, with the weight being the relative asset values of the two firms. The values of the combined firm are used in the post acquisition period. Pre- and postacquisition control group returns are control target and control acquirer values, weighted by the relative asset values of the two corresponding sample firms at the beginning of the year prior to acquisition (year-1). Control adjusted values are computed for each firm and year as the difference between the firm value in that year and the value of the control firm in the same industry during that period. Significance tests refer only to the control-adjusted returns. b Significantly different from zero at the 5% level, using a two-tailed test. c Significantly different from zero at the 10% level, using a two-tailed test.
158
STEVE BURT AND ROBIN LIMMACK
Table 3.
Pre-Bid Operating Performance of Targets and Acquirers Relative to Controls.
Panel A: Operating performance for acquiring and control firms in years prior to acquisition1 Year relative to acquisition (Year 0)
Year –5 Year –4 Year –3 Year –2 Year –1 Median annual performance over years –5 to –1
Acquirer
Control
Difference
Observations
Median %
Mean %
Median %
Mean %
Median %
Mean %
N
23 19 25 24 29
44 31 122 –45 39
21 20 20 25 27
22 23 24 25 28
3 1 1 4 2
21 9 98 –70 11
36 37 41 44 44
26
28
23
26
1
2
44
Panel B: Operating performance for acquired and control firms in years prior to acquisition1 Year relative to acquisition (Year 0)
Year –5 Year –4 Year –3 Year –2 Year –1 Median annual performance over years –5 to –1 1
Target
Control
Difference
Observations
Median %
Mean %
Median %
Mean %
Median %
Mean %
N
0 17 22 21 21
2 20 22 29 24
16 21 19 21 21
18 23 23 23 22
–15a –3 0 0 –1
–15a –1 –1 6 1
36 36 39 42 44
19
18
20
22
–4
–3
44
Operating performance is defined as operating cash flow deflated by the book value of operating assets. Operating cash flow is operating profit before tax and extraordinary items, adjusted for depreciation and goodwill and changes in working capital. Operating assets at the beginning of the year is the book value of equity plus debt less cash and marketable securities. Control adjusted values are computed for each firm and year as the difference between the firm value in that year and the value of the control firm in the same industry during that period. Significance tests refer only to the control-adjusted returns. a Significantly different from zero at the 1% level, using a two-tailed test. b Significantly different from zero at the 5% level, using a two-tailed test. c Significantly different from zero at the 10% level, using a two-tailed test.
The Operating Performance of Companies Involved in Acquisitions
159
measured over both pre and post-acquisition periods, using the weights described earlier.17 The first two columns of data in Table 4 provides a summary of the operating performance of the pro-forma combined firm in the eleven year period centred on the bid year, but excluding the bid year itself. Columns 3 and 4 provide the corresponding data for the pro-forma control while columns 5 and 6 summarise the control-adjusted operating performance. As the results in Panel A of Table 4 indicate the weighted average operating performance of the pro-forma combined firm appears to be similar to that of the control firm in each of the five years prior to the bid. The only significant difference in performance is for year –2 in which the combined firm earns a 9% significantly higher mean operating performance than the pro-forma control.18 The lack of significant difference in pre-acquisition performance is not surprising given the chosen matching criteria. What we are more concerned with is in identifying any significant change in operating performance following the acquisition. As the results in Panel B of Table 4 demonstrate, although the operating performance of the new combination is never less than that of the controls, it is significantly higher only in the first full year after the year of acquisition. In no postacquisition year does the average performance of the newly combined firm fall below that of the pro-forma control.19 In order to test whether the operating performance of the combined firm has improved following acquisition we undertake a regression of post acquisition operating performance on pre-acquisition operating performance. The median value of control-adjusted operating performance for each pro-forma combined firm is identified over the five years prior to the bid year. A similar figure is then identified for the five years following the bid year.20 The model adopted is as follows:21 APcposti = + APcprei + I (1) where APcposti is the median (or average) annual adjusted cash flow returns for company i for the post acquisition years. APcprei is the pre-acquisition median (or average) for the pro-forma company i. The abnormal adjusted cash flow returns is interpreted as the component of post acquisition operating performance that is not explained by the preacquisition operating performance but is caused by the acquisition. The intercept is used as the measure of the abnormal adjusted cash flow return (changes in performance caused by acquisition). The slope coefficient represents the correlation in operating performance, if any, between pre- and posts acquisition years.
160
STEVE BURT AND ROBIN LIMMACK
Table 4. Operating Performance for Combined Acquiring and Target Firms in Years Surrounding Acquisitions Completed in the Period 1977–19921. Year relative to acquisition (Year 0)
Combined Firm
Control Group
Control-Adjusted
Observations
Median %
Mean %
Median %
Mean %
Median %
Mean %
N
20 17 22 24 26
24 22 27 33 27
19 21 24 26 26
20 22 24 24 28
3 –3 1 2 1
5 –1 3 9c –1
36 36 39 42 46
23
24
23
25
1
1
46
34 34 23 28 24
43 30 29 30 34
26 25 24 23 20
28 30 27 25 29
7b 3 2 7 9
15b 0 2 6 5
46 46 44 40 36
25
31
24
27
4
4
46
Panel A: Pre-Bid Year –5 Year –4 Year –3 Year –2 Year –1 Median annual performance over years –5 to –1 Panel B: Post-Bid Year + 1 Year + 2 Year + 3 Year + 4 Year + 5 Median annual performance over years + 1 to + 5 1
Operating performance in the pre-acquisition period are pre-tax operating cash flow return on assets of target and acquirer, weighted by the relative asset sizes of the two firms. Post-acquisition performance used data of the combined firms. Pre-acquisition control group returns are control target and control acquirer values, weighted by the relative asset values of the acquirer and target firms at the beginning of the year. Post-acquisition control group returns are control target and control acquirer values, weighted by the relative asset values of the acquirer and target firms at the end of year t 1. Control adjusted measures of performance are computed for each combined firm as the difference between the firm measure in that year and the measure for the control ‘firm’ in the same industry during that period. Significance tests refer only to the control-adjusted returns. a Significantly different from zero at the 1% level, using a two-tailed test. b Significantly different from zero at the 5% level, using a two-tailed t-test. c Significantly different from zero at the 10% level, using a two-tailed t-test.
The Operating Performance of Companies Involved in Acquisitions
Table 5.
161
Abnormal Adjusted Operating Cash Flow Returns. APcpost i = 0.06 + 0.266 APcpre i (2.08) (3.42) R2 = 19.9% F-statistic = 11.70
APcpost,i and APcpre,i are the median annual adjusted operating cash flow returns in the post- and preacquisition period for firm i.
The results of the regression on median abnormal control adjusted cash flow returns are described in Table 5. Both the slope coefficient , of 0.266, and the intercept, , of 0.06, are significantly different from zero. The results show that acquirers experience an improvement in operating performance post-bid once pre-bid performance is controlled for. The above results may be affected by a number of factors including the definition of operating performance employed in the study. In particular these results are sensitive to the choice of weighting adopted for control companies. The current study uses the same weights as those adopted by Healy et al. (1992), namely the relative value of operating assets of acquirer and target at the end of the year prior to the bid. This choice helps avoid bias caused by imprecise size matching. The use of alternative weights based on the relative operating assets of the control companies at the beginning of each year results in a small but insignificant alpha. We also examine the potential impact of the use of fixed asset write-downs following the acquisition.22 Examination of the financial statements of the newly combined group in the year of the acquisition reveals reference to fair value adjustments. In a number of cases these were not identified separately to the goodwill that had been written-off to reserves. However we were able to identify 39 cases in which there was specific reference to asset revaluations, with the majority (35 cases) involving writing down the value of the assets acquired. We therefore examined the notes to the group financial statements for the year of acquisition in order to identify the amount of the revaluation adjustment, if any. In 26 cases for which details were provided the total value of asset write-down’s identified were £274.9 million with £122.4 million of upward revaluations. The net amount of these revaluations amounted to less than 0.3% of the combined fixed assets of the target and bidders prior to the acquisition. We are therefore confident that any asset revaluation to fair value will have had an insignificant effect on the results reported above.
162
STEVE BURT AND ROBIN LIMMACK
One additional potential problem associated with this data set is the impact of leased assets on our measure of performance. As Beattie et al. (1998) point out, exclusion of leased assets from measures of performance is liable to affect inferences on profitability.23 We were, however, unable to undertake any adjustment for the impact of leases as much of our data was extracted for the period prior to July 1987 when the disclosure requirements of SSAP21 became effective. Any change in the level of use of leased assets by the newly combined group post-acquisition is however likely to offset in part at least by a change to the retail sector as a whole, including our control companies. We continue our analysis with an investigation of other factors that may have led to the perceived improvement in post-acquisition operating performance. One possibility is that the acquiring firms’ may have used the opportunity provided by the acquisition to undertake asset write-offs, thus reducing the post-acquisition denominator.24 It is possible that the lower rate of growth in operating assets post-acquisition, recorded in Table 2, reflects the above rather than an improvement in asset utilisation that is justifiably reflected in the measure of operating performance. We test for the above possibility in two ways. First we test for any significant change in the level of depreciation and asset write-offs, as opposed to disposals, over the post-acquisition period. Secondly we examine the level of write-offs and asset disposals in the bid year itself. In our analysis we compare the post-acquisition level of depreciation for the control-adjusted pre-acquisition level. First we calculate the combined level of pre-bid depreciation for the acquirer and control sample in each of the five years prior to the bid, weighted by their net fixed asset values at the beginning of each year. We then calculate a similar rate for the relevant control firms. Control adjusted rates of depreciation are then calculated for each year and median rates also calculated for each combination over the five years prior to the bid year. In a similar fashion post-bid depreciation rates are then computed for the acquirer firms, together with the relevant combined acquirer plus target control rates, in each of the five years following the year in which the bid was completed together with a median value over the five years. We then regress the median post-bid depreciation rate against the pre-bid rate with the results described in Table 6. The results in Table 6 suggest that the rate of depreciation has significantly reduced, rather than increased, in the five years following the bid after controlling for pre-bid performance and external factors. As the lower depreciation charge will reduce the rate of change of the denominator, this will induce a slight bias against finding any increase in post-bid performance. We also test for and find no significant difference in the level of depreciation write-
The Operating Performance of Companies Involved in Acquisitions
Table 6.
163
Adjusted Depreciation Rates.
DRcpost i = –0.018 + 0.541 DRcpre i (–2.79)a (5.89)a R2 = 43.4% F-statistic = 36.47
N = 45
DRcpost,i and DRcpre,i are the median annual adjusted depreciation rates in the post- and preacquisition period for firm i. a Indicates significant at 1% level.
offs in the bid year compared to the pre-bid period. We therefore conclude that the improvement in post-acquisition operating performance is not a function of higher depreciation but reflects an improvement in the utilisation of operating assets.
4. POTENTIAL SOURCES OF OPERATING EFFICIENCY GAINS AND LOSSES In the current section we analyse components of our measure of operating performance in order to identify whether there are any specific sources of change in operating performance. Initially we test for changes in operating return on sales and then test for changes in sales turnover. Operating return on sales in the current study is measured as operating cash flow divided by sales. Table 7 provides summary statistics of the control-adjusted operating return on sales for acquirer and target firms and the pro-forma combined firms in the five years prior to acquisition and for the acquiring firms for the five years postacquisition. As shown in Panel A of Table 7, acquirers and targets each experience significantly lower control-adjusted operating return on sales in each year prior to the year of acquisition. A similar result is also obtained for the pro-forma combined acquirer and target firms, with median adjusted operating return on sales of –3.09% over years –5 to –1. The pattern of lower operating return on sales is repeated in each of the five individual pre-bid years. The consistency of the pattern is rather surprising but demonstrates lower operating margins earned by both acquirer and target pre-bid. The results for the post-acquisition period are reported in Panel B of Table 7. In the post acquisition period, years + 1 to + 5, the newly combined firm continues to earn a lower rate of operating cash flow on sales with a median control-adjusted return of –2.87% over the five years. Although not reported in
164
STEVE BURT AND ROBIN LIMMACK
Table 7. Operating Return on Sales for Combined Acquiring and Target Firms and their Control Companies for Acquisitions in the U.K. Retail Industry, Completed in the Period 1977–19921. Panel A: Pre-bid Operating Return on Sales Year relative to acquisition (Year 0)
Year –5 Year –4 Year –3 Year –2 Year –1 Median annual performance over years (as Table 4): –5, –1
Control-Adjusted Acquirer Return
Control-Adjusted Target Return
Control-Adjusted Combined Return
Mean %
Mean %
Mean %
Median %
Median %
Median %
Observations
N
–1.04 –3.25a –5.28a –4.41a –5.55a
a
–3.35 –2.57b –3.18a –2.30a –2.90a
a
–3.30 –5.03a –4.03a –4.55a –5.79a
a
–2.60 –5.62a –3.28a –3.68a –4.44a
a
–2.95 –4.36a –4.10a –3.59a –5.47a
a
–2.54 –3.67a –3.14a –3.17a –2.78a
37 40 42 46 49
–5.10a
–2.50a
–5.05a
–3.30a
–4.88a
–3.09a
49
Panel B: Post Acquisition Operating Return on Sales Year relative to acquisition (Year 0) Year + 1 Year + 2 Year + 3 Year + 4 Year + 5 Median annual performance over years + 1 to + 5 1
Combined Firm Return
Pro-forma Control
Control-Adjusted Return
Observations
Median %
Mean %
Median %
Mean %
Median %
Mean %
N
11.15 9.07 7.94 8.90 7.38
10.16 9.32 9.01 10.00 9.44
11.96 13.37 11.63 11.71 12.06
11.59 12.93 12.26 12.26 13.51
–1.29 –4.15b –3.84b –2.62c –4.30b
–1.43 –3.61b –3.25a –2.28c –4.07b
49 49 47 43 39
8.57
9.65
11.60
12.31
–2.87b
–2.77b
49
Operating return on sales is defined as operating cash flow as a percentage of sales. Performance measures for the combined firm and control in the pre-acquisition period are weighted by the relative sales of the two firms. Post-acquisition performance used data of the combined firms. Post-acquisition control group returns are control target and control acquirer values, weighted by the relative sales. Control adjusted values are computed for each firm and year as the difference between the firm value in that year and the value of the control firm during that period. a Significantly different from zero at the 1% level of significance. b Significantly different from zero at the 5% level of significance. c Significantly different from zero at the 10% level of significance. Significance tests refer only to the control-adjusted returns.
The Operating Performance of Companies Involved in Acquisitions
165
detail we also test for any change in operating return post-bid but are unable to find any significant improvement. We next turn our attention to asset utilisation and in particular sales turnover on operating assets. We have already provided evidence, in Table 2 that operating asset growth rates were lower post-acquisition. Table 8, Panel A, provides summary data on acquirer, target and control firm sales turnover rates for each of the five year periods before and after the year of acquisition. Sales turnover is defined as the ratio of sales for the year to operating assets at the beginning of that year. As the data shows the pro-forma acquirer, target and combined firms have significantly higher sales turnover ratio in each of the five years prior to the bid year relative to their controls. Similarly the combined firm achieves a significantly higher level of sales turnover than the pro-forma combined control in each of the five years post-acquisition. Once again the consistency of the result is surprising. In order to test whether the turnover ratio has changed significantly from the pre- to the post-bid period, we regress the median value for the five years post-acquisition adjusted asset turnover on the five year median pre-acquisition control adjusted asset. As shown in panel C of Table 8, the slope coefficient from this regression, of 0.645, is significantly different from zero at the 5% level. The intercept, of 1.43, is also significant at the 5% level, indicating that there is a significant improvement in the combined firms’ asset turnover in the post acquisition period. In summary the results of the above analysis indicate that, relative to their controls, both acquirers and targets have significantly lower operating profit margin both prior to and following acquisition but significantly higher levels of sales turnover. The improvement in operating performance, reported in Table 5, appears to have been generated by an increase in sales turnover, whereas casual observation would have suggested a greater potential for improvement in the profit margin. This unexpected pattern may find an explanation in the nature of the retail sector. Discussions with management in this sector suggest that the prime motive is the acquisition of stores and floor-space. Planning controls together with characteristics of the property market can create a scarcity value for floor-space. The acquisition of additional floor-space will (or should) lead to an immediate improvement in sales turnover. The scope to leverage profit growth is more limited, and as noted from our preliminary results is more likely to be dependent on asset utilisation rather than disposals or write-offs. The desire to acquire sites and floor-space plus the scarcity value and, for high street stores in particular, rental/leasing arrangements make asset disposal both unlikely and at certain times difficult. Pointers to other potential sources of improved margin may be found in the cost structure of retail companies. Although retail cost structures do vary from
166
STEVE BURT AND ROBIN LIMMACK
Table 8. Sales Turnover for combined acquiring and target firms and their control companies for acquisitions in the U.K. retail industry, completed in the period 1977–19921 Figures in table represent mean values, followed by medians. Panel A: Pre-Bid Turnover Ratios Year relative to acquisition (Year 0)
Year –5 Year –4 Year –3 Year –2 Year –1 Median annual performance over years –5 to –1
ControlAdjusted Acquirer
ControlAdjusted Target
Combined Firm
Pro-forma Control
ControlAdjusted Return
Observations
%
%
%
%
%
N
3.63b, 1.09a 3.03a, 1.16a 2.20a, 0.94a 2.86a, 1.29a 2.53a, 1.12a
2.27a, 1.32a 1.85a, 1.21a 2.12a, 1.03a 1.51a, 0.83a 1.40b, 1.33a
4.55, 3.19 4.29, 3.41 4.64, 3.27 5.17, 3.30 5.29, 3.57
2.42, 2.24 2.31, 2.08 2.26, 2.07 2.43, 2.16 2.49, 2.31
2.13a, 0.66a 1.98a, 0.78a 2.38a, 1.03a 2.74a, 1.16a 2.80a, 1.43a
39 41 43 44 47
2.34a, 1.40a
1.76a, 1.12a
4.67, 3.56
2.34, 2.16
2.29a, 1.33a
48
Panel B: Post-Acquisition Turnover Ratios Year Relative to Acquisition (Year 0) Year + 1 Year + 2 Year + 3 Year + 4 Year + 5 Median annual performance over years + 1 to + 5
Combined Firm
Pro-forma Control
Control-Adjusted Return
Obser– vations
Median
Mean
Median
Mean
Median
Mean
N
3.24 2.97 2.71 2.79 2.99
5.20 3.81 3.55 3.46 3.73
2.08 1.88 1.99 1.86 2.00
2.59 2.32 2.50 2.24 2.30
1.09a 0.96a 0.92a 0.96a 0.98a
2.61b 1.49a 1.05b 1.22a 1.42a
49 49 47 44 43
2.73
3.62
2.04
2.41
0.92a
1.24a
49
The Operating Performance of Companies Involved in Acquisitions
167
Table 8. Continued. Panel C: Abnormal adjusted post acquisition sales turnover (t-values in parentheses) ATcpost i = 1.43 + 0.645 ATcpre i (2.03) (2.37)
R2 = 11.1% F-statistic = 5.64
ATcpost,i and ATcpre,i are the median annual control adjusted sales turnover ratios in the post and prior acquisition period for firm i. 1
Sales turnover is calculated as sales for the year divided by operating assets at the beginning of the year. Performance measures for the combined firm and control in the pre-acquisition period are weighted by the relative book value of operating assets of the two firms. Post-acquisition turnover uses data for the combined firms. Post-acquisition control group turnover rates are based on control target and control acquirer values, weighted by the relative operating assets. Control adjusted values are computed for each firm and year as the difference between the firm value in that year and the value of the control firm during that period. a Significantly different from zero at the 1% level of significance. b Significantly different from zero at the 5% level of significance. c Significantly different from zero at the 10% level of significance. Significance tests refer only to the control-adjusted returns.
