Investment Counseling for Private Clients V

Overview James H. Gilkeson, CFA Associate Professor of Finance University of Central Florida Orlando, Florida Managing private client assets is barely different from managing institutional assets—just add a few tax issues, throw in a bit of investor behavior talk, and you are set, right? If only that were true. In practice, investment professionals who work with high-networth individuals and families may feel as if they have developed split personalities—part securities specialist, part tax expert, and part therapist. As the presentations at the AIMR conference “Private Wealth Management: New Developments in Investing and Advising,” held in Atlanta on 17–18 March 2003, make clear, working with private clients does have many facets. But the skills a private client manager needs in addition to a thorough grounding in investment management are a general understanding and appreciation of the tangential issues of income and estate tax planning, business succession planning, tax-efficient investment strategies, and so on, not expert status in each of them. The presentations in this conference proceedings provide valuable insights into each of the three key facets of investing for high-net-worth clients: the characteristics of new investment opportunities, tax complexities, and client needs.

New Investment Opportunities Investment professionals need to move beyond a single-period framework (i.e., invest today and liquidate in 20 years) to a life-cycle approach, according to Zvi Bodie. Private clients acknowledge the cyclical pattern of earning, spending, and leisure, as well as a concern for wealth accumulation during their lifetime. This new approach requires more complex modeling of portfolio construction and of risk. It also implies a need for new, more sophisticated investment products and more cost-efficient channels of distribution. As an example, Bodie points out that two common life-cycle needs of investors are longterm-care insurance and life insurance. Bundling these together creates a product with risk characteristics that better meet client needs and are easier for issuers to manage. Hedge funds are increasingly being included in the portfolios of high-net-worth individuals. Jean Brunel explains that it is a mistake to conclude that ©2003, AIMR®

hedge funds are a separate asset class. Simply put, the hedge fund universe is not homogeneous. Instead, Brunel argues, hedge funds are an extension of active management that represents many different asset classes. The key to allocating them properly in a private client portfolio is to understand how they change the risk profile of the portfolio. The key issues are liquidity and manager risk. Hedge funds are much less liquid than traditional actively managed investments (mutual funds) but allow much greater manager flexibility.

Tax Complexities Whereas Brunel concentrates on the unique investment characteristics of hedge funds, Edward Dougherty explains their tax effects. He provides a primer on the differing tax concerns of individuals, corporations, and “tax-exempt” organizations and how a hedge fund’s structure affects an investor’s tax burden. Critical structural decisions include corporation versus partnership or limited liability partnership, onshore versus offshore, and master-feeder versus side-by-side. It is important that portfolio managers and advisors understand the tax effects of each of these decisions before placing private client assets with a hedge fund manager. Of course, taxes are important for all manner of private client portfolios, not just hedge funds. As Christopher Luck explains, there are a number of ways to create tax alpha by more efficiently managing portfolios. In a world of uncertain outcomes, however, determining which strategies will be most effective is more difficult. As a solution to this problem, Luck shows how simulations can be used to judge the relative effectiveness of various approaches to tax management. Christine Todd takes a fresh look at tax-exempt securities. In recent years, this asset class has provided surprisingly strong returns with little risk. There is, however, a concern lurking on the horizon for an increasing number of high-net-worth (and not so high-net-worth) investors—the alternative minimum tax (AMT). Although a relatively minor percentage of the tax-exempt securities in the market are subject to AMT, Todd explains that these securities www.aimrpubs.org • 1

Investment Counseling for Private Clients V often provide the highest yield but are often avoided because more and more taxable investors are nearing proximity to AMT liability. Todd also explains that the muni market’s volatility is increasing as nontraditional (taxable) investors are beginning to participate in the market through crossover strategies— investing in munis when they look cheap relative to taxable bonds and reversing the trade when the yield relationship reverts to a more normal spread.

Client Needs Aside from taxes, the primary difference between working with institutional investors and working with private clients is dealing with behavioral issues. When working with high-net-worth individuals and families, it is as crucial to understand each client’s distinct point of view as it is to understand the characteristics of the financial marketplace. As Marty Carter explains, an individual’s approach to investing is driven by his or her learned attitudes toward wealth, saving, and spending. Wealth can critically impact—often negatively— personal and family relationships. Although it may seem odd to suggest that an investment professional should be concerned if a large family or a couple cannot communicate effectively about wealthrelated issues, it will be next to impossible to work with them if these sensitive issues are not tackled. Carter recommends involving an expert facilitator to help open the doors of honest communication before proceeding with an investment plan that lacks client buy-in. Scott Welch points out the growing need for private client advisors who can bypass the conflicts

2 • www.aimrpubs.org

of interest, biases, and other problems frequently associated with traditional investment channels. He explains that private clients seek six traits from an advisor: objectivity, expertise, discipline, access, education, and service. Welch suggests that advisors can particularly demonstrate their value to clients through manager search and selection and by creating performance reports that are easily understood and address each client’s unique concerns. Thomas Giachetti also emphasizes clear communication with clients. He focuses on the regulatory examination process but makes it clear that the keys to avoiding client complaints (as well as surviving regulatory examinations) are creating and following specific procedures, making repeated and complete disclosures to clients, and maintaining good record keeping. Clients who are informed do not get surprised, and clients who do not get surprised are likely to remain clients.

Conclusion Understanding client parameters is only the beginning of the story, of course. As Zvi Bodie observes, “Investment decision making that is dynamically tailored to the needs of an individual is highly complicated.” Indeed, the considerable difficulty of integrating client needs and goals into a comprehensive, effective investment strategy is why clients hire managers and advisors in the first place. The presentations in this proceedings offer valuable insights for investment professionals who seek to put their wisdom and expertise to better use for their clients.

©2003, AIMR®

Applying Financial Engineering to Wealth Management Zvi Bodie Professor of Finance and Economics School of Management, Boston University Boston

The traditional investment paradigm, the Markowitz mean–variance model, focuses only on total wealth at the end of a single (long) period. The assumption is that time diversification will make equities “safe.” A more appropriate approach is the Samuelson–Merton life-cycle paradigm, a multiperiod model that stresses the need for hedging and insurance in addition to precautionary saving and diversification. A corollary of this new model is that the investment industry ought to provide affordable products tailored to suit the average investor’s means and needs.

rises give rise to opportunities,” as the old saying goes. Now is the time that the services of investment professionals are probably most needed by private clients. During the bull market, investors could do no wrong. Now, they desperately need professional help. During the past 50 years, investment management has become a science in the same way that medicine is a science. In both cases, some practices long predated the development of the underlying science. In medicine, although disagreements may exist about what treatments are appropriate for certain types of problems, most people can distinguish between a well-trained doctor and a quack. In investment management, the distinction is not always so clear. I have written extensively about the fallacy of the long run—the belief that holding extremely risky stocks for a long time makes them safe.1 This assertion has made me rather unpopular with practitioners because I am challenging an almost universally held belief on their part. Certainly, such a rift has existed between what academics think and what investment managers and investors believe for the more than 30 years since Samuelson and Merton wrote ground-

“C

1 Zvi Bodie, “On the Risks of Stocks in the Long Run,” Financial Analysts Journal (May/June 1995):18–22.

©2003, AIMR®

breaking articles on this subject.2 A number of the concepts described in this presentation have come from my collaboration with Robert Merton.3 The modern science of investment management (and finance in general) can help practitioners add a tremendous amount of value for their clients. Much value can come from creating new types of investment products and educating practitioners in their use and how to explain them to their clients. The time has come to move beyond the rather simplistic choices of which mutual fund to buy and of the right mix of stocks, bonds, and cash for a client’s portfolio to a discussion of investment products designed to meet specific goals of ordinary investors. The challenge is to use financial engineering to produce a new generation of user-friendly investment products that can be customized at reasonable cost. 2

See Robert C. Merton, “Lifetime Portfolio Selection by Dynamic Stochastic Programming: The Continuous Time Case,” Review of Economics and Statistics (August 1969):247–257 and Paul A. Samuelson, “Lifetime Portfolio Selection by Dynamic Stochastic Programming,” Review of Economics and Statistics (August 1969):239–246. 3 For further discussion of some of the topics in this presentation, see Zvi Bodie, “Thoughts on the Future: Life-Cycle Investing in Theory and Practice,” Financial Analysts Journal (January/February 2003):24–29 and Robert C. Merton, “Thoughts on the Future: Theory and Practice in Investment Management,” Financial Analysts Journal (January/February 2003):17–23.

www.aimrpubs.org • 3

Investment Counseling for Private Clients V

Shortcomings of Current Investment Products Virtually no one can understand today’s investment choices, whether the source of the information is an online advice engine or a brochure from a mutual fund company. Such concepts as mean–variance analysis or probability of loss/gain mean nothing to most people. I am often approached by fellow professors who are quite intelligent yet do not have a clue about what this information means or how they should use it. Investors are struggling with the decision of how to allocate among various investment options that currently exist, yet the mutual fund industry continues to add more products. If investors do not know how to allocate their funds among three choices, offering 30 choices will not help them. A good example of this is the explosion of investment choices offered by the Teachers Insurance and Annuity Association—College Retirement Equities Fund (TIAA-CREF). The investment management industry must rethink the choices they give to investors. Today’s offerings are like a restaurant that requires diners to choose from a menu of eggs, flour, butter, and milk, rather than a list of prepared desserts. Investors, like diners, want to choose from a list of completed products, not a list of ingredients. Such a menu should be tailored to the goals and needs of individuals and should enable them to make informed choices. Consider the way computers are sold online. The major manufacturers offer standard configurations tailored to home or business users. Using a menu-like interface, customers can click on items to learn what various features do and thus can easily customize the standard configurations to their specific needs and desires. The investment management business should use a similar approach. Such an approach could not have been followed 30 years ago, but because of advances in financial engineering, it can be accomplished today. Financial engineering is used extensively by banks, pension funds, and other financial institutions. It is used to manage currency, interest rates, and other risk exposures. To date, however, it has had little impact on investment management decisions. Today, financial engineering is commonly used in the world of private banking. Multimillionaires use custom-tailored derivatives to help them manage their risk. For example, they might desire a contract that puts a collar around a large exposure (such as an undiversified equity holding), but they have to pay dearly for this service. These kinds of private banking services are by far the most lucrative of a bank’s activities. But the costs are far beyond the means of 4 • www.aimrpubs.org

ordinary people who are saving for retirement, their child’s education, or other common goals. The investment management industry should strive to provide this type of financial engineering to average investors—but on a mass scale.

Insights of Modern Finance Five key insights of modern investment science should be considered in redesigning standard investment products. First, a person’s welfare depends not only on end-of-period wealth—what they will own in a year or 30 years or whatever their “end point” is— but also on consumption of goods and leisure over a lifetime. An investment portfolio should be tailored to the investor’s personal objectives. An individual who is concerned about a certain minimum standard of living in retirement is extremely risk averse, and the way to achieve his or her goal is not through diversification but through hedging. Such a person should invest all of his or her money in inflationindexed Treasury securities (commonly referred to as “TIPS”) because doing so will lock in a real rate of return that provides a real standard of living. The majority of personal investment advisors say that the most basic principle of investing is diversification. For example, as an experiment, I tried a large number of online investment advice engines. For each one, when answering the set of questions that helped determine my risk tolerance, I made it perfectly clear that I am extremely risk averse. Nevertheless, every engine recommended that I hold at least 30 percent of my money in equities. Why? Because I have a long time horizon. No matter how risk averse I made myself appear, these advisors still wanted me to put one-third or more of my wealth into the riskiest class of securities. Advisors still cling to the mistaken idea that over a long-enough time period, the magic of diversification will somehow do away with risk. Actually, a risk-averse investor with a long time horizon should hold long-term U.S. Treasury bonds that are inflation protected. One of the first discussions in basic finance courses on the term structure of interest rates is on determining the preferred habitat of the investor. An investor with a short horizon should focus on short-term bonds; a long-horizon investor should focus on long-term bonds. This principle is not reflected in any of the advice engines that I found on the Internet. Second, when thinking about investment risk and the creation and preservation of wealth, it is extremely important to consider what is likely to happen to real rates of return in the future. For example, consider a simple choice. Which portfolio would ©2003, AIMR®

Applying Financial Engineering to Wealth Management you prefer at retirement: $5 million in cash or $10 million in cash? The obvious answer is $10 million. But what if you could have $5 million with a real interest rate of 5 percent per year or $10 million with a real interest rate of 1 percent per year? A real interest rate of 1 percent a year on $10 million generates a perpetual annual income of $100,000 a year in real terms, whereas 5 percent on $5 million generates $250,000 per year. Most people care less about the size of their retirement portfolio than the standard of living that the portfolio can produce to sustain them in retirement. Most people would prefer the $5 million with the 5 percent real interest rate. Of course, time horizon (or life expectancy) affects the optimal choice. The greater wealth wins in the short term; the greater rate of return wins in the longer term.4 This approach to investment management differs from the standard approach, which measures risk in terms of ending wealth rather than interest rate (earning power) uncertainty. Virtually all of the quantitative tools currently used for asset allocation measure risk in terms of ending wealth; interest rate (earning power) uncertainty is ignored. Third, how should investors deal with multiple periods? The Markowitz mean–variance model developed in the 1950s, which is at the analytical core of every advice engine being used today, is a singleperiod model designed to optimize end-of-period wealth. In the Markowitz model, lengthening the time period always results in a higher allocation to equities (risky securities). In 1969, the Markowitz model was superseded in the academic literature by the Samuelson–Merton life-cycle model. The lifecycle model produced some interesting results that contradict the single-period Markowitz model. In particular, what happens as the length of the planning horizon gets longer and longer? Samuelson demonstrated that length of time horizon has no predictable effect on the optimal fraction of assets that should be invested in equities. The Markowitz model relies on time diversification to manage market risk; what is really needed is multiperiod hedging. The life-cycle model shows that diversification alone is insufficient to ensure the investor’s goals. Investments that limit downside risk must be held in the portfolio, regardless of the investor’s time horizon. Fourth, the value, riskiness, and flexibility of a person’s labor earnings are of first-order importance in determining optimal portfolio selection at each stage of the life cycle. The rule of thumb among financial planners and personal investment advisors 4

Robert Merton, “Thoughts on the Future: Theory and Practice in Investment Management,” Financial Analysts Journal (January/ February 2003):17–23.

©2003, AIMR®

is that the fraction of an individual’s portfolio that should be allocated to equities is 100 minus the person’s age. A 30-year-old investor should have a 70 percent allocation to equities, but a 70-year-old investor should have a 30 percent allocation to equities. Although simplistic, this rule may be appropriate for the average individual. However, consider those who work on Wall Street or are entrepreneurs starting new businesses. They already have a huge exposure to the market by virtue of the risk to their human capital. That is, if the market takes a dive, not just their wealth but also their earning power may be damaged. An analogous and more common situation is an individual who holds the stock of his or her employers (perhaps in retirement accounts). The additional exposure to the market is undesirable because the employee already has a large exposure to the risk of the company as a result of working for it. In fact, the company-specific exposure from employment alone is probably larger than is optimal, so perhaps the best strategy is to short the stock of the employer. I am not suggesting that if you work in the investment management industry and your future income depends on the stock market, you should immediately short the market. Buying some puts on the S&P 500 Index, however, might not be a bad idea. Generally, some people, especially those in their 20s or 30s, have exposure to the stock market as a whole or to the stocks in a particular industry. Depending on the occupation and the industry, that equity exposure may be great or small. But whatever the degree of exposure, it should be an important determinant of the composition of the individual’s investment portfolio. Why, out of thousands of choices, are there no mutual funds configured for people who work in certain industries? Fifth and finally, habit formation is a basic aspect of human nature and can give rise to a demand for guarantees against a decline in investment income. People want to maintain the standard of living to which they have become accustomed. A need exists for ratcheting or escalating products. These products rise in value when the market does well and are stable when the market drops. Life insurance companies sell some products of this nature, but they are incredibly expensive. The major problem is selling costs. If these products were mass produced, they could be offered at much lower margins. If professional investment advisors and financial planners respond positively to these products and recommend them to clients, it will help lower the huge marketing costs associated with selling them. Image problems must also be overcome. For example, variable annuities are a terrific idea in concept but have acquired a pejorative meaning www.aimrpubs.org • 5

Investment Counseling for Private Clients V because of their high fees. Although most variable annuity products are offered by life insurance companies, they are essentially mutual funds in various wrappers (with a few added features).

Comparing the Samuelson–Merton and Markowitz Models Exhibit 1 compares the two paradigms of life-cycle investing. The old paradigm is the Markowitz model, which currently dominates the world of personal investment management. The new paradigm is the Samuelson–Merton model. Welfare. What is the best measure of welfare? In the old paradigm, it is wealth at the end of the planning horizon. In the new paradigm, it is lifetime consumption (or some component of consumption, such as consumption during retirement). A key to the new paradigm is that welfare is measured by the flow of consumption over time, not the stock of wealth at a particular point in time. To understand the difference between the two concepts, remember my example of whether you would rather have $5 million earning 5 percent or $10 million earning 1 percent. Time Frame. With the old paradigm, a singleperiod model, the time horizon could be 1 year or 30 years. When a user of an online financial engine performs a Monte Carlo simulation for his or her chosen time horizon, no dynamic rebalancing is involved. In effect, an investor makes a single decision and runs with it. The structure of the model is such that the longer the time horizon, the greater the equity allocation. The results are quite different for a multiperiod model. In a multiperiod setting, investors make dynamic adjustments depending on unfolding circumstances. They may change their asset allocation as markets rise or fall or as their income rises or falls.

They may even change their long-term life plan. If the market does particularly well, they may choose to retire early. If it does particularly poorly, they may have to work longer than expected. Dynamic adjustments are also at play throughout the retirement period, because investors may alter the speed with which they draw down their wealth in response to changing market conditions and personal circumstances. Risk Management Techniques. The old paradigm rests on two risk management techniques: precautionary saving and diversification. The wealthy are able to build a large cash reserve to protect against the occurrence of a downside event (precautionary saving). For most, however, it is a challenge to save the small amount that they do, and then in an attempt to better the return on those savings, invest some of it, if not all, in the market. Fortunately, or so the story goes, diversification provides an acceptable approach to risk management for these investors—over a longenough time horizon, risky stocks become magically safe. In the old paradigm, diversification is the touchstone. The new paradigm also includes precautionary saving and diversification as risk management techniques but adds two more important, fundamental risk management techniques: hedging and insuring. Hedging involves a conscious decision to forgo some of the upside of a risky investment in order to eliminate the downside. For example, an investor might decide to lock in a return, maybe using TIPS. Currently, the 30-year TIPS is paying a real rate of interest of 2.4 percent. (A few years ago, investors could lock in a real rate of return of 4 percent.) The price paid for this protection against falling stock or corporate bond markets is forgoing the ability to participate in rising markets.

Exhibit 1. Comparison of Life-Cycle Paradigms Feature

Old Paradigm

New Paradigm

Measure of welfare

Wealth

Lifetime consumption

Time frame

Single period (stocks seem safe in the long run)

Many periods (stocks are risky in short and long run)

Risk management techniques

Precautionary saving Diversification

Precautionary saving Diversification Hedging Insuring

Retail financial products

Cash Insurance policies Mutual funds

Targeted savings accounts (e.g., tuition-linked certificates of deposit) Structured standard-of-living contracts

6 • www.aimrpubs.org

©2003, AIMR®

Applying Financial Engineering to Wealth Management Insuring means paying a premium today to eliminate the future downside—but with the ability to retain the benefit of the upside. For example, buying a put option on a stock portfolio is insuring against a price decline below the exercise price of the option. This insurance has a cost, but the investor retains the upside less the cost of the put. Most of the new products I envision, designed for ordinary people with relatively limited wealth, would incorporate some kind of hedging or insurance features to protect against downside risk. Of course, investors would have to pay for this protection, either with an insurance (put) premium or with less upside participation. Retail Financial Products. The retail financial products in the old paradigm are cash, traditional insurance policies, and mutual funds. In the Markowitz world, client goals have a limited impact on asset allocation recommendations. The goals are only incorporated in determining the time horizon; if the time horizon is long, the standard formulation (i.e., a large equity allocation) is considered optimal. As the horizon moves shorter, the allocation weights are changed to favor cash and bonds. This approach is demonstrably wrong. If the strategy is driven only by age and has nothing to do with expected market performance over the planning horizon, the result may be anything but optimal. I call the conventional model “faith-based investing” because it is based on the idea that the stock market will always rise. The stock market, however, can decline, and the conventional model does not use dynamic feedback to guide the investor’s decision making if, and when, the market does decline. For the new paradigm, I envision structured savings and investment accounts that are designed to lock in specific client goals over the client’s life cycle. TIAACREF (of which I am currently a member) offers variable annuity payouts that can increase or decrease over time. I like the increasing part, but I do not want to experience the decreasing part during my retirement! Because the annuity provides income that lasts for life, the purchaser of the annuity is insured against living too long, not against a decline in the market or a decline in interest rates. The same is true of most annuity products offered today. It is important that new standard-of-living contracts, or variable annuities, include provisions for escalation. Ultimately, the goal should be that the standard form of retirement annuity will be an annuity that can only increase. Consider an example of an escalating life annuity that provides a minimum standard of living (indexed for inflation) that increases if the stock market performs well. A 65-year-old investor with $1 million in savings puts $900,000 into an inflation-proof real annuity paying $55,000 a year and invests the other $100,000 in equities for potential growth in real income. Each year, she transfers some of the money ©2003, AIMR®

from the risky equity portfolio to buy an additional life annuity with an escalation. The upside leverage of this strategy could be increased by investing the $100,000 in a series of equity call options maturing in each of the next 10 years. If, on the annual expiration date, that year’s call is in the money, the proceeds would be used to increase the guaranteed income floor (the lifetime annuity). If the call is out of the money, the floor remains unchanged for another year. We can take this a step further and consider another product. People worry about two things in their old age: outliving their resources (assuming they are healthy) and needing assisted living or nursinghome care. Life annuities are available to take care of the first problem, and insurance products are available to provide for long-term care. But problems exist with these two products. On the one hand, many people do not qualify for long-term-care insurance because they are too sick or their health history puts them in a high-risk category. On the other hand, companies that sell life annuities have to protect themselves against adverse selection; that is, only healthy people will buy life annuities because they are the ones for whom it makes sense actuarially (so the price of lifetime annuities is too high for all but the healthiest). A bundled-risk product that provides a lifetime annuity and long-term-care insurance will help solve both problems. A healthier person will receive the lifetime annuity payment for a longer time but will not use the long-term-care protection for many years. A less-healthy person will need longterm care sooner but is unlikely to receive annuity payments for as long. From the insurance company’s point of view, the adverse selection problems have cancelled out, and the investor receives desired protection from both the risk of living too long and the risk of needing expensive long-term care.

Conclusion The old investment paradigm, the Markowitz mean– variance model, is a single-period model in which a longer investment time horizon invariably leads to a higher equity (risky security) allocation. The focus of the model is end-of-period wealth. It mistakenly assumes that time diversification will make equities safe enough for all investors, no matter how risk averse they are. The new paradigm, the Samuelson– Merton life-cycle model, is a multiperiod model that stresses the need for hedging and insurance in addition to precautionary saving and diversification. The new focus is on the investor’s income, consumption, and leisure patterns throughout life. For investment advisors and managers, a primary implication of this new paradigm is the need for affordable structured products that are tailored to fit the average investor’s goals and needs. www.aimrpubs.org • 7

Investment Counseling for Private Clients V

Question and Answer Session Zvi Bodie Question: What are the educational implications of this new paradigm for investment managers and clients? Bodie: We will not see radical changes in or rewriting of books intended for public consumption. In a way, the literature developed during the bull market is all wishful thinking. Glassman and Hassett, two reputable market commentators, wrote a book called Dow 36,000: The New Strategy for Profiting from the Coming Rise in the Stock Market, in which they said stocks are safer than inflationindexed Treasury securities.5 No one believes that premise any longer because it was predicated on the notion that stocks can never go down. The views I expressed in my presentation derive from mainstream academic thinking. The Investments textbook I co-authored with Kane and Marcus makes the same points in several places, and all of the top 30 Business Week 5 James K. Glassman and Kevin A. Hassett, Dow 36,000: The New Strategy for Profiting from the Coming Rise in the Stock Market (Three Rivers, MI: Three Rivers Press, 2000).

8 • www.aimrpubs.org

schools use this text.6 So, my views are not marginal in the academic community. Among practitioners, however, these ideas are considered quite controversial. Question: Could you comment on the lack of growth in the reverse mortgage market relative to your claim that new products are needed to meet life-cycle investing goals? Bodie: The problem with the reverse mortgage market is high cost. The hedging products that reduce volatility are quite transparent in terms of cost. In contrast, one of the nice features (from the seller’s viewpoint) of equity mutual funds is that fees are buried in the volatility, so the investor doesn’t notice a 200 bp fee in the midst of 10, 15, or 20 percent annual volatility. But when the investor is earning 2.4 percent or paying the interest rate on a reverse mortgage, the fee is noticeable. The fee is a fixed sum, so the big challenge is lowering the cost, which is an engineering problem (hence my use of the term “financial engineering”). 6 Zvi Bodie, Alex Kane, and Alan Marcus, Investments (New York: The McGraw-Hill Companies, 1999).

The same is true with escalating annuities. The challenge is to use new technology to make the escalation option more accessible and affordable. Professional financial planners and investment advisors have a huge role to play in addressing this challenge. Perhaps 10 percent of sales will be direct sales. The rest will be intermediated because most investors are uncomfortable buying these products on their own. The stakes are too high. They want validation from a trained professional. Thus, the practitioner community needs to educate itself. I don’t understand why ordinary people, unless they’re interested in the intricacies of investing, should have to become educated about it any more than I see why ordinary people should have to become educated about cars. One needs to know the basics of car maintenance, but one should not have to be knowledgeable enough to diagnose and repair car malfunctions. Car mechanics should do that for us. Investment decision making that is dynamically tailored to the needs of an individual is highly complicated.

©2003, AIMR®

A New Perspective on Hedge Funds and Hedge Fund Allocations Jean L.P. Brunel, CFA Managing Principal Brunel Associates, LLC Edina, Minnesota

The hedge fund universe is not homogeneous and should not be considered a coherent asset class. Rather, hedge fund management is an extension of traditional active management, with some critical differences. Thus, if investors seek to place portfolio constraints on hedge fund allocation, they should do so not to limit general exposure to an asset class but rather to achieve a certain level of exposure to manager and liquidity risk.

edge funds are not a homogeneous universe; therefore, characterizing hedge funds as a generic asset class is an error. Hedge funds are simply an extension of traditional forms of active management, with several notable differences. In this presentation, I will share an approach I have recently developed to help managers make decisions about optimal hedge fund allocations. By definition, any statistical analysis that is performed on historical manager results is subject to a number of caveats, particularly survivorship bias. The data used in this presentation are almost certainly affected by this problem. But to paraphrase Winston Churchill’s famous statement on democracy, this was the absolutely worst possible approach I could take—except for all the others.

