REITS and the Property Market (Journal of Property Investment and Finance, Vol. 25, No. 3, 2008)

The current issue and full text archive of this journal is available at www.emeraldinsight.com/1463-578X.htm Depreciati...

Author: Goodchild | Robin

150 downloads 755 Views 669KB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

The current issue and full text archive of this journal is available at www.emeraldinsight.com/1463-578X.htm

Depreciation, income distribution and the UK REIT

Depreciation, income and the UK REIT

Andrew Baum Department of Real Estate and Planning, University of Reading Business School, Reading, UK

Steven Devaney University of Aberdeen Business School, Aberdeen, UK

195 Received November 2006 Accepted April 2007

Abstract Purpose – The purpose of this paper is to consider prospects for UK REITs, which were introduced on 1 January 2007. It specifically focuses on the potential influence of depreciation and expenditure on income and distributions. Design/methodology/approach – First, the ways in which depreciation can affect vehicle earnings and value are discussed. This is then set in the context of the specific rules and features of REITs. An analysis using property income and expenditure data from the Investment Property Databank (IPD) then assesses what gross and net income for a UK REIT might have been like for the period 1984-2003. Findings – A UK REIT must distribute at least 90 per cent of net income from its property rental business. Expenditure therefore plays a significant part in determining what funds remain for distribution. Over 1984-2003, expenditure has absorbed 20 per cent of gross income and been a source of earnings volatility, which would have been exacerbated by gearing. Practical implications – Expenditure must take place to help UK REITs maintain and renew their real estate portfolios. In view of this, investors should moderate expectations of a high and stable income return, although it may well still be so relative to alternative investments. Originality/value – Previous literature on depreciation has not quantified amounts spent on portfolios to keep depreciation at those rates. Nor, to our knowledge, has its ideas been placed in the indirect investor context. Keywords Depreciation, Income, Real estate, Investments, Trusts, United Kingdom Paper type Research paper

Introduction Over the last three years, there has been serious debate regarding the introduction of a quoted tax-transparent real estate investment vehicle into the UK. This has culminated in the passing of legislation in the Finance Act 2006 that enables real estate investment companies to convert to tax-transparent status provided they meet certain rules. The new vehicles will be known as UK REITs. It has been argued that the introduction of UK REITs will increase investment in property (particularly by small investors), encourage regeneration, facilitate the supply of housing and enable assets currently held offshore for tax reasons to be brought back within the UK regulatory system (British Property Federation/Investment Property Forum/Royal Institution of This paper originated from a larger project on depreciation (Investment Property Forum, 2005a) that was commissioned and funded under the auspices of the IPF Educational Trust and IPF Joint Research Programme. The authors are grateful for the support of the IPF and IPF Educational Trust, and for the permission to reproduce empirical results first presented there.

Journal of Property Investment & Finance Vol. 26 No. 3, 2008 pp. 195-209 q Emerald Group Publishing Limited 1463-578X DOI 10.1108/14635780810871597

JPIF 26,3

196

Chartered Surveyors, 2004a). Whether any of these ambitions can be realised depends on the structure of the new vehicles and the success of their launch. Long-established examples of tax-transparent vehicles exist in the shape of Real Estate Investment Trusts (REITs) in the USA and the Australian Listed Property Trusts (LPTs). However, it was proposals for such vehicles in other European countries that helped accelerate the debate in the UK. In particular, the successful introduction in 2003 of Socie´te´s d’Investissements Immobiliers Cotees (SIICs) in France led to arguments in the UK about the need to maintain national competitiveness. Official proposals soon followed, with the Treasury publishing a consultation document in March 2004 (HM Treasury, 2004) and a proposed framework for a tax-transparent vehicle in the 2005 Budget (HM Treasury, 2005). Each of these invited feedback from the real estate industry and this subsequently helped shape the 2006 legislation, which enables UK REITs to appear for accounting periods beginning on or after 1 January 2007. The UK REIT structure should provide investors with returns that are closer to those of the direct property market in comparison with shares in current UK listed companies. This is because of the obligation on UK REITs to pay out a high proportion of taxable income, the removal by government of “double taxation” on that income and other rules about their activities and investments. Therefore, it is likely that they will be popular with investors, especially in the case of commercial real estate, which in recent years has seen a rising volume of direct investment and the launch of many new unlisted products. In particular, it is perceived that commercial real estate offers investors a stable and relatively high income return, as well as having other advantages such as its role in portfolio diversification (Investment Property Forum, 2005b). However, it is also well recognised that property investments depreciate in value through time and require expenditure to maintain their condition and income earning potential. The argument of this paper is that these costs may be higher than generally appreciated, because significant expenditure is required to maintain and renew real estate portfolios. For indirect investors, this is important because UK REIT distributions are set at a high proportion of net income, i.e. income after property related expenditure and other costs such as debt interest. Hence, high and stable rental cash flows may not translate into a high and stable dividend income stream. This proposition is tested empirically within the paper. The rest of the paper is organised as follows. First, it discusses how depreciation affects properties and is thus able to impact on the earnings and value of property investment vehicles. The paper then sets out the specific context of the UK REIT. Following this, the experience of US REITs with respect to expenditure and distributions is examined, but differences between US and UK accounting regimes and markets mean that the main analysis is of historic cash flow data for UK portfolios. The implications of depreciation and expenditure for UK REIT investors are then summarised in the conclusion. The effects of depreciation Depreciation can affect the performance and value of real estate vehicles in a number of ways. This section sets out how it affects commercial real estate directly and then considers how, in turn, depreciation influences real estate vehicle performance and management.

Depreciation and direct property investment In the context of real estate, depreciation has been defined by Law (2004) as follows: . . . the rate of decline in rental [capital] value of an asset (or group of assets) over time relative to the asset (or group of assets) valued as new with contemporary specification (Law, 2004, p. 242).

This definition relates to the economic depreciation of assets, with which this paper is primarily concerned. This is distinct from depreciation as used in corporate accounting, i.e. a method of reducing the book value of assets through time to reflect in company profits the consumption of capital assets that will need to be replaced. In the case of the UK, such methods have not been allowed for properties held as investments, which must be shown in accounts at market value[1]. Whilst this clarifies the concept underlying the discussions that follow, the definition does not indicate the causes of such depreciation, which it is helpful to briefly consider. In particular, Baum (1991) sets out following causes of depreciation in commercial property values: . Physical deterioration. This relates to the wearing out of the building through time. . Building obsolescence. This refers to changes in what is expected from buildings, the most obvious source of which are changes in technology that impact upon occupier requirements and may render a building’s design or configuration redundant. . Site value changes. These result from changes in planning, the local environment and supply and demand within sub-markets[2]. Technically, they are excluded under the definition of depreciation given above, but they also impact upon values and are, in practice, extremely difficult to exclude from any measurement of depreciation. These factors can affect rental values and capital values, as they will influence the rent that a potential tenant would be prepared to pay and they will have consequences for future cashflows either in the form of reduced rents and/or capital expenditure in order to combat their effects. Several studies have attempted to measure the impact of depreciation on property values. Those conducted for the UK commercial real estate market include CALUS (1986), Baum (1991, 1997), Jones Lang Wootton (1987), Barras and Clark (1996) and College of Estate Management (1999). The results of these studies are summarised in Investment Property Forum (2005a), which itself also conducts a measurement exercise. Whilst the rates found are interesting, Investment Property Forum (2005a) noted that all of these results were measured after the effects of expenditure. In other words, they display a “net” rate, which does not reflect the efforts by investors to restrict depreciation to that rate. The full cost of depreciation includes expenditure needed to maintain and renew the properties in a portfolio. Later in this paper, expenditure is examined in detail so that the effects of depreciation on the income and distributions of UK REITs can be better understood. Depreciation and indirect property investment Rental depreciation creates the most basic effect on vehicle performance by reducing income received from properties through time. This can be through falls in property

Depreciation, income and the UK REIT 197

JPIF 26,3

198

rental values or through slower rental growth compared to benchmarks, which may make a vehicle perform less well than its competitors. Depending on the country and sector, lease structures may be able to protect investors from rental depreciation to some extent, but eventually property rents will need to adjust either at rent reviews or re-letting. The implications are that unless rental depreciation is tackled through renewal of the property or portfolio, earnings growth will be affected, with knock-on impacts on dividends and equity valuation. Addressing rental depreciation is likely to require expenditure on properties. A certain amount of spending may be needed before a building can be re-let, while other repairs and improvements can be important in maintaining or improving rental value, particularly in the context of competition from other properties and schemes. This spending will have an effect on dividends too, as income will be reduced. However, it is worth noting that the consequence for future earnings and dividends of not undertaking expenditure could be greater than the present cost. Both rental depreciation and expenditure requirements influence the yields and capital values of vehicle assets. For instance, where depreciation is being caused by functional obsolescence, it may make a building of less or no value to occupiers now (reducing ERV), require expenditure in order to put right (reducing future income) and mean that future cash flows are more uncertain (causing the yield to rise, as the property is more risky). This will then feed through into company net asset value (NAV). It will also affect vehicle share price, irrespective of whether prices are related directly to NAV, as is the case for current UK quoted property companies, or to the future expected cash flows of the firm, as is the case for US REITs. So these relatively simple property-level effects can have far-reaching implications for the vehicles that own properties and the investors holding the shares or units. The size of the impact will depend on many factors, including the size of the vehicle, the nature of the properties it owns, the particular causes of depreciation at any one time, and a vehicle’s structure and flexibility to deal with such risks. This last point is particularly significant for tax-transparent vehicles, where distribution rules may constrain expenditure to combat depreciation or where other restrictions may prevent declining properties from being redeveloped or traded out of the portfolio. Current and future earnings can be affected, as well as the realisable value of the asset base. UK rules on income distribution In considering the possible impact of depreciation on UK REITs, it is now important to set out the rules under which they must operate. These include requirements to: . Be a “property rental business”, with 75 per cent of income arising from that business and 75 per cent of assets dedicated to it (Finance Act, s 108). However, as long as these tests are satisfied, a UK REIT may carry out taxable ancillary activities, which can include property (re-)development. . Distribute at least 90 per cent of the profits of the property rental business (Finance Act, s 107). This does not include capital gains from selling property, though, which can either be distributed or reinvested in the portfolio (Finance Act, s 118). The qualifications to each requirement are important, as they appear to give UK REITs flexibility to manage depreciation and renew their portfolios. Also important is the fact that distributions will be from profits rather than gross income. It had been initially

suggested by government that UK REITs should distribute at least 90 per cent of income “before depreciation” (HM Treasury, 2004, p. 21), but this was later recognised as unsuitably high, with 95 per cent and then 90 per cent of profits proposed instead. While these steps have, on the face of it, reduced the level of dividends to investors, they were critical for the proper management of assets. Otherwise, UK REITs could have been left with insufficient cash to maintain properties and could thus have been forced to defer expenditure, sell assets or make regular capital issues (British Property Federation/ Investment Property Forum/Royal Institution of Chartered Surveyors, 2004b; Lizieri, 2004), with potentially serious impacts on vehicle performance and success. The rules address concerns raised during the consultation process about the ability of the UK REIT to manage depreciation. However, questions remain about what the effect of depreciation on vehicle returns might be, particularly with regard to distributions. To answer these, a natural step would seem to be to explore the experiences of REIT structures in other countries, particularly those of the long-established US REIT. Hence, the paper now examines the US situation as a prelude to analysis of UK portfolio data, but, in doing so, it highlights a number of reasons why UK REITs will differ. Distributions – the US experience US REITs are one of the longest standing and most researched tax-transparent vehicles in existence. There are a large number of articles in the academic literature relating to them, but it should be noted that much of this literature is focused on topics such as performance, diversification and the nature of returns – to what extent US REITs are real estate and to what extent stocks (see reviews of the literature in Corgel et al., 1995; Worzala and Sirmans, 2003). Structural aspects, such as distribution rules, receive less attention despite their influence on returns and investor attractiveness. However, there are some studies on US REIT distribution policy from which insights can be gained and which provide a link into the wider financial literature on dividends and earnings retention. REIT income distributions When real estate investment trusts were introduced in the USA in 1960, a high compulsory income distribution level of 90 per cent was set, later increased to 95 per cent. These distributions were to be made from net income after the deduction of expenses, interest and a depreciation allowance, recognising that: . real estate vehicles need the ability to retain some earnings through which reinvestment and renewal of the stock can be made; and . in a particular year, a real estate owner can be faced with significant costs to repair and maintain buildings for existing or prospective occupiers. More recently, in the REIT Modernisation Act 1999, the compulsory distribution level was reduced back to 90 per cent of net income. Yet the change did not appear to have a big impact on US REIT values, although Howe and Jain (2004) found positive share price reactions to the Act as a whole. The reason for this limited impact was that REIT dividends are only partially constrained by the minimum distribution requirement. As noted by Campbell and Sirmans (2002), the average payout by US REITs is often over 100 per cent of accounting earnings. In other words, not only are dividends paid in excess of the compulsory level, but some payouts are also above recorded net income. This is confirmed by the studies of Wang et al. (1993) and Bradley et al. (1998), with the

Depreciation, income and the UK REIT 199

JPIF 26,3

200

former reporting an average payout ratio of 1.65, where payout ratio is the ratio of dividends to reported net income. This was compared to an expected 0.95 if regulation were the only driver. Meanwhile, for 2003, Investment Property Forum (2005a) reported a similar average payout ratio of 1.64, whereas regulation alone would point to an average of around 0.9. The reason for these high proportions is that there is a big difference between reported net income and the net cash flow that an equity (property owning) REIT has available to distribute. This difference is created by large provisions for depreciation and amortization, associated with a cost treatment of commercial real estate assets in accounts. Such provisions are not available to mortgage REITs, and data from Investment Property Forum (2005a) shows that these have an average payout ratio of 0.91, close to the statutory limit. In the US REIT industry, it is recognised that the accounting measure of net income is thus not a useful measure of equity REIT profitability (Yungmann and Taube, 2001). Therefore, other measures have been developed, such as funds from operations (FFO), adjusted funds from operations (AFFO) and funds available for distribution (FAD). While these different metrics offer alternative ways of assessing REIT earnings, there is a lack of consistency in how they are calculated between firms. In the absence of uniform measures, it becomes difficult to assess how much of the gross income from properties is required for US REITs to operate and how much can be paid out. Furthermore, the difference between net income and net cash flow gives US REITs flexibility to pursue distribution policies, the reasons for and motivations behind which are not straightforward. For instance, Wang et al. (1993) suggest several hypotheses to explain the high payout ratios observed and differences between the ratios of different firms. One strand of argument is rooted in agency cost theory and states that shareholders prefer managers to return cash flows rather than keep them within the company. Then, when future decisions have to be taken, such as whether to undertake a major refurbishment, managers must submit their proposals to the capital markets since there are no retained funds. Therefore, greater monitoring of management by the investors can take place. However, in examining differences between firms, Bradley et al. (1998) argued that their results were inconsistent with agency explanations. They related dividend ratios and policies to the anticipated volatility in future cash flows, as proxied by variables such as firm size, gearing and portfolio diversification. Firms with higher volatility might be expected to have higher agency costs and thus pay out more than other firms, but this was found not to be the case. The alternative explanation put forward by the authors was that dividends are used to convey information about expected earnings, with higher payouts signalling more confidence in future cash flows. Differences between the UK and the USA Whether UK REITs will encounter a similar situation and be able to pursue such policies is a matter of interest, but it can be quickly shown that this is unlikely to be the case. Earlier it was stated that UK REITs would be required to show their properties at market value in their accounts. Under such a regime, no allowance for depreciation is made to offset against taxable income; values change with market conditions rather than being written down each year. This difference from the US treatment means that the distributable profits of a UK REIT will be much closer to its cash flow, and so there is no decision to be made about whether some or all of an allowance should be retained.

Even if accounting differences were not a factor, other differences between the countries exist that may influence distribution and retention decisions. In particular, a key difference lies in the nature of the leases granted in each country. The terms and conditions of leases granted will determine whether the REIT or the tenant is responsible for repairs and maintenance. This, in turn, not only influences the pattern of income and expenditure, but potentially also the extent and amount of depreciation in the portfolio (Baum and Turner, 2004). For many years, the following lease terms were typical for prime property space in the UK. Leases were agreed for long periods, often 25 years, with five-year rent review intervals and upward-only rent reviews as standard. In addition, repairing and insuring costs were passed on to tenants through full repairing and insuring (FRI) clauses. Although in recent years leases have become shorter and opportunities to break have increased, these repairing and rent review provisions still predominate (Crosby et al., 2005). In the USA, leases are shorter on average[3] and more of the repairing obligations are borne by the landlord. This means that, in terms of tackling depreciation, there are more opportunities for the US owner to actively manage its buildings and more incentive to do so owing to the need to achieve re-lettings more often. In contrast, UK leases have tended towards more passive management of the stock. Responsibility for regular maintenance to combat physical deterioration is passed to the tenant, especially in single-let buildings, but there is no guarantee that the tenant will perform these obligations in the same way and, often, they are discharged through payment of a dilapidations charge at the end of the lease instead. While, in theory, this compensates the landlord for lost value, the impact of not performing work when necessary may mean greater depreciation and cost overall. More important from the perspective of this paper are the implications that this has for income returns on the vehicle. The different lease terms mean differences in the nature of the income received (see Kennedy et al., 2004). In the USA, there is a wide difference between the gross income received and the net operating incomes (NOI) from which distributions are made. In the UK, gross income and NOI are very close together. However, the restrictions on tackling depreciation caused by the UK lease structure may mean that this small difference between gross income and NOI is artificial, with income returns being higher than they should be at the expense of capital return in the short term. An alternative angle offered by Baum and Turner (2004) is that under the shorter and more flexible (in this case, US) leases, more reinvestment in the stock is taking place, which suggests better total returns in the long run, though this is very difficult to prove empirically at present. Hence, the experiences of the USA and indeed other jurisdictions are only of limited value when assessing the likely returns from the UK REIT because of both accounting and property market differences. Yet, important insights can be drawn from the differences. In particular, a failure to appreciate leasing differences and their effects on income return could put a UK REIT at risk of over-distribution in the short term at the expense of its long-term capital value. Income from and expenditure on UK real estate An alternative approach to exploring the impact of depreciation and property expenditure on distributable earnings is to look directly at the cash flows of UK

Depreciation, income and the UK REIT 201

JPIF 26,3

202

property portfolios. In effect, the analysis examines the portfolio as if it had been that of a UK REIT operating over the period in question. As noted already, depreciation can affect portfolio income in two particular ways: (1) it can influence income received through its effect on rental values; and (2) it can influence net income through its impact on expenditure to maintain and renew the asset base. Whilst several previous studies have tried to measure the influence of depreciation on values (especially rental values), neither of the above aspects has been explored directly in the property investment literature. In this analysis, both gross and net income through time for a held sample of properties is presented, where net income is net of expenditure, but before debt interest. Data and method The analysis draws on a dataset created for Investment Property Forum (2005a) from the UK database of IPD, a commercial organisation that provides performance measurement and benchmarking services to real estate investors. It comprises a sample of 624 properties that were held over the period 1984-2003 (19 years). The sample provides a long run of income and expenditure data and is spread across different property types and locations. It also includes buildings of different ages as at the start of the period and so simulates a diversified portfolio held for 19 years. The data relates to institutional grade property, since contributors to IPD in the UK are, predominantly, insurance companies, pension funds and publicly listed property companies. However, it would be expected that UK REIT portfolios would be comprised of institutional-grade property as well. As the analysis refers to a held portfolio of properties, the potential influence of trading assets is not reflected. A vehicle may be able to improve its income and returns, and avoid large costs, by replacing and renewing its portfolio of properties. While this means that the possible benefit of new purchases does not show through in the figures above, this is counterbalanced by the fact that there are no retirements in the data either (buildings that may have had very bad performance, but which were sold out rather than held). Another potential feature is survivor bias, with the sample reflecting properties retained by their owners because of superior performance. This was tested in Investment Property Forum (2005a, pp. 50-2) and some out-performance relative to a control sample of properties was noted, being 1 per cent per annum in terms of total returns. For all properties in the sample, records are held on the amount of income received and the amount of regular expenditure paid out by owners, including the payment of property management fees, as well as the amounts of any irrecoverable capital expenditure that takes place. However, work undertaken by tenants in performance of lease obligations or costs that are reimbursed to the owner through dilapidations payments or service charges are not recorded. Whilst these do not affect the earnings of the owner, it does mean that the expenditure recorded probably understates the true running costs of the properties. In addition, if the practical consequences of such arrangements are that under-investment takes place in the long run, then any long-term impacts on earnings will be masked. Nonetheless, from these records, the gross and net income for the sample of properties can be calculated. The margin between the two can then be examined,

giving investors an insight into how much of the income is likely to be required to meet the maintenance and refurbishment needs of a UK REIT portfolio. However, it must be borne in mind that, while expenditure may alleviate depreciation, it cannot eliminate it entirely. There is a well-recorded difference between curable and incurable depreciation (Baum, 1994) and redevelopment or renewal of the stock may be required even if expenditure has been applied to the curable depreciation element. For each year over the period 1984-2003, the total income for all the properties was computed, as well as two measures of net income. These were: Net income ¼ total income 2 revenue expenditure; Net cash flow ¼ total income 2 ðrevenue þ capital expenditureÞ: Total income is mostly the rent receivable on the sample, although there are small amounts of other property-related income and occasional instances of capital receipts, which have also been included. Expenditure is then divided between “revenue” and “capital” expenditure. It is classified as revenue expenditure if the spending by the owner is for the regular management of the properties. It is classed as capital expenditure where funds are for the refurbishment or improvement of the property. IPD also record “development” expenditure, where cash flows would be entered if any property had been redeveloped. However, while the measurement sample contains refurbishments, redevelopments are not included. So the analysis reflects revenue and capital expenditure to combat curable depreciation, but not the complete replacement of buildings which may be needed to solve incurable factors. This also points to understatement of property/portfolio running costs and indicates that the net income figures calculated here should be taken as an upper bound, since, clearly, this redevelopment activity would impact earnings further. As it is based directly on property records, the analysis also excludes the impact of gearing. Results The results of the income and expenditure investigation for all properties are shown in Table I. Further results for the shop, office and industrial sectors of the UK commercial property market are then summarised in Table II, with full results presented in the Appendix. Shops are considered rather “All retails” owing to the small number of assets in the “Retail warehouse” and “Shopping centre” segments that were held continuously over the period in question. One of the most striking things about the figures in Table I is the rise in total income over the period concerned. The selected period spans a complete property cycle in the UK of boom, recession and recovery, but whilst decline in the property market had set in by 1990, total income on this set of properties does not fall until 1995, and then only by a small amount. This illustrates the effect of UK lease structures in protecting from both depreciation and market conditions over the short- and medium-term and, hence, it is difficult to draw conclusions about the former from the total income series[4]. Figure 1 further illustrates this effect, displaying in real terms the changes in income against a market rental growth series. It is also interesting to note that the real total income of the sample fails to grow from 1993 onwards. The analysis also shows the proportions of income taken up by different categories of expenditure. Around 10 per cent of the gross income in each year was used for revenue expenditure, a proportion that has grown slightly but has generally been stable over the period. In the case of capital expenditure, though, the proportion varies