sector to sector, the largest items are the cost of goods sold (typically 70–80% of sales in the grocery sector) and labour costs (8–10%). Whilst increased scale clearly contributes to the ability of companies to administer purchasing activities more effectively, the scope for immediate profit improvement from other areas is more limited. Although productivity gains may accrue from distribution efficiencies, these make up a relatively small proportion of total costs. Any immediate benefits from the acquisition will also be offset in many cases by the costs of rebranding the acquired chain with the acquirers facia, store concept and reporting systems. Evidence on the internal investment activities of acquiring firms suggests that management groups are often optimistic about the future prospects for their company. The optimism is reflected to a certain extent in the relatively high level of new investment undertaken in these firms in the period surrounding the acquisition. The optimism is often shared by other groups, including analysts. Evidence from studies on security return behaviour of acquiring companies, however, suggests that this evidence may often be misplaced. We examine the rate of new fixed asset acquisitions as a proxy for this optimism. The rate of fixed asset acquisition is defined as new fixed assets acquired during the year divided by net fixed assets at the beginning of the year. We report summary statistics for both acquirers and targets in our sample together with that of the control companies in Table 9. In addition we provide
168
STEVE BURT AND ROBIN LIMMACK
Table 9. Rate of Acquisition of Fixed Assets for combined acquiring and target firms and their control companies.1 Figures in table represent mean values, followed by medians. Panel A: Pre-Bid Acquisition Rates ControlAdjusted Acquirer
ControlAdjusted Target
Combined ControlAdjusted Return
Observations
%
%
%
N
14.2, 3.3 43.1b, 4.0 21.0, 3.4 18.7a, 6.1a 21.3a, 13.8a
21.9b, 12.3 9.3, 10.5 21.3a, 6.1a 7.7, 7.0b 50.3c, 16.8a
5.67, –0.28 12.64c, 1.78 14.94c, 2.88 11.97a, 3.03c 18.58b, 3.81c
33 37 41 44 47
17.5a, 3.6b
30.8b, 9.6a
14.04b, 3.67c
48
Year relative to acquisition (Year 0)
Year –5 Year –4 Year –3 Year –2 Year –1 Median annual performance over years –5 to –1
Panel B: Post-Acquisition Fixed Asset Acquisition Rates Year relative to acquisition (Year 0)
Year + 1 Year + 2 Year + 3 Year + 4 Year + 5 Median annual performance over years + 1 to + 5
Combined Firm
Pro-forma Control
Control-Adjusted Return
Observations
Median %
Mean %
Median %
Mean %
Median %
Mean %
N
29.0 20.8 25.7 15.7 18.1
45.4 31.3 30.2 21.9 26.8
29.4 29.0 23.8 25.0 22.5
34.5 28.0 27.9 32.3 27.5
4.8 1.2 0.6 –5.7c –2.2
10.6 3.3 2.2 –10.4 –0.7
36 34 31 18 12
22.5
25.8
26.2
27.2
–3.0
–1.4
36
The Operating Performance of Companies Involved in Acquisitions
169
Table 9. Continued. Panel C: Abnormal adjusted fixed asset acquisition rates (t-values in parentheses): ATcpost i = –0.017 + 0.031 ATcpre i (–0.70) (0.50)
R2 = 0% F-statistic = 0.25
ATcpost i and ATcpre i are the median annual control adjusted rates of new fixed asset acquisition in the post- and prior-acquisition period for firm i. 1
New fixed asset acquisition rate is calculated as new fixed asset acquisitions for the year divided by net fixed assets at the beginning of the year. Acquisition rates for the combined firm and control in the pre-acquisition period are weighted by the relative net fixed assets of the acquirer and target. Post-acquisition turnover uses data for the combined firms. Post-acquisition control group acquisition rates are weighted by target and acquirer net fixed assets at time t 1.27 Control adjusted values are computed for each firm and year as the difference between the firm value in that year and the value of the control firm during that period. a Significantly different from zero at the 1% level of significance. b Significantly different from zero at the 5% level of significance. c Significantly different from zero at the 10% level of significance. Significance tests refer only to the control-adjusted returns.
a pro-forma estimate of the rate of new asset acquisition by the combined company in the five years prior to the financial year in which the acquisition occurred and compare this with the results for the combined control companies. We weight the rate of acquisition of the acquirer and target by the relative value of their fixed assets and apply the same weighting to the control companies for the pre-acquisition years. In the post-acquisition period we apply the relative fixed asset values of acquirer and target at the end of year t 1 as weights to the control companies. As the results in Panel A of Table 9 demonstrate both acquirers, targets, and the pro-forma combined companies have significantly higher rates of fixed asset acquisition than their control companies for a number of years prior to the year of acquisition. However in the post acquisition period (Table 9, Panel B) there is only one year (year 4) in which the rate of asset acquisition is significantly higher than that of the combined control companies. The results may be interpreted in a number of ways. One interpretation of the results is that they demonstrate an excessive level of investment prior to the acquisition in both acquirer and target companies, reflecting over-optimism. An alternative interpretation is that they represent an investment programme in a period of high growth for the companies involved that has subsequently produced benefits in terms of the higher sales turnover as reported earlier in Table 8. As floor-space is typically the main asset that a retailer will invest in, the period
170
STEVE BURT AND ROBIN LIMMACK
following the acquisition of new sites and floor-space is likely to result in management turning its attention to integration of that asset rather than the search for further “new” assets. Such activities would, as noted above, include rebranding and refitting of stores together with changes in systems and practices. Furthermore, as the number of potential sites is theoretically finite (a certain catchment area is required to justify operating a store), one could argue that with each successive takeover the scope for further investment is reduced. A number of studies of takeover activity have questioned whether shareholder groups have gained at the expense of other stakeholder groups, in particular employees. It is clear that, in the USA in particular, takeovers (or the threat of takeover) have been used to justify changes, i.e. reductions in wage rates.25 In addition to the above, if potential economies of scale were to arise as result of the takeover then one source of gains would be through more efficient use of employees. In the current study we investigate changes in the level of employment of the firms involved in takeovers. We use employment in the control groups to compare the numbers employed in acquirer and target over the eleven years centred on the takeover year, but excluding that year. The median value of numbers employed in acquirer and target combined is calculated separately over the five years prior to the bid and the five years following the bid year, using data extracted from the annual accounts. We then adjust each of these figures for the level of employment in control groups in the corresponding periods. Finally we regress the median control-adjusted using the following model: EMPcposti = + EMPcprei + I
(2)
EMPcposti is the median annual (or average) control-adjusted number of employees for company i for the post acquisition period (t = 1, t = 5) and EMPcprei is the corresponding figure for the pre-acquisition period (t 5, t 1). As an alternative we calculate changes in employment, deflated by the opening level of employment, both for our combined acquirer and target companies and also for their respective controls. We then run the regression using control-adjusted change in pre and post-bid employment rates as the variables. The results of this regression are shown in Panel B of Table 10 and indicate a high correlation between pre- and post-bid control-adjusted levels of employment. Neither the results in Panel A for levels of employment, nor those in Panel B, based on changes, indicate that, after controlling for external factors, takeovers of themselves reduce the level of employment in the firms involved. Again this finding is less of a surprise for the retail sector. The labour deployment characteristics of the sector differ markedly from that of
The Operating Performance of Companies Involved in Acquisitions
Table 10.
171
Comparison of Pre- and Post-bid Control-adjusted Employment for Firms Involved in Retail Mergers.
Panel A: Levels of Employment (t values in parentheses): EMPcposti = 13 + 0.989 EMPcprei (0.0) (7.67) Adjusted R2 = 59.1%
F = 58.85
N = 41
Panel B: Change in Employment Levels Changecposti = –0.008 + 0.0053 Changecprei (–0.99) (1.45) Adjusted R2 = 3.0%
F = 2.09
N = 36
manufacturing. Retailing is a multi-site business with a relatively small proportion of staff employed in central functions. The bulk of the work-force is dispersed at store level. Employment losses following an acquisition are most likely to occur in the relatively less-numerous head office and middlemanagement functions. Furthermore with some central support functions such as distribution and IT support often sub-contracted to third parties, any immediate job losses may occur outside of the immediate business boundaries. Employment would only be significantly affected if a major part of the store network were to be closed after acquisition. Given the stated motives for retail acquisitions, the mass disposal of store assets is unlikely. As the final stage in our analysis we consider whether the results obtained in the current study are unrepresentative of the performance and characteristics of takeovers in the retailing sector in general. As explained earlier in the paper, the sample selected for analysis in the current study is biased towards larger acquisitions. It is possible that this explains why the negative post-outcome performance reported in studies using security returns data is not repeated in the current study. In order to test this possibility we compare the post-bid security return performance of companies in the current sample with that of acquirers in a more comprehensive data set (reported in Burt & Limmack, 2001). This latter set consists of 216 acquisitions by U.K. quoted companies in the retail sector during the period 1977–1992. We use two alternative control benchmarks for measuring expected security returns, the first being the corresponding return on the FT All Share Index and the second being the return
172
STEVE BURT AND ROBIN LIMMACK
on size-matched portfolios.26 The analysis is somewhat restricted as we possess security returns for the full data set only for the two-year period following the bid month. However the aim of this analysis is not to compare the two alternative performance matrices but rather to determine whether the subsample of acquirers used in the current study has similar security performance characteristics to those in alternative studies. As the results in Table 11 demonstrate, the two-year post-bid excess returns for the companies in the current data set is negative (although not statistically significant), using both control models and whether measured using means or medians. The excess returns for the group of retail acquirers not included in the analysis of operating performance are significantly negative. However the excess returns for the latter sub-group are not significantly different from the returns for the sub-sample of acquiring companies analysed in the current study. We are therefore unable to conclude that the sub-set of retail acquirers used in the current study have superior performance to retail acquirers generally.
Table 11. Control-adjusted acquirer returns for the sub-sample of acquirers used in the current study compared to those on a more comprehensive set of returns on acquirers in the retailing sector, 1977–1992. Returns are measured over the period from bid month to two years thereafter, and adjusted relative to (a) Size-matched portfolios; (b) FT All Share Index. Size Decile Control
Market Relative
Observations
Mean %
Median %
Mean %
Median %
N
–10.48
–14.44
–8.34
–11.76
49
–14.76a
–16.5
–16.45a
–13.31
167
Operating Performance sub-sample: CAR (0,24) Remaining sub-sample: CAR (0,24) F (H) statistic
0.31
0.03
0.99
0.35
Probability
0.58
0.85
0.32
0.56
a
Significantly different from zero at the 1% level, using a two-tailed test.
The Operating Performance of Companies Involved in Acquisitions
173
5. SUMMARY AND CONCLUSIONS The current paper has examined operating performance of a sample of companies involved in acquisitions in the retail sector which were initiated and completed in the period January 1 1977 to December 31 1992. We use a measure of operating performance based on operating cash flows, similar to that used in Manson et al. (1994). Our results suggest that neither the pre-bid operating performance of the acquirer or the target is significantly different to that of their control companies. However we find that the post-acquisition operating performance of the combined company improves relative to a pre-bid control-adjusted benchmark. While this result is partly sensitive to the choice of control benchmark (i.e. the weighting used) we believe that the approach adopted here, and that by Healy et al. (1992), is the most appropriate. In addition to the above we find that acquirers and targets on average earn a lower operating margin on sales prior to the bid and that there is no change in control-adjusted profit margin post-bid. By contrast, however, we find that both acquirers and targets achieve a higher rate of sales turnover on total assets in the five years prior to the year of acquisition and that control-adjusted sales turnover improves significantly post-acquisition. We are unable to report any significant change in employment levels following acquisition, after controlling for other factors. The data requirements for the current study mean that the sample of acquisitions for which data was available are likely to be larger in absolute size than the full population of acquisitions, even in the retail sector. It is possible, therefore, that the results reported here are not representative of retail mergers in general. We provide one test of this by comparing the abnormal security returns to shareholders in this sub-sample of acquiring firms over two years post-acquisition to those of a larger sample of retail acquirers. Although the results provide some evidence that our sub-sample may perform less badly, when measured using security returns, than retail acquirers generally, the differences do not appear to be sufficient to invalidate our results. The conclusion of the study is that takeovers in the retailing sector lead to an improvement in operating performance over the five years post-acquisition that is mainly generated by improved asset utilisation. The results do not, of themselves, contradict those involving analysis of security returns. Rather they suggest that the benefits of the improved operating performance are likely to have been passed on to the owners of the target company, rather than to have been shared with the owners of the acquiring firm. One caveat to our results is the potential differential impact of operating leases on the companies involved. Given the data limitations in the current study we were unable to determine whether this would effect the above conclusions.
174
STEVE BURT AND ROBIN LIMMACK
NOTES 1. See for example Agrawal, Jaffe and Mandelker (1992), Gregory (1997). 2. See for example Meeks (1977), Singh (1971, 1975), Utton, (1974). 3. There is also evidence that the threat of takeover can force management to introduce changes that might not otherwise have occurred. Denis (1994) describes the leveraged buy-out of Safeway Stores Inc. in the USA, and subsequent restructuring activities, following an unsuccessful takeover bid from the Dart Group. 4. Or to be more precise, security returns behaviour. 5. See for example Agrawal, Jaffe and Mandelker (1992), Healy, Palepu and Ruback (1992), Manson, Stark and Thomas (1994) and Jarrell (1995). 6. Obviously excluding targets for the latter period. 7. Their results also suggested that target firm managers did not manipulate earnings (possibly because of the lack of warning prior to the bid). 8. Quoting from Chemical Week, May 8, 1991, p. 28. 9. This introduces a potential measurement error in our analysis because of the likelihood of off-balance sheet assets and financing. However the extent of the potential measurement error will be a function of the extent to which our sample are more likely to use this form of financing than the respective controls. Lack of publicly-available data over the relevant period prevents any formal test of this. 10. Although Barber and Lyon (1996) also advocate the use of operating income, rather than operating cash flows, they do also recognise that the former may be more appropriate in situations in which there is a potential for earnings management. 11. The exclusion of cash and marketable securities, although common practice in this type of analysis, may produce an understatement of assets used in retailing and hence to an overstatement of operating performance. 12. We select the firm that is closest in size (based on equity market value) from within the Retail Sector which is not itself an acquiring or acquired firm. 13. Healy et al. (1992) actually use industry median values, rather than matched firms. 14. Barber and Lyon (1996) also recognise that the power of performance matching decreases as the lagged interval increases as in the current study. 15. We test the sensitivity of our results to the choice of alternative deflators, as discussed later. 16. Healy et al. (1992) use industry groups as their chosen control performance and also weight the performance of each group by the relative asset value of acquirer and target at the end of year t 1. 17. Weights used are the relative operating assets of acquirer and target at the end of the last financial year preceding the year of the acquisition. 18. At the 10% level of significance. 19. When the post-acquisition performance of the control companies is weighted by their relative operating assets the results are even in stronger in favour of significantly positive post-acquisition control-adjusted operating performance. 20. To avoid problems associated with survivorship bias we take a median value over all years for which data is available. 21. See also Healy et al. (1992), Manson et al. (1994) and Ghosh (1998). 22. This analysis was undertaken following a suggestion by Alan Gregory.
The Operating Performance of Companies Involved in Acquisitions
175
23. The authors recognise helpful comments by Alan Goodacre on this issue. 24. As we use operating cash flows in the numerator, rather than operating profits, there would be no impact from an consequent reduction in the depreciation charge. 25. See, for example Baker (1992) and Denis (1994). 26. Further details of these may be obtained from Burt and Limmack (1999). 27. The results are similar if the weights are based on the net fixed assets of the control companies at the beginning of the relevant years.
REFERENCES Agrawal, A., Jaffe, J. F., & Mandelker, G. N. (1992). The post-merger performance of acquiring firms: A re-examination of an anomaly. The Journal of Finance, 67, 1605–1621. Anand, J., & Singh, H. (1997). Asset redeployment, acquisitions, and corporate strategy in declining industries. Strategic Management Journal, 18 (Summer Special Issue), 99–118. Appleyard, A. R. (1980). Takeovers: accounting policy, financial policy and the case against accounting measures of performance. Journal of Business Finance and Accounting, 7, 541–554. ASC (1984). Accounting for leases and hire purchase contracts. Statement of Standard Accounting Practice No. 21. London: Accounting Standards Committee. Baker, G. P. (1992). Beatrice: a study in the creation and destruction of value. The Journal of Finance, 47(3), 1081–1119. Barber. B. M., & Lyon, J. L. (1996). Detecting abnormal operating performance: the empirical and power specification of test statistics. Journal of Financial Economics, 41, 359–399. Beattie, V., Edwards, K., & Goodacre, A. (1998). The impact of constructive lease capitalisation on key accounting ratios. Accounting and Business Research, 28, 233–254. Bowen, R., Burgstahler, D., & Daley, L. (1986). Evidence on the relationship between earnings and various measures of cash flow. The Accounting Review, 65, 713–725. Burt, S., & Limmack, R. J. (2001). Takeovers and shareholder returns in the retail industry. International Review of Retail, Distribution and Consumer Research, 11, 1–21. Clark, K., & Ofek, E. (1994). Mergers as a means of restructuring distressed firms: an empirical investigation. Journal of Financial and Quantitative Analysis, 29(4), 541–565. Cosh, A., Hughes, A., & Singh, H. (1980). The causes and effects of takeovers in the U.K.: An empirical investigation for the late 1960s at the micro-economic level. In: D. C. Mueller (Ed.), Determinants and Effects of Mergers (pp. 227–270). Cambridge, Mass.: Oelgeschlager, Gunn and Hain. Dechow, P. M. (1994). Accounting earnings and cash flows as measures of firm performance: The role of accounting accruals. Journal of Accounting and Economics, 18, 3–42. Denis, D. J. (1994). Organizational form and the consequences of highly leveraged transactions. Kroger’s recapitalization and Safeway’s LBO. Journal of Financial Economics, 36, 193–224. Erickson, M., & Wang, S-W. (1999). Earnings management by acquiring firms in stock for stock mergers. Journal of Accounting and Economics, 27, 149–176. Franks, J., Harris, R., & Titman, S. (1991). The post-merger share price performance of acquiring firms. Journal of Financial Economics, 29, 81–96. Ghosh, A. (1998). Does accounting-based performance really improve following corporate acquisition? Working Paper. Zicklin School of Business, Baruch College, (CUNY), New York, USA.
