Investment Counseling for Private Clients

Foreword Private clients are one of the most demanding and one of the highest-growth segments in the investment management business. Managing their accounts takes time, expertise in a variety of disciplines-including psychology-and an eye for details. Still, managing private clients has and will continue to have its rewards as wealth is created and transferred between generations. Private clients may be more than one person-a husband and wife or a family spanning several generations. In order to identify and address successfully the needs of private clients, investment counselors must know who their clients are, determine their investment objectives, and identify unique circumstances and constraints that may influence the investment decision-making process, which quite often requires education but always requires the ability to communicate effectively. The excellent manager will build a lasting relationship with the client-by balancing client needs, goals, and even idiosyncrasies, by aligning those factors with the realities of the markets, by communicating technical information in ways each client will understand, and by staying in close communication with the client. Increased focus on after-tax performance has placed greater demand on investment managers. As a result, managers often find themselves working with professionals in accounting, estate planning,

and other areas to meet the clients' goals. Thus, the ability to work as part of a team and understand one's "position" on the team is critical. We are pleased to bring you this proceedings as part of AIMR's series of conferences and publications on the topic of managing accounts for private clients. We think you will find it full of insights, strategies, and tools to help you in building successful client relationships and in dealing with the myriad details and factors that must be taken into account when managing accounts for taxable investors. We are especially grateful to Eliot P. Williams, CFA, of The New England Guild, Inc., for serving as conference moderator. We also extend our thanks to the following authors for their contributions to this book: Jean L.P. Brunel, CFA, Morgan Guaranty Trust Company of New York; Maureen Busby Oster, CFA, MBa Advisors; R.B. Davidson III, Sanford C. Bernstein & Company, Inc.; Joanne M. Hill, Goldman, Sachs & Company; Nancy L. Jacob, Windermere Investment Associates; William R. Levy, Brown Brothers Harriman Trust Company of Pennsylvania; Christopher G. Luck, CFA, First Quadrant L.P.; Philip Penaloza, Merrill Lynch Asset Management; James M. Poterba, Massachusetts Institute of Technology; and Meir Starman, Leavey School of Business, Santa Clara University.

Katrina F. Sherrerd, CFA Senior Vice President Educational Products

i"

©Association for Investment Management and Research

Overview: Investment Counseling for Private Clients Terence E. Burns, CFA Vice President, Educational Products The enormous wealth created by a robust u.s. economy, technological advances, and the almost 300 percent increase in If.S, stock market indexes during the 1990scombined with demographic trends have made the management of taxable accounts for private clients one of the highest-growing business segments in investment counseling. Investment management organizations recognize the tremendous opportunity to increase fee income by attracting private clients; they also recognize that this client segment presents the greatest challenges for investment managers. Tax, legal, and client constraints must be taken into account when developing investment policies and portfolio strategies for taxable clients. Managers must also integrate tax and estate planning with the investment strategy and must often work with other professionals to accomplish client goals. Unlike accounts for institutional investors, taxable-private-client accounts are often composed of different generations-each with its own objectives. Thus, asset allocation strategies for private clients must address multiple objectives and multiple investment horizons. Perhaps the most important challenge is working with wealthy individuals and families who have unique needs, circumstances, and idiosyncrasies and who have differing levels of investment sophistication. Private clients can range from very sophisticated individuals who have great interest in portfolio strategies and investment results to the less knowledgeable and less interested clients. Because the investment management profession is a global business, managers and firms must also understand the complexities that global investors face and investment firms must organize themselves so that they can successfully serve global investors, whether the investors are based in the United States or elsewhere. Gaining the client's trust and confidence and building a lasting relationship with the client is an absolute necessity. Often, client education must go hand in hand with investment management as managers strive to align client expectations and views of risk with the realities of the marketplace. Communicating investment decisions and technical information to clients in ways that each client will understand is vital. At times, managers may have to get clients to look at investment decisions or tax "problems" from an entirely different perspective from what the clients are accustomed to.

©Association for Investment Management and Research

Asset allocation and portfolio construction play an integral part in achieving above-average performance, as does tax management. Therefore, investment managers need to understand how taxawareness or tax management can affect the overall asset allocation decision and portfolio construction process. Tax awareness requires using the total portfolio approach; that is, determining the appropriate asset allocation and constructing an overall portfolio that leads to the most rewarding, and most practical, investment strategies for each component of the portfolio. Investment managers can use a host of strategies that are designed to enhance returns for taxable clients and to eliminate, reduce, and delay the tax consequences for taxable equity and fixed-income portfolios. Derivatives can play an important role in that regard and also can be used to manage investment risks and to create customized strategies for taxable investors. Performance measurement and evaluation is not the final step in the investment decision-making process but, rather, an opportunity to provide feedback during an ongoing process about the success or failure of a given strategy. As clients increase their focus on after-tax investment results, investment managers will have to respond to client demands by comparing after-tax returns with after-tax benchmarks and implementing tax-efficient strategies that meet client objectives. John C. Bogle encourages investors to think along three dimensions: risk, return, and taxes. Taxes matter, and the old adage rings true: A little planning goes a long way in terms of maximizing after-tax wealth. Investment managers need to realize that proper planning and tax awareness throughout the investment management process will keep the U'.S, Internal Revenue Service (IRS) from appropriating a large portion of a private client's hard-earned wealth. This proceedings focuses on how taxes (income and estate) can influence the investment decision-making process for private clients and to highlight how important it is for investment counselors and their clients to think in an after-tax world. The authors draw on their extensive experience to provide a strong understanding of the key issues involved in dealing with private clients, to explain how to derive the optimal after-tax asset allocation and construct 1

Investment Counseling for Private Clients tax-efficient portfolios, and to discuss what strategies are available to investment managers for maximizing after-tax returns. Finally, the authors focus on what investment managers need to know about estate and tax planning and on the relatively new concept of after-tax performance evaluation.

Private clients are a demanding and highly diverse client segment. They demand a high level of service and want managers to pay extra special attention to their needs. A successful private-client relationship involves communication, knowing the client, and education. Effective communication allows managers to identify correctly each client's investment objectives, constraints, and unique circumstances; to develop an appropriate investment policy; to determine the optimal asset allocation; and to construct the portfolio that meets the client's after-tax investment objectives. Asking the right questions, having good listening skills, and providing feedback are invaluable in this regard, because they help managers become aware of any unique constraints that may influence asset allocation, portfolio strategy, and estate and tax planning. Through close communication, a manager will come to know his or her clients and his or her own limitations. At times, the manager will act as teacher, educating the client about investment alternatives, strategies, or taxes. At other times, the manager will act as pupil, learning about income tax regulations and estate planning. And sometimes, the manager will be part of a team of investment professionals, acting either as "coach" or "player." Regardless of what position the manager plays on the team, the manager must know what role he or she plays and be a team player for the client to come out a winner. Maureen Busby Oster, CFA, examines the challenges in dealing with private clients. Taxable-privateaccount managers face different challenges from those found on the institutional side. On the institutional side is a formal decision-making process; an investment committee or board of trustees is responsible for overall investment decisions. In dealing with private clients, the investment manager has to determine who is the ultimate decision maker-especially in the case of husband and wife teams or whole families. Busby Oster emphasizes that the manager must know his or her client and identify the person with "veto power" when it comes to investment decisions. If managers know what types of clients they have, then they can figure out the best ways to deal with the clients. Clients can have high incomes and/or high net worth, be delegators or controllers, and be risk takers or risk avoiders. Certain types of clients are easier to deal with than others. For example, controllers may be

reluctant to relinquish total control of their portfolios to portfolio managers. The type of relationship the manager has with each client can also vary. Depending on how the client or prospect came to the firm, the manager might have to work harder to gain that client's confidence and ultimately the account. Busby Oster points out the importance of knowing each person's role when working on a team of professionals. Finally, she discusses the benefits of having a written investment policy statement, the best ways to communicate with clients, how to resolve conflicting agendas with family clients, and how to present results to clients on an after-tax basis. Philip Penaloza discusses the challenges and complexities of dealing with"global clients" (i.e.,nonU'S, clients with global investments). Global clients are similar to domestic clients; they expect a high level of service, additional handholding from their portfolio managers, and low investment management fees. Although global clients may be "controllers," they are highly flexible and open to creative investment strategies and tax and estate-planning solutions. Penaloza warns that global clients have complex needs. They are often looking for a safe haven for their assets, secrecy, portability of asset control, tax benefits, and a stable regulatory regime. The reward for meeting the complex needs of, and providing value-added services to, global clients is a satisfying long-term relationship. Managers up for the challenge of meeting the tangible and intangible needs of global clients will be rewarded and increase assets under management. Understanding the way individual investors think about risk and taxes is important in helping to align client expectations with the reality of the market and the tax code. Many people view risk as the possibility of losing money rather than as the standard deviation of returns, and most rational people believe they pay much more than their fair share in taxes. Meir Statman examines the psychology of risk and taxes from a behavioral finance perspective. He suggests that normal people are as averse to risk as they are to taxes and that client views about risks and taxes are influenced by framing and season of the year. Investors see losses as less painful in December than in other months because the losses become a tax deduction. People always feel cheated by the IRSbut especially so in December because the tax bite materializes. In order to show that people are averse to losses rather than to risk, Statman demonstrates that people reach different decisions when identical risky choices are framed differently. That finding has important implications for investment managers because an investor's perception of risk influences his or her investment strategy. Therefore, investment managers must encourage their clients to correctly

2

©Association for Investment Management and Research

Counseling Private Clients

Overview frame the concepts of risk and taxes to avoid suboptimal investment strategies.