Depreciation, income and the UK REIT 203

JPIF 26,3 Year

204

Table I. Income and expenditure for portfolio of 624 properties held over the period 1984-2003 (figures in millions)

Table II. Sample sizes, rental depreciation and net cash flow by property type

1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Average

Total income

IPD defined revenue expenditure

Net income

91.3 98.4 103.4 117.1 131.5 155.1 181.7 208.7 216.6 226.9 227.4 224.4 232.3 242.9 236.0 243.1 250.4 268.7 280.7 287.6

7.5 7.0 6.6 7.0 7.8 10.1 14.1 18.4 22.0 25.7 23.5 23.9 22.3 21.4 23.6 21.2 19.8 21.5 25.9 25.9

83.7 91.5 96.8 110.1 123.7 145.0 167.5 190.2 194.6 201.2 203.9 200.5 210.0 221.4 212.5 221.9 230.6 247.1 254.8 261.7

As a percentage of gross income

IPD defined capital expenditure

Net cash flow

As a percentage of gross income

92 93 94 94 94 93 92 91 90 89 90 89 90 91 90 91 92 92 91 91 91

12.7 17.3 12.2 13.9 19.1 47.4 12.9 15.4 8.2 7.6 29.2 41.0 36.5 13.4 14.1 28.5 33.3 36.6 25.0 17.5

71.1 74.2 84.6 96.3 104.6 97.6 154.7 174.8 186.4 193.6 174.7 159.5 173.5 208.0 198.4 193.4 197.3 210.6 229.8 244.2

78 75 82 82 80 63 85 84 86 85 77 71 75 86 84 80 79 78 82 85 80

Sample Rental depreciation per annum (per cent) Net income as a percentage of gross income (average) Net cash flow as a percentage of gross income (average) Minimum reduction Maximum reduction

Shops

Offices

Industrials

339 0.1 92 81 90 67

165 1.0 90 75 88 42

120 0.6 94 86 95 66

Source: Depreciation rates taken from Investment Property Forum (2005a)

quite significantly between years, with it accounting for around 5 per cent of gross income in some years, but up to 30 per cent in 1989[5]. This means that net cash flow varies considerably around its average level of 80 per cent of gross income. High levels of gearing could then further exacerbate this net income volatility. At a sector level, some differences between property types emerge. Table II shows that offices have received the most expenditure out of income and industrial properties the least over the period. The perhaps surprising place of industrials in this ranking is probably related to the type of industrial properties held by institutional investors – modern light industrial and warehousing premises rather than factories and sites for heavy industry. Meanwhile, offices have not only required most expenditure, but have also shown the greatest volatility in expenditure and net cash flow. Twice during the period, the net cash flow of the office properties was less than 50 per cent of their gross income. Yet, despite this higher rate of spending, offices have experienced the highest

Depreciation, income and the UK REIT 205 Figure 1. Total income from the study sample compared with rental growth: 1984-2003

rates of (post-expenditure) rental depreciation, a result found by Investment Property Forum (2005a), Baum (1991) and the College of Estate Management (1999). The sector balance of a UK REIT can, therefore, further influence its distributable income and its exposure to depreciation. However, before concluding that an office REIT will be most severely affected of all, it should be remembered that depreciation may strike property types differently in the future to how it has done in the past. Conclusions This paper has set out to explore issues relating to the effects of depreciation on the income and distributions of a UK REIT. Such considerations are important in assessing how they will perform and whether the UK REIT market is likely to be successful. The rules for UK REITs announced in the Finance Act 2006 appear to give them flexibility in portfolio and, hence, depreciation management. However, depreciation will still affect the returns of such vehicles through its effects on income received and cash flow, the extent of the impact depending on factors such as the size of the vehicle, the properties it holds and the causes of depreciation at any one time. Investors in and analysts of UK REITs should be interested in these effects because of the close link between cash flow and distributions in a REIT structure. Their potential impact may be explored through reference to the experience of other vehicles, but in looking to the USA for lessons, where REITs have been established for many years, few conclusions can be drawn. This is because of differences in accounting regimes which, for US equity REITs, can cause net income to be much lower than net cash flow and so allow more discretion in distribution policy. Even if these did not exist, though, structural differences between the two markets may make UK properties more prone to building depreciation, although greater restrictions in land supply may, conversely, support the values of their sites. Therefore, this paper has also examined income and expenditure on UK properties to see what lessons might be learned. Although the dataset has a number of limitations, these are such that the results are likely to understate rather than overstate the costs of depreciation for a portfolio. The analysis demonstrates the importance of expenditure on properties, hitherto left out of empirical depreciation measurement in the property literature. Capital expenditure, in particular, reduces income available for distribution

JPIF 26,3

206

and could be a source of volatility from year to year, making “high and stable” cash flow perhaps less stable than investors expect. Notes 1. Although International Accounting Standards outline both cost and fair value approaches to accounting for property investments (International Accounting Standards Committee, 2000), the UK government has specifically legislated that UK REITS must use the fair value (i.e. market value) approach (Finance Act 2006, section 107(6)), consistent with previous UK rules for investment property (Institute of Chartered Accountants in England and Wales, 1981). 2. It is, of course, possible for these changes to be positive or negative. 3. Devaney et al. (2007) show average US lease terms in 2002 to be 5.8, 5.0 and 3.6 years for the retail, office and industrial sectors, compared to UK lengths in 2002 of 9.8, 7.6 and 7.3 years, respectively (British Property Federation/Investment Property Databank, 2004). These are equal- rather than value-weighted averages as no US value-weighted figures were available to the authors. 4. There are several analytical difficulties in assessing the effect of depreciation on gross income. It has a different profile to rental values because of issues such as vacancies and lease structures. Rental depreciation rates will not, therefore, feed immediately though into income flows. Meanwhile, it is unclear how an appropriate income growth benchmark could be constructed. 5. These fluctuations show some cyclical relationship, the correlation with year on year changes in rental growth being 0.49. References Barras, R. and Clark, P. (1996), “Obsolescence and performance in the Central London office market”, Journal of Property Valuation and Investment, Vol. 14 No. 4, pp. 63-78. Baum, A. (1991), Property Investment Depreciation and Obsolescence, Routledge, London. Baum, A. (1994), “Quality and property performance”, Journal of Property Valuation and Investment, Vol. 12 No. 1, pp. 31-46. Baum, A. (1997), Trophy or Tombstone? A Decade of Depreciation in the Central London Office Market, Lambert Smith Hampton and HRES, London. Baum, A. and Turner, N. (2004), “Retention rates, reinvestment and depreciation in European office markets”, Journal of Property Investment & Finance, Vol. 22 No. 3, pp. 214-35. British Property Federation/Investment Property Databank (2004), BPF IPD Annual Lease Review, British Property Federation, London. British Property Federation/Investment Property Forum/Royal Institution of Chartered Surveyors (2004a), A Guide to Real Estate Investment Trusts – And Why We Need Them, British Property Federation/Investment Property Forum/Royal Institution of Chartered Surveyors, London. British Property Federation/Investment Property Forum/Royal Institution of Chartered Surveyors (2004b), Industry Response to HMT & Inland Revenue Consultation Document, British Property Federation/Investment Property Forum/Royal Institution of Chartered Surveyors, London. Bradley, M., Capozza, D.R. and Seguin, P.J. (1998), “Dividend policy and cash-flow uncertainty”, Real Estate Economics, Vol. 26 No. 4, pp. 555-80. CALUS (1986), Depreciation of Commercial Property, College of Estate Management, Reading.

Campbell, R. and Sirmans, C. (2002), “Policy implications of structural options in the development of Real Estate Investment Trusts in Europe: lessons from the American experience”, Journal of Property Investment & Finance, Vol. 20 No. 4, pp. 388-405. College of Estate Management (1999), The Dynamics and Measurement of Commercial Property Depreciation in the UK, College of Estate Management, Reading. Corgel, J.B., McIntosh, W. and Ott, S.H. (1995), “Real estate investment trusts: a review of the financial economics literature”, Journal of Real Estate Literature, Vol. 3 No. 1, pp. 13-43. Crosby, N., Hughes, C. and Murdoch, S. (2005), Monitoring the 2002 Code of Practice for Commercial Leases, Office of the Deputy Prime Minister, London. Devaney, S.P., Lee, S.L. and Young, M.S. (2007), “Serial persistence in individual real estate returns in the UK”, Journal of Property Investment & Finance, Vol. 25 No. 3, pp. 241-73. HM Treasury (2004), Promoting More Flexible Investment in Property: A Consultation, March, HM Treasury and Inland Revenue, London. HM Treasury (2005), UK Real Estate Investment Trusts: A Discussion Paper, March, HM Treasury and Inland Revenue, London. Howe, J.S. and Jain, R. (2004), “The REIT Modernisation Act of 1999”, Journal of Real Estate Finance and Economics, Vol. 28 No. 4, pp. 369-88. International Accounting Standards Committee (2000), International Accounting Standard IAS 40: Investment Property, International Accounting Standards Committee, London. Institute of Chartered Accountants in England and Wales (1981), Statement of Standard Accounting Practice No. 19: Accounting for Investment Properties, Institute of Chartered Accountants in England and Wales, London. Investment Property Forum (2005a), Depreciation in Commercial Property Markets, Investment Property Forum/IPF Educational Trust, London. Investment Property Forum (2005b), Understanding Commercial Property Investment: A Guide for Financial Advisers, 2005 edition, Investment Property Forum, London. Jones Lang Wootton (1987), Obsolescence: The Financial Impact of Property Performance, Jones Lang Wootton, London. Kennedy, P., Haddock, M. and Sauer, A. (2004), “European capitalisation rates – towards a consistent approach”, paper presented at the PREA Conference, Los Angeles, CA, October. Law, V. (2004), “The definition and measurement of rental depreciation in investment property”, unpublished PhD dissertation, University of Reading, Reading. Lizieri, C. (2004), “Promoting more flexible investment in property: response to HM Treasury/Inland Revenue consultation paper”, July, University of Reading, Reading. Wang, K., Erickson, J. and Gau, G.W. (1993), “Dividend policies and dividend announcement effects”, Journal of the American Real Estate and Urban Economics Association, Vol. 21 No. 2, pp. 185-201. Worzala, E. and Sirmans, C.F. (2003), “Investing in international real estate stocks: a review of the literature”, Urban Studies, Vol. 40 Nos 5/6, pp. 1115-49. Yungmann, G. and Taube, D. (2001), “FFO – earnings or cash flow?”, Real Estate Portfolio, National Association of Real Estate Investment Trusts, Washington, DC.

Further reading CB Richard Ellis (2006), Prime Rent and Yield Monitor, CB Richard Ellis, London.

Depreciation, income and the UK REIT 207

JPIF 26,3

Appendix. Sector level income and expenditure results

Year

208

Table AI. Income and expenditure for portfolio of shop properties held over the period 1984-2003 (figures in millions)

1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Average

Year

Table AII. Income and expenditure for portfolio of office properties held over the period 1984-2003 (figures in millions)

1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Average

Total income

IPD defined revenue expenditure

Net income

29.2 27.6 28.0 33.4 36.7 46.1 55.2 62.0 66.6 68.3 69.4 70.5 71.5 72.4 76.5 79.4 83.4 88.5 92.1 94.7

1.5 2.0 1.7 1.9 2.2 3.7 4.0 4.3 5.9 6.2 8.1 8.8 7.2 6.3 6.6 6.3 6.4 7.5 7.8 7.6

27.7 25.6 26.4 31.5 34.5 42.4 51.2 57.8 60.7 62.0 61.3 61.7 64.3 66.0 69.9 73.1 77.0 81.0 84.3 87.1

Total income

IPD defined revenue expenditure

Net income

39.0 44.2 46.9 53.7 62.0 72.5 85.0 99.2 100.3 106.1 104.7 99.4 104.5 107.3 103.7 108.0 109.9 119.1 123.1 129.7

4.2 3.5 3.6 4.0 4.1 4.8 8.2 11.5 11.7 16.5 12.3 11.8 10.6 12.4 13.4 10.8 10.0 10.2 13.1 13.2

34.8 40.7 43.3 49.6 57.9 67.7 76.8 87.7 88.7 89.7 92.4 87.6 94.0 94.9 90.3 97.3 99.9 108.9 110.0 116.5

As a percentage of gross income

IPD defined capital expenditure

Net cash flow

As a percentage of gross income

95 93 94 94 94 92 93 93 91 91 88 88 90 91 91 92 92 92 92 92 92

3.4 4.0 6.1 4.5 9.2 9.7 3.4 6.8 4.6 0.3 3.6 0.6 1.8 5.6 9.1 6.7 21.3 7.3 10.4 5.1

24.4 21.6 20.3 27.0 25.3 32.7 47.8 51.0 56.1 61.7 57.7 61.1 62.5 60.5 60.8 66.3 55.7 73.7 73.8 82.0

83 78 72 81 69 71 87 82 84 90 83 87 87 84 79 84 67 83 80 87 81

As a percentage of gross income

IPD defined capital expenditure

Net cash flow

As a percentage of gross income

89 92 92 93 93 93 90 88 88 84 88 88 90 88 87 90 91 91 89 90 90

7.8 12.1 5.9 8.5 7.5 36.9 9.3 7.4 3.6 6.0 10.3 39.4 33.8 3.2 2.3 11.1 8.8 12.2 6.4 2.9

27.0 28.6 37.4 41.2 50.4 30.8 67.4 80.2 85.1 83.6 82.1 48.3 60.1 91.7 88.0 86.1 91.1 96.7 103.5 113.6

69 65 80 77 81 42 79 81 85 79 78 49 58 85 85 80 83 81 84 88 75

Year 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Average

Total income

IPD defined revenue expenditure

Net income

23.0 26.6 28.5 30.0 32.8 36.5 41.5 47.5 49.7 52.5 53.3 54.5 56.3 63.2 55.8 55.7 57.1 61.1 65.5 63.2

1.9 1.5 1.3 1.1 1.5 1.6 1.9 2.7 4.4 3.0 3.1 3.4 4.5 2.7 3.6 4.1 3.4 3.8 5.0 5.1

21.2 25.1 27.2 29.0 31.3 34.9 39.6 44.8 45.3 49.5 50.2 51.2 51.8 60.5 52.3 51.6 53.7 57.2 60.5 58.2

As a percentage of gross income

IPD defined capital expenditure

Net cash flow

As a percentage of gross income

92 94 95 96 95 96 96 94 91 94 94 94 92 96 94 93 94 94 92 92 94

1.5 1.2 0.2 0.9 2.4 0.8 0.2 1.2 0.0 1.3 15.3 1.1 0.9 4.6 2.7 10.7 3.2 17.0 8.1 9.6

19.7 24.0 27.0 28.1 28.9 34.1 39.4 43.5 45.2 48.2 34.9 50.1 50.9 55.8 49.6 40.9 50.5 40.2 52.4 48.6

86 90 95 94 88 93 95 92 91 92 66 92 90 88 89 73 88 66 80 77 86

Corresponding author Steven Devaney can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

Depreciation, income and the UK REIT 209

Table AIII. Income and expenditure for portfolio of industrial properties held over the period 1984-2003 (figures in millions)

The current issue and full text archive of this journal is available at www.emeraldinsight.com/1463-578X.htm

JPIF 26,3

Time-varying performance of four Asia-Pacific REITs

210

Department of Building and Real Estate, Hong Kong Polytechnic University, Hong Kong, China

Chiang Yat-Hung, So Chun-Kei Joinkey and Tang Bo-Sin

Received October 2006 Accepted April 2007

Abstract Purpose – The aim of the paper is to determine the dynamic relationships between REIT returns and those of other financial and real unsecuritized assets internationally. Design/methodology/approach – Using a multi-factor model the flexible least squares (FLS) coefficients of REIT returns against stock, bond and direct property returns are derived for the REIT markets of the USA, Australia, Japan and Singapore. Findings – The correlation between REIT returns and those of other financial and real assets varies not only across countries but also inter-temporally. REITs can certainly provide diversification benefits to a multi-asset investment portfolio. However, due to the time-varying nature of the correlation, active management is advised and REITs should be not be viewed as a complete substitute for direct property investment. Research limitations/implications – There are two major limitations to the study. Firstly, the sampling periods used are not the same across the countries due to differing market maturity. Secondly, there are also sheer differences in market sizes. However, as REIT markets around the world continue to grow and become more mature in terms of their breath and depth, there will be a richer set of data available for more in-depth analyses based on the methodology presented here. Practical implications – The conclusions on both mature and emerging REIT markets could provide some ideas for international investors as to how they should formulate their time-varying investment strategies and reconstruct their portfolios as mature markets become more efficient and emerging ones more mature. Originality/value – The inclusion of Asian markets enables investigation of the correlation between REITs and different assets in respect not only of different market conditions, but also different geographical locations and market maturity. The international dimension of this paper may appeal to readers and investors who are interested in identifying diversification opportunities around the globe, especially so when the capital and property markets around the world are becoming more integrated and globalized. Keywords Real estate, Investments, Trusts, Least square approximation, Financial risk Paper type Research paper

Journal of Property Investment & Finance Vol. 26 No. 3, 2008 pp. 210-231 q Emerald Group Publishing Limited 1463-578X DOI 10.1108/14635780810871605

Introduction This paper examines the dynamic relationships between REIT returns and those of other financial and unsecuritized real assets internationally. The objectives of this paper are: . to identify the causes of the dynamic nature of these relationships; and . to determine whether the relationships vary geographically. The authors would like to thank The Hong Kong Polytechnic University for funding this study (project code: A-PE21).

The findings will give us insight into the predictability of REIT returns, and will help investors and market practitioners make better informed decisions. Literature review The return and risk profile of REITs has interested both academics and practitioners. Generally speaking, REITs are believed to have a volatility in between direct property and stock, and so their average return (for a comprehensive review of 128 papers on returns and risk on real property and REITs, see Benjamin et al., 2001; for another comprehensive review of 240 studies on REITs in particular, see Zietz et al., 2003, which extends the work of Corgel et al., 1995). Han and Liang (1995) covered a period between 1970 and 1993 and concluded that the performance of REITs was “similar to that of a passively managed portfolio consisting of three-month Treasury bills and a stock market portfolio”. Further, as an asset class, REITs provide not only a “middle of the road” alternative for investors, but also serve as a diversifying asset in a multi-asset portfolio. Chiang and Lee (2002) consider the REIT to be a class of its own, which should be included as a diversifying asset in a multi-asset portfolio “even when unsecuritized real estate is a viable investment”. Lee and Stevenson (2005) add that the diversification benefits increases as the holding period increases. However, we need to understand the time-varying nature of the return relationship between REITs and the other assets before we can construct an optimal portfolio that also includes REITs. Chen and Peiser (1999) found that over a four-year period between 1993 and 1997, the relations between REITs and the stock market were not strong. The beta of equity REITs under capital asset pricing model analysis is generally very low, suggesting that REITs are exposed to risk factors different from stocks in the USA. Wang et al. (1995) attempted to explain the weak relation. They attributed the different behaviour of REITs vis-a`-vis stock to market efficiency. The stock market does not “provide the same level of services, such as information dissemination, monitoring activities and the pricing mechanism, for REIT stocks as it does for other stocks in the market”. Kuhle and Alvayay (2000) also suggested “a degree of inefficiency in REIT prices”. They applied runs and autocorrelation tests to the prices of 108 randomly selected equity REITs between 1989 and 1998. However, in a more recent study, Jirasakuldech and Knight (2006) covered a much longer period, from 1972 to 2004. Based on serial correlations, variance ratio tests and a non-parametric runs test, they concluded that market efficiency had increased over time for both equity REITs and small cap stocks. However, as they also rightly pointed out, efficient or not, it is the correlation with other assets or the marginal contribution of risk that should be considered when REITs are to be included in a mutli-asset portfolio, echoing the similar view of Seiler et al. (1999). There appears to be some recent evidence suggesting that REITs have correlated more with direct property than stocks, making REITs “a liquid means of adding exposure to the unsecuritized property market” (Jirasakuldech and Knight, 2006). As the REIT market becomes mature, some argue that the risk return profile of REITs will become more like that of direct property than stock. When the REIT market becomes more information efficient, REIT prices will more accurately reflect the fundamentals of its underlying assets, i.e. direct property. Khoo et al. (1993) found a structural change in equity REIT betas, which decreased in the 1980s with respect to the equity market due to improvements in REIT information efficiency. Conover et al. (2000) established that between 1978 and 1994, the beta alone was a significant factor

Asia-Pacific REITs

211

JPIF 26,3

212

explaining REIT returns during “bull market months”, but not so during “bear market months”, when betas are allowed to vary over bull and bear months. Chatrath et al. (2000) reported a REIT beta asymmetry similar to the case of small cap stocks, but were not able to explain this. In a later paper, Chiang et al. (2004) claimed to have solved the puzzle of beta asymmetry by resorting to factors controlling capitalization and book-to-price variables. According to their study, REIT betas are symmetrical when the Fama-French three-factor model (Fama and French, 1993) is used to estimate performance. He (2000) found a “very strong contemporaneous causality between apartment REIT stock returns and new house prices”, suggesting that prices of REIT and direct property are subject to the same fundamentals such as interest rates. More importantly, Clayton and MacKinnon (2001) concluded that the return relationship between REITs and four other assets changes over time and under different market conditions. They regressed REIT returns on bonds, small cap stocks, large cap stocks and direct properties using a flexible least squares method developed by Kalaba and Tesfatsion (1989) to allow the coefficient estimates of independent variables to change over time. Their findings have major implications on risk control under different market conditions and rebalancing portfolios. Methodology Our study extends the work of Clayton and MacKinnon (2001) by including three more Asian REITs from Australia, Japan and Singapore. There are emerging REIT markets across Asia (Chiang and So, 2006; Ooi et al., 2006). The inclusion of Asian markets enables us to investigate the correlation between REITs and different assets in respect of not only different market conditions, but also different geographical locations and market maturity. The Australian market is probably the most established and mature in the world outside the USA, while Japan and Singapore are newcomers. Australia had its first REIT listed in 1971; this did not occur in either Japan or Singapore until after 2001. Our aim is to identify the drivers of REIT market movements. With the use of large cap index return, we can study how REITs behave in different economic environments, and then formulate portfolio strategies by including optimal amounts of REITs. Previous studies, including those of Gyourko and Nelling (1996) and Goldstein and Nelling (1999), have reported that REIT return has a higher correlation to small cap stocks than to larger cap stock; we also test whether this phenomenon still held over recent years by including small cap stock in our regression model. We employ the multi-factor model as often used before (Clayton and MacKinnon, 2001; Sing and Ling, 2003; Chiang et al., 2004). The multi factor model is specified as follows: RREIT;t ¼ a þ bLC RLC;t þ bSC RSC;t þ bB RB;t þ bRE RRE;t þ 1t ;

ð1Þ

where RREIT is the quarterly return to local REITs index, a is a constant, bLC is the regression coefficient for the quarterly local large cap stock index return, bLC,t is the quarterly return to local large cap stock index at time t, bSC is the regression coefficient for the quarterly local small cap stock index return, bSC,t is the quarterly return to local small cap stock index at time t, bB is the regression coefficient for the quarterly long-term bond price index return, bB,t is the quarterly return to bond price index at time t, bRE is the regression coefficient for the quarterly return of the local unsecuritized real estate investment, bRE,t is the quarterly return of the local unsecuritized real estate investment at time t, and 1t is the error term.