176
STEVE BURT AND ROBIN LIMMACK
Gregory, A. (1997). An examination of the long run pattern of U.K. acquiring firms. Journal of Business Finance and Accounting, 7 & 8, 971–1002. Harford, J. (1999). Corporate cash reserves and acquisitions. Journal of Finance, 54 (December), 1969–1997. Healy, P. M., Palepu, K. G., & Ruback, R. S. (1992). Does corporate performance improve after mergers? Journal of Financial Economics, 31, 135–175. Holl, P., & Pickering, J. F. (1988). Determinants and effects of actual, abandoned and contested mergers. Managerial and Decision Economics, 9, 1–19. Jarrell, S. (1995). Long-term performance of corporate takeovers: an improved benchmark methodology. Working Paper. University of Chicago, Chicago, Illinois. Lev, B., & Mandelker, G. (1972). The microeconomic consequences of corporate mergers. Journal of Business, 45(1), 85–104. Loughran, T., & Ritter, J. R. (1997). The operating performance of firms conducting seasoned equity offerings. The Journal of Finance, 52(5), 1823–1850. Loughran, T., & Vijh, A. (1997). Do long-term shareholders benefit from corporate acquisitions? The Journal of Finance, 52(5). Manson, S., Stark, A. W., & Thomas, H. M. (1994). A cash flow analysis of the operational gains from takeovers. The Chartered Association of Certified Accountants, Research Report 35, CAET, London. Martin, K. J. (1996). The method of payment in corporate acquisitions, investment opportunities, and management ownership. The Journal of Finance, 51(4), 1227–1246. Meeks, G. (1977). Disappointing marriage: a study of the gains from mergers. Cambridge University Press, Occasional Paper 51. Mueller, D. C. (1980). The U.S. 1962–72. In: D. C. Mueller (Ed.), Determinants and Effects of Mergers (pp. 271–298). Cambridge, Mass, Oelgeschlager, Gunn and Hain. Ravenscraft, D., & Scherer, F. M. (1987). Mergers, sell-offs, and economic efficiency. Brookings Institution, Washington. Rayburn, J. (1986). The association of operating cash flow and accruals with security returns. Supplement to The Journal of Accounting Research, 24, pp. 112–133. Shleifer, A., & Vishny, R. W. (1997). A survey of corporate governance. The Journal of Finance, 52(2). Singh, A. (1971). Takeovers. Cambridge University Press. Singh, A. (1975). Takeovers, economic natural selection and the theory of the firm: evidence from the post-war U.K. experience. Economic Journal, 85(339), 497–515. Utton, M. A. (1974). On measuring the effects of industrial mergers. Scottish Journal of Political Economy, 21(1), 13–28.
SHAREHOLDER WEALTH EFFECTS OF DIVERSIFICATION STRATEGIES: A REVIEW OF RECENT LITERATURE Robin Limmack ABSTRACT Diversification strategies, in particular those associated with acquisition activity, offer companies an opportunity for growth in a less restricted environment than is available to companies pursuing a policy of focused growth. Not only are there likely to be fewer legal restrictions but also the choice of potential target company is much greater. However easier access to growth through diversification also provides more opportunity for managerial opportunism and the pursuit of policies that are unlikely to benefit shareholders. In addition to the standard agency problems there are likely to be additional problems associated with cross-subsidisation of inefficient segments. Potential benefits from diversification strategies include increased debt capacity together with creation of internal capital markets. Whether benefits outweigh costs is an empirical issue. As with many such issues, however, the empirical evidence is not unambiguous. While a number of studies have suggested that diversified firms generally trade at a discount to non-diversified firms, it appears for some periods of time the benefits of diversification have outweighed the costs. Additionally it appears that companies who choose the diversification route may come from industries that were experiencing problems and that any ‘discount’ may be present in these companies before they make
Advances in Mergers and Acquisitions, Volume 2, pages 177–205. © 2003 Published by Elsevier Science Ltd. ISBN: 0-7623-1003-0
177
178
ROBIN LIMMACK
the diversification decision. It may be that for these companies at least the diversification decision is not the source of the value loss but the result. The current paper provides a review of contemporary literature on the perceived benefits and costs of diversification strategy and of the empirical evidence surrounding this. While no definitive conclusions can be reached from the evidence, it does appear that an assumption that diversification is a value-reducing activity in all circumstances is unwarranted.
1. INTRODUCTION One of the many strategic issues facing corporate management relates to the degree of specialisation that their firm should attempt to achieve. Companies have varied in the extent to which they have followed a focused rather than a diversified strategy. The focusing decision may be related either to a particular product or service or indeed to the geographical market in which the firm attempts to operate. Examples exist of companies that have at various times adopted different strategies, at one time diversifying then later refocusing.1 Changes in strategy may reflect the time varying nature of the relative benefits and costs or may simply indicate the correction of previous errors in strategy. In the current paper we will examine both the claimed benefits and costs of diversification (and refocusing strategies) and the evidence provided by empirical research on the impact on shareholders’ wealth. As with many corporate strategic decisions these benefits and costs will impact on both shareholders and management. However the impact on the latter group is often unobservable except in relation to remuneration packages. Our review is therefore confined mainly to a discussion of shareholder wealth effects. Although diversification may be achieved either through internal growth or external acquisition much of the published literature on this topic concentrates on diversification through external acquisition. There are a number of different reasons why acquisition may be the preferred motive for diversification, including speed, tax motives, and the use of equity financing. Our review will therefore also include reference to literature on the relative merits of related versus unrelated acquisitions. In addition, the continued survival of diversified firms may be seen as evidence of the efficiency of that particular organisational form, although Jensen (1989) suggests that survival may be more dependent on the high costs of corporate control transactions. Evidence from valueincreasing refocusing activities in recent years does, however, suggest that these latter costs may be worth bearing in some circumstances at least. Section 2 provides an overview of the motives for diversification while section 3 examines in more detail those motives that are related to shareholders
Shareholder Wealth Effects of Diversification Strategies
179
interests. The view adopted by many commentators is that the diversification decision is potentially highly correlated with the pursuit of management objectives. We therefore proceed in Section 4 to a review of agency problems and diversification. In Section 5 we explore issues raised in the literature relating to the methodological approaches to the measurement of diversification benefits and costs. In the penultimate section we provide a review of literature on refocusing activities and end this review with a brief summary of what we appear to know (and what we do not know) about the benefits, or otherwise, of diversification.
2. CLASSIFICATION OF MOTIVES FOR DIVERSIFICATION Montgomery (1994) provides a threefold classification for considering motives for diversification. The first set of motives, based on market-power, views a diversification strategy as a means of exploiting ‘conglomerate power.’ Conglomerate power refers to the opportunity for a firm to exercise power in a particular market by ‘virtue of the scope and character of its activities elsewhere’ (Edwards, 1955). Exercise of conglomerate power may be achieved through cross-subsidisation, for example to support predatory pricing, mutual forbearance, to reduce the degree of competitiveness, and reciprocal buying, in order to deny markets to smaller competitors.2 In many countries the regulatory framework of the corporate sector includes provisions to limit opportunities for the exercise of conglomerate power. Economists also differ as to whether market power may be enhanced by greater concentration rather than through diversification. The second set of motives, based on resource utilisation, argues that rentseeking firms diversify in response to excess productive capacities (Penrose, 1959). Penrose articulates a theory that assumes firms will not achieve equilibrium positions because of problems caused by indivisibility of resources, alternative uses of resources in different circumstances, and the creation of new productive services. Within this theory firms have incentives to expand, although not necessarily through diversification. If resources are fairly specific then expansions should be in related areas. With less-specific resources benefits may also be obtained through diversification (Montgomery & Wernerfelt, 1988). In a similar vein Drucker (1981) suggests that economically sensible mergers must follow a certain set of rules which include the requirement for a ‘common core of unity’. This common core may include tangible factors (Weston, 1986) or intangibles, including the corporate culture (Barney, 1988)
180
ROBIN LIMMACK
The final set of motives, based on agency relationships, views the diversification decision as a convenient means for management to pursue their own interests rather than those of the owners of the firm (Mueller, 1969; Jensen, 1986; Shleifer & Vishny, 1989). Problems arise according to this perspective from use of free cash flow in over-investment (Jensen, 1986), increasing management entrenchment (Shleifer & Vishny, 1989), and diversification of risks associated with human capital (Amihud & Lev, 1981). The above framework is not the only possible classification of diversification motives (see also Cable, 1977). Nevertheless it provides a useful framework for separation of those motives that are most likely to be aligned with shareholder wealth maximisation from those that are not. It therefore provides a useful framework for discussion of relevant research in the remainder of this paper.
3. EXPLORING MARKET POWER AND RESOURCE UTILISATION MOTIVES 3.1. Market Power Motives Motives that are related to the exercise of market power may also be consistent either with goals that consider shareholder interests or with those that favour management interests, although they may be inconsistent with the interests of other potential stakeholder group. Evidence relating to increased market power through diversification is relatively sparse. In addition it may be unwise to infer motive from the results of such actions, although the later is generally all that researchers are able to measure. A number of studies have attempted to distinguish between market power and efficiency arguments for related acquisitions, but with only limited success.3 While Eckbo (1983) identified a positive relationship between the merger-induced change in concentration ratios and bidders abnormal returns this was not, of itself, sufficient to distinguish between the two sets of motives. Even evidence of a positive share price reaction for the competitors of the acquiring firm was not considered to be sufficient evidence to distinguish between the two sets of motives as Eckbo argued that the acquisition may convey information to investors about the potential for general efficiency improvements. Eckbo (1985) did, however, suggest that examination of the share price reaction of competitors to the instigation of an anti-trust complaint over a proposed related acquisition may distinguish between the two sets of motives. In this latter instance a negative share price response was viewed as evidence that the potential for increased collusion had disappeared. As Eckbo (1985) was unable to identify a negative response he concluded that there was
Shareholder Wealth Effects of Diversification Strategies
181
no evidence for collusion. However, as Limmack and McGregor (1995) suggest, a negative reaction to competitors may also be a response to the likely challenge to any further acquisitions. In addition, a positive response to the anti-trust complaint may reflect a larger anticipated gain to competitors through the loss of market share by the original bidder (Huth & Macdonald, 1989). While the above refers to related acquisitions, a number of studies have identified a negative relationship between diversification and corporate performance, rather than the anticipated positive relationship that may be expected if the market power motive dominated.4 While these results are suggestive of the view that market power is not the dominant motive they do not rule out the possibility that this motive may also be present together with managerial motives, with the impact from the latter dominating the eventual performance of the company. 3.2. Resource Utilisation (Efficiency) Motives Weston et al. (1998) suggested that the diversification decision should be made within a strategic framework that is based on examination of organisational strengths and weaknesses. Anslinger and Copeland (1996), for example, identified intensive acquirers who consistently outperformed the S&P 500 index over the period 1985–1994 while Weston (1999) illustrated the strategy behind the acquisition policy of the General Electric Company in making 509 acquisitions and 310 divestitures over the period 1981–1997. Prahad and Bettis (1986) have examined the issue of related vs. unrelated diversification and conclude that it is the selection of an appropriate strategy rather than diversification per se that will produce superior performance. Pandaya and Rao (1998) suggest that ‘dominant firms operating with core competencies and operating in less competitive environments are better off concentrating on one business segment’ (p. 76). Other firms, however, may be better off diversifying. One motive, in particular, often claimed for diversifying acquisitions is the utilisation of skills developed by one management team to the control of assets currently managed by a less efficient team. While Jensen and Ruback (1983) refer to the application of this ‘market for corporate control’ as a disciplinary mechanism for acquisitions in general, firms following a diversified strategy will be able to consider a greater number of opportunities for the application of their perceived superior skills. By contrast there will also be a greater opportunity to exercise poor judgement or ‘hubris’ (Roll, 1986) and the potential for over-valuation based on asymmetric information may also be greater (Chatterjee & Lubatkin, 1990).5 Superior management skills may either
182
ROBIN LIMMACK
be those developed in specific management functions or in general management capabilities. One example of application of the latter may be found in the acquisition strategy adopted by Beatrice Company (USA) in the 1960s. According to Gazel (1990) the most important aspect to the management of Beatrice of any acquisitions in that period was ‘the quality of the company’s incumbent management’ rather than the acquisition of inefficiently managed firms. The acquisition of relatively small, well-managed companies allowed the management of Beatrice to apply their existing decentralised management structure to the newly acquired organisation and to concentrate senior management resources on strategic planning and hands-off monitoring (Baker, 1992). It is interesting to note that one of the reasons identified by Baker for the subsequent failure of Beatrice was a movement away from this particular organisational structure towards one with a more centralised decision-making structure. Matsusaka (1993) similarly found that firms making diversifying acquisitions in the period 1968 to 1974, but who also retained the management of the target company, achieved positive abnormal returns.6 The resource-based theory, developed by Penrose (1959) among others, is perhaps a more generalised form of the above in which excess capacity in valuable resources (including management ability) is transferable across industries and leading to situations in which the diversified firm may emerge as the most efficient form of organisation. While the resource-based motive for diversification expressed above appears to focus mainly on economies of scale in the utilisation of various resources,7 other potential sources of efficiency gain have also been identified for diversifying activities including the following: • Increased debt capacity (Lewellen, 1971). • Utilisation of tax losses (Levy & Sarnat, 1994). • Reduction of information asymmetries in capital markets (Alchian, 1969; Williamson, 1970; Weston, 1970; Stein, 1997; Hubbard & Palia, 1999). • Reduction in the under-investment problem (Stultz, 1990). Potential costs include: • Over-investment (Stultz, 1990; Berger & Ofek, 1995). • Subsidisation of loss-making segments (Meyer et al., 1992; Berger & Ofek, 1995). According to Lewellen (1971) increased debt capacity may arise from diversification through the combination of companies with imperfectly correlated cash flow streams.8 The increased debt capacity then provides the
Shareholder Wealth Effects of Diversification Strategies
183
opportunity for greater utilisation of tax shields from interest payments and hence higher shareholder wealth. Asquith and Kim (1982), however, found that, while target shareholders obtain gains from diversifying acquisitions, it was not at the expense of bondholders. Shleifer and Vishny (1992) suggested that debt capacity may also be increased in conglomerates that are more likely to be able to identify marketable assets for disposal when faced with liquidity problems. By contrast to the above Mansi and Reeb (2001) suggested that wealth losses to shareholders in diversified firms are related to the level of debt financing with all-equity firms suffering no wealth losses. The latter authors suggested that diversification involves a transfer of wealth from equity investors to bondholders. In addition a recent paper by Lamont and Polk (2001) suggests that, rather than reducing risk, diversified firms appear to have a higher variability of expected future cash flows and expected future returns than portfolios of single segment firms. Porter (1985) also suggested that related acquisitions are more likely to reduce systematic risk through the increased ability of the merged firm to defend its market position. Possible tax gains from acquisition include the use of otherwise underutilised tax losses. Where one company has experienced taxable losses that are unlikely to be offset against taxable profits in the foreseeable future, combination with another firm that is currently earning taxable profits will provide an increase in shareholder wealth that is equivalent to the present value of the tax saved.9 While such wealth increases may not be restricted to unrelated acquisitions tax savings are more likely to be obtained through the combination of profit streams with a relatively low correlation. Levy and Sarnat (1994) refer to research by Blume et al. (1984) in which tax considerations feature among the main reasons for corporate acquisition. Evidence on the tax benefits of diversification is however sparse and while Berger and Ofek (1995) report some evidence of tax savings from diversification, the amount involved was considered to be too small to have any significant impact on firm value. The utilisation of more efficient internal capital markets to eliminate problems caused by information asymmetries has been cited by a number of authors as a motive for diversified acquisition, both in the USA during the 1960s and, more recently, in relatively under-developed economies. Alchian (1969) and Williamson (1970, 1975), among others, have suggested that as managers of a firm possess information advantages over external providers of capital, internal capital markets may produce a more efficient allocation of resources than is possible with external markets. Myers and Majluf (1984) refer to the under-investment problem that can result from information asymmetry although Stein (1997) also suggests that similar asymmetries may be present in large diversified firms. Matsusaka and Nanda (2000) also suggested that
184
ROBIN LIMMACK
internal markets might be a cheaper source of finance in the presence of high external financing costs. In particular they argued that ‘internal flexibility allows them to avoid costly external financing in more states of the world than separated firms’ (ibid p. 30). Against this, however, must be set the potential disciplinary power of the external capital market and the potential overinvestment problem when access is available to internal financing. Hubbard and Palia (1999) reported that higher returns were available for diversifying acquirers in the 1960s when financially unconstrained acquirers bought constrained targets.10 In addition they identified that the management of target firms were generally retained, suggesting that the motives for acquisition did not include replacement of an inefficient management team. Their results are consistent with those reported by Baker (1992) who includes in his paper an analysis of the source of diversification benefits for Beatrice Company. Support for the benefits of an internal market is also provided in the study by Khanna and Palepu (2000) of the financial performance of affiliates of diversified Indian companies.11 While they report that accounting and stock market measures of performance initially declined with diversification, this pattern was reversed at higher levels of diversification. It appears that in less-well developed economies there may still be benefits from the presence of an internal market in finance and management expertise. Further support for the benefits of an internal capital market is also provided in a recently reported study by Klein (2001) of the valuation of conglomerate acquirers over the period 1966–1974. For the earlier period studied Klein reports a diversification premium that becomes a discount in the later period. The higher valuation in the earlier period is ascribed to the benefits of internal markets. Once the relative advantages of internal markets disappear, however, the costs of diversification will outweigh the benefits (Bhide, 1990; Stultz, 1990). A recent paper by Billet and Mauer (2000) suggested that benefits of an internal market may have persisted for some U.S. companies into the 1990s. The authors investigated stock market reaction to the announcement by a sample of U.S. companies of the introduction of tracking stock into their capital structure. Tracking stock is a form of equity capital that has been designed to be linked to the performance of a particular line of business within a company.12 The use of such stock allows different rights to be assigned to ownership of different lines of business and also allows managerial compensation to be linked to the performance of the line of business tied to the particular tracking stock. The use of tracking stock allows a company to continue to operate an internal capital market, by contrast with other forms of corporate restructuring. From their analysis Billet and Mauer (2000) identified that the announcement of tracking stock issues was seen as a positive event by the stock
Shareholder Wealth Effects of Diversification Strategies
185
market and that diversified companies that issue tracking stock are valued at a lower discount than other diversified companies.13 The authors concluded that, for some companies at least, there are still benefits to be provided from internal capital markets. However as the authors are only able to identify a small number of companies that use tracking stock they suggested that ‘ a drive toward focus . . . is probably more value-enhancing for the vast majority of diversified firms’ (ibid p. 1489). Berger and Ofek (1995) reported that one of the sources of value-loss for diversified companies was over-investment in poorly performing segments. Their proxy for over-investment was based on the sum of depreciation-adjusted capital expenditures, scaled by total assets, for segments operating in industries whose median Tobin’s q is in the lowest quartile. Cross-subsidisation of poorly performing subsidiaries was also identified as a potential source of value loss in diversified firms. While the maximum loss suffered for a stand-alone company is restricted to the total value of that unit because of limited liability, within a diversified organisation the loss may also reduce the value of other segments. Myers et al. (1992) also reported greater value losses through crosssubsidisation of unprofitable lines of business. Similarly Berger and Ofek (1995) also report results that are consistent with the view that diversified firms suffer a significant value-loss through cross-subsidisation. In the latter case, however, the authors also acknowledged that their results were also consistent with the view that the security market interprets a negative cash flow in one or more segments as a signal of poor management quality. Further evidence of inefficient transfer of resources across divisions was supplied in recent papers by Lamont (1997), Shin and Stultz (1998), and Rajan, Servaes and Zingales (2000). A number of studies have also suggested that power struggles amongst managers of divisions of diversified firms leads to a sub-optimal allocation of resources (Rajan et al., 2000; Scharstein & Stein, 2000). However Whited (2001) suggested that prior evidence of inefficient investment by divisions of conglomerates was a consequence of methodological error in measurement of q values. Using generalised method of moment estimators (Hansen, 1982) Whited reports that any evidence of cross-subsidisation of inefficient divisions in diversified firms disappears. Published research of the impact of diversification on shareholders wealth suggests that many of the claimed benefits may not have materialised or else may have been more than offset by greater agency costs. Indeed recent years have produced a reversal of diversification strategies with many firms following a more focused approach. One possible reason for the diversification discount and subsequent reversal in strategy is that ‘shocks’ in competitive and regulatory conditions occurred in the 1980s and reduced the previously
186
ROBIN LIMMACK
obtainable benefits from diversification. Potential shocks include changes in the external capital markets, technological change, and in the regulation of competition and industrial concentration. Support for the impact of shocks on the relative merits of focused and diversified strategies is provided in literature on the benefits of internal vs. external markets (Baker, 1992; Hubbard & Palia, 1999), limited availability of superior management skills (Baker, 1992), and changes in the regulatory environment (Shleifer & Vishny, 1991). Khanna and Tice (2001) document the response of firms to the specific ‘shock’ induced by the entry of Wal-Mart into their market. The authors find that diversified firms appear to respond faster to the threat, either by exiting the discount market or by instituting a campaign against the new entrant. Khanna and Tice (2001) also reported that diversified firms were more sensitive to the productivity of their discount division than were more focused firms and that internal markets transferred funds efficiently to more profitable sectors. Contrary evidence to that reported above is found in those studies that have identified the presence of a diversification discount in the 1960s and 1970s (Lang & Stultz, 1994; Servaes, 1996) and those that have suggested that diversified firms react slowly, and generally only in response to externally imposed pressure, to diversification costs (Berger & Ofek, 1996; Dennis, Denis & Sarin, 1997). In addition Chevalier (2000) reported that the investment behaviour documented in diversified firms was present prior to the absorption of the individual firms rather than being a function of diversification per se. In the next section we examine in more detail the impact of agency issues on the diversification decision.