Asset Allocation and Portfolio Construction Extensive research has focused on the asset allocation decision and portfolio construction process. In their seminal article in 1986, Brinson, Hood, and Beebower found that slightly more than 95 percent of the variance in returns from quarter to quarter during the 1974-83 period was explained by the pretax asset allocation decision. 1 That research focused on the asset allocation decision for nontaxable pension funds. Recent research has examined how taxes influence the asset allocation and portfolio construction process because private clients are focusing more and more on after-tax results. For example, the first volume of the Journal ofPrivate Portfolio Management contains 1.5 articles that analyze how taxes influence some aspect of asset allocation and portfolio construction. Income and estate tax expenses are far greater than investment advisory and custodian fees. Thus, Nancy Jacob adds another dimension-taxes-to mean-variance optimization and analyzes how the asset allocation decision changes in an after-tax world. Many wealthy private clients have large blocks of low-cost-basis stock, which may have been passed down from previous generations or may have been the reward for hard work at a startup company. Think about how many private clients have, and will have, large blocks of low-cost-basis stock in technology companies as a result of technological advances in computers, the Internet, and e-commerce. How will those investors diversify their portfolios? Traditional mean-variance analysis would sell off such positions and create huge capital gains taxes. Jacob analyzes how those investors can diversify their portfolios and minimize taxes. After-tax asset allocation optimizes the portfolio by taking into account onetime gains, ongoing tax impacts, unique constraints, the client's tax status, and multiperiod investment horizons. The result is a mean-variance tax-efficient portfolio that increases tax efficiency, saves tax dollars, and produces different portfolios from those generated by conventional optimization programs. Depending on the client's objectives, constraints, and investment horizon, the portfolio might even retain some of the low-cost-basis stock. lGary P. Brinson, L Randolph Hood, and Gilbert 1. Beebower, "Determinants of Portfolio Performance," Financial Analysts Journal (July/August 1986):39-44.

©Association for Investment Management and Research

Jean Brunel, CFA, believes that tax awareness is the key to portfolio construction and warns managers that tax-oblivious portfolio construction for private clients can lead to underperformance. Brunel discusses the concept of tax-aware portfolio construction using the total portfolio approach. Elements of the total portfolio approach include making optimal use of a loss regardless of who manages that particular asset, optimizing the asset location between a client's taxable and tax-exempt accounts, and synchronizing asset-class and security decisions. Brunel goes on to explain that simultaneous rather than sequential optimization produces the most tax-efficient portfolio. That approach involves avoiding asset-class-driven solutions, may require a coach to maintain the tax awareness of a multiple-manager portfolio, and requires minimizing tax friction. Managers need to realize that two kinds of transactions can add value for private clients: alpha generating, which is the overall goal of all securities transactions, and alpha enabling, which is driven by tax awareness. A critical distinction between the two is that alpha-enabling transactions create after-tax wealth. Brunel provides guiding principles that portfolio managers should follow when constructing portfolios, but overall, the key is not to let taxes drive the portfolio construction process but to incorporate taxes in the process.

Portfolio Strategy and Taxes Maximization of after-tax returns for equity and fixed-income investors is attainable in two ways: Investors can focus on generating enough excess return to more than pay for the taxes, or they can focus on minimizing the tax burden. The majority of managers cannot generate sufficient alpha to cover the taxes they generate, but they can try to minimize the tax bite, which is the basic idea behind tax-efficient investing. Christopher Luck, CFA, believes that taxsensitive investing can benefit taxable investors and examines tax-efficient strategies for equity investors (individuals and corporations). For equity investors, the key driver is turnover. Luck shows the connection between turnover and taxes by comparing the returns of passive and active managers. Luck shows how, given a long enough investment horizon, a lowturnover, passive strategy will outperform a highturnover, zero-alpha strategy. The big disappointment is that earning sufficient alpha to cover taxes creates such a tremendous hurdle for active managers that few managers can consistently jump the hurdle. Luck discusses how a high-alpha strategy used by highly skilled hedge fund managers, a relatively low-turnover strategy, and a tax-sensitive strategy (such as avoiding dividend-paying stocks)

3

InvestmentCounseling for Private Clients can help individual investors deal with the relationship between turnover and taxes. Finally, he offers straightforward advice on how to reduce the largest cost of managing a portfolio-taxes. For high-net-worth investors, R.B. Davidson III explores a variety of tax-efficient fixed-income strategies, which generally involve using a combination of taxable and tax-exempt securities. Davidson believes that active fixed-income portfolio managers can add value for taxable investors; the key is to know when to recognize gains and realize losses. In a rising-interest-rate environment, harvesting losses and reinvesting in higher-yielding municipal bonds can add value over a buy-and-held strategy. How much value is added depends on the maturity of the portfolio and the investment horizon, but tax management of municipal bond portfolios can add up to 121 basis points to overall performance. Managers can also add value by changing the portfolio's allocation among bond sectors as the yield ratio changes. Nevertheless, managers need to consider the client's tax bracket; clients subject to the top tax bracket rather than the alternative minimum tax will have different sector allocations and earn different after-tax premiums. Finally, Davidson explains how managers need to balance tax implications with expected returns in a disciplined framework that factors in transaction costs. Managers can use derivatives to manage investment risks and to create customized strategies for private clients. Joanne Hill discusses a variety of derivative strategies that can help managers create tax-efficient portfolios. Before using these instruments, managers must fully understand how tax laws affect the use of derivatives. Managers need to understand whether derivative transactions constitute a straddle, whether the premium received or paid for a stock option transaction is considered a capital item, how and when to use qualified covered calls and Section 1256 contracts, and how to work around the wash-sale rule. Hill explains how derivatives can be used to alter the general characteristics of any portfolio, make asset allocation shifts, create index exposure, or implement synthetic index strategies. Portfolio managers can also use derivatives to convert short-term gains to long-term gains, manage market risk, and harvest tax losses. The key to maximizing after-tax returns and creating tax-efficient portfolios is to understand the limitations of derivatives, the costs and benefits of each strategy, and the applicable tax rules.

the IRS. An integrated approach to wealth management often requires guidance from experts who have a detailed understanding of estate- and tax-planning tools. William Levy points out that the federal estate tax rate is a whopping 50 percent (after the unified credit exemption is used up) and discusses a variety of estate-planning tools to reduce death taxes and provide value-added services for private clients. Levy strongly encourages lifetime gifting because it is the easiest way to reduce estate taxes and avoid having the appreciation of assets taxed. If the client is able, aggressive gifting of assets in the form of annual exclusion gifts, tuition and health care gifts, or splitinterest gifts is a first line of defense against estate and gift taxes and a great way to transfer wealth while the client is still living. Levy also explains how making taxable gifts and making sure that each spouse has sufficient assets in his or her name to meet the unified credit equivalency can reduce estate and transfer taxes. Estate and tax planning may be beyond the portfolio manager's expertise, so Levy recommends bringing in qualified experts to make sure that the legal details are handled correctly. Failure to do so could put the client at risk and create serious tax problems. Levy also encourages private clients to make use of charitable deductions, which have great potential to reduce taxes-especially when they are funded with retirement plan assets.

After-Tax Benchmarks

Integrating investment objectives with estate and tax planning is critical to preserving wealth for future generations and keeping it out of the open arms of

Monitoring performance is an important step in the portfolio management process and provides an opportunity to critically evaluate a manager's performance relative to the predetermined benchmark. As more and more private clients demand presentations of after-tax investment performance, portfolio managers will be forced to develop customized after-tax performance benchmarks. James Poterba focuses on how taxes interact with performance management and how to measure aftertax returns. Private clients are becoming more tax conscious and want to know how portfolio managers' decisions influence the amount of taxes clients pay. Therefore, managers must understand how portfolio turnover, cash inflows versus withdrawals, and dividend yield versus capital gains affect the tax efficiency of portfolios. Private clients have unique income and estate tax constraints, so portfolio managers need to understand how individual federal tax rates, estateplanning issues, marginal tax rates, capital gains taxes, and special circumstances should influence portfolio management decisions-especially the realization of capital gains. Unrealized capital gains pose a special challenge for portfolio managers who are measuring after-tax performance because they represent a future contin-

4

©Association for Investment Management and Research

Estate and Tax Planning

Overview

gent tax liability. Poterba presents an algorithm-the accrual equivalent tax rate-that can take into account different future tax rates and demonstrates the value of deferring capital gains over different investment horizons. This algorithm provides portfolio managers with a way of comparing the after-tax value of unrealized gains in a retirement account, for example, with realized gains from assets invested in a taxable account. The key to maximizing after-tax performance is the integration of tax planning and portfolio management; a separate tax-management "overlay" is suboptimal for the client. Communication between the portfolio manager and tax accountant is necessary in this regard.

Conclusion Managing portfolios for private clients will remain a high-growth business segment for portfolio managers

©Association for Investment Management and Research

in the 21st century, especially as new wealth is created and existing wealth is transferred from one generation to the next. Managing private clients' portfolios presents interesting challenges for portfolio managers, but it also presents an opportunity for significant rewards for those who can add value on an after-tax basis. In this proceedings, the authors bring together a unique blend of experience, skills, and foresight to help portfolio managers deal with the challenges associated with managing taxable accounts. The authors highlight the importance of knowing the client, identifying the client's objectives and constraints, determining an optimal after-tax asset allocation, constructing a tax-aware portfolio, implementing taxefficient strategies, and presenting performance using an after-tax return framework. To be successful, portfolio managers must be tax conscious throughout the investment decision-making process.

5

orking with Taxable Clients Maureen Busby Oster, CFA President MBO Advisors

Investment counseling for taxable investors is a rewarding but challenging experience. Clients are diverse; they make decisions differently and demand different levels of understanding with respect to the performance of their investments. So, managers need to understand their clients in terms of who makes decisions, types of clients, and the client relationships they have. High-net-worth taxable clients, in particular, require extra managerial attention to tax issues and reporting investment results.

hen working with clients-individuals, husband and wife teams, whole families, or their advisors or office managers-the first concern is to know your client. Some investment advisors deal primarily with taxable wealthy families, and others deal largely with individuals who have substantial wealth in tax-deferred rollover individual retirement accounts (IRAs). This presentation provides a clientfocused overview of the concerns that are important for both groups.