Since all our explanatory variables share some common macroeconomic drivers, it is necessary to examine whether there are strong correlations between these return series so that spurious regression induced by multicollinearity, if any, can be avoided or rectified. The major limitation of the multi-factor model is the constant coefficient assumption of ordinary least squares (OLS). Previous studies have suggested a time-varying nature of the return relationship between REITs and other different asset classes, although the cause of the structural breaks in the REIT beta is not certain. Therefore, in of our study, we shall attempt to examine the impact of market conditions on the coefficients by employing time-varying linear regression using the flexible least squares (FLS) technique. Instead of having to presume when the structural breaks are by pre-defining separate sub-samples, as in the Chow test, FLS has the advantage of being able to recognize changing correlations, thus identifying a structural break if there is one. Data The data used were the quarterly return series of different assets from the selected REIT markets. We are limited to using quarterly data because all property price indicators are available only on a quarterly basis. The study periods are different across different markets due to their different maturities. The USA is the most mature among the markets that we study. Its long time series of data enables us to adopt a sampling period of ten years between 1995 and 2005. Data on equity REIT is provided by the National Association of Real Estate Investment Trusts (NAREIT), and their Equity NAREIT index is used to calculate the return of REIT investment. S&P 500 and Russell 2000 are used to represent the large cap and the small cap index, respectively. Lehman Brothers’ US Treasury Index is chosen to calculate the return on fixed income. For unsecuritized property investment, we employ a “de-lagged” NCREIF NPI index. Unlike financial products traded on a stock exchange, properties are not traded continuously, and the property price indices that we use as a proxy of direct property investment are based on property appraisals. The problem of appraisal is that it has smoothed the actual returns (Geltner, 1991) as well as lagged the true market returns (Giliberto et al., 1993). Thus, for a more realistic proxy of direct property investment, we compute and use a “de-lagged” NPI index based on the finding of Fisher and Geltner (2000) that the NPI index is lagging the true market for about three quarters in general. Based on de-smoothed indices, Feldman (2003) suggests that, for the period 1987-2001, the “significant overweighting” of both direct and securitized property is justified and they were complementary investments. Although the history of Australian listed property trusts dates back to the 1960s, it did not become popular until the 1990s. The return on Australian REITs (LPT) is computed using the Dow Jones LPT index between a five-year period between 2001 and 2005. The returns on ASX 50, ASX small ordinaries and MSCI Australia 7-10 year bond index are used as proxies for large cap, small cap, and bond market performance respectively. Similar to the USA data, we use a de-lagged house price index (HPI) published by the Australian Bureau of Statistics for computing the return on unsecuritized property investment in Australia. The Japanese REIT (J-REIT) made its debut in 2001 and expanded rapidly, with 39 listed J-REITs now trading on the Tokyo Stock Exchange (see Japan-REIT.com). The J-REIT index published by STB Research Institute is used here to calculate the

Asia-Pacific REITs

213

JPIF 26,3

214

quarterly returns on equity REIT investment in Japan between the years 2001 and 2005. Returns on Nikkei 225 and NGSC Japan small cap index are used as proxies of returns on large cap and small cap stock investment, respectively. Return on J.P. Morgan’s Japan benchmark ten-year bond index is used to calculate fixed income return. Return on Japanese unsecuritized property investment is derived from the average urban house price index provided by the Japan Real Estate Institute. The Singapore REIT market is the youngest among the four countries selected for this study. Since there is no readily available REIT index for Singapore, we have to create our own to calculate REIT returns. Using monthly closing price of all available Singapore REITs (namely Ascendas REIT, Ascendas REIT A, Capita Commercial Trust, Suntec REIT, Mapletree Logist Trust and Fortune REIT), we have created a capital-weighed Singapore REIT index for a four-year period between 2002 and 2005. The Straits Times Index (STI) and the Nomura Global Small Cap (NGSC) index are used to compute the large cap and small cap stock investment returns, respectively. The Singapore government long-term bond index and the property price index from the Real Estate Information System (REALIS) of the Urban Redevelopment Authority of Singapore are used to compute returns on Singapore bond and direct property investment, respectively. Results and analysis Results from our regression model are presented as follows. Tables I and II show the descriptive statistics of the investment performance of the assets across the sample periods. Their means and standard deviations (volatility) of returns are presented. Tables III-VI show the correlation matrices among the explanatory variables to identifying any possible multicollinearity within the model. We were not surprised to find that large cap and small cap stocks are highly correlated in all samples. To rectify this problem, we orthogonalized the small cap returns with respect to the large cap ones. The residuals of the regression of small cap stock returns on large cap stock returns were used instead to provide a pure small cap factor for the multi-factor model. Results from the multi-factor model regression analysis are presented in Tables VII and VIII. Plots of FLS coefficients estimated on different assets against REIT return are illustrated in Tables IX and X and Figures 1-16. For comparison, we also plot time series of annual returns of large cap stock against the estimated FLS coefficients across Full sample 1995-2005 Mean Volatility (percent) (percent)

Table I. Descriptive statistics of quarterly data calculated on price indices of REIT, large cap stock, small cap stock, bond and property: USA

NAREIT S&P 500 Russell Bond NPI

3.52 2.64 2.84 0.57 2.77

6.80 8.61 0.32 8.36 10.80

Sub-periods 1995-2000 2001-2005 Mean Volatility Mean Volatility (percent) (percent) (percent) (percent) 2.29 6.09 4.11 0.58 2.90

6.69 6.94 9.66 8.27 0.80

4.69 20.64 1.63 0.56 2.65

6.71 8.75 11.66 5.81 1.14

Sources: National Association of Real Estate Investment Trusts (NAREFT), National Council of Real Estate Investment Fiduciaries, Datastream

time. We do this to obtain a graphical idea of how the relationships between REITs and other different assets vary under different market conditions. USA market We split the period between 1995 and 2005 into two sub-periods to see whether there is a significant change in the return and volatility of different asset investments. The first sub-period is 1995-2000 and the second is 2001-2005. The first sub-period witnessed the technology bubble, and the second was the aftermath of it bursting. Table I shows that Australia (2001-2005) Mean Volatility (percent) (percent) LPT ASX 50 ASX small Bond HPI

1.84 1.30 2.60 1.51 2.74

3.74 7.05 8.23 2.68 15.50

4.21 2.42 4.67 0.22 2 1.80

5.07 8.44 9.34 1.78 0.38

SREIT STI NGSC Bond HPI

5.61 4.37 5.90 0.69 6.58

5.22 5.24 5.89 3.49 7.02

Sources: Australia: Australian Bureau of Statistics, Datastream; Japan: STB Research Institute, Japan Real Estate Institute , Datastream; Singapore: Urban Redevelopment Authority of Singapore, Datastream

S&P 500 Russell 2000 Long term bond index Adjusted NPI index

ASX 50 ASX small cap MCSI bond index Adjusted HPI

S&P 500

Russell 2000

Long-term bond

Adjusted NPI

1 0.88 0.12 0.28

1 0.10 0.16

1 0.10

1

ASX 50

ASX small cap

MCSI bond index

Adjusted HPI

1 0.88 20.28 20.07

1 20.34 20.07

1 2 0.06

1

Nikkei 225 MCSI Japan small cap index Long term bond index Adjusted urban house price index

215

Singapore (2001-2005) Mean Volatility (percent) (percent)

Japan (2001-2005) Mean Volatility (percent) (percent) J-REIT Nikkei 225 Small cap Bond HPI

Asia-Pacific REITs

Nikkei 225

MCSI Japan small cap

Long-term bond

Adjusted UHPI

1 0.50 20.088 20.17

1 2 0.21 2 0.36

1 0.17

1

Table II. Descriptive statistics of quarterly data calculated on price indices of REIT, large cap stock, small cap stock, bond and property: Australia, Japan and Singapore

Table III. Correlation matrices between quarterly returns on different assets classes: USA, 1995-2005

Table IV. Correlation matrices between quarterly returns on different assets classes: Australia, 2001-2005

Table V. Correlation matrices between quarterly returns on different assets classes: Japan, 2001-2005

JPIF 26,3

216

Table VI. Correlation matrices between quarterly returns on different assets classes: Singapore, 2002-2005

Table VII. Return correlations of REIT with other assets: USA

the mean return on NAREIT is higher in the second sub-period than the first, while both large cap and small cap stock investment suffer lower returns in this recessionary period. Contrary to some previous studies claiming that volatility in REIT return decreases as its market becomes mature and liquidity increases, we do not find any sign of volatility change across both sample periods. In addition, the volatility that we find is about the same magnitude (7 percent) of those that date back to the 1970s (see, for example, Chan et al., 1990). Our results, however, show that the NAREIT return has a lower volatility than stock when the stock market gets particularly volatile during a recession period. We also find that the NPI return has the lowest volatility among all assets. However, it should be remembered that this low volatility may be caused by the smoothing effect from the property appraisals process as mentioned previously. Results on the correlation between REIT and other assets are shown in Table III. We find that the small cap index return has the highest correlation with NAREIT return in both sub-periods, closely followed by S&P 500 as a proxy for large cap stocks. This finding is consistent with many previous studies (e.g. Gyourko and Nelling, 1996).

STI NGSC Long term bond Adjusted HPI

STI

NGSC

Long term bond

Adjusted HPI

1 0.42 0.04 0.16

1 20.17 0.81

1 2 0.22

1

Full sample 1995-2005

Correlation of NAREIT returns with Sub-period 1995-2000

Sub-period 2000-2005

0.36 * * 0.57 * * 0.10 0.05

0.34 0.53 * * 0.28 0.06

0.47 * * 0.58 * * 20.20 0.07

S&P500 Russell 2000 Long bond Adjusted NPI

Notes: *Significant at 90 percent level of confidence; * *significant at 95 percent level of confidence

Australia (2001-2005)

Table VIII. Return correlations of REIT with other assets: Australia, Japan and Singapore

Return ASX correlation of 50 LPT with ASX small cap Bond HPI

0.36

0.05 * 2 0.62 2 0.07

Japan (2001-2005) Return correlation of J-REIT with

Singapore (2002-2005)

Nikkei Return 20.05 correlation of 225 S-REIT with Small 0.31 Cap Bond UHPI

0.20 20.16

STI

0.50

NGSC

0.38

Bond 2 0.10 Adjusted PPI 0.39

Notes: *Significant at 90 percent level of confidence; * *significant at 95 percent level of confidence

0.49901 * * 0.44 0.37 8.17 * * 0.40949 * 0.37 0.20 1.95 * 0.53594 * * 0.49 0.37 6.47 * *

bS&P 500 (2 4.393)

(2 2.281)

(2 2.4)

2 1.0754 * *

2 0.92859 *

2 1.1097 * *

(4.381)

(1.866)

(3.69)

bRussell 2000

20.0799

0.14380

0.056326

bBonds

(0.31)

(0.8278)

(0.4667)

0.63794

0.43494

0.43745

(0.58)

(0.2482)

(0.4778)

bAdjusted NPI

(0.3624

(0.1608)

(0.09)

0.95896

0.82674

0.68990

bConstant

Notes: *Significant at 90 percent level of confidence; * *significant at 95 percent level of confidence; t-statistics are shown in parentheses

1995-2005 R2 Adj. R 2 F-test 1995-2000 R2 Adj. R 2 F-test 2000-2005 R2 Adj. R 2 F-test

Period

Asia-Pacific REITs

217

Table IX. Regressions of equity REIT returns on large and small cap stock, bond and unsecuritized property returns: USA

JPIF 26,3

218

bASX Australia 2001-2005 R2 R2 F-test Japan 2001-2005 R2 Adj. R 2 F-test

Table X. Regression of LPT (Australia), J-REIT (Japan) and S-REIT (Singapore) returns on large and small cap stock, bond and unsecuritized property returns

Singapore

bASX small cap

0.02 (0.14) 0.44 0.26 2.376 *

0.72 * *

bNikkei 225 0.416 0.23 0.11 2.54 *

bSTI

2002-2005 0.22 0.3025 R2 2 0.096 Adj. R 2 F-test 2.92 *

(2.54) 21.7

bNGSC

(1.17) 2 0.58

0.55

bAdj. HPI

bConstant

(20.18) 2 0.02 (20.32) 4.2

bBond

(2 1.48)

bNGSC (0.31)

bBond

0.77

bBond

bUHPI

bconstant

(1.03) 2 0.08 (20.01) 3.07

bPPI

(0.58) 20.25 (20.47) 2 0.09

(0.32)

1.84

bConstant (0.20) 4.19

1.62

Notes: *Significant at 90 percent level of confidence; * *significant at 95 percent level of confidence; t-statistics are shown in parentheses

Figure 1. Time path of flexible least squares coefficient on S&P 500 versus S&P 500 annual returns: USA

These relationships are all significant at the 5 percent level. The return correlation between REIT and bond changes from positive in the first sub-period to negative in the second. However, the relationship is statically insignificant in both sub-periods. In addition, we do not find any significant correlation between adjusted NPI and NAREIT returns. In fact the magnitude of the correlation that we find is back to the low level during 1979-1991 as calculated by Clayton and MacKinnon (2001). Our finding is contrary to some previous studies (McIntosh and Liang, 1998; Clayton and MacKinnon, 2001) claiming that REIT would behave more like real property and the link between REIT and stock would weaken as the REIT market becomes mature.

Asia-Pacific REITs

219

Figure 2. Time path of flexible least squares coefficient on Russell 2000 versus S&P 500 annual returns: USA

Figure 3. Time path of flexible least squares coefficient on US bond versus S&P 500 annual returns: USA

The results derived from our multi-factor model for US data are illustrated in Tables IX and X. The adjusted R 2 of our multi factor model is 0.37, which is similar to results found in previous studies. For example, Chiang et al. (2005) report an adjusted R 2 of 0.32 for their three-factor model covering the 1992-2002 period. Clayton and MacKinnon (2001) reported an adjusted R 2 of 0.27 with a sampling period covering 1992-1998. F-tests statistics are all significant at the 10 percent level. We can assume that the model is valid. Overall, we find that only returns on S&P 500 and Russell 2000 appear to be related to NAREIT return with a statistical significance of 95 percent. Further, we find little trace of REIT return being related to bond and direct real estate from all sample periods. Contrary to the finding of Khoo et al. (1993), we do not see any sign of a weakening link between stock and NAREIT returns over time. Instead, this link grew stronger during the second sample period when the stock market suffered from its technology bubble burst. Comparing the results obtained from the two

JPIF 26,3

220

Figure 4. Time path of flexible least squares coefficient on adjusted NPI versus S&P 500 annual returns: USA

Figure 5. Time path of flexible least squares coefficient on ASX 50 versus ASX 50 index annual returns: Australia

sub-periods, we notice an increase in both large and small cap stock market betas: The large cap beta increases from 0.41 to 0.54 from the first sub-period to the second, and the small cap beta also increases by 20 percent at the same time. The relationship between equity REIT and bond returns changes sign, and it is also noteworthy that the sensitivity to real estate returns increases in the second period. Our multi-factor model explains 37 percent of the variation in NAREIT returns in the second period compared to just 21 percent in the first period. We believe that this is mostly attributable to the link between NAREIT and stock returns becoming stronger in the second sub-period. To gain a better perspective on the dynamic nature of the correlation coefficients, we now turn our focus to the results from the FLS analysis. Figures 1 and 2 show the time path of FLS coefficients on S&P 500 plotted against S&P 500 and Russell 2000 annual returns. We can see a moving pattern with “anti-node” between the two time

Asia-Pacific REITs

221 Figure 6. Time path of flexible least squares coefficient on ASX 50 small ordinaries versus ASX 50 index annual return: Australia

Figure 7. Time path of flexible least squares coefficient on Australia bond versus ASX 50 index annual returns: Australia

series, which tallies with the study of Goldstein and Nelling (1999). The coefficients between equity REITs and stocks are high during a bear market and low during a bull market. Similar results are found between the returns on the small cap stock and the NAREIT index. However, we do not find any trace of decline of this relationship over time, as suggested by previous studies. The FLS coefficients on bond with respect to NAREIT returns hover around zero from 1995 to 2000, when the US interest rate stayed at rather stagnated level. However, the coefficients start becoming negative when the interest rates start decreasing from the year 2000. However, contrary to some market practitioners considering E-REIT a fixed income product, which is then interest-rate sensitive, the correlation between bonds and E-REIT is not statistically significant. Finally, regarding equity REITs, we do not find any clear increasing trend in the correlation between E-REIT and direct property investment. Myer and Webb (1993) found that during the earlier period of 1978-1990, REIT returns were inter-temporally related much more to direct property than to stock. However, from our

JPIF 26,3

222 Figure 8. Time path of flexible least squares coefficient on adjusted HPI versus ASX 50 index annual returns: Australia

Figure 9. Time path of flexible least squares coefficient on Nikkei 225 versus Nikkei 225 annual returns: Japan

plot we do find an increase in correlation between the two assets during the property boom, with the exception of the more recent one in the first half of 2004. To explain these puzzling return patterns and correlation between E-REIT and different asset classes, we resort to the fundamental causes of E-REIT returns variability within our sampling period. While the stock market recorded an annual return of more than 20 percent during 1995, there was a significant drop in return on E-REIT after the REIT IPO boom of 1993-1994. Investors started to see E-REITs overvalued in terms of their price when compared with their cash flows from operation. P/FFO (price per funds from operation, similar to price per earnings as in stocks) was 15 and it dropped further to 12.6 by the end of 1995, delivering a mere annual return of 3.6 percent in the second quarter of 1995 and causing the low correlation with other assets during this period. However, between 1996 and 1997, REIT prices bottomed out, and REITs was getting popular among institutional investors and fund mangers again. As portfolio mangers gradually began to invest again in REITs in addition to their

Asia-Pacific REITs

223 Figure 10. Time path of flexible least squares coefficient on Japan small cap index versus Nikkei 225 annual returns: Japan

Figure 11. Time path of flexible least squares coefficient on Japanese bond versus Nikkei 225 annual returns: Japan

direct property investment, a huge demand was created on REITs. The total capitalization of the REIT market grew, and the REIT market also became more liquid. During the same time many REITs expanded through acquisitions and mergers. Their FFO became better than what the market had expected. Investors were also willing to pay a higher premium on the net asset value (NAV) of E-REITs, being fuelled by the bull market of the underlying direct property investment. Feeling that the stocks were overvalued after the boom of early 1996, momentum investors also joined the bandwagon of REIT investing. Consequently, there was an amazing annual return of over 30 percent from Q4 1996 until it peaked at 40.5 percent in the third quarter of 1997. It was clear that as 1997 ended, the REIT price came under pressure when it was trading at a 30 percent premium over its NAV. The REIT market fell and did not recover until 2000 when the annual return went back to positive, while the stock market had stayed healthy. This recovery was mainly attributed to the low interest

JPIF 26,3

224

Figure 12. Time path of flexible least squares coefficient on Japan UHPI versus Nikkei 225 annual returns: Japan

Figure 13. Time path of flexible least squares coefficient on STI index versus annual return on STI index: Singapore

rate environment mentioned earlier. Being perceived as a fixed income product, the higher yields of REITs relative to low interest rates attracted investors (Figure 3). Yet, in 2001, the technology bubble burst, which along with the 9/11 terrorist attack, pushed the prices of not only stock but also REITs as well down. This triggered a three-year period of low interest rate, which went down from a peak of 6.57 percent in Q3 2000 until it bottomed out in Q1 2004 at 1.1 percent. Unlike previous cases, both the NAREIT and stock market plunged this time. For REITs, it was one year after the market recovered in 2000 from the panic sell out earlier (Figures 1 and 2). The US economy began to show some sign of strengthening in Q3 2003, and the Federal Reserve decided to increase the interest rate again for the first time since 2000 to curtail inflation. By Q2 2004, the yield of T-bond increased to 4.4 percent, thus making REITs unattractive again because their yield was only about 5 percent. Thus we found a

Asia-Pacific REITs

225

Figure 14. Time path of flexible least squares coefficient on NGSC returns versus annual return on STI index: Singapore

Figure 15. Time path of flexible least squares coefficient on Singapore bond versus annual return on STI index: Singapore

negative return of 2 5.81 percent on NAREIT index after factoring in dividends. There was the threat of higher interest rates to come and it had a cooling effect on both REITs and stocks. The increases in our FLS coefficients on stocks demonstrated the effect of increasing the interest rate as a common threat to both assets. As regards a comparison between REITs and direct property investment, what happened is a little counterintuitive and an example of how REITs and direct properties move in opposite directions. When the occupancy rates of both residential and office properties and consequently their rentals began to fall starting in 2001, we would expect an adverse effect on the FFO of REITs and their prices should fall. Yet, the market seems to have taken an opposite view. This time, REITs were cheap, being traded at 20 percent discount of their estimated net asset value (NAV). With further lowering of interest rates expected, investors saw an opportunity to get exposure to