4. DIVERSIFICATION AND AGENCY ISSUES According to a number of commentators diversification strategies provide clear examples of the agency problems described by Jensen and Meckling (1976), among others. In particular diversification provides greater opportunity for growth and managerial labour capital risk reduction than is likely to be available in more focused strategies. Growth opportunities may be followed without many of the restrictions imposed by competition within a primary industry or the threat of government legislation.14 In addition diversification offers the opportunity for senior managers to reduce the risk to their own personal labour capital (Amihud & Lev, 1981; May, 1995) although Aggarwal and Samwick (2001) find little evidence that mangers diversify their firms to reduce their risk exposure. Jensen and Murphy (1990) suggest that diversification may provide greater opportunity for managers to increase their power and prestige as well as their
Shareholder Wealth Effects of Diversification Strategies
187
level of remuneration. Hoskisson and Hitt (1990) support this argument by pointing to the high correlation between diversification, firm size, and management remuneration. Avery et al. (1998) report that the main benefit to CEO’s from acquisition is the opportunity to gain outside directorships, rather than increased compensation. This result holds irrespective of whether the acquisition increased shareholder wealth or was diversifying in nature. They conclude that acquisitions increase the prestige and standing of CEO’s in the business community. Private benefits obtained from managing a more diversified firm may also include increased career prospects (Gibbons & Murphy, 1992) and increased prestige from dealing with the greater complexity (Jensen, 1986; Stultz, 1990). In addition Shleifer and Vishny (1989) suggest that managers may become more indispensable (entrenched) in diversified firms and may therefore undertake unrelated acquisitions when their position is threatened. Denis, Denis and Sarin (1997) identified a strong negative relationship between the degree of diversification and inside share ownership based on analysis of a sample of U.S. companies in 1984. However their results also found ‘little support for the hypothesis that the value loss from diversification is greater in firms with low managerial ownership’ (ibid p. 144). Their results appear to be inconsistent with the finding by Servaes (1996) that diversification increased in the 1970s among firms with high inside ownership ‘when the costs to shareholders was negligible’ (ibid p. 1203). The result is also inconsistent with the finding by Lewellen, Loderer and Rosenfeld (1989) who reported that diversifying acquisitions led to higher abnormal returns in firms with higher management share ownership. Rajan, Servaes and Zingales (2000) demonstrated that in diversified firms resources flow to inefficient investments to the detriment of investment in other, more profitable, divisions. Similarly Scharfstein and Stein (2000) identified misallocation of resources by divisional management. As reported earlier, however, Whited (2001) suggested that the relationship between the diversification discount and inefficient investment may have been produced as a result of measurement error in the use of Tobin’s q as a proxy for investment opportunities. After correction for the measurement error Whited finds no evidence of inefficient allocation of investment funds.15 Aggarwal and Samwick (2001) tested a model that combines both the risk reduction and private benefits explanation for diversification. They initially reported that firm performance increased with incentives but decreased with the level of diversification. However when they controlled for firm-specific features they then found that managers’ incentives were positively related to diversification in contrast to reported findings of earlier papers (Denis et al.,
188
ROBIN LIMMACK
1997). In addition the authors concluded that ‘diversification decisions are driven by the private benefits managers receive from greater diversification’ (op cit p. 26). Aggarwal and Samwick did however reject the inference by May (1995) that the positive relationship between diversification and incentives provided support for a risk-reducing motive for diversification. Instead they suggested that managers with the greatest need for risk reduction would have the lowest incentives and that there should in those circumstances be a negative relationship between the two. However in both their paper and that by May (1995) a positive relationship had been identified. Hence Aggarwal and Samwick (2001) concluded that management risk reduction was not a primary motive for diversification.
5. MEASUREMENT ISSUES Examination of diversification effects involves decisions relating to at least two measurement issues. The first involves the measurement of the degree of diversification while the second relates to performance issues. On the first issue most studies have tended to separately identify diversified vs. undiversified firms based either on the number of separate segments identified in financial reports or by reference to identified Standard Industrial Classification (SIC) codes. Classification errors may arise through the absence of segmental information.16 Alternative measures of diversification include the number of reporting segments (Berger & Ofek, 1995; Servaes, 1996) and measures of concentration such as the Herfindahl index (Berger & Ofek, 1995) or the Specialisation ratio (Rumelt, 1974, 1982; Pandya & Rao, 1998). Studies of the impact of diversification on firm performance have generally used accounting measures and/or security market measures. Early research into the effects of diversifying acquisitions used measures of accounting performance. Thus Reid (1968) reported that conglomerate acquirers had higher levels of profitability than other types of acquirer. Weston and Mansingka, (1971) found that conglomerate firms had lower operating profit margins during the early part of the 1960s, but that profitability improved during the later half of that decade. The suggestion that the improvements in performance reported by them may have been an artifact of some form of profit averaging, in the acquisition of more profitable targets (Mueller, 1977, p. 322) was rejected in papers by Conn (1973) and Melicher and Rush (1973). Conn found no difference in profitability between acquirer and acquired firm for conglomerate acquisitions made in the period 1954 to 1969. Melicher and Rush (1973), found that for acquisitions conducted in the period 1960 to 1969 conglomerate acquirers were less profitable than non-conglomerates, pre-acquisition, but that
Shareholder Wealth Effects of Diversification Strategies
189
there was no significant difference in pre-acquisition profitability of the two groups of target firms. Improvements in profitability did not therefore appear to come from acquiring more profitable targets. In a more recent study of the impact of diversification on U.S. company performance, Pandya and Rao (1998) found that risk-adjusted accounting performance increased with the degree of diversification. For U.K. acquisitions in the period 1964–1971, Meeks (1977) reported greater improvements in profitability for non-horizontal acquirers relative to horizontal acquirers. By contrast Cosh, Hughes and Singh (1980) identified a post-acquisition decline in the performance of unrelated acquisitions. More recently much of the research into the wealth effects of diversification strategies has adopted one of two methodological approaches based on analysis of market prices. The first approach uses an ‘events’ framework and is concerned with measuring the wealth effects for the acquiring firm shareholders of the decision to make a diversifying acquisition. In this type of study it is common to find comparison made with the wealth effects of related acquisitions, sometimes using multi-variate analysis. Thus Elgers and Clark (1980) concluded that diversifying acquirers earned higher returns than related acquirers. Similarly Matsusaka (1993) reported that in the years 1968, 1971, and 1974 acquisition announcements by diversified bidders produced positive abnormal returns. By contrast Morck, Shleifer and Vishny (1990) found no difference in returns in related and unrelated acquisitions in the 1970s but identify negative abnomal returns for acquirers of unrelated targets in the 1980s, while Kaplan and Weisbach (1992) reported mixed returns and Hyland (1999) identified positive returns to unrelated acquisitions. Limmack and McGregor (1995) found no significant difference in returns to U.K. bidders in related versus unrelated acquisitions in the period 1977–1986 although they did identify a positive relationship between the change in concentration ratio and bidding company returns in related acquisitions. Gregory (1977), however, reported that for U.K. acquisitions in the period 1984 to 1992 conglomerate acquirers achieved lower returns than non-conglomerate acquirers, although he also noted that his results were sensitive to the choice of benchmark. More recently Pandaya and Rao (1998) report that both absolute and risk-adjusted monthly returns are positively related to the degree of diversification for U.S. companies over the period 1984–1990. The latter authors use the Sharpe Index as their measure of risk adjustment and make no distinction between related and unrelated diversification. One problem with the above approach relates to the identification of the relevant event period. In particular conglomerates may be formed in the process of a programme of acquisitions that has been previously announced by the
190
ROBIN LIMMACK
company. Schipper and Thompson (1983) suggested that the expected value of an acquisition programme should be reflected in the share price of the acquirer around the time of the programme announcement. Share price reaction to individual takeovers subsequent to that event will reflect the incremental effect of each acquisition.17 They examined the share price behaviour of 30 companies that announced acquisition programmes during the period 1953–1968. While there was some ambiguity about the exact date of announcement of the programme for a number of these companies, Schipper and Thompson nevertheless reported significantly positive abnormal returns in the announcement month. Although their study was not explicitly concerned with conglomerate acquirers many of those companies most active in the takeover market are likely to be involved in unrelated (diversifying) acquisitions. An alternative approach to the above has been to compare the value of diversified companies at specific points in time with the value of stand-alone equivalents. This approach first requires the establishment of a relationship between market value and financial characteristics of non-diversified companies and subsequently, the application of that relationship to observable segmental information of the diversified companies. A discount (or premium) is then identified for the diversified company based on the difference between the actual market value and the combined value based on the imputed value of individual segments. The approach is limited by the availability of segmental information for the diversified firm and is generally based on quasi valuation ratios such as market value to sales, total assets or profits. In one of the first studies to use the valuation approach Lang and Stultz (1994) calculated Tobin’s q values for diversified firms and compared these with the combined stand-alone equivalent value of the separate divisions.18 They found that Tobin’s q for diversified firms was lower than that for singleindustry firms in the late 1970s and 1980s. Although adjustment for industry factors reduced the relative discount on diversified firms, it did not eliminate the discount completely. The authors also reported that their result held even after controlling for other factors that may explain cross-sectional differences in q values, including size, access to capital markets and intensity of research and development expenditure.19 Lang and Stulz (1994) also reported that the q value falls as the level of diversification increases. Firms that diversify also appear to have poor performance relative to other firms that do not, but that they are not significant under-performers relative to non-diversifying firms in the same industry. The evidence provided by Lang and Stultz (1994), while indicating that diversified firms are valued at a discount relative to nondiversified firms also suggests that the reasons for the low valuation may in part
Shareholder Wealth Effects of Diversification Strategies
191
be a function of the industry and other conditions in which these firms operate. The authors are cautious about inferring from their analysis that diversification ‘hurts performance’. Rather they conclude that their evidence suggests that diversification ‘is not a successful path to higher performance’ (ibid p. 1278). One possible explanation proposed for their results is that firms seek growth through diversification because they have exhausted profitable growth opportunities in their existing activities. Recent research (discussed in the next section) has explored a number of the issues raised by Lang and Stultz. Using a similar methodology to the above, Berger and Ofek (1995) also tested for the presence of a diversification discount. Their approach was based on the identification of industry segments for diversified firms and the calculation of median ratio for single segment firms in the same industry based on the relationship between total value and (separately) one of three accounting variables, namely assets, sales and earnings.20 The valuation ratio computed is then applied to the relevant segmental information of the diversified firm. The sum of these segmental values is identified as the ‘pseudo’ stand-alone value of the diversified company. The natural log of the firms actual value to its imputed value is then identified as the valuation premium (or discount). Using a sample of companies from the Compustat Industry Segment database for 1986 to 1991, Berger and Ofek (ibid) reported significant discounts in the median (and mean) value of diversified companies, whichever valuation multiple was used. They also reported that the value loss increased with the number of segments (ibid p. 49). Berger and Ofek also distinguished between related (at the SIC twodigit level) and unrelated diversification and found that the value loss was reduced in the former. Servaes (1996) adopted a similar methodology to the above in testing for the presence of a diversification discount in the period 1961–1976. However his analysis was made more difficult by the absence of disclosed segmental information for many of the companies in his sample. Servaes solves this problem by calculating Q values for multi-segment firms based on two alternative assumptions.21 First he calculated industry-adjusted values after deducting the mean (median) of the q-ratio of single segment companies in the industry identified as the primary industry for each firm. Second he calculated the equally weighted industry-adjusted value after deducting the mean (median) of the q-ratio of single segment companies in each segment separately. The first approach may give excess weight to the primary industry of the diversified firm while the second approach may give undue weight to other segments. An average of the two should therefore reduce the impact of the individual sources of measurement error. Servaes identified a significant mean and median industry-adjusted diversification discount to the firms in his
192
ROBIN LIMMACK
sample in 1961, 1967, and 1970 whichever method of adjustment was applied. In 1964 the discount was found to be negative but insignificantly different from zero. His results suggested, however, that ‘there is little evidence of a discount penalty’ for 1973 and 1976.22 Overall Servaes found that the diversification discount had reduced significantly from the period 1961–1970 to 1973–1976. While this pattern is contrary to what might be expected if diversification was motivated by the comparative advantage of internal capital markets, Servaes suggested an alternative explanation based on changes in the level of insider ownership. In the first time period examined diversified firms had lower inside share ownership than single segment firms. However in the later time period the difference in inside share ownership between the two groups was insignificant. Servaes concluded that in the later periods firms choose to diversify when there was no financial penalty. In the earlier years, however, the lower level of inside ownership did not prevent firms from diversifying even when there was a significant penalty. By contrast with the results from Berger and Ofek (1995), Servaes found no evidence that the discount increased with greater diversification beyond two segments. He was also unable to find any explanation for the diversification discount in differences in profitability, capital structure or investment policy. Nor did it appear from the above analysis that the diversifying decision was caused by lower prior profitability. His results also conflict with those reported by Klein (2001) on a restricted sample of 39 conglomerate acquirers over the period 1966–1974. While Klein reported a diversification premium for the period 1966–1968 he found that this turned into a discount in the later period of his study. Klein ascribed his results to the change in benefits from internal markets. Other papers have suggested that there may be exogenous factors that lead to the decision to diversify and that any observed discount (or premium) is not caused by the diversification decision alone. Research into the latter hypothesis is reviewed in the next section.
6. EXOGENOUS FACTORS AND DIVERSIFICATION DISCOUNTS A number of recent papers (Burch et al., 2000; Campa & Kedia, 2000; Villalonga, 2000; Graham et al., 2001) have suggested that the diversification decision ought to be considered as related to the economic environment in which the decision is made, rather than as a completely endogenous decision. In particular this approach requires controlling for the characteristics of the diversifying firms, or their industries, and possibly also of the firms that are acquired. The argument is that firms do not diversify at random but in response
Shareholder Wealth Effects of Diversification Strategies
193
to various factors relating either to their own characteristics, those of the target companies, or to the economic environment in which they operate. Burch, Nanda and Narayan (2000) provide a framework in which the diversification decision is viewed as a value-increasing response to industry conditions that had produced a discount in the value of the firm prior to the diversification decision. Their model is based in an agency setting in which managers in a declining industry are reluctant to reduce assets under their control and instead diversify into more profitable sectors.23 The firms that are most likely to diversify are those that are described as ‘less innovative’ in the sense that they are less able to cope with changes to their primary industry, whether caused by new technology, marketing and distribution patterns, or regulatory framework. Such firms are likely to be valued at a discount to their more innovative peers prior to any diversification decision. All other things being equal, the decision of such firms to diversify will be inversely related to the availability of profitable investment opportunities within their primary industry. Burch et al. (2000) find support for their hypotheses using panel data of fifty of the largest U.S. industries over the period 1978–1997. In particular industry market-to-book ratios and excess value measures are lowest in those industries with a higher proportion of firms that have chosen to diversify. Support for their hypotheses is also present in the finding reported earlier by Lang and Stultz (1994) that diversified firms previously operated in poorly performing industries. Similarly in a study of U.S. corporate performance in the 1950s Weston and Mansingka (1971) concluded that conglomerates were achieving a policy of ‘defensive diversification’ that was designed both ‘to preserve the values of ongoing organisations as well as restoring the earnings power of the entities’ (ibid p. 928). Villalonga (2000) also reported that diversifying firms trade at a significant industry-adjusted discount prior to the diversifying decision and that these firms were present in industries with lower q values than their non-diversifying counterparts. Institutional and inside share ownership was also found to be lower in diversifying firms. Villalonga also reports that diversifying firms had higher risk and were more likely to diversify into industries with lower leverage. Campa and Kedia (2000) suggested that it was the characteristics of diversifying firms that produces the diversification discount, not the decision to diversify per se. They began their analysis by comparing firm characteristics for diversified firms in years in which they were not diversified (single segment years) with the characteristics of firms in other samples. Using a data set of companies from the Compustat database for the period 1978–1996 the authors reported that in their single segment years conglomerate companies were likely
194
ROBIN LIMMACK
to be larger, have greater leverage and have lower research and development expenditure than multiple segment firms.24 In addition single segment years of diversified firms had higher capital expenditure rates and lower sales margins than single segment years of refocusing firms. Diversified firms in multisegment years tended to invest more in research and development and to have higher rates of capital expenditure and higher profitability than multiple segment years of refocusing firms.25 However diversified firms that do not change their status seem to be in mature industries with lower growth rates, lower research and development but higher profitability and, perhaps surprisingly, higher rates of capital investment. As stated earlier the authors concluded that it was the characteristics of the diversified firms that appeared to account for the discount rather than diversification itself. Campa and Kedia (2000) also found that there was a higher exit rate of single segment firms from industries in which diversified firms operate and that the exiting firms generally had lower values than remaining firms. A benchmark based on average returns of surviving firms is therefore likely to be biased. After controlling for exogenous factors that appear to predict the likelihood that a firm will diversify the authors reported that the diversification ‘discount’ becomes a premium. Their evidence therefore suggests that diversification may be a value increasing decision that results from changes to the economic environment in which diversifying firms operate. A number of authors, including Hyland (1999) and Villalonga (2000) have also suggested that the discount in firm value was present prior to the decision to diversify. Matsusaka (2001), however, appears to go beyond Campa and Kedia (2000) above by suggesting that the causal relationship is such that the presence of a discount in potential targets provides the reason for making diversifying acquisitions. This view is consistent with the theory that takeovers provide a disciplinary mechanism for the management of inefficiently performing firms (Jensen & Ruback, 1983). Graham, Lemmon and Wolf (2001) compared value changes for acquiring firms making related and unrelated acquisitions. They adopted two measures of value change, the first based on standard event study methodology and the second based on the excess value multipliers used by Berger and Ofek (1995), among others. The authors identified significantly positive three-day combined abnormal returns around the acquisition announcement date for both related and unrelated acquisitions, indicating that these events were not viewed as value-destroying on average, at least initially. However when identifying excess values based on sales multiples the authors reported that, while acquirers appear to possess significantly positive excess values one year prior to acquisition, there was a significant fall in these values post-acquisition.26 The
Shareholder Wealth Effects of Diversification Strategies
195
reported value loss was found to be greatest in related acquisitions in contrast with expectations based on the results of previous research. Indeed the median value loss in unrelated acquisitions was not significant at the 10% level. One confusing finding from their analysis was that firms reporting an increase in the number of business segments following acquisition suffered greater value losses than those reporting no increase. However it also appeared that many of the firms making unrelated acquisitions did not report any increase in the number of business segments. The results of this latter analysis are therefore difficult to interpret. As indicated above, one implication of the analysis by Graham et al. (2001) above is that acquisitions lead to value losses but that these losses are mainly confined to related acquisitions. The authors, however, suggested that their result may be driven by acquisitions of targets that had previously suffered value losses. The apparent discounted value of the subsequent combination will have therefore appear as the result of averaging the excess values of the two previously separate entities rather from the market response to the acquisition. In support of this hypothesis the authors reported that target companies had significantly negative mean (and median) excess values one year prior to acquisition. Their interpretation is also consistent with the reported greater post-acquisition reduction in value following related acquisitions, as targets in related acquisitions suffer the largest pre-acquisition discount. The authors also reported that there was no significant difference between the actual change in excess value for the combined companies and the change projected through a weighted-average of pre-acquisition values. In summary Graham et al. (2001) concluded that ‘the characteristics of acquired units are an important factor in determining the valuation discount’ and that the standard methodology used in diversification studies for comparing values with stand-alone companies may be flawed. Additional explanations for the previously reported diversification discount have also been offered in recent papers by Lamont and Polk (2001) and Mansi and Reeb (2001). Lamont and Polk (2001) suggested that much of the diversification discount might be explained by greater risk, both in higher variability of future expected cash flows and of future expected returns. Finally a recent paper by Mansi and Reeb (2001) found that the diversification discount disappeared when account was taken of leverage. They found that all-equity diversified firms in the period 1988 to 1999 appear to suffer no diversification discount while leverage was associated with the diversification discount in firms with debt financing. They also suggested that the use of book value for debt in valuation ratios of diversified firms leads to a downward measurement
196
ROBIN LIMMACK
bias and reported that the discount disappeared if market values were adopted for measuring the value of debt rather than book values. There is little international evidence on the relationship between the valuation of diversified firms and other factors. Lins and Servaes (1999) found no evidence of a diversification discount in a sample of German companies for 1992 and 1994 but did identify a discount in Japanese and U.K. companies. They found no evidence that inside share ownership mitigated against valuereducing diversification for the latter two countries but did report that a diversification discount was present in German companies in which inside ownership was below 5%. By contrast Fleming, Oliver and Skourakis (2001) reported that diversified Australian companies appeared to trade at a premium in the period 1988 to 1998 when adjustment was made for excess profitability. The study, referred to earlier, by Khanna and Palepu (2000) of diversified companies in India also suggest that diversification may be positively valued in a country with a less well-developed financial market.