W

Decision Making

between you, because sharing sometimes requires a mediator." The second thing I do, particularly when dealing with what I call "recently liquefied wealth," is give them articles and recommend books. The last thing I do is find out if anyone in the family has veto power. Often, particularly in multigeneration families, there is a key individual whose opinion carries more weight than others'. If so, I want to know that information right away.

Types of Clients

An important first task for advisors is to identify who makes the decisions in the client group and how the decisions are made. Sitting down with a new client or a potential client the marketing person has brought in is an interesting experience. For example, the client could be a husband and wife team who have built a business together and have just sold it. They are 50 plus years old; they have children; they have a lot of cash and a lot of low-cost-basis stock because the sale was a combination cash-and-stock deal, As you talk to them a bit about risk, return, and investment objectives, you discover that she wants to build a home in Arizona, take a cruise, and do some shopping, then she will think about other matters. However, he wants to build a hunting lodge in the mountains and put a chunk of money aside for a new business venture. What do you do? The first thing I do in such a situation is say, "Here are two questionnaires, one for each of you to complete. Take them home, think about them, answer them separately, and then we will sit down and talk about them together. Do not share them

Big differences exist between the taxable client who has a high income and the taxable client who has a high net worth. Some people who live like the wealthy do not have high net worth; they have high debt. Some very high-net-worth people do not live as if they are at an rich. As for the client's assets, the advisor needs to know whether the money is recently liquefied money or not. Is it diversifiable? Somebody who just sold the family business for a combination of cash and stock has low-cost-basis stock that mayor may not be diversifiable. Advisors must be accustomed to sitting with the tax attorney, going through how long a person is likely to live, and considering the trade-offs in terms of paying the tax now or later. Understanding of the client's psychological profile is important. First, who created the wealth? Whether it is the client, the just-previous generation, or an earlier generation is important. For instance, I live in Milwaukee, where some of the old industrial businesses were created 150 years ago. So, in many cases, the wealth builder lived several generations ago. Wealth builders have a very different psycho-

6

©Association for Investment Management and Research

Working with Taxable Clients logical profile from those who have inherited wealth. Early work on client profiling was presented at an Institute of Chartered Financial Analysts confer1 ence in 1986. This work, a 13-year study of a large cross-section of professionals, concluded that these people could be classified along four dimensions: active to passive, risk taker to risk avoider, wealth builder to income seeker, and controller to delegator. Subsequent research has found that entrepreneurs are risk takers and controllers. They will take big risks with something over which they have control. So, private clients who built their own businesses are likely to be extremely resistant to giving up financial risk management to someone else. Corporate executives delegate all day long; as clients they tend to be delegators rather than controllers. Certified public accountants at the Big Five accounting firms are different from CPAs at regional firms. Those at the Big Five are similar in profile to surgeons; surgeons take big risks but want to control risk. Those at regional firms are more apt to be risk avoiders. Athletes and entertainers come in all types of profiles. One of the groups MBO Advisors works with is the National Football League, and we have found that some football players are great investors and entrepreneurs whereas others have managers and delegate everything.

Client Relationships The first aspect to consider when defining your relationship with a client is how the client came to your firm. In particular, did you know the client previously, or is the client a referral? Advisors may play golf or tennis with potential clients or take their children to the same soccer games. Working with someone who was a friend before becoming a client is different from working as a hired hand for a stranger. Some of my clients are also people with whom I interact socially. So, I work hard to not mix business discussions with the social discussions. When we need to have a business discussion, we have a business meeting. Consider also whether the referral came from an existing client, the client's attorney, or the client's tax advisor. Depending on how you or your firm acquired the client, you might use a different approach to gain the client's trust and confidence and to establish a level of comfort with the client. Next, you want to be clear about what you are being retained to do. For example, were you retained IMarilyn MacGruder Barnewall, "Psychological Characteristics of the Individual Investor," Asset Allocation for the Individual Investor, edited by William G. Drams (Charlottesville, VA: Institute of Chartered Financial Analysts, 1987).

©Association for Investment Management and Research

to advise or to manage assets? If you were retained to coach, advise, and counsel, your job is not to pick stocks, evaluate individual investments, or decide asset allocation. The portfolio manager who interacts with a high-net-worth client may think differently from the advisor, or coach, about how information is to be conveyed to the client, particularly a high-networth client whose money was made in a nonfinancial business and who has a hired financial advisor. Clients may not understand everything a sophisticated portfolio manager is talking about. For example, from the media today, clients have heard all the bad aspects of derivatives and hedge funds. As a result, some clients may want to have nothing to do with those instruments. They also may not want to bother with extensive, detailed explanations. Considering these aspects is important when you are preparing presentations in order to carry out what you are hired to do. It is important to know with whom you will be working and to whom you will be explaining (and occasionally defending) results. Finally, the advisor's role and responsibilities may depend on other advisors. Are you on your own or part of a team that, perhaps, includes accountants, lawyers, and a family office manager? If you are on a team, the issues of roles and coordination of activities arise. Is there a coach, as well as a portfolio manager, on the team? Have you been hired to deal with the whole portfolio or only a portion of it? Is somebody providing tax analysis? Is a small-capitalization/ value stock manager needed, and if so, do you have one? The team may need to include other specialized managers. For example, today, advisors are making more and more use of the concept of a "completion fund" for the very wealthy. If a client has a pension fund and other assets, a completion fund can be used to fill in the gaps to reach the desired investment goal. Managers are also using hedge funds because they provide the ability to defer taxes. In addition, some members of the client family may not be U.S. residents, which raises a new set of issues. In these situations, it may be desirable to have specialized expertise on the team. An aspect of relating to other advisors that should be considered is whether the other advisors are from your firm or other firms. If the other managers are from your firm, coordination will, of course, be relatively easy. But needed skills, such as a smallcap/value manager, may be easier to find if several firms are involved. Assuring that the team interacts on a regular basis will help the client and the team members a great deal. Financial advisors need to clearly understand their individual roles when they are working as part of a team. For example, I am in a situation where I am

7

Investment Counseling for PrivateClients a member of one team and the captain of another. I am a member of the team with the lawyers, the accountants, and the family office manager, but I am the leader of the team that deals with multiple portfolio managers who have different assignments.

mation that comes out of this process. In that way, all the managers in a multimanager situation know from the beginning what their rules are and what everyone else's rules are. They all receive the same documents and information, including a copy of the full investment policy statement. They know whether they are going to have to produce income or pay taxes out of their portions of the portfolio or whether income and taxes are going to be paid elsewhere.

Investment Policy Statement. The most effective way to build a good long-term relationship with taxable clients is to have a written investment policy statement. Advisors need to invest the time up front to find out about their clients and educate their clients. These tasks often involve getting them to respond to questions for which the answers cannot be quantified-"How do you feel about ..."-and incorporating those answers in the policy statement. The process and the policy statement have the following elements: • Establishing goals and objectives. Goals may vary among the members of a client family. For example, as noted previously, when dealing with a husband and wife, the advisor should have them respond separately to the questionnaire about investment policy and then sit down together with them to reconcile differences in the answers. • Outlining asset allocation targets. • Discussing the client's concept of risk. Advisors need to establish, with their clients, a definition of risk. Investment professionals talk about risk from an academic standpoint-risk as standard deviation of returns-which leaves people who are not familiar with the investment business glassy-eyed. To the client, risk can mean losing money, missing the portfolio return objective, or underperforming a benchmark. • Establishing benchmarks for the evaluation of portfolio performance. • Delineating the investment time frame. Advisors will want to have periodic (at least annual) reviews of the investment policy statement with the client. The purpose of the annual review is to remind clients of what the advisor is doing, the advisor's role, and what the original arrangement was. When building and maintaining wealth for a multigeneration family, what happens at the moment is only a blip on the screen. The family or family foundation may continue forever. • Defining the spending policy. Specifying the spending policy in the investment policy statement is a good idea: It minimizes the possibility of misunderstanding in the future, it identifies what the cash demands will be, and so on. • Describing the annual (or more often) review process. All the parties should receive copies of the infer-

Resolving Conflicting Agendas. In addition to differences among family members about the amount of detail to report, an advisor should keep the dynamics of multigeneration families in mind when dealing with family clients. Increasingly today, as businesses are bought and sold more often than in the past and people are living longer, advisors are dealing with multiple generations. Generational conflicts arise because of the different time frames of the family members. Advisors need to find a way to deal with the conflicts between the generations. Anybody who has worked in a trust department has heard about marital and residual trusts and about the tension between income beneficiaries and principal beneficiaries. I am the oldest of II, so when the family attorney sat down with us, she had to consider the concerns of multiple generations. The first time she met with all 11, she got up and left the room and then came back and sat down again because she needed

8

©Association for Investment Management and Research

Communicating with Clients. Remember that communications come in the form of written reports, phone calls, and meetings. One of the things that helps immensely in dealing with clients is to update them periodically by phone. One of the first considerations in reports and other communications is whether the client wants to know only the bottom line or wants every detail, perhaps in the form of an executive summary followed by a detailed report. Are you reporting to a family, an individual client, or an office manager? For example, if the client is a family, you should be aware that some family members will want a minimum of information and some will want all the details. Both types will be present at the annual family meeting, so you have to present the material in such a way that all types can see or hear what they want to know. For example, I worked at one time for the Johnson Wax family; they had an annual family meeting of many generations. Their most significant net worth is Johnson Wax, which is a nonpublic company, so they had time devoted to presentations by the operating businesses as well as the money managers who were building a new business for the family. As money managers, we focused on explaining what we were doing to build new business.