H

The Problem with Traditional Thinking The main problem confronting portfolio managers when making asset allocation decisions about hedge funds is assigning them the right strategic weight. Because hedge funds are viewed as a different type of investment vehicle, constraints are often imposed on allocations to them. Some of these constraints are wise, but frequently, they are highly arbitrary. And whether the constraints are informed or not, the typical result is that managers do not focus on what makes a hedge fund truly different from other types of assets. Hedge funds are a highly heterogeneous universe, not a single asset class with homogeneous char-

©2003, AIMR®

acteristics. Panel A of Figure 1 plots (for the five-year period ending 31 December 2002) the risk and return for indexes representing all of the distinct hedge fund strategies covered by the Hedge Fund Research (HFR) database, composites (some of which are provided by HFR and some of which I calculated), the traditional asset classes used in strategic asset allocation (i.e., U.S. Treasury bills, Salomon Brothers Broad InvestmentGrade [BIG] Index, high-yield bonds, corporate bonds, S&P 500 Index, Europe/Australasia/Far East [EAFE] Index, and Russell 2000 Index), and an index of market-neutral strategies. Panel A illustrates why categorizing hedge funds as a separate asset class is a mistake; the hedge fund universe comprises a wide range of strategies and outcomes. As evidence that the wide range of outcomes shown in Panel A is not period specific, Panel B plots risk and return for the 10-year period ending 31 December 2002. Panel B also shows that hedge funds have provided a higher level of return than traditional asset classes for any level of risk. Nevertheless, the differences between various hedge fund strategies are so great that to think of them as a single aggregate category makes no sense. But if hedge funds are not a single asset class, then what are they? Figure 2 separates various parts of the hedge fund universe into clusters of commonality. Panel A shows that the so-called market-neutral strategies share risk and return outcomes that are much closer than those outcomes for the full hedge fund universe. Note that the range of risk and return levels in Panel A of Figure 2 is much smaller than in Panel A of www.aimrpubs.org • 9

Investment Counseling for Private Clients V Figure 1. Risk–Return Scattergrams of Hedge Funds and Major Indexes, December 1992–December 2002 A. Most Recent 5 Years Return (%) 20

15

10

HF Sector Total

BA MN Index

JPM EMBI+

Salomon BIG Composite MN Strategies 5

HF Fixed Total

U.S. T-Bills

HF Equity Unhedged HF EM Total

S&P 500

Merrill HY

0

Composite EM Equities and Bonds Russell 2000

MSCI EAFE

MSCI EM

5 0

5

10

15

20

25

30

35

Risk (%) Hedge Fund Strategy

Composite

Traditional Asset Class

B. 10 Years Return (%) 25

20

15

Composite MN Strategies

HF Sector Total HF Equity Unhedged

BA MN Index

HF EM Total

10

S&P 500

HF Fixed Total Salomon U.S. BIG T-Bills

5

JPM EMBI+ Russell 2000

Merrill HY MSCI EAFE Composite EM Equities and Bonds

MSCI EM

0 0

5

10

15

20

25

30

Risk (%) Hedge Fund Strategy

Composite

Traditional Asset Class

Note: BA MN = Brunel Associates market-neutral composite; EM = emerging markets; HF = hedge fund; HF EM = hedge fund emerging markets; JPM EMBI+ = J.P. Morgan Emerging Markets Bond Index Plus; Merrill HY = Merrill Lynch High-Yield Index; MN = market neutral; MSCI EM = MSCI Emerging Markets Index.

10 • www.aimrpubs.org

©2003, AIMR®

A New Perspective on Hedge Funds and Hedge Fund Allocations Figure 2. Risk–Return Scattergrams for Various Strategies, 10-Year Period Ending 31 December 2002 A. Market-Neutral Strategies

B. Fixed-Income Strategies

Return (%) 16

Return (%) 11 Event Driven

14 12

Distressed Convertible Arbitrage Merger Arbitrage

10

Equity Market Neutral

8 6

Convertibles

9

Total

8 Statistical Arbitrage Fixed-Income Arbitrage

Salomon BIG

7

FI High Yield

Arbitrage

6

Benchmark (60% U.S. T-Bill, 40% BIG)

4

Merrill Lynch High Yield

5

2 0

FI Mortgage Backed

10

Relative-Value Arbitrage

4 0

1

2

3

4 Risk (%)

5

6

7

8

2

0

4

C. Equity Strategies

16

12

14

16

D. Sector-Specific Strategies

Equity Hedge (Long/Short) Equity Unhedged

14

20

Technology

Healthcare and Biotech

12 S&P 500 Equity Market Neutral Russell 2000

Energy

Financial Sector

Sector Total Miscellaneous S&P 500

15

8

8 10 Risk (%)

Return (%) 25

Return (%) 18

10

6

Nasdaq 10

Nasdaq Russell 2000

Real Estate 5

6 Short Sellers

4 0

5

10

15 20 Risk (%)

25

30

35

0 5

0

10

15 20 Risk (%)

25

30

35

E. Emerging Market Strategies Return (%) 16 14

EM Global

JPM EMBI+

EM Latin America

EM Total

12 10 8

EM Asia

6

Benchmark (50% EM Debt, 50% EM Equity)

4

MSCI EM-Free

2 0 0

5

10

15 Risk (%)

20

25

30

Note: EM = emerging markets; FI = fixed income; JPM EMBI+ = J.P. Morgan Emerging Markets Bond Index Plus; MSCI EM-Free = MSCI Emerging Markets Free Index. Source: Based on data from HFR.

©2003, AIMR®

www.aimrpubs.org • 11

Investment Counseling for Private Clients V Figure 1. Fixed-income strategies, as shown in Panel B, are less closely related than market-neutral strategies but are still more closely related than the universe as a whole. This wider range is understandable, given that various fixed-income strategies differ greatly in their exposure to equity prices, interest rate volatility, and so on. Equity strategies, as shown in Panel C, are relatively well clustered but with a higher than expected level of dispersion. The “equity market-neutral” index, the market-neutral universe, appears in Panel A but clearly does not belong in the equity strategy world, despite its name. The higher level of dispersion evident in Panel C is likely the result of returns being above the long-term norm for the last 10-year period. And as would be expected in such an environment, short sellers had unusually poor performance. Sector-specific equity strategies, as shown in Panel D, share little commonality. In my view, these

strategies lack similarity in their risk–return profiles because no single tendency dominates their performance. Finally, Panel E shows emerging market strategies, which exhibit a relatively high level of commonality. Differences in this universe are most pronounced between region-specific strategies, such as Latin America and Asia. Now, consider these strategy clusters in a composite view, as shown in Figure 3. Most strategies fall within well-defined groups with common risk– return characteristics, although there are exceptions. As noted earlier, the sector-specific cluster, denoted by the dotted line, is extremely large—stretching from the lower left to upper right of the scattergram— and exhibits only a minimal degree of commonality. In contrast, the market-neutral and fixed-income clusters have relatively tight dispersions. The obvious conclusion is that hedge funds constitute not a homogeneous universe but rather a conglomeration

Figure 3. Risk–Return Scattergram Composite of All Strategies, 10-Year Period Ending 31 December 2002 Return (%) 25

20

15

Nasdaq

S&P 500

10

Russell 2000

5

0 0

5

10

15

20

25

30

35

Risk (%) Market Neutral

Equity

Sector Specific

Macro and Market Timing

Emerging Markets

Fixed Income

Source: Based on data from HFR.

12 • www.aimrpubs.org

©2003, AIMR®

A New Perspective on Hedge Funds and Hedge Fund Allocations of distinct strategies whose common characteristics lead to similar risk and return outcomes, with wide dispersion at times when the features of the strategies differ dramatically. Thus, the best approach in an asset allocation framework that includes hedge funds is to think of them as a set of distinct manager strategies rather than an asset class.

A Different Approach What makes hedge funds different from more traditional active strategies such as mutual funds? I propose that we think of all active strategies as having two components: the risk–return profile of the base asset class and the overlay of the active strategy of the manager. Hedge funds are simply an extension of the active management continuum. At one end of the continuum is the strategy of buying the index—a totally passive strategy. Managers move along the continuum away from a passive approach either to reduce risk versus the index or increase return versus the index. Traditional active managers and hedge fund managers take much the same approach with the fundamental asset—but with two notable differences. First, the structure of the investment, typically a limited partnership with an initial lock-up period, means that hedge funds are considerably less liquid than traditional investment structures. Second, both traditional and hedge fund managers can buy attractive securities, but hedge fund managers can also short unattractive securities. As they manage the balance between attractive and unattractive securities, they can also manage their net exposure to the market index. Some manage the exposure in a systematic, top-down manner; others take a bottom-up approach. For example, a market-neutral strategy can be viewed as a cash or short bond portfolio to which the manager adds security selection alpha. This analogy is attractive both intuitively and theoretically. The approach taken to eliminate market risk in a marketneutral strategy is effectively a hedge. Imagine a long General Motors/short Ford Motor Company pairs trade. The alpha the manager expects to capture is not the excess return that General Motors will earn relative to the equity market or the auto sector as a whole but simply its performance relative to Ford. Another market-neutral strategy is a convertible arbitrage trade, in which a manager trades the difference between the expected return on a company’s convertible bond and its stock. Yet, another market-neutral strategy is a capital structure arbitrage in which the manager might be long a company’s high-yield bond and short its underlying equity. ©2003, AIMR®

Regardless of the market in which the marketneutral strategy is pursued, the strategy encompasses the two components I outlined earlier: the risk–return profile of the base asset class and the risk–return profile of the active strategy. The manager uses the active strategy to generate alpha. In the case of an arbitrage strategy, the manager can generate alpha if she accurately judges that the spread between the expected returns of two assets varies from the norm, she locks in that spread, and the spread does, in fact, revert to a more normal relationship over a reasonable time period. For example, over the past five years, fixed-income arbitrage has proven to be an extremely poor strategy because fixed-income investors were unwilling to take risk during that time, preferring high-quality issuers. Thus, although fixed-income arbitrageurs saw abnormally wide spreads that were expected to revert to historical levels, spreads actually grew wider by the day. Recently, spreads have started to shrink, but the damage has been done. The point is that a manager’s alpha depends on the ability to find an attractive spread, trade on it, and be right. Consider a few other strategies in light of this perspective. In a hedged equity strategy, the portfolio manager has a long equity portfolio and a short equity portfolio, with some net exposure to the market if the market value of the two portfolios differs. The average net exposure for long–short equity managers is about 35 percent net long over time, suggesting some directional bias. The manager’s alpha (added value) is driven by his ability to correctly pick winners and losers as well as on the portfolio’s net exposure to the market. That net exposure may depend on the available opportunities (a bottom-up approach) or the manager’s view of the market (a topdown approach). Fixed-income arbitrage strategies also require the manager to pick winners and losers. In addition, however, they involve decisions on how much interest rate risk to take and how much exposure to other market features, such as convexity, or prepayment risk in a mortgage portfolio, is appropriate given anticipated market conditions. Once the base asset class that best represents a hedge fund manager’s strategy is identified, the next step in the forecasting process is twofold—(1) forecast return and risk for the asset class and (2) estimate expected manager alpha and tracking error, as shown in Table 1. (Note that this capital market forecast is not intended to be definitive. The point is to illustrate the process, not to argue the accuracy of the estimates.) The method for generating these forecasts should be consistent with those used for other, nonhedge fund strategies, and it is the same approach that should be used for a traditional large-cap equity fund, www.aimrpubs.org • 13

Investment Counseling for Private Clients V Table 1. Example Forecasts for Asset Class Return, Manager Alpha, and Manager Tracking Error Expected Return Strategy

Base Asset/Strategy

Total

Expected Risk

Manager Alpha

Base

Total

Base

Manager Tracking Error

Hedge fund strategy Various arbitrage

Short bonds

Event driven

Short bonds

9.5%

Fixed income

Bonds

Equity long–short

35% equity

U.S. equity

Equity

Emerging markets

Emerging market composite

3.5%

6.0%

4.5%

2.0%

4.0%

3.5

7.0

6.3

2.0

6.0

9.0

5.0

4.0

7.2

4.0

6.0

13.1

5.1

8.0

10.5

5.4

9.0

12.0

8.0

4.0

17.0

15.0

8.0

14.0

10.0

4.0

21.5

20.0

8.0

10.5

Equity sectors

Equity

13.0

8.0

5.0

19.2

15.0

12.0

Macro/market timing

Short bonds

11.5

3.5

8.0

10.6

3.5

10.0

Managed futures

Cash

5.0

1.0

4.0

8.1

1.0

8.0

Bonds

5.5

5.0

0.5

4.1

4.0

1.0

Traditional strategy U.S. fixed income U.S. large-cap equities Indexed

Equity

7.8

8.0

–0.3

15.0

15.0

0.5

Core

Equity

9.0

8.0

1.0

15.1

15.0

2.0

Active

Equity

11.0

8.0

3.0

16.2

15.0

6.0

incorporating forecasts of the equity market return and the particular manager’s alpha and tracking error.

employs only two people. I therefore chose a simpler approach that I will explain. The Morningstar database provides not only the statistics for each mutual fund but also its best-fit index (i.e., the index that best correlates with the manager’s strategy). With a best-fit index and an R2 between the performance of the manager and the best-fit index, we could infer the relative tracking error of the mutual fund; a higher R2 implies a lower tracking error.

Comparing Hedge Funds and Mutual Funds To further evaluate the difference between hedge fund and traditional managers, I performed an analysis of tracking error—the measure that managers have been taught is their risk—for equity and fixed-income mutual funds using the Morningstar database. The ideal data for this analysis would have been the monthly return stream of every mutual fund in that universe. Although the information is available, the amount of data (for about 1,700–2,000 funds) would be overwhelming for a firm as small as mine, which

E q u i t y F u n d s . For U.S. large-cap equity funds, we divided the universe into four broad categories—traditional index funds, active/core funds, concentrated portfolios, and mandates—as shown in Table 2. The first category, traditional index

Table 2. Index Correlation and Risk–Return Analysis for Traditional U.S. Large-Cap Equity Mutual Funds for Two Periods Ending 31 December 2002

Fund Type

Number of Funds

Correlation with S&P 500 R2

5 Years

10 Years

<

Return

Risk

199

0.99

1.00

–1.55%

19.34%

8.61%

16.84%

Subgroup 1

292

0.95

0.99

–1.78

17.05

7.12

15.43

Subgroup 2

437

0.90

0.95

–1.76

17.37

7.07

15.55

Subgroup 3

470

0.80

0.90

–1.93

18.53

7.23

15.88

Subgroup 4

382

0.65

0.80

0.02

16.91

8.32

15.27

Concentrated

102

0.50

0.65

1.61

17.05

10.16

16.09

Opportunistic

70

0.00

0.50

2.45

17.19

11.27

16.82

Traditional index

>=

R2

Return

Risk

Active/core

Source: Based on data from Morningstar.

14 • www.aimrpubs.org

©2003, AIMR®

A New Perspective on Hedge Funds and Hedge Fund Allocations funds, has an R2 of 0.99 or more. The next category, active/core funds (which is made up of four subgroups based on their observed R2), includes managers who seek to match the index or beat it by up to 2 percent, with a tracking error usually less than 4 percent. The third category, funds classified as “concentrated” (with an R2 between 0.65 and 0.5), only holds about 40–45 stocks, a relatively concentrated portfolio when the fund’s goal is to match or beat the S&P 500. Finally, the “mandates,” or opportunistic funds, category is a residual group, with an R2 of less than 0.5. Remember, all the funds selected for this analysis had a large-cap focus and had the S&P 500 as their best-fit index (according to Morningstar). The funds in this fourth category have a low correlation with the S&P 500 because they might be weighted 80 percent in technology one year and 80 percent in energy the next year. Their managers are the ultimate stock pickers. In Panel A of Figure 4, this universe of mutual funds is presented on a risk–return basis, with the three equity hedge fund strategies (hedged [long– short], unhedged, and market neutral) from Panel C of Figure 2 added for comparison. For the five-year period ending December 2002, the unhedged equity hedge fund strategy, the opportunistic mutual funds, and the concentrated mutual funds have similar risk– return profiles. The opportunistic and concentrated mutual fund strategies exhibited roughly the same level of risk as the unhedged equity hedge fund, with a positive, but much lower, return. The 10-year picture shown in Panel B of Figure 4 better illustrates the similarity in risk and return of the concentrated and opportunistic mutual funds to the unhedged equity hedge fund strategy. Over the period, all the active mutual fund managers closely tracked the equity index mutual fund in terms of risk. And four of the mutual fund strategies actually had

a lower return than the equity index mutual fund. In my view, this analysis shows that equity hedge funds, in exhibiting similar risk–return profiles to traditional active equity managers, are simply an extension of the active equity management universe, not a separate asset class. Fixed-Income Funds. In Table 3, the same categorizing analysis using R2 values is applied to the Morningstar fixed-income universe. Over the fiveyear period ending December 2002, as shown in Panel A of Figure 5, the best a fixed-income hedge fund manager could hope for was to produce the same return as a traditional manager but with substantially higher risk. Fixed-income arbitrage was clearly the worst strategy to follow during this period. Remember, I have defined a fixed-income hedge fund manager as one who looks for an attractive spread—for example, between two credit issues—that is expected to change in the future. Given this definition, the story depicted in Panel A makes sense because, over the past five years, the major feature in fixed-income markets has been the continual widening of credit spreads. Any manager who bet that spreads would narrow back to traditional levels would have performed poorly. The 10-year results shown in Panel B of Figure 5 reveal a different story. Fixed-income portfolios range from short to long durations so that the fixedincome indexes that are available in terms of “best fit” are quite limited. During this period, the “highest tracking error” fixed-income mutual funds were those with short- to intermediate-term portfolios. The active management contribution to the fixed-income manager’s performance was actually a bit misleading in that the manager’s relative performance was not greatly affected by security-selection skills. Instead, the duration of the portfolio made the greatest difference in performance.

Table 3. Index Correlation and Risk–Return Analysis for Traditional U.S. Fixed-Income Mutual Funds: Two Periods Ending 31 December 2002 Fund Type

Number of Funds

Search Criterion R2

5 Years Risk

10 Years

<

Return

6

0.99

1.00

7.16%

Subgroup 1

86

0.95

0.99

6.66

3.69

6.75

4.07

Subgroup 2

153

0.90

0.95

6.35

3.59

6.62

3.96

Subgroup 3

51

0.80

0.90

6.27

3.27

6.19

3.49

Subgroup 4

29

0.65

0.80

6.11

2.95

6.14

2.87

Concentrated

8

0.50

0.65

6.52

2.87

7.37

3.06

Opportunistic

4

0.00

0.50

5.73

2.27

5.84

2.62

Traditional index

>=

R2

3.58%

Return 7.29%

Risk 3.95%

Active/core

Source: Based on data from Morningstar.

©2003, AIMR®

www.aimrpubs.org • 15

Investment Counseling for Private Clients V Figure 4. Risk–Return Scattergrams for Various Equity Hedge Fund Strategies and Traditional Mutual Funds, December 1992– December 2002 A. Equity Hedge Fund Strategies (5 years) Return (%) 14 12

Equity Hedged

10 8

Equity Unhedged

Equity Market Neutral

6 4 2

Equity Concentrated

Equity Highest Tracking Error

0 Equity Active/Core

2

Equity Index Fund

4 0

5

10

15

20

25

Risk (%) B. Equity Hedge Fund Strategies and Traditional Mutual Funds (10 years) Return (%) 18 16

Equity Hedged

14 12

Equity Market Neutral

10 8

Salomon BIG

6

Equity Unhedged

S&P 500

Equity Highest Tracking Error Nasdaq Equity Concentrated Equity Index Fund

Equity Active/Core EM Debt/Equity Composite

4 2 0 0

5

10

15

20

25

30

35

Risk (%) Note: EM = emerging markets. Source: Based on data from Morningstar.

An important caveat, as I mentioned at the beginning of the presentation, is that all of my analysis is based on historical data. Thus, I have no idea whether my findings will hold in the future. The differences between the 5- and 10-year periods suggest that the patterns I have discussed are not always stable. Perhaps an even longer time period would provide more convincing evidence of the similarity in hedge fund and other active management strategies. 16 • www.aimrpubs.org

Another caveat is that Morningstar includes only those managers with a complete 10-year performance history. Managers who went out of business during the period are not included. A manager who takes on more business risk (in this case, more potential for tracking error) relative to his competitors is more likely to “guess wrong” and consequently go out of business. His poor performance will not be included in the analysis of the 10-year period. This ©2003, AIMR®

A New Perspective on Hedge Funds and Hedge Fund Allocations Figure 5. Risk–Return Scattergrams for Various Fixed-Income Hedge Fund Strategies, December 1992–December 2002 A. Recent 5 Years Return (%) 8 Fixed-Income Index

7 Fixed-Income Concentrated 6

Fixed-Income Highest Tracking Error

5

Fixed-Income Mortgage Backed

Fixed-Income Active/Core

FixedIncome Total

4 3 Fixed-Income Arbitrage

2 1 0 0

1

2

4

3

5

6

7

Risk (%) B. 10 Years Return (%) 12 Fixed-Income Mortgage Backed

10

Fixed-Income Total

Fixed-Income Index

8 Fixed-Income Concentrated 6

Fixed-Income Arbitrage

Fixed-Income Highest Tracking Error

Fixed-Income Active/Core

4 2 0 0

1

2

3

4

5

6

Risk (%) Source: Based on data from Morningstar.

means that my analysis of the 10-year period is likely biased in favor of higher-risk strategies. Because the sample includes all winners but not all losers, it show a higher expected return and a lower risk than would have been experienced by someone invested in the average manager.

Manager Risk and Liquidity If hedge funds truly represent nothing more than an extension of the active management spectrum for a ©2003, AIMR®

given asset class, what are the fundamental differences between hedge funds and other types of strategies? Essentially, the differences are the two that I mentioned earlier—manager risk (which comes in part from the ability to sell short and from a tendency to hold fewer, more concentrated positions) and lower liquidity. Manager Risk. The ability to sell short and, more generally, the freedom to pursue a wider variety of strategies mean that hedge fund and mutual fund managers take on different kinds of risk. Consider the www.aimrpubs.org • 17

Investment Counseling for Private Clients V return distribution statistics for 1990 through 2002, as shown in Table 4. (Note that this 12-year period is not selected to be convenient for my purposes; it is the longest monthly data series I could obtain for hedge funds.) Two differences are apparent between the hedge and nonhedge strategies. First, the return per unit of risk tends to be higher for hedge fund strategies than for the various indexes. Second, and more significant, hedge fund strategies exhibit substantially more kurtosis than the indexes (i.e., their return distributions have fat tails). Consider the return distributions of the lowervolatility hedge fund strategies shown in Figure 6 compared with the Salomon BIG Index. Using the returns for the 12-year period, I let the computer decide what buckets should be used to create a histogram based on the distribution of the Salomon BIG

Index. Then, I forced all the other indexes (hedge and nonhedge) into those buckets and graphed them together. I have two observations to make. First, I created an average of all market-neutral strategies because I wanted to prove a point—for a large enough number of managers and a large enough number of strategies, the high kurtosis may be reduced. (I will return to this thought later.) Second, the other two strategies graphed in the histogram— equity hedge and fixed-income average—exhibit huge tails, which means that accidents will happen more frequently than one might anticipate. Now consider my point about normality and the relationship between kurtosis and manager risk—the traditional way to measure manager risk is tracking error to the index. In fact, the more decisions a

Table 4. Selected Hedge Fund Return Distribution Statistics, 1990–2002 Market-Neutral Strategies Composite

Equity Hedge

Fixed-Income Average

Sector Total

Equity Nonhedge

Salomon BIG

S&P 500

Nasdaq

Average return

0.93%

1.44%

0.67%

1.57%

1.28%

0.67%

0.87%

1.11%

Standard deviation

0.92%

2.70%

1.18%

4.15%

4.29%

1.09%

4.41%

7.52%

Return per unit of risk

1.009

0.533

0.570

0.378

0.297

0.616

0.197

0.147

–1.522

0.147

–0.705

0.051

–0.490

–0.234

–0.458

–0.372

7.198

1.152

3.036

2.654

0.421

0.099

0.445

1.076

Measure

Skew Kurtosis

Source: Based on data from Morningstar.

Figure 6. Distribution of Hedge Fund Monthly Returns, 1990–2002 (lower-volatility strategies) Frequency 50 40 30 20 10 0

M or e

3. 43

2. 90

2. 36

1. 82

1. 28

0. 75

0. 21

0. 33

0. 87

1. 41

1. 94

2. 48

10

Return Basket (% range) Average Market-Neutral Strategies

Equity Hedge

Fixed-Income Average

Salomon BIG

Source: Based on data from HFR.

18 • www.aimrpubs.org

©2003, AIMR®

A New Perspective on Hedge Funds and Hedge Fund Allocations manager makes, the more likely it is that he or she will make both extremely good and extremely bad decisions. In concentrated portfolios, the manager makes fewer decisions but each decision is critical and thus that much more important. Therefore, in concentrated portfolios, extreme outcomes are more likely. It is also likely, however, that more concentrated traditional mutual funds would also exhibit a significant level of kurtosis. Liquidity. Hedge fund investments are considerably less liquid than traditional mutual funds. Hedge funds require investors to provide advance notice of intended withdrawal. Withdrawals are usually allowed only once per quarter and sometimes only once per year. Hedge funds also typically impose lock-up periods, meaning that investors must stay in the fund for a certain amount of time before they can withdraw their investment. In a traditional mean–variance framework, an asset class is defined by three sets of important data— return expectations, expected volatility, and expected covariances (or correlations) with other asset classes. In the same traditional framework, an investor’s preferences, defining a preferred risk–return trade-off and investment horizon, are used to select a point on the efficient frontier (the set of portfolios that provides maximum expected return for each level of risk) that specifies the appropriate portfolio allocation. In practice, the selection of a portfolio is subject to many other influences. For example, “agency risk”—the possibility that the manager may care more about job sustainability than the welfare of the client—is a concern. The investor may also impose constraints that reflect biases and preferences described by behavioral finance. Furthermore, investors are exposed to what I call decision risk, which is the risk of not unwinding a position at the point of maximum pay. More important, the traditional framework does not allow a real investigation of the exposure to overall manager risk and overall portfolio liquidity. This is true because the traditional framework does not incorporate a distinction between market and manager risk: All asset classes or strategies are described in terms of total-strategy expected return and risk. Thus, an investor cannot explicitly trade lower manager risk in one strategy for higher manager risk in another. Measuring Manager Risk and Liquidity. I have seen investors set an arbitrarily low allocation to hedge funds and yet place their entire equity allocation with a single highly concentrated mutual fund. Such a decision makes no sense because they are accepting an extremely high level of manager risk in ©2003, AIMR®

the traditional world while being unwilling to accept any manager risk in the hedge fund world. The only reasonable explanation would be that mutual funds trade daily and thus provide much greater liquidity than hedge funds. But I do not believe these investors are motivated by liquidity concerns. Rather, I believe the process is driven by an incomplete understanding of the world of hedge funds and thus by prejudice and unhealthy labeling. An alternative approach is to evaluate each strategy for its exposure to both expected market and manager risk. The idea is to evaluate the combined components of base asset class and manager strategy. To determine how much market risk is taken relative to manager risk, consider the expected risk columns in Table 1. Table 5 shows the squares of the two risk forecasts that provide a breakdown of the relative exposures. For two assets, the variance of a portfolio composed of those two assets is equal to the weighted-average sum of the variances of each of the assets plus an additional component that is calculated by multiplying the correlation coefficient between the two assets, their portfolio weights, and their respective standard deviations. I have made one critical assumption for the sake of simplicity, but it is not necessary for the method to produce correct results. I have assumed that no correlation exists between market and manager risk. If the extent to which a manager outperforms his or her market is independent of the market’s rise or fall, the correlation part of the equation is eliminated and the variance of the manager’s return is equal to the sum of the squared volatility of the asset class plus the square of the manager’s tracking error. The “various arbitrage” strategy provides a good example of this calculation. The strategy has a base (asset class) risk of 2 percent and a manager expected tracking error of 4 percent. The square of these is 0.0004 and 0.0016, respectively. Thus, I calculate that the strategy’s risk allocation is 20 percent market/80 percent manager. Table 5 shows that hedge funds tend to have more manager risk than market risk, whereas the reverse is true for traditional strategies. A similar approach can be used with respect to liquidity, as shown in Table 6. Gauging liquidity requires estimating the time needed to liquidate the portfolio. The most important information in Table 6 is the column labeled “Effective Liquidity Period,” which I define as the average number of days needed to find liquidity without causing a market impact. Hedge funds have a much longer effective liquidity period than traditional funds. www.aimrpubs.org • 19

Investment Counseling for Private Clients V Table 5. Estimated Market vs. Manager Risk Expected Risk

Strategy

Base Asset/Strategy

Total

Base

Strategy Risk Allocation

Manager Tracking Error

Base Risk

Tracking Error

Market

Manager

Hedge fund Various arbitrage

Short bonds

4.5%

2.0%

4.0%

0.0004

0.0016

20.0%

80.0%

Event driven

Short bonds

6.3

2.0

6.0

0.0004

0.0036

10.0

90.0

Fixed income

Bonds

7.2

4.0

6.0

0.0016

0.0036

30.8

69.2

Equity long–short

35% equity

10.5

5.4

9.0

0.0029

0.0081

26.5

73.5

U.S. equity

Equity

17.0

15.0

8.0

0.0225

0.0064

77.9

22.1

Emerging markets

Emerging markets composite

21.5

20.0

8.0

0.0400

0.0064

86.2

13.8

Equity sectors

Equity

19.2

15.0

12.0

0.0225

0.0144

61.0

39.0

Macro/market timing

Short bonds

10.6

3.5

10.0

0.0012

0.0100

10.9

89.1

Managed futures

Cash

8.1

1.0

8.0

0.0001

0.0064

1.5

98.5

Bonds

4.1

4.0

1.0

0.0016

0.0001

94.1

5.9

Traditional U.S. fixed income U.S. large-cap equities Indexed

Equity

15.0

15.0

0.5

0.0225

0.0000

99.9

0.1

Core

Equity

15.1

15.0

2.0

0.0225

0.0004

98.3

1.7

Active

Equity

16.2

15.0

6.0

0.0225

0.0036

86.2

13.8

Table 6. Estimated Hedge Fund Liquidity Fund Type

Required Notice Period

Liquidation Period

Effective Liquidity Period

Liquidity Factor 50.7%

Hedge fund Absolute return

Quarter

Annually

180

Semidirectional

Quarter

Annually

180

50.7

Directional

Quarter

Annually

180

50.7

Fund of funds

Quarter

Quarterly

135

63.0

Traditional Individual portfolios U.S. fixed income

None

1 Week

7

98.1

U.S. large-cap equity

None

1 Day

1

99.7

U.S. small-cap equities

None

1 Week

7

98.1

EAFE equities

None

1 Day

1

99.7

Emerging markets equity

7

98.1

None

1 Week

Commingled funds

Two weeks

None

15

95.9

Mutual funds

None

1 Day

1

100.0

Note: The “Liquidity Factor” is calculated by dividing the number of days required to liquidate the portfolio by the number of days in a year and can thus be seen as the fraction of any year during which liquidity is lost.