JPIF 26,3

226 Figure 16. Time path of flexible least squares coefficient on adjusted PPI index versus annual return on STI index: Singapore

properties by buying REITs at low prices and at low costs of debts. Driven by investors’ interest, the total capitalization of REITs doubled between 2001 and 2004. We can conclude that, most of the time, REITs provide investors with an alternative to both stock and direct properties. Their prices do not usually move together in magnitude and direction, except in crisis when the prices of all assets plunge, something that the Asian Wall Street Journal (2001) described in its headline as “[d]iversification has its limits during crisis”. However, the relationship between REITs and other assets is not static. Our results suggest that these relationships change due to changes in performance drivers under different market conditions. There are time-varying characteristics in these relationships. However, we cannot conclude whether these variations are systematical, as claimed by Clayton and MacKinnon (2001). Asian markets As regards the three Asian markets, the analyses should be interpreted as exploratory only due to the paucity of data. We shall report on the cases of Australia, Japan and finally Singapore in the following discussion. Although Australian listed property trusts date back to the early 1970s, they did not become popular until 20 years later. The property market collapsed in the early 1980s (for an account of what happened in the Sydney property market over a time span of more than 100 years until 1981, see Daly, 1982). The market remained in the doldrums and only started to recover in the mid-1990s. True acceleration took place in 1997 when institutional investors saw LPT as a more robust investment vehicle after the Asian economic crisis. The total market capitalisation grew to US$40 billion in 2004, according to the Australian Stock Exchange. Table II presents the descriptive statistics computed from the Dow Jones Index, which has been compiled to track the LPT performance since 2001. During the period 2001-2005, LPT yielded a 1.84 percent quarterly return on average, which is 41 percent higher than large cap stock. Further, it managed to generate a relatively stable stream of returns. The volatility is 3.74 percent, only about half of the 7.05 percent volatility of ASX 50 quarterly returns. The ASX small cap index, on the other hand, generates an average quarterly return that is 30 percent higher than LPT. However, it also has a

much higher volatility, at 8.23 percent. The return on long-term bond is expectedly rather stable, offering an average quarterly return of 1.51 percent with a 2.68 percent volatility. Finally, direct property investment offers the highest average quarterly return of 2.74 percent, but at the same time it also has the largest volatility of 15.5 percent. On a risk-adjusted return basis, the best performer is the bond, followed by LPT. Their coefficients of variation (volatility over mean) are 1.77 and 2.03, respectively. From the correlation matrix of Table IV, we are not surprised to find that returns on large cap stock and small cap stock are highly correlated. To rectify the multicollinearity problem in our regression we have used the same procedure as described for the US data. Table X presents results from the multi factors model on Australian data. The explanatory power of this model is similar to the US case. Together, the four asset classes explain 44 percent of the LPT returns movement. The correlation between LPT and ASX50 returns is 0.2, but this is not statistically significant. The correlation between LPT and small cap is 0.72, and is significant at the 95 percent confidence level. It appears that LPT and small cap stock are subject to similar factors affecting their performance, suggesting the small-firm effect as identified by McIntosh et al. (1991) in their study of US-REITs traded between 1974 and 1988. Return on LPT is related neither to bond or direct property based on correlations. Due to the short series of data, we are unable to analyze changes in correlation between LPTs and other financial and real assets by dividing our data into sub sampling periods. The time paths of FLS in Figures 5-8 demonstrate the changing patterns of LPT’s relations with stock and direct property. The ripple of the global economic downturn and the political unrest in the Middle East reached Australia in 2002. ASX 50 did not show a positive return until Q2 2003. The market lost direction and when investors were seeking more stable investments, the property market, helped by a low interest rate environment, continued its boom. LPT became a good alternative for “yield-player” investors seeking liquid property exposure. Our results support the view that LPT performance is more related to direct property than stock over this period (see Figures 5, 6 and 8). There is a reverse of the relationship in 2004, when the return on direct property started to level off and then gradually went into a slow decline after the stunning performance of 2003. The annual return on HPI up to December 2004 is negative at 2 2 percent. Yet, helped by the low vacancy rate brought about by a strong recovery from the first half of 2004, LPT manages to generate positive revenue and produce earnings growth. LPT continues to thrive alongside stock. As our FLS analysis shows, LPT is more related to stock and less to direct property during the latter half of our sample period. In Japan, ever since J-REIT made its debt on the Tokyo Stock Exchange in September 2001, it has been the best investment in terms of risk-adjusted return. It provides investors with an average 4.21 percent quarterly return, with lower volatility compared to both large cap and low cap stocks (Table II). We believe that the warm reception for J-REIT is largely due to high demand from “yield-hungry” investors under an extremely low interest rate environment. The ten-year government bond yield was a mere 1.5 percent when J-REIT made its debut. As of September 2005, the number of J-REITs had already expanded to 22 with a total market capitalization of about $25 billion. J-REITs have capitalized on the bottomed property price to offer higher yields

Asia-Pacific REITs

227

JPIF 26,3

228

to their investors. The slow but steady recovery in the property market has also given investors the confidence they need to invest in REITs. As shown in Figure 12, the relationship between J-REITs and direct property has been getting stronger over the period. There are steady increases in the FLS coefficients on direct property. Indeed, acquisitions by J-REITs account for nearly half of all officer property transactions during 2003 to 2004 in central Tokyo (Nikkei Real Estate Management Information, 2003/2004), pushing up property prices and bringing down rental yields. Our multi factors model (Table X) explains 23 percent of the J-REIT returns variation over the sample period and the F-test is significant at the 5 percent level. The model shows that the J-REIT return is not statistically significantly related to any other asset (Table X). However, it should be noted that the correlation with bond is relatively high when compared to other countries, indicating the particular importance of interest rate factors on J-REIT returns. The warm reception of J-REITs by the market has also caused a low correlation with other financial assets. However, as the yield gap between J-REITs and bond narrows, J-REIT price would be more exposed to interest rate risks, and we can see this happening already in Q3 2003 (Figure 11). We also believe that the disappearance of this effect in the later part of 2004 contributes to the new flux of J-REIT IPOs, thus widening the average disparity of the yield spread between two assets. Singapore has the youngest REIT market in our study. Like the J-REIT, the S-REIT has gained a lot of attention from both local and international investors, and the market had grown to nearly 13 billion Singapore dollars as of December of 2005. The average quarterly return is 5.61 percent since S-REITS began in December 2002 until September 2005 (Table II). As presented in Table X, we find from our multi-factor model that S-REIT is not correlated with any of the four asset classes with statistical significance. We believe that the reasons for this are similar to the case of Japan. A very large part of the growth in S-REITs is demand-driven. Initial public offerings (IPO) of S-REITs are always over-subscribed, thus giving them amazing returns on the first days of their listing. However, the main drivers for S-REIT returns are still ambiguous, especially for the early part of our data, though we expect an increasing correlation between S-REITs and other asset classes as the S-REIT market becomes mature. In fact, from Figures 13 and 14, we can see that the FLS coefficients on stock are on a rising trend. Our results from the FLS analysis also reflect how the market has reacted on interest rates. S-REITs looked particularly attractive in 2002 when they were offering an average 7 percent yield. As S-REIT prices inflated, the average yield fell to an average of only 3.8 percent in Q3 2005, and investors started to be attracted to the Singapore government’s ten-year bond, which yielded 2.7 percent at the same time. From Figure 15, we can see that as the difference between the yields of these two assets narrows, a rise in interest rates would make S-REITs less attractive, especially if they cannot sustain their growth. In fact the S-REIT price went down by 5.1 percent on average in Q4 2005, and clearly the interest rate increase was the main driver of this downturn. Limitations and suggestions for further studies The main objective of this paper is to gain an insight into the changing inter-relationship over time between the REIT returns and returns of stock, fixed income instruments and unsecuritized property. We have drawn some conclusions

based on findings from the more mature markets which could provide ideas to investors in emerging European and Asian markets about how they should formulate their time-varying investment strategies and reconstruct their portfolios as their market matures. Sing and Ling (2003) used the historical relationships between the returns of stocks, bonds and a selection of listed property trusts (LPTs) in Australia to simulate a hypothetical property trust in Singapore over the same studied period to project a Singapore REIT profile. They identified the risk return characteristic of a REIT asset, thus devising optimized investment portfolios with the inclusion of REITs. When more local data become available, investors should be able to make more informed decisions about constructing their portfolio using our methodology. They do not have to simulate a hypothetical trust again. However, it should also be noted that our study does not aim to going into detail on the fundamentals that have affected the performance of the assets discussed above. Besides, comparisons are also made difficult due to two reasons: (1) our sampling periods are not the same across the countries due to different market maturities; and (2) there are also sheer differences in market size. Yet, as REIT markets around the world continue to grow and become more mature in terms of their breath and depth, there will be a richer set of data available for more in-depth analyses. Conclusion In a mature REIT market like the USA, REIT return generally follows stock return, in particular that of small cap stock. However, just like stock, the performance of REITs is subject to specific risk factors, such as the underlying property investment and interest rates. Both REITs and stock are affected by interest rates, respectively indirectly and directly, but the effect appears to be less on REIT than on stock. The interrelationship between REIT return and that of other assets varies across countries. Given the availability of data, we can identify the relationship more easily from a more mature REIT market. The findings from these mature markets may shed some light on how REITs would perform in new markets. It is suggested that the performance of these new markets would depend on REIT supply and demand and other indigenous issues of the REIT itself rather than following the path of mature markets. However, in general, we believe that it is not wise to try making a hard and fast rule that REITs will behave more like stock, or a fixed income asset, or its underlying direct property investment. Its correlation with other assets is varies over time. REITs can certainly provide diversification benefits to an investment portfolio. However, due to the time-varying nature of the correlation, active management is advised and REITs should not be be viewed as a complete substitute for direct property investment. References Asian Wall Street Journal (2001), “Diversification has its limits during a crisis”, Asian Wall Street Journal, September 24. Benjamin, J.D., Sirmans, G.S. and Zietz, E.N. (2001), “Returns and risk on real estate and other investments: more evidence”, Journal of Real Estate Portfolio Management, Vol. 7 No. 3, pp. 183-214.

Asia-Pacific REITs

229

JPIF 26,3

230

Chan, K.C., Hendershott, P.H. and Sanders, A.B. (1990), “Risk and return on real estate: evidence from equity REITs”, AREUEA, Vol. 18 No. 4, pp. 431-52. Chatrath, A., Liang, Y. and Mclntosh, W. (2000), “The asymmetric REIT-beta puzzle”, Journal of Real Estate Portfolio Management, Vol. 6 No. 2, pp. 101-11. Chen, J. and Peiser, R. (1999), “The risk and return characteristics of REITs – 1993-1997”, Real Estate Finance, Vol. 16 No. 1, pp. 61-8. Chiang, C.H.K. and Lee, M.L. (2002), “RElTs in the decentralized investment industry”, Journal of Property Investment & Finance, Vol. 20 No. 6, pp. 496-512. Chiang, Y.H. and So, C.K.J. (2006), “Devising a conducive regulatory framework for Hong Kong real estate development trusts”, Journal of Financial Management of Property and Construction, Vol. 11 No. 2, pp. 91-104. Chiang, C.H.K., Lee, M.L. and Wisen, C.H. (2004), “Another look at the asymmetric REIT-beta puzzle”, Journal of Real Estate Research, Vol. 26 No. 1, pp. 25-42. Chiang, C.H.K., Lee, M.L. and Wisen, C.H. (2005), “Over the time-series properties of real estate investment trust betas”, Real Estate Economics, Vol. 33 No. 2, pp. 381-96. Clayton, J. and MacKinnon, G. (2001), “The time-varying nature of the link between REIT, real estate and financial asset returns”, Journal of Real Estate Portfolio Management, Vol. 7 No. 1, pp. 43-54. Conover, M.C., Friday, S. and Howton, S.W. (2000), “An analysis of the cross section of returns for EREITs using a varying-risk beta model”, Real Estate Economics, Vol. 28 No. 1, pp. 141-63. Corgel, J.G., McIntosh, W. and Ott, S.H. (1995), “Real estate investment trusts: a review of the financial economics literature”, Journal of Real Estate Literature, Vol. 3 No. 1, pp. 13-43. Daly, M.T. (1982), Sydney Boom, Sydney Bust: The City and its Property, 1850-1981, Allen & Unwin, Sydney. Fama, E.F. and French, K.R. (1993), “Common risk factors in the returns on stocks and bonds”, Journal of Financial Economics, Vol. 33, pp. 3-56. Feldman, B. (2003), “Investment policy for securitized and direct real estate”, Journal of Portfolio Management, September, pp. 112-21. Fisher, D. and Geltner, D. (2000), “De-lagging the NCREIF index: transaction prices and reverse-engineering”, Journal of Real Estate Finance, Vol. 17 No. 1, pp. 7-22. Geltner, D. (1991), “Smoothing in appraisal-based returns”, Journal of Real Estate Finance and Economics, Vol. 4 No. 3, pp. 327-45. Giliberto, S.M. (1993), “Measuring real estate returns: the hedged REIT index”, Journal of Portfolio Management, Spring, pp. 94-9. Goldstein, A. and Nelling, E.F. (1999), “REIT return behavior in advancing and declining stock markets”, Real Estate Finance, Vol. 15, pp. 68-77. Gyourko, J. and Nelling, E. (1996), “Systematic risk and diversification in the equity REIT market”, Real Estate Economics, Vol. 24 No. 4, pp. 493-516. Han, J. and Liang, Y. (1995), “The historical performance of real estate investment trusts”, Journal of Real Estate Research, Vol. 10 No. 3, pp. 235-62. He, L.T. (2000), “Causal relationships between apartment REIT stock returns and unsecuritized residential real estate”, Journal of Real Estate Portfolio Management, Vol. 6 No. 4, pp. 365-72. Jirasakuldech, B. and Knight, J.R. (2006), “Efficiency in the market for REITs: further evidence”, Journal of Real Estate Portfolio Management, Vol. 11 No. 2, pp. 123-32.

Kalaba, R. and Tesfatsion, L. (1989), “Time-varying linear regression via flexible least squares”, Computers & Mathematics with Applications, Vol. 17 Nos 8/9, pp. 1215-45. Khoo, T., Hartzell, D. and Hoesli, M. (1993), “An investigation of the change in real estate investment trust betas”, Journal of the American Real Estate and Urban Economics Association, Vol. 21 No. 2, pp. 107-30. Kuhle, J.L. and Alvayay, J.R. (2000), “The efficiency of equity REIT prices”, Journal of Real Estate Portfolio Management, Vol. 6 No. 4, pp. 349-54. Lee, S. and Stevenson, S. (2005), “The case for REITs in the mixed-asset portfolio in the short and long run”, Journal of Real Estate Portfolio Management, Vol. 11 No. 1, pp. 55-80. McIntosh, W. and Liang, Y. (1998), REITs: What Are They?, Prudential Real Estate Investors, Parsippany, NJ. McIntosh, W., Liang, Y. and Tompkins, D.L. (1991), “An examination of the small-firm effects within the REIT industry”, Journal of Real Estate Research, Vol. 6 No. 1, pp. 9-18. Myer, F.C.N. and Webb, J.R. (1993), “Return properties of equity REITs, common stocks and commercial real estate: a comparison”, Journal of Real Estate Research, Vol. 8, pp. 87-106. Nikkei Real Estate Market Information (2003/2004), Reading Trends in Real Estate Transactions, Nikkei, Tokyo. Ooi, J.T.L., Newell, G. and Sing, F.T. (2006), “The growth of REIT markets in Asia”, Journal of Real Estate Literature, Vol. 14 No. 2, pp. 203-22. Seiler, M.J., Webb, J.R. and Myer, F.C.N. (1999), “Are EREITs real estate?”, Journal of Real Estate Portfolio Management, Vol. 5 No. 2, pp. 171-81. Sing, T.F. and Ling, S.C. (2003), “The role of Singapore REITs in a downside risk asset allocation framework”, Journal of Real Estate Portfolio Management, Vol. 9 No. 3, pp. 219-35. Wang, K., Erickson, J. and Chan, S.H. (1995), “Does the REIT stock market resemble the general stock market?”, The Journal of Real Estate Research, Vol. 10 No. 4, pp. 445-60. Zietz, E.N., Sirmans, G.S. and Friday, H.S. (2003), “The environment and performance of real estate investment trusts”, Journal of Real Estate Portfolio Management, Vol. 9 No. 2, pp. 127-65. Further reading Mills, E.S. (1988), “Are real estate markets becoming more efficient?”, Journal of Real Estate Finance and Economics, Vol. 1, pp. 75-83. Corresponding author Chiang Yat-Hung can be contacted at: bschiang@ inet.polyu.edu.hk

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

Asia-Pacific REITs

231

The current issue and full text archive of this journal is available at www.emeraldinsight.com/1463-578X.htm

JPIF 26,3

German property investment vehicles and the introduction of G-REITs: an analysis

232

Ulrich Schacht

Received May 2007 Accepted July 2007

Finance and Accounting Group, WHU – Otto Beisheim School of Management, Vallendar, Germany, and

Jens Wimschulte Chair of Financial Services, University of Regensburg, Regensburg, Germany Abstract Purpose – Germany is the biggest real estate market in Europe. Although some established vehicles for indirect property investments are available, the German real estate market is dominated by direct investments and lags behind its international peers in capital market integration. The purpose of this study is to examine whether the recent launch of German REITs may improve this situation. Design/methodology/approach – Existing indirect property investment vehicles and the new G-REIT are analysed and compared along the dimensions of transparency, liquidity and risk/return characteristics. In addition, potential capital flows into G-REITs are investigated and economic implications derived. Findings – The study identifies the limitations of existing German indirect real estate investment vehicles and demonstrates the superior characteristics of the new G-REIT. Substantial short-term capital flows from existing vehicles to G-REITs are, however, unlikely. Instead the temporary exit tax will foster an economically beneficial reallocation of capital by private companies and public authorities through property sales to new domestic and international investors via G-REITs. Originality/value – The results indicate that G-REITs have the potential to attract substantial funds in the medium term and facilitate a more integrated and developed German property and capital market. Keywords Real estate, Investments, Trusts, Germany Paper type Research paper

Journal of Property Investment & Finance Vol. 26 No. 3, 2008 pp. 232-246 q Emerald Group Publishing Limited 1463-578X DOI 10.1108/14635780810871614

1. Introduction Real estate investment trusts (REITs) were first introduced in the USA in 1960 to offer retail investors a possibility to invest in diversified property portfolios. Overregulation and unfavourable tax treatment compared to direct property investments, however, resulted in a sluggish start of REITs and it took until the reforms in 1986 to remove these obstacles and to kick off the rapid growth of the market. Since then, REITs have established themselves globally as the primary vehicle for indirect property investments, both for private and institutional investors. The launch of REITs in The Netherlands in 1969 formed the start of this investment vehicle in Europe. A couple of countries have followed the Dutch example since then: Luxemburg (1988), Spain (1994), Italy (1994), Belgium (1995), France (2003) and lately the UK (2007)[1]. The introduction of REITs in Germany has been requested by almost all stakeholder groups for several years, but unresolved details prevented an agreement. The Deutsche

Bundesrat (upper house) finally passed a German REIT Act on March 30, 2007 and established the legal basis for German REITs (G-REITs) retroactively to January 1, 2007. The new G-REIT has to compete with the existing German indirect property investment vehicles. These vehicles, however, show specific weaknesses, especially with respect to transparency, liquidity and transaction costs, which are addressed by REITs. Although large capital transfers from existing indirect vehicles to G-REITs are unlikely in the short term, REITs will attract substantial funds in the medium term from private firms and public authorities selling their non-core property holdings and opportunity funds exiting their real estate investments. As a consequence, the introduction of REITs will change the structure of the German real estate market towards more exchange-traded investments and more participation of international investors. Since the second half of 2006 the German market has experienced some IPOs of real estate companies that will supposably transfer into REITs. The remainder of the paper is organized as follows. In Section 2, we first give an overview of the German market for property investments. We then discuss and review the existing vehicles for indirect property investments in Germany, focusing on open-end and closed-end real estate funds as well as listed real estate corporations. Section 3 is devoted to the presentation of the legal and tax regime for G-REITs based on the German REIT Act passed in March 2007 and an illustration of the characteristics of REITs compared to existing products. The expected changes in the German property investment market and potential for REITs are presented in Section 4. The final section summarizes and considers possible future directions. 2. The German market for property investments 2.1 Overview The German market represents the largest property market in Europe. While playing an important role in the overall economy, for example, due to its asset size or number of employees, the sector is seen as underdeveloped compared to its European peers with regard to liquid exchange-traded property investment vehicles. Based on the 100 largest listed European real estate companies the ratio of real estate firms’ market value to total stock market capitalization was 0.5 percent for Germany at the end of 2006, which is clearly below markets like The Netherlands (2.8 percent), the UK (2.6 percent) and France (1.1 percent)[2]. In addition, only e260 billion or roughly 4 percent of the e7 trillion invested in real estate in Germany at the end of 2005 is held indirectly[3]. The dominant direct real estate holdings are typically investments in a single or small number of properties directly held by an investor without the use of any intermediary. Indirect holdings in contrast are co-investments of several investors executed via an investment vehicle that is handled by an intermediary who takes care of the management of the properties[4]. The three most common indirect investment vehicles in Germany account for almost 66 percent of all indirect investments: the fairly illiquid closed-end funds amount to more than 36 per cent, open-end funds to roughly 28 percent, and quoted real estate corporations to more than 3 percent. Other vehicles like leasing companies/funds (29 percent) or foreign opportunity funds usually focus on special investor groups and are not available to the general public[5]. International investors showed little interest in the German real estate market for a long time and only started investing significantly again in recent years.