Working with Taxable Clients time to digest the variety of ages, not to mention the needs and wants of the variety we presented. Dealing with generational conflict may be delicate, particularly when conflicts arise between who is getting favored and when-for example, when an advisor is working with a wealthy family in which the surviving spouse is not the mother or father of the children. As noted previously, an approach that is increasingly being used is to set spending policies, just as an endowment fund would. Unit trusts also help defuse conflicts in this area because the normal distributions are based on total return, not simply current income. This may be a good approach to overcome the issue of maximizing current income versus maximizing future income (or protecting future purchasing power) through long-term growth of capital.

Tax Issues One of the major issues that arises for individual clients is that some of their wealth may be in taxdeferred accounts and some in taxable vehicles. With pension portability what it is today, multimillion dollar rollover lRAs are not unusual. So, although the person is "taxable," the assets an advisor is managing may not be. Asset Allocation. For an advisor who is dealing with both taxable and tax-deferred pools of assets, the first question is where to put the high-risk assets versus where to put the low-risk assets. Should the bonds, core stocks, and high-turnover assets go in the tax-deferred account? Because of changing spreads on the tax-exempt yield on municipal bonds in a taxable account versus the tax deferral of a retirement account, what types of bonds are owned and where they are best placed should be reviewed. On the one hand, the long-lived, low-turnover assets, such as limited partnerships, should go in the taxable account because such assets use appraisal accounting and until the companies are either taken public or the whole fund is liquidated, they generate no tax event. On the other hand, high-turnover assets should go in the tax-deferred account. An example is small-cap/ value stocks: The style discipline for a small-cap/ value manager requires selling a stock that has grown beyond the definition of small cap, which creates frequent turnover. Small-cap/value managers may also have losers, however, which can add more value on the tax-management side when those losses can be realized and utilized to offset other gains. Estate Taxes. Advisors who are acting as general coaches for their wealthy taxable clients need to have at least a working knowledge of estate tax issues

©Association for Investment Management and Research

so that they are not caught off guard when estate planning is being discussed. For example, both a grantor annuity trust and a charitable remainder unit trust have a charity as the terminal beneficiary, but the annuity trust specifies a dollar amount to be distributed annually whereas the charitable remainder unit trust determines the annual distribution based on a percentage of each year's beginning value. Just remember to rely on the tax and legal experts on the team for advice. Charitable Giving. The difference between gift taxes and estate taxes is obvious when the calculations are done but may not be obvious otherwise. For example, if I give my daughter $10,000 this year, I have no gift tax to pay because that amount is within the limits. If I give her $250,000, I must pay a 55 percent tax on the gift, but at that point, the basis is written up, the tax is paid, and the issue is closed. If the money is left to my daughter in my estate, my estate pays taxes based on the total value of my estate, including what was used to pay the taxes. Giving through an estate is simply much more expensive, so when possible, making a gift earlier makes a lot of sense. In their role as coach, advisors should be aware of the issues that surround charitable giving. Wealthy people are sometimes interested in making some sort of permanent gift. The decision between making the gift through a community trust (or a private foundation) or a supporting organization has consequences. The donor and/or the donor's family may control a private foundation, whereas supporting organizations and "donor advised" charitable gift trusts may not be controlled by the donor. However, a private foundation does not qualify for tax treatment that is as generous as is the case with a gift to a public charity through a charitable gift trust or a supporting organization. In addition, a private foundation is subject to an excise tax on net investment income, but neither supporting organizations nor charitable gift trusts pay excise taxes. Because of excise taxes on the private foundation and the tax deductibility of grants, gift givers get a bigger bang for the buck by giving to a community foundation, and community foundations have become much more active in conveying that information than they were in the past. Giving through supporting organizations provides some of the benefits of a community trust, such as avoiding the excise tax, but still allows the donor the opportunity to have some control over how the money is given away. One of the most interesting challenges in coaching high-net-worth clients, particularly those who find themselves wealthier than they ever imagined, involves helping them be charitable. (Another inter-

9

Working with Taxable Clients •

Calculation of after-tax returns for tax-exempt bonds must include amortization and accretion of premiums or discounts. • Cash-basis accounting must be used if required by law. The AIMR-PPS standards recommend that benchmark returns be calculated using the actual turnover in the benchmark index, if available; otherwise, an approximation is acceptable. Required disclosures. In accordance with the AIMR-PPS standards, investment firms must also make the following disclosures: • For composites of taxable portfolios, disclose the composite assets as a percentage of total assets in taxable portfolios (including nondiscretionary assets) managed according to the same strategy as the client. • If performance results are presented after taxes, disclose the tax rate assumptions. • If adjustments are made for nondiscretionary cash withdrawals, disclose both client average and manager average performance. Presentation ofresults. If returns are presented after taxes, client-specific tax rates may be used for each portfolio but composite performance should be based on the same tax rate for an clients in the composite. In addition, the AIMR-PPS recommendations are that the following presentations should be made for composites:

©Association for Investment Management and Research

• • • o

• •

•

beginning and ending market values, contributions and withdrawals, beginning and ending unrealized capital gains, realized short-term and long-term capital gains, taxable income and tax-exempt income, the accounting convention used for the treatment of realized capital gains (for example, highest cost, average cost, lowest cost, first in/first out, or last in/first out), and the method or source for computing the after-tax benchmark return if a return is shown.

Conclusion Advising people is a "relationship business." No recipe exists for dealing with individual persons or families and the members of those families. Each situation is unique, and advisors will miss something if they do not concentrate on each client. Still, investment counselors should know who is the ultimate decision maker for husband-and-wife or family clients, understand what type of client and client relationships they have, and most importantly, understand the tax issues for taxable investors. In addition, the investment manager must have a detailed understanding of how to present performance results to clients. Finally, investment managers must realize that the stewardship of wealth is an ongoing process.

11

Investment Counseling for Private Clients

Question and Answer Session Maureen Busby Oster Question: How do you determine who is the decision maker in a client group and who has veto power? Busby Oster: You cannot find out in one meeting; it takes multiple meetings. If you have been brought in by the attorney or the accountant, he or she may have insights because of prior experience with the client. If you've been brought in by someone involved in the investment banking deal that liquefied a family business, that person may also have some insights to share. I use the process of the questionnaire and meetings to find out who makes decisions and who may veto decisions. Question: Have you found that certain types of client communications work better than others? Busby Oster: I make a conscious effort to use analogies the particular client will understand. For example, most clients are parents, so I've found over the years that using analogies to parenting is helpful. Sports analogies do not work well with some clients. We try to get down to the basic issues. For example, for a couple who worked their whole lives in their business and just recently liquefied their holdings by selling the business for $50 million, when we're dealing with the issues of risk and return, we have to remember that their company was not valued every day. That couple did not receive a report once a month on what the resale value was. Now, they're going to get monthly performance reports. So, talking through the reporting process is useful,

12

You will find that some people do not look at their statements. They will say, "Send them to my accountant. Send me a summary; send me the income." Question: What tools do you use to help clients assess their own risk tolerances? Busby Oster: The most basic tool we have is the six-page questionnaire, which reveals a lot about how clients feel about losses. Clients are asked about risk in terms of percentage losses. For example, do you feel bad/not so bad if you lose 5 percent, 15 percent, 15 percent in a market that went down 15 percent, and so on. We ask questions that may, in fact, be restating the same questions in different ways on different pages. The purpose is to uncover their attitudes toward risk and be able to discuss the ramifications. For instance, we can say, "Welt you don't want to lose any money but you always want to beat everybody else. Do you know you can't have both?" You need simple examples to use in talking to people who are very wealthy but who may not be financially sophisticated. For example, they may listen to the evening news, and you can talk about what it means when the market goes down 100 points, how it might affect their small-cap portfolio. Question: Do you try to assess the client's desire for tax efficiency versus investment efficiency? Busby Oster: Clients tend to want to be tax efficient but also to have returns that are better than the market's. The only way to

change those conflicting desires is by continuously working to educate clients. As the coach, you may have to keep reminding them about what each team member is supposed to do. For example, the small-cap/value manager isn't supposed to own 5&P 500 Index stocks even though the S&P 500 is up in a particular year. You remind the client that the small-cap manager is not to be compared with the performance of the S&P 500. Question: Do you include in the policy statement a specific expectation for return or a target return for the portfolio? Busby Oster: Most target returns that I use have both a real rate of return number and an expression of risk that relates to the asset allocation-for example, 3 percent real return plus 5-7 percent over the benchmark or relative to a peer group that is managing the same kind of money. The return number depends on the asset class. For example, with an all-bond portfolio, you cannot ask for much over the benchmark. The benchmark also depends on the asset class; a large-cap / core manager would be evaluated in relation to a largecap / core peer group. All of the targets are after fees and expenses. The evaluations are not necessarily made after taxes in the case of multiple managers. For example, if the coach tells a largecap manager with gains, "We want to take some money from large cap and move it to small cap; don't worry about gains, because there are ample offsetting losses elsewhere," the coach is forcing selling, and thus tax events, on the individual portfolio manager. So, measuring that manager on an after-tax