Optimizing. In addition to the tracking error, return, and covariance, I now have a vector that shows exposure to manager and market risk and an estimate of expected liquidity. With such information, an optimizer, whether a mean–variance or a more complicated mathematical form, can be used to calculate the whole portfolio’s exposure to manager and market risk as well as expected liquidity for any portfolio. Constraints can be set on maximum allowable manager risk and required minimum liquidity 20 • www.aimrpubs.org

(rather than setting an arbitrary limitation on hedge funds), requiring that optimal portfolios must have a manager risk exposure lower than the level established by the investor and higher liquidity than the minimum set by the investor. Hedge fund investment will still be limited by these constraints, but this decision-making approach correctly focuses on the investor’s goals and needs rather than simply being an arbitrary decision.

©2003, AIMR®

A New Perspective on Hedge Funds and Hedge Fund Allocations When a traditional optimizer is applied with the liquidity and manager risk constraints, it leads to a barbell allocation. As shown in Figure 7, the portfolio will seek liquidity from traditional sources and take on manager risk from other sources (including hedge funds) to provide the highest possible alpha. If the binding constraint is manager risk, most of the equity exposure will be in index or index-like strategies and most of the active portfolio management will be in hedge fund-type strategies. Sometimes, problems of liquidity arise, in which case the binding constraint that limits the hedge fund allocation will be required liquidity rather than maximum manager risk. Interestingly, this is the same structure that I encountered earlier in trying to construct tax-efficient portfolios using a traditional approach focused on after-tax return and risk—a barbell structure that promoted high tax efficiency at the expense of manager alpha and high alpha strategies at the expense of tax efficiency. Both approaches thus converged on the conclusion that tax-efficient and manager-risk/liquidity-

efficient portfolios are often best achieved through barbell structures.

Conclusion The hedge fund universe is broadly viewed as a homogeneous asset class for purposes of asset allocation. But hedge fund management is an extension of traditional active management with several important differences—less liquidity; the ability to short and to have higher levels of portfolio concentration (in terms of securities, sectors, or other risk characteristics), resulting in higher levels of manager risk; and manager-risk-induced kurtosis. Constraints on the allocation to hedge funds in the asset allocation process should be set not because hedge funds are a separate asset class, which, in my view they are not, but rather to achieve the desired portfolio exposure in terms of manager risk and the investor’s liquidity needs, together with the usual concern for efficiency in the trade-off between after-tax return and risk for the whole portfolio.

Figure 7. Optimal Asset Allocation with Manager Risk and Liquidity Constraints Frequency 60 50 40 30 20 10

M or e

9. 27

7. 12

4. 96

2. 81

0. 65

1. 51

3. 66

5. 82

7. 97

10 .1 3

12 .2 8

14 .4 4

0

Return Bucket (% range)

©2003, AIMR®

Equity Nonhedge

S&P 500

Sector Total

Nasdaq

www.aimrpubs.org • 21

Investment Counseling for Private Clients V

Question and Answer Session Jean L.P. Brunel, CFA Question: Given the problems with hedge fund reporting, how can private wealth managers have confidence that reported results are credible? Brunel: Take that question and substitute corporate financial reporting for hedge fund reporting. I don’t mean to be flippant, because this issue is truly important. My point is that reporting problems are not limited to hedge funds. I tend to serve families that are extremely wealthy. My average family has close to $1 billion and yet, as for most families, we are constrained to using only funds of funds. I know that it is extremely painful to pay 1 percent in addition to what the hedge fund managers are charging, but I see that 1 percent as a reasonable insurance policy. This approach, however, will not avoid all catastrophes. For example, Morgan Stanley has a team with a long history of investing in hedge funds. They came from the Weyerhaeuser pension fund, which until 1999, I think, was the best-performing pension fund in the United States for the previous 15 years and was also invested entirely in alternative

22 • www.aimrpubs.org

assets. Yet, this team managed to stumble twice over the last year. One position basically blew up and the other, whose problems have just been discovered, has not yet produced audited returns for 2001 (and we’re in 2003). Mistakes will occur, but by investing in funds of funds and doing a serious amount of due diligence, one can rely on the assumption that the reported returns make some sense. If you can understand what a hedge fund does and develop a good level of comfort with the fund-of-funds manager, I believe the risks are manageable. A large amount of discipline and quite a bit of homework go a long way toward providing confidence. In this case, as in many other situations, remember that if it looks too good to be true, it probably is. Be prepared to invest in, say, market-neutral funds that generate a long-term return of about 10–11 percent. If a fund says it has produced much more, ask yourself how it could continue to do so over the long term. Question: What situations do you see as red flags when you are considering a new manager?

Brunel: One well-known fund uses extremely high diversification, with more than 100 securities plus leverage. I do not understand how this fund has continued to generate a high double-digit return. Even last year, the fund was in the 15 percent range. I will not invest in this fund, despite the fact that I cannot point at anything it does wrong. I am simply not confident that the fund can continue to produce such results. If you stay away from the funds with extraordinary returns and focus more on the funds with sustainable longterm, low-double-digit returns, you will probably do well. Being cautious, however, can lead to its own problems. I recently made a big mistake in this regard. A mortgage arbitrage fund gave me all the access I wanted and allowed me to spend a half-day sitting at its desk so that I could see not only the whole portfolio but also all of the trade variations it uses. But I am an equity jockey by trade and because I simply could not understand the nature of these backward, forward, interest-only, and so forth transactions, I could not be convinced to pull the trigger. Last year, this fund had a 22 percent return.

©2003, AIMR®

Tax Implications of Hedge Fund Investments Edward H. Dougherty Senior Tax Manager Deloitte & Touche LLP New York City

Different investment vehicles have different implications for individual, corporate, and tax-exempt investors. Hedge funds in particular pose significant tax considerations. In order to achieve the benefits of tax planning for U.S.-based investors, a thorough understanding of the major differences in tax treatment of three common hedge fund structures (U.S. master-feeder, offshore master-feeder, and side-by-side) and private foreign investment companies is necessary.

edge fund investments have varying tax implications for their investors, depending on whether the investor is an individual, a corporation, or a tax-exempt entity, such as a pension fund. Each investor has a unique tax situation. In turn, the structure of the hedge fund itself has tax ramifications that can affect each type of investor differently. A hedge fund can be structured as a partnership, a limited liability company (LLC), or a foreign corporation. Accordingly, for prospective and current hedge fund investors, understanding the potential tax impact of an investment is imperative. My presentation begins with a brief review of the fundamental tax rules governing investors. I will then examine the after-tax performance reporting requirements for hedge funds and mutual funds and provide a comparison of the types of funds in terms of tax implications and other issues. This analysis will lead into a comparative discussion of the partnership versus corporation hedge fund structures and an explanation of the significant features of the three primary multi-entity hedge fund structures (U.S. master-feeder, offshore master-feeder, and side-byside). Finally, I will describe a variety of ways that hedge funds can engage in tax planning for their investors.

H

Tax Fundamentals of Investors In the United States, an individual’s tax rate varies with the type of income being taxed. Ordinary

©2003, AIMR®

income and short-term capital gains are currently taxed at 35 percent.1 Long-term capital gains (holding period greater than one year) are taxed at a preferential rate of 15 percent.2 Individuals cannot carry back capital losses to previous years but have an unlimited carryforward period. Throughout the presentation, when I talk about tax-advantaged income, I usually mean long-term capital gains, but other tax attributes are important to investors as well. For example, although U.S. government bond interest is not exempt from federal income tax, it usually confers a benefit for state and local income tax purposes. In the United States, a Subchapter C corporation pays tax at the marginal rate of 35 percent on all income and has a three-year capital loss carryback and a five-year capital loss carryforward. In contrast to the flow-through entities of a Subchapter S corporation, LLC, or partnership (whose individual items of income or loss, as well as deductions, flow through to the individual shareholder or partner for reporting and taxation and are not taxed at the entity level), a 1 Note that the Jobs and Growth Tax Relief Reconciliation Act of 2003 (the “2003 Tax Act”) was passed by the U.S. Congress and signed by President George W. Bush in May 2003. The Act reduced the top marginal tax rate for individuals to 35 percent; at the time this presentation was made, the top marginal rate was 38.6 percent. The reduction in the ordinary tax rate is retroactive to 1 January 2003 and is scheduled to “sunset” after 2010. 2 The 2003 Tax Act reduced the top tax rate applicable to long-term capital gains from 20 percent to 15 percent, effective for gains recognized after 6 May 2003. The reduction in the long-term capital gains rate is scheduled to sunset after 2008.

www.aimrpubs.org • 23

Investment Counseling for Private Clients V C corporation is taxed on its net income at the entity level and makes distributions of this net income to its owners in the form of a dividend, which is taxed again when reported on the shareholder’s tax return.3 A tax-exempt entity in the United States, such as a pension fund or an endowment, contrary to the implication of its name, is subject to tax on unrelated business taxable income (UBTI), which can arise from two sources. The first is income from a trade or a business unrelated to an entity’s tax-exempt purpose. The second is income generated through the use of leverage in a trading portfolio. Therefore, before a tax-exempt entity invests in a hedge fund, it should find out whether the fund expects income to be generated from either of these sources, and if so, the choice of investment vehicle needs to be reconsidered. As I will explain later, a tax-exempt entity generally would prefer to invest in a hedge fund structured as a non-U.S. corporation because the foreign corporation would serve as a “blocker” for UBTI.

The partnership structure gives fund managers (general partners) the ability to take performance fees as a profit allocation rather than as fee income. A hedge fund’s return contains tax-advantaged components, such as long-term capital gains and unrealized gains, which retain their character when they are passed through in a profit allocation. Structuring the performance fee as a profit allocation within a partnership also reduces the investors’ adjusted gross income (AGI). This can be an advantage because itemized deductions on the investors’ individual returns are limited at higher levels of AGI. If a hedge fund is structured as a corporation, the performance fees earned by the portfolio managers (along with fund expenses) are simply a deduction in determining the net income, or return, due the shareholders, or investors, in the fund. Such fee income is taxable to the fund manager as ordinary income taxed at the highest marginal rate. If the corporation is an offshore corporation, some opportunity exists to defer recognition of the income from performance fees for the portfolio managers. This practice is common in the hedge fund industry.

Fundamentals of Hedge Fund Structures

After-Tax Performance Reporting and Tax Planning

As I mentioned earlier, the income of a partnership or LLC is not taxed at the entity level; rather, the entity is treated as a conduit, and all of the income or loss it generates flows through to the owners (or investors, in the case of hedge funds) in the year of recognition. (Typically, LLCs are treated as partnerships for tax purposes.) The big draw for hedge fund managers is the performance fee they can generate for themselves. The partnership flow-through structure is particularly valuable in this regard. Typically, in a partnership fund structure, a hedge fund manager will charge two types of fees: a management fee (which is a percentage of assets under management) and an incentive or performance fee (which is based on any positive return the manager generates for the portfolio). Although performance fees are standard in the hedge fund arena, some portfolio managers are required to achieve a predetermined return, commonly referred to as a “hurdle rate,” before they can collect the performance fee.

In February 2002, the U.S. SEC began requiring mutual funds to report their performance on an aftertax basis for 1-, 5-, and 10-year periods. This after-tax performance must be calculated for the highest individual tax rate (currently 35 percent). At the present time, hedge funds are not similarly required to report performance on an after-tax basis. After-tax reporting, as well as tax planning, would be extremely cumbersome for a hedge fund because it would entail accounting for all the different stakeholders in the fund—namely, high-networth (taxable) individual investors, corporations that may be invested as general partners or limited partners, and tax-exempt investors. In the hedge fund world, the two most common ways of measuring the after-tax performance of an investor are the current tax liability divided by economic income and the taxable income divided by economic income. For individual investors, the first method makes more sense because it takes into account the preferential rate for long-term capital gains and other taxadvantaged income. Neither method takes into account the future tax liability (benefit) potentially arising from unrealized gains (losses) in the fund. A hedge fund can incorporate tax planning as part of its overall investment strategy, but to do so imposes two types of costs on the fund: transaction costs and opportunity costs. For example, if a hedge

3 The 2003 Tax Act also reduced the tax rate applicable to “Qualified Dividend Income” (QDI); prior to passage of the 2003 Tax Act, dividends were taxed at the same rate as ordinary income (i.e., 38.6 percent). Under the new law, dividends that meet certain tests will be considered QDI and will be taxed at the long-term capital gains rate of 15 percent. This provision of the law is scheduled to sunset after 2008.

24 • www.aimrpubs.org

©2003, AIMR®

Tax Implications of Hedge Fund Investments fund manager chooses to execute a basket, or total return, swap to avoid triggering capital gains in the portfolio, the investment bank with whom the swap is structured is going to want a piece of the action— a transaction cost for the hedge fund. And because this tax-planning activity, the basket swap, reduces the time that the manager would otherwise spend working to maximize the portfolio’s market-related performance, an opportunity cost is also incurred by the hedge fund. Do investors really want their portfolio managers distracted from managing the portfolio on an economic basis so that they can engage in tax planning instead? Whether hedge funds will continue to pursue more aggressive tax-planning opportunities is an open question. More fundamentally, should they engage in tax planning in the first place? In my view, the answer has to be yes. But given that an opportunity cost is involved, the best solution is for someone other than the portfolio manager to handle tax planning. This person can be either internal or external, perhaps a consultant, and should not be responsible for executing tax-related transactions. Rather this person should be charged with the task of determining what tax-related transactions are feasible and how they should be implemented.

Hedge Fund Structures Three common types of hedge fund structures exist— the U.S. master-feeder, the offshore master-feeder, and the side-by-side structure. The tax implications differ for each depending on the type of investor. I will briefly describe these differences and the reasons a portfolio manager would choose one structure versus another. U.S. Master-Feeder. The standard U.S. masterfeeder hedge fund structure is shown in Figure 1. Two feeder funds exist—a U.S. partnership or LLC that accepts investments from taxable U.S. investors and an offshore corporation or partnership (often domiciled in the Cayman Islands) that accepts investments from tax-exempt U.S. investors and non-U.S. investors. Both feeders make investments in the master fund, which is a U.S. partnership or LLC. All trading activity occurs at the master level. Income earned by the master fund is allocated to the two feeder funds. U.S. source dividends earned by nonU.S. investors in the offshore feeder are subject to a 30 percent U.S. withholding tax. Offshore Master-Feeder. The offshore masterfeeder structure, as shown in Figure 2, is similar to the U.S. master-feeder, except that the master fund is domiciled offshore, often in the Cayman Islands. As ©2003, AIMR®

Figure 1. U.S. Domestic Master-Feeder Hedge Fund Structure U.S. Investors

Non-U.S. Investors and U.S. Tax-Exempt Investors

U.S. Partnership (Feeder)

Cayman Corporation (Feeder)

U.S. Partnership (Master)

Figure 2. Offshore Master-Feeder Hedge Fund Structure U.S. Investors

Non-U.S. Investors and U.S. Tax-Exempt Investors

U.S. Partnership (Feeder)

Cayman Corporation (Feeder)

Cayman Partnership (Master)

with the U.S. structure, all trading takes place at the master level and income is allocated between the two feeder funds. In a typical offshore fund, the master fund is structured as a corporation under local law, but an election in the U.S. tax code allows investors to treat the structure as a partnership. This is called the “check-the-box election” and is made by most U.S. investors (for reasons that I will explain shortly). Offshore master-feeder structures are subject to the same withholding issues as U.S. structures, that is to say that U.S. source dividends earned by non-U.S. investors in the offshore feeder are subject to a 30 percent U.S. withholding tax. Side-by-Side (Mirror). The side-by-side fund, also called a “mirror” or “clone fund,” is shown in www.aimrpubs.org • 25

Investment Counseling for Private Clients V Figure 3. In this structure, the trading activity takes place in two different vehicles. Typically, the investment vehicle for U.S. taxable investors is a U.S. partnership or LLC, whereas the investment vehicle for tax-exempt U.S. investors and non-U.S. investors is an offshore corporation.

Comparison of Structures Multiple considerations, both tax and nontax, are involved in choosing the structure a hedge fund wishes to use, as shown in Exhibit 1. Tax planning and tax-sensitive investments are facilitated with the side-by-side structure. When the master-feeder structure is used, however, a conflict arises between the tax-planning needs of taxable U.S. investors and the lack of such a need on the part of both non-U.S. and tax-exempt U.S. investors. For example, under the new U.S. dividend tax rules, U.S. taxable investors would prefer their stock not be loaned so they can potentially earn qualified dividend income (QDI), whereas non-U.S. and tax-exempt investors, who do not qualify to earn QDI, would prefer their stock be loaned to generate additional income. So, the fundamental question is: Should a master fund engage in tax planning, even though such planning offers no benefit to non-U.S. and tax-exempt investors? In a side-by-side structure, targeted tax planning can be done in the domestic partnership or LLC without affecting non-U.S. and tax-exempt U.S. investors. All three structures provide incentive-fee flexibility to the fund manager. Incentive fees can be taken as a profit allocation from the master fund, which is a partnership. More commonly, however, the fees are taken from the feeder funds, where they are structured as a profit allocation from the domestic partnership and as a straight fee from the offshore corporation. The latter approach is preferred because the investment manager wants to retain the ability to

make elections that allow the manager to defer recognition of the income and thus the payment of the tax liability associated with the performance fees earned. All three structures provide for the ability to eliminate UBTI and passive foreign investment company (PFIC) issues for U.S. tax-exempt and taxable investors. As I stated earlier, tax-exempt investors typically invest in the offshore (Cayman) corporation of the offshore master-feeder because the corporate structure blocks the flow-through of tax attributes to the investors. Thus, if a hedge fund manager uses a leveraging strategy, the offshore corporation blocks the UBTI-tainted income that would otherwise flow to the tax-exempt investor. A similar advantage is provided to U.S. taxable investors who can invest in the U.S. partnership of the offshore master-feeder and thus avoid dealing with the complex PFIC rules. Two non-tax-related business issues arise with the side-by-side structure that generally do not occur with either of the master-feeder structures. The first issue is trade allocation. In a master-feeder structure, all of the trading activity takes place in the master fund, so no trade allocation issues are involved. The economic income is allocated to the feeder funds at the end of the accounting period based on initial investment levels, which is the job of the administrator or the fund accountants. But in a side-by-side structure, at the end of each day, the trades made by the investment manager must be allocated between the domestic fund and the offshore fund. The second issue is that side-by-side structures do not provide for economies of scale in terms of account aggregation.

PFICs Under U.S. tax law, a foreign corporation will be treated as a PFIC if either 75 percent of its income or 50 percent of its assets are deemed to be passive.

Figure 3. Side-by-Side (Mirror) Hedge Fund Structure

Investment Manager LLC U.S. Tax-Exempt Investors Non-U.S. Investors

U.S. Limited Partners

U.S. Domestic Limited Partnership (Investment Vehicle)

26 • www.aimrpubs.org

Cayman Corporation (Investment Vehicle)

©2003, AIMR®

Tax Implications of Hedge Fund Investments Exhibit 1. Tax and Nontax Considerations of Different Hedge Fund Structures Tax/Nontax Consideration Target tax efficiency

Sideby-Side

Non-U.S. MasterFeeder

U.S. MasterFeeder

Yes

No

No

Make tax-sensitive investments

Yes

No

No

Incentive-fee flexibility

Yes

Yes

Yes

Avoid UBTI and PFIC issues for investors

Yes

Yes

Yes

Avoid trade allocation issues

No

Yes

Yes

Account aggregation

No

Yes

Yes

Passive income includes interest, dividends, capital gains, rents, and royalties, and passive assets are those that generate such passive income. Most offshore hedge funds structured as corporations are PFICs. Prior to 1986, high-net-worth individuals could put their money in an offshore corporation and avoid paying tax on any income from the investment until they disposed of it. This changed in 1986, when PFIC rules were enacted to establish what is essentially an antideferral regime. These rules are designed to dissuade investors from deferring recognition of the income earned in a passive investment vehicle. The income earned by an investor in a PFIC can be taxed in one of three different ways. Qualified Electing Fund. The first (and most common) way that the income from a PFIC investment is taxed is under the qualified electing fund (QEF) regime. In this case, a U.S. investor makes an election on his or her income tax return to pay tax on a current basis on the ordinary income and net capital gains from the offshore corporation (as if the corporation were a partnership).4 In other words, the investor does not take advantage of any deferral opportunities that are provided because the foreign entity is structured as a corporation. This election can only be made, however, if the offshore fund issues to each U.S. investor an annual information statement that details the investor’s share of ordinary earnings and net capital gains generated by the fund during the year. The election is made at the U.S. investor level and cannot be made by the offshore corporation itself. Without such a statement, the QEF election is not permitted and taxes cannot be paid on a current basis. This statement must also include an agreement by the fund to allow the investor and the U.S. IRS to inspect its records so that the income can be verified. 4 Note that capital gains and losses do not flow through to the investor as they do in a partnership—retaining their individual character—but are instead netted.

©2003, AIMR®

Excess Distribution. The second way that income from a PFIC can be taxed is under the excess distribution regime. In this case, investors are taxed on a PFIC investment when they receive a distribution in the form of a dividend or when they receive cash from the redemption or sale of shares. Because an implicit deferral has occurred (the investor has not been declaring the income on a current basis), there is an interest charge in addition to the tax liability. The rules are extremely complex, but in short, they are intended to determine over the investor’s holding period when and how much tax the investor would have paid on a current basis. Because of these complex calculations, investors will generally make the QEF election if they can get a PFIC annual information statement from the fund. Mark to Market. The investor can also make a mark-to-market election with respect to his or her PFIC stock if the stock is traded on an organized exchange. Thus, an investor who is not issued an annual information statement can still be currently taxed on the PFIC investment. Unrealized gain is treated as ordinary income and unrealized loss is treated as an ordinary loss.

Mutual Funds vs. Hedge Funds Mutual funds and hedge funds vary in their tax implications for investors, as shown in Exhibit 2. The first column in Exhibit 2 represents mutual funds, the second and third columns represent trader hedge funds and investor hedge funds, respectively, and the fourth column represents PFICs. The difference between a trader hedge fund and an investor hedge fund is determined by the strategy of the fund—that is, whether the fund is focused on active trading or on making investments for long-term capital appreciation. Most tax issues are the same for both types of hedge funds. Current Taxation. A shareholder of a mutual fund generally pays tax on a current basis on his or her share of all the taxable income that is earned by the mutual fund. Similarly, an investor in a hedge fund structured as a partnership will pay tax on a current basis on his or her pro rata share of income as reported on his or her Schedule K-1. For a PFIC, current taxation depends on whether the PFIC investor makes a QEF election. If the election is made, tax is paid on a current basis. If no election is made, there is no current tax liability; however, when a dividend is paid or when an investor redeems shares, taxes plus an interest charge will be due. www.aimrpubs.org • 27

Investment Counseling for Private Clients V Exhibit 2. Comparison of Tax Attributes for Mutual Funds and Hedge Funds Tax Attribute Current taxation Mandatory cash distribution Flow-through of tax attributes Capital gains flow-through

Mutual Fund

Trader Hedge Fund

Investor Hedge Fund

PFIC

Yes

Yes

Yes

Depends

Yes

No

No

No

Modified

Yes

Yes

Modified Modified

Yes

Yes

Yes

Maybe

Yes

Yes

No

Yes

Yes

No

Yes

Incentive fee as profit allocation

No

Yes

Yes

No

Blocking of UBTI taint

Yes

No

No

Yes

Foreign tax credit flow-through Expenses reduce AGI

Partial sale taxable

Yes

No

No

Yes

Securities distributions for tax planning

No

Yes

Yes

No

Flow-through of losses

No

Yes

Yes

No

Mandatory Cash Distribution . The only one of these four investment vehicles that provides for mandatory cash distributions is a mutual fund. This distinction is important. Prospective hedge fund investors should be aware that, although they must pay tax each year on the allocated income from the hedge fund reported on their Schedule K-1, they will not necessarily receive a cash distribution with which to pay the associated tax liability. Some partnership agreements, however, do stipulate that the hedge fund will make a cash distribution just for this purpose. Other partnership agreements stipulate that to get cash out of the investment, an investor must request a redemption 45 or 90 days in advance. In many cases, the partnership agreement gives the general partner the discretion to deny the withdrawal request. Before investing in a hedge fund, investors need to satisfy themselves that they will have the liquidity necessary to pay taxes on the hedge fund investment. Flow-Through of Tax Attributes . The flowthrough of tax attributes is limited for mutual funds because it only applies to three types of income: ordinary earnings (including net short-term capital gains), net long-term capital gains, and now QDI. Under certain circumstances, U.S. government interest is also specified as a pass-through item. For a hedge fund structured as a partnership, all of the tax attributes flow through to the investor. Similar to the treatment of a mutual fund, a PFIC reports only limited income information—the investor’s share of ordinary earnings and net capital gains; other tax attributes do not pass through to the PFIC investor.5 5 It is not clear at this time whether QDI income will retain its character when earned by a PFIC. Under the 2003 Tax Act, dividends paid by a PFIC will not be treated as QDI for a U.S. individual.