7. REVERSING THE DIVERSIFICATION DECISION Indirect evidence on the costs and benefits of diversification has been imputed from examination of situations in which companies change their strategy to a more focused approach. Indeed there is now considerable evidence to suggest that many of the diversifying acquisitions in the 1960s and 1970s have been reversed subsequently (Ravenscraft & Scherer, 1987; Porter, 1987; Bhagat, Shleifer & Vishny, 1990; Kaplan & Weisbach, 1992). Thus Berger and Ofek (1996) identified that higher value loss for diversified firms led to a greater probability of takeover. They also found a relationship between the likelihood of takeover in the form of leveraged buyout (LBO) and the size of the value loss previously suffered by the diversified firm. Finally they identified a greater prior value loss to those firms that are ‘broken up’ subsequent to takeover. Overall Berger and Ofek (1996) suggested that their results were consistent with ‘the market for corporate control targeting and breaking-up combinations of business lines which have a greater value when each line is operated on a stand-alone basis’ (ibid p. 1178). Comment and Jarrell (1994) also reported that the 1980s in the USA were characterised by a trend towards refocusing with a positive relationship identified between stock returns and an increase in focus. John and Ofek (1994) reported that the positive impact on the sellers share price at the time of the divestiture announcement was related to the subsequent improvement in performance of the remaining assets of the seller. In addition the latter authors found that sellers share price gains appeared to be
Shareholder Wealth Effects of Diversification Strategies
197
related in part to an improved fit between the divested assets and those of the buyer. Subsequent studies have also identified that buyers of divested assets tend to be companies operating in related areas. Increased focus appears to provide potential efficiency gains both in the divested and the remaining assets. Denis, Denis and Sarin (1997), for example, reported that a sample of companies that decreased their level of diversification in the period 1985 to 1989 experienced significant negative excess values prior to the decrease in diversification. They also identified that 54% of the cases of refocusing were in response to a market disciplinary event in the year prior to refocusing. They concluded that there was little evidence to support the hypothesis that diversified firms with large negative excess values refocused voluntarily. A recent study by Gillan et al. (2000) of the actions of Sears, Roebuck & Co. provides a detailed examination of some of the tactics adopted by entrenched management in an attempt to overcome shareholder pressure to divest. While the evidence from U.S. studies generally supports the notion that poor performance leads to divestment and refocusing, evidence from a study undertaken by Haynes, Thompson, and Wright (2000) of a sample of U.K. companies in the period 1985 to 1989 found no such relationship. The authors did, however, find that the level of divestment was related to the size and degree of diversification of U.K. firms. In addition they identified a significant relationship between divestment and a number of governance variables, including board composition, management share ownership and gearing. The reduction in level of diversification appeared to be related to the extent to which the firm was monitored effectively. An alternative view of the reason for corporate refocusing was provided by Miller (1995), who argued that the higher value from refocusing was attributable to a ‘clientele’ effect. He suggested that the diversification discount identified in various studies was a consequence of investors’ lack of preference for diversified firms in which they hold less confidence or for risk-reducing aspects that they have no need of. However when a spin-off occurs the price of each component is then set by the group of shareholders who wish to invest in that particular segment. Miller also cited tax reasons why this shareholderoriented motive may be under-stated by managers. Although Miller’s views have produced very little reaction from others in this field, additional support for the clientele effect was provided in an earlier paper by Kudla and McInish (1988), which suggested that the greatest potential for post-spin-off benefits would arise in firms in which owners and non-owners differed most about value.
198
ROBIN LIMMACK
Weston (1989) also argued that companies often divest previously acquired firms for non-performance related reasons, including changes in external factors that have reduced synergistic benefits. Kaplan and Weisbach (1992) reported that 44% of large acquisitions completed between 1971 and 1982 were subsequently divested but that only 34% of these could be classified as unsuccessful based on accounting performance. The latter authors did, however, report that divestitures were four times more likely following unrelated acquisitions. While related acquisitions were more likely to be identified as ‘successful’ there was a similar proportion of divestments reporting a gain on divestment following both related and unrelated acquisitions.
8. SUMMARY AND CONCLUSIONS Motives ascribed to the diversification decision range from those relating to shareholder wealth maximisation (e.g. improved resource utilisation) to those aligned with management interests. Empirical evidence is available to support both sets of motives. While there is little evidence in support of the exercise of monopoly power as a diversification motive, this may be a function of the difficulty in obtaining relevant data and constructing unambiguous tests. Early evidence supported the hypothesis that diversified firms trade at a discount to their stand-alone equivalent. This evidence, and that relating to managerial shareholdings, suggested that the diversification discount was likely to be aligned with managerial interests, including the pursuit of growth and risk reduction. When management interests were more closely aligned with those of shareholders then the result was either a reduced probability of diversification or a lower value loss. The absence of a diversification discount for some recent time periods suggests that there may be external factors that have a different impact on the balance of costs and benefits at different points in time. However the evidence is also consistent with the ‘hubris’ hypothesis, with management pursuing a strategy that is designed to allow their (assumed) ‘superior’ skills to be applied to inefficiently managed firms, irrespective of their industry. In the latter case the motive may well have been to act in the interests of shareholders but the implementation of the decision appears to have led to the reverse. Support for the hypothesis that management interests have driven the diversification decision has been provided in the literature describing the subsequent ‘break-up’ of previously diversified firms. One of the main features of this literature is the apparent reluctance of senior management to reverse the diversification decision in the absence of external threat.
Shareholder Wealth Effects of Diversification Strategies
199
Evidence that provide support for the resource utilisation (shareholder wealth maximisation) motive for diversification may be found mainly in that literature relating to the greater efficiency of internal capital markets and to the elimination of information asymmetries. According to this literature the diversification discount appeared when the comparative advantage of internal markets was overtaken by the agency costs arising from the pursuit of management objectives. However the finding by Servaes’ (1996) that the diversification discount disappeared in the 1970s appears to be inconsistent with this view. An alternative, and potentially more plausible explanation, has been provided in recent literature with the suggestion that the lower value associated with diversified firms may have been present prior to the decision to diversify. The non-survival of other firms in these industry that decided not to diversify has also produced a valuation bias against diversifying survivors. Although there is as yet only limited evidence in support of the latter hypothesis it does appear to offer a plausible explanation for the diversification decision and of the time-varying nature of the discount (or premium). The relatively high number of companies that continue in their diversified state also suggests that for many companies any costs associated with diversification may be more than outweighed by the benefits or by the relatively higher costs involved in refocusing.
NOTES 1. Hanson Trust was for many years identified as an example of successful diversification but has, in recent years, transformed itself into a specialised building products company. 2. See also Dugger (1985). 3. Chappell and Cottle (1983), Eckbo (1983, 1985). 4. See for example Utton (1977), Montgomery (1985), Palepu (1985), Montgomery and Wernerfelt (1988). 5. The impact of information asymmetry leading to under-valuation will not be as easily observable. 6. See also Elgers and Clark (1980). 7. See also Chandler (1977). 8. See also Stultz (1990). 9. Majd and Myers (1987). 10. A related explanation proposed by Fluck and Lynch (1999) is the inability of a marginally profitable entity to obtain financing as a stand-alone entity because of agency problems between managers and potential financial sources. 11. See also Fauver et al. (1999). 12. Hass (1996). 13. See also Zuta (1997).
200
ROBIN LIMMACK
14. See for example the diversifying actions of Beatrice in the 1950s following antitrust scrutiny of their related acquisitions. (Baker, 1992.) 15. See also Erickson and Whited (2001). 16. Although various Companies Acts and Accounting standards require U.K. companies above a certain size to report segmental information there is a considerable discretion as to the form. Business segment information was made available in 1974 from Compustat for U.S. companies. 17. Asquith, Bruner and Mullins (1983) reported that acquisitions subsequent to the announcement of the programme also produced gains. 18. The definition of Tobin’s q adopted in the study was based on a numerator defined as the market value of equity plus the book value of debt and preference capital, rather than the replacement cost of assets. The numerator was calculated as a combination of book values and replacement costs (computed using the Lindenberg & Ross, 1981, algorithm). The authors also state that their results are similar when using the ratio of market value to book value of total assets. 19. The adjustment for size is based on the possibility that size and corporate efficiency are related (Lichtenberg, 1992). Firms with heavy investments in research and development (R&D) will have larger q values if the R&D is not capitalised. Differences in R&D between diversified and non-diversified firms would therefore lead to differences in q values, irrespective of the relative merits of diversification. Finally lack of access to capital markets by specialised firms may reduce the investment opportunities for these firms and maintain an artificially high q value. 20. Defined as earnings before interest and taxes. 21. Q ratios are calculated using the algorithm developed in Lindenberg and Ross (1981) and Hall et al. (1988). 22. Although in Table III of his paper the median discount, adjusted by the weighted average of all industries, was reported to be significant. 23. Refer also to Lang and Stultz (1994). 24. It is possible that differences in the accounting treatment of Research and Development expenditure may produce some bias in the calculation of valuation ratios. 25. Earlier studies have also identified the presence of Research and Development expertise as one aspect of non-specific resources that may be used efficiently in diversified expansion (Montgomery, 1994). 26. A similar result is also observed when asset multipliers were applied.
REFERENCES Aggarwal, R. K., & Samwick, A. A. (2001). Why do managers diversify their firms? Agency reconsidered. Dartmouth College Working Paper No. 01–01. Alchian, A. (1969). Corporate management and property rights. In: H. Manne (Ed.), Economic Policy and the Regulation of Corporate Securities. Washington D.C.: American Enterprise Institute. Amihud, Y., & Lev, B. (1981). Risk reduction as a managerial motive for conglomerate mergers. The Bell Journal of Economics, 12, 605–617. Anslinger, P. L., & Copeland, T. E. (1996). Growth through acquisition: a fresh look. Harvard Business Review, (Jan.–Feb.), 126–135.
Shareholder Wealth Effects of Diversification Strategies
201
Asquith, P., Bruner, R. F., & Mullins, D. W. (1983). The gains to bidding firms from merger. Journal of Financial Economics, 11, 121–139. Asquith, P., & Han Kim, E. (1982). The impact of merger bids on the participating firms’ security holders. Journal of Finance, 37(5), 1209–1228. Avery, C., Chevalier, J. A., & Schaefer, S. (1998). Why do managers undertake acquisitions? An analysis of internal and external rewards for acquisitiveness. Journal of Law, Economics, and Organization, 14(1). Baker, G. (1992). Beatrice: A study in the creation and destruction of value. Journal of Finance, 47, 1081–1119. Barney, J. B. (1988). Returns to bidding firms in mergers and acquisitions. Strategic Management Journal, 9, 71–78. Berger, P. G., & Ofek, E. (1995). Diversification’s effect on firm value. Journal of Financial Economics, 37, 39–65. Berger, P. G., & Ofek, E. (1996). Bustup takeovers of value destroying firms. Journal of Finance, 51(4), 1175–1200. Bhagat, S., Shleifer, A., & Vishny, R. W. (1990). Hostile takeovers in the 1980s: the return to corporate specialization. Brookings Papers on Economic Activity: Microeconomics, 1–72. Bhide, A. (1990). Reversing corporate diversification. Journal of Applied Corporate Finance, 3, 70–81. Billet, M. T., & Mauer, D. C. (2000). Diversification and the value of internal markets: the case of tracking stock. Journal of Banking and Finance, 24, 1457–1490. Blume, M. E., Friend, I., & Westerfield, R. (1984). Factors affecting capital formation. Rodney L. White Centre for Financial Research, The Wharton School, University of Pennsylvania. Burch, T. R., Nanda, V., & Narayanan, M. P. (2000). Industry structure and the conglomerate “discount”: theory and evidence. Working paper. Cable, J. (1977). A search theory of diversifying merger. Recherches Economiques de Louvain, 43, 225–243. Campa, J. M., & Kedia, S. (2000). Explaining the diversification discount. Working paper, Harvard Business School. Chandler, R. B. (1977). The visible hand. Cambridge, MA: Belknap Press. Chappell, W. F., & Cottle, R. L. (1985). Sources of concentration related profits. Southern Economic Journal, 51, 1031–1037. Chatterjee, S., & Lubatkin, M. (1990). Corporate mergers, stockholder diversification and changes in systematic risk. Strategic Management Journal, 11, 225–268. Chevalier, J. A. (2000). What do we know about cross-fertilization? Evidence from the investment policies of merging firms. Working paper, University of Chicago. Comment, R., & Jarrell, G. (1994). Corporate focus and stock returns. Journal of Financial Economics, 37, 67–87. Conn, R. L. (1973). Performance of conglomerate firms: comment. Journal of Finance, 28(3), 754–758. Cosh, A., Hughes, A., & Singh, A. (1980). The causes and effects of takeovers in the U.K.: An empirical investigation for the late 1960s at the micro-economic level. In: D. C. Mueller (Ed.), The Determinants and Effects of Mergers. Cambridge, MA: Oelgeschlager, Gunn and Hain. Denis, D. J., Denis, D. K., & Sarin, A. (1997). Agency Problems, Equity Ownership, and Corporate Diversification. Journal of Finance, 52(1), 135–160. Drucker, P. (1981). The five rules of successful acquisition. Wall Street Journal, 198, Part 15.8.
202
ROBIN LIMMACK
Dugger, W. (1985). The shortcomings of concentration ratios in the conglomerate age: new sources and uses of corporate power. Journal of Economic Issues, 19, 343–353. Eckbo, B. E. (1983). Horizontal mergers, collusion and stockholder wealth. Journal of Financial Economics, 11, 241–273. Eckbo, B. E. (1985). Mergers and the market concentration doctrine: evidence from the capital market. Journal of Business, 38, 325–349. Edwards, C. D. (1955). Conglomerate business as a source of power. In: National Bureau Committee for Economic Research, Business Concentration and Price Policy (pp. 331–360). Princeton: Princeton University Press. Elgers, P. T., & Clark, J. J. (1980). Merger types and shareholder returns: Additional Evidence. Financial Management, 9, 66–72. Erickson, T., & Whited, T. M. (2001). Measurement error and the relationship between investment and q. Working paper. Fauver, L., Houston, J., & Naranjo, A. (1999). Capital market development, legal systems and the value of corporate diversification: a cross-country analysis. Working paper, Columbia University, New York. Fleming, G., Oliver, B., & Skourakis, S. (2001). The valuation discount of diversified firms in Australia. Working paper, Australian National University. Fluck, Z., & Lynch, A. W. (1999). Why do firms merge and then divest? A theory of financial synergy. Journal of Business, 72, 319–346. Gazel, N. R. (1990). Beatrice from buildup through breakup. Chicago, Illinois: University of Illinois Press. Gibbons, R., & Murphy, K. (1992). Optimal incentive contracts in the presence of career concerns: theory and evidence. Journal of Political Economy, 100, 468–505. Graham, J. R., Lemmon, M. L., & Wolf, J. G. (2001). Does corporate diversification destroy value? Journal of Finance, (forthcoming). Gillan, S. L., Kensinger, J. W., & Martin, J. D. (2000). Value creation and corporate diversification: the case of Sears, Roebuck & Co. Journal of Financial Economics, 55, 103–137. Gregory, A. (1997). An explanation of the long-run performance of U.K. acquiring firms. Journal of Business Finance and Accounting, 24, 971–1002. Hall, B., Cummins, C., Laderman, E., & Mundy, J (1988). The R&D master file documentation. National Bureau of Economic Research. Boston, MA. Hansen, L. P. (1982). Large sample properties of generalised method of moments estimators. Econometrica, 50, 1029–1054. Hass, J. J. (1996). Directorial fiduciary duties in a tracking stock equity structure; the need for a duty of fairness. Michigan Law Review, 94, 2089–2177. Haynes, M., Thompson, S., & Wright, M. (2000). The determinants of corporate divestment in the U.K. International Journal of Industrial Organization, 18, 1201–1222. Hoskisson, R. E., & Hitt, M. A. (1990). Antecedents and performance outcomes of diversification: a review and critique of theoretical perspectives. Journal of Management, 16, 461–509. Hubbard, R. G., & Palia, D. (1999). A re examination of the conglomerate merger wave in the 1960s: An internal capital markets view. Journal of Finance, 54(3), 1131–1152. Huth, W. L., & Macdonald, D. N. (1989). The impact of anti-trust litigation on shareholder return. Journal of Industrial Economics, 37, 411–426. Hyland, D. C. (1999). Why firms diversify: an empirical examination. Working paper, University of Texas at Arlington. Jensen, M. C. (1986). Agency costs of free cash flow, corporate finance, and takeovers. American Economic Review, 76, 323–329.