©Association for Investment Management and Research

Workingwith Taxable Clients basis would be unfair. That approach is not completely in compliance with the AIMR guidelines for performance presentation, but it is how we deal with the fact that things may be happening to the individual pieces that may not be under the control of those managers but are in the client's best interest when viewed in totality. Question: What guidelines do you apply for rebalancing the portfolios of taxable clients? Busby Oster: From everything I've read, whether taxes are involved or not, rebalancing more than once a year isn't useful. We use wide bands around target allocations-20 percent on either side of the target-so for an allocation to move outside a band requires a large disparity in the rates of return in the various classes. We will rebalance at any time if something seems egregiously out of balance among asset classes. When U.S. Treasury security returns dropped under 5 percent, we stopped (perhaps partly because we remembered when Treasuries went through 6 percent on their way to 12 percent) and said, "Wait a minute. This is

too out of line. We should consider rebalancing the portfolio. We may decide not to rebalance, but we take a look at the issue. In 1998, we have been assessing the issue of large-cap stocks versus small-cap stocks. Whether rebalancing is done in consultation with the client or at our own discretion depends on the client. Generally, we can rebalance at our own discretion, but we find it makes a lot of sense to inform people at the onset rather than three months later. The phone and e-mail are acceptable and efficient ways of informing clients about their portfolios. (Keep in mind that if you work for anybody who is in any way regulated by brokerage industry rules, you need to keep hard copies of your e-mails to your clients. The requirements for e-rnails are the same as for written letters.) 11

Question: What is the difference between the coach relationship and the investment manager relationship with clients? Busby Oster: If you are the coach, you are managing the managers. The first thing to remember is that the client or the

©Association for Investment Management and Research

general manager is the person who is paying you to be the coach. You have probably played a significant role in hiring the managers. You don't want to blindside a manager in a meeting, so you may want to prepare managers ahead of time if you are going to be asking them some difficult questions. It is also a good idea to compliment them for what they've done, particularly in a difficult environment. As the coach, you also can't second-guess the managers. There is a fine line between looking out for the client's interests and micromanaging. When everything is wonderful, the relationships between coach and managers go smoothly; the problems arise when something contentious must be dealt with. An important aspect of managing managers, I have found, is assuring that the clients get what they are paying for. For example, if someone is a closet indexer, the client should not be paying active management fees. If the asset is a limited partnership that has a 1 percent fee and profit participation, I want to see that the general partners have a meaningful stake--say, 20 percent of their own money--on the line in the partnership.

13

The Complexities of Global Clients Philip Penaloza Vice President, Portfolio Manager Merrill Lynch Asset Management

Although global clients share common elements with domestic clients, global clients have unique characteristics and complex needs that require special attention from investment counselors. Establishing a successful long-term relationship requires identifying and addressing key issues. Meeting tangible and intangible needs requires communication, a lot of handholding, and making the client feel special.

xp erience teaches that investment managers should not have any preconceived notions or assumptions about who their clients are, especially if the clients come from a variety of cultural and national backgrounds. A "typical" global client does not exist, just as a typical U.S. client does not exist. Every client who seeks investment counsel has unique needs and circumstances. Hence, each client seeks unique services, and providing these services to address complex individual needs makes the investment profession thrive. This presentation examines the characteristics of global clients, identifies some of their complex needs, and addresses the issues that are most important for and most common among global clients.

E

Characteristics of Global Clients For the purposes of this presentation, a global client is defined as a client who is a non-U.S. national and whose investment planning needs and solutions cross borders, currencies, and national jurisdictions. U'.S, and global clients share two elementsdeath and taxes-which necessitate estate and tax planning. In general, global clients are open to more flexible and creative tax, estate, and investment planning options than are Ll.S, clients. Also, global clients have needs that are usually taken for granted by the typical U.s.-based investment manager. The ideal global investor is well read and knowledgeable about world financial securities and markets and appreciates risk-return trade-offs and efficient diversification of assets. In other words, the ideal client is a longterm investor. Compared with the average U'S. eli-

14

ent, the average global client is more leveraged and has a more sophisticated understanding of trading and foreign currencies. Despite the uniqueness of individual clients, a few generalizations are useful. Clients of different nationalities and different cultural backgrounds have very different appetites for risk. One extreme consists of the ultra risk averse, who do not want to invest in anything other than time deposits. European global clients commonly fit into this group. They tend to be professionals and are usually trying to avoid local taxes on their "idle" funds. They tend to prefer fixed-interest investments. Managers have to slowly educate such clients, work step by step, and dabble in equities and bonds until the clients feel comfortable. The other extreme is risk seekers. Asian clients are predominantly business entrepreneurs or merchants who want to avoid local taxes but also seek to use leverage. For these investors, "idle" funds do not exist because their money is always working for them. Most of the tycoons in Asia are overseas Chinese or South Asians from the subcontinent and are highly risk tolerant. As a matter of fact, they are risk lovers, but only if the act of risk taking is under their control. One of the challenges of marketing to such clients is making sure they get the message about discretionary portfolio management. They are reluctant to hand over control to a manager. On the other hand, these clients do very risky things with their portfolios, such as trading on commodities, futures, and options. The key is to offer them alternatives to diversify their portfolios and manage the portfolios for future generations. Even though they understand the concept of rational portfolio management and the

©Association for Investment Management and Research

The Complexities of Global Clients risk-return trade-off, getting them to relinquish control is difficult. Once Merrill Lynch can convince the clients we encounter in Asia that they need to diversify into more-conservative alternatives and we develop a globally diversified portfolio for them, they usually ask us not to invest in their country of domicile because they trade on those exchanges for themselves. We manage their offshore diversification for them. Usually they have global portfolios, primarily European, U.S., and other North American stocks, with sometimes a few Latin American stocks. The more risk-averse investors will have global fixedincome portfolios in which, if the portfolio is large enough, we try to hedge the currency risk into the benchmark reporting currency. Global clients also tend to be more short-term oriented than U.S. clients. They are often multilingual, but even though many of them were educated abroad and speak English, they feel more comfortable speaking in their native tongues. If we have a direct relationship with a client, we deal frequently with and must be diplomatic toward the client's assistants, accountants, and attorneys. A client's accountant could even be a minister of finance!

Upsides and Downsides For financial professionals, an attractive aspect of dealing with global clients is that they have large pools of money. In this business, developing a relationship with a global client takes a long time, but once formed, the relationship becomes much deeper than with U.S. clients. Moving accounts from one manager to another is more difficult, which creates a need for a stable organization and staff continuity. Developing goodwill is expensive, but once developed, it lasts a longtime. One of the downsides of dealing with global investors is that they always want you to manage for free. Our business is not a retail business, because these clients did not get to where they are by paying retail prices. So, the margins are tight. They always want to have the cheapest management fees, and they compare their deals with those of their friends and associates. When you present them with the fee schedule, they are not pleased to see it. They would prefer to tell you how much they want you to charge them or how much they want to pay. This attitude means that many firms may not be in compliance with the AIMR Performance Presentation StandardssM. Firms publish a high fee on their sheets but do a lot of discounting to make clients feel that they are getting a big discount on fees. At Merrill

©Association for Investment Management and Research

Lynch, the fee we publish is the fee we tell the clients. If we are talking about large amounts, a declining fee

scale allows a discount of a few basis points in certain places that will make clients feel special. In terms of fund or portfolio management, we charge a fee that can then be unbundled in various parts of the firm, whether it is custodial fees or trust fees. We also do international work on a wrap-fee basis, where commissions and management fees are bundled together. Because our main jurisdiction is the United States and we are subject to U.S. securities regulations, we are not able to offer the total wrap fees that a lot of European and Asian investment houses are able to offer. Global clients are becoming more common among U.S. investment management firms. The average high-net-worth Ll.S. immigrant travels first class and has significant wealth. Many global clients commute among multinational business interests. The average immigrant no longer stays in the United States indefinitely but commutes between his or her home country and the United States; they travel back and forth as often as four times a year. Some global clients are simply frequent visitors who maintain second homes in the United States.

Complex Needs Global clients have complex needs that must be successfully identified and addressed if the manager is to have a successful long-term relationship with the client. The following sections do not provide an exhaustive list of needs but will identify certain "hot button" items that can be used to gauge the attitudes and values of global clients. Safe Haven. Global clients may seek a safe haven for corporate and trust incorporation and for asset custody. Usually, global clients are looking for countries with stable, nonconfiscatory national regimes where they can park their assets and set up their personal holding companies and trusts. They want countries in which the legal, accounting, and tax codes are established and rational. They also require a body of impartial and professional courts, judges, and lawyers to settle disputes. When clients' trusts go into perpetuity-that is, when they die and their trusts go into active implementation-they want a reliable body of lawyers executing their wishes. The subject of safe havens thus boils down to which court system the client is most comfortable with. Most clients settle for countries governed by English common law. Secrecy. The main question related to secrecy is how much clients are willing to pay for privacy,

15

Investment Counseling for Private Clients for anonymity, and to have their assets untraceable. This kind of secrecy is needed to guard against political risk, creditors, and lawsuits. In addition, clients may wish to establish multitier personal holding companies-that is, to have a personal holding company on top of one or more other personal holding companies. According to an old Chinese saying, a wealthy person needs fame as much as a pig needs to fatten up. Many international clients are very sensitive to this issue, and the secrecy involved is tremendous. For example, Merrill Lynch uses a Swiss bank in Geneva, SWitzerland, that caters to clients that require ultrasecrecy and need many layers of corporations to preserve the anonymity of their assets. The need for secrecy is one of the reasons our international marketing representatives wear many hats. Depending on the needs of the client, they may represent a Swiss bank or our recent acquisition, Mercury Asset Management, where the custodian used is an offshore bank. We once had a client in Central Europe who opened a $5 million trust account with a Merrill Lynch representative. A week later, the client's brother, who was a long-time client of this Merrill Lynch private banker, called the banker and, engaging in casual banter, made references to his brother's newly opened trust account. Naturally, he drew a blank response from the private banker, who deftly ducked every attempt to become drawn into such a conversation or even to acknowledge the brother's new account. The next day, the new client added $100 million to his trust account. The call the previous day had been a ruse to test the code of confidentiality. Portability. In the event of a national emergency, clients may need to transfer company registration or the trust domicile to another bank or another jurisdiction. Many times, the trusts of global clients include a fallback or escape clause in which, in the event of a national calamity, the trust is transferred to another jurisdiction. For example, during the height of the Cold War, when clients feared an invasion from Eastern Europe, they wanted to be assured that a trust or personal holding company established in the Jersey Islands or the Channel Islands would transfer to the jurisdiction of the Cayman Islands. Often, the solution is to have backup corporations or parallel overlapping trusts.