28 • www.aimrpubs.org

Capital Gains Flow-Through. All of these four structures provide for the pass-through of long-term capital gains. To determine the amount of long-term gains that pass through to a PFIC investor, however, net short-term capital losses are netted against longterm capital gains rather than against ordinary income. Flow-Through of Losses. Losses generated by mutual funds and PFICs do not flow through to the investor except as a reduction to ordinary income (short-term losses) or capital gains (long-term losses). In contrast, losses generated by a partnership (trader or investor hedge fund) do flow through to the investor in the period in which they are generated and, in retaining their character as losses, are deductible on the investor’s individual income tax return, whether on page 1 of Form 1040 or on Schedule A (Schedule of Itemized Deductions) of Form 1040. Foreign Tax Credit Flow-Through . Foreign withholding taxes (that is, foreign from the standpoint of a U.S. investor) can only be passed through a mutual fund to investors if 50 percent or more of the assets in the mutual fund are international in nature. In contrast, all foreign withholding taxes paid by a partnership flow through to the partner (investor) to be claimed as a credit on the partner’s Form 1040. Foreign withholding taxes do not flow through to the investor in a PFIC, but are an adjustment to ordinary earnings—a deduction before the income is reported to the investor. This treatment does not provide a benefit to the taxpayer equal to that of a tax credit. Expenses Reduce AGI . Expenses in a mutual fund reduce the ordinary dividend that the fund pays to its investors, thus reducing the individual investor’s AGI. Similarly, with a trader hedge fund, all fund expenses reduce the investor’s AGI because they are treated as associated with the trade or ©2003, AIMR®

Tax Implications of Hedge Fund Investments business of the hedge fund and are reportable as a reduction to AGI on page 1 of the investor’s individual income tax return, Form 1040. But an investor in an investor hedge fund must report the expenses of the fund on Schedule A of Form 1040 rather than as a reduction to his or her AGI. Such expenses are treated as a miscellaneous itemized deduction, meaning that these expenses can be deducted for most high-networth investors only if they exceed 2 percent of AGI. So, for an investor hedge fund, the deduction of expenses is effectively lost because the investor’s share of the expenses is seldom large enough to overcome the 2 percent hurdle. Expenses do reduce the ordinary earnings that flow through to the investor of a PFIC, so they effectively result in a reduction in the investor's AGI. Incentive Fee as a Profit Allocation. An incentive fee is not typically taken by a mutual fund, but if it were, it would be treated as an expense of the fund without the favorable tax attributes associated with the profit allocation treatment of a partnership. As I stated earlier, the incentive fee associated with hedge funds is the main attraction from the point of view of the hedge fund manager. And the fact that incentive fees can be treated as a profit allocation in a trader hedge fund or an investor hedge fund, either of which can be structured as a partnership, makes these structures particularly appealing to managers. Likewise, any fees generated by a PFIC are considered to be like any other expense and cannot be treated as a profit allocation. As with the other expenses of a PFIC, the performance fee reduces the ordinary earnings that flow through to the investor of a PFIC, so they effectively result in a reduction in the investor’s AGI. Blocking UBTI Taint. If a mutual fund or a PFIC generates UBTI, a tax-exempt investor will not suffer any negative results because the corporate shell blocks the pass-through of UBTI. For a hedge fund with a partnership structure, however, the UBTI taint passes through to the investor so that tax-exempt investors would then be liable for tax on the tainted income. Partial Sales Taxable . Selling shares in a mutual fund, whether a partial or a complete position, is a taxable event. Usually, the investor has received dividend distributions and capital gains distributions over the investment period, which (if reinvested) increased his or her basis in the shares, so the capital gains realized at liquidation and thus the tax due at liquidation are not likely to be great. A partial redemption of a partnership interest in a hedge fund, however, is subject to fundamentally different ©2003, AIMR®

income recognition rules. Under U.S. tax law, a cash withdrawal from a partnership is not a taxable event unless the withdrawal exceeds the investor’s tax basis in the investment. If cash withdrawn exceeds basis, gain is recognized. In no event is a loss generated on a partial redemption. The partial or complete sale of a PFIC investment is a taxable event, much like it is for a mutual fund. Securities Distributions for Tax Planning. Technically, using securities distributions to fund withdrawals is allowed for mutual funds, but for a large mutual fund, distributing securities to satisfy withdrawals is not practical. The practice is common in hedge funds that are structured as partnerships, and it provides some tax-planning opportunities, notably as a tool to take highly appreciated securities out of the fund. This type of distribution is typically not done by PFICs. Other Business Issues. The other business differences between hedge funds and mutual funds are summarized in Exhibit 3. ■ Short selling and leverage. Hedge funds are free to engage in short selling, whereas mutual funds typically may do so only in a limited manner. Also, most mutual funds have only a limited ability to use leverage or margin borrowing, whereas hedge funds have a much greater ability to use leverage to generate additional returns for investors. ■ Incentive fee income/ability to defer. The mutual fund fee structure usually does not include an incentive or performance fee component—a standard in the hedge fund industry. The ability of managers under a partnership structure to take profit allocations for their performance fees allows hedge fund managers to share in tax-advantaged income, such as long-term capital gains, but always on a currentperiod basis. The only structure that provides a fund manager with the ability to defer recognition of fee income is the PFIC structure. ■ After-tax performance and reporting. Currently, a mutual fund is the only vehicle required to provide after-tax performance reporting to the SEC, although in the future, certain hedge funds may also be required to follow suit. ■ U.S. tax filing/privacy. No U.S. tax filing is required by a PFIC. A PFIC may provide an annual information statement to its investors (in order for investors to make the QEF election), but this statement is not required to be filed with the IRS. In contrast, mutual funds, as well as hedge funds structured as partnerships, are closely monitored by the IRS through required annual filings. The superior privacy offered by PFICs explains why most non-U.S. www.aimrpubs.org • 29

Investment Counseling for Private Clients V Exhibit 3. Comparison of Business Differences between Mutual Funds and Hedge Funds Mutual Fund

Trader Hedge Fund

Investor Hedge Fund

PFIC

Short selling allowed

Limited

Yes

Yes

Yes

Leverage/borrowing

Limited

Yes

Yes

Yes

Manager earns incentive fee

No

Yes

Yes

Yes

Business Practice

Manager earns management fee

Yes

Yes

Yes

Yes

Manager can defer incentive fee

No

No

No

Yes

Manager can take fee as profit allocation

No

Yes

Yes

No

After-tax reporting required by SEC

Yes

No

No

No

Privacy maintained

No

No

No

Yes

U.S. tax filings

Yes

Yes

Yes

No

investors prefer to invest through a PFIC or a corporate investment vehicle.

Tax Planning for Hedge Fund Investors Different types of investors have different tax considerations, as shown in Exhibit 4. High-net-worth individuals and corporations are interested in taxplanning arrangements that defer the recognition of gains or accelerate the recognition of losses. Highnet-worth individuals who invest in an investor hedge fund rather than a trader hedge fund are concerned about the deductibility of expenses. Taxexempt entities typically make investments through offshore vehicles and thus are interested in tax planning that minimizes withholding taxes on dividends (and UBTI, if they are not investing through a foreign corporate vehicle). Only high-net-worth individuals care about transactions that convert ordinary income and short-term capital gains into long-term capital gains, because they are the only investors who enjoy the commensurate benefit. Several common and relatively straightforward tax-planning techniques used by hedge funds

Exhibit 4. Tax-Planning Considerations for Different Hedge Fund Investors Tax-Planning Consideration

High-NetWorth Individual

Corporation

Tax-Exempt Entity

Gain deferral

Yes

Yes

No

Loss acceleration

Yes

Yes

No

Expense deductibility

Yes

No

No

Withholding taxes

No

No

Yes

Leverage/UBTI issues

No

No

Yes

Gain conversion (to long-term capital)

Yes

No

No

30 • www.aimrpubs.org

include tax-lot identification, use of broad-based stock index options, constructive-sale avoidance through total return swaps and long-term hedging, the doubling up of a position to avoid the wash-sale rules, forward conversion transactions, collars, and special allocations to partners (investors) withdrawing from the partnership. Tax-Lot Identification. The most basic form of tax planning is tax-lot identification. Generally, if a hedge fund sells a partial position and it has multiplecost lots of the stock, the first lot purchased is the one reported as the sale. Unfortunately, the first lot generally has the lowest cost basis and thus triggers the highest capital gain. At Deloitte & Touche, we advise our clients to minimize the gain or maximize the loss by designating the lot(s) with the highest cost basis as the lot(s) being sold. Broad-Based Stock Index Options. Broadbased stock index options, such as the S&P 500 Index contract, are referred to in the tax law as Section 1256 contracts and are taxed at a preferential rate. By statute, 60 percent of any gain from one of these contracts is treated as long-term, even if the contract was held for only a day, and the remaining 40 percent is treated as short-term regardless of the holding period. Therefore, an index fund interested in tax efficiency for the benefit of its investors may opt to use 1256 contracts rather than buying all of the securities in the index to take advantage of the preferential tax treatment. Total Return/Basket Swaps. A total return swap is typically not subject to the constructive-sale rules (which address gain acceleration) and the washsale and straddle rules (which address loss deferral). For example, when a stock is sold at the end of the year for the express purpose of harvesting losses as part of the fund’s tax planning and the fund ©2003, AIMR®

Tax Implications of Hedge Fund Investments purchases the same stock within 30 days of, either before or after, the date of sale, the wash-sale rule will cause the realization of the losses to be deferred until the next lot of the same stock is sold. The solution is a basket swap designed to provide a return that is expected to emulate the return of the stock that was sold as if the stock had been repurchased by the fund. For tax purposes, this arrangement is typically not treated as a wash sale; thus, deferral of the loss is not required. Doubling Up. Doubling up is another planning strategy that is typically not subject to the wash-sale rules. For example, suppose that a hedge fund is long 100 shares of General Motors (GM) with an unrealized loss in the position. The manager wants to realize the loss but also wants to continue to hold GM from an economic standpoint. The doubling-up technique would involve buying another 100 shares of GM and selling the original lot 31 or more days after the date of the second purchase to trigger the loss. Forward Conversion Transaction. If the exposure to the extra 100 shares of GM for the 31-day period is unacceptable, another approach that is not typically subject to wash-sale treatment is a forward conversion transaction. The investor buys a put and writes a call at the same strike price on the second lot of GM stock that is purchased. Economically, the result is the same as if the second block of stock was simultaneously bought long and sold short. The net exposure is still only 100 shares of GM. After 31 days, the original lot is sold (triggering the tax loss) and the option positions are liquidated. The change in the value of the stock, the call, and the put over the 31day period will essentially cancel each other out (ignoring transactions costs and time decay in the options). Collar. Another common tax-planning strategy is a collar transaction, which is used to hedge a significantly appreciated stock without triggering gain recognition. In this case, the investor writes a call and

©2003, AIMR®

buys a put. The income from writing the call pays for (or helps to pay for) the cost of the put. As long as there is a difference between the strike prices of the call and the put (we typically recommend at least a 20 percent difference), this transaction will not trigger gain recognition under the constructive-sale rules. Partnership Special Allocations. Prospective investors in hedge funds should carefully read the partnership agreements that govern the hedge fund investment they are contemplating in order to understand how the economic and tax allocations to the partners are handled. For example, many partnership agreements provide that partners who withdraw some or all of their capital get a special allocation of income to eliminate the difference between the book basis and the tax basis of their partnership interest. The effect is to give more income to the partner who is withdrawing and reduce income allocated to the continuing partners. Of course, disposing of the entire interest in a partnership is always a taxable event. The opportunity for tax planning is in determining the kind of income that is allocated to the withdrawing partner—long-term capital gains or a mix of income types. Investors need to focus on these provisions in the partnership agreement, particularly if they are not likely to be in the partnership for the long term.

Conclusion The tax implications of a particular investment vehicle—in this case, mutual funds, hedge funds, and PFICs—differ for individual, corporate, and taxexempt investors. As hedge fund investing increases in popularity with all types of investors, a thorough understanding of the major differences in the tax treatment of three common hedge fund structures (U.S. master-feeder, offshore master-feeder, and sideby-side) as well as PFICs is needed to achieve the benefits of tax planning under current U.S. tax law.

www.aimrpubs.org • 31

Investment Counseling for Private Clients V

Question and Answer Session Edward H. Dougherty Question: What would a capital loss carryback mean for a corporate investor? Dougherty: If a corporation experienced a capital loss in the current year, the corporation could use the loss to offset income in the prior three years and potentially get a tax refund. Of course, the capital loss carryback is available only for Subchapter C corporations. In contrast, a Subchapter S corporation is a pass-through entity; any losses, whether capital or operating, flow through to the owners of the S corporation (who can only be individuals) and are thus taxed on the shareholder’s individual return. The loss carryback is not applicable for a capital loss in the case of an S corporation because individuals have no carryback ability. Question: Is it possible to trigger the wash-sale rule twice? Dougherty: The wash-sale rule is a loss-deferral rule. In effect, the deferred loss adjusts the tax basis of the new security. When the new security is sold, the loss becomes deductible (unless you buy back

32 • www.aimrpubs.org

the security within the 30-day period). This question may refer to the desire to convert long-term losses into short-term losses. On an individual’s Form 1040 income tax return, long-term gains are netted against long-term losses and shortterms gains are netted against short-term losses and then net long-term and net short-term are netted together. If the ultimate result is a net long-term capital gain, then the preferential tax rate of 20 percent (15 percent for sales after 6 May 2003) applies. So, all things being equal, individuals would prefer to have long-term gains and short-term losses. We do use a technique for converting long-term losses to shortterm losses. It works as follows: You sell a long-term position (with a holding period longer than one year) for a loss. You buy a call option on the same security, which triggers the wash-sale rule. The loss is disallowed, thus adjusting the basis of the option. Then, you exercise the option and take delivery of the underlying stock. According to the way the tax rules are written, the loss is embedded in

the stock position, which has a new holding period. You can dispose of the stock the next day and generate a short-term loss. Question: Does the Cayman Islands impose a withholding tax on assets domiciled there? Dougherty: No. The Cayman Islands imposes no withholding taxes that are applicable to hedge funds, which is one of the reasons why so many hedge funds are domiciled in the Cayman Islands. Question: In light of new mutual funds billing themselves as hedge funds, what investment restrictions do registered funds have that private funds do not? Dougherty: The most significant difference is that most registered funds are limited in their ability to engage in short selling and thus cannot use that method to manage risk. Mutual funds might, in certain circumstances, be able to use derivatives to achieve the effect of short sales, but rarely do they engage in outright short selling because of investment limitations; this limits their risk management opportunities.

©2003, AIMR®

Capturing Tax Alpha in the Long Run Christopher G. Luck, CFA Partner First Quadrant, LP Pasadena, California

For taxable investors, tax alpha (the value added from tax management) is far more valuable than pretax alpha. The benefits of tax management strategies, such as loss harvesting and HIFO (highest in, first out) accounting, can be quantified by using Monte Carlo simulations that are based on various assumptions about tax rates, return, volatility, and so on. Regardless of the market environment, a portfolio managed in a taxefficient manner should consistently outperform a buy-and-hold portfolio.

onte Carlo simulations are an excellent tool for analyzing strategies designed to maximize risk-adjusted after-tax return. Although this presentation focuses on equity portfolios, the principles are applicable to any asset class. In essence, our approach at First Quadrant is simple. Instead of using historical data to evaluate a strategy, we make assumptions about various parameters—return, volatility, tax rates, dividends, and so forth—and apply these assumptions in Monte Carlo simulations to determine the strategy’s after-tax performance under a wide range of possible outcomes.

M

Tax Alpha For taxable clients, the goal is to maximize riskadjusted after-tax return. Unfortunately, the focus (both for mutual funds and separate accounts) is typically on pretax returns. For active managers, the real challenge is balancing uncertain stock alpha versus relatively certain tax alpha. Tax alpha is the value added from tax management. In other words, it is the difference between a manager’s pretax added value and after-tax added value. There are three value-adding components to any process focused on taxable management. First, managers can use the traditional method to generate pretax excess return through superior stock selection. Second, they can use risk management to minimize

Editor’s note: This presentation uses the tax rates and assumptions of a U.S. investor.

©2003, AIMR®

tracking error and extreme outcomes. Third, they can use tax management to minimize or eliminate taxes. Tax management itself has three main components: managing taxes on dividends, deferring capital gains taxes, and harvesting losses. Currently, dividend management is a relatively unimportant source of tax alpha because dividends tend to be low in both the U.S. and non-U.S. equity markets. Because the greatest source of negative tax alpha tends to be realization of capital gains, gain deferral is a very important source of tax alpha. The opposite of realizing capital gains is realizing capital losses, loss harvesting, which is also a significant source of tax alpha. At First Quadrant, we use Monte Carlo simulations to evaluate the benefits of all three sources of tax alpha. Pretax alpha is difficult to achieve. Even an extremely skilled manager has only a relatively modest advantage over the market. Generating tax alpha, however, is a far easier task. The reason is that dividend rates and unrealized capital gains and losses are known with almost complete precision. Thus, estimating the tax hit or the tax advantage of a particular trade is straightforward. This observation is not intended to be an argument against active management. At First Quadrant, we believe in active management. But we also recognize that adding value through stock selection is far more difficult than adding value through tax management. Furthermore, we understand that once pretax excess return is achieved, the challenge becomes keeping it in the portfolio. Tax management is the key to meeting this challenge.

www.aimrpubs.org • 33

Investment Counseling for Private Clients V

Managing Taxable Assets There are two complementary approaches to tax management: the accounting approach and the investing approach. The accounting approach—the highest in, first out (HIFO) method—involves evaluating the tax impact of every trade by looking at the applicable short- or long-term capital gains tax rate and picking the cost lot that offers the best tax outcome. This is an accounting decision: Given multiple blocks of the same stock with different costs purchased at different times, which one should be identified as having been sold? Thus, this approach does not alter the underlying portfolio management. The second approach involves delaying the sale of a stock solely to avoid recognizing a capital gain (gain deferral) or selling a stock that has dropped in price solely to use the loss to negate a realized capital gain (loss harvesting). In short, this approach means managing the portfolio (making trading decisions) so as to minimize the tax burden. At first glance, using the accounting approach seems to be an obvious choice. After all, any manager who works with taxable clients should choose accounting conventions that maximize tax savings, particularly when this choice does not alter any portfolio management decisions. Applying the HIFO method is essentially Pareto optimal; that is, it almost always produces a tax cost that is less than or equal to that of any other method. In the case of the investing approach, altering investment decisions to minimize taxes is a fundamental change for many managers. Deferring gains and harvesting losses clearly provides a tax benefit to the client, but it also introduces risk in the form of tracking error. First, selling stock for the sole purpose of harvesting losses incurs transaction costs, which reduce pretax portfolio returns. Tracking error is typically measured by comparing the pretax returns of the portfolio and its benchmark, so a reduction in the portfolio return becomes a component of tracking error. Second, from an active management point of view, loss harvesting may lead to stock-selection shortfall over the 31-day wash-sale period (when the manager has to be out of the stock in order to be able to recognize the loss). The advantage, of course, is that it can maximize client after-tax return. To maximize both tax savings and after-tax return, the loss harvesting/gain deferral approach requires a continuous, disciplined, systematic process. To achieve maximum tax savings, a manager should dispose of all stocks for which the tax benefit of the loss exceeds the related transaction costs. This discipline should be used continuously, not only at year-end or quarter-end. Furthermore, it should take into account the 34 • www.aimrpubs.org

total portfolio “utility,” including risk (and stock selection, if appropriate).

Monte Carlo Simulations The Monte Carlo simulations discussed in this presentation are also presented in an article co-authored by several investment professionals at First Quadrant.1 In a Monte Carlo simulation, assumptions about the mean and standard deviation of a variety of inputs are made. For each simulation, the computer randomly generates values for each input based on the assumptions. The outcome of the simulation (in this case, the pre- and after-tax returns) is determined based on the randomly generated inputs, our fixed assumptions about the world (such as tax rates and dividend yields), and the particular strategy being followed. For our study, we ran a large number of simulations, and we examined the range of outcomes to determine how a particular strategy performs relative to other strategies. For the base case in our simulations, we assumed an average annual benchmark return of 8 percent with a standard deviation of 15 percent for the equity market, consistent with historical norms. Part of that total return comes from a fixed annual dividend yield of 1.44 percent (i.e., 0.12 percent a month, which is close to the actual average dividend). We assumed a tax rate of 35 percent; in other words, the assumption was for a corporate entity, not an individual who faces both long- and short-term capital gains tax rates. We created 300 months of results for each simulation. The investment universe consisted of the 500 stocks in the benchmark. The return for each stock was assumed to be the benchmark return plus a normally distributed random variable with a mean of zero and a standard deviation of 9 percent, which again is consistent with the long-term mean and standard deviation of individual stocks in the S&P 500 Index. This means that each stock had the same expected return as the benchmark but a larger standard deviation. We assumed that each month, one new company was added to the benchmark and one was removed. Transaction costs were assumed to be zero, and no wash-sale rule was in effect. The portfolio was rebalanced monthly, and the tax alpha and dividends were reinvested in the portfolio quarterly. No cash contributions or withdrawals were allowed after the initial cash contribution. We ran each Monte Carlo simulation 500 times in an effort to show the range of likely outcomes. 1 Andrew L. Berkin and Jia Ye, “Tax Management, Loss Harvesting, and HIFO Accounting,” Financial Analysts Journal (July/August 2003).

©2003, AIMR®

Capturing Tax Alpha in the Long Run We used the base-case simulations to evaluate three tax management strategies: • Strategy 1. Buy-and-hold portfolio with averagecost accounting. • Strategy 2. Buy-and-hold portfolio with HIFO accounting. • Strategy 3. Tax-managed portfolio with HIFO accounting and loss harvesting. All three strategies resulted in a portfolio that fully duplicated the benchmark at the end of each month. The difference between Strategies 1 and 2 is purely an accounting convention. The difference between Strategies 2 and 3 is that in Strategy 3, when a stock had to be sold (because it had left the benchmark) and the sale resulted in a capital gain, a search was made for other stocks that currently had unrealized losses. If available, these other stocks were sold, the loss was realized, and the stock was immediately replaced at its current market price. (Recall that we assumed zero transaction costs and no wash-sale rule.) We measured the value of the portfolio each month in both gross and net terms. The gross value was the market value of the shares held. The net value was the gross value less any capital gains taxes that would be paid if the portfolio was liquidated immediately. Interpretation of the portfolio values is easy. In short, subtracting Strategy 1’s value from Strategy 2’s value shows the benefit of HIFO accounting. Subtracting Strategy 2’s value from Strategy 3’s value shows the benefit of loss harvesting. The 500 Monte Carlo simulations using the parameters I have described produced the following results: The median cumulative (over 300 months) tax benefit from HIFO accounting is 8.6 percent before liquidation taxes and a little more than 4 percent after liquidation taxes. The median cumulative tax benefit from loss harvesting is 115 percent before liquidation taxes and 58 percent after liquidation taxes. Although the use of HIFO accounting is essentially Pareto optimal, it confers only a slight advantage over using a non-tax-sensitive accounting method. If the management approach is changed in favor of gain deferral/loss harvesting, the tax benefit is much greater. Of course, these data are for the median of the 500 simulations. Panel A of Figure 1 shows the cumulative benefit over time for the 25th, 50th (median), and 75th percentile simulation for the HIFO accounting strategy, and Panel B shows the same information for the tax-managed (lossharvesting) strategy. In all cases, the tax-managed strategy provides a much larger benefit than the tax-accounting strategy. The benefit of harvesting losses and not blindly fol©2003, AIMR®

Figure 1. Cumulative Alpha for HIFO and LossHarvesting Strategies A. HIFO Cumulative Alpha (%) 16 14 12 10 8 6 4 2 0 0

50

100

150

200

250

300

Months 25th Percentile

50th Percentile 75th Percentile

B. Loss Harvesting Cumulative Alpha (%) 180 160 140 120 100 80 60 40 20 0 0

50

100

150

200

250

300

Months 25th Percentile

50th Percentile 75th Percentile

lowing the index is apparent. Remember that this simulation is based on a monthly turnover rate of 2.5 percent, which is consistent with historical turnover in the S&P 500. The reason the curves rise over time is compounding; the tax benefit is invested back into the portfolio. Figure 2 shows that most of the benefit of loss harvesting occurs early on. The reason is that we assumed the benchmark (and each individual stock) increased in market value at an average rate of 8 percent a year. Initially, our assumption that individual stocks had an independent standard deviation of 9 percent over and above the standard deviation of the benchmark meant that quite a few stocks (out of the 500) would experience a few bad months in a row or even a bad year or two. Such losses can be harvested. After a few years, however, these initial losses will have been harvested and (because our assumption was a rising market) most other stocks will have risen above their initial values. The portfolio will contain fewer and fewer unrealized losses. www.aimrpubs.org • 35

Investment Counseling for Private Clients V Figure 2. Alpha of the Loss-Harvesting Strategy before Liquidation Taxes (annualized)

Figure 3. Turnover of the Loss-Harvesting Strategy (annualized)

Annualized Alpha (%)

Annualized Turnover (%)

10

350 300

8

250

6

200

4

150

2

100 50

0 0

50

100

150

200

250

300

Months 25th Percentile

50th Percentile 75th Percentile

Because loss-harvesting activity will diminish over time, its marginal added value will diminish. After three years, its benefit is less than 2 percent a year, and after five years, its benefit is less than 1 percent. Interestingly, after 25 years, the benefit is still positive, although small. A manager will still be harvesting losses after 25 years, even though the benchmark is growing at 8 percent a year, for two reasons. First, the dividends that flow into the portfolio every year will be reinvested, creating blocks of each stock that will have a generally higher cost basis. Even if, on average, the cost basis of a particular stock is less than its current market value, individual lots purchased with dividends may be underwater. Second, the benchmark has turnover. When turnover occurs, the proceeds from the required sale are reinvested in the new stock at its current market price. Some of these new purchases will generate losses that can be harvested. Just as the tax benefit is greatest in the beginning, turnover is also highest in the early years—more than 100 percent in the first year, as shown in Figure 3. Clearly, little difference exists in turnover between the 25th, 50th, and 75th percentile of the Monte Carlo simulation distribution.