German property investment

233

JPIF 26,3

234

2.2 Direct real estate investments With a share of roughly 60 percent of all direct real estate investments, private households represent the predominant investor group in Germany, clearly surpassing the public sector (13 percent) and churches (8 percent). Institutional investors (5 percent), corporates, charities and foreign investors play a minor role, but nevertheless hold sizeable real estate assets in absolute terms[6]. Although private households own almost 30 million residential units, half of them are rented out to third parties, resulting in a home ownership ratio that is lower only in Switzerland of all European nations. This is partly the result of strong public sector activity in residential real estate construction after the Second World War, when most housing was destroyed, refugees poured into the country and the high demand for affordable housing could not be fulfilled quickly enough by the private sector. Consequently, the public sector is an important landlord today, especially in the Eastern part of Germany, where privatisation of property did not begin until German reunification in 1990. In addition, strict German tenancy laws, relatively high property prices in the past due to the extensive building code, the limited availability of mortgages[7], and the lack of freely available real estate transaction data, favored renting over acquiring apartments. 2.3 Indirect real estate investments vehicles 2.3.1 Closed-end real estate funds. German closed-end real estate funds (“geschlossene Fonds”) are indirect investment vehicles usually in the form of a civil or limited partnership, with the latter being the predominant form today[8]. The general partner of a limited partnership, mostly itself a limited liability company controlled by the initiator of the fund, takes over all liability of the partnership and receives management fees as a percentage of annual net rent income. In addition investors have to pay upfront fees based on the amount of subscribed capital. The initiator might also act via another subsidiary as the founding limited partner who sells its shares to new limited partners, i.e. private investors. A closed-end fund is initiated to pursue only a single or small number of specific, usually commercial, real estate projects, leaving diversification aspects to investors. The underlying projects – which can be the development or acquisition of properties – and the invested capital are fixed before the fund is marketed to the public via a prospectus prepared by the fund initiator. The capital is typically drawn from investors as needed for the project. After the specified maturity the properties are sold and the money is returned to the investors, while other mechanisms like a project rollover are also possible. Minimum investment requirements are typically rather high – around e20-30k – thus targeting high net worth individuals as limited partners. Closed-end funds are tax-exempt and taxation takes place just at the investor level. This has made it possible to structure closed-end funds to offer high tax advantages for investors by generating tax losses during the first years, which could be offset with rental income from other sources. Tax losses were generated via tax depreciation and partly high leverage (no formal limit), usually without influencing possible tax-free gains when these investments are sold after the ten-year holding period. These tax-driven fund structures made investors stick to their investment until maturity, as a sale would have resulted in a retroactive loss of the tax advantages. While the tax advantages have been the main reason to invest in closed-end funds for many years, this has changed gradually as tax authorities have reduced tax incentives. This change in investment

objective has also fostered the rise of a secondary market as the obstacle of losing tax incentives has diminished in importance. Most initiators of closed-end funds today offer some kind of trading platform to buy and sell their funds after issuance. The regional exchange of Hamburg even established a market segment for various types of closed-end funds. While volumes and number of trades are still quite low, this initiative is a first step to creating a liquid secondary market for closed-end funds. Closed-end funds are fairly intransparent compared to other indirect investment vehicles, because they are not regulated and little information is published in a standardized way. Even though a prospectus has had to be published since mid-2005 due to the EU abusive directive and corresponding German law, the content of the prospectus is not subject to any official examination. As the consequence of long-term committed capital and a fixed investor base, closed-end funds often invest in riskier property types such as, for example, shopping centers or property development projects[9]. In terms of volume closed-end funds are the most successful German indirect real estate investment vehicle, with total assets under management reaching e94 billion of equity capital at the end of 2005. Their strong growth, especially in the 1990s, was partly driven by huge tax incentives provided by the government after German reunification to refurbish existing real estate and build new commercial properties in Eastern Germany. However, the extensive construction activity finally resulted in substantial excess capacities, and consequently losses for many investors. 2.3.2 Open-end real estate funds. German open-end real estate funds (“offene Immobilienfonds”) have the legal form of a special asset pool. They are managed by an investment firm, which is a limited liability company owned by banks or insurance companies, and which usually manages several separate open-end property funds at the same time. Two types of open-end funds exist: (1) open-end mutual funds designed for retail investors with a low minimum investment; and (2) open-end special funds targeted at institutional investors[10]. Open-end special funds differ from mutual funds with regard to the number of investors, which is restricted to 30, the absence of a liquidity reserve requirement and different transaction costs. In addition, special open-end funds might offer tax advantages to some institutional investors. The focus here is on open-end mutual funds that have roughly three times the assets under management as special funds. Open-end mutual funds serve as an acquisition vehicle to invest deposited money in its own name but on the account of the investors in a large diversified property portfolio[11]. Properties for the fund are typically bought after the equity capital has been transferred to the investment company. The equity capital placed in the fund is variable as the investment firm serves as an intermediary, issuing and redeeming fund shares on a daily basis, with minimum investments rather low at around e50. The take-back obligation, however, rests with the fund and not the investment firm. As open-end funds invest mid- to long-term in illiquid properties but issue liquid fund shares to investors, they perform a liquidity transformation function, which might put a huge burden on the funds. To be able to fulfill the take-back obligation at any time, the fund has to own sufficient liquid assets and is thus obliged by law to hold between 5 percent and 49 percent of its portfolio in liquid/money market assets. For the period

German property investment

235

JPIF 26,3

236

from 2001 to 2004 the ratio effectively fluctuated between 22 percent and 31 percent. In case of very large redemptions of fund shares the investment firm is allowed to stop any redemption for up to two years[12]. This protective action might be necessary to prevent existing investors from extensive losses due to forced the liquidation of real estate assets. The required liquidity reserve, however, reduces a fund’s performance due to low returns of money market assets. In addition this feature results in pro-cyclical investment behaviour as investments are liquidity- rather than opportunity-driven. One mechanism to prevent investors from trading actively in open-end mutual funds is a relatively high up-front fee that is in the range of 5 percent of a fund share’s redemption price. The taxation of open-end mutual funds takes place only at the investor level and the fund itself is tax-exempt. The investor’s income from these funds is, in contrast to closed-end funds, classified as capital income. Open-end funds act in a highly regulated environment governed by the German Banking Act and Investment Law, which set among others specific diversification levels and minimum disclosure requirements, making this instrument fairly transparent. Compliance is monitored by the Federal Financial Supervision Authority. Moreover, open-end funds are the only vehicles whose properties have to be appraised by independent third parties on a regular basis. Open-end fund volumes have grown tremendously over the last 30 years with especially strong inflows since 2001 due to worsening stock markets conditions. Declining commercial real estate yields and a positive development of the stock markets resulted in substantial redemptions of fund shares by investors in 2004 and 2005. In this situation some funds could not provide enough cash and often the investment firms stepped in to avoid selling assets at large discounts and prevent a liquidity crisis. In December 2005, however, one open-end real estate mutual fund had to abandon redemption and finally close the fund due to massive outflows after announcing a property revaluation. This fund closure raised the awareness of the general public of the liquidity risk and fostered an ongoing discussion about changes to the open-end mutual fund structure. 2.3.3 Real estate corporations. Publicly listed real estate corporations (“Immobilien AGs”) are indirect property investment vehicles in the legal form of a corporation with the only difference to other listed corporations of having real estate investments and other property related activities as their main business purpose[13]. Theses companies were either directly founded as real estate companies or changed their business model to the real estate area. Minimum investment requirements and transaction costs are low and no upfront fees exist as for funds, thus making this investment vehicle also attractive for short-term oriented real estate investors. Real estate corporations are regulated amongst others by the stock corporation act. With some exceptions their disclosure lags behind large listed companies as they are usually not members of any indices or stock exchange segments with high disclosure requirements. The taxation of real estate corporations takes place at the corporate and investor level according to the half income system, resulting in some double taxation. Profits are first taxed at the corporate level and once again half the distributed profits are charged at the individual tax rate of the investor. Currently there are roughly 50-60 listed real estate corporations in Germany, with most of them being remnants of former industrial firms. Their market capitalization

and trading volume are mostly too low to attract institutional investors. Both figures, however, have increased in the recent past due to the expected introduction of G-REITs and potential conversion of these corporations to G-REITS. The former discounts to net asset value have switched to premiums. In addition some IPOs of real estate corporations have taken place since 2006 with the IPO of residential property company Gagfah SA, owned by private equity investor Fortress, being by far the largest IPO for the time being with an issuing volume of e850m and a market capitalization of roughly e5bn. 2.3.4 Other instruments for indirect investments. Other instruments for indirect real estate investments in Germany are primarily opportunity funds, real estate leasing funds and structured real estate products. For some years private equity investors have been actively acquiring real estate assets in Germany via so called opportunity funds that are not marketed to the general public but instead target pension funds or very wealthy individuals as investors. Since 2001 these funds have bought residential real estate assets for roughly e50bn, whereof up to 90 percent have been financed with debt to achieve their ambitious return targets. Another approach to achieve these targets is to take advantage of the low homeownership ratio in Germany by selling residential units to their tenants for a premium compared to the original purchase price. Given the typical investment horizon of these funds of three to five years, several exits can be expected for the near future, either via (block) sales to strategic or financial investors or via IPOs as real estate corporations or REITs. Real estate leasing funds usually invest in a small number of properties, which are rented out to a single lessee for the full duration of the fund. These investment vehicles allow corporates to free up capital tied in real estate for their core business activities but at the same time still have the right to use these properties and often also repurchase the assets at the end of the lease contract. Leasing funds typically invest in similar real estate assets as closed-end funds, and their development was also mainly driven by tax incentives. Several changes in German tax laws during recent years have eliminated the basis for pure tax-driven leasing funds. Their equity value at the end of 2005 was e77bn and was thus comparable to open-end funds[14]. A new type of property investment vehicles in Germany are exchange traded structured products. These “certificates” are targeted at retail investors and represent bank bonds whose redemption value depends on the development of property price indices or baskets of real estate corporations. Their outstanding volume, however, is rather low. 3. The design of German REITs An intensive and controversial discussion on the introduction and detailed design of REITs in Germany has been going on since the Initiative Finanzstandort Deutschland (IFD), an initiative for the promotion of Germany as a financial centre launched by the German financial sector, presented a first proposal for a German REIT Act in June 2004[15]. Various parties from academia, business and politics have contributed to this discussion in the meantime. The German government has recognized the trend towards REITs in other countries and the importance of their launch for the international competitiveness of the national property and capital market, and has therefore signalled its support. On March 30, 2007 the Deutsche Bundesrat (upper house) passed the German REIT Act and thus implemented the legal basis for G-REITs

German property investment

237

JPIF 26,3

238

retroactively to January 2007. The essential policies and procedures set forth in this act are discussed in the following[16]. To qualify for the G-REIT status and use of the protected REIT label as well as associated benefits, property investment companies have to register their legal and actual headquarters in Germany and have to comply with a specific set of requirements, primarily along the three dimensions of business activities, investor structure and financial covenants. Compliance with these rules has to be certified by chartered accountants annually. G-REITs have to focus their business activities on the purchase and management (i.e. renting, leasing and related service activities) as well as sale of real properties and equivalent rights. Investments in real estate limited partnerships and corporations are included in this definition, whereas residential properties built in Germany before January 2007 are not. The reason for this exemption is political concern over potential adverse effects on rents, social housing and urban development. To concretize the admitted business activities some criteria for the composition of assets and revenues of REITs are defined. Total assets at fair value (according to IAS 40) after deduction of distributable profits and potential capital gain reserves have to consist of at least 75 percent of real property. Tax authorities fine non-compliance with a penalty payment of at least 1 percent and at most 3 percent of the amount the actual portion falls short of the specified one. Moreover, gross proceeds have to originate at least 75 percent from rent, lease and sale of real estate. A breach of this condition results in penalty payments to the tax authorities of at least 10 percent and at most 20 percent of the difference between the specified and actual amounts. However, if one of these two conditions is missed three fiscal years in a row by a REIT, the special tax status is suspended after the third year. In addition, to avoid pure dealing in real estate, gross proceeds from sales of real property are limited to 50 percent of average real assets of REITs over a period of five fiscal years, whereas REITs’ subsidiaries are taken into account. As a sanction the special tax status is cancelled in the year of non-compliance. REITs are allowed to provide real estate related services for third parties like facility management or project development, but only via a specific fully owned subsidiary. The business of these subsidiaries is, however, restricted to 20 percent of total assets at fair value and 20 percent of gross proceeds of REITs. Non-compliance three fiscal years in a row results in cancellation of a REIT’s special tax status. The subsidiaries themselves do not possess the same special tax status as REITs. On the investor side, REITs are required to dispose of a minimum nominal capital of e15m, to issue only fully paid in shares with equal voting rights, and to list on an organized market, i.e. regulated exchange, in the European Union. They furthermore have to possess a free float of at least 25 percent at the date of listing and to permanently ensure a minimum free float of 15 percent. Individual investors’ holdings with less than 3 percent of shares classify as free float. Compliance with this standard has to be demonstrated to the Federal Financial Supervisory Authority (BaFin) at the end of each calendar year. Moreover, no individual investor may directly own 10 percent or more of shares. In case of non-compliance only those shares below 10 percent are entitled to the special tax treatment, also with respect to double-tax treaties. Aggregated indirect stakes above 10 percent through separate holding companies, however, are permitted. If REITs do not meet the stake limits three fiscal years in a

row, the special tax status is suspended after the third fiscal year and those shareholders with less than 3 percent of shares are entitled to financial compensation. A loss of exchange listing results in a cancellation of the tax status for the respective year. These rules assure on the one hand a sufficient German tax base in the light of existing double-tax treaties and on the other hand fungible property investments, especially for retail investors. Against this background two financial covenants for REITs are also introduced. First, not less than 90 percent of annual distributable profits have to be paid out to shareholders until the end of the following fiscal year. In order to calculate distributable profits only straight-line depreciation is permitted and half the capital gains might be allocated to a reserve for at most two years. In case REITs do not conform to this norm, tax authorities impose a fine on them, with penalty payments of at least 20 percent and at most 30 percent of the amount the actual distribution falls short of the specified amount. These penalty payments can be supplementary to those for non-compliance with the criteria for composition of assets and revenues. If one of these criteria or the profit distribution rule is violated at a time five years in a row or if the profit distribution rule is violated three years in a row by REITs, the special tax status is suspended after the respective year. Second, the equity capital has to amount to not less than 45 percent of total assets at fair value. If this norm is missed three fiscal years in a row, the favorable tax status is suspended after the third year. The loss of the special tax status is associated with a waiting period of four years before the status can be granted again. Two differences of German REIT rules compared to major international regimes are notable[17]. First, REITs are required to list at an exchange to qualify for the G-REIT status. This is in line with regulations in Belgium or France, but contrasts with well-established markets like Australia, The Netherlands or the USA, where privately held REITs obtain the same tax benefits as publicly traded ones. Whether this rule adversely affects the development of the G-REIT market is not clear ex ante[18]. Second, the exclusion of existing domestic residential properties is unparalleled in other regimes and objectively not justified. Substantial rent increases and a loss of influence on urban development as main arguments typically put forward for this exclusion do lack a sound basis, since the German tenancy law clearly limits rent increases and municipalities still dispose of the same legal framework to direct urban development. The exclusion in addition puts listed residential property corporations at a tax disadvantage to other real estate companies allowed to convert to G-REITs, which, all else being equal, should result in lower relative valuations. This tax disadvantage also becomes relevant when German residential real estate firms compete with foreign REITs for German residential properties, because they can all else equal only offer lower acquisition prices. Compliance with the set of requirements presented and the related registration of property companies as G-REITs result in two tax advantages for investors: (1) an ongoing simplification and reduction of the taxation of profits; and (2) a temporary decrease of taxes on hidden reserves (exit tax) that might be released when real properties are transferred into REITs. Currently the income of German investors from corporations and property companies is taxed according to the half income system. No imputation of corporation tax credit is

German property investment

239

JPIF 26,3

240

available. REITs, however, are exempt from corporate and trade taxes and taxation takes place only at the investor level, where all distributed profits are taxed at the individual tax rate. In addition to a more transparent and less complex taxation that is in line with international standards, this results in a reduced tax burden for REIT investments compared to non-REIT property companies[19]. This taxation method is, however, applied to open-end real estate mutual funds too. In order to attract companies to transfer their real estate holdings into G-REITs the basis for capital gains taxes is reduced by 50 percent over a three-year period starting January 1, 2007 for real properties held for at least two years at this date. If a company decides to sell its real estate to a G-REIT the assets have to be held for at least five years at this date. Due to accounting and tax rules the book value of corporate real estate in Germany is typically significantly below its market value and the realized difference or capital gain is normally fully taxed at a rate of 26.4 percent. This temporary tax relief provides a clear incentive for the contribution of real properties into G-REITs, but is associated with a lock-up period for the real properties of four years after their transfer. The relief is also available to Pre-REITs. These Pre-REITs were introduced to give property companies, which intend to gain the G-REIT status and fulfill all but the listing and investor requirements, the opportunity to benefit from the temporary tax break. In return they have to register with the federal tax authorities and are moreover obliged to apply for an exchange listing within three years and gain the G-REIT status within four years after inception. Non-compliance is penalized with a loss of the Pre-REIT status and a retroactive cancellation of the tax relief. In addition to tax advantages, and unlike open-end real estate funds, REITs are not subject to product supervision by the Federal Financial Supervisory Authority (BaFin). The multitude of detailed rules might give the impression that G-REITs are over-regulated and their policies are too restrictive, but their regulation is in fact in line with major international REIT regimes and less restrictive in practice than it appears at first glance. In addition, they have superior characteristics compared to other German indirect property investment vehicles. As laid out, G-REITs are basically more transparent and most likely more liquid than closed- and open-end real estate funds and most, especially smaller, German real estate corporations. Their transaction costs are furthermore clearly lower than the upfront fees of both types of fund. Whether they are more efficient in managing the day-to-day business, for example, by realizing economies of scale, remains to be seen. Finally, G-REITs have a temporary tax advantage compared to other indirect property vehicles due to the exit tax and an ongoing tax relief versus property corporations. Table I summarizes the main characteristics of G-REITs and the other indirect real estate investment vehicles discussed in section 2.3[20]. 4. Expected changes to the German property market The introduction of G-REITs will change the real estate investment landscape in Germany. Several estimates of the potential medium-term capitalization of REITs have been published, with figures reaching up to e100bn. The rationale for these estimates, however, is typically limited and not fully convincing. Therefore we aim at presenting reasonable arguments for potential capital flows between real estate investment vehicles, instead of providing another ball-park figure.

Gearing/leverage

Exchange listing

Liquidity

Medium: diversification level dependent on investment focus, i.e. diversified versus sector REIT High to medium (expected): minimum free float criterion and high (international) investor awareness

Medium: diversification via investments in many properties but higher volatility due to listing High to low (in practice): usually increases with market capitalization and free float

Low Private, institutional Not required; usually investments in several properties Investor-specific

Development, rents, capital gains on sale of property Capital income Tax-exempt at corporate level; taxation at investor level; netting only with capital income in current year Low Private, institutional Not required; usually investments in several properties Investor-specific

AG/AktG, HGB, IAS, REIT law Small: , 1 percent per transaction, IPO costs reduce NAV Stock exchange

G-REIT

Development, rents, capital gains on sale of property Capital income Taxation at corporate and investor level (half income system)

Small: , 1 percent per transaction, IPO costs reduce NAV (if listed) Stock exchange (if listed)

AG/AktG, HGB, IAS

Real estate company

Low: stake can be sold usually High (but artificial): fund only in OTC market at high shares can be returned discount to NAV anytime; redemption can be postponed by fund for some time No; secondary market No; guaranteed redemption via Possible; usually continuously Mandatory established by regional daily bid/ask quotes by fund traded on stock exchange exchange company No formal limit; usually 50 Maximum loan-to-value level No formal limit Maximum loan-to-value level percent of 50 percent of 55 percent

GbR, KG/HGB

Open-end funds

Separate estate/InvG, KWG, KAGG Transaction costs Very high: 5 percent front load High: 5-6 percent front load plus up to 10 percent additional soft costs Regulator No real regulations Federal Financial Supervisory Authority Income at vehicle Rents, capital gain at sale of Rents, capital gain at sale of property property Income at investor Rental income Capital income Taxation Tax-exempt on fund level; Tax-exempt on fund level; losses can be netted with other taxation at investor level; rental income of investor in netting only with other capital current and future years income in current year Minimum investment High Low Investors Private HNI, institutional Private, institutional Diversification Not required; usually Required; usually investments requirements investment in a single or few in several properties properties Investment horizon Long term due to very high Medium to long term due to transaction costs and tax high transaction costs issues Risk High: low number of properties Low: diversification via investments in many properties

Legal vehicle/basis

Closed-end funds

German property investment

241

Table I. Characteristics of German indirect real estate investment vehicles

JPIF 26,3

242

G-REITs are in principle a substitute for existing indirect property vehicles and have some clear advantages. Large short- to medium-term capital flows from those vehicles to REITs are nevertheless unlikely. Existing investments in closed-end real estate funds are largely bound by high transaction cost, i.e. high discounts when selling stakes in the OTC market, and the loss of tax incentives probably still associated with some of the fund structures. In addition, it is important for many investors to receive rental income instead of capital income as losses of closed-end funds might be offset with gains from other property holdings and vice versa. These investors will be reluctant to support the conversion of closed-end funds into REITs. The holders of open-end property fund shares also have no eminent reasons to switch immediately to REITs, since the liquidity issue currently causes no concern and the tax treatment is very similar to G-REITs. Moreover conservative investors prefer their relatively low volatility[21]. A transfer of open-end funds into a G-REIT is not in the best interest of the fund management firms as well, because they are most likely better off with the existing structure. When it comes to reinvestment of distributed fund profits and repayments of closed-end funds as well as investment of fresh money, the above mentioned arguments against REITs only partly hold. As a result G-REITs might gain from these sources at the expense of closed- and open-end real estate funds. Real estate corporations will finally convert to G-REITs on a big scale. The only real reasons against such a status change are a concentrated ownership structure and focus on German residential properties, which render a transfer impossible under current REIT regulation[22]. In the first case large homogenous portfolios might be listed as separate G-REITs. Given the conversion costs, small real estate corporations will not transfer and will instead be acquired by a few big players. In France the major listed real estate companies converted to REITs in 2003, the year of their launch. In the UK, where REITs were introduced in January 2007, REITs directly reached a market capitalization of e50bn due to the transfer of nine large listed real estate corporations (currently 14 listed UK-REITs). This potential is, however, clearly lower in Germany, with quoted property companies currently valued at roughly e12bn and only a few big firms fulfilling all G-REIT criteria. For opportunity funds G-REITs present a primary new exit tool, whereas for German residential real estate assets foreign REITs or similar structures have to be used for the time being. Both investment vehicles and the G-REITs formed out of them will lead the consolidation of the real estate sector and also pick up assets from closed- and open-end property funds. The largest medium-term asset suppliers to G-REITs will, however, most likely be the public sector and corporates through the sale of non-core real estate assets or the conversion of property subsidiaries into REITs. Real estate directly held by private households is typically too diverse and small for REITs and the other investment vehicles. The relevant real estate assets of public authorities, mainly office buildings, will not be directly listed as G-REITs, but rather sold to existing REITs or other investors who bundle several assets and then restructure the portfolios. German industrial corporates currently hold substantial real estate assets, which consist mainly of office buildings and special industrial assets. Special industrial assets like complex production facilities, are however, not suitable for investors due to limited third party usage. The remaining office and retail space properties are typically well suited for a sale or transfer to REITs, as the high standardization allows a diverse usage if the lessee terminates the lease. Office buildings were still owned roughly 70

percent by their users as of 2002[23]. Investors in REITs might additionally look for specific other property assets like hotels and logistic properties[24]. In the USA, for instance, such specialized sector REITs exist and will evolve in Germany over time, too. The temporary exit tax will act as a catalyst for corporates to sell their non-core real estate assets to G-REITs. Given that this tax advantage is available only to G-REITs, industrial companies, insurance firms and pension funds will be reluctant to sell or transfer their assets to indirect real estate investment vehicles other than G-REITs. Real estate corporations and opportunity funds who want to participate in this asset flow might feel impelled to convert to G-REITs. Whether this will actually happen or whether they are able once again to create some international tax structures to overcome this disadvantage remains to be seen. The reallocation of freed-up capital by private companies and public authorities to their core businesses will without much doubt result in an economically beneficial rise in investment activities and thus economic growth. A change from direct property investments to G-REITs will expose a larger part of the property market to the discipline of capital markets, which should result in a more transparent and efficient capital allocation and more professional management of these assets. Investors will also benefit from a broader and more diverse real estate investment universe. 5. Conclusion This study has examined the design of the dominant German indirect property investment vehicles and the new G-REITs, and has compared the products along the dimensions liquidity, transparency and risk/return profile. G-REITs show some clear advantages and as a result will change the structure of the German real estate market. Large short-term capital flows from existing indirect vehicles to G-REITs are, however, unlikely and substitution effects will materialize only over time. Instead G-REITs will in the medium term absorb the real estate assets currently directly held by public authorities and especially corporates. Whether big real estate corporations or opportunity funds at first acquire these assets to bundle them and restructure the portfolios before listing them as REITs remains to be seen. In the long run REITs will become the primary indirect real estate investment vehicle in Germany and lead to a better integration of the real estate market with the capital market. Notes 1. For this introduction and further details on the development of REITs see Bausch et al. (2005, pp. 6ff). 2. See World Federation of Exchanges and Euronext statistics and European Public Real Estate Association (2007). 3. Real estate held by insurance firms, pension funds and foreign investors are not treated as indirect investments here, except for those held indirectly via opportunity funds. For leasing companies/funds no further break-down and thus no detailed classification as direct or indirect investments is available. 4. For a discussion of the merits of direct versus indirect real estate investments, see McAllister (2000). 5. See Rat der Immobilienweisen (2006, pp. 41 and 53) and Federal Statistical Office Germany (2006, p. 645) for these figures.