Shareholder Wealth Effects of Diversification Strategies
203
Jensen, M. C. (1989). Eclipse of the public corporation. Harvard Business Review, 67, 61–74. Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3, 305–360. Jensen, M. C., & Ruback, R. S. (1983). The market for corporate control: the scientific evidence. Journal of Financial Economic, 11, 5–50. Jensen, M. C., & Murphy, K. J. (1990). Performance pay and top management incentives. Journal of Political Economy, 98, 225–264. John, K., & Ofek, E. (1994). Asset sales and increase in focus. Journal of Financial Economics, 37, 105–126. Kaplan, S. N., & Weisbach, M. S. (1992). The success of acquisitions: evidence from divestitures. Journal of Finance, 47, 107–138. Khanna, T., & Palepu, K. (2000). Is group affiliation profitable in emerging markets? An analysis of diversified Indian business groups. Journal of Finance, 55, 867–891. Khanna, T., & Tice, S. (2001). The bright side of internal capital markets. Journal of Finance, 56, 1489–1531. Klein, P. G. (2001). Were the acquisitive conglomerates inefficient? Rand Journal of Economics, (forthcoming). Kudla, R. J., & McInish, T. H. (1988). Divergence of opinion and corporate spin-offs. Quarterly Review of Economics and Business, 20, 20–29. Lamont, O. (1997). Cash flow and investment: evidence from internal capital markets. Journal of Finance, 52, 83–109. Lamont, O., & Polk, C. (2001). The diversification discount: cash flows vs. returns. Journal of Finance, 56, 1693–1721. Lang, L. H. P., & Stultz, R. M. (1994). Tobin’s Q, corporate diversification, and firm performance. Journal of Political Economy, 102, 1248–1280. Levy, H., & Sarnat, M. (1994). Capital investment and financial decisions (5th ed.). Prentice Hall International Lewellen, W. G. (1971). A purely financial rationale for the conglomerate merger. Journal of Finance, 26, 521–537. Lewellen, W. G., Loderer, C., & Rosenfeld, A. (1985). Merger decisions and executive stock ownership in acquiring firms. Journal of Accounting and Economics, 7, 209–231. Lichtenberg, F. R. (1992). Industrial de-diversification and its consequences for productivity. Journal of Economic Behavior and Organization, 18, 427–438. Liebeskind, J., & Opler, T. (1995). The causes of corporate refocusing: evidence from the 1980s. Working paper. Southern Methodist University, Dallas. Limmack, R. J., & McGregor, N. (1995). Industrial relatedness, structural factors and bidder returns. Applied Financial Economics, 5, 179–190. Lindenberg, E., & Ross, S. (1981). Tobin’s q and industrial organization. Journal of Business, 54, 1–32. Lins, K., & Servaes, H. (1999). International evidence on the value of corporate diversification. Journal of Finance, 54, 2215–2239. May, D. O. (1995). Do managerial motives influence firm risk-reduction strategies? Journal of Finance, 50, 1291–1308. Majd, S., & Myers, S. C. (1987). Tax asymmetries and corporate income tax reform. In: M. Feldstein (Ed.), Effects of Taxation on Capital Accumulation. Chicago, USA: University of Chicage Press. Mansi, S. A., & Reeb, D. M. (2001). Corporate diversification: what gets discounted? Working paper.
204
ROBIN LIMMACK
Matsusaka, J. G. (1993). Takeover motives during the conglomerate merger wave. Rand Journal of Economics, 24, 357–379. Matsusaka, J. G. (2001). Corporate diversification, value maximisation, and organizational capabilities. Journal of Business, (forthcoming). Matsusaka, J. G., & Nanda, V. (2000). Internal capital markets and corporate refocusing. Working Paper. University of South California. Melicher, R. W., & Rush, D. F. (1973). The performance of conglomerate firms: recent risk and return experience. Journal of Finance, 28(2), 381–388. Melicher, R. W., & Rush, D. F. (1974). Evidence on the acquisition-related performance of conglomerate firms. Journal of Finance, 29, 141–149. Meeks, G. (1977). Disappointing marriage: a study of the gains of merger. Cambridge and New York: Cambridge University Press. Meyer, M., Milgrom, P., & Roberts, J. (1992). Organizational prospects, influence costs, and ownership changes. Journal of Economics and Management Strategy, 1, 9–35. Miller, E. (1995). Why the break-up of conglomerate business enterprises often increases value. Journal of Social, Political and Economic Studies, 20, 317–341. Morck, R., Schleifer, A., & Vishny, R. W. (1990). Do managerial objectives drive bad acquisitions? Journal of Finance, 45, 31–48. Montgomery, C. A. (1985). Product-market diversification and market power. Academy of Management Journal, 28, 789–798. Montgomery, C. A. (1994). Corporate diversification. Journal of Economic Perspectives, 8, 163–178. Montgomery, C. A., & Wernerfelt, B. (1988). Diversification, Ricardian rents, and Tobin’s q. RAND Journal of Economics, 19, 623–632. Mueller, D. C. (1969). A theory of conglomerate mergers. Quarterly Journal of Economics, 83, 643–659. Mueller, D. C. (1977). The effects of conglomerate mergers: a survey of the empirical evidence. Journal of Banking and Finance, 1(4), 315–347. Myers, S., & Majluf, N. (1984). Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics, 13, 187–221. Palepu, K. (1985). Diversification strategy, profit performance and the entropy measure. Strategic Management Journal, 6, 239–255. Pandya, A. M., & Rao, N. V. (1998). Diversification and firm performance: an empirical evaluation. Journal of Financial and Strategic Decision, 11, 67–81. Penrose, E. (1959). The theory of the growth of the firm. Oxford: Basil Blackwell. Porter, M. (1985). Competitive strategy. New York: Free Press. Porter, M. (1987). From competitive advantage to corporate strategy. Harvard Business Review, 43–59. Prahalad, C. K., & Bettis, R. A. (1986). The dominant logic: a new linkage between diversity and performance. Strategic Management Journal, 10, 523–552. Rajan, R., Servaes, H., & Zingales, L. (2000). The cost of diversity: the diversification discount and inefficient investment. Journal of Finance, 55, 35–80. Ravenscraft, D. J., & Scherer, F. M. (1987). Mergers, sell-offs, and economic efficiency. Washington D.C.: The Brookings Institution. Reid, S. R. (1968). Mergers, managers and the economy. New York: McGraw-Hill. Reid, S. R. (1968). The conglomerate merger: as pecial case. Anti-trust Law and Economics Review, 2, 141–166.
Shareholder Wealth Effects of Diversification Strategies
205
Reid, S. R. (1971). A reply to the Weston/Mansinghka criticisms dealing with conglomerate mergers. Journal of Finance, 26, 937–946. Roll, R. (1986). The hubris hypothesis of corporate takeovers. Journal of Business, 59, 197–216. Rumelt, R. P. (1974). Strategy, structure, and economic performance. Boston, MA: Harvard Business School Press. Rumelt, R. P. (1982). Diversification strategy and profitability. Strategic Management Journal, 3, 359–369. Servaes, H. (1996). The value of diversification during the conglomerate merger wave. Journal of Finance, 51, 1201–1225. Scharfstein, D., & Stein, J. (2000). The dark side of internal capital markets: divisional rentseeking and inefficient investment. Journal of Finance, 55, 2537–2564. Schipper, K., & Thompson, R. (1983). Evidence on the capitalized value of merger activity for acquiring firms. Journal of Financial Economics, 11. Shin, H-H., & Stultz, R. (1998). Are internal capital markets efficient? Quarterly Journal of Economics, 113, 531–552. Shleifer, A., & Vishny, R. (1989). Management entrenchment: the case of manager-specific investments. Journal of Financial Economics, 25, 123–139. Shleifer, A., & Vishny, R. (1991). The takeover wave of the 1980s. Journal of Applied Corporate Finance, 4, 49–56. Stein, J. C. (1997). Internal capital markets and the competition for corporate resources. Journal of Finance, 52, 111–133. Stultz, R. M. (1990). Managerial discretion and optimal financing policies. Journal of Financial Economics, 26, 3–27. Utton, M. A. (1977). Large firm diversification in British manufacturing industry. Economic Journal, 87, 96–113. Villalonga, B. (2000). Does diversification cause the “diversification discount?” Working paper, University of California, Los Angeles, Anderson Graduate School of Management. Weston, J. F. (1970). The nature and significance of conglomerate firms. St. John Law Review, 44, 66–80. Weston, J. F. (1986). The rules for successful mergers. In: J. M. Stern & D. H. Chew, Jr. (Eds), The Revolution in Corporate Finance. New York: Basil Blackwell. Weston, J. F. (1989). Divestitures: mistakes or learning. Journal of Applied Corporate Finance, 2, 68–76. Weston, J. F. (1999). Mergers and performance. Seminar sponsored by The Martindale Center for the Study of Private Enterprise, Lehigh University. Weston, J. F., & Mansinghka, S. K. (1971). Tests of the efficiency performance of conglomerate firms. Journal of Finance, 26, 919–936. Weston, J. F., Chung, K. S., & Siu, J. A. (1998), Takeovers, restructuring, and corporate governance (2nd ed.). New Jersey: Prentice Hall. Whited, T. (2001). Is it inefficient investment that causes the diversification discount? Journal of Finance, 56, 1667–1691. Williamson, O. E. (1970). Corporate Control and Business Behaviour. New Jersey: Prentice Hall. Williamson, O. E. (1975). Markets and hierarchies: analysis and anti-trust implications. New York: Free Press. Zuta, S. (1997). Diversification discount and targeted stock: theory and empirical evidence. Working paper, New York University, New York.
INTERNATIONAL MERGERS AND ACQUISITIONS: PAST, PRESENT AND FUTURE Pervez N. Ghauri and Peter J. Buckley INTRODUCTION Mergers and acquisitions (M&As) have played a dominant role in the world economy for the past ten years. In 1998, mergers were worth US$2.4 trillion worldwide, a 50% increase on 1997, which was itself a record year. In 1999, this figure exceeded US$3.3 trillion and in 2000 US$3.5 trillion. In Europe, a major M&A activity area, the value of M&As rose to $1.2 trillion in 2002. The pace of M&A activity has slowed down in the last two years, mainly because no mega mergers, over $50 billion, were announced in these years. We define a merger as combination of assets of two previously separate firms into a single new legal entity. In a takeover or acquisition, the control of assets is transferred from one company to another. In a complete takeover, all the assets of the acquired company are absorbed by the acquirer; and the takeover “victim” disappears. In fact, the number of mergers in ‘mergers and acquisitions’ is almost vanishingly small. Less than 3% of cross border M&As by number are mergers (UNCTAD, 2000, p. 99). Full or outright acquisitions (100% control) accounted for more than half of all cross border M&As in 1999, although the proportion was lower in developing countries, largely because of legislation. In reality, even when mergers are supposedly between equal partners, most result in one partner dominating the other. The number of “real” mergers is so insignificant, that for practical purposes “M&As” should be simply referred to as “acquisitions” (UNCTAD, 2000, p. 99). Advances in Mergers and Acquisitions, Volume 2, pages 207–229. © 2003 Published by Elsevier Science Ltd. ISBN: 0-7623-1003-0
207
208
PERVEZ N. GHAURI AND PETER J. BUCKLEY
Friendly M&As can be distinguished from hostile ones. In a friendly M&A, the Board of the target firm agrees to the transaction (this may be after a period of opposition to it). Hostile M&As, on the contrary, are undertaken against the wishes of the target firm’s owners. The price premium tends to be higher in hostile than in friendly M&As even in situations where only one bidder is involved. The overwhelming proportion of both international and domestic M&As are friendly. In 1999, there were only 30 hostile takeovers out of 17,000 M&As between domestic firms. Hostile cross-border completed M&As accounted for less than 5% of total value and less than 0.2% of the total number of M&As during the 1990s (UNCTAD, 2000, p. 105; Thomson Financial Securities Data Company). There are of course some high profile battles which attract attention to hostile M&As. M&As are conventionally grouped into three categories (Buckley & Ghauri, 2002): • Horizontal – between competing firms in the same industry. • Vertical – between firms in buyer-seller, client-supplier and value chain linkages. • Conglomerate – between companies in unrelated businesses. Acquisitions differ according to the size of the acquired assets relative to the acquiring firm. In situations where the acquirer can leverage its assets to buy a firm which is as big or bigger than itself, the problems posed are quite different from ‘bolt-on’ deals where the new assets can be integrated with an existing part of the buying firm. A firm that takes over another firm makes two assumptions. The first is that the acquiring firm can extract more value from the same assets than can the current owners. This is a strong statement of comparative management ability. The second assumption is that not only the acquiring firm can extract more value from the same assets than the present owners, but also that the value extracted will be more than the market price paid for the assets. In order words, an acquiring firm is saying “our valuation of the assets is superior to the current valuation”. The fact that this assumption is often erroneous is the core reason why many acquisitions fail as profit enhancing tools – if the market price fully reflects the future profit stream of the acquired assets, then there is no scope for profit from the acquisition. However, opportunities for profit arise in situations where assets do not have a market price, as is the case with private firms or divisions of multi-unit companies. Here, ‘guesstimates’ of the market price have to be made, which provides scope for some firms to be more skilled than others in this estimation (Buckley & Ghauri, 2002).
International Mergers and Acquisitions: Past, Present and Future
209
The uncertainty in the valuation of future income streams is an important issue in takeovers. Superior competitive performance may well be unique to the firm, viewed as a team, which is not obtainable by others except by the purchase of the whole firm (Alchian & Demsetz, 1972). The firm may have a reputation or goodwill that is difficult to separate from the firm itself. Possibly members of the employee team derive their productivity from the knowledge they possess about each other in the particular environment of that unique firm. This gives rise to concern about “knowledge management” within firms. Knowledge may be difficult to transfer piecemeal to other firms by M&A or by any other means. The complexity of the modern firm defies easy analysis; so the inputs responsible for (long term) success may be difficult to identify and may be overvalued or undervalued for some time. The success of firms will be reflected in higher returns and stock prices, not higher input process. This is compounded by the fact that inputs are acquired at historic cost, but the use made of these inputs yields only uncertain outcomes. The acquisition cost of these inputs may fail to reflect their value to the firm at some future date. By the time their value is recognised, they are beyond acquisition by other firms at the same historic cost, and meanwhile the shareholders of this lucky firm have enjoyed higher profits. When such input acquisition decisions are made, they can give rise to high accounting returns for several years (Demsetz, 1973, p. 1). By the same token, once assets are embodied into a firm, their value becomes difficult to separate from the other assets which comprise the firm. Takeovers often involve the search for undervalued assets packaged into an existing firm. In a takeover, the potential acquirer is seeking some otherwise unavailable assets. Key examples of such assets are brand names, distribution networks, R&D facilities, management team skills, a loyal customer base, or specific knowledge. Often, the only way to acquire such assets is to purchase the whole firm. On the positive side for the acquiring firm, all the other assets of the acquired firm which are not required can be sold off. This gives rise to the phrase “asset stripping”. On the negative side, the remaining assets may not provide the value which the purchaser envisaged. This may be due to a simple misvaluation of the assets or it may result from the inability of the acquirer to release the synergy between the new assets and the firm’s existing ones. The purpose of this paper is to review the field of Mergers and Acquisitions and to analyse the present boom in order to provide guidance and evidence for managers engaged in M&A activities. The purpose is also to bring forward underpinning theoretical and societal implications of M&As, thus stimulating further research in the field.
210
PERVEZ N. GHAURI AND PETER J. BUCKLEY
MERGER WAVES M&A activity has experienced growth and downfall in different time periods over the last century, giving rise to the notion of ‘merger waves’. Table 1 shows successive waves of merger activity in the U.S. industry and features five such waves – the latest running from 1997 until 2000 and continuing into 2001 (Buckley & Ghauri, 2002, p. 4). The chief reason why M&As occur is “that control of corporations is a valuable asset; that this asset exists independently of any interest in either economies of scale or monopoly profits; that an active market for corporate control exists; and that a great many mergers are probably the result of the successful workings of this special market” (Manne, 1965, p. 112). Thus, management teams compete for the right to control corporations, and operating efficiency is ensured by a natural selection mechanism in which the threat and act of a takeover by raiders ensures the survival of the fittest management teams. The key mechanism works as follows: Inefficiency or abuse of managerial discretion by present managers leads to weak share prices. Potential raiders see the opportunity to alter policies and make capital gains as share prices respond to their improved management of the victim’s assets, and try to take over the company. Several conditions are necessary for this stock market selection process to work (Hughes, 1993): (1) Share prices should reflect the relative expected profitability of firms. (2) Raiders should be able to distinguish managerial shirking from poor performance arising from external circumstances. (3) Raiders should be motivated by the desire to remove policies that do not maximize shareholder value. Table 1.
1. 2. 3. 4. 5.
Merger Waves in U.S. Industry.
Merger Wave Approximate Dates
Current $ Amount Billion
Constant (2000) $ Amount Billion
Number of Deals*
1898–1902 1926–1939 1966–1969 1983–1986 1997–2000
6.9 7.3 46 618 4,500
136 69.3 236 Na 4,500
3012 4828 Na 9617 31,152
* Involving U.S. companies. Source: Buckley and Ghauri (2002), p. 4
International Mergers and Acquisitions: Past, Present and Future
211
(4) Raiders should be able to obtain a sufficient payoff to make their activities worthwhile. If (1) does not hold, scope for changes in corporate control occurs because of differences of opinion about the accuracy of stock market valuations between seller and raiders, rather than (potential) changes in managerial objectives or corporate efficiency. Failures to generate acceptable returns in the market for corporate control are also caused by transaction costs and because acquiring companies cannot capture all the benefits of the raid. Acquisitions and takeovers incur large transaction costs (legal, advisory, and information costs). Defensive tactics on the victims’ part aim to increase these costs. They include: “shark repellents;” constitutional or voting arrangements to protect incumbents; “golden parachutes” to raise the costs of firing incumbent managers; counterbidding; seeking a “white knight” to contests an unwanted bid; or raiding a third party. It has been suggested (Grossman & Hart, 1980) that the incentive structures in bids may lead to too few M&As for them to serve as a disciplining force for inefficient managers. There can be an incentive for individual shareholders to “free ride” on post-merger gains rather than to sell out. In this case, the private return to the raider is less than the overall return which is split between the raider and the free riders. However, the presence of huge “merger waves” suggests that free riding may not be a major problem. Alternatively, partial bids (e.g. 51% acquisitions) may be undesirable given that while incurring transaction costs they may merely redistribute wealth between the raiders and the “oppressed” minority shareholders. Smaller companies with substantial owner control may be overvalued by the owners, making the costs of disciplinary M&A action prohibitive.