Stress Testing After running the simulations with the base case, we decided to examine how our results would change when individual assumptions changed. In other words, we wanted to know what would be the impact of changing market conditions on after-tax liquidation values. First, we increased stock-specific volatility. One could argue that the 9 percent stock-specific volatility 36 • www.aimrpubs.org

0 12

60

108

156

204

252

300

Months 25th Percentile

50th Percentile 75th Percentile

in our base case is not representative of recent experience. We chose 9 percent because it is the average for the S&P 500 over the past 20 years. When we increased individual stock volatility to 10, 12, or 14 percent, after-tax liquidation values improved. The reason is that significant loss harvesting continued for a longer time. Increasing volatility resulted in a wider distribution of stock returns and, correspondingly, stocks owned at a larger amount of unrealized loss—and thus more opportunities for tax mitigation. Because greater stock-specific volatility leads to better after-tax performance, tax management is more valuable in such an environment. Tax management matters less during periods of low stock-specific volatility. A second test was to assume that the range of outcomes at the individual stock level would remain unchanged but the benchmark itself would become more volatile. In other words, cross-sectional stock volatility held steady but total market volatility increased. This assumption had virtually no effect on after-tax performance. For purposes of tax management, what matters is how volatile stocks are relative to each other, not market volatility per se. As I explained earlier, the base case assumed an 8 percent annual market return (1.44 percent from dividends and 6.56 percent from capital gains). When we increased the market return by increasing the capital gains component, keeping the dividend rate the same, the opportunities for loss harvesting diminished and the value of capital gains deferral increased. Nonetheless, the compounding of lossharvesting benefits at a higher rate of return over 25 years led to superior performance. An important point to note here is that the investment time horizon ©2003, AIMR®

Capturing Tax Alpha in the Long Run matters for tax-advantaged strategies; the longer the horizon, the greater the value of tax management. Testing the sensitivity of tax alpha to the level of market return produced some of our more interesting results, as shown in Figure 4, which indicates the tax alpha (after liquidation taxes) for an average market return of 11, 8, and 5 percent. (Note that the data are for the 50th percentile, or median, of the 500 simulations run for each market return assumption.) In the first 10 years, little difference in cumulative alpha is apparent. At the 10-year point, however, the compounding effect took over and the benefit of a higher market return becomes clear.

Figure 4. Impact of Market Return: Median Cumulative Alpha after Liquidation Taxes Cumulative Alpha (%) 140 120 100 80 60 40 20 0 0

50

100

150

200

250

300

Months 11% Average Market Return 8% Average Market Return 5% Average Market Return Note: The data are for the 50th percentile, or median, of the 500 simulations run for each market return assumption.

We tried both increasing and decreasing the benchmark turnover rate and observed relatively little impact on after-tax returns, although higher benchmark turnover, not surprisingly, provided a slight benefit. The advantage is slight because higher turnover leads to more gain realization, with the offset that replacing high-gain stocks with new stocks allows for more ability to harvest losses. Increasing the dividend return while holding the total benchmark return constant led to slightly higher after-tax performance. The dividend return generated a little more cash, which was invested in the portfolio at a current-cost basis year after year, providing more opportunities for loss harvesting. The advantage was slight because the change in assumption was slight—from 1.4 percent to 1.8 percent. A larger increase would have led to a greater after-tax advantage. Allowing cash contributions over time had a slightly positive effect on after-tax performance. ©2003, AIMR®

Receiving (or increasing) cash contributions is equivalent to earning a higher dividend. Larger cash inflows provide greater scope for loss harvesting. But because additional cash also went into the passive benchmark, the impact was extremely slight. Similarly, allowing cash withdrawals over time also had only a slight negative impact, leading to lower aftertax performance. Using HIFO accounting to meet withdrawals, however, always led to improved performance relative to using average-cost accounting. Of course, increasing the amount of withdrawals would make an increasingly more significant difference because doing so would reduce the effective time horizon, thus leading to a greater similarity between the actively managed tax-aware portfolio and the buy-and-hold strategy. Finally, adding (or increasing) both contributions and withdrawals (i.e., a net neutral effect) led to slightly higher after-tax performance. Interestingly, the difference was greater than was the case for altering either contributions or withdrawals alone. Rather than canceling out each other, these effects appear to be additive. Increasing the tax rate led to significantly higher after-tax performance compared with the simple buy-and-hold strategy; the higher the tax rate, the greater the advantage for tax management. Interestingly, the improvement was most pronounced in moving from a low to a medium tax rate, as shown in Figure 5. The benefit of tax management when moving from the 20 percent to the 35 percent tax rate is clearly greater than when moving from the 35 percent to the 50 percent tax rate. We tested only these three tax rates, but we believe that they adequately describe the range of investor exposure to tax liability. A number of corporate institutions are subject to a 20 percent rate. The 35 percent rate applies to other

Figure 5. Impact of Tax Rate: Median Cumulative Added Value before Liquidation Taxes Cumulative Value Added (%) 180 160 140 120 100 80 60 40 20 0 0

50

100

150

200

250

300

Months 35% Tax Rate

50% Tax Rate 20% Tax Rate

www.aimrpubs.org • 37

Investment Counseling for Private Clients V corporate entities and is fairly descriptive of federal individual rates. The 50 percent rate is relevant when high state and local rates are added to federal rates.

Conclusion The past three years in the equity markets have not produced an environment conducive to superior tax alpha. But when equity market returns are positive, a tax-managed portfolio should consistently outperform a buy-and-hold portfolio. At First Quadrant, we have published several articles finding that taxoblivious active management would have to deliver tremendous alpha to overcome the after-tax hurdle presented by the simple buy-and-hold approach.2 2 Robert D. Arnott, Andrew L. Berkin, and Jia Ye, “How Well Have Taxable Investors Been Served in the 1980s and 1990s?” Journal of Portfolio Management, vol. 26, no. 4 (Summer 2000):84–94; Robert D. Arnott and Robert H. Jeffrey, “Is Your Alpha Big Enough to Cover Its Taxes?” Journal of Portfolio Management, vol. 19, no. 3 (Spring 1993):15–26.

38 • www.aimrpubs.org

Tax management can also add a lot of value relative to a simple buy-and-hold strategy but is often overlooked. Taxes matter greatly, especially over long time horizons. Taxes amount to a heavy transaction cost— whether a 35 percent short-term capital gains rate, a 35 percent corporate rate, or a 15 percent long-term capital gains or qualifying dividend rate. Unfortunately, most managers focus solely on beating the benchmark on a pretax basis, which is an uncertain game to play because consistently generating this sort of alpha is very difficult to do. Taxes are perhaps the one aspect of investment management that can be known with virtual certainty. With such information, anyone who uses proper discipline can dramatically reduce the impact of taxes. For taxable clients, the focus should be on after-tax return. Whether active or passive management is applied, tax management can provide a tremendous benefit.

©2003, AIMR®

Capturing Tax Alpha in the Long Run

Question and Answer Session Christopher G. Luck, CFA Question: Your base-case scenario assumes a positive market return. Are there benefits from loss harvesting in a negative market return environment? Luck: In the past three years, few equity managers have had difficulty avoiding capital gains. But even when the market is down, loss harvesting makes sense. The benefit of loss harvesting, however, depends critically on the ability to balance a loss against a realized gain. The constraint is the client who insists that she doesn’t need any more losses because her portfolio has no gains to protect. But if the client invests in hedge funds, which seem to generate a lot of short-term gains, she may well benefit from consistent loss harvesting, even in down markets. Overall, although a negative market affords more opportunities for loss harvesting, the compounded benefit over a long-term horizon is not particularly high if the market is declining year after year. Question: Is it better to harvest tax losses throughout the year rather than only at year-end? Luck: Harvesting only at yearend is better than doing no harvesting at all, but if clients can actually use losses, the opportunity set is much larger if loss harvesting is done throughout the year. For example, think back to 1999, to a market in which stock prices were rising rapidly. A tax-oblivious manager with a lot of capital gains to offset might have decided to harvest losses at year-end but then found few positions with losses that could be harvested. In contrast, a manager who harvested losses throughout the year would have been better able to offset gains with realized losses. When the market is declining, year-end versus constant harvest©2003, AIMR®

ing does not make much difference. When the market is rising, it can make a dramatic difference. Question: Is the major risk of loss harvesting having to be out of the security for 31 days, and have you attempted to quantify this risk in your simulations? Also, do you use a market index to reduce tracking error? Luck: Switching from a different process to active tax management clearly introduces risks, particularly higher transaction costs and greater tracking error. Recognizing a loss means being out of the stock for 31 days before you can buy it back, which tends to introduce some tracking error for the portfolio, especially for highly volatile stocks. When we test our portfolios, we look at the tracking-error impact of loss harvesting and try to juggle the risk. For example, imagine that you are holding Microsoft Corporation at benchmark weight and it is at a loss. Harvesting the entire loss will provide a tax benefit, but it will also introduce 50 bps of tracking error. We probably would start slowly reducing the Microsoft position over time because, although we do not want to add 50 bps of tracking error, we are willing to add 10 bps of tracking error in return for picking up some of the tax benefit. You also might buy calls and puts on Microsoft to hedge your position during the wash-sale period. We tend not to buy index options because it means buying the index. Of course, you could buy Standard & Poor’s Depositary Receipts or derivatives. Such methods, however, do not offer a good hedge against idiosyncratic risk when selling Microsoft or some other stock.

Question: At what level of turnover in a portfolio does tax management stop adding alpha, and if it always adds alpha, at what level are the expenses greater than the value added? Luck: Obviously, as portfolio turnover increases, so do transaction costs. In the base case I described, we make an unrealistic simplifying assumption—zero transaction costs. So, in that sense, the benefit I’ve reported is overstated. In actuality, managers have to figure out the costs of a trade. For us, it is a mathematical question. We set aside the stock-selection alpha, although we believe in it, and focus on the tax benefit. For example, if I sell a stock at a loss, calculating the tax impact is easy. The tax benefit may then be compared with the incremental risk and transaction costs imposed by selling the stock and buying it back (including the fact that I have to buy another stock for the portfolio to remain fully invested). For us, if the tax benefit exceeds twice the transaction cost, plus the risk impact times a penalty function (representing my dislike for a higher tracking error), we will do the trade—assuming that the client can use the tax benefit. Whether turnover is 80 percent or 10 percent does not matter because we look at the tax benefit of each individual trade. In a rising market, however, turnover levels for a tax-managed portfolio are low because half of the battle is deferring gains. For example, in the late 1990s, our turnover was about 20–25 percent a year. In the past few years, our turnover has been almost 100 percent a year because the market has been down and opportunities to harvest losses have been numerous. www.aimrpubs.org • 39

Strategic and Tactical Allocation of Municipal Bonds in Private Client Portfolios Christine L. Todd, CFA Senior Vice President Standish Mellon Asset Management Boston

In the current market, municipal bonds offer attractive after-tax yields combined with high credit quality and low price volatility, but investors’ heavy use of munis in shortterm tactical allocations (such as crossover trades) has added volatility to the market in recent years. Given the probability of yield-curve flattening and a return to a more normal short-term/long-term yield spread, investors should consider a barbell strategy in which the portfolio is divided approximately equally between long-term and short-term maturities. Investors also should consider the implications of the alternative minimum tax, which will affect an increasing percentage of taxpayers if the tax code is not repealed or modified.

few years ago, municipal bonds, federally taxexempt bonds issued by states and municipalities in the United States, were underappreciated by investors and managers alike. In the past few years, however, their stature has appreciated along with their market value. Municipal bonds have performed quite well on an absolute and relative basis, especially considering their high credit quality, low volatility, excellent after-tax equivalent yield, and tremendous diversification benefits. In addition to the strategic importance of municipal bonds for taxable investors, this presentation will concentrate on certain tactical maneuvers that can be used within the municipal bond (“muni”) market as well as between munis and other fixed-income markets. Finally, I will examine the significance of the alternative minimum tax (AMT) for the tax-exempt bond market.

A

Advantages of Municipal Bonds The tax-exempt bond sector is known for its high credit quality, as shown in Figure 1. Even in today’s environment, with so many headlines about states under fiscal pressure because of a weak economy, this market remains strong, especially when compared with the early 1990s. A comparison of default rates for municipal versus corporate bonds is telling. Table 1 compares the default rates of municipal and corporate bonds 15

40 • www.aimrpubs.org

years after issue. The municipal market simply has a higher standard of quality. The default rate in AA and AAA municipal bonds is zero; furthermore, the recovery rate in the event of a skipped payment is 100 percent. The same cannot be said of corporate bonds. In fact, BBB municipal bonds have a default rate essentially equal to AAA corporate bonds. Munis have about half the volatility of U.S. Treasury bonds, as shown in Figure 2. Stock market investors who see Treasuries as a flight-to-quality vehicle may want to consider munis instead. Of course, munis have lower volatility in part because they are not as liquid as Treasuries. An investor who believes that interest rates might rise should consider buying munis rather than Treasuries because munis lag Treasuries not only in a rally but also in a sell-off. So, munis are a more defensive high-quality, fixedincome vehicle than Treasuries. The after-tax yield advantage of munis versus taxable bonds is quite attractive in today’s environment, as shown in Table 2. (Note that this is calculated for the 38.6 percent marginal tax bracket. The after-tax yield for an AMT filer is shown in Table 3.) In only one case are munis less attractive than other fixedincome vehicles—the two-year maturity for A rated corporates. The difference of 17 bps, however, is not sufficient to compensate a fixed-income investor for the significantly lower quality in the corporate market (see Table 1).

©2003, AIMR®

Strategic and Tactical Allocation of Municipal Bonds in Private Client Portfolios Figure 1.

Upgrades vs. Downgrades for Tax-Exempt Bonds, 1991–2002

Number 1,000 900 800 700 600 500 400 300 200 100 0 91

92

93

94

95 Upgrades

96

97

98

99

00

01

02

Downgrades

Source: Based on data from Standard & Poor’s.

Table 1. Standard & Poor’s Cumulative Default Rates for Municipal vs. Corporate Bonds, 1986–2000 Rating

Municipal

AAA

0.00%

Corporate 0.51%

AA

0.00

1.07

A

0.16

1.83

BBB

0.52

4.48

BB

1.34

16.36

Note: Rolling 15-year periods, 1981–2000. Source: Based on data from Standard & Poor’s.

Table 2. Municipal After-Tax Yield Advantage over Taxable Bonds, as of 20 February 2003 (non-AMT taxpayer; bps) Bond Sector

1 Year

2 Year

5 Year

10 Year

Treasury

34

34

78

135

Agency

30

30

60

105

Asset-backed

14

13

45

92

Corporate A

21

–17

17

61

Note: After-tax yield advantage calculated for the 38.6 percent marginal tax bracket. Source: Based on data from Municipal Market Data and J.P. Morgan Chase & Company.

©2003, AIMR®

Munis are cheap not only in today’s market but also on a historical basis. As of January 2003, at certain longer maturity points along the yield curve, specifically the 30-year maturity, the yield of munis even exceeds the yield of comparable-maturity Treasuries. Panel A of Figure 3 shows the yield on 10-year AAA munis as a percentage of the yield on 10-year Treasuries. The higher the ratio, the closer the muni yield is to the Treasury yield (i.e., the higher the ratio, the cheaper munis are). Compared with the 1993–97 period, munis have become extraordinarily cheap versus the Treasury market; in fact, farther out the yield curve, past the 10-year point, munis are even cheaper than shown in Panel A. The reverse situation has occurred in the corporate market, however, as shown in Panel B of Figure 3. Corporate yields have increased considerably relative to muni yields, as corporates have shown themselves to be of lower quality than munis. As a result, munis are less attractive (not as cheap) on a pure yield measure in comparison with the corporate market than in comparison with the Treasury market. Panel A of Figure 3 shows that, as of 31 January 2003, the yield of 10-year AAA rated munis was 96 percent of the 10-year Treasury yield. This means that any investor with a marginal tax rate of 4 percent or more should own munis rather than Treasuries www.aimrpubs.org • 41

Investment Counseling for Private Clients V Figure 2.

Volatility for the S&P 500, 30-Year Municipal Bonds, and 30-Year Treasury Bonds, 1999–2003 (12-week moving average)

Moving-Average Volatility (%) 50

40

30

20

10

0 99

00

01

30-Year Treasury

02

03 30-Year Municipal

S&P 500

Note: Data through 7 February 2003. Source: Based on data from Bloomberg and Thomson Financial.

Figure 3.

10-Year AAA Municipal Bond Yields as a Percentage of Treasury and Corporate Bond Yields, 1 December 1993 to 31 January 2003 A. Versus 10-Year Treasury Bonds

Percent

90

75

1/ D ec /0 2

1/ D ec /0 0 1/ D ec /0 1

1/ D ec /9 8 1/ D ec /9 9

1/ D ec /9 7

1/ D ec /9 6

1/ D ec /9 5

1/ D ec /9 3 1/ D ec /9 4

60

B. Versus 10-Year AAA Rated Corporate Bonds Percent

85 70

1/ D ec /0 2

1/ D ec /0 0 1/ D ec /0 1

1/ D ec /9 8 1/ D ec /9 9

1/ D ec /9 7

1/ D ec /9 6

1/ D ec /9 5

1/ D ec /9 3 1/ D ec /9 4

55

Note: Merrill Lynch Municipal Bond Index for 7- to 12-year maturities; Merrill Lynch U.S. Corporate Bond Index for 7- to 10-year maturities. Source: Based on data from Municipal Market Advisors and Merrill Lynch & Company.

42 • www.aimrpubs.org

©2003, AIMR®

Strategic and Tactical Allocation of Municipal Bonds in Private Client Portfolios because the after-tax equivalent yield will be higher from the muni investment. The historical average yield relationship over the last 10 years is only 80 percent—the current level is three standard deviations above this average. Munis are also a great diversifier. Consider the correlations with the following indexes for the period from 1 January 1990 to 31 December 2002: Nasdaq Composite Index, 0.03; Dow Jones Industrial Average Index, 0.10; Wilshire 5000 Index, 0.11; and S&P 500 Index, 0.14. The correlation of municipal bonds with equity markets is negligible. And in fact, munis have exhibited negative correlation with equities in 2003. This low level of correlation is driven partly by emotional factors and partly by technical factors. Many retail investors who fear further stock market sell-offs are selling equity mutual funds and buying muni mutual funds. Nonetheless, these low correlations are based on more than just the poor equity market performance of 2000–2002.

Tactical Allocation From a risk–return standpoint, munis are the best place to be. Figure 4 shows the after-tax risk and return (based on a 28 percent marginal federal tax bracket) of four fixed-income indexes with comparable durations for the 10 years ending 31 December 2002—Merrill Lynch Agency 5–7 Year Index, Merrill Lynch Corporate 5–7 Year Index, Lehman 5-Year Muni Index, and the Merrill Lynch 5-Year Treasury Index. During this period, munis provided about as high a return or higher with far less price risk. Given that a strategic allocation to munis makes sense, what can managers

Figure 4.

do tactically to add value to fixed-income portfolios for high-net-worth (i.e., taxable) clients? Crossover Allocation. Crossover allocation means using all fixed-income sectors to maximize after-tax returns. At Standish Mellon Asset Management, when we see that munis are becoming rich (although, of course, the opposite has been true recently), we crossover into the taxable markets. The starting point for determining the appropriate crossover portfolio weighting is the client’s income tax situation. The trade-off between tax-exempt and taxable bonds will vary according to investment guidelines, market conditions, relative value, and liquidity. The excess yield provided by munis versus taxable bonds is typically lower in the short end of the curve, where the excess yield advantage of munis tends to be much greater. Thus, a tactical crossover allocation to taxable bonds makes sense in the short end of the curve. Managers need to know two things: the historical after-tax yield of each taxable bond sector as a reference point for relative value versus munis and the likely direction of the price relationship between munis and the other sectors going forward. Even if a manager decides that corporates are cheap at the present time and their after-tax yield is attractive, it still might be dangerous to crossover. For example, during the past two years, while corporates offered compelling after-tax yield versus munis, the prices of corporates kept depreciating versus munis as quality concerns plagued the sector. In this environment, the crossover strategy would not have accomplished the

After-Tax Risk and Return for Municipal, Agency, Corporate, and Treasury Bonds, 10 Years Ending 31 December 2002

Annualized Return (%) 6.5 Merrill Lynch Agency 5−7 Year Index Merrill Lynch Corporate 5−7 Year Index Lehman 5−Year Muni Index 5.5 Merrill Lynch Treasury 5-Year Index

4.5 2.5

3.5

4.5

5.5

6.5

Standard Deviation (%) Note: Taxable returns netted down at top federal tax rate (28 percent). Source: Based on data from Lehman Brothers and Merrill Lynch & Company.

©2003, AIMR®

www.aimrpubs.org • 43

Investment Counseling for Private Clients V desired outperformance. In other words, it is extremely important for managers to assess the future price direction of the relative sectors as well as the simple after-tax yield advantage of doing the crossover trade. Figure 5 shows the ratio of the muni yield to the pretax yields of three taxable bond sectors for the 10year period ending December 2002. This crossover historical yield comparison shows munis becoming cheaper relative to Treasuries and agency debt. The opposite has been true for corporate bonds since the credit crisis began in late 1998. In December 1999, corporate bond yields were below the AMT breakeven point at 72 percent (i.e., a marginal tax rate of 28 percent; the non-AMT breakeven point for a 39.6 percent federal taxpayer is 61.4 percent). Although this might have appeared to be a good opportunity to add after-tax yield by selling munis and buying corporates, as I explained earlier, it would have been a bad trade because corporates cheapened even further. Late in 2002, however, if a taxable investor made a crossover trade into corporates, it would have been a good trade because the investor would have captured both the favorable after-tax yield and the principal appreciation from the relatively greater price appreciation of the corporate sector from that point forward. As with every other market, timing and relative value cannot be ignored in the muni market. Consider the after-tax yield advantage of munis versus taxable bonds for an AMT filer, as shown in Table 3. As in Table 2, at the short end of the curve,

Figure 5.

Table 3. Municipal After-Tax Yield Advantage vs. Taxable Bonds, as of 20 February 2003 (AMT taxpayer; bps) Bond Sector

1 Year

2 Year

5 Year

10 Year

Treasury

21

17

47

94

Agency

16

12

26

59

Asset-backed

–2

–7

9

43

Corporate A

6

–42

–24

8

Note: After-tax yield advantage calculated at 28 percent tax rate. Source: Based on data from Municipal Market Data and J.P. Morgan Chase & Company.

it is possible to capture a bit more after-tax yield by crossing over into the corporate market, but this pattern does not hold as one goes further out the curve. It may appear that two-year A rated corporates— with a 42 bp after-tax yield advantage—are attractive versus munis, but given the greater volatility of the corporate market, this yield advantage alone is not enough to compensate investors for lower quality and possibly lower liquidity. The Yield-Curve Trade. Although a crossover trade might not be best pursued in the current environment, the yield-curve trade absolutely will be the trade of the coming year, second half 2003 to first half 2004, because the yield curve is so steep. At 346 bps, the slope of the curve (measured as the difference between the yields of the 30-year and 2-year maturities) is more than double its 8-year average of 175

Crossover Historical Yield Comparison for 10-Year Maturities, December 1993–December 2002

Muni/Taxable Yield (%) 100 95 90 85 80 75 AMT

70 65 60

Non-AMT

55 50 12/93

12/94

12/95

12/96

Treasury

12/97

12/98

12/99

Agency

12/00

12/01

12/02

Corporate

Note: AMT/Non-AMT = after-tax breakeven point for individual investor. Source: Based on data from Merrill Lynch & Company and Bloomberg.

44 • www.aimrpubs.org

©2003, AIMR®

Strategic and Tactical Allocation of Municipal Bonds in Private Client Portfolios bps, as shown in Figure 6, and also two-and-a-half standard deviations above that average. The 2-yearto-10-year slope is also remarkably steep at 245 bps relative to, for example, the flat slope of 32 bps on 31 December 2000 just prior to a major rally in the short end of the yield curve and to the average of 100 bps. At the front end of the curve, a lot of incremental yield can be gained without adding much duration by

Figure 6.

taking advantage of the steepness in intermediate maturity muni yields. For example, in Figure 7, the bars represent the incremental yield pickup of moving from one maturity to the next in AAA rated munis. The greatest incremental yield pickup for extending maturity is in maturities of less than 10 years. The yield curves shown in Figure 7—the dotted line at the beginning of 2003 and the solid line at mid-February—show

Slope of Municipal Bond Yield Curve for 2- to 30-Year Maturities, 1995–2002

Yield (bps) 400 350

300 250 200

Average

150 100 95

96

97

98

99

00

01

02

Source: Based on data from Thomson Financial.

Figure 7.

Municipal AAA Yield-Curve Comparison, 31 December 2002 vs. 18 February 2003

Yield Pickup (bps)

Yield (%) 5.0

60 18/Feb/03

50 42 38

40

37

31/Dec/02 30

30

3.0 25

23

17

20

12

12

10 0

4.0

43

10

10

2.0 9

8

7

0

0 1.0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

20

30

Maturity (years) Source: Based on data from Thomson Financial.

©2003, AIMR®

www.aimrpubs.org • 45

Investment Counseling for Private Clients V that yields have remained unchanged at both the long and short ends of the yield curve but that the market sold off in the belly of the curve. Hedge funds often put on the crossover trade from the taxable bond sectors to the muni sector to exploit the relative value of markets. Believe it or not, the muni market has become extremely interesting to hedge funds, as well as to the broader investing public, because of its cheapness relative to the other liquid fixed-income sectors. The muni sector has seen a lot more issuance than other sectors, the deal size is much bigger, and credit quality has remained high. These factors have supported the liquidity of the muni sector. Thus, the potential exists for a bit more profitability in arbitrage situations. When munis were trading close to 100 percent of Treasury yields, hedge funds bought tax exempts aggressively, focusing on the intermediate part of the curve because of the lack of callability in those maturities and the number of deals of large size being issued with intermediate maturities. Once demand drove muni yields down to, say, 95 percent of intermediate Treasury yields, hedge funds dumped their munis, leading to a sell-off in the market. This crossover trade was later repeated after the sell-off successfully returned munis to a position of relative value versus the taxable bond sectors. It is likely that the muni sector will become more volatile (but more efficient) as it becomes increasingly driven by large arbitrage accounts. The traditional muni investor needs to be aware of this potential toward greater volatility to avoid getting caught on the wrong end of the trade. The slope of the yield curve is important, but how it can be manipulated by crossover investors is becoming even more important. Panel A of Figure 8 shows the annualized total return for municipal bonds at different points on the yield curve for the 1990–2000 period. Not surprisingly, total return increases as maturity is lengthened. Since 31 December 2000, an anomaly has occurred, however, with short-term bonds outperforming long-term bonds, as shown in Panel B. Reversion to the mean would suggest that long-term muni bonds now hold greater relative total return potential—in a stable interest rate environment or in the long run—than muni bonds at other parts of the curve because they have underperformed short-term bonds versus historical experience. Looking ahead, should the fear of rising interest rates encourage the selling of long-term bonds and the reinvesting of those proceeds in short-term bonds? Two arguments contradict such a view. The first is reversion to the mean, as I have just described. The second is that the muni yield curve is so steep that the opportunity cost of owning short-term bonds can be 46 • www.aimrpubs.org

Figure 8.