clients have multiple residences around the world, which enables them to choose the most advantageous resident jurisdiction for tax purposes. The same phenomenon occurs in the United States when clients choose the state with the most tax-efficient jurisdiction for their residences. In addition, many global clients, usually the tycoons, have ongoing business operations that allow them to enjoy tax credits (such as those that are offered to encourage enterprise tax zones), tax holidays, and tax shelters, which makes their planning more flexible. Regulatory Regime. Clients seek regulatory regimes and financial systems with integrity. For example, they may investigate whether opening an account in Hong Kong and having that trust looked after by solicitors in Hong Kong or Singapore will be monitored by the relevant banking or professional supervisory authorities to ensure high professional standards. Global clients are very concerned that their accountants, lawyers, and bankers be well regulated. Domicile. Clients often have complex domicile requirements. The clients, the grantor, or the settler of a trust may live in one country, but the trust may be established in another jurisdiction, and the trustees may be located in yet another jurisdiction. Finally, beneficiaries may be located all over the world. Reporting. Reporting performance is a complex matter involving multiple currencies and multiple benchmarks. A report is frequently a moving target. Clients often choose to change the benchmark for judging performance from year to year; sometimes they want multiple benchmarks. Language can also pose a challenge in reporting. Performance reporting and report statements may need to be multilingual, just as corporate presentations, annual reports, or quarterly reports are pre~ pared in local languages. Investment Medley. Many global clients have only one savings account with multicurrency deposits. Many universal banks outside the United States have combined on their monthly statements multicurrency deposits, stocks, bonds, and the riskier or more volatile futures, options, derivatives, and commodities. Many global investors also dabble in moretangible assets, such as real estate and golf dub debentures. Art, precious metals, and gemstones may be some of the most sought-after investments overseas.

Taxation Benefits. Managers need to be aware that global clients have the flexibility to benefit from international tax treaties between other countries. In addition, a client with multiple nationalities has the luxury of picking the country that would be most tax efficient for his or her situation. Also, many global

Intangibles. Global clients seek various intangible qualities in their investment managers. Prestige

16

©Association for investment Management and Research

The Complexities of Global Clients

is important. Some clients look for a global brand name, a name that is advertised in markets around the world. Many clients are looking for cachet firms with names that may be recognized only by people in the know.

Value-Added Services Globally connected firms offer significant advantages to their clients. For example, we can refer our clients to the Merrill Lynch global network of services. We work closely with the Merrill Lynch financial consultant (FC), who is our primary contact in terms of referring clients and prospective clients to us. The FCs, who are based in local markets at local Merrill Lynch offices,have good local contacts. The FCs provide the first source or line of communication with clients. We also have portfolio managers based in Hong Kong, Geneva, London, Princeton (New Jersey), and other cities across the United States and Japan who can, in conjunction with the FCs flying into their home country, give clients regular reports on their portfolios to provide the interaction and handholding that is so crucial to global business relationships. We often refer people to certain value-added services that can facilitate movement of block funds. Within Merrill Lynch, opportunities abound to arrange cross-border parallel loans and back-to-back letters of credit. We also refer clients to other clients who may have opportunities because their funds or real assets are blocked; we can provide them with countertrade and barter opportunities. Often, clients also need insurance services as an add-on to their portfolio management services. Tycoons and wealthy individuals running their own businesses need our investment bankers for initial public offerings, mergers, or acquisitions. They also own large chunks of buildings and properties, and Merrill Lynch is able to securitize these hard assets in the global capital markets in order to provide clients with liquidity.

©Association for Investment Management and Research

Another service we provide clients is referrals to and coordination with specialists-for example, immigration lawyers and offshore accounting, taxation, and legal counsels. Clients may even ask us to refer them to schools for their children. We also refer them to medical professionals, and we network with clients in their art and charitable circles.

Establishing Relationships Investment managers must follow three steps to satisfy global clients and start a successful long-term relationship with them. First, go over the list of hot buttons described in the previous section. Second, get an international tax accountant and lawyer. Third, if the portfolio is already in hand, immunize it from market volatility during the transition and planning phase. Ideally, the portfolio should be tax, currency, and market neutral. Once the appropriate tax and trust structures have been set up and the portfolio has become an offshore account, capital gains taxes are never an issue. Interest income from certain bonds is always subject to withholding taxes because it is withheld at the source. The challenge is to find the best situation among the various jurisdictions to which the client is subject. To make this determination, managers must know their clients' risk-efficient diversification objectives, risk and comfort levels, and various tax constraints.

Conclusion Global clients have universal needs. No matter how complex their situations, their needs can be split into two categories: tangible and intangible. Tangible needs boil down to the basic requirement for a satisfactory investment return, subject to the client's profile. Intangible needs concern clients' desires to feel that they are unique and special. Both needs involve a lot of handholding and communication.

17

The Psychology of Risk and Taxes Meir Statman Glenn Klimek Professor of Finance Leavey School of Business Santa Clara University

Risk and taxes are linked by aversion; people are as averse to risk as they are averse to taxes. The twin aversions to risk and taxes, which are features of perfectly normal people, teach us about the psychology of normal people. The twin aversions also teach us about the pitfalls on the road that financial advisors travel as they manage their clients.

M

y accountant told me about a client who came to him following a very profitable real estate transaction. "Congratulations," said the accountant, "you made a million dollars." "Thank you very much," said the client sarcastically, "I made a million and a half dollars before I stepped into your office." This client, like many people, sees her advisor as a messenger from the US. Internal Revenue Service (IRS). It is as if financial advisors had the power to banish taxes and deliver high returns without risk. "Taxes are the price that we pay for a civilized society," said U.S. Chief Justice Oliver Wendell Holmes. Well, yes. We know that we have to pay taxes, but why do we have to pay so much? The search for tax shelters is long standing. The title of an article in The World's Work magazine of May 1914 is "Of Buying Stocks to Dodge Taxes." It tells the sad story of a clergyman searching for tax-free returns who ended with no returns at all. The article states, "Doubtless the time will come when all the states will recognize the necessity of working out, along sensible and scientific lines, the problem of the taxation of investment securities." Keep on wishing. We will never work out taxes along sensible and scientific lines. There is something impossible about taxes. It is not only that we have taxes but that we have many kinds of taxes: income taxes, sales taxes, property taxes, and more. Why so many kinds? Ask the tax authorities, and they will tell you the reason is that people cheat on their taxes. We have so many kinds of taxes, they say, for the same reason we have so many tollbooths on roads: Drivers are able to evade some tollbooths, but they cannot evade them all. Ask taxpayers the same question, and they will tell you that they are good, patriotic citizens. It is the

IRS that always cheats, they say, and they are only trying to protect themselves from the IRS's cheating. Evolutionary psychology provides a window into the nature of the war between taxpayers and the IRS. Evolutionary psychology is built on the Darwinian idea that everything about us, including our brains and psychology, has been shaped by the forces of evolution. Our hearts are specialized machines, not general-purpose machines. Hearts are designed to pump blood, not water or oil. Similarly, our brains are specialized machines, not general-purpose machines. Brains are more like simple calculators than they are like computers. Computers are generalpurpose machines; you can use them to run any software program-s-from a spreadsheet to word processing. But calculators are specialized machines, you can use them for some programs, such as adding and multiplying, but not for others, such as word processing. The modules of specialized machines do have some flexibility. For example, you can use the multiplication module of a simple calculator to find, by trial and error, the square root of a number. But the trial-and-error process is tedious, and it is more prone to error than the direct process of finding a square root number with a computer or with a calculator with a square root module. Our brains developed mostly in the hunter / gatherer period, through the forces of evolution, to solve problems that conferred evolutionary advantage. We have a brain module ready to acquire language. The module can acquire many languages, but it will not acquire gibberish. In turn, acquisition of language affects the structure of the brain. For example, the accent someone hears underneath my English is the accent of Hebrew, my first language, embedded in my brain.

18

©Association for Investment Management and Research

The Psychology of Risk and Taxes

Framing Taxes One module of the brain is designed to acquire language. Another is designed to detect cheaters. A brain module that detects cheaters was as important to hunters and gatherers as it is important to us modem people, because both our ancestors and we engage in transactions that are not simultaneous. Nonsimultaneous transactions open the door to cheating: If I give you a portion of the deer I hunted today, will you reciprocate next week when your hunt is successful or will you cheat? Here is an experiment examined by Cosmides (1985) that highlights the "look for cheaters" module in our brains. The Cosmides experiment built on the following selection task: Imagine that you are a new clerk at a high school. Part of your new clerical job is to make sure student documents have been processed correctly. Specifically, you are to make sure the documents conform to the following alphanumeric rule: If a person has a D rating, then his or her document must be marked Code 3. You suspect that the clerk you replaced did not categorize the students' documents correctly. You have cards containing information about the documents of students who are enrolled at this high school. Each card represents one student, and one side of a card tells a student's letter rating while the other side tells that student's number code. You have pulled out the following four cards:

Your task is to indicate only the card(s) of these four that you definitely need to tum over to see whether the documents of any of these students violate the alphanumeric rule. Logic dictates that you tum over two cards, Card D and Card 7. Card D has to be turned over because it might have a number other than 3 on the other side, which would violate the rule. Card 7 has to be turned over because it might have the letter D on the other side, which would also violate the rule. You have no reason to tum over Card F or Card 3 because there is no way for them to violate the rule. (The rule does not require that all 3's be D's, only that all D's be 3's.) Nevertheless, people commit cognitive errors as they solve the clerk problem; only 4-25 percent of people in the original experiment chose to tum over the correct cards, D and 7. The most common answers were D and 3 or D and F. Cosmides asked: Why do most people fail to solve the clerk problem correctly? The answer, she concluded, is that the clerk problem does not fit well