MOTIVES FOR M&As There are several motives for choosing M&A as a strategy. The most important, as seen from the experience of the last decade, has been economies of scale and scope. Companies aim to achieve economies of scale by merging the resources of two companies, or create economies of scope by acquiring a company allowing product/market diversification. Daimler’s merger (acquisition) with Chrysler and AOL’s merger with Time Warner are good examples. Other motives include accessing each other’s technology or market reach, achieving a dominant position in the industry, and consolidation of the industry. Overall, value creation for either company, or at least for the acquiring company, is the underlying reason. This value is created through either cost cutting or through
212
PERVEZ N. GHAURI AND PETER J. BUCKLEY
increased scope. However, not all the possible motives can be considered rational from the business perspective. Fashion, too often, has led managers to pursue a strategy of M&A. As a consequence, M&A strategy has become an end in itself or a means of empire-building. Another motive for M&A is differences in asset valuation between different national markets. In examining cross-border mergers, Gonzalez et al. (1998) suggest that assets at a national level are systematically undervalued. Imperfections that cause friction in the product and service markets lead to asset undervaluation. This leads to acquisitions being the most cost-effective strategy for penetrating certain national markets. The idea that knowledge acquisition is frequently the motive of corporate acquisitions has been pointed out by many studies (Buckley & Ghauri, 2002). Knowledge is a key focus because often it cannot be acquired in efficient factor markets as it is bundled with other assets and because of asymmetric information. The range of buyer responses to the possibility of overbidding includes the following: lower bid premia, contingent buy-outs, and lengthy negotiations designed to elicit tacit information. Where the firms draw on unrelated forms of expertise, these strategies are not used, leading to the conclusion that either lower post-merger integration is envisaged or that unrelated buyers are simply unaware of extra information needs. Seth et al. (2000) suggest three motives for the foreign acquisitions of U.S. firms – synergy, managerialism, and hubris. The synergy hypothesis proposes that M&As take place when the value of the combined firm is greater than the sum of the values of the individual firms. The managerialism hypothesis suggests that managers embark on an acquisition to maximise their own utilities at the expense of the firm’s shareholders. The hubris hypothesis suggests that bidding firm managers make mistakes in evaluating the target firm, but undertake the acquisition presuming that their valuations are correct. There is something of a conflation of these motives with outcomes in the empirical testing where positive and negative gains are attributed to acquirers and targets, and total gains and losses are distributed among motives on a somewhat arbitrary basis. In addition to all the possible motives listed in the previous section, M&As may be the result of empire-building behaviour by managers. Growth maximizing implies that some management teams have a lower discount rate than the market as a whole. Such teams are then faced with a wealth of “undervalued” takeover opportunities. In the international arena, the explanation of takeovers can have a two-fold aspect. If we use the simple formula for capitalising a future income stream:
International Mergers and Acquisitions: Past, Present and Future
C=
213
1 r
where “C” is the value of a capital asset, “I” is the stream of income it produces, and “r” is the rate of return on investment, then a foreign firm can successfully buy out a domestic one if either “I” is higher or “r” is lower (or both). International business theory suggests that “I” will be higher where the foreign firm can raise the income stream by infusing the victim with better management, technology, organisational skills, or superior marketing (Kindleberger, 1969; Buckley & Ghauri, 1999). Alternatively, the capital market may apply a lower discount rate to the asset when it comes under foreign ownership, perhaps in the belief that higher returns will result. Thus, Aliber (1970, 1971) suggests that a foreign asset owned by a U.S. firm will be evaluated by the capital market as if it were a dollar asset. Differences in capitalisation ratios set by the market on the basis of ownership may provide a rationale for foreign takeovers even if no increase in earnings occurs. Morck et al. (1988) align the categories of hostile and friendly takeovers with disciplinary and synergistic motives. “Disciplinary takeovers” are aimed at correcting non-value-maximising practices of the managers of target firms. Here, the actual integration of the acquirer and the target firm is not essential. A takeover is the best way to change control and therefore strategy. “Synergistic takeovers” are motivated by the possibility of benefits of combining the businesses of the two firms. Morck et al strongly suggest that hostile and friendly takeovers should not be lumped together in an analysis of M&As, but that they should be separated and inferences about one type should not be drawn on takeovers undertaken for the other motive. The empirical work samples from the Fortune 500 in 1980 and examines takeovers between 1981 and 1985. They found that firms which were the target for hostile take-over bids compared to the universe of firms, were smaller, older, more slowly growing, and had lower Tobin’s q, more debt, and less investment of their income. They were less likely to be run by the founding family and had lower officer ownership than the average firm. Compared to the universe of Fortune 500 firms, friendly targets were smaller and younger but had comparable Tobin’s q values and growth rates. The friendly targets were more likely to be run by a member of the founding family and had higher officer ownership than the average firm. The decision of a CEO to retire with a large stake in the firm or with a relationship to the founder often precipitated a friendly takeover – high officer ownership was the most important attribute predicting friendly acquisitions.
214
PERVEZ N. GHAURI AND PETER J. BUCKLEY
ARE M&As THE RIGHT CHOICE? Kogut (1991) examines joint ventures (JVs) as an option to acquire. He seeks to explain JVs as “real options” to expand in response to future technological and market developments. In this sense, the exercise of the option is the move to acquisition. JVs thus reduce the risks associated with uncertainty in entry into a new (national) market or technology. Investment in acquisition then becomes a second stage choice made after the firm acquires additional information from its initial venture. Such a strategy is useful in times of great volatility (Buckley & Casson, 1998), but entrant firms should be aware of the risk of being locked into a joint venture where it is captured by the local partner. Brouthers and Brouthers (2000) examine the impact of institutional cultural, and transaction cost factors on the “buy or build” (acquire or establish new facilities) choice. As joint ventures and wholly owned subsidiaries are ownership choices, the “buy or build” decision can be related to the entry mode choice of acquiring existing assets vs. establishing new facilities or a “greenfield” site. Greenfield entry is more likely to take place in high-tech industries where the entrant firm has strong intangible competitive advantages. The option perspective is widely recognised as a means of understanding the dynamics of JVs as the partners have the option to acquire, divest, and expand their stake. However, as Chi (2000) points out, the nature of the option, its value to the “buyer” and the “seller” varies markedly with its structure. “Even the most commonly recognised option in a JV – the option to acquire the partner’s stake – tends to have some unique structural attributes” (Chi, 2000, p. 665). A typical JV contract, as Kogut (1991) points out, does not give either partner the right to acquire or divest the venture at an ex ante specified price. Thus, the partners have to negotiate a price ex post making the options exercise price indeterminate ex ante. Partners have good reasons for believing that new information will arrive during the operation of the JV, which can result in one of them moving to a full acquisition. Hennart and Reddy (1997, p. 1) examine the choice between “two alternative methods of pooling similar and complementary assets: the merger/acquisition and the greenfield equity joint ventures”. This conflates the ownership choice (wholly owned versus joint venture) with the entry choice (takeovers versus greenfield entry). In fact, there are technically four potential choices of entry mode viz: (1) Greenfield/Wholly Owned;
International Mergers and Acquisitions: Past, Present and Future
215
(2) Greenfield/Joint Venture; (3) Takeover/Wholly Owned; and (4) a partial takeover, Takeover/Joint Venture. Hennart and Reddy compare (3) with (2) leaving out wholly owned greenfield entry and partial acquisitions presumably because they are not methods of “pooling similar and complementary assets”. Hennart and Reddy contrast “two competing theories of why joint ventures exist” (p. 1): asymmetric information in the pre-merger situation versus “indigestibility” which refers to the (postmerger) problems of acquiring indivisible resources. In a commentary, Reuer and Koza (2000) point out that these views are complementary and overlapping. Kogut and Singh (1988) examine the effect of national culture on choice of entry mode after accounting for firm and industry level variables. They classify entries into “greenfield” (equals wholly owned greenfield ventures), “joint ventures” (greenfield shared ownership), and “takeovers” (purchase of stock in an already existing company in an amount sufficient to confer control). They claim that all acquisitions in their sample conferred a controlling equity share so the partial takeover (takeover/joint venture) category conveniently disappeared. After allowing for the impact of firm and industry variables, they found that cultural distance was significant in influencing the entry mode choice into the U.S. They used a conflation of Hofstede’s (Hofstede, 1980) four cultural dimensions (power distance, uncertainty avoidance, masculinity/ femininity, and individualism) as a measure of cultural distance. It should be noted that this measure is open to great objections even if the separate relevance of the four dimensions to the decision is accepted. Baumann (1975) sets out a theory of (international) mergers. He examines “monopoly power, economies of scale, synergistic effects” and “non-profit maximising behaviour, differences in attitudes towards risk” and economic disturbance theory as alternative explanations for mergers, and applies property rights theory to foreign direct investment. The theoretical base, following an early version of internalisation theory (McManus, 1972, following Coase, 1937, see also Buckley & Casson, 1976) is tested on U.S. FDI in Canada. Both the improvement in the profit stream and the lower discount rate for assets denominated in stronger currencies are included as explanatory variables as are improvements due to superior management post takeover. Although Baumann’s model is unable to distinguish between M&As and greenfield FDI, he is able to show the close relationship between the determinants of M&As and of FDI.
216
PERVEZ N. GHAURI AND PETER J. BUCKLEY
EVALUATION AND TACTICS Mergers and Acquisitions face two rather distinct groups of issues and problems: the motives and valuation at pre-merger stage and then integration and performance evaluation at post-merger stage. This section discusses the questions of valuation that arise in the pre-merger stage, post-merger integration and performance evaluation. Valuation has been and is one of the most difficult problems in M&As. Keenan (1980) examines a situation in which real economic adjustments evolve from the mergers and focuses on the valuation problems related to synergistic benefits arising from M&A. He claims that capital budgeting theory is not developed enough to handle such “multi-period” problems where risk levels are shifting and where project-tied financing consideration exists. Moreover, the theory is rather generalised at the firm level, considering equations for demand, supply, and budget constraints in the economy. The valuation of M&As becomes particularly complex because the real economic markets for labour and capital are not strongly efficient. Most of the research on M&As since the 1960s focused on two questions: (1) Are there any real benefits to conglomerate type mergers? and (2) What are the portfolio implications of a merger between two firms? According to this research, statistically there are few valuation benefits, and many M&As are in fact detrimental to profit maximising goals. While the stockholders of acquired firm may reap some benefits, the stockholders of the acquiring firm do not. Keenan (1980) also demonstrates the complexities of valuation in a hypothetical case where a firm (A) wants to take over a firm (B) and becomes a new firm (C). He claims that a merger between two firms with different constant growth rates yields a firm that does not have a constant growth rate. This leads to the question of whether mergers generate any real benefits. Kennan (1980) also builds up a scenario with acquisition of labour-intensive firms and demonstrates how the value of the acquired firm may, in fact, decrease. This suggests that there may be serious problems in the expectedvalue determination, negotiated-price determination, and accounting standards applied to the acquisition of a service firm where the value of the labour product is greater than the wage rate paid. Singh and Montgomery (1987) find that acquired firms in related acquisitions have higher returns than acquired firms in unrelated acquisitions. For acquiring firms, abnormal returns directly attributable to the acquisition transaction are not significant. However, announcement effects are less easily detectable for the acquirers rather than targets because the acquisition affects only part of the acquiring firm but the whole of the target firm. In addition, as
International Mergers and Acquisitions: Past, Present and Future
217
the acquisition is an event in a series of an implicit diversification programme, its effect as a unique market signal is mitigated. Expected gains for acquiring firms are thus often competed away in the bidding process, while stockholders of the target firm obtain high proportions of the gains. M&A literature cites culture as one of the dominating factors that influence performance. Buono et al. (1985) studied a merger between two banks from the perspective of organisational culture. Data on organisational culture is collected and analysed through pre- and post-merger interviews. The interesting point is that the study shows that culture is an important issue even when the firms come from the same country and same industry. The study thus enhances our understanding of culture and its consequences on the merger process. It also clarifies and helps us understand how a new culture emerges in a newly formed organisation. It looks at objective and subjective organisational culture and suggests that although both aspects of culture are important, subjective culture provides more distinctive interpretations of similarities and differences among people in different groups, as it is particularly unique to an organisation. Another issue discussed is the organisational climate. These are the issues that influence behaviour, satisfaction, expectations of organisational members and thereby success or failure of a merger. Bresman et al. (1999) focus on the factors facilitating the post-acquisition transfer of knowledge over time. They find that the transfer of technological know-how is facilitated by communication, visits, and meetings and increases as time goes by. The transfer of patents is associated with the articulability of the knowledge, the size of the acquired unit, and the recency of the acquisition. High quality reciprocal knowledge transfer takes time to develop but it gradually replaces the one-way transfer of knowledge from acquirer to acquired. This parallels the increasing transfer of tacit knowledge which is more time consuming to transfer than knowledge which is more articulated. The human relations processes across the merged unit develop in similar fashion (Buono & Bowditch, 1989). Calori et al. (1994) examine the influence on national culture on the choice of integration mechanisms used by French and U.S. entrants into Britain and British and U.S. acquirers in France. They find that firms are influenced by their national administrative heritage when they acquire foreign firms. In the U.K. comparison, the French acquiring firms relied more on formal control of strategy and of operations than U.S. firms. The American acquiring firms relied more on informal communication and cooperation (teamwork) than the French. In examining the acquisition of French firms, it was found that the U.S. acquiring firms relied more on formal control by procedures than the British, and American managers became more personally involved, suggesting a
218
PERVEZ N. GHAURI AND PETER J. BUCKLEY
“hands off” attitude from the managers of British acquiring firms. In terms of impact on post-merger performance, the article is suggestive. It points out that the health of the acquired firm prior to the merger is negatively correlated to improvements in attitudinal performance, and both informal communication and cooperation (teamwork) and informal personal efforts from the managers of the buying firms are positively correlated with improvements in attitudinal performance. This confirms the view that socialisation is a key factor in reducing post-merger conflicts and demotivation (Haspeslagh & Jemison, 1991). Two other variables are significant in economic performance of the acquired firms – time elapsed since acquisition and the level of shared resources and transfers. Thus the implementation of synergies between the merging firms contributes to economic performance but there seems to be a delay before the economic performance becomes evident. Two measures of control mechanisms are also correlated with economic performance. Informal personal effects of the buying firm’s managers positively affect performance, and the level of control of operations exercised by the buying firm is negatively correlated with performance. Chatterjee et al. (1992) deal with the compatibility of the merging firms and its relationship to shareholder gains and relatedness. They surveyed the perceptions of top management teams in 168 M&As. Based on the response from 73 firms, their findings show an intense relationship between the acquired managers’ perceptions of cultural difference and shareholder value. Their results support the hypothesis that changes in shareholder value of acquiring firms are directly related to the degree to which buyer’s top management team tolerates multiculturalism. Their findings also show that a lower cultural tolerance will influence shareholder value negatively. This study provides systematic evidence linking equity and human capital in M&As. Morosini et al. (1998) examine cross-border acquisition activity in Italy. They measure post acquisition performance using the percentage rate of growth of sales over the two-year period following the acquisition. This is clearly a less satisfying measure of performance than stock market pricing or profitability. The results of the study are quite surprising. The authors found that national cultural distance enhances cross-border acquisition performance. They explain this result by the argument that cross-border acquisition in more culturally distant countries might provide a mechanism for multinational companies to access diverse routines and repertoires which have the potential to enhance the combined firm’s performance over time. Post-acquisition strategies are also adduced to enhance this result. Deepak Datta (1991) studies the relationship between organisational and acquisition performance. Using a sample of 173 U.S. acquisitions he evaluates
International Mergers and Acquisitions: Past, Present and Future
219
post-merger performance of the firms involved in acquisition. This study demonstrates that differences in top management styles of two firms have a negative impact on post-merger performance as well as integration. He also examines difference in reward and evaluation systems and suggests that these differences do not have the same kind of negative impact on performance as differences in management styles. A possible reason is that differences in reward and evaluation systems are more easily and quickly reconciled following an acquisition than are differences in management styles. Although the study only takes two aspects of organisational fit into account and there are some flaws in data collection, it brings forward important issues in postacquisition integration of firms. Villinger (1996) deals with post-acquisition learning by Western firms in Hungary, Czech Republic, Slovakia, and Poland. By focusing on these countries, he covers the most important market area for M&A activity in Central and Eastern Europe. He points out that in these countries, cross-border management skills seem to be more important than general business skills for post-acquisition management efforts. The knowledge of acquired partner’s language and sensitivity about cultural issues are crucial. He claims that learning is the central aspect rather than integration, at least in the earlier phases of the post-acquisition stage, as it is more important to transform the East European target than to integrate it in its Western parent. It then proposes three levels of learning based on Senge (1968) and Child et al. (1992) as “single-loop learning,” “double-loop learning,” and “deuterolearning.” These different levels pose different problems for learners and learning entity but in the Eastern Europe context and for a successful transition, this framework is considered to be the most “elegant.” Empirically examining 35 cases, Villinger (1996) concludes that general business competence dominates cross-border skills, which creates post-acquisition problems. A lack of communication and understanding is generally blamed for insufficient learning. Language and cultural awareness is thus considered the most important issue for postacquisition integration in Eastern Europe.
THE IMPACT OF M&As ON COMPANY STRATEGIES Many problems can arise in the assessment of the impact of M&As on efficiency and economic performance. In terms of the stock market selection process, the choice of the control group is critical. Control groups are classified into acquirers, acquired, and neither acquiring nor acquired. Some firms may belong to both the first two categories. Characteristics of firms may affect the relevant comparisons – the size of firms, industry, and time may affect
220
PERVEZ N. GHAURI AND PETER J. BUCKLEY
the choice of the most appropriate control group. M&A performance, too, can be measured in various ways – profitability, growth, changes in profitability, and market value are possible measures. The impact on technical change may also be relevant in certain contexts; and again, it is important to choose the appropriate counterfactual. Total gains in cross-border acquisition are estimated to be 7.6% of the preacquisition value of the combined firm (Seth et al., 2000). This is similar to the estimate of total gains for domestic acquisitions (Bradley, Desai & Kim, 1988) but larger than that reported by Eun et al. (1996). Seth et al. find that positive total gains occur in 74% of the acquisitions in their sample, similar to the proportions reported by Berkovitch and Narayanan (1993) at 76% and Bradley, Desai and Kim 75%. However, Seth et al. find that targets realise the majority of the gains, whilst acquirers appear to neither gain or lose on average. Successful acquirers in single bidder acquisitions retain around 40% of the total gains on average, whilst acquirers in multiple bidder transactions make small losses. Researchers find that only about one-third of recent M&As have positive effects on the shareholder value for the buying firm (Schenk, 2000). Performance results, reported by earlier research, are quite mixed (Cowling et al., 1980; Magenheim & Mueller, 1988; Sirowar, 1997). Dickerson et al. (1997) show that company growth through acquisition yielded a lower rate of return than growth through internal investment. A survey of 22 accounting data studies from nine countries showed that the average acquiring firm does not earn a significantly higher profit than the industry average (Bild, 1998). Earlier results from Mandelker (1974) found that the market for acquisitions can be regarded as perfectly competitive, supporting the hypothesis that information regarding mergers is efficiently incorporated into the stock price. Stockholders of acquiring firms seem to earn normal returns from mergers as compared to other ventures of similar risks. Stockholders of acquired firms earned abnormal returns of approximately 14% on average in the seven months preceding the merger. The fact that the market anticipates mergers, at least three months on average according to Franks et al. (1977), before announcement further complicates the analysis. Aiello and Watkins (2000) suggest that leveraged buyouts (LBO) are more successful than M&As in general. “Between 1984 and 1994, some 80% of LBO firms exported that their fund investors had received a return that matched or exceeded their cost of capital, even though in many cases the prices paid for the companies those funds acquired were pushed up by competing bidders” (p. 101). This, they suggest, is due to the fact that senior managers at financial investors, like the most successful corporate acquirers, approach potential acquisitions with sensitivity and a well established process.
International Mergers and Acquisitions: Past, Present and Future
221
They contrast this approach with that of “most corporate managers” who try to beat down prices and often delegate actual deal management to outside experts, investment bankers and lawyers. Literature from the strategic management and international business fields suggests that M&As are not a uniform phenomenon (Lubatkin, 1983). Outcomes of M&As depend on the strategic fit between the firms and on the quality of pre- and post-merger management. In addition, market conditions influence M&A outcomes. Halpern (1973) showed that since most mergers occur in rising markets, this effect needs to be removed in estimating gains and premia. Size of company is a factor too. If we accept that acquisitions are often a search for a unique asset embedded in a company, then returns from that asset will vary as the size of the unique element varies as a proportion of the total company. Because quality of management, too, is a factor, uncertainty about when that management will be replaced will be reflected in the current stock price. When the merger is announced, there is a positive total adjusted gain which reflects the markets’ re-evaluation of the company under its new management. Bargaining strategies also will affect the distribution of gains.