Annualized Historical Yield-Curve Return for Municipal Bonds A. 1 January 1990 to 31 December 2000

Return (%) 7.91

8 7

8.00

7.43 6.36

6 5 5

10

20

Long

Maturity (years) B. Since 31 December 2000 Return (%) 8

7.50

7

6.78

7.01

6.99

10

20

Long

6 5 5

Maturity (years) Note: Panel B data through 31 January 2003. Source: Based on data from Lehman Brothers.

punishing. The recent rally in interest rates has had little effect on the long-term end of the yield curve. As of the end of the first quarter of 2003, to forgo a 30-year maturity for a 2-year maturity would mean a give-up of 340 bps in yield, and to forgo a 10-year maturity for a 2-year maturity would mean a give-up of about 225 bps. This give-up in yield represents a great deal of money for a sizable fixed-income portfolio, especially given today’s historically low yields. I keep focusing on 31 December 2000 because it predates the significant yield-curve shift shown in Figure 9. This dramatic change in the shape of the yield curve between then and now is called a “snapdown” in short-term yields, which pivoted off the intermediate part of the curve. This event was driven by the U.S. Federal Reserve Board’s surprisingly aggressive easing in January 2001, a policy it has so far maintained. The result was a significant steepening of the yield curve because the yields at the long end of the curve remained virtually unchanged. The curve will have to flatten at some point and return to a more normal shape, and so this snap-down will be reversed by a “snap-up” should again pivot off the intermediate part of the curve. This means that shortterm yields will rise relatively more than long-term yields. Current real yields in the municipal sector are favorable to those in the Treasury market, which also ©2003, AIMR®

Strategic and Tactical Allocation of Municipal Bonds in Private Client Portfolios Figure 9.

Decline of Short-Term Municipal Bond Yield, 31 December 2000 vs. 31 December 2002

Yield (%) 5.00 4.50 4.00 3.50 3.00 2.50 2.00 1.50 1.00

31/Dec/00

31/Dec/02

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

20

30

Maturity (years) Source: Based on data from Municipal Market Data.

minimizes the risk of remaining invested at the long end of the muni market. The challenge is in structuring a municipal bond portfolio to take advantage of the anticipated flattening of the yield curve. Figure 10 shows two extremes: a bullet approach and a barbell approach. The bullet is a concentrated exposure at the intermediate part of the curve, which would capture more yield than the barbell approach but would fail to take advantage of the expected flattening of the yield curve. The barbell provides broader exposure along the yield curve, with a concentration in the maturity buckets that will offer the best roll if the curve flattens as expected. Consider two market scenarios. In the first scenario, the yield curve flattens, returning to its average 175 bp slope between 2-year and 30-year maturities. In the second scenario, the yield curve exhibits only

modest flattening, returning to a slope of 205 bps. In the first scenario, the barbell strategy will outperform the bullet by slightly more than 100 bps. In the second scenario, the barbell’s excess return over the bullet is 90 bps. The reason the barbell outperforms in both scenarios is its exposure to the long-term end of the curve, whereas the bullet is concentrated in the intermediate range of the curve, the pivot point for yields.

The AMT The AMT poses a problem for taxable investors in the United States. Table 4 shows that by 2010, if tax law remains unchanged, the proportion of taxpayers subject to the AMT will grow significantly. Political pressure, however, may lead to a change in the law before such a scenario develops. The AMT is not indexed for

Figure 10. Duration Bucket Exposure: Barbell vs. Bullet Approach Portfolio Allocation (%) 80

60

40

20

0 0−2

2−3

3−4

4−5

5−6

6−7

7−8

8−9

9−10

10+

Duration (years) Barbell

©2003, AIMR®

Bullet

www.aimrpubs.org • 47

Investment Counseling for Private Clients V now than it has been in recent years, the spread will continue to widen as more investors avoid buying AMT-subject bonds as they anticipate AMT liability. Figure 13 illustrates that the pickup in yield that AMT-subject bonds provide is not sufficient to offset paying the AMT on the coupon and that, net–net, the investor is better off in all maturities to buy AMTexempt AAA munis or Treasuries. The challenge for a portfolio manager working with an AMT-subject client is to mitigate the impact of the AMT. The most important way to do this is to avoid buying municipal bonds that are subject to the AMT. Unfortunately, many managers do not have good control over the AMT status of bonds that are put into their clients’ portfolios and, in many instances, learn only at the time of tax filing that their clients owe the AMT on some of their municipal bonds. A wide variety of tax management strategies that are beyond the scope of this presentation can also be used to mitigate the effects of the AMT. For example, because the exercise of stock options is considered regular income, which is subject to the AMT, AMT-subject taxpayers need to carefully time their exercise of stock options. Two deductions that are still allowable in the calculation of the AMT are charitable giving and home mortgage interest. Maximizing these deductions is a tax-saving measure that can mitigate the impact of the AMT.

Table 4. Estimated Increase in Filers Subject to AMT (percent) Adjusted Gross Income per Year Year

$75–$100K

$100–$200K

$200–$500K

2002

3

11

36

2010

79

94

97

Source: Based on data from Lehman Brothers.

inflation, so each year an increasing number of taxpayers becomes subject to the tax. This is a challenge for private wealth management because both managers and clients must be prepared if the law does not change and, at the same time, be prepared if the law does indeed change. Issuance of bonds subject to the AMT accounts for about 10 percent of total issuance in the muni market, as shown in Figure 11. So, it is quite easy to find muni bonds that are not subject to the AMT for those investors who currently are subject to the AMT. Unfortunately, some of the higher-yielding sectors, such as industrial development bonds, housing bonds, high-quality intermediate-term structures, and student loan bonds, are also subject to the AMT, so avoiding them may mean sacrificing yield. The steady supply of bonds subject to the AMT combined with decreasing demand for them (as more investors anticipate becoming liable for the AMT) has caused yield spreads on bonds subject to the AMT to widen versus yields on muni bonds that are not subject to the AMT (such as insured bonds), as shown in Figure 12. Even though the spread between yields on AMT-exempt and AMT-subject bonds is wider

Conclusion In the current market, municipal bonds typically offer higher after-tax yields than other fixed-income sectors. Furthermore, these attractive yields are bundled with high credit quality (lower levels of default than corporate bonds at every credit rating) and low price

Figure 11. Issuance of Bonds Subject to AMT vs. Total Municipal Bond Issuance, 1989–2002 Issuance ($ billions) 500 400 300 200 100 0 89

90

91

92

93

94

95 Total

96

97

98

99

00

01

02

AMT

Source: Based on data from Thomson Financial.

48 • www.aimrpubs.org

©2003, AIMR®

Strategic and Tactical Allocation of Municipal Bonds in Private Client Portfolios Figure 12. 10-Year Yield Spread for Insured Municipal Bonds vs. AMT-Subject Municipal Bonds, July 1994–July 2002 Yield Spread (bps) 33 28 23 18 13

client’s portfolio is approximately equally divided between long-term and short-term maturities. And the AMT lurks on the horizon as a thorn in the side of taxable investors. Will it or will it not be repealed or adjusted? In the meantime, steady issuance of AMT-subject bonds and a lessening of demand for them are causing spreads on AMT-subject bonds to widen but not enough to compensate for the AMT if owed. All told, the muni market offers many potential rewards for taxable investors, as well as a few wooden nickels.

7/ 02

7/ 01

7/ 99 7/ 00

7/ 97 7/ 98

7/ 96

7/ 95

7/ 94

8

Note: Insured muni curve is virtually AMT free. Source: Based on data from Bloomberg.

Figure 13. Yield–Tax Trade-Off for AMT-Subject and AMT-Exempt Bonds Yield (%) 5

volatility. Their exceptional characteristics make them an attractive fixed-income vehicle for investors in almost any tax bracket. Munis can also be used in short-term tactical allocations. In a crossover trade, taxable investors can move from munis to Treasuries to corporates and back to capture superior after-tax yields in anticipation of changes in the relative yields of each sector. This type of activity has added volatility to the muni market in recent years. In anticipation of a flattening of the yield curve and a return to a more normal short-term/long-term yield spread, the trade that would best position investors to take advantage of such a flattening is a barbell strategy in which a

©2003, AIMR®

4 3 2 1 0 2 4

10

20

30

Maturity (years) AAA AMT-Exempt Muni Yield After-Tax Treasury Yield After-Tax AAA AMT-Subject Muni Yield Source: Based on data from Bloomberg and Thomson Financial.

www.aimrpubs.org • 49

Investment Counseling for Private Clients V

Question and Answer Session Christine L. Todd, CFA Question: Would you combine mutual funds with individual bond selection in managing a muni portfolio for a private client? Todd: I would avoid buying mutual funds. With the decline in interest rates, they contain a lot of accumulated capital gains and they do not allow investors the ability to control their tax situation. I would buy separate holdings of munis, even for a small portfolio. Question: For active management, what is the minimum number of issues you would buy in order to achieve adequate diversification, and what would be the shortest and longest maturities? Todd: We strongly believe that a $2.5 million to $3.0 million portfolio with 30 or 40 issues can be actively managed—although not full-blown active management on the order of 30–40 percent annual turnover because position size would be too small. A $5 million portfolio is more suitable for fullblown active management because it allows 3 percent positions that are large enough to be liquid and the ability to do relative-value trading—concentrated sector overweights, state overweights, and yield-curve overweights. The ideal is a portfolio large enough for positions of 250,000 par value, but positions of 100,000 par value should still provide reasonably good liquidity. With a $1 million portfolio, you should buy 10 high-quality, noncallable bonds with maturities from two to seven years. You should search for anomalies that enable you to gain an extra 10 bps in yield because of excess supply in the issuance of a particular state or through odd-lot trading at a regional brokerage firm. Buy bonds with those anomalies and hold them. Collect your coupons 50 • www.aimrpubs.org

until they mature and reinvest with a similar strategy. Active trading of a $1 million portfolio would cause transaction costs to erode returns, so we don’t recommend active management for such a small portfolio.

the portfolio, providing a better total portfolio return.

Question: What is the current yield-curve slope telling us about the economy?

Todd: We’re seeing a lot of headlines about bonds being issued to bail out certain states and thus more issuance than was anticipated. The wider spreads have been driven by an emotional reaction to the glut of supply in the marketplace. Thus, we are not afraid of these credits whose yields have widened out. At Standish Mellon, we see this situation as a tremendous opportunity for relative-value trading. These state governments are well run and, in many cases, have zerodeficit requirements. They also have a great record with not a single default, so we see a strong recovery rate.

Todd: The current steep slope is saying that the Fed is expected to keep short rates extremely low. We don’t believe that another easing is likely, but the Fed could further lower short-term interest rates to support the economy if unfortunate geopolitical events were to occur. Of course, the Fed has other means to boost the economy without lowering short-term rates. Such a disparity between longterm and short-term rates can mean that the market is anticipating a heating up of inflation in the future, and in certain sectors, inflation has shown signs of bubbling up. Not long ago, however, when the curve was just as steep as it is today, people were talking about a deflationary scenario. So, I don’t believe that the current steepness is signaling inflation but rather investor concerns about the economy and weak consumer confidence. A flight to quality has occurred and no more so than at the short end of the yield curve because that is where volatility is the lowest. Question: Why is monthly correlation with other asset classes important for a long-term investor? Todd: The lack of correlation between munis and other financial asset classes means that the total portfolio will perform better on a risk-adjusted basis. For example, when equities are selling off, your allocation to munis will stabilize

Question: Given the deteriorating budgetary outlook for state governments, what is your outlook for the supply of both high- and lowquality munis over the next year?

Question: What is your opinion on the relative quality of general obligation bonds and revenue bonds? Todd: The market has traditionally considered general obligation bonds to be of higher credit quality than revenue bonds because of the taxing power of the issuer. At Standish Mellon, we don’t take that view. We actually believe that revenue bonds are a better source of quality because they’re easier to analyze. In the past 18 months, the market has seen that event risk applies to general obligation bonds because legislation to increase taxes or freeze tax decreases is hard to pass. Many states now regret tax reductions they made late in the economic boom. Fortunately, some states had sizable rainy-day funds. States that are experiencing greater deficits or greater revenue reduction will have to rely more heavily ©2003, AIMR®

Strategic and Tactical Allocation of Municipal Bonds in Private Client Portfolios on new issuance, which will cause their spreads to widen relative to other states. Investors have already seen a dramatic example of this effect in New York and California. Question: What is your opinion of the credit quality of munis issued by New York and California? Todd: Our belief is that the California economy is now stronger than it was in the early 1990s. The budget situation is a bit more severe, but fixes are in place, and not just one-time fixes, but fixes

©2003, AIMR®

that will correct matters for the long term. The yield spread between California and general market AAA munis is 20 bps wider now than it was when Orange County filed for bankruptcy. The wide spreads indicate that the combination of the heavy issuance to bail out the state and the recent utility crisis is driving spreads beyond what credit factors would justify. After all, California is an A rated credit and the sixth largest economy in the world.

For states with heavy issuance, that issuance is causing spreads to widen between the yields of those states and the yields of other comparable quality states. For example, after September 11, New York City took all of its issuance off the table. Now, the pent-up supply is bursting through, bringing a lot of issuance from both New York state and New York City. So, their bonds are trading 1 percent cheaper than others of similar maturity and quality.

www.aimrpubs.org • 51

The Psychology of Wealth Marty Carter Family Communications Advisor Charles D. Haines, LLC Birmingham, Alabama

Investment advisors tend to focus on the “hard” side of managing private client wealth, but they should give equal attention the soft side (i.e., the psychological factors). Understanding the psychology of wealth can help advisors build more effective relationships with their clients. Advisors who do not understand their clients’ needs and objectives are unlikely to serve them effectively.

dvisors who try to understand the psychology of wealth, rather than ignore it, will have more productive relationships with their clients. Although a “soft” issue, the psychology of wealth is a major factor in family dynamics, especially money conflicts, that can have important implications for those charged with managing a family’s wealth. I will review the most common challenges facing wealthy clients and the psychological roadblocks that inhibit their ability to manage their assets. I will also discuss how, in certain circumstances, a facilitator can help to break barriers and build more effective client relationships. At my firm, Charles D. Haines, LLC, for example, we have adopted a multidisciplinary, collaborative approach that compels clients to establish meaningful goals and devise appropriate solutions.

A

Common Wealth Issues A wealthy individual faces many challenges peculiar to his or her status, whether he or she is ultrawealthy, a “millionaire next door,” or an aspiring millionaire.1 Typically, the wealthy talk about money a great deal but rarely talk about the purpose of money or how it might affect their relationships. Some issues affect certain wealth levels more than others, but recognition of certain factors can help advisors gain insight into what is going on in their clients’ lives.

1

For a discussion of the “millionaire next door” type, see Thomas J. Stanley and William D. Danko, The Millionaire Next Door: The Surprising Secrets of America’s Wealthy (Atlanta, GA: Longstreet Press, 1998).

52 • www.aimrpubs.org

Culture. One of the most significant issues is the culture of wealth. In many families, the descendents of those who made the money tend to assume that the well will never run dry. As a result, parents, who perhaps themselves never acquired sound money practices, do not instill appropriate attitudes and behaviors in their children, which may engender a sense of entitlement. For example, the children might think their parents are spending their inheritance, even though the money is not yet theirs to spend. Seeing the family name over the door of a new art center can create at best a selfless spirit of giving back to the community and at worst a sense of entitlement. Those who feel entitled to their wealth tend to be cheap and expect discounts and immediate service. In my experience, the wealthiest people are often the most reluctant to pay for what they want. In addition, wealthy people are often placed on pedestals and receive deferential treatment in their communities. They are viewed as possible donors or potential clients, and such community adulation, although welcomed by some individuals, causes many to retreat. Problem Values. For ultrawealthy families, inculcating in their children a sense of responsibility about money is a significant challenge. In short, parents who do not set boundaries and clear expectations tend to rear children who do not have a strong sense of responsibility or self-worth. If the message is that the money will last forever and no one in the family needs to worry about money, then the children have little or no incentive to do well in school and prepare for a job.

©2003, AIMR®

The Psychology of Wealth Self-Esteem. Wealthy individuals are often uncertain whether they are liked for who they are or for how much money others think they have. A lack of self-esteem is thus often exacerbated by wealth. Some individuals feel undeserving or self-conscious of their inherited wealth or isolated because of the deferential attention they receive. A common complaint among young wealthy individuals in particular is that they are often expected to pick up the tab. A 10-year-old child recently told me about an incident that will likely affect him for years to come. A classmate borrowed his tennis racket and returned it damaged. When he asked the classmate to fix it or buy a new one, the response was, “Buy yourself a new tennis racket. You’re rich.” Another stumbling block to a healthy sense of self-esteem occurs in families in which success is measured by money. Little attention is given to other accomplishments, such as intellectual ability, the capacity to help others, or commitment to family and friends. Money Beliefs. Everyone has what I call “money beliefs,” or attitudes and behaviors about money that stem from childhood. Helping clients to identify the messages they have received about money from their families provides insight into their current attitudes about money. These beliefs also play a significant role in determining how they will relate to an advisor. I have heard every cliché imaginable about money from clients when I ask them about their beliefs: for example, “money doesn’t grow on trees,” “money is the root of all evil,” “marry money, don’t bury it,” or “marry for love or money, but mostly money,” all of which are messages that have been forever ingrained in clients’ psyches. One of the most common answers, however, is that their families simply do not talk about money, which is a message that can be particularly detrimental to effective investment management. Money Personality Polarities. In order to establish a rapport and build a successful working relationship with clients, advisors may find it useful to identify a client’s “money personality.” For example, people tend to be either generous or miserly, straightforward or secretive, cautious or impulsive, intrusive or respectful. People also tend to be more or less dictatorial, more or less financially informed, and more or less risk averse. Lifestyle. Surprisingly, many wealthy individuals live beyond their means. One of my clients will be insolvent by the third quarter of 2003 if she does not drastically cut her spending. For years, her advi©2003, AIMR®

sors have been warning her about the possibility of running out of money, but she has refused to heed their advice. She recently purchased a villa in Rome and continues to make unsustainable philanthropic commitments. She can choose to sell some of her assets or have a family member bail her out, as she has done before, but she seems to be incapable of altering her extravagant lifestyle. Many people are ashamed of their wealth, however, particularly if it was inherited, so they hide how much money they have. Others flaunt their wealth for the sake of appearances, which often leads to particularly lousy financial decisions. In New Orleans, for example, families will take out a second mortgage to pay for a daughter’s debutante ball. The cost of a wedding pales in comparison to that of a debutante ball, but appearances mean a lot to some people. Relationships. When spouses are from different socioeconomic backgrounds, many problems can arise. The person who grew up without much money may continue to be unnecessarily frugal or, at the opposite extreme, become a spendthrift. One daughter of a wealthy family married a man who had no income. Worried about the balance of power in the relationship, the family gifted the groom $1 million. Unfortunately, he developed a sense of entitlement to the money. He spent nine months sailing and is now building an airplane and merely engaging in the proverbial eating of bonbons on the couch. Needless to say, parents of newlyweds are often concerned that the spouse of the inheritor will take advantage of the newfound access to money. One of the factors that most negatively affects relationships is the reluctance to talk constructively about money. I was recently asked to help complete a prenuptial agreement—on a Tuesday before a Saturday wedding—because the couple had reached an impasse. The problem was that, even though both the bride and groom were from extremely wealthy families and knew they needed an agreement, neither one had been asked by his or her attorney to specify what was necessary and doable. Once I was able to get them talking, we constructed a satisfactory agreement within 45 minutes and sent it to their respective attorneys. Relationships are inevitably affected by power and control issues associated with money. I know a woman who was worried about her grandson’s poor academic performance. Without checking with the child’s parents, she hired a tutor for him, an act that created resentment in the parents. Jealousy between family members who have varying levels of wealth can also have a significant impact on relationships. Furthermore, those who are wealthy sometimes feel put upon by family members who are less well-off. www.aimrpubs.org • 53

Investment Counseling for Private Clients V And wealthy people are frequently puzzled by relatives who refuse to ask for financial help, even for a worthy cause. Financial Literacy. Contrary to popular belief, wealthy individuals are not always financially literate nor do they always recognize the need to have a longterm strategy. Many people hire financial advisors but take little or no responsibility for managing their own money. Some of my clients have never possessed a checkbook because they rely on the family office to pay their bills. As James Hughes stresses, the danger of such an attitude lies in going from shirtsleeves to shirtsleeves in three generations: The first generation toils, the second generation is frugal (although it has access to money the first generation did not have), and the third generation depletes the rest of the fortune. By the fourth generation, the family is back to nothing but shirtsleeves. Therefore, to preserve wealth, families should consider long-term planning, even for a time frame of 100 years or longer.2 My campaign to help families become financially literate and talk about money stems partly from personal experience. My daughter was accepted at two colleges, an A-level school and a B-level school. Only later did we learn that she chose the B-level school primarily because of a scholarship the school offered. I regret that I somehow had given her the message that we could not afford a top-notch school. To avoid such an issue, parents should begin discussing financial issues when children are young. For example, children should be encouraged to do their own due diligence in regard to paying for their education.

Psychological Roadblocks Once advisors understand these common wealthrelated issues, they can help clients move past the psychological roadblocks that inhibit the ability to manage money effectively. Advisors need to ask probing questions in order to delve behind the numbers and better understand what is going on in their clients’ lives. Such a strategy helps to foster long-term relationships that will steer clients toward thinking about the “why” of investing rather than merely the “how” of investing. Attitudes and Behaviors. Many factors govern client attitudes and behaviors in relation to money, but anxiety about money is an overriding concern, no matter how wealthy someone is. Wealthy parents fear their children will be lazy, or if they think they have too little money, they worry about running 2 James E. Hughes, Family Wealth: Keeping It in the Family (Hughes and Whitaker, 1997).

54 • www.aimrpubs.org

out of money. Advisors can make a significant contribution by allaying clients’ fears and letting them know how long they can expect their money to last given all the variables, such as current spending plans and stated goals. Advisors can also help clients overcome their fear of conflict, which is one of the most significant factors that prevents people from talking about money. Some people simply refuse to share financial information with the next generation. Instead, their strategy is to let everyone fight about the money after they are gone. But the fear of conflict often leads to conflict. Consider the case of Sarah, who hid from her husband that she earned more money than he did. Her assumption that he would be uncomfortable with this fact led to her assumption that he did not need to know anything about her spending habits. When her husband had to sign their joint income tax return, however, he discovered her secret and was not happy about being duped. In time, I found out that Sarah’s father used to harangue her mother about expenses. In order to avoid conflict, she and her mother grew accustomed to hiding their purchases. Clearly, Sarah was repeating the same pattern with her husband and needed to alter her behavior. Client Capacity. Another roadblock is client capacity. The habit of buying something as a kind of emotional panacea is a major hurdle for many clients to overcome. Others struggle with disinterest; they have cultivated the idea that they do not need to worry about money. The desire to use money as a weapon, to get even, is another limitation. A couple in the midst of household renovation argued so much about the amount of money to spend that the husband refused to spend another cent. Furious, the wife got even by “forgetting” to write their quarterly estimated income tax payments to the U.S. government. Keep in mind as well the psychological impact of the recent bear market, particularly for clients with liquidity concerns or for those who are financially overextended. Clients are no longer as interested in philanthropy and gift giving as they were in the 1990s. Anxious about the economy, clients may retreat from interaction with their investment manager because they do not want to hear bad news, or they may decide to micromanage their investments or second-guess their advisor’s recommendations. Furthermore, a client’s low risk tolerance may reflect not only values learned at home but also concern over significant downturn in the value of a portfolio. Age is also an important factor in a client’s willingness to listen to advice. Young adults are generally not interested in planning for retirement or setting long-term goals. Nonetheless, advisors can help by ©2003, AIMR®

The Psychology of Wealth encouraging parents to establish financial plans for their adult children. Advisor Attitudes. Another roadblock, which is often unrecognized, is the effect that a client’s wealth may have on an advisor. Some advisors are envious of their clients’ wealth. I admit that hearing wealthy clients complain about their cruise ships being too crowded grates on my nerves. Nonetheless, advisors need to be aware of their own attitudes and prejudices in this regard and keep their personal judgments to themselves.

Solutions My discussion of the “soft” side of money will make the most sense to so-called “right brainers.” People who use the left side of their brain analyze information by breaking it down into units and examining causes and effects (e.g., financial analysts), whereas right brainers look at the big picture by synthesizing information and looking for patterns, relationships, and connections (e.g., psychologists). This preference for using one side of the brain more than the other directly affects the way a person makes decisions; challenging this natural preference can be disconcerting. Because of this dichotomy in approach, advisors should consider hiring a facilitator to help advise certain clients on the soft issues. Clients who are analytical, logical, and think in terms of the bottom line are likely to be easy to deal with, but not all clients fit that mold. If an advisor needs to delve into soft issues with a client who is a left brainer, for instance, the client may become uncomfortable and accuse the advisor of being “touchy feely,” thus putting the relationship at risk. Having a facilitator involved in the communications process can be helpful in managing the client– advisor relationship. Advisors should never open a wound they cannot close. If, for example, a couple fights in the advisor’s presence or a client is secretive or appears to be troubled or hurt, seek professional help in facilitating. And do not worry that a facilitator will try to take away clients. When I first started working with advisors outside my firm, many financial professionals were afraid I would steal their clients. Only gradually did they come to trust that I was interested only in making their client relationships run more smoothly and effectively. Breaking Barriers. Advisors can help clients move beyond their psychological roadblocks by listening and paying attention to the client’s stories about money. In addition, advisors should seek clients’ feedback, a lesson I learned from the following example. A couple hired my firm to update their ©2003, AIMR®

financial plan because they were selling part of their business. The husband asked for a plan that addressed the worst-case scenario. The wife was not happy with the resulting plan. Although she had never voiced her opinion in my presence, I eventually learned that she thought the plan meant that, even though she now had gazillions of dollars, she was going to have to shop at Kmart. Clearly, advisors need to specifically ask whether clients are comfortable with their recommendations. The worst outcome possible—to have no action taken after a plan is devised—is bound to occur if client feedback is insufficient. Both the head (business factors) and the heart (emotional factors) should be acknowledged as having equal weight in financial decisions. Advisors are trained to encourage clients to make sound business decisions, but they need to recognize that, occasionally, emotional factors interfere. For example, some clients ignore the advice not to accelerate their mortgage payments and insist on paying off their mortgage. They are more interested in the psychological satisfaction of ownership than the benefit of low interest rates. Advisors need to take the time to find out about the life experiences that influence their clients’ choice of investments, their priorities, and how much they save or spend. For example, one wealthy couple could not meet their financial goals, and only through a story the husband told was I able to discover why. When I asked the husband what word came to mind when I said the word “money,” he said, without hesitating, “food.” He was the oldest of five brothers. His father had died when he was 11, and he always worried whether the family would have enough food on the table. He did not think this association influenced his relationship to money, but his wife disagreed, suggesting that his reaction was to give his daughters everything they wanted and more. Although he had never made the connection before, he sheepishly agreed that his experience had led him to indulge his daughters to the point of hampering his own lifestyle. Although left brainers may cringe, advisors need to encourage clients to ask themselves the following questions to ascertain their relationship to money: What is meaningful in my life? What are the five most important things in my life? What do I want my money to do? How can I help my children and grandchildren learn to be responsible with money? Sample Model. The multidisciplinary financial planning model Charles D. Haines, LLC, has adopted is not based on rocket science, but it does require an investment of time and money. Our clients are enthusiastic about our collaborative approach and benefit www.aimrpubs.org • 55