©Association for Investment Management and Research

within a specific-purpose brain module. Cosmides demonstrated her insight with a "drinking age" problem, a problem that i~ identical to th~ cle~k problem in substance but not in frame. Cosmides s frame created a fit between the problem and the "look for cheaters" specific-purpose brain module. . Consider the drinking age problem: In 1tS crackdown against drunk drivers, Massachusetts law enforcement officials are revoking liquor licenses left and right. Suppose you are a bouncer in a Boston bar and you will lose your job unless you enforce the following law: If a person is drinking beer, then she or he must be over 20 years old. Now, the following cards have information about four people sitting at a table in your bar. Each card represents one person. One side of the.card tells what a person is drinking and the other side of the card tells that person's age. Indicate only those card(s) you definitely need to tum over to see if any of these people are breaking this law. 16 Years

Old

Whereas only 4-25 percent of subjects solved the clerk problem correctly, about 75 percent solved the drinking age problem correctly ("Drinking B.ee.r" an~ "16 Years Old"). The drinking age problem 1S identical in substance to the clerk problem, but its frame fits the "look for cheaters" module. Subjects are asked to look for cheaters of the liquor law. People who lack a "look for cheaters" module do not fare well because they are not able to tell that they are being cheated again and again. Unfortunately for the IRS, it appears to us to be a cheater. Ask anyone if they get their taxes' worth from the government. We all know, in some abstract way, that tax revenues are for the common good, but our "look for cheaters" module tells us that we are being cheated. Framing taxes involves more than the activation of the "look for cheaters" module. It also involves the variations that accompany the seasons of taxes. Some glasses that are framed as half empty in Nove~be: are framed as half full in December. If your clients mutual funds are up 30 percent in November, they frame the 30 percent profit as entirely theirs, as if there were no IRS. In December, your clients add the tax bite of the IRS into the frame, as if the IRS were just born. Similarly, a 30 percent loss i.s framed as a loss in November while the same loss 1S framed as a tax deduction in December.

Framing Risk Risk, like taxes, is framed. Risk is framed in standard 19

Investment Counseling for Private Clients finance as the component that, together with expected returns, determines rational choices among investment alternatives. Rational people are assumed to prefer high expected returns to low expected returns and low risk to high risk. It is customary in standard finance, following Markowitz, to frame risk as the variance of returns. But normal people, the people described in behavioral finance, are not always averse to variance. Consider the following experiment by Kahneman and Tversky (1979).

choice between A 2 and B2 in Problem 2, the overall choice is between A4 : a sure gain of $1,500 and B4 : a 50 percent chance to gain $2,000 and a 50 percent chance to gain $1,000. Problems 1 and 2 are identical in substance. In other words, they are identical when judged as a whole. But people do not integrate the parts of a problem into a whole before making judgments. Rather, people frame the parts separately into mental accounts, and they make choices about the mental accounts separately, one at a time. So, subjects in the experiment placed the $1,000 received in Problem 1 into one mental account and made the choice between the sure amount and the gamble without reference to the $1,000 received. They behaved similarly with the $2,000 received in Problem. 2. Such framing led subjects to the sure amount in Problem 1 and the gamble in Problem 2. Markowitz taught us to frame portfolios as a whole-to consider the correlations among assets, not simply the variance of each asset. People who really learned Markowitz's lesson are not confused by the frames of Problems 1 and 2. But normal people are rarely Markowitz people. As Figure 1 shows, normal people frame assets within distinct mental accounts, as if mental accounts are layers in a pyramid, each associated with its own goals. One layer is for downside protection, to protect us from being poor. This layer is where we put U.S. Trbills, money we cannot afford to lose. Another layer is for upside potential, the potential for being rich. This layer is where we put Internet initial public offerings (IPOs) and lottery tickets. For some people, aspirations of upside potential are low relative to their resources. They have $500,000 now and want $1 million when they retire. Such people use stocks for the upsidepotential layer. Other people, those with high aspirations and little money, buy lottery tickets. They have $1 now and want $20 million. Buyers of lottery tickets are not stupid; they know that the expected value of the dollar they pay for a lottery ticket is 50 cents. It is not the variance of lotteries they like; it is the chance to be rich.

One group of subjects receives Problem 1: 1. In addition to whatever you own, you have been given $1,000. You are now asked to choose between A 1: a sure gain of $500 and B1 : a 50 percent chance to gain $1,000 and a 50 percent chance to gain nothing. Another group of subjects receives Problem 2: 2. In addition to whatever you own, you have been given $2,000. You are now asked to choose between A 2 : a sure loss of $500 and B2 : a 50 percent chance to lose $1,000 and a 50 percent chance to lose nothing. Kahneman and Tversky found that 84 percent of subjects chose A v the sure amount, over B1, the gamble, in the first problem set. The choice of the sure $500 in Problem 1 is consistent with aversion to variance. Al has a variance of zero, whereas Bv the alternative, has the same $500 expected value together with a positive variance. In Problem 2, however, most subjects chose as if they loved variance; 69 percent of subjects chose B2, the gamble, over the sure amount, A2 . The choice of the B2 gamble is puzzling under the assumptions of standard finance because the gamble has higher variance than the sure amount and an equal and expected value. Are people averse to variance, or do they love it? It turns out that people are not really averse to variance; it is losses that they hate. They prefer the gamble for a loss of $1,000 or a zero loss to the sure $500 loss in Problem 2 because the gamble gives them a fighting chance to break even and avoid a loss. The sure loss means kissing $500 goodbye. Problems 1 and 2 contain another insight. The problems are, in fact, identical in substance, but the difference in frames leads to radical differences in choices. When the initial $1,000 is integrated into the choice between Al and B1 in Problem 1, the overall choice is between A3: a sure gain of $1,500 (the sum of the initial $1,000 and the sure $500) and B3 : a 50 percent chance to gain $2,000 and a 50 percent chance to gain $1,000. Similarly, when the initial $2,000 is integrated into the

People in standard finance begin the process of choice among investment alternatives by finding the risk and expected returns of each investment. But what is risk? Risk surely means more than variance. Indeed, the concept of risk is problematic for two reasons. First, risk means many things, and second, we each have specific ideas about the meaning of risk. So, discussions about risk are all too often discussions among people who are deaf but not mute.

20

©Association for Investment Management and Research

The Box of Risk

The Psychology ofRiskandTaxes Figure 1. Markowitz Mean-Variance Portfolios and Behavioral Portfolios Mean-Variance Portfolio

Behavioral Portfolio

Mean-variance portfolios are constructed as a whole, and only the expected return and the variance of the entire portfolio matter. Covariance between assets is crucial in determination of the variance of the portfolio.

Behavioral portfolios are constructed not as a whole but layer by layer, where each layer is associated with a goal and is filled with securities that correspond to that goal. Covariance between assets is overlooked.

~r---

Upside-Potential Layer (contains, for example, foreign stocks, aggressive growth funds, /-----'\ WOs, lottery tickets)

Downside-Potential Layer (contains, for example, 'l-bills, ....- ' t - - certificates of deposit, money market funds)

Source: Statman (1999). Reprinted by permission of the Financial AnalystsJournal.

The debate about the meaning of risk goes beyond the debate about the merits of variance and semivariance. We regularly talk about inflation risk and liquidity risk, management risk and market risk. The word "risk" has become a synonym for "factor." We could talk, with greater clarity, about investment choices affected by inflation factors, liquidity factors, management factors, and market factors. Trying to squeeze many factors into risk and expected returns does not work. Risk is many factors, and it needs a large box to contain them all. The desire to simplify the concept of risk makes it incomprehensible. Picture the engine compartment of your car. Is it comprehensible? The engine compartment of my car makes little sense to me. I know where the engine is, and I know where the transmission is. Two factors are good enough for me. Why don't I just Simplify the engine compartment by cutting away the jumble of hoses and wires that clutter it? This is what we do when we simplify investment choices by limiting factors to risk and expected returns and by defining risk as variance. Our desire to make choices simple is so great that we are willing to give up their essence. To a trained technician, the engine compartment of a car is perfectly comprehensible; it is not a jumble. A technician can tell us the function of each of the many hoses and wires. Similarly, investment decisions are comprehensible to the great technician in the sky. Someday, we may understand investment decisions as well as that great technician does. Meanwhile, we should not pretend, in the name of simplification, that risk is nothing but variance. So, what is risk? What are some of the wires and hoses that fill the box? I will discuss a partial list:

©Association for Investment Management and Research

losses, regret, familiarity, and thrill seeking. I will illustrate these factors within the context of two investment strategies, namely, time diversification and dollar-cost averaging (see Statman 1995 and Fisher and Statman, 1999b). These strategies serve as Rosetta stones in understanding the factors of investment choice. Investors tell us that both time diversification and dollar-cost averaging reduce risk. Our challenge is to decipher the meaning of risk from these investor perceptions. Risk and Losses. Imagine that you begin with $1,000 and have a 50/50 chance for a 20 percent gain or a 10 percent loss in each period. Does your risk increase with the number of periods, or does it decrease? This is the question of time diversification. There is a 50 percent probability that you will lose money if your time horizon is one year. But as Figure 2 shows, this probability declines to 25percent if your horizon is two years. So, if risk is the probability of loss, risk declines when the time horizon increases. This argument is the essence of the pro-time-diversification camp-those who argue that time reduces risk. "But look at the amount of loss!" cries the antitime-diversification camp. Figure 2 also shows that the maximum loss you can sustain after a year is only $100,but you can sustain a higher loss, $190,after two years. So, if risk is the amount you can lose, risk increases as the time horizon increases. So, what is risk: the probability of losses, the amount of losses, or some combination of probability and amount? The attraction of time diversification to its proponents has to do with framing. The framing of the pro-time-diversification camp focuses attention on the probability of losses and downplays the