German property investment

243

JPIF 26,3

244

6. See Schulte and Holzmann (2005, pp. 172, 185ff) and Rat der Immobilienweisen (2006, p. 53). Data was available for different years and if necessary was adjusted by the average value appreciation rate of total real estate assets during the respective period. It has to be noted, however, that figures on direct real estate investments are rough estimates due to the low transparency of the German real estate market, the large number of small assets and the limited availability of recent appraisal or market values for most assets. For a discussion of the transparency of the German real estate market, see Schulte et al. (2005). 7. Pfandbriefe (covered bonds) are the predominant refinancing vehicle for mortgages in Germany and the Pfandbrief Act limits the underlying mortgages to 60 percent of property value (loan-to-value ratio). In contrast, ratios of 80 percent or more are common for mortgage-backed securities used internationally. A broader use of these instruments in Germany in recent years has reduced the minimum capital requirements for private property loans. Real estate credit is usually less flexible for private households in Germany, as only fixed loans are usually offered. In other countries variable loans are common (see Organization for Economic Co-operation and Development, 2004, p. 138. 8. See Bausch et al. (2005, pp. 30ff) and Klug (2004, pp. 13-18) for the following discussions and also a more detailed description of closed-end funds. 9. See Berry et al. (1999, pp. 438-41). 10. See Bannier et al. (2006, pp. 5f, 14-16), Bausch et al. (2005, pp. 31ff) and Klug (2004, pp. 7-13, 28-48) for the following discussions and also a more detailed description of open-end funds. 11. See Maurer et al. (2004, pp. 210-15). 12. This situation is comparable to a bank run; see Bannier et al. (2006, pp. 11-14). 13. See Black et al. (2006, pp. 5ff), Maurer et al. (2004, p. 210) and Bausch et al. (2005, pp. 33ff). for the following discussions and also a more detailed description of real estate corporations. 14. See Rat der Immobilienweisen (2006, p. 53) and Schmidt-Breitung (2005, pp. 173ff). 15. A summary of the IFD’s final recommendations can be found at www.finanzstandort.com. For a detailed and regularly updated chronology of the German REITs discussion, see www. reits-in-deutschland.de 16. See Deutscher Bundesrat (2007). 17. See European Public Real Estate Association (2004) for information on major REIT regimes worldwide. 18. For an analysis of the relevance of private REITs for the success of REITs in Germany, see Bundesverband Freier Immobilien- und Wohnungsunternehmen (2006). 19. For an analysis of the wealth effects of a transfer of property companies into G-REITs, see Richter (2005). 20. See Bausch et al. (2005, pp. 35ff) and Barkow and Stanislawek (2004, p. 35) for the summary in Table I. 21. See Po¨rschke (2007). In France, similar open-end property funds structures were introduced after REITs. 22. It is, however, fair to assume that over time residential real estate assets will be allowed for G-REITs too. 23. See Bausch et al. (2005, p. 66). 24. For a survey of investor preferences with respect to G-REITs, see Knepel (2007).

References Bannier, C.E., Fecht, F. and Tyrell, M. (2006), “Open-end real estate funds in Germany: genesis and crisis”, working paper, University of Frankfurt, Frankfurt. Barkow, P. and Stanislawek, I. (2004), “German real estate, a German REIT (G-REIT) structure to leverage the importance of real estate into the listed sector”, December, HSBC Trinkaus & Burkhardt, Du¨sseldorf. Bausch, S., Goronczy, S. and Siebels, K. (2005), “Der German Real Estate Investment Trust (REIT) und der deutsche Immobilienmarkt”, October, HSH Nordbank, Hamburg. Berry, J., McGreal, S., Sieracki, K. and Sotelo, R. (1999), “An assessment of property investment vehicles with particular reference to German funds”, Journal of Property Investment & Finance, Vol. 17, pp. 430-43. Black, R.T., Rottke, N. and Becker, M. (2006), “A comparison of US real estate investment trusts and German real estate investment entities”, Working Paper No. 06-001, European Business School, Wiesbaden. Bundesverband Freier Immobilien- und Wohnungsunternehmen (2006), “Bedeutung von nicht-bo¨rsen-gehandelten REITs fu¨r den Erfolg von REITs in Deutschland”, May, Bundesverband Freier Immobilien- und Wohnungsunternehmen, Berlin. Deutscher Bundesrat (2007), “Entwurf eines Gesetzes zur Schaffung deutscher Immobilien-Aktiengesellschaften mit bo¨rsennotierten Anteilen”, March 30, Deutscher Bundesrat, Berlin. European Public Real Estate Association (2004), “EPRA global REIT survey: a comparison of the major REIT regimes in the world”, September, European Public Real Estate Association, Schiphol. European Public Real Estate Association (2007), “Top 100 European listed real estate companies”, in Dijkman, M. (Ed.), Europe Real Estate Yearbook 2007, Europe Real Estate Publishers, The Hague, pp. 638-45. Federal Statistical Office Germany (2006), Statistical Yearbook 2006 for the Federal Republic of Germany, SFG Servicecenter Fachverlage, Wiesbaden. Klug, W. (2004), Offene Immobilienfonds, Fritz Knapp Verlag, Frankfurt am Main. Knepel, H. (2007), “Welche REITs wollen Investoren?”, Immobilien & Finanzierung, Vol. 58, p. 186. McAllister, P. (2000), “Is direct investment in international property markets justifiable?”, Property Management, Vol. 18, pp. 25-33. Maurer, R., Reiner, F. and Rogalla, R. (2004), “Return and risk of German open-end real estate funds”, Journal of Property Research, Vol. 3, pp. 209-33. Organization for Economic Co-operation and Development (2004), “Housing markets, wealth and the business cycle”, OECD Economic Outlook, No. 75, pp. 127-47. ¨ Porschke, F. (2007), “REITs und Offene Immobilienfonds im internationalen Kontext”, Immobilien & Finanzierung, Vol. 58, pp. 183-5. Rat der Immobilienweisen (2006), Fru¨ hjahrsgutachten 2006, Immobilien Zeitung Verlagsgesellschaft, Wiesbaden. Richter, F. (2005), “Determinanten einer Wertsteigerung durch REIT”, working paper, University of Ulm, Ulm. Schmidt-Breitung, P. (2005), “Leasinggesellschaften”, in Scha¨fer, J. and Conzen, G. (Eds), Praxishandbuch der Immobilien-Investitionen, C.H. Beck, Mu¨nchen, pp. 173-92.

German property investment

245

JPIF 26,3

Schulte, K.-W. and Holzmann, C. (2005), “Institutionelle Aspekte der Immobilieno¨konomie”, in Schulte, K.-W. (Ed.), Immobilieno¨konomie, 3rd ed., Band I, R. Oldenbourg, Mu¨nchen, pp. 167-208. Schulte, K.-W., Rottke, N. and Pitschke, C. (2005), “Transparency in the German real estate market”, Journal of Property Investment & Finance, Vol. 23, pp. 90-108.

246

Further reading Rat der Immobilienweisen (2007), Fru¨hjahrsgutachten 2007, Immobilien Zeitung Verlagsgesellschaft, Wiesbaden. Corresponding author Ulrich Schacht can be contacted at: [email protected].

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

The current issue and full text archive of this journal is available at www.emeraldinsight.com/1463-578X.htm

Securitising China real estate: a tale of two China-centric REITs

Securitising China real estate

Michael C.H. Quek and Seow Eng Ong Department of Real Estate, National University of Singapore, Singapore Abstract

247 Received April 2007 Accepted July 2007

Purpose – There is currently no real estate investment trust (REIT) listed in China. As of date, only two REITs – GZI REIT of Hong Kong and CapitaRetail China Trust (CRCT) of Singapore – have securitised Chinese property assets. The purpose of this paper is to examine the driving forces and the obstacles surrounding China REITs, and evaluate REIT securitisation as an exit strategy for Chinese properties. Design/methodology/approach – The paper analyses the performance of the two cross-border REITs and investigates whether REITs holding Chinese assets outperform other listed REITs. Research limitations/implications – CRCT outperforms GZI REIT as well as some of the other Singapore REITs, while GZI REIT ranked second lowest in terms of price performance when compared to other Hong Kong REITs. The limited history of CRCT suggests that when a well-structured REIT holding Chinese assets can perform very well. We also infer that performance is closely linked to portfolio composition and diversification, growth story and originator reputation. Originality/value – The study shows that there is indeed a strong local demand for China REITs, and that REITs can provide an alternative source of real estate financing for Chinese developers and promote a better regulated Chinese real estate market. Keywords Real estate, Investments, Trusts, Securities, China Paper type Research paper

1. Introduction Asia has witnessed two major phenomena over the past decade. The first is the emergence of the China real estate market and the second is the rapid growth in real estate investment trusts (REITs). The Chinese economy has outperformed most developed economies over the past decade, and this is commensurately reflected in the performance of the China real estate market. REITs in Asia have grown exponentially since the first REIT was introduced in 2001. The commonality in these two phenomena – China-centric REITs – is the focus of this paper. Currently, two REITs are listed on portfolios of assets from China – the Guangzhou Investment (GZI) REIT and the CapitaRetail China Trust (CRCT). Both REITs were well received upon listing and were highly oversubscribed. The GZI portfolio comprises four office/retail buildings concentrated in the Guangzhou region, while CRCT holds seven retail malls spanning five cities across China. While both REITs promise stable distributions through sustainable long-term growth, they follow different business models and strategies. The first objective of this paper is to examine REIT securitisation – both on- and off-shore – as an exit strategy for Chinese properties. The second objective is to evaluate the performance of the two REITs holding Chinese properties. Our study shows that there is indeed a strong local demand for China REITs, and that REITs can provide an alternative source of real estate financing for Chinese developers and

Journal of Property Investment & Finance Vol. 26 No. 3, 2008 pp. 247-274 q Emerald Group Publishing Limited 1463-578X DOI 10.1108/14635780810871623

JPIF 26,3

248

promote a better regulated Chinese real estate market. CRCT outperforms the GZI REIT as well as some Singapore REITs, while the GZI REIT ranked second lowest in terms of price performance when compared to other Hong Kong REITs. The limited history of CRCT suggests that a well-structured REIT holding Chinese assets can perform very well. We also infer that performance is closely linked to portfolio composition and diversification, growth story and originator reputation. 2. Overview of China real estate market Between 1990 and 2005, China experienced an overall average GDP growth of 9.7 per cent, far in excess of growth rates in developed economies. Based on forecasts by the Economist Intelligence Unit (EIU), China will still continue to experience a GDP growth rate of around 4-6 per cent higher than average world GDP (see Figure 1). The growth of China’s urban population is both a driving force behind and a consequence of rapid economic development. Over the past ten years, the urban population has grown by 21 million per year (see Figure 2), at an average annual rate of 4.8 per cent. By 2005, urban residents totalled 562 million, accounting for 43 per cent of the population. EIU projects China’s urban population to grow at an average of about 3 per cent (19.3 million) per year over the next decade. By 2015, urban residents are estimated to number 755 million, accounting for 54.4 per cent of the total population. The urbanisation rate is expected to rise by 1.1 per cent per year. The growing population and urbanisation rate will spur demand for new homes in urban city centres. With growing demands and economic affluence, investment in Chinese real estate has been increasing rapidly as well in China at an average growth rate of 25.5 per cent (see Figures 3 and 4, based on official statistics from the China National Bureau of Statistics). “Other sources” of funding[1] account for the highest percentage of real estate funding for Chinese real estate investments (Figure 4). Around 29 per cent of the funding was self-raised by the developers themselves. They include listed property companies who are able to raise funds through the issuance of shares and state-owned

Figure 1. GDP growth: China versus the world

Securitising China real estate

249

Figure 2. Urbanisation rate of China

Figure 3. Real estate investments in China

Figure 4. Sources of funding for Chinese real estate investments

JPIF 26,3

250

developers funded by the government. The majority of the small- to medium-scale Chinese developers still obtain financing through commercial banks. Domestic lending by China’s commercial banks accounts for roughly 22 per cent of funding for real estate investments. Foreign-injected funds have been increasing over the years. This could be due to the entrance of many foreign institutional investors in the 2000s. Some of these players include Morgan Stanley Real Estate Funds, Hines, ING Real Estate, Deutsche Bank, Macquarie Bank, Goldman Sachs, Merrill Lynch and Lehman Brothers. Singapore’s Government Investment Corporation (GIC) and CapitaLand also entered the market back in the late 1990s. These foreign institutions entered into the China real estate market via active acquisitions of commercial properties (Sing, 2006). However, foreign institutions account for only 1.5 per cent of the total market. Commensurate with economic growth, the annual disposable income per capita of Chinese citizens has been growing at 9-12 per cent annually since 2001 (see Figure 5). Retail sales have also been growing steadily over the years, with a high of RMB 7,641 billion ($US932 billion) of sales in 2006 (Figure 6). The increasing disposable income growth and higher retail sales signalled a higher purchasing power on the part of the Chinese. The increased consumer spending power is a key factor underpinning the strong demand for the two China-centric REITs. 3. Brief review of securitisation in China and Asian REITs Ooi et al. (2006) examined the growth of REIT markets in Asia. Favourable regulatory changes by governments have provided the necessary catalyst for REIT growth. Two main factors driving the success of REITs in the USA and Australia are their tax exemption status and high dividend payouts. The ability of Asian REITs to replicate these factors will stimulate further growth of REITs in Asia. The introduction of REITs has also improved the liquidity and efficiency of the real estate market, as well as the corporate governance of real estate organizations in these countries. However, as more REITs are listed on stock exchanges across Asia, investors will be given a wider variety of choices. Hence discernment will be required to pick the right REITs, since

Figure 5. Annual disposable income per capita

Securitising China real estate

251

Figure 6. Retail sales in China

the availability of quality properties to acquire will be reduced. New REIT markets (e.g. China, India) are also expected to emerge in the near future, bringing further depth to Asian REIT markets. Ooi et al. (2006) also noted that REITs in Japan and Singapore had performed very well, with price appreciation dominating the total returns for most listed REITs. Newell et al. (2005) conducted a study on the dynamics of direct and indirect real estate markets in China. The study examined the risk-adjusted performance of the direct and indirect real estate markets in China from 1995 to 2002 and revealed evidence of portfolio diversification benefits for both the direct office markets and real estate companies in China against the stock market. Other than the above, there has been limited study of securitised China real estate. 4. Securitisation in China Although the China real estate market has flourished, the main route of raising funds by developers is through traditional debt financing. Commercial bank loans typically fund up to 70 per cent of the project cost, while the remaining 30 per cent[2] is required by banking regulations to be funded by the developer’s own equity. The lack of a secondary market for real estate related mortgages meant an over-reliance on commercial bank lending by developers The Chinese securitisation market truly started in 2005 and is still in its infancy. Securitisation projects are mostly focused on industrial securities rather than real estate assets. On 13 June 2003, the People’s Bank of China (PBOC) issued Circular 121 with the aim of tightening financing in the real estate sector in order to promote a more sustainable growth. The Circular 121 imposed stricter controls on loans originated by commercial banks for real estate related activities like project developments, land banking, construction and individual home mortgages[3]. Following this circular, new restrictions were introduced by commercial banks, requiring developers to obtain four official permits (namely project commencement,

JPIF 26,3

252

land use planning, construction and land use right) before banks could disburse loans. These were tough measures, as only the financially strong and well-established developers, or politically linked state-owned companies, could afford the 30 per cent registered capital up front and obtain the necessary official permits in advance. This restriction effectively restricted the main source of funding for these real estate developers and forced them to explore alternative sources of funding for real estate projects. In response to the need for alternative financing by Chinese developers, real estate trust companies emerged to fill this financing gap. Real estate trust companies, which are also known as “quasi-REITs”, operate like mortgage trusts. They are not listed entities but raise capital through private arrangement. The operations of these trust companies are governed by provisions in the “Trust Investment Companies Management Guidelines” published by the People’s Bank of China on 5 June 2002. However, there are a number of prohibitions with regards to direct or indirect acceptance deposit savings, capital issuance, foreign debts/loans borrowings, and number of units (no more than 200). A recent amendment to the Trust Investment Companies Management Guidelines was released on 24 January 2007 by the China Banking Regulatory Committee to be enforced on 1 March 2007. The new ruling brings good news for trust companies as the cap of 200 investors will be lifted for institutional investors. The new ruling now only limits individual investors of trust companies to 50 persons, and they must have an annual income exceeding RMB 200,000 and net worth of more than RMB 1 million. This raises the barrier for individual investors, making it harder for the public to invest. These Chinese “quasi-REITs” definitely still lack the market transparency and public availability associated with listed REITs. The implementation of credit tightening policies following the Circular 121 issued by China’s central bank in 2003 has forced real estate developers to seek alternative financing sources. Trust companies have come in to provide short-term bridging facilities to help real estate developers overcome the liquidity crunch caused by the retreat of commercial banks from real estate lending. However, the supply of loans by trust companies is limited and is subject to tight regulatory controls by the China Banking Regulatory Commission (CBRC)[4]. An alternative and more direct way for real estate developers to raise funds is through the direct sale of real estate assets held in their portfolios for long-term investment. However, developers may be reluctant to part their high-grade income-generating properties in prime location through the direct sale option. As a form of securitisation, REITs may therefore offer a more acceptable alternative for divestment where developers can raise funds and yet retain claims on the cash flows by keeping a controlling stake in the REIT units. 4.1 Barriers in developing a REIT market in China Potential barriers that hold back the development of a REIT market in China are discussed below. 4.1.1 Security of legal title. There is no sound legal system to ensure that the investor can obtain the legal title of the acquired properties. Investors could face great hurdles in obtaining the certificates of title of the properties if the vendor disappears or is

uncooperative after receiving the purchase price. There is always a risk that the vendor may use the purchase price for other purposes, instead of discharging the mortgages. 4.1.2 Lack of professional fund and asset management expertise. Development of an active REIT market requires the support of a strong pool of well-qualified and experienced fund and asset managers. This is currently lacking in China. In the short term, experienced fund and asset managers could be brought in by allowing foreign REITs to operate in China. Transfer of knowledge and expertise could be systematically arranged so that a pool of domestic expertise could be developed, which is essential for the long-term development of the REIT market in China. 4.1.3 Different property valuation techniques. The method of pricing of real estate assets adopted by developers or vendors in China can be different from the practice of international purchasers. This might eventually lead to disagreement on the final acquisition price. Vendors in China usually adopt the market comparable approach, which can be distorted due to the lack of efficient market information. On the other hand, international fund managers usually adopt the income approach by utilising the discounted cash flow method. 4.1.4 Lack of market transparency. Even with the current open door policy towards economic development in China, transparency in the real estate market is still very much lacking. Limited or unavailable information hinders institutional investors from conducting extensive due diligence work which is required to satisfy legal and financial requirements before a REIT listing. 4.1.5 Absence of legislative framework for REITs. Currently, there is no regulatory framework facilitating REIT listings in China. It is therefore important for relevant Chinese authorities to work closely to formulate a set of guidelines governing REIT listing and operation in China. 4.1.6 High taxation. Taxes are the main expenses for foreign investors when evaluating the options of investing or setting up REITs in China. Major taxes applicable to foreigners, foreign investment enterprises (FIEs) and foreign enterprises (FEs) doing business in China are summarized in Table I. 4.2 Benefits of REIT securitisation in China REITs are designed to encourage public ownership of commercial real property, through a tax efficient pass-through vehicle. In China, business tax, land appreciation tax and other associated documentation taxes on acquisition and disposal of real estate assets, and urban real estate taxes levied on rental income could reduce the after-tax cash flows available for distribution to REIT investors. The high taxation costs incurred in multiple categories of taxes will adversely affect the returns of investment in REITs, and make them less competitive in the face of other Asian REITs. Tax transparency is hence a necessary first step by the Chinese government to attract REIT listings in China. Other drivers of REITs in China include a large pool of quality real estate assets, strong potential demand from Chinese institutional investors and the ability of REITs to stimulate a better regulated property market in China through the influx of other real estate professional services that comes along with a REIT market. The success of a REIT listing in China will be both beneficial to the Chinese developers as well as the overall property market.