Investment Counseling for Private Clients V from the extensive relationships with many team members (investment advisors, financial planners, a philanthropy advisor, a communications advisor to businesses and foundations, and a psychologist). We also work closely with clients’ own advisors, investment managers, estate attorneys, accountants, and insurance representatives. Before drawing up a financial plan, we require that clients take a test to determine their goals and priorities. The first draft of the plan is a reality check to determine whether the plan goals are doable. My favorite part of the process is the two-hour meeting I conduct with each client. I spend a small amount of time ensuring that the plan is doable and that we are, in fact, meeting client expectations. But the majority of time is spent learning the client’s story and determining how it affects his or her relationship to money. I want to know how financial decisions are made in the family and whether one person bears most of the fiscal responsibility. And given that all families fight about money, I want to gain insight into how the person’s family fights about money. Throughout the meeting, I try to resolve any particular psychological roadblocks I encounter. Thus, while others create the financial plan and put together all the financial pieces, I help individuals understand their relationship to money so they can adopt prudent attitudes and achieve their goals. After discussing with key family members their concerns about money and their goals, I facilitate a series of family meetings focusing on family history, attitudes and behaviors about money, and financial education for both children and adults. I encourage all family members to participate in these meetings. At the first meeting, I draw a family tree to encourage storytelling. A grandfather in a recent meeting shared a family secret that his Uncle Harry was a horse thief. His revelation led to a lively discussion about the values passed down by that ancestor. By taking notes and composing the stories on large sheets of paper around the room, various patterns begin to emerge (e.g., the family does not talk about money, or women are not supposed to know anything about money). We then work on eliminating undesirable patterns or attitudes about money. The second family meeting (what I call the money meeting) goes into more depth, although I have designed it to be engaging, not intimidating. I remember my first such meeting. The family had 16 members, with ages ranging from 8 to 74, and the unspoken rule in this family despite its enormous amount of wealth was not to talk about money. I composed 150 questions about money and put the questions in a hat so they could be passed around. If 56 • www.aimrpubs.org

someone picked a question he or she did not like, another question could be chosen. Imagine the silence in the room when the 8-yearold drew the question, “Grandma, what is the dumbest thing you have ever done with money?” Equally provocative was the question, “What has money done for our family, whether good or bad?” The longest silence was produced by the question, “In our family, are there strings attached to money?” I had them write their answers on a piece of paper so I could look at them later. As a facilitator, I do not tolerate confrontation. I prefer to work with individuals behind the scenes. In this case, I discovered that, indeed, in this family, many strings were attached to money. The next task is to identify the patterns that need to be changed in order to develop positive behaviors and attitudes about money. I track these on a chart, identifying benchmarks that will be used to measure progress. I later meet individual family members with their investment manager, estate attorney, or whomever, depending on the topic that needs to be addressed. I also work with any young adults who have, or will soon have, access to a trust fund to help that young adult (the beneficiary) and the trustee to agree on their relationship. I follow the recommendation that Hughes makes in his book to clearly define the roles, rights, and responsibilities of each party. Even though young adults do not tend to think about long-term goals and are not likely to read Hughes’ book, two questions invariably pique their interest: How is their portfolio doing, and what are their family members’ attitudes and behaviors about money? Inheritance can be a touchy subject. While many parents are comfortable sharing information about their estate plan because they appreciate the value of sharing the information so that the next generation can develop its own strategies, other clients refuse to discuss it. I recommend treating this topic gingerly by acknowledging and honoring client wishes in this regard. Sometimes, I have to tell potential inheritors that no information about their future inheritance is available, and I help them deal with that frustration. For families that are philanthropically inclined, family meetings are a perfect opportunity to focus on the importance of charitable giving. To hear about a grandmother’s experience serving as a volunteer firewoman can be important for a child’s self-esteem. The discussions are also designed to help the family identify its philanthropic interests and create vehicles for charitable grant making. With young children, I hold special meetings to discuss allowances, spending plans, wish lists, ©2003, AIMR®

The Psychology of Wealth delayed gratification, and how to be successful with money. I encourage the idea of investing together as a family because it teaches concepts and skills to young family members that they might not learn otherwise. My partner, Charlie Haines, remembers being invited to his mother’s investment committee meeting at the age of 14. Inspired, he purchased two shares of stock. Unfortunately, they were shares in Pan Am Corporation. Nevertheless, the committee members guided him by emphasizing what he gained from the experience of losing money. Letting children make mistakes with money is an important learning tool. A family has hired me and a colleague, Charles Collier, a senior philanthropic advisor at Harvard University, to coach its next generation of nine children on how to be responsible with money. Through cash flow analysis, goal setting, and benchmarks, we are trying to impart a sense of responsibility to them. To do this, we created a series of financial education programs with one of the family’s investment managers, and this approach has worked well. One of the children, a 21-year-old trust fund beneficiary who has been kicked out of three private preparatory schools, initially said the purpose of his trust fund was to allow him to buy anything he wanted. I asked him to be more serious, but he had no other response. I knew I had my work cut out for

©2003, AIMR®

me because he did not say it allowed him to do anything he wanted. In time, after several meetings in which I managed to get him to talk about his goals, his values about money, and his family’s values about money, he decided he wants to do things in a different way from his family. He is wondering how many houses, cars, and boats he really needs. I have also worked with him on changing the nature of his relationship with the family trustee from that of “money cop” to valued resource.

Conclusion By giving equal consideration to the soft side of money as well as to the hard side, advisors can build more effective relationships with their clients. Advisors should stress to their clients that money is merely the means to an end. If goals are unclear or nonexistent, the money serves no purpose. Recall the Cheshire cat’s response to Alice in Alice’s Adventures in Wonderland when she is lost in the garden and asks the cat which way she should go. “That depends a good deal on where you want to go,” says the cat. When Alice responds that she does not care, the cat replies, “Then it does not matter which way you go.” Likewise, if an advisor does not know which way a client wants to go, then the advisor cannot optimally manage that client’s money.

www.aimrpubs.org • 57

Investment Consulting and Its Evolving Role in Advising Family Offices (as corrected September 2003)

Scott D. Welch Managing Director Lydian Wealth Management Rockville, Maryland

The rise of “insti-viduals”—individuals whose needs are virtually institutional in their complexity—means that the demand for investment consultants is likely to grow. But it also means that consultants must meet the requirements of ever more sophisticated clients. Indeed, the key to success for consultants may be the ability to differentiate themselves not through quantitative performance but through the qualitative aspects of the services they provide.

ore and more family offices are relying on investment consultants to provide a variety of services, from investment plan development and asset allocation to manager selection, performance measurement, and more. The impetus behind this presentation was a posting that was made to the “member channels” of the website for the Institute for Private Investors (IPI), a noncommercial consortium of wealthy families and advisors.1 Based in New York City, IPI has between 200 and 300 members, approximately 40 percent of whom have $200 million or more in assets. The following comment struck a particular chord with me and led to a rather lengthy “e-mail thread” on the member channels:

M

Most of my colleagues at the IPI with whom I am friendly do not use advisors/consultants for one or more of the following reasons. I would very much appreciate any comments on these reasons: • lack of value added, • lack of “wattage,” • not being “on the line,” and • lack of candor in advising.

This comment brings to mind the old saying, “Other than that, Mrs. Lincoln, how was the play?” It prompted much discussion within my firm (and, I am sure, at other consulting firms) as to how we can 1 The channels, available only to members of the IPI, can be accessed at www.memberlink.net. Editor’s note: Lydian Wealth Management was formerly CMS Financial Services.

58 • www.aimrpubs.org

prove our worth. To that end, I will discuss several important trends that are making the investment consulting business more vital than ever before. In addition, I will list the attributes that consultants should bring to the table, illustrated by our perspective at Lydian. In particular, consultants’ value added can be discerned by their clients through the manager’s selection process that the consultants follow and the consultants’ performance reporting capabilities. Finally, although identifying the right consultant is not easy, I will offer some suggestions on how a family office might go about finding the best consultant for its needs.

Industry Trends A generally accepted hypothesis is that asset allocation and discipline in adhering to an investment plan are the greatest contributors to long-term investment success. In my view, a number of trends in the private wealth management industry are making consultants instrumental in achieving these goals. As long as consultants can prove to their clients that they are doing a good job designing allocations and maintaining discipline, they should be able to justify their fees. In the words of Dr. Alan Starkie (a consultant specializing in placing business development officers and client relationship managers for high-net-worthfocused firms), wealthy families are “insti-viduals”—

©2003, AIMR®

Investment Consulting and Its Evolving Role in Advising Family Offices individuals who have institutional needs in terms of complexity and sophistication.2 As a result, more of these families are forming family offices, more family offices are starting or joining multifamily offices (MFOs), and more family offices and MFOs are looking to outsource expertise—all of which bodes well for the investment consulting business. Family Offices. Many families retain extraordinary wealth, despite the recent market downturn, and many of them are forming or have formed family offices. According to recent data, the number of families with intergenerational wealth is growing at 12 percent a year and more than 3,500 dedicated family offices now exist in the United States.3 Furthermore, based on IPI’s annual survey of member families, the percentage of respondents who have a family office rose from 62 percent in 1999 to 80 percent in 2000 (the most recent data available).4 Alternative Advice. Families are not only more willing to form family offices; they are also more likely to seek advice from alternatives to such traditional service providers as banks, trust companies, and brokerage firms. The reasons for this transition are numerous and include the following: ■ Perceived conflicts of interest. Media reports about conflicts of interest between the research and investment banking divisions at major brokerage firms have been discouraging. ■ Captive product mix. Many firms claim to offer a broad mix of investment alternatives to their clients through an “open architecture” platform but do not. ■ Constant acquisitions. Constant acquisitions tend to create uncertainty and a lack of continuity within an organization, both of which make clients uncomfortable. ■ High turnover rate. The turnover rate for those involved in relationship management is high in the corporate world. The problem is that just as someone becomes adept at understanding client needs, he or she moves on to a different job, either within the same organization or with a rival firm. ■ Poor investment performance. Poor investment performance is one of the main reasons clients are seeking alternatives to traditional service providers. More than any single factor, the recent market down2

Alan Starkie, “The Myth of the Million Dollar Revenue Producer,” Private Asset Management (4 February 2001).

3

Based on data from the Merrill Lynch/Gemini Consulting World Wealth Report 2000 (www.foxexchange.com/public/fox/news/ industry_trends/world_wealth_report_2000.pdf) and the Family Office Exchange (www.familyoffice.com).

4

IPI Year 2000 Member Profile Survey. This survey is available only to members of IPI and can be accessed at www.memberlink.net.

©2003, AIMR®

turn has focused investors’ attention on the suboptimality of using only captive or internal asset managers. ■ Wrong business model. The salesman/portfolio manager model—whereby a business development officer gathers clients and assets while a portfolio manager manages the assets and serves as the primary contact for clients—has not proven to be successful. When clients have questions or concerns, especially in times of poor market performance, portfolio managers may not be the best professionals to turn to for answers. (Put less delicately, the very characteristics and personalities that make someone a good portfolio manager often make that same person a less than optimal contact point for concerned clients.) According to some independent studies, a relationship manager/support team model is better in terms of both gathering assets and retaining clients.5 A relationship manager remains with a client throughout the investment decision-making process and maintains knowledge of the client’s investment needs and preferences. A team of specialists in such areas as portfolio management, concentrated wealth, trusts and estates, and accounting provides the necessary support. The previously mentioned surveys indicate that clients much prefer this business model, as they have a consistent contact person who understands their needs and objectives and who can bring to bear the necessary experts within the organization to address specific needs. ■ Sellers of products. Many clients perceive traditional service providers as sellers of (often captive) products rather than trusted providers of advice. Outside Consultants. As another indication of a trend toward the use of consultants, according to IPI member surveys, the number of families retaining outside consultants increased from 35 percent in 2001 to 42 percent in 2002.6 A majority of families still do not use consultants, but survey respondents are, by definition, self-selected, so the results need to be viewed in context. I suspect we will see an increase in the use of outside consultants as the following factors come into play: ■ Buy-or-build decision. When high-net-worth investors form family offices, a decision has to be made whether to “buy” the necessary support structure (i.e., enlist outside experts) or to build it internally. 5

Based on data from studies conducted by Tiburon Strategic Advisors and the Creating Equity Group. These surveys can be accessed at www.tiburonadvisors.com and www.cegworldwide.com.

6 IPI Year 2002 Member Profile Survey. This survey is available only to IPI members and can be accessed at www.memberlink.net.

www.aimrpubs.org • 59

Investment Counseling for Private Clients V Buying the support structure through a consultant is often the best choice. Building the structure can be quite a lengthy and expensive process. For example, it has taken Lydian 10 years to build the business we now have. ■ Generational changes. As generational changes occur within families, family members may become less attached to their historical providers of advice and less interested in directly managing their assets. If the third, fourth, or sixth generation has no interest in running a family office or even being a part of it, then hiring a consultant is an obvious alternative. ■ Acquisitions. Consider some noteworthy events of the past few years: SunTrust Bank acquired Asset Management Advisors, Wilmington Trust Corporation acquired Balentine & Company, and Atlantic Trust Group acquired Pell Rudman & Company (prior to being acquired itself by Invesco). All of the acquired firms were independent investment consulting and wealth management firms prior to being bought. These are just three examples of a trend in which more and more trust companies seem to be acquiring consulting firms. One possible explanation for these acquisitions is that the trust companies—all of which had historically captive investment management products—realized the “flight risk” associated with younger beneficiaries who are seeking the optimal (i.e., objective) investment solution. Given the choice of building a business internally or buying one, buying one made more sense. As an alternative to the acquisition model, however, I will note that Lydian has successfully partnered with several regional bank and trust companies that sought to expand their wealth management offering without necessarily acquiring another company. ■ Specialization. No firm can possibly excel in all areas. Yet, some firms that brand themselves as MFOs attempt to position themselves as being capable of providing “best-of-breed” performance across multiple services: investment management, wealth management, and family office services (such as tax return preparation, accounting, and record keeping), which implies they can perform these services better than an outsourced solution. True consulting, however, entails relying on specialists—the best providers for whatever the service may be—rather than maintaining all services in-house. ■ Lack of omniscience. People often ask whether the past three years have been difficult with respect to maintaining existing client relationships and developing new ones. The answer, at least for Lydian, is no. The past three years were not nearly as difficult as the period from 1996 to 1999, when many investors, confusing a bull market for investment brilliance, thought they did not need professional 60 • www.aimrpubs.org

assistance. Over the past three years, however, the concepts of objectivity, discipline, diversification, and asset allocation have regained popularity, which has been good for the consulting business. ■ Independent expertise. Clients have come to recognize the need for independent expertise and access to the “best and brightest.” ■ Improved technology. Technological advances have made “virtual” family offices a reality. Family members can communicate directly with each other and run a successful operation without having to be in the same location. MFOs. More families are either joining or forming MFOs. They are being launched by both individual family offices and institutional firms (banks and brokerages)7 for a variety of reasons: ■ Outside investors and profitability. As more families start hedge funds internally, discover hedge funds, or create funds of funds within their own family offices, they find they can attract new clients with this expertise and increase profitability. ■ Economies of scale and cost reduction. Another driver has been the need to contain costs. Sometimes, a family office’s fixed costs are too high to efficiently serve only its dedicated family. Accepting additional clients allows the office to leverage its structure and hired expertise, reducing the dedicated family’s costs. ■ Consolidated buying power. Consolidated buying power allows consultants to negotiate better deals for their clients. The goal is to have large firms treat all clients of a firm as one big client. Placing $500 million in collective assets with one manager, for example, should reduce costs for all of the individual clients that make up that $500 million. ■ Threat of stagnation. MFOs can help individual family offices retain and attract top talent. Often, once the investment plan is in place and the portfolio is implemented, talented managers get bored. Opening the doors to other families creates new challenges, which create a more dynamic workplace.

What Consultants Should Bring to the Table To meet the demands of these trends, consultants need to offer objectivity, expertise, discipline, access (to managers who might otherwise be off-limits and to reduced investment minimums), education, and 7

For more information on the involvement of banks and brokerages, see “Family Office Roundtable,” Private Asset Management (17 September 2001).

©2003, AIMR®

Investment Consulting and Its Evolving Role in Advising Family Offices service. The most critical component is service, but everything a consultant does should be driven by all six characteristics. An issue that arises, however, is how to measure a consultant’s performance. Consultants, after all, are not sellers of products but rather providers of advice. How can they prove they are adding value and providing “wattage”? How can they demonstrate to clients that they are “on the line” with respect to results and being candid in their advice? To answer these questions, I have identified two tangible (i.e., measurable) areas in which consultants can add value: manager search and selection and performance reporting.8 Manager Search and Selection. If consultants can deliver a portfolio construction approach that integrates and optimizes the performance of managers and these managers consistently add alpha and offer tax efficiency and reduced fees, then consultants will be adding value and providing wattage. Therefore, when vetting managers, consultants should do the following: • objectively and proactively monitor the universe of managers, • provide access to any manager (not restricted to legacy, wrap, or platform managers), • use consolidated assets to negotiate lower fees (manager, custodial, and trading) and greater access (reduced minimums and closed managers), • evaluate return in a risk-adjusted context, • evaluate managers using benchmarking and asset class, peer, and universe comparisons, and • evaluate the performance of the entire portfolio as well as each individual component. Most people in the industry would agree that the previous summary describes the baseline for competent consulting. At Lydian, investment goals and asset allocation come first in the manager selection process. In other words, the process is driven by the objectives of the investment policy. When a new client asks the classic question, “What should I do? I have $25 million and a wife and two kids,” the correct answer is “I have no idea.” Before recommending a particular allocation strategy, the consultant should find out the client’s lifestyle needs, philanthropic goals, generational transfer issues, and so on. Moreover, even though we conduct separate manager searches for each asset class, we are careful to view the portfolio as a whole, not as a collection of individual managers. A portfolio construction approach 8 For

more information on these topics, see Scott Welch and Jamie McIntyre, “Show Time,” Bloomberg Wealth Manager (April 2003) and Scott Welch, “The Role of the Investment Consultant in Advising Family Offices,” Bloomberg Wealth Manager (July 2003).

©2003, AIMR®

allows us to evaluate each manager in isolation and then put together a portfolio that considers manager correlations both within and across asset classes. Rather than trying to pick the best manager, the goal is to select the best combination of managers in terms of optimal risk–return characteristics. Manager performance should be monitored relative to benchmarks and peer groups, keeping in mind that good asset allocation combines art and science. In other words, trust the math but use common sense. For example, if a mathematical allocation model with no investment constraints was used to construct a hypothetical portfolio, the result could be a 100 percent allocation to an absolute-return hedge fund strategy because of its consistent expected return and low volatility. The math, of course, does not mirror the intuitive reasoning of an experienced advisor. Such a portfolio might be appropriate for a few clients but not for the majority. We advocate a “core-satellite” approach because certain asset classes seem to be more efficient than others. We favor tax-sensitive indexing managers for the asset classes we deem to be efficient and active managers for those we believe are less efficient. The performance of the managers we select is far more important to us than the performance of those we do not select, no matter how stellar. We concentrate on making sure that our managers are doing the job we hired them to do. We focus on long-term results, not quarterly performance, and try to be unemotional; how much we personally like a manager does not override his or her results. Ironically, in a couple of cases, families who had long-standing relationships with managers they wanted to fire, but were uncomfortable doing so, hired us to “do the dirty work”; they hired us to fire them. Finally, the following quantitative and qualitative factors are important in the manager selection process: ■ Quantitative factors. On the quantitative side, look for long-term, risk-adjusted outperformance relative to the index. We take into consideration all the usual statistics, such as alpha, beta, standard deviation, Sharpe ratios, and upside and downside capture ratios. We are not interested in managers who alternate between hitting home runs and striking out. We prefer managers who hit singles and doubles year after year and consistently beat their benchmarks in both up and down years. ■ Qualitative factors. On the qualitative side, consider how managers generated their performance (was it on the broad portfolio or because they picked one or two stocks that drove the performance?), whether they can repeat their performance, and how www.aimrpubs.org • 61

Investment Counseling for Private Clients V they fit with other managers in the portfolio in terms of holdings and investment style. Another important factor is whether the current team managing the assets is the same team that was responsible for the historical track record. We do not change managers often and are fairly loyal to the managers we select, but if someone who was responsible for a manager’s success leaves, that manager could be fired; likewise, if a manager claims to be a value manager but includes growth stocks in a portfolio for no justifiable reason (i.e., inexplicable style drift), that manager is likely to be fired. Other decision criteria include who owns the firm (particularly whether the manager is an owner) and how quickly the firm has grown (or not grown). We are not likely to be interested in a successful manager who has gathered a lot of assets quickly because the larger the amount of assets, the more difficult it can be to replicate historical results. As a general rule, we tend to select boutique managers or asset class specialists more often than large multistrategy firms. Interestingly, investment management divisions of banks, brokerage firms, and insurance companies are rarely selected. And many of the best managers neither belong to a wrap program nor have any interest in being in one. They do not want to cut their fees for the sake of gathering assets. They are not necessarily interested in having billions of dollars under management. They simply want to do what they do best and make money. No single wrap program contains all the best managers, a point consultants should make when they talk to families. Performance Reporting. Basically, consultants can prove they are “on the line” for results and being candid with respect to advice if they are managing portfolios in accordance with investment plans; exercising discipline in rebalancing and adjusting portfolios as required to bring them into alignment with the plans; providing objective results via consolidated performance reports; documenting results in a comprehensive, segmented manner; meeting with clients at least quarterly; and communicating when portfolio changes are necessary. ■ Value. Often overlooked by consultants is the fact that the quarterly performance report is the single most tangible piece of information the client receives from the consultant. A performance report serves several purposes. It is the best tool for comparing actual performance to the original investment plan and for objectively assessing how the plan is performing. It also provides a framework for determining the factors driving portfolio performance. A good report helps to identify the need for manager changes or a re-evaluation of risk–return parameters as well as 62 • www.aimrpubs.org

highlighting new opportunities. Performance reports are thus more than numbers on a page; they are the most tangible product a consultant has to offer. A performance report is also the best tool for evaluating whether a consultant is adding any value with respect to discipline, costs and fees, tax management, and performance. Note that performance is last on the list. As I mentioned earlier, good performance is expected; otherwise, a consultant would not be hired in the first place. Furthermore, consultants cannot control the overall performance of the market, so focusing on performance sets false expectations as to what the consultant actually can control and diminishes the focus on those areas where they can add tangible value—taxes, costs and fees, access to managers, and overall service. ■ Format. A comprehensive performance report should analyze all of a client’s investment assets (regardless of custodian), not just those under the consultant’s advisement. Clients should also be able to view the performance of each entity, as well as the family office as a whole, because a family office typically has multiple investment entities, including trusts, family limited partnerships, and crossgenerational structures (i.e., the consultant should be able to deliver a performance report for any and all of the client’s “composite” portfolios). For example, one of our clients is a family office with 54 family members across three family branches and about 150 trust structures. We provide quarterly reports for each of them as well as for the aggregate family portfolio. In addition to being quarterly, reports should be as user-friendly as possible and follow a “hierarchy” that allows the investor to drill down to his or her desired level of information. We begin with a onepage portfolio summary that describes the portfolio value at the beginning of the quarter, contributions, withdrawals, fees incurred, portfolio value at the end of the quarter, and why the values changed. Many clients find that information to be sufficient (i.e., all they really want to know is how much they are worth, how much that value has changed, and why), but other clients want to know how each asset class performed or what their hedge fund exposure is. Some clients also want to know the performance of each asset style or which manager added the most value or caused the most drag. And although we do not encourage clients to ask for security-specific reporting, we can provide the data for each security that was bought or sold by each manager within the portfolio. By segmenting the performance report into these sections, our report allows clients to review areas in which they have interest, depending on their objectives, knowledge base, and level of sophistication. Note that in our performance report, the performance ©2003, AIMR®

Investment Consulting and Its Evolving Role in Advising Family Offices of individual managers is fairly low in the “drill down hierarchy.” This is intentional; it is typical for individual investors to focus on how specific managers are performing rather than on their portfolios as a whole. By placing this information fairly deep into the report, we are signaling (reinforced whenever we speak with the client) that the real focus should be on how the portfolio is performing, not on any single component within that portfolio. Most important, clients should be able to see not only how the investment plan is performing but also what is driving that performance. They should be able to compare the portfolio results with the plan that was originally developed to see the sources of any deviation and whether they are temporary or permanent. They need to decide whether any adjustments are needed to bring the portfolio back in line with the plan. Each manager’s performance should therefore be reported net of fees, relative to benchmarks, relative to peers, and relative to weighted-index results to determine how much alpha is being added. (After-tax performance is also important but is beyond the scope of this presentation.) Reconciliation of all cash inflows and outflows should be clear because clients often lose track of how much money they are withdrawing to finance their lifestyle. Finally, time- and dollarweighted returns should appear on the report for multiple time periods (quarterly, year-to-date, threeyear, five-year, and since inception).

Identifying the Right Consultant Identifying the right consultant is not easy for families. The problem is that most competent consultants offer similar core services—investment plans, asset allocation recommendations, portfolio construction, manager search and selection, and performance reporting (although not all offer consolidated reporting on all investments)—and generate similar investment results for a given level of risk over a reasonable time horizon. Thus, distinguishing one consultant from another by focusing solely on performance is difficult. One solution is to differentiate among consultants according to the services they offer. The best consultant for a specific family office is the one that offers the specialized services best matched to the family’s needs. For example, in addition to core investment management products, a family might want a specialist in concentrated wealth strategies, consolidated performance measurement, trust services, or alternative investments. Families should also consider the other services a consultant can provide, such as consolidating management fees, managing tax liability, or providing reduced investment mini©2003, AIMR®

mums and access to managers who would not be available otherwise. Once the family determines its specific needs and goals, it can make decisions accordingly and identify the most appropriate consultant. One of the most difficult challenges facing true consultants is in differentiating themselves and breaking through the “noise” of other, larger firms that claim to offer independent and objective advice but really do not. A family office should be highly wary of advertising campaigns from organizations offering a wide array of products and should clarify whether a consultant truly offers an open architecture platform, not just on the management side but also on the full spectrum of family office services— accounting, estate planning, alternative investments, life insurance, mortgage financing, and so forth. For example, one national firm claims to be an “open architecture provider with proprietary asset class products,” while another, also purporting to be an open architecture provider, claims that it will not use its own internal managers—a true case of being “reverse conflicted.” Service Differentiators. To distinguish among consultants, families should first consider which wealth management services they need, whether concentrated wealth strategies, access to collateralized lending, or creative financing strategies for such items as mortgages, luxury goods, collectibles, or travel expenses. Family office services, such as trust capabilities, bill paying, record keeping, and family governance, are also important considerations. Another distinguishing feature is whether the consultant is client centered or investment centered. A study by John Bowen, principal of the Creating Equity Group, found that during the recent market decline, investment-centered firms focused on investments, portfolio construction, manager selection, and performance improvement, whereas clientcentered firms focused on communication—making sure the client understood the investment plan, fostering a client perspective that extended beyond the current market cycle, and, in essence, doing serious “hand holding” as investors faced declining portfolios.9 Client-centered firms had a much higher client retention rate as well as greater success in attracting new clients than investment-centered firms. For firms that focus on investment results, this may be difficult to accept, but the data support it. Segmentation. Families should be aware of how the consulting business has been segmented into various (unofficial) niches, each of which offers a 9 This

survey can be accessed at www.cegworldwide.com.

www.aimrpubs.org • 63

Investment Counseling for Private Clients V slightly different array and level of services. Some firms follow the external chief investment officer model. The historical business of such firms was the institutional market, with later expansion into the high-net-worth market. These firms typically limit their services to investment management-related areas: investment policy statements, asset allocation, portfolio construction, manager search and selection, and performance reporting. Wealth advisory firms or wealth managers (such as Lydian) offer services that go beyond basic investment management consulting. They offer estate planning and insurance coordination, philanthropic planning coordination, wealth management (e.g., concentrated wealth strategies and access to loans), tax planning and management, and special projects management (concierge services). MFOs frequently build on a wealth management platform by offering administrative functions, such

64 • www.aimrpubs.org

as bill paying, cash management, tax management, and family governance.