THE IMPACT OF M&As ON SOCIETY Siegfield and Sweeney (1981) discuss the most important characteristics of M&A, namely, market concentration, firm size, product diversification, and geographic dispersion, and discuss their impact on society. Based on empirical evidence, they then assess other social consequences such as effect on workers, distribution of income, and on community welfare and explain reasons why M&A might increase and/or decrease social goals. Finally, they stress the need for more systematic research in this area that will assist policy makers in handling these problems. Franks and Mayer (1993) examine the relationship between capital markets and corporate control. They compare the relationship in France, Germany, and the U.K. and include buy-out and buy-ins by managers. They find that regulation (company laws, competition policy, stock exchange rules, and labour laws) explains much of the difference in control and ownership changes across countries, even within an economic area such as the European Union (EU). The EU’s Competition Commissioner heads a competition authority which is currently “seen by business as overburdened and unpredictable in its decision making” (Hargreaves, Financial Times 18.9.2000, p. 16). The EU’s merger regulation has a theoretically clear remit to prevent mergers which inhibit competition and allow others to proceed. In practice, European mergers still need clearance from several national competition authorities. The unit that
222
PERVEZ N. GHAURI AND PETER J. BUCKLEY
reviews mergers, the Merger Task Force, is under-resourced. The absence of a single European regulatory organisation, let alone a global one, vastly increases the transaction costs of M&As. In practice, European authorities veto very few mergers (of 1500 deals in the decade of the 1990 only 13 were vetoed and 12 more withdrawn after it became clear that Commission would intervene over excessive market power). The increase in number and complexity of deals is putting pressure on regulatory authorities – particularly those like the EU which wants to examine ever smaller sized M&As. In Europe in particular, labour organisations play a role in M&As. Given the high rate of failure because of the human factor, this is perhaps to be welcomed. The International Labour Office (ILO) is currently calling for increased dialogue between management and workers in the M&A process in an attempt to minimise the negative effects. The number of the significant regulatory authorities with which merging companies have to comply in the world has mushroomed. Reports of the deal between Coca-Cola and Cadbury-Schweppes suggest that their lawyers sought anti-trust approval in more than 40 jurisdictions around the world. In the failed effort to merge, Alcan Aluminium of Canada, Pechiney of France and Algroup of Switzerland, lawyers from 35 firms filed for regulatory approval in 16 jurisdictions and 8 languages. Pernod Ricard’s £2.1 billion purchase of 38.6% of Seagram’s drinks portfolio (with Diageo) has required filing for regulatory approval in 70 countries (Kemeny, Sunday Times 21.1.01, pp. 3/6). Simplification of competition laws and creating a consensus among regulators to make antitrust laws coherent and predictable is the current extent of feasibility. There is little current prospect of a new international bureaucracy to oversee mergers. The United States has recently opposed a European proposal to give the World Trade Organisation (WTO) a central role in policing mergers. However, harmonisation is made urgent by the increased number of merger cases which amounted to 26,000 filings between 1994 and 2000, with the U.S. Department of Justice and the Federal Trade Commission (FTC). In case of regulations and antitrust laws, firms often jump into bed with each other without even considering regulatory concerns. MCI and WorldCom, two American telecommunication firms were totally surprised by objections to their $37 billion marriage from European regulators (The Economist, Jan. 9, 1999). The liberalizations in FDI regimes has encouraged companies to expand through M&As. The international regulatory framework, bilateral investment, and double taxation agreements and the WTO have also supported these trends. In Europe, the EU has stimulated restructuring for national, regional, and crossborder M&As which has resulted in major changes in corporate ownership. One impact of all these changes is that there is an increased acceptance of
International Mergers and Acquisitions: Past, Present and Future
223
M&A all over the world, even though there are mixed opinions on the impact of M&A on industrial concentration. According to one opinion, cross-border M&As can have a positive effect on competition if the foreign firm takes over ailing domestic firms that would otherwise have been forced out of the market. They can also challenge established domestic firms to create effective rivals. The opposing view is that “monopolizing M&As” can occur in the following situations (UNCTAD, 2000, p. 193): (1) The acquiring firm was exporting substantially to a market before it buys a competing firm in it; (2) A foreign firm with an affiliate already in the market acquires another, thereby acquiring a dominant or monopolistic market share; (3) The foreign firm acquires a market leader with which it has previously competed; (4) The acquisition is intended to suppress rather than develop the competitive potential of the acquired firm; (5) A foreign firm with an affiliate in a host country acquires an enterprise in a third country that has been a source of import competition in the host country market; (6) Two foreign affiliates in a host country merge, although their parent firms remain separate, eliminating competition between the two and leading to a dominant market position. Although we can accept that higher concentration by itself does not indicate anti-competitive conduct, the crucial issue is that it differs from market to market and from industry to industry. In the United States and the European Union, only a small number of M&As are scrutinized to assess negative impacts on competition. An even a smaller number are asked to sell off parts of the business or are completely ruled out. In the United States, for example, only 1.6% of 4,679 M&As notified to antitrust authorities resulted in enforcement actions, with only about 1% being challenged in the end (U.S. Department of Justice, 2000). In the EU, the situation is not much different. In 1999, only 14 out of 292 M&As (less than 5%) were challenged or subject to a second phase investigation, while an additional 19 cases were reported but cleared in the first phase of investigation. In Japan, all 3,813 M&As notified in 1998 were cleared, only 2 transactions were revised during pre-notification consultation (ibid, p. 7). This means that the authorities do not believe M&A to be against the public interest even if concentration does increase. This really needs further research.
224
PERVEZ N. GHAURI AND PETER J. BUCKLEY
CONCLUSIONS Mergers and acquisitions have become the most dramatic demonstration of vision and strategy in the corporate world. With one single move you can change the course of your company, the careers of your managers, and create value for your shareholders (Puranam et al., 2000). Time and again, we have seen how share values increase or decrease due to the mergers announced and completed. Daimler-Chrysler, a $38 billion merger, is a good example where the share values dropped by almost 40% from the time of the announcement of merger, May 1998 to December 2000 (Bert & Trait, 2000). More than 50% of the mergers so far have led to a decrease in share value and another 25% have shown no significant increase. As well as the strategies of actual and potential acquiring firms, the defensive strategies of potential victims also impact taxpayers, employees, managers, and workers in other firms, consumers and the pace of technological change. M&As have a profound impact on social and political processes as well as the purely business and economic ones. The merger wave, which in the early 1990s was considered an American trend, became a worldwide trend and the latter part of the 1990s and the year 2000 has shown increased M&A activity in other parts of the world. Europe led in M&A activities in 2000. In 1999–2000, the total value of M&A reached $3.5 trillion (from $0.4 billion in 1990). Out of this, the USA accounted for about $1 trillion and Asia for $144.6 billion, the rest came from Europe (The Economist, Jan. 27, 2001, p. 59). In spite of this popularity, more than 50% of the M&As are reporting post-merger problems. While in the past, firms have justified these deals arguing for synergistic benefits, more and more M&A are reporting synergistic losses. Only about one-third of M&As were found to have positive effects on the shareholder value for the buying firm in one study (Schenk, 2000). Performance results, reported by earlier research, are quite mixed (Magenheim & Mueller, 1998; Sirower, 1997). A survey of 22 accounting data studies from nine countries showed that the average acquiring firm does not earn a significantly higher profit than the industry average (Bild, 1998). As for achieving efficiency gains, the most cited justifications for M&As, the scope for rationalization and improving company performance by achieving an international specialization of the value chain can be particularly high in M&As (UNCTAD, 2000). Moreover, size can have a large influence on R&D and the expansion of distribution networks as well as adoption of new information and innovations.
International Mergers and Acquisitions: Past, Present and Future
225
More than 50% of the total M&A activity in the last few years have been between cross-border firms, and a lot of these mergers and acquisitions experienced problems stemming from cultural and institutional country differences. For example, in Upjohn-Pharmacia, a lot of time and energy was spent on “American” versus “European” practices rather than on achieving synergistic benefits. Daimler Chrysler was taken to court by its biggest shareholder for misleading information. The fact that cross-border M&As place host country assets under the governance of TNCs and contribute to the growth of an international production system does not necessarily mean that it creates synergistic benefits. In addition, as a result of these cross-border mergers, concentration has increased in many industries such as: banking, pharmaceuticals, automobiles, telecommunication, insurance, and energy. After the Asian crisis, cross-border M&As in the five main crisis-hit countries accounted for more than 60% of the total Asian M&As in 1998–1999. Since cross-border M&As have become an important element in the expansion of the international production system, there is a need for a better understanding of what impact they have on countries, especially the host country (UNCTAD, 2000). A great omission in the literature on M&As and their performance is that too little attention is given to cultural clashes, while many mergers faced enormous problems due to cultural differences. Bankers Trust and Deutsche Bank, Upjohn and Pharmacia, and Daimler Chrysler are good examples. As put by Mr. Hubert of Daimler Chrysler, “We are absolutely happy with the development of the merger, we have a clear understanding: one company, one vision, one chairman, two cultures.” This is also due to the fact that in most M&As more attention is given to the financial and legal aspects of the deal rather than to the integration of the companies involved. In many cases, the merged company ends up with two bosses, two accounting systems and neither partner really know what the other is doing (The Economist, Jan. 9, 1999). The literature so far (see e.g. Cooper & Gregory, 2001; Buckley & Ghauri, 2002) contains many fascinating insights from experienced M&A executives. M&As possess several important advantages: speed of entry and speed to market, the ability to acquire critical assets, platforms for future growth, and entry into new geographical markets. Speed of entry is felt to be particularly critical in the internet age (UNCTAD, 2000). At the same time, disadvantages are significant: asset values go down as well as up, cultural barriers in international acquisitions often mean that M&As do not replace internal growth and alliances. However, the failure of M&As is often very visible whereas startups and new venture failures often go unnoticed. Industry differences are obviously important – it may be that where the purpose of the M&A is to cut costs, these benefits are more obvious and
226
PERVEZ N. GHAURI AND PETER J. BUCKLEY
quantifiable than revenue enhancing strategies, where there may be more risk. R&D synergies are often to be the major reason for M&As, as in pharmaceuticals. Keeping the focus on the acquisition of products and technologies rather than companies may be important. Venture capital operations in biotechnology, for instance, are alternatives to M&As. This discussion and the views on the need to integrate internet start-ups strongly suggest that M&A strategy should be part of an overall strategy, and not separate from it. Further research is needed to fully understand the impact of Mergers and Acquisitions on companies and societies. M&As can be approached from a number of perspectives (Buckley & Ghauri, 2002): (1) (2) (3) (4) (5)
the market for corporate control; transaction cost theory; the resource based view of the firm; the impact of national and organizational cultures; processes: pre- and post-acquisition strategies;
There are a number of key constituencies whom M&As directly impact: (1) (2) (3) (4)
the acquiring firm including managers, workers, and its shareholders; the acquired firm including managers, workers, and its shareholders; potential rival acquirers and non-acquired firms; in international M&As, the host country and source country (competitive environment) and welfare impacts.
REFERENCES Aiello, R. J., & Watkins, M. D. (2000). The fine art of friendly acquisition. Harvard Business Review, 78(6), 101–107. Alchian, A. A., & Demsetz, H. (1972). Production, information and economic organisation. American Economic Review, 62, 777–795. Aliber, R. Z. (1970). A theory of direct foreign investment. In: C. P. Kindleberger (Ed.), The International Corporation. Cambridge, Mass: MIT Press. Aliber, R. Z. (1971). The multinational enterprise in a multiple currency world. In: J. H. Dunning (Ed.), The Multinational Enterprise, London: George Allen & Unwin. Ashkenas, R. N., & Francis, S. C. (2000). Integration managers: Special leaders for special times. Harvard Business Review, 78(6), 108–116. Baumann, H. G. (1975). Merger theory, property rights and the pattern of U.S. direct investment in Canada. Weltwirtschaftlices Archiv., 11(4), 676–697. Berkovitch, E., & Narayanan, M. P. (1993). Motives for takeovers: An empirical investigation. Journal of Finance and Quantitative Analysis, 28(3), 347–362. Bert, T., & Trait, N. (2000), Kerkorian versus Schrempp may add to Chrysler’s woes. Financial Times, (November 28), 24.
International Mergers and Acquisitions: Past, Present and Future
227
Bild, M. (1998). Valuation of Takeovers. Ph.D. dissertation, Stockholm School of Economics, Stockholm: Elander Gotalo. Bradley, M., Desai, A., & Kim, E. H. (1988). Synergistic gains from corporate acquisitions and their division between the stockholders of target and acquiring firms. Journal of Financial Economics, 21(1), 3–40. Bresman, H., Birkinshaw, J., & Nobel, R. (1999). Knowledge transfer in international acquisitions. Journal of International Business Studies, 30(3), 439–462. Brouthers, K. D., & Brouthers, L. E. (2000). Acquisition or greenfield start-up? Institutional, cultural and transaction cost influences. Strategic Management Journal, 21, 89–97. Buckley, P. J., & Casson, M. (1976). The future of the multinational enterprise, London: Macmillan. Buckley, P. J., & Casson, M. (1998). Models of the multinational enterprise. Journal of International Business Studies, 29(1), 21–44. Buckley, P. J., & Ghauri, P. N. (Eds) (1999). The internationalization of the firm: A reader. London: Thompson Buckley, P. J., & Ghauri, P. N. (Eds) (2002). International Mergers and Acquisitions: A Reader. London: Thompson. Buono, A. F., Bowditch, J. L., & Lewis, J. W. (1985). When cultures collide: The anatomy of a merger. Human Relations, 38(5), 477–500. Buono, A. F., & Bowditch, J. L. (1989). The human side of mergers and acquisitions: Managing collisions between people and organizations, San Francisco: Jossey Brass Chatterjee, S., Lubatkin, M. H., Scheweiger, D. M., & Weber, Y. (1992). Cultural differences and shareholder value in related mergers, linking equity and human capital. Strategic Management Journal, 13, 319–344. Chi, T. (2000). Option to acquire or divest a joint venture. Strategic Management Journal, 21, 665–687. Child, J., Markocy, L., & Cheung, T. (1992), Management adaptation in Chinese and Hungarian strategic alliances with culturally different foreign partners, Working Paper, The Judge Institute Management Studies at St. John College, University of Cambridge, U.K. Coase, R. H. (1937). The nature of the firm. Economica, 4, 386–405. Cooper, C., & Gregory, A. (Eds) (2001). Advances in mergers and acquisitions. Connecticut: JAI Press. Cowling, K., Stoneman, P., Cubbin, J., Cable, J., Hall, G., Domberger, S., & Dutton, P. (1980). Mergers and economic performance. Cambridge: Cambridge University Press. Datta, D. K. (1991). Organizational fit and acquisition performance: Effects of post-acquisition integration. Strategic Management Journal, 12, 281–297. Demsetz, H. (1973). Industry structure, market rivalry, and public policy. The Journal of Law and Economics, 16, 1–3. Dickerson, A. P., Gibson, H. D., & Tsakalotos, E. (1997). The impact of acquisitions on company performance: Evidence from a large panel of U.K. firms. Oxford Economic Papers, 49, 344–361. The Economist (1999). How to merge (January 9, pp. 19–21). The Economist (2001). The great merger wave breaks (January 27, pp. 59–60). Eun, C. S., Kolodny, R., & Scheraga, C. (1996). Cross-border acquisitions and shareholder wealth: Tests of the synergy and internalisation hypotheses. Journal of Banking and Finance, 20, 1559–1582. Farrell J., & Shapiro, C. (1990). Horizontal mergers: An equilibrium analysis. American Economic Review, (March), 107–126.
228
PERVEZ N. GHAURI AND PETER J. BUCKLEY
Franks, F. R., Broyles, J. E., & Hecht, M. J. (1977). An industry study of the profitability of mergers in the United Kingdom. The Journal of Finance, XXXII(5), 1513–1525. Franks, J., & Mayer, C. (1993). European capital markets and corporate control. In: M. Bishop & J. Kay (Eds), European Mergers and Merger Policy. Oxford: Oxford University Press. Grossman S. J., & Hart, O. D. (1980). Takeover bids, the free-rider problem, and the theory of the corporation. Bell Journal of Economics, 11(1), 42–64. Halpern, P. J. (1973). Empirical estimates of the amount and distribution of gains to companies in mergers. Journal of Business, 46(4), 554–575. Haspeslagh, P., & Jemison, D. (1991). Managing acquisitions: Creating value through corporate renewal, New York: The Free Press. Hennart, J. F., & Reddy, S. (1997). The choice between mergers/acquisitions and joint ventures: The case of Japanese investors in the United States. Strategic Management Journal, 18, 1–12. Hofstede, G. (1980). Culture’s consequences: International differences in work related values. Beverly Hills: Sage Publications. Hughes, A. (1993). Mergers and economic performance in the U.K.: A survey of the empirical evidence 1950–90. In: M. Bishop & J. Kay (Eds), European Mergers and Merger Policy. Oxford: Oxford University Press. Kindleberger, C. P. (1969). American business abroad. New Haven: Yale University Press. Kogut, B. (1991). Joint ventures and the option to expand and acquire. Management Science, 37, 19–32. Kogut, B., & Singh, H. (1988). The effect of national culture on the choice of entry mode. Journal of International Business Studies, (Fall), 411–432. Kosnik, R. D., & Turk, T. A. (1994). The antecedents and consequence of resistence to takeovers. In: G. von Krogh, A. Sinatra & H. Singh (Eds), The Management of Corporate Acquisitions: International Perspectives (pp. 170–190). London: Macmillan. Lubatkin, M. (1983). Mergers and the performance of the acquiring firm. Academy of Management Review, 8(2), 218–225. Mandelker, G. (1974). Risk and return: the case of merging firms. Journal of Financial Economics, 1, 303–335. Magenheim, E., & Muller, D. C. (1988). Are acquiring firm shareholders better off after an acquisition? In: J. Coffee, L. Lowenstein & S. Rose-Ackerman (Eds), Knights, Radiers and Targets. Oxford: Oxford University Press. McManus, J. C. (1972). The theory of the international firm. In: G. Pacquet (Ed.), The Multinational Firm and the Nation State. Toronto: Collier-Mcmillan. Manne, H. G. (1965). Mergers and market for corporate control. Journal of Political Economy, 73(2), 110–120. Morck, R., Shleifer, A., & Vishny, R. W. (1988). Characteristics of targets of hostile and friendly takeovers. In: A. J Auerbach (Ed.), Corporate Takeovers: Causes and consequences (pp. 101–118). Chicago: University of Chicago Press. Puranam, P., Singh, H., & Zollo, M. (2000). Bringing some discipline to M&A mania. Financial Times: Mastering Management, (October 30), 2–4. Reuer, J. J., & Koza, M. P. (2000). On lemons and indigestibility: Resource assembly through joint ventures. Strategic Management Journal, 21, 195–197. Schenk, H. (2000). Are international acquisitions a matter of strategy rather than wealth creation? International Review of Applied Economics. Senge, P. (1968). The leader’s new work: Building learning organisations. Sloan Management Review, 7–23.
International Mergers and Acquisitions: Past, Present and Future
229
Seth, A., Song, K. P., & Pettit, R. (2000). Synergy, managerialism or hubris? An empirical examination of motives for foreign acquisitions of U.S. firms. Journal of International Business Studies, 31(3), 387–405. Sirower, M. I. (1997). The synergy trap: How companies lose the acquisition game. New York: Free Press. UNCTAD (2000). Cross-border mergers and acquisitions. World Investment Report. New York: United Nations. United States Department of Justice (2000). International Competition Policy Advisory Committee to the Attorney General and the Assistant Attorney General for Antitrust: Final Report, http://www.usdoj.gov/ Villinger, R. (1996). Post acquisition managerial learning in central East Europe. Organisation Studies, 17(20), 181–206. Von Krogh, G., Sinatra, A., & Singh, H. (1974). The management of corporate acquisitions: International perspectives. London: Macmillan. Williamson, O. E. (1968). Economics as anti-trust defence: The welfare trade-offs. American Economic Review, 58 (March), 18–36.