Securitising China real estate

253

JPIF 26,3

Category

Type of tax

Tax on income

Enterprise income tax (“EIT”) Standard tax rate is 33 per cent, but the tax rate could be reduced to 24 per cent for enterprises located at the coastal cities or 15 per cent for those located at the special economic development zone (Note: Currently, separate EIT systems are applicable to foreigners, FIE and FE, as contrasted with those applicable to local Chinese and domestic companies) Withholding income tax on payments to non-residents A concessionary rate of 10 per cent is currently applicable to interest, rental, royalty and other income Individual income tax (IIT) Progressive rates range from 5 per cent to 45 per cent

Tax on transactions (turnover tax)

Value-added tax General tax rate is 17 per cent, certain necessities are taxed at 13 per cent Consumption tax Tax rates range from 3 per cent to 50 per cent Business tax Tax rates range from 3 per cent to 20 per cent

Tax on specific objective

Land appreciation tax Enterprises or individuals who generate gains realized from transfer of the real property and its attachments will be taxed at the progressive rates ranging from 30 per cent to 60 per cent

Tax on property

Urban real estate tax A tax imposed on the owners, users or custodians of houses and buildings at the rate at either 1.2 per cent of the original value with certain deduction or 12 per cent on the rental value

Tax levied by finance department

Deed tax Transferees or assignees of the ownership of land use rights or real properties will be taxed and the general tax rates are ranged from 3 per cent to 5 per cent

254

Table I. Major taxes applicable to foreigners, foreign investment enterprises (FIEsa) and foreign enterprises (FEsb) doing business in China

Notes: aFIEs include Chinese-foreign equity joint ventures, Chinese-foreign contractual joint ventures, and wholly foreign-owned enterprises established in the People’s Republic of China (PRC). bFEs include foreign companies, enterprises and other economic organizations that are not Chinese legal entities, but have establishment or places in the PRC. Source: PriceWaterHouseCoopers

However, before a REIT listing can take place in China, legislative support from the Chinese government is crucial. A regulated framework and pro-taxation policies will boost investors’ confidence in the market. Also, other issues like legal titles, market transparency, property valuation methods and lack of professional expertise need to be looked into. Already, things are on the way to make REIT listing a reality in China. According to the Shanghai Daily (2007), authorities from the Shanghai Stock Exchange (SSE) are already reviewing the possibility of a REIT listing on the Shanghai Stock Exchange. The SSE set up a specialised research unit in March 2007. No timetable has been given, but SSE is determined to become one of the REIT hubs of Asia and further its development of derivatives. Also, China Business News (2007) reported that the mayor of Tianjin City has announced the submission of a REIT proposal to the central government for approval. The Chinese REIT will be known as “LingYue” REIT, and will focus on industrial properties. LingYue REIT is the product of LingYue Asset Management Company, a conglomerate of various major state-owned enterprises. The initial plan will be to procure a series of industrial properties using the model of asset-backed securities (ABS) before converting it into a REIT structure when a REIT listing can be made on the Chinese stock exchange. However analysts are skeptical on Ling Yue’s claim of being a REIT as the legislative framework is still needed for a REIT listing. Hence they predict it will still take some time for a “true” China REIT. Both the recent announcements by Shanghai Stock Exchange and LingYue REIT signify a closer step to the reality of a China REIT listing. 5. China-centric REITs In this section, we examine two REITs – Guangzhou Investment (GZI) REIT and CapitaRetail China Trust (CRCT) – that have been listed in Hong Kong and Singapore, respectively. In particular, we examine the portfolio composition, dividend yields, IPO and ex-IPO performance to gain insights on how investors view China-centric REITs. 5.1 Guangzhou Investment (GZI) REIT GZI REIT was the first REIT made up of a portfolio of properties from China. Property developer Guangzhou Investment Company Limited (GZI), an investment arm of China’s Guangzhou municipal government entity, packaged four prime shopping and office buildings in Guangzhou into a REIT for a $US230 million initial public offering (IPO) in Hong Kong. When GZI REIT was listed on 21 December 2005, it was the third REIT to be listed on the Hong Kong Stock Exchange and created a novel template for investors to tap into the value of commercial properties in China. GZI REIT was oversubscribed by 496 times in the public offering and 74 times in the institutional placement. It offered an attractive dividend yield of 6.54 per cent to unit holders, the highest among listed Hong Kong REITs back then. The four properties under GZI REIT are commercial properties, with office and retail components, located in the city of Guangzhou, Guangdong province of China. The competitive strengths of the properties include good location in Guangzhou’s Tianhe Central Business District and Yue Xiu district and the high occupancy rates achieved by the established commercial buildings. Given the prime location of the

Securitising China real estate

255

JPIF 26,3

256

properties in Guangzhou, the manager is confident of delivering a good distribution payout to unit holders. The properties will be initially focused in the Guangdong province of China before plans will be made to expand to the other provinces of China. Guangzhou is located along the Pearl River and is the capital city as well as the economic, scientific and cultural centre of Guangdong province. The Guangzhou economy is the PRC’s third largest city-level economy and enjoyed double-digit growth in the past 15 years, averaging 15.3 per cent per annum. With continued buoyant fixed asset investment, solid domestic demand, robust trade activities and high inflows of foreign direct investments, the outlook for the economy remains positive. These factors have combined to increase demand for commercial properties in Guangzhou and the properties are well positioned to benefit from such demand. GZI REIT believes that in such an economic environment, GZI REIT has considerable potential for growth. Presently, GZI REIT has three commercial buildings (office and retail) located in Guangzhou’s Tianhe CBD and one commercial building (White Horse) in Yue Xiu district. Figure 7 shows the location of the Tian He and Yue Xiu districts in Guangzhou. According to GZI’s prospectus and announcements, GZI REIT plans to grow its assets through active asset management. The manager will endeavour to increase the property yield of GZI REIT’s property portfolio and, correspondingly, the NAV per unit by actively managing the portfolio. Some of the measures which the manager intends to take include optimising rental and occupancy rates through managing lease renewals effectively, strengthening the properties’ competitive positions and capitalising on Guangzhou’s economic growth through procurement of other Guangzhou commercial properties. While GZI listed four projects for future acquisition in its prospectus, it has not exercised any of its rights to procure the properties to date even though the construction for one project (Yue Xiu New Metropolis) was completed at the end of 2006. The next property in line for completion will be the two office blocks behind Victory Plaza retail mall, which are expected to be completed at the end of 2007. In addition, all the identified future properties available to GZI REIT through GZI are located within Guangzhou. This might not be an attractive portfolio to investors as all the existing and future pipeline of commercial properties are all located in the same city, meaning little diversification. 5.2 CapitaRetail China Trust (CRCT) CapitaRetail China Trust (CRCT) was the second listed REIT that comprised a portfolio of Chinese properties. CRCT is the fourth REIT by CapitaLand Limited, after the successful listing of CapitaMall Trust (July 2002), CapitaCommercial Trust (April 2004) and Ascott Residence Trust (March 2006). It is the fourteenth REIT listed on the Singapore Stock Exchange and is made up of a portfolio of seven retail properties across five cities in China. CRCT’s public offering was oversubscribed by 39 times and attracted an institutional placement of 196 times. It was listed on 8 December 2006 in Singapore and successfully raised over $SG218.4 million (approximately $US142 million). CRCT offered a dividend yield of 5.42 per cent (first year) and 5.78 per cent (second year) to unit holders. The seven retail malls are geographically spread out

Securitising China real estate

257

Figure 7. Map of Guangzhou and its major districts

across five cities in China, namely Beijing, Shanghai, Zhengzhou, Wuhu and Huhehaote. The competitive strengths included quality retail properties in strategic locations, geographical diversification (reduces CRCT’s dependence on any single regional market), stable and quality tenant base (well-established brand names of anchor tenants like Wal-Mart, Carrefour and Beijing Hualian Group), favourable lease structure with upside potential (up to 20 years with annual step-up in base rent) and the potential for future asset enhancement to the properties. The sponsor of CRCT, CapitaLand Limited, is one of the largest listed real estate companies in Asia. The sponsor has extensive experience in creating, managing and investing in property funds and real estate financial products. It has leveraged on opportunities within the asset base of the CapitaLand Group in the origination of, and investment in, several REITs including CapitaMall Trust, CapitaCommercial Trust and Ascott Residence Trust. The sponsor also acted as the strategic partner to Hong

JPIF 26,3

258

Figure 8. Map showing location of malls under CRTC’s acquisition pipeline

Kong’s first REIT, The Link REIT, and holds an approximate 4.0 per cent stake in The Link REIT, which listed on 16 November 2005. CapitaRetail China Limited’s partnership with SZITIC and Beijing Hualian Group allowed CapitaRetail China Trust to secure seven retail malls for the initial offering. The partnerships also paved the way for CRCT’s attractive acquisition pipeline of 20 retail malls across 24 Chinese cities by year 2010. Being the second cross-border REIT to contain a portfolio of Chinese properties, CRCT had the advantage of tapping onto GZI REIT’s template when structuring its offering. The same professional team of lawyers, which was critical in structuring GZI REIT, was also hired to help with the legal structuring of CRCT (Paul Hastings, 2006). Furthermore, CapitaLand had already gained experience and credibility after successfully floating three REITS in Singapore and assisting Hong Kong’s Link REIT. Hence CapitaLand was definitely well positioned to structure CRCT. A strong acquisition growth strategy had already been highlighted in CRCT’s prospectus, where there is a potential to triple CRCT’s initial portfolio of seven properties and increase CRCT’s presence to more than 24 cities in China (see Figure 8). This is possible as CRCT has been granted rights of first refusal over properties from CapitaRetail China Development Fund (the “Development Fund”) and CapitaRetail China Incubator Fund (the “Incubator Fund”). The sponsor has secured, for investment by the Development Fund, a portfolio of 19 retail mall developments through a equity joint venture with Chinese state-owned Shenzhen International Trust & Investment Co., Ltd and its subsidiaries and associates (the “SZITIC Group”). Through the sponsor’s strategic alliance with the Beijing Hualian Group (the sixth largest retail group in China), the Incubator Fund may invest

in a potential pipeline of retail malls to be anchored by the Beijing Hualian Group, including the prime Xihuan Plaza Retail Mall in Xizhimen, Beijing, which was acquired at approximately $US162.4 million. 6. Performance of GZI REIT and CapitaRetail China Trust As mentioned earlier, both GZI REIT and CRCT attracted a huge subscription for its offering. Overall, GZI REIT was oversubscribed by 117 times and CRCT by 167 times, far exceeding the interest generated by fellow listed Hong Kong REITs (H-REITs) and Singapore REITS (S-REITs). Both REITs saw a price increase on their first day of IPO listing, with GZI REIT seeing an increase of 13.82 per cent and CRCT an increase of 59.29 per cent. This section will analyze the performance of GZI REIT and CRCT and determine which REIT performed better (see Table II for a brief comparison). 6.1 Subscription level Table III and Figure 9 show the overall subscription rate (averaging demand from both public offering as well as institutional demand) of the various H-REITs and S-REITs during their IPO offering. Both REITs attracted an overwhelming demand that far exceeded what their peer REITs achieved. 6.2 Dividend yield The yield of the various REITs (reported distributable income of individual REITs by their listing IPO price) is compared with other yield indices in Table IV and Figure 10. Indices providing stable yield distributions include the ten-year government bond yield and fixed deposit rates. The NAREIT Asia REITs yield index computes the average yield of REITs in other parts of ASIA (Japan, Australia, Singapore, Hong Kong, New Zealand and South Korea). GZI REIT promised a higher dividend yield of 6.54 per cent in its prospectus as compared to Link REIT’s 5.53 per cent and Prosperity REIT’s 5.31 per cent. Its actual annualised distributions in the first year increased by 0.18 per cent to 6.72 per cent, which was higher then Link REIT’s actual yield of 6.09 per cent but less than Prosperity REIT’s actual yield of 7.48 per cent. Presently, the H-REIT that promises the highest yield to unit holders is Hong Kong Henderson Land Development’s Sunlight REIT at 8.49 per cent. One point to highlight is the high ten-year government bond yield and fixed deposit rates in Hong Kong. On the day of GZI REIT’s listing, GZI REIT’s promised distribution yield to unit holders was only at a difference of 2.25 per cent for the ten-year government bond yield (at 4.29 per cent) and 2.09 per cent difference for fixed deposit rates (at 4.45 per cent). The average H-REIT promised yield was 6.26 per cent (calculated based on yield promised on IPO and IPO prices) and 4.33 per cent for government bond yield, a difference of only 1.93 per cent. With this small difference in yield, investors would rather choose to park their investments in other higher yielding vehicles such as stocks or conversely invest in government bonds if they want a stable return. This might explain why investors’ response to H-REITs has been lukewarm recently, following the lacklustre performance of both Champion and Sunlight REITs (the fourth and fifth H-REIT listings). The average promised yield for the 15 S-REITs was 6.5 per cent (calculated based on yield promised on IPO and IPO prices) while the Singapore ten-year government

Securitising China real estate

259

JPIF 26,3

GZI REIT

CRCT

Date of IPO listing Location of REIT listing Sponsor company

21 December 2005 Hong Kong Guangzhou Investment Company (GZI) (Guangzhou Municipal Government Entity)

Location of offshore companies to hold project companies Property portfolio

British Virgin Islands

December 2006 Singapore CapitaLand Group (One of Asia’s largest listed property developers, successful prior listing of three REITS in Singapore. Subsidiary of Temasek Holdings of Singapore) Barbados

260

Number of properties available Location of properties

Commercial buildings (mixture of office and retail) 4 All located in Guangzhou

Retail malls

7 Across five Chinese cities: Shanghai (1), Beijing (3), Inner Mongolia (1), Henan (1) and Anhui (1) Province Quality retail properties in Competitive strengths of the Quality commercial buildings strategic locations portfolio located in prime areas of Geographically diversified in Guangzhou High occupancy rates achieved five cities across China Well-established anchor tenants Favourable lease structure with upside potential Asset enhancement strategy Growth strategy First right of refusal granted for First right of refusal granted for investments by GZI to GZI REIT projects under CapitaRetail by year 2010 China Development Fund and Acquisition pipeline of for foura Incubator Fund by year 2010 Acquisition pipeline of 20a other commercial properties under development by GZI retail assets across 24 Chinese Active enhancement of existing cities, potentially tripling properties CRCT’s initial portfolio Initial focus on commercial Active asset enhancements properties in Guangzhou Asset value at IPOb $HK4,005,000,000 RMB3,451,000,000 ($US513,461,538) ($US441,304,348) Size of IPO offering 583 million units 193.3 million units Value of IPO offering raised $HK1,792,725,000 $SG218,429,000 ($US229,836,538) ($US141,837,013) 167 (overall) 117 (overall) Subscription rate for IPO 39 (public offering) 496 (public offering) offering (number of times 196 (institutional placement) 74 (institutional placement) oversubscribed) Price performance on day of IPO Increased 13.82 per cent Increased 59.29 per cent listing Promised dividend yield to unit 6.54 per cent (first year) 5.42 per cent (first year) holders (as of IPO prospectus) 5.78 per cent (second year) Percentage of distributable 100 per cent for first three years At least 90 per cent Income promised to unit holders (2006, 2007 and 2008 financial year. Thereafter at least 90 per cent) Table II. A brief comparison of GZI REIT and CRCT

Notes: aMalls located in the locality. b$US1 ¼ $HK7:80, December 2005; $US1 ¼ RMB7:82, December 2006) Source: GZI & CRCT Prospectus, Datastream, author’s compilation

No.

Name of REIT

Hong Kong 1 The Link REIT 2 Prosperity REIT 3 GZI REIT 4 Champion REIT 5 Sunlight REIT Singapore 1 CapitaMall Trust 2 Ascendas REIT 3 Fortune REIT 4 CapitaCommercial Trust 5 Suntec REIT 6 Mapletree Logistics Trust 7 Macquarie MEAG Prime REIT 8 Allco Commercial Trust 9 Ascott Residence Trust 10 K-REIT Asia 11 Frasers Centrepoint Trust 12 CDL Hospitality Trust 13 Cambridge Industrial Trust 14 CapitaRetail China Trust 15 First REIT

Public offering

Institutional placement

Overall subscription level

19.0 260.0 496.0 6.5 8.3

19.0 19.0 74.0 3.5 5.0

19.0 43.0 117.0 3.8 5.3

2.0

3.1

1.7

2.0

10.6 26.3 12.3

13.7 17.5 33.4

6.0 6.6 11.3 39.0 26.3

2.0 1.9 1.8 196.0 6.6

5.0 5.0 3.7 N/A 9.6 44.0 35.0 3.9 N/A N/A 2.2 2.0 2.0 167.0 7.6

Note: N/A, IPO subscription rate unavailable as REIT offering was conducted via “distribution in species” Source: Company data, newspaper reports, author’s compilation

bond yield was 3.29 per cent. This works out to a difference of 3.21 per cent, which is 1.28 per cent higher than the difference of average H-REITs’ yield to the Hong Kong government bond yield. CRCT promised a dividend yield of 5.42 per cent in its first year to unit holders. This was actually the third lowest promised yield among the other 14 S-REITs during their IPO. CRCT’s promised dividend yield was only 2.38 per cent higher than the government bond and 3.62 per cent higher than the fixed deposit rates. Despite this, in terms of subscription rate and price performance, it outperformed other S-REITs which promised higher yield. 6.3 Price performance on first day of IPO Table V and Figure 11 show the IPO price performance on the first day of listing. GZI REIT is compared to the other H-REITs and CRCT with other S-REITs. GZI REIT achieved a positive price increase of 13.82 per cent on the first day of IPO. This, however, was less then the first-day IPO price increase achieved by the earlier Link and Prosperity H-REITs. One thing to highlight was the subsequent poor price performance of Champion REIT (listed on 24 May 2006) and Sunlight REIT (listed 21 December 2006).

Securitising China real estate

261

Table III. IPO subscription rates of H-REITs and S-REITs

JPIF 26,3

262

Figure 9. IPO subscription level of H-REITs and S-REITs

CRCT, however, performed very well on the first day of listing, with a price increase of 59.29 per cent, placing it as the second best performer out of 15 S-REITS. This result is just behind fellow CapitaLand’s Ascott Residence Trust of 69.12 per cent. 6.4 Price performance after IPO compared to other REITs in their country Figures 12 and 13 show the price performance of GZI REIT and CRCT after their IPO, compared over a period from 21 December 2005 (GZI REIT’s IPO listing) until 30 March 2007. GZI REIT and CRCT’s price performances were compared to the other H-REITs and S-REITs, respectively. The broader stock market index for the respective countries – i.e. the Hang Seng Composite Index and the SGX All Shares Index – are included as a reference to the stock market’s performance.

July 2002 November 2002 August 2003 May 2004

11 December 2006

8 December 2006

19 July 2006 25 July 2006

28 April 2006 5 July 2006

31 March 2006

30 March 2006

20 September 2005

9 December 2004 28 July 2005

17 19 12 11

6.35

6.00

8.15

7.50

N/A N/A

5.85 6.71

5.50 6.37

5.42 8.62

9.37 6.50

6.01

8.99 5.22

5.77

5.71

6.81 6.37

5.68 5.82

5.12

7.66 8.60 7.03

6.09 7.48 6.72 N/A N/A

7.06 8.00 6.50

5.53 5.31 6.54 5.46 8.49

Actual yield for first year (per cent, based on IPO price)

1.80 1.80

1.80

1.80 1.80

1.80 1.80

1.80

1.70

1.70

0.69 1.29

0.81 0.75 0.81

4.44 4.39 4.46 4.74 3.99

Fixed deposit rate (per cent)a

Notes: aRates as of date of REIT listing. N/A, company reports on distributable income not yet available

Singapore 1 CapitaMall Trust 2 Ascendas REIT 3 Fortune REIT 4 CapitaCommercial Trust 5 Suntec REIT 6 Mapletree Logistics Trust 7 Macquarie MEAG Prime REIT 8 Allco Commercial Trust 9 Ascott Residence Trust 10 K-REIT Asia 11 Frasers Centrepoint Trust 12 CDL Hospitality Trust 13 Cambridge Industrial Trust 14 CapitaRetail China Trust 15 First REIT

25 16 21 24 21

Hong Kong 1 The Link REIT 2 Prosperity REIT 3 GZI REIT 4 Champion REIT 5 Sunlight REIT November 2005 December 2005 December 2005 May 2006 December 2006

Date of listing

No. Name of REIT

Expected yield for first year (per cent, Based on IPO price)

3.04 3.06

3.61

3.57 3.64

3.59 3.51

3.54

2.93

2.65

3.41 2.75

3.80 2.90 3.43

4.38 4.29 4.29 4.88 3.83

Ten-year government bond (per cent)a

5.08 5.04

5.57

5.56 5.64

5.49 5.53

5.46

5.66 5.00

NAREIT Asia REITS (per cent)a

Securitising China real estate

263

Table IV. Dividend yield of H-REITs and S-REITs compared with other yield indices

JPIF 26,3

264

Figure 10. Dividend yield of H-REITs and S-REITs compared with other yield indices

From Figure 12, we can clearly see that during the period, The Link REIT was the best performing H-REIT, constantly performing at least 40 per cent above its IPO price. The performance of the other four H-REITs was lacklustre, with GZI REIT being the second best, performing close to its IPO price while the others all performed below their IPO prices. On the whole, the price performance of H-REITs has been poor, with three out of five H-REITs performing below their IPO prices.

No.

Name of REIT

Hong Kong 1 The Link REIT 2 Prosperity REIT 3 GZI REIT 4 Champion REIT 5 Sunlight REIT Singapore 1 CapitaMall Trust 2 Ascendas REIT 3 Fortune REIT 4 CapitaCommercial Trust 5 Suntec REIT 6 Mapletree Logistics Trust 7 Macquarie MEAG Prime REIT 8 Allco Commercial Trust 9 Ascott Residence Trust 10 K-REIT Asia 11 Frasers Centrepoint Trust 12 CDL Hospitality Trust 13 Cambridge Industrial Trust 14 CapitaRetail China Trust 15 First REIT

IPO price

Price at end of first trading day (local currency)

IPO price performance (per cent)

10.30 2.16 3.08 5.10 2.60

11.80 2.60 3.50 4.30 2.43

14.56 20.37 13.82 (2 15.69) (2 6.54)

0.96 0.88 4.75a 1.00 1.00 0.68 0.98 1.00 0.68 1.04 1.03 0.83 0.68 1.13 0.71

0.99 0.93 4.65a 1.18 1.10 0.88 1.05 0.98 1.15 1.49 1.05 0.86 0.68 1.80 0.73

3.13 5.11 (2 2.11) 18.00 10.00 29.41 7.14 (2 2.00) 69.12 43.27 1.94 3.01 0.00 59.29 2.11

Note: aIn $HK Source: Datastream (author’s compilation)

For S-REITs, CapitaMall Trust was the top performer, witnessing a four-fold price increase since its IPO. Approximately four months since its listing, CRCT performed well with a price increase of over 2.7 times and outperformed several S-REITs in terms of price performance to rank in fifth place. Unlike H-REITs, the majority of S-REITs performed above their IPO prices. Although GZI REIT was the second best performing H-REITs, its price only rose close to 20 per cent above its IPO price in the initial few months and started dipping after the third month to perform at around its IPO price level. On the other hand, CRCT enjoyed superior performance, rising more than 250 per ent above its IPO price within a span of two months and outperforming ten other S-REITs by its fourth month. 6.5 Price performance after IPO compared to other price indices A comparison was made (see Table VI and Figures 14 and 15) between the performance of GZI REIT and CRCT compared with other price indices. Stock market indices like Hong Kong’s Hang Seng Index and Hang Seng Composite Index (the broader index) and Singapore’s Straits Times Index as well as the SGX All Shares Index were used to compare the REITs against the stock market performances. The S&P/Citigroup REIT Index is a REIT index compiled by Standard & Poor’s and is used

Securitising China real estate

265

Table V. Price performance on first day of IPO

JPIF 26,3

266

Figure 11. Price performance on first day of IPO

as a comparison of GZI REIT and CRCT against the overall REIT market performance of Hong Kong and Singapore, respectively. NAREIT stands for the National Association of REITs (NAREIT). The NAREIT Asia REITs index tracks the performance of REITs in the major Asia-Pacific countries, including Japan, Australia, Singapore, Hong Kong, New Zealand and South Korea. NAREIT Asia Non-REITs index was included as a benchmark of the performance of other property stocks. By comparing both REITs in their initial four months of listing, the price returns of GZI REIT increased by almost 20 per cent above its IPO listing but started to fall nearing the second month and descended below the S&P/Citigroup H-REIT price index. On the other hand, CRCT was already outperforming all the other indices by over 50 per cent within the first few weeks of listing. Its price performance index continued to ascend rapidly and by the second month, it already outperformed all the other price indices by over 200 per cent, exceeding 270 per cent by 30 March 2007. In comparison with other price indices, it is evident that CRCT outperformed GZI REIT in the initial months of listing.

Securitising China real estate

267

Figure 12. Price performance of H-REITs from 21 December 2005 until 30 March 2007

Looking at GZI REIT’s performance from IPO until 30 March 2007, we can see that after the fourth month, GZI REIT was starting to perform below all the other price indices. Since then, GZI REIT has been underperforming all the other indices for the majority of the time. Due to the short time frame of data for CRCT, it is uncertain whether CRCT might face a similar weakened performance after the initial few months. 6.6 Cumulative market adjusted returns for GZI REIT and CRCT In order to determine whether GZI REIT or CRCT provides better excess returns for investors, a cumulative market adjusted returns graph was plotted for GZI REIT and CRCT for the period 21 December 2005 to 30 March 2007 (see Figure 16). The market measure was based on the broader stock indices of the Hang Seng Composite Index for GZI REIT and SGX ALL Shares for CRCT. GZI REIT performed slightly above the mean of 1.0 in the initial few weeks before dipping below the line. Since then, the cumulative excess returns have been dropping steadily, reaching around 0.6 on 30 March 2007. On the other hand, CRCT’s excess return index dipped slightly below the mean in the initial few weeks before climbing exponentially in the subsequent few months to reach more than 1.5 times its mean. This analysis showed that CRCT provided a much higher excess return from the stock market for investors as compared to GZI REIT, whose excess return is heading South. 7. Conclusion China’s phenomenal rate of economic growth and its emergence as one of the world’s largest economies continue to drive real estate development and prices. This fundamental transformation creates tremendous opportunities for the real estate industry in China, which is still funded primarily through domestic bank loans.