Conclusion The family office market is rapidly evolving, with more family offices, more MFOs, more demands on providers of services, and more outsourcing. To keep pace and take advantage of the myriad opportunities, good consultants need to differentiate themselves in the industry through their objectivity, specialized services, product and service mix, and technological sophistication. Rather than focusing on performance, they should concentrate on providing a level of service commensurate with the demands of “insti-viduals.” If they fail to do this, the perception will remain that consultants lack value added and wattage, are not “on the line” for results, and are not candid in their advice.

©2003, AIMR®

Investment Consulting and Its Evolving Role in Advising Family Offices

Question and Answer Session Scott D. Welch Question: Why don’t you choose managers from the investment divisions of banks, brokerage firms, or insurance companies? Welch: The bottom line is that, typically, they don’t meet our selection criteria. It is nothing personal. We don’t care where a manager works or how he or she runs the business, as long as he or she produces good numbers. Question: ratio?

What is the Sortino

Welch: The Sortino ratio is a variation of the Sharpe ratio that was developed to differentiate between “good” and “bad” volatility. It attempts to adjust for the fact that what clients really don’t like is downside volatility—they typically don’t mind upside volatility at all. Question: What is your opinion of MyCFO? Welch: MyCFO was a wellfinanced concept that was perhaps ahead of its time. As an Internetbased business, however, the bursting of the Internet bubble doomed it. The firm hired a lot of people quickly who commanded large compensation packages and wanted major responsibilities, but family office services typically are fairly low margin. People don’t want to pay much for record keeping, bill paying, and tax preparation. The firm offered far more than those services, of course, but its cost structure relative to its ability to gather assets was simply not viable. When the firm was launched in

©2003, AIMR®

1999, a couple of our large tech clients were interested in it because they knew the founder, Jim Clark, or some of the other initial investors in the firm. But we never lost a client to MyCFO and never felt direct competition from it. Question: Does Lydian plan to ever report, as individual managers do, a composite performance number? Welch: We won’t report composite numbers, because each allocation is unique to the client. At the family office level, unlike the institutional level, allocations vary widely. New clients frequently ask what the “typical” performance of our portfolios was for the past year, but there is no typical performance; performance depends on the individual allocation. Instead, we determine an appropriate allocation for the client and then provide either historical data or references to existing clients with similar portfolios. Clients can judge for themselves how well the managers we have selected have done. Our clients are our best marketing assets. Question: Over a long period, do client-centered firms have better investment performance than investment-centered firms? Welch: I don’t know. The study I mentioned did not address performance, but I believe that any competent consulting firm will have good performance numbers. Some variation from year to year will occur among consultants and

advisors, but if two firms reach similar conclusions about the risk tolerance, goals, and objectives of a particular client, their performance should be similar. The point made in the study is that with clientcentered firms, clients have a better understanding of their overall investment plan and they’re more comfortable with the process, even in periods when their portfolio has underperformed. Question: How does Lydian provide investment performance data for noncustodied assets? Welch: I should make it clear that we don’t have a “master” custodian—our clients and managers can custody assets wherever they please. We have our preferences, but at last count, we were currently working with around 90 different custodial relationships. About half of our staff is dedicated to our technology reporting and reconciliation group. Our preference for certain custodians is based only on a couple of criteria. First, a custodian should offer favorable custody and trading fees. Second, it should have reliable electronic feeds into our performance system so that we can run reports as efficiently as possible. Of course, clients often have their own preferences. We accept that as part of the family office business, even if it entails obtaining duplicates of monthly brokerage statements and manually inputting them into our performance system.

www.aimrpubs.org • 65

The Importance of Proactive Compliance Thomas D. Giachetti Chairman, Securities Practice Group Stark & Stark Attorneys at Law Princeton, New Jersey

For financial advisors and investment managers, taking the proper precautions to ensure regulatory compliance should be viewed not as an inconvenience or disservice to clients but as an essential protection against punitive action. Firms need to understand the regulatory examination process, how to handle complaints, and how to make proper disclosures to their clients. By being proactive and taking the right steps today in three key areas—preparing for compliance examinations, averting complaints, and handling disclosures—firms can avoid serious problems in the future.

his presentation will focus on three topics: first, how to prepare for a compliance examination; second, how to avert customer and regulatory complaints; and third, how to ensure that disclosures are handled appropriately. My central point is that financial advisors and investment managers should be proactive in all of these areas; taking the right steps today will avoid serious problems in the future.

T

Preparing for a Regulatory Examination The scope of the examination process has definitely changed in the past two years. It has become much more comprehensive, requiring me to spend a great deal of time traveling throughout the country helping our clients at Stark & Stark prepare for examinations. Ten steps in particular are crucial to improving examination readiness. Obtain the Current U.S. SEC Examination Checklist. The first step is to get a copy of the current U.S. SEC examination checklist. Be warned that the format of the checklist may differ according to which branch office of the SEC will perform your firm’s examination. The Miami office uses a different format from the one the New York office uses, and so forth. Most of the questions, however, are the same. And on any given day, the questionnaire provided might be one they used four years ago rather than the questionnaire that they are using now. Unfortunately, the version a firm gets depends on what the SEC happens to send.

66 • www.aimrpubs.org

The SEC may give 48 hours notice of an examination or none at all. If a firm is close to a branch office, the SEC examiners are likely to provide no notice. Firms farther away will most likely get some notice (typically, on Thursday for an examination on Monday). Either way, a firm will have little time to prepare; thus, advance preparation is important. The best way to be ready is to know the questions on the checklist and the answers to those questions. A firm that is prepared will be able to provide better service to its clients during the examination, rather than focusing most of its energy on gathering information for the examiners. A prepared firm will also have more appropriate responses to the concerns of examiners. For example, unprepared firms often supply documentation that was not requested, which is a waste of everyone’s time. Review Prior Examination Deficiency Letters. Firms should review their prior examination deficiency letters. The examiners will start with previous deficiencies. The key is to make sure that any earlier deficiencies have been corrected. If a firm indicated in its deficiency response letter that it would take certain steps to ameliorate a deficiency, it would be wise to implement those changes before examiners return. A firm should have a good reason for any failure to make the promised reforms. Maintain Good Record Keeping. Good record keeping is essential. By necessity, large institutions generally understand the importance of good record keeping and assign that responsibility and oversight

©2003, AIMR®

The Importance of Proactive Compliance to specific professionals within the firm. Although smaller firms have fewer resources, being able to retrieve and produce documents for the examiners sends a message about how well prepared the firm is and how well its operations are run; it also improves the odds for an examination that is concluded in a timely fashion. A firm should not entrust record keeping to a part-time employee or an employee who does not have its complete confidence. When asked, firms need to be able to go straight to the right file and find the requested documents. Furthermore, files should have a consistent method of organization. The examiners will ask to see client files in a certain format (as indicated on the checklist). Typically, they will ask for about half a dozen, depending on the size of the firm. The number of files they will actually investigate can vary. If the first few files are organized without rhyme or reason, the examiners might review a few more. If the files are organized in the same manner and kept in the proper format and the examiners can readily understand them (because the firm has explained its recordkeeping system), the examiners most likely will not spend an entire day inspecting files. Insist on an Entrance Interview. A firm should insist on an entrance interview because it is an opportunity to tell the examiners what the firm does and, perhaps more important, does not do. Remember, at the time of examination, two years will have passed since a firm filed its last written disclosure statement (i.e., Part II and Schedule F) with the SEC. Thus, the examiners do not have a clue about what a firm has been doing during the past two years, except for the Form ADV filing in the Investment Advisor Registration Depository (IARD). Although the IARD was supposed to improve reporting, I admit that I still do not understand Form ADV. In short, examiners probably need a concise explanation of the nature of the firm. The most valuable information to give examiners is what the firm is not. This information narrows the scope of the examination. The goal is to remove certain potential conflicts of interest (such as soft dollars, referral fees, relationships with broker/dealers, and so forth) that examiners may be interested in and to dispel any preconceived notions they might have about the firm. Also, firms should tell examiners that the concerns raised by the SEC during the previous examination have been addressed in the manner the firm described in its response letter. Firms that take this step are likely to find that their examination will go much more quickly. They will earn the respect of the examiners and will be able to carry on with normal business during the period that the examination is ongoing. ©2003, AIMR®

Use a Point Person. Firms should name a point person responsible for all direct interaction with the examiners. A primary reason for having a point person is to prevent the examiners from using a divide-and-conquer strategy, asking the same question of three or four different people and possibly getting as many different answers. This does not occur because of problems in the firm but because different people have different levels of understanding about various practices and procedures. When I meet with client firms, I usually arrange a quick staff meeting to tell them how they should address the examiners and that they should not engage the examiners in conversation. At some point, people may be asked questions that are important in ways they do not understand. The key is to make it clear to the examiners during the entrance interview that the point person is the only one who will respond to substantive questions or requests for copies of any documents. The message should be conveyed in a clear, nonconfrontational manner. Effective use of a point person can give a firm some control over the examination process. Involve Legal Counsel in the Examination Process. Firms should have legal counsel involved in the process. We usually call our clients two or three times a day during an examination—to help them manage the process, to help them solve issues that have been raised by the examiners, and so forth. If an examiner says a firm should not be engaging in a certain activity, legal counsel can help the firm decide how to respond. If the examiners are correct, a prompt solution may be possible. Correct Deficiencies during the Examination Process. Firms should not wait until they get a deficiency letter; they should correct as many deficiencies as possible during the examination process. The challenge is finding out what are the deficiencies. Firms cannot always trust examiners to bring deficiencies to their attention. The point person should talk to the examiners two or three times a day: “How are things going? Anything we need to address? Anything you want to bring to my attention?” This communication is critical because they usually will tell the point person what they have found. Some issues can be addressed immediately, eliminating examiners’ concerns. They will bring other issues to the firm’s attention. After discussion with counsel, the firm can then respond to concerns or, in some cases, not respond— if the issues raised are not deemed to be legitimate. Maintain Document Control. Examiners are likely to request documentation in addition to what www.aimrpubs.org • 67

Investment Counseling for Private Clients V is indicated on the checklist. Firms should maintain a detailed list of all documents that are provided. Firms should not give examiners access to their files. The point is not to tell examiners that they do not have access to files but to avoid the problem. Bring them the documents they request. Make copies if they want copies. It is to a firm’s advantage to know what examiners are looking at and what they are concerned about. The point person will better understand the process and the issues raised and be able to address concerns as they arise. Insist on an Exit Interview. Firms should insist on an exit interview, with an eye toward knowing what examiners will probably identify in a deficiency letter. The point person should not be the only one involved. One or two other people should take copious notes because the point person needs to listen. When appropriate, the point person should emphasize that concerns raised during the examination have already been resolved by the firm. All firms get deficiency letters. The goal is to narrow the scope of the letter. If the process has been managed effectively, the exit interview should hold no surprises. If new issues are raised, at least the exit interview can reveal them in a timely manner. Firms may not be able to address all deficiencies during the examination. Some issues may be too broad or may involve enforcement. The SEC has changed in recent years. Matters that I formerly could resolve with a phone call or a quick meeting now may go straight to enforcement. For this reason, firms need to be prepared for this eventuality. Respond to Deficiency Letters Carefully and in a Timely Fashion. Few firms escape receiving a deficiency letter that raises multiple issues. If a letter cites 10 deficiencies, 3–5 of them may be inaccurate and others may be without merit. The examiners may have misconstrued some of the information provided to them or found it to be insufficient. Inexperienced examiners may not understand the particular way that a firm does business. In some cases, examiners do not understand the SEC’s own questions. When a firm’s leaders receive a deficiency letter, they should not respond hastily. They need to understand how to respond. They should not say they will do something that is contrary to their best interests if it is not required by statute or regulation. This requires a clear understanding of the statutes and regulations the examiners are quoting. Firms should develop a deep appreciation as to what their response may commit them to do. They should never put anything in writing until they understand its implications. 68 • www.aimrpubs.org

Firms should respond to a deficiency letter in a timely manner, usually within 30 days. An extension can be requested but usually should not be needed. Firms should also follow up on their response. The examiners will not send a letter indicating that everything is fine. After the examination is over and before sending a deficiency letter (or even after sending it), examiners may ask for additional information. Once in a while, they will raise questions about aspects of the deficiency letter response, an occurrence that has happened much more often in the past year or two than was the case five years ago.

Disaster Preparedness In addition to the steps I have discussed, firms should have disaster procedures and policies. Although disaster provisions are not required, they are recommended. This topic is now included in every SEC examination checklist, and the examiners want to see such policies. As a result of September 11, we now prepare a disaster policy for all our clients in support of their fiduciary duty to their clients. The most likely scenario is not a major calamity but a smaller, localized event. If firms lose power or their computer system or access to their facilities, they need a comprehensive policy to address the problems that will arise.

Averting Regulatory and Client Complaints Averting regulatory and client complaints is an issue of critical importance. Over the past year and a half, we have seen a tremendous influx of complaints against investment advisors. Broker/dealers have always been popular targets of complaints from retail customers, but never before have we seen these types of complaints against investment advisors. This change in circumstance is a result of the severity of recent losses. The legal theory behind the complaints seems to be: “I gave you money; you lost money. You now have to pay me money.” Whether the investment was consistent with client objectives seems to be beside the point. Still, firms can take certain compliance and disclosure actions to help avoid such complaints. When a client complains to the SEC, the typical result is an SEC request for information. The SEC will rarely take action against a firm on behalf of a disgruntled client of that firm unless the regulators believe the complaint rises to the enforcement level; the SEC is not the proper forum for dealing with damages from market losses. Nor will state examiners get involved in market-loss cases, although they may investigate whether a firm was acting in a manner contrary to its ©2003, AIMR®

The Importance of Proactive Compliance accepted procedures or policies, which then exacerbated or contributed to the client’s loss. The greatest concern arising from most marketloss complaints is a lawsuit. Right now, we are defending cases in probably two-dozen states throughout the country, ranging from small to multimillion-dollar requests for judgment. How do advisors deal with this threat? First, advisors should distinguish between macro and micro events. In the case of a macro event, such as a major terrorist attack, advisors should send letters to their clients explaining the firm’s response to the event and how the event could affect the clients’ investments. Micro events involve issues concerning a particular client or portfolio. A micro event should not be hidden away in the dark like a mushroom. The advisor should be proactive. If the problem is not addressed, a minor matter can turn into a liability claim. In dealing with a micro event, a letter will not suffice. The advisor needs to talk directly to the client—personally, not through a subordinate. Advisors inadvertently do certain things to get themselves in trouble. Typical areas of vulnerability arise from such actions as administering poorly defined risk-tolerance questionnaires, disclosing performance inadequately, and using canned documents. But if advisors understand the areas in which firms are vulnerable to complaints and take the necessary steps to safeguard themselves, many complaints can be avoided. Administer Well-Defined Risk-Tolerance Questionnaires. Many firms administer risktolerance questionnaires to their clients. These questionnaires often use terms that may not be well defined. For example, the questionnaire might refer to aggressive versus moderate risk. My advice to firms is to pinpoint what the different risk categories mean for their clients by tying them to particular indexes. For example, indicate that a certain risk level is commensurate with, say, the Russell 1000 Index and provide the index’s average annual return—the highs, the lows, and so forth—for a reasonably long history before asking the client to answer the question. Adequately Disclose Performance. Consider clients who, in 1999, were shown only the past four years of equity performance. Did the advisor do an adequate job in terms of disclosure? If a four-year window of performance was all that was shown to clients, should the advisor be held liable for losses? Were clients counseled properly? These are the sorts of issues that might be scrutinized during SEC examinations and are also often the cause of lawsuits. One of the most common ways for advisors to get them©2003, AIMR®

selves in trouble is by hanging themselves with their own documents. If a client responds to a questionnaire by saying on one page that he or she wants a 10–12 percent annual return but on another that the maximum acceptable loss is 5 percent a year, there is an obvious problem. Examiners or lawyers (and, ultimately, courts) might later take an interest in how this contradiction was addressed. Avoid Canned Documents. Firms should avoid “canned” documents purchased from a third party. Keeping such documents in a file as proof that a policy exists is not an effective procedure in the eyes of the SEC. Believe it or not, firms are not currently required to have a compliance manual, but canned compliance manuals are being marketed to firms as necessary tools. I am not a proponent of comprehensive manuals unless a firm is actually reading them and implementing them effectively. The worst scenario is to have a lawsuit expose that the firm is not doing half of the things it says it should do in its own manual. Understand the Importance of Client Intake Documents. Probably the most important document in the client’s file is the investment policy statement (IPS), investment profile, or client questionnaire. The professionals at the firm should read each client’s IPS and understand the obligations that are entailed. The IPS is not a rudimentary, routine document. In fact, having an IPS can give a firm a false sense of security. The IPS is sometimes the best place for the SEC to assess liability. For example, it is not unusual for the IPS to state that the firm will not change a client’s portfolio weights (or anything else) unless the request from the client is in writing but, in practice, changes frequently will be made to the portfolio in the absence of a written request in direct contradiction to the IPS. I developed a one-page letter for a firm to give its clients, a mini-IPS. It was essentially a form letter stating that the firm would allocate client assets among specified asset classes in accordance with the client’s designated investment objectives, and requesting that the client advise the firm of any change in his or her financial or personal situation or of a desire to impose new restrictions on the management of the account. Every document sent to clients should have such a disclosure in it, putting the onus on clients to inform the firm if something has changed that may require a reassessment of the investment management process, consulting arrangements, or financial planning recommendations. This approach prevents clients from claiming that the firm never contacted them because the firm will have ample documentation to refute the complaint. Attention to www.aimrpubs.org • 69

Investment Counseling for Private Clients V such details can provide considerable protection for a firm. The IPS is not a document that is cast in stone. Like compliance, maintenance of the IPS is an ongoing process. Firms should not get locked into an IPS that can cause problems for them later. These documents should be reviewed regularly. Another important document is the investment management agreement. Investment management agreements serve one purpose—to protect advisors and managers, not to give the client warm feelings about the advisor or manager. Out of 100 clients, 98 will sign the agreement. Clients have been signing brokerage agreements for years, and if they will sign those, they will sign virtually anything. The agreements should tell the client what the manager will do and will not do for them. If a firm does not offer certain services (such as financial planning, estate planning, and insurance planning), its investment management agreement should declare that the firm does not offer these services. Clients need to understand the scope of the engagement because if the client dies, the heirs may blame the investment manager, complaining that the proper estate or insurance planning was not in place. The manager needs to be able to document that clients were informed that such advice and services would not be provided. The financial planning agreement, also an important document to have in the client’s file, should say that the engagement does not include ongoing investment implementation, monitoring or reviewing, supervision, or management services. It should say that clients who want other services must contract them pursuant to the terms and conditions of a separate engagement, for which the client will pay a separate, additional fee. A good disclosure statement specifies what the firm does and does not do. It states that services are limited to certain activities and describes how they will be executed. The point is to protect the firm. Managers are being sued every day, and most of them cannot point to documents indicating that clients were notified of the issues or concerns that gave rise to the suit. A firm that receives a complaint should be able to respond that it gave notice not only in the initial agreement but also in disclosures made in invoices, correspondence, and newsletters. For those who truly hate to give disclosure statements to clients unless requested, the minimum is to give them a complete disclosure every year, not every two to four years. Documentation of regular disclosure is the best defense against charges of conflicts of interest or complaints about how the advisory business is run. Firms should not change their agreements and disclosures to accommodate the wishes of a few individuals. In fact, losing such clients is probably desirable because 70 • www.aimrpubs.org

they will become a problem sooner or later. Agreements are designed to protect advisors and managers, as are disclosures. Most people do not even read such documents and sign them without a quibble. When the market takes a turn for the worse, firms should remind clients that their assets have been allocated in a manner that is consistent with their own objectives, that the advisor does not know what the market will do in the future, and that the advisor is not the cause of market behavior. It is fair and appropriate to remind clients that, although the advisor may have helped them understand their objectives and allocate their assets properly, what goes up may eventually come down (or at least stop going up) and that the market is no exception. By the way, a primary function of advisors is to help clients define reasonable objectives. Clients who insist on unreasonable objectives should be avoided because they eventually will come back to bite the advisor. Recognize the Risks of Firm Newsletters. Many firms send newsletters to clients, and these newsletters pose serious risks. Unless a firm’s only business is writing a newsletter, it should avoid forward-looking statements. It should not make declarative statements about the future. No matter how great their expertise or their previous success, firms do not know what will happen next. And any newsletter or marketing letter should contain disclosures for everything. For example, I am defending a case in Wisconsin in which an advisor is being sued for $3.5 million. For two years, the firm told its clients that the technology sector would return to its near former levels. Month after month, the firm repeated this same refrain. I do not know how to defend the case, because the firm clearly told its clients to stay put. Review Liability Coverage Exclusions. Another important issue is not having sufficient insurance liability coverage. Firms need to read their insurance policies. Premiums are going sky-high, but coverage may be lacking. Firms should focus not on what is included in the coverage, which usually amounts to only a few lines of text, but on what is excluded. Most policies have several pages of exclusions. I review all of the policies for my clients because I want to know that they have appropriate coverage for the activities in which they engage. It is important to have an insurance carrier that respects an investment manager’s fiduciary duty. If a client loses money because the manager made a mistake, the manager has a fiduciary duty to inform the client. But such a disclosure may be treated by the carrier as a breach of the terms of the policy. Managers need to review their policies in this regard, question their agents, and know who is underwriting the policy. ©2003, AIMR®

The Importance of Proactive Compliance Most important, when a client makes a claim and demands to be compensated, the manager should take absolutely no action before reporting the claim to the insurance carrier. Managers should do nothing to jeopardize their coverage because, in its current condition, the insurance industry is eager to avoid liability. I recommend letting the manager’s legal counsel report the claim to the carrier. The policy may or may not allow the manager to appoint counsel, and this step will allow the manager to have greater influence on the decision regarding appointment. If the manager’s counsel has insurance defense experience and is willing to work at the rates the insurance carrier will pay, the manager’s appointment may be accepted. This is an important concession because the manager’s counsel is focused on the manager’s interests. Counsel appointed by the insurance company may be more focused on the carrier’s interests than the manager’s.

The third type says, “We purchase IPOs for our proprietary hedge funds only.” Regardless of the policy type to which a firm adheres, all firms need to have a definite IPO policy.

Firms should not use canned agreements or canned disclosure documents. Using canned documents is like trying to put a square peg in a round hole. Disclosure documents should specify how each service is handled, from asset allocation to block trades. They should not contain language that is inappropriate to the particular firm. Using a canned document may mean one day being held to a standard that the manager does not understand or one that is not applicable to the particular manager’s business. Most of a manager’s activities require disclosures, but some activities have fallen under stronger scrutiny in recent years, including IPOs, client-directed trades, websites, client-directed brokerage, and referrals.

Client-Directed Trades. Clients may call managers and ask them to buy shares of a particular stock. Most managers will oblige, but they should use an accommodation letter stipulating that the security was purchased as an accommodation, that the manager prefers that the client use a separate retail account for such acquisitions, and that the manager did not choose the security and will neither be responsible for its performance nor monitor the security as part of assets under management. During the bull market, many of my clients were inundated by calls from their clients begging them to buy certain stocks. When the stock prices later plunged, the managers were often blamed. The solution is to set up an accommodation or courtesy account for such transactions and to document how the matter was and will be handled. In such cases, clients may get preferential trading prices through the manager’s clearing firm, but the manager does not take responsibility for the investment. Similarly, clients often ask managers to handle stocks they already own or that they inherited. Managers should use an excluded assets letter that stipulates that such stocks are not part of assets under management and that the manager will be available to consult with the client but will not be responsible for making decisions regarding the holding. The goal is liability prevention. Firms will find it much more difficult to protect themselves after a problem has occurred.

IPOs. IPOs are a subject of particular interest to examiners. If a firm plans to get involved with IPOs, it should establish a policy before beginning to allocate IPOs to its clients. During the market bubble, many investment management firms were buying IPOs on a regular basis, including smaller firms that got access to the deals through their clearing arrangements. Instead of having a policy in writing or a procedure for how to allocate these IPOs, they were giving them to their best clients only and not rotating the allocation. Such discrimination is inappropriate and can get a firm in real trouble. Generally, there are three types of IPO policy; one of the three should be added to the client’s documents. The first type says, in effect, “We do not recommend IPOs, but if we get an allocation through our clearing arrangement and you ask for an allocation and we agree that it is suitable, you will get an allocation.” The second type says, “We recommend IPOs, and we have a procedure for allocating them.”

Websites. Many firms have their own websites, which are often purchased as canned products from outside providers. A website requires comprehensive disclosures. Clients may have access to such third-party links as economic calculators and newsletters from other sources. Disclosures should be all over the website. The disclosures should be accurate. For example, many websites have disclosures that say, in effect, “I cannot do business in your state unless I am registered.” This claim is not true. An SEC-registered advisor can have a website in any state and can have up to five clients in almost any state without filing notice with that state. Only four states have no de minimus exemption: Vermont, New Hampshire, Texas, and Louisiana. So, an advisor with even one client domiciled in any of these four states must file with the respective state. (Most states require filing if the firm maintains a place of business in the state.) All firms should regularly review their clients’ state

Disclosures

©2003, AIMR®

www.aimrpubs.org • 71

Investment Counseling for Private Clients V of residence to make sure that they have made all necessary regulatory filings. Client-Directed Brokerage. Client-directed brokerage is a serious topic. Some clients request that a manager direct the client’s brokerage to a particular broker. I am involved in an enforcement matter that centers on directed brokerage. Although the SEC has not yet provided concrete guidance regarding this issue, it will soon do so. In such cases, managers need to tell the client in writing—ideally in a disclosure statement in the investment management agreement—that the client has directed the manager to use a particular brokerage and that this direction may result in higher transaction fees or commissions than may have occurred with the clearing arrangement available through the manager. Referrals. Firms should beware of receiving referrals from broker/dealers. Directing clients to those broker/dealers is a problem. It is a problem even if the client directs the manager to use the broker/dealer. The SEC is saying that brokerage directed to the referral source is a significant conflict of interest. The firm must tell the client that using the broker that referred the client may not be in the client’s best interest. Firms need to negotiate best execution because if referred clients are paying up for brokerage and disclosures are inadequate, there is real potential for trouble. Solicitors or third-party referral agents need to understand that the rules differ from state to state. Rule 206(4)-3 under the Investment Advisers Act of 1940, “Cash Payments for Client Solicitations,” states

72 • www.aimrpubs.org

that if the manager is going to pay a referral fee, there must be a written agreement between the manager and the solicitor.1 At the time that the solicitor introduces the manager to the prospective client, the client must receive not only a disclosure statement from the manager but also a statement from the solicitor disclosing the solicitor’s relationship with the manager. The client must be told that if the client engages the manager, the solicitor will receive a certain fee. Even if the manager takes such precautions, the manager must ascertain whether the solicitor has any individual registration requirements for the state in which clients are solicited. Few people understand the scope of the solicitor rules. The rules differ significantly from state to state. If four examiners from a state are asked the same question about solicitor requirements, it would not be unusual for them to provide four different answers.

Conclusion A financial advisor or investment manager’s job is not to be a hero for clients but to allocate client assets in a manner that is consistent with designated investment objectives. Taking the proper precautions should be viewed not as an inconvenience or disservice to clients but as an essential protection against regulatory or legal action. Firms need to understand the regulatory examination process, how to handle complaints, and how to make proper disclosures to their clients. 1 Rule 206(4)-3, “Cash Payments for Client Solicitations,” can be accessed at www.sec.gov/rules/extra/iarules.htm#20643.

©2003, AIMR®