JPIF 26,3

268

Figure 13. Price performance of S-REITs from 21 December 2005 until 30 March 2007

21 December 2005 12 January 2006 24 January 2006 15 February 2006 21 February 2006 13 March 2006 22 March 2006 3 April 2006 12 April 2006

Date (weekly)

HangSeng Index 100.00 103.27 102.03 101.33 102.67 102.11 102.77 105.53 107.16

GZI REIT

100.00

113.82

117.07

111.38

110.57

111.38

110.57 110.57 106.50

106.86 110.70 111.65

105.32

106.14

104.88

104.41

104.43

100.00

113.70 113.02 115.88

112.56

118.14

111.39

101.79

99.86

100.00

Hong Kong HangSeng Composite S&P/Citigroup REIT Index Index (Hong Kong)

101.71 105.11 103.54

100.40

102.40

101.29

105.79

101.81

100.00

NAREIT Asia REITS

96.46 99.41 96.68 (continued)

93.83

98.26

98.28

109.92

107.43

100.00

NAREIT Asia Non-REITS

Securitising China real estate

269

Table VI. Price performance of GZI REIT and CRCT (approximately four months from IPO)

Table VI. 100.00 101.94 104.21 109.95 112.31 112.98 108.34 109.72 108.79 112.78

100.00

158.41

185.84

255.75

253.10

267.26

261.06 256.64

255.75

276.11

Straits Times Index

116.20

111.44

109.60 110.50

114.20

114.13

111.37

104.11

101.25

100.00

123.16

120.13

117.15 119.38

120.07

122.95

115.34

111.32

101.44

100.00

Singapore S&P/Citigroup REIT Index SGX All SHARES (Singapore)

Source: Datastream, S&P/Citigroup REIT Index, NAREIT, author’s compilation

8 December 2006 20 December 2006 02 January 2007 24 January 2007 08 February 2007 20 February 2007 28 February 2007 9 March 2007 20 March 2007 30 March 2007

Date (weekly)

CapitaRetail China Trust

112.46

113.00

112.85 111.33

115.06

115.81

109.05

107.38

101.98

100.00

NAREIT Asia REITS

115.87

113.99

112.87 114.11

114.95

111.78

107.49

102.58

99.35

100.00

NAREIT Asia Non-REITS

JPIF 26,3

270

Securitising China real estate

271

Figure 14. Price performance of GZI REIT compared to other price indices

Figure 15. Price performance of CRCT compared to other price indices

Although plans are still under way to introduce REIT legislation in China, two China-centric REITs have been listed offshore in Hong Kong and Singapore. Both REITs were very well subscribed at offering and posted good first-day returns. However, CRCT strongly outperformed GZI at and after IPO. Our study shows that there is indeed a strong local demand for China REITs, and that REITs can provide an alternative source of real estate financing for Chinese developers and promote a better

JPIF 26,3

272

Figure 15. GZI and CRCT cumulative excess returns

regulated Chinese real estate market. CRCT outperforms GZI REIT as well as some of the other Singapore REITs, while GZI REIT ranked second lowest in terms of price performance when compared to other Hong Kong REITs. The limited history of CRCT suggests that a well-structured REIT holding Chinese assets can perform very well. We attribute the outperformance to portfolio composition and diversification, growth story and originator reputation. Therein are valuable lessons to be learned for potential REIT originations. Notes 1. Other sources include funding from state-owned enterprises, undeclared financing by state-owned banks and equity investments by investors whom wish to remain anonymous for tax reasons. 2. This was increased to 35 per cent in 2004. 3. One of the main rules stipulated that developers would have to put in 30 per cent of the total investment from their own capital prior to application for bank financing, a ratio that was increased to 35 per cent in 2004. 4. A noticed issued by the People’s Bank of China in September 2005 regarding the possibility of enacting Circular 212 may dampen the growing real estate trust sector. The new Circular 212, not yet officially enacted, proposed to enforce equally strict requirements like that in Circular 121 for developers borrowing from trust companies. This means that developers will also need to finance 35 per cent of the project with equity and be required to obtain all four certificates before they are allowed to borrow from trust companies. If this rule is enacted, Chinese developers will once again be short of another financing alternative.

References China Business News (2007), “Tianjin’s LingYue REIT will be the first REIT in China”, China Business News, January, p. 12.

Newell, G., Chau, K.W., Wong, S.K. and McKinnell, K. (2005), “Dynamics of the direct and indirect real estate markets in China”, Journal of Real Estate Portfolio Management, Vol. 11 No. 3, p. 263. Paul Hastings (2006), “Paul Hastings advises GZI REIT on first Chinese property REIT to list in Singapore”, available at: www.paulhastings.com/NewsDetail.aspx?NewsId ¼ 284 (accessed 16 January 2007). Shanghai Daily (2007), “Shanghai Stock Exchange may launch REITs”, Shanghai Daily, 1 February. Sing, T.F. and various authors from Pacific Star and Chinese Real Estate Chamber of Commerce (2006), A Practical Guide to Real Estate Investment Trusts (REITs): Prospects for establishing a REIT Market in China. Further reading Ascott Residence Trust (2006), “Opening speech by Ms Ho Ching”, 31 March 2006, available at: www.temasek.com.sg/news_room/press_speeches/31_03_2006.htm (accessed 20 February 2007). Brueggeman, W.B., Chen, A.H. and Thibodeau, T.G. (1984), “Real estate investment funds: performance and portfolio considerations”, Journal of the American Real Estate and Urban Economics Association, Vol. 12 No. 3, pp. 333-54. Chen, J. and Peiser, R. (1999), “The risk and return characteristics of REITs – 1993-1997”, Real Estate Finance, Vol. Spring, pp. 61-8. China Banking Regulatory Commission (2007), available at: www.cbrc.gov.cn/chinese/home/jsp/ index.jsp (accessed 22 February 2007). Chinese Real Estate Information Centre (2005), available at: www.realestate.cei.gov.cn/ (accessed 27 January 2007). Clayton, J. and MacKinnon, G. (2001), “The time-varying nature of the link between REIT, real estate and financial asset returns”, Journal of Real Estate Portfolio Management, Vol. 7 No. 1, pp. 43-54. Eichholtz, P.M.A. (1997), “Real estate securities and common stocks: a first international look”, Real Estate Finance, Vol. 14 No. 1, pp. 70-4. European Public Real Estate Association (2007), “FTSE EPRA/NAREIT Asia Index data”, available at: www.epra.com/indices.jsp (accessed 8 February 2007). Ghosh, C., Miles, M. and Sirmans, C.F. (1996), “Are REITs stocks?”, Real Estate Finance, Fall, pp. 46-53. Glascock, J.L., Lu, C. and So, R.W. (2000), “Further evidence on the integration of REIT, bond, and stock returns”, Journal of Real Estate Finance and Economics, Vol. 20 No. 2, pp. 177-94. Han, S.S. (1998), “Real estate development in China”, Journal of Real Estate Literature, Vol. 6, pp. 121-33. Hartzell, D.J., Stivers, H.M., Ludgin, M.K. and Pire, T.J. (1999), “An updated look at constructing a public and private real estate portfolio”, Real Estate Finance, Summer, pp. 49-57. Ibbotson Associates (2006), “Portfolio diversification through REITs”, NAREIT, September/October. Liang, Y. and McIntosh, W. (1998), “REIT style and performance”, Journal of Real Estate Portfolio Management, Vol. 4 No. 1, pp. 69-78. Liang, Y. and Whitaker, W. (2000), “Style attributes of equity REITs”, Real Estate Finance, Spring, pp. 31-6.

Securitising China real estate

273

JPIF 26,3

274

Liang, Y., McIntosh, W. and Webb, J.R. (1995), “Intertemporal changes in the riskiness of REITs”, Journal of Real Estate Research, Vol. 10 No. 4, pp. 427-43. Mueller, G.R. and Laposa, S. (1996), “PREIT returns: a property-type perspective”, Real Estate Finance, Spring, pp. 45-55. Mueller, G.R., Pauley, K.R. and Morill, W.K. (1994), “Should REITs be included in a mixed-asset portfolio?”, Real Estate Finance, Vol. Spring, pp. 23-8. National Association of Real Estate Investment Trusts (2007), “NAREIT Index data”, available at: www.nareit.com (accessed 8 February 2007). Paul Hastings (2005), “Paul Hastings advises GZI REIT on first Chinese property REIT to list in Hong Kong”, available at: www.paulhastings.com/NewsDetail.aspx?NewsId ¼ 85 (accessed 16 January 2007). People’s Daily Online (2006), “Total assets of Chinese insurance industry to hit 2 trillion yuan this year”, 15 November, available at: http://english.people.com.cn/200611/15/ eng20061115_321560.html (accessed 21 February 2007). Pramerica Real Estate Investors (2005), “Global REITs: a new platform of ownership”, available at: www.investmentmanagement.prudential.com/view/page/8483 (accessed 20 January 2007). PriceWaterHouseCoopers (2007), “Overview of PRC taxation system”, available at: www.pwchk. com/home/eng/prctax_corp_overview_taxation.html (accessed 12 February 2007). Reuters (2007), “Insurers poised to spur China commercial property”, 28 February. Standard & Poor’s (2007), “S&P/Citigroup REIT Index data”, available at: www.globalindices. standardandpoors.com/sandp/index.jsp (accessed 20 February 2007). State-Owned Assets Supervision and Administration Commission of the State Council (2003), “Enterprise statistics and auditing in year 2002”, p. February, available at: www.sasac. gov.cn/eng/eng_qytj/eng_qytj_0001.html (accessed 12 February 2007). Surveyor Times (2006), Selection of Papers from HKIS Annual Conference 2006, Surveyor Times, Hong Kong, pp. 37-44. (The) Standard Newspaper (2007), “Home run: Initial public offerings soared in 2006, particularly in December, when 17 listings floated shares”, The Standard Newspaper, 2 January. Trombly, M. (2006), “Weak legislation hindering China’s securitization market”, Asset Securitization Report, 26 June, pp. 22-3. Zhu, J. (2004), “From land use right to land development right: institutional change in China’s urban development”, Urban Studies, Vol. 41 No. 7, pp. 1249-67. Corresponding author Michael C.H. Quek can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

Sustainable Brownfield Regeneration: Liveable Places from Problem Spaces

Book Review : Sustainable Brownfield Regeneration: Liveable Places from Problem Spaces The Reviewers Frances Plimmer, Senior Research Officer, The College of Estate Management, Reading, UK

RR 2008/1

Review Subject: Sustainable Brownfield Regeneration: Liveable Places from Problem Spaces Edited by Tim Dixon, Mike Raco, Philip Catney and David N. Lerner

Publisher Name: Blackwell Publishing

Place of Publication: Publication Year: 2007

ISBN: 13: 978-1-4051-4403-2

Sustainable Brownfield Regeneration: Liveable Places from Problem Spaces

Article type: Review

Keywords: Emerald Journal: Journal of Property Investment & Finance

Volume: 26

Number: 3

Year: 2008

pp. 277-279

Copyright: © Emerald Group Publishing Limited

ISSN:

Sustainable Brownfield Regeneration: Liveable Places from Problem Spaces

1463-578X The UK government's declared policy of ensuring three million homes by 2020 seeks to ensure that 60 per cent of this development will take place on brownfield sites. The need to understand the UK policies, processes and practices in brownfield regeneration is therefore vital for all stakeholders involved in the provision of the government's stated housing target. However, this book is not merely about the provision of residential accommodation. Perhaps more importantly, it recognises the wider need of urban environments for sustainable outcomes and is therefore of relevance for those involved in securing urban regeneration for a range of social, environmental, recreational, commercial and financial outcomes. The stated aims of this book are two-fold. Firstly the aim is “to examine the ways in which science and social science research disciplines can be brought together to help solve important brownfield regeneration issues”; and the second is “to assess the efficiency and effectiveness of different types of regeneration policy and practice, and to show how ‘liveable spaces’ can be produced from ‘problem places’” (p. 5). The text reports on the outcome of a research project funded by the Engineering and Physical Science Research Council (EPSRC), called SUBR:IM (Sustainable Urban Brownfield Regeneration: Integrated Management). SUBR:IM was a research consortium formed in 2003, designed to deal with the perceived discontinuities of previous attempts at dealing with the technologies of regeneration of brownfield sites, which had often done so in a deterministic way and without incorporating an understanding of the impact on communities and other stakeholders, and vice versa . SUBR:IM brought together ten major research institutions to work on 18 inter-related projects: the researchers involved had a range of relevant experience and were drawn from a wide spectrum of science and social science disciplines. SUBR:IM had four principal themes:the property development and investment industries and their role in brownfield regeneration;the process of governance and multi-level decisionmaking relating to brownfield regeneration, including institutional structures and community engagement, risk, trust and systems of democratic representation;the development of robust technical solutions to contamination and the examination of the impact of climate change within this; andthe ways in which integrated solutions to brownfield renewal can be developed, including how the greening of former brownfield spaces can open up new opportunities for urban regeneration.The book documents the core evidence and findings from the SUBR:IM research projects in a series of chapters, grouped around the main headings of Regeneration, Remediation, and Joined up Solutions. The contents are presented, not as a collection of research papers, but as a framework within which the various chapters can be integrated and developed in a theoretically focused and robust manner. “[The book] adopts an integrated approach to the subject by drawing on the lessons learned from the research, not only from the individual projects themselves, but also from the synergies established through

Sustainable Brownfield Regeneration: Liveable Places from Problem Spaces

the process of working together in a multidisciplinary team” (p. 4). All of the SUBR:IM projects developed cross-cutting methodologies and approaches, with some which were designed to weave together the broader research findings. There is an English focus to the book, with case studies drawn from the Thames Gateway and Greater Manchester but it is clear that the problems dealt with within the text are not uniquely English, although the organisational policies and processes may be. Of particular interest to researchers is the Introductory section of the text, which uses the experience of the project to provide a reflective and instructive review of the team's experience in multi-disciplinary and cross disciplinary research which, as they say, “is likely to become more and more significant across the social and technical sciences as the value of such work in tackling multiple urban problems becomes more apparent”. This section discusses what it calls “the challenge to disciplinarity” in order to break down boundaries which limit outlooks and perspectives and which undermine the potential for researchers to develop broader and more imaginative perspectives and ways of viewing the world. The text also introduces the concept of transdisciplinarity “which has as its core concept a break away from disciplinary thinking by not only unravelling complexities and confronting disjointed knowledge, but also by creating new knowledge paradigms for solving problems” (p. 13). This Introduction concludes with “four core findings” which have significance for other work in this field:such consortia have the potential to make researchers think more actively about their own expertise, what it consists of and how it relates to other disciplines;multi-disciplinary research needs to be a key element in the very early phases of the process if it is to develop into anything like an effective work programme and genuinely transdisciplinary ways of working; as well as common conceptions of key terms;the formation of effective user networks and dissemination strategies is a core part of the programme; andresources need to be available for “consortium-building”, including the organisation of leadership, and for the early management processes which are both time-consuming and distracting from the intellectual and academic questions raised by the research.Inevitably and appropriately, however, the focus of the text is on the very specific issues of regeneration, remediation and evaluation of the greening of brownfield land and will be recognised as an extremely valuable contribution to an ongoing and important process. As a substantive and authoritative investigation and demonstration of one country's approach to a very widespread issue and process, this book should attract international interest.

REITs - a global phenomenon

REITs - a global phenomenon Article Type: Guest editorial From: Journal of Property Investment & Finance, Volume 26, Issue 3. This Special Issue focuses on real estate investment trusts (REITs) and their operation in four different markets across the world, two from Europe and two from Asia. While property markets are always essentially local, the REIT structure has recently become a truly global vehicle and its dominance is set to spread further. As a result, investors can now get ready exposure to global real estate, though some would question that statement, of which more below. REIT vehicles are now available in 25 countries around the world, including all the major markets USA, Japan, Germany, the UK and France. But their REIT vehicles are not identical though they share a number of common features. These features are: ●

● ●

the entity derives the bulk of its income/profit from property ownership/investment (not development); most of that income/profit is distributed to shareholders; and then the entity pays no tax on its income and capital gains (though its shareholders may well pay tax on the distributions they receive).

As a result the REIT achieves the vision of President Eisenhower[1], who wanted the “man in the street” to be able to invest in real estate in a similar manner to the wealthy individual or large insurance company. In particular, REIT shareholders are not subject to the double taxation that occurs where a conventional corporation owns real estate, just as a direct property owner pays tax on the net income once. The REIT world has seen phenomenal growth over the last 15 years. While the chief catalyst for that growth has undoubtedly come from the USA, the roles of the Australians and the Dutch in the story must not be ignored. Arguably the Australian “Listed Property Trust” market is the most developed globally, with LPTs being the preferred route for most investors into Australian real estate. Moreover, LPTs comprise 10 per cent of the Australian stock index, a much larger proportion than is found anywhere else, and Australia punches above its weight in the global REIT market cap (11 per cent). The Dutch REIT vehicle the “belegging instelling” was instigated in 1969, so Europe has had some exposure to REITs for a while too. But The Netherlands’ role in the story is not just as a pioneer in Europe; their institutions have led the way as global investors in REITs, and it is no coincidence that

REITs - a global phenomenon

LaSalle’s first client with a global REIT portfolio was a Dutch pension fund. From these pioneers, there has been a steady increase over the last ten years in the number of countries with a REIT product and in the total market cap. The latter has increased from US$53bn in 1995 to US $146bn in 2000 and to US$518bn at the end of 2007, so REITs now comprises a very significant market in their own right. The four papers in this edition are: 1. Andrew Baum and Steven Devaney’s “Depreciation, income distribution and the UK REIT”; 2. Chiang Yat-Hung, So Chun-Kei Joinkey and Tang Bo-Sin’s “Time-varying performance of four Asia-Pacific REITs”; 3. Ulrich Schacht and Jens Wimschulte’s “German property investment vehicles and the introduction of G-REITs an analysis”; and 4. Michael Quek and Seow Eng Ong’s “Securitising China real estate: a tale of two China-centric REITs”. These papers provide a good example of the range of research opportunities REITs can provide as well as providing important insights. Baum and Devaney focus on a relatively narrow, but critical, aspect of the new UK REIT vehicle whether the depreciation of the properties held will adversely affect the operation of those REITs by constraining their cash flow. Chiang, So and Tang extend the literature on the characteristics of REIT returns in a multi-asset portfolio using data from Australia, Japan, Singapore and USA. Both Schacht and Wimschulte and Quek and Ong examine the details of national entities. Schacht and Wimschulte provide a detailed critique of the German REIT a vehicle for which, to date, take-up has been slow because of the restricted nature of the legislation. Quek and Ong analyse the relative merits, etc., of the Hong Kong and Singapore vehicles for investing in Chinese real estate an example of international vehicle arbitrage which is likely to become more prevalent. Yet it should be unsurprising that new REIT vehicles take some time to gain traction. Governments are naturally cautious at foregoing tax revenues (even if there may well be higher capture elsewhere in the system), so it is difficult to get the rules right first time. It must not be forgotten that it took over 30 years for the US REIT to start to fulfil its potential. But none of these papers directly examine arguably the key issue about REITs are they a substitute for direct property? The US literature (see also Chiang, So and Tang) has generally concluded that REITs there have similar characteristics to small cap stocks, but “with a twist”, which may be the real estate component. Research on this issue in the USA is severely hampered by having such a lagging direct property market index. While it is clear that REITs have a dynamic that is not always reflected in the direct market (e.g. the boom and correction of 1996 to 1999) the two markets have been much more highly correlated recently, and particularly in 2006 when large numbers of public REITs were taken private. Yet the only material difference between a public REIT and a private real estate fund is the way in

REITs - a global phenomenon

which the trading price is determined; for the latter it is based on the net asset value (NAV) of the vehicle, while for a public REIT it is fixed in the market. Both vehicles hold properties for investment, within a financial structure operated by dedicated management. But public REITs (and quoted property companies) often trade at significant discounts or premia to their underlying NAV, so generating more volatility. Does that price volatility merely reflect the much greater liquidity REITs offer and is that sufficient to create a different return profile? If the IPD UK monthly total return index (the best direct market index in the world) is compared with, for example, the FTSE 350 Property total return index over the 21-year period since the IPD index started in January 1987, the similarities are more striking than the differences. Real estate now has a truly global vehicle in the REIT. In many markets, the public and private property markets have been surprisingly separate. That state of affairs is sustainable no longer, particularly with the growth of investment in private real estate funds. Most of the world’s investors have a choice between buying highly liquid units from a screen at the press of a button or through the more complex structures of a private fund. That choice and the interactions of the public and private markets will provide plenty of opportunities for research as well as extraordinary investment opportunities in the future. President Eisenhower was responsible for the original legislation that created the US REIT in 1960, though that legislation has been significantly amended since, particularly with the creation of the “Up REIT” in the 1980s. Robin Goodchild Guest Editor

Reversionary freehold valuations by spreadsheet: introducing flexibility

Reversionary freehold valuations by spreadsheet: introducing flexibility Article Type: Education briefing From: Journal of Property Investment & Finance, Volume 26, Issue 3. Introduction In the previous article in this series (Vol. 25 No. 3), I looked at the construction of a flexible valuation spreadsheet model to value a rack-rented property by an implicit and explicit method. I will now amend that model to capture changes in the explicit method when the property is reversionary.

The example I will amend the example in the previous article. The existing information applies but instead of the property being rack rented, it is reversionary. Thus I need to add the extra reversionary information in addition to the existing information, i.e.: ● ● ● ● ●

market rent: £1m; all risk yield (equivalent yield): 8 per cent; target rate (equated yield): 10.75 per cent; rent review: five years; and calculated annual growth: 3.2 percent

Assuming that the property has a passing rent of £750,000 for the next three years, we now have the additional information of: ● ●

rent passing: £750,000; and term: three years.

We can therefore amend our INPUT/OUTPUT page as shown in Figure 1 (changes are shown in bold). The existing DCF model can be amended with a few simple changes: ● ● ●

the first period is for the term, not the rent review; the YP for the first period is for the term, not the rent review; and the trigger for the “perp” should reflect the point at which the number above equals the lease length minus the years gone. This is ll−(rr−t).

All these changes are shown in bold in Figures 2 and 3.

Reversionary freehold valuations by spreadsheet: introducing flexibility

Figure 1 Spreadsheet input/output page (showing formulae)

Figure 2 Spreadsheet DCF method (years columns as hidden formulae)

Figure 3 Flexible (years columns 3 to 7 as hidden formulae)

Reversionary freehold valuations by spreadsheet: introducing flexibility

Nick French