21

Investment Counseling for Private Clients Figure 2. Relationship between Investment Horizon and Probability and Amount of loss $1,440

$1,200

$1,080

$810 One-Year Horizon (50% chance for a $100 loss)

Two-Year Horizon (25% chance for a $190 loss)

amount of losses. The probability that stocks will underperform T-bills, based on historical data, is more than 30 percent over a I-year horizon but is only 0.1 percent over a 4D-year horizon. People who focus on the probability of losses might conclude they have little to worry about if their horizon is 40 years. But people who focus on the amount oflosses will tell you that 40 years gives you a chance to drive your investment all the way into the ground. The amount you might lose over one year is smaller. Framing is very important when you deal with low probabilities. Imagine that you tell a client that there is 0.1 percent probability that the return on his or her investment will be less than that of T-bills over a 40-year horizon. The client might conclude that 0.1 percent is so close to zero that it might as well be zero. Tversky and Kahneman (1992) found that funny things happen when people assess small probabilities: They either reduce small probabilities to zero, or they exaggerate the probabilities well beyond their true magnitudes. People who buy lottery tickets often exaggerate their small probabilities of winning. People who think about the chance that their stocks will underperform T-bills over 40 years often set that probability at zero. For example, de Fontenay (1996), arguing for time diversification, wrote, A positive return [on stocks] in the long rW1 is near certainty.... There is no reason to expect a negative return on the broadest possible stock index.... Samuelson (1994) has been vocal in reminding investors that small probabilities are different from zero probabilities and that the amounts of losses have to be considered along with their probabilities. Stocks

are not 40-year T-bonds that assure payment at maturity. Moreover, there is a positive probability of default even on U.S. T-bonds. The notion that stocks guarantee a happy ending over the long run is an illusion. Low probabilities are sometimes perceived as zero probabilities, but at other times, low probabilities are exaggerated into very high probabilities. In her description of life as an obsessive-compulsive, Colas (1998) wrote about a woman who, out of boredom, peeled off the label of a can of tuna from which she was eating. Underneath the tuna label, she found another label-a label for cat food. Now, what are the chances you will accidentally eat cat food, Colas asks. Well, maybe 0.001 percent, she answers. So, to be safe, Colas became a follower of the 0.001 percent 100 percent theory of life: If there is a 0.001 percent chance that an event can occur, then the chance might as well be 100 percent. And life has to be lived accordingly. As you can see, the consequences of exaggerating small probabilities can be as sad as the consequences of assuming that small probabilities equal zero. This story leads us to the notions of "normal," "smart," and "rational." It also leads us to the role of financial advisors. Imagine that you are the advisor of "Paul Jones," a 30-year-old man with a horizon of more than 40 years. In your assessment, Jones should have a portfolio consisting of 90 percent stocks and 10 percent T-bills. But Jones resists. "I am afraid of stocks," he says, "because stocks might go down. I prefer a portfolio that is 100 percent in 'l-bills." Is Jones normal? Is he smart? Is he rational?

22

©Association for Investment Management and Research

;;;;0

The Psychology of Risk and Taxes Nothing in standard finance suggests that Jones is irrational. All degrees of risk aversion are rational in standard finance. "No one can prove to me that I am too risk-averse," wrote Samuelson (p. 24). Samuelson knows his mind as well as Colas and Paul Jones know theirs, and they are all entitled to the presumption of rationality. Perhaps Colas and Jones are simply very risk averse. Very risk-averse people choose portfolios on the mean-variance-efficient frontier with low risk and commensurately low returns. We know that Colas is not normal, but Jones is normal. He holds a steady job, and you know that, together with his wife, he provides a good home for their two children. But is Jones smart? Many advisors would conclude that he is not. They might teach him the historical returns of stocks and T-bills, and if this lesson failed, they might resort to the argument of time diversification. "Look," they would say, "don't be afraid of stocks. There is only a 0.1 percent chance that you will come out behind T-bills with stocks if you hold them for 40 years." The role of financial advisors is to turn normal investors into smart investors. They use time diversification arguments to encourage timid investors to invest more in stocks by emphasizing the low probabilities of losses. And they restrain ebullient investors who want to invest everything in stocks by emphasizing the high amounts of potential losses. Regret. Jeffrey (1984)wrote that "the real risk of holding a portfolio is that it might not provide the owner ... with the cash he requires to make essential outlays" (p. 34). It seems like a good definition of risk. But wealthy investors might lose 50 percent, even 90 percent of their wealth without losing the cash they need to make essential outlays. So why do the rich care about losses? It turns out that losses inflict pain in two places-in the pocketbook and in the ego. The pain to the ego is the pain of regret. To understand the pain of regret, imagine that your flight from San Francisco to New York arrives late. You run to the gate of your flight to Paris carrying a heavy bag. As you stand at the gate breathless, the agent says the flight to Paris left on time, an hour ago. Now, imagine an alternative scenario: As you stand at the gate breathless, the agent says that the flight to Paris was delayed and left the gate just a minute ago. You see the plane taxiing on the runway. The two scenarios are identical in substance. You had every reason to expect that you were going to miss the plane, and indeed, you missed it. But you kick yourself harder in the second scenario because it carries an extra punch of regret-the pain that you feel when you find out after the fact that had you run just a bit faster, you would have fared better. The pain of regret affects many choices, including financial choices.

©Association for Investment Management and Research

Think again about time diversification. The time diversification debate is typically framed as if people have particular time horizons that are fixed before investment choices are made. But in fact, people do not have fixed horizons. Sometimes, investment outcomes determine horizons. For example, imagine that you bought your house for $400,000 with the intention of selling it in seven years. Seven years have passed, but now you find that the house will fetch only $300,000. Will you sell? Many investors conclude that the market is simply "slow" and do not sell. Selling today means the realization of a loss. Realization of a loss brings the pain of regret, and many investors avoid the pain of regret by extending their time horizons. The fear of regret is the feature that makes dollarcost averaging attractive. The usual story is that the attraction of dollar-cost averaging is in its ability to reduce risk. But think about the following scenario: You inherit $100,000 in the form of a stock portfolio. Because you are very averse to risk, you want it all in T-bills. So, why don't you also sell all your stocks today, in a lump sum, and buy T-bills? Because you would feel a searing pain of regret if the stock market were to rocket as soon as you sold. Dollar-cost averaging out of the market, like dollar-cost averaging into the market, is motivated by aversion to regret. Hindsight interacts with regret in the perception of risk. Risk has meaning only when looking at the future. But looking at the past brings hindsight and regret. Following a major tumble in the market in the summer of 1998,historian Ron Chernow (1998)wrote a newspaper article in which he stated that "we shouldn't have been surprised when the stock bubble burst." Well, I was surprised by the tumble, and I hope that you were surprised as well. I hope that Chernow was surprised when the market moved to new highs later in 1998. Be surprised, be very surprised, I say, because surprise teaches us a good lesson. It teaches us that hindsight is much more precise than foresight. We should know that crystal balls are cloudy. Hindsight is a major problem in the work of advisors. In hindsight, the decline in the Japanese market over the last decade seems inevitable. Your clients tell you now that they have known since 1989 that the Japanese market would go down, so why did you put some of their money in Japanese stocks? Here is what could have actually happened. If you had $1,000 in 1987 and invested it all in the U.S. stock market, you would have had $5,350 at the end of 1997. With the same investment in the Japanese stock market, you would have roughly broken even

23

Investment Counseling for Private Clients by the end of 1997. Had you moved the money optimally with perfect tactical asset allocation throughout each year to the best performing asset class, you would have had $11,482, but there is also a chance that you did your tactical asset allocation precisely the wrong way and ended up with only half your initial $1,000.All of these possibilities existed in 1987. But at the end of 1997, the possibilities shrink. It is very clear to your clients that if you had half a brain, you would have made them $11,482. Hindsight brings the pain of regret, and investors are quick to transfer the pain of regret to their advisors. "I am not stupid," they say, "my advisor is stupid." Risk and Familiarity. Is an investment in Japanese stocks more risky to Ll.S, investors than an investment in u.s. stocks? Not according to standard finance. Foreign stocks bring advantages to meanvariance-efficient portfolios of U'S, investors because foreign stocks have low correlations with domestic stocks. So, why do U.S. investors allocate to foreign stocks so much less than the proportion of foreign stocks in the world portfolio? Because foreign stocks are unfamiliar, and lack of familiarity is associated with risk Lack of familiarity underlies the "home bias" of investors everywhere, not only in the United States. In September 1998, I asked a group of Swiss investment professionals about their portfolios. The median allocation to Swiss stocks among all stocks was 60 percent, quite a bit more than the proportion of Swiss stocks in a world portfolio. Consider the experiments of Heath and Tversky (1991). They gave people a choice of bets on politics and bets on sports. They found that people who are familiar with politics prefer to bet on politics even when they consider the odds of bets on polities identical to the odds of bets on sports. Lack of familiarity seems like risk Huberman (1997)found that the home bias exists even within the United States. New York residents tend to buy NYNEX (the local New York State telephone company) shares and California residents tend to buy Pacific Bell shares. The combination of lack of familiarity, hindsight, and regret can be especially lethal. Lowenstein (1997) bashed investment advisors who diversified globally in an article titled '''97 Moral: Drop Global-Investing Bunk" "We have no money in the former Yugoslavia," he crowed, "none in the present Argentina, none in the future Republic of Antarctica, none in Zambia, Belgium or Kazakhstan."

tion costs, a heavily traded portfolio of 100 stocks is no more likely to lead to losses than a buy-and-hold portfolio of 100 stocks. Both portfolios, at the end of the day, have about the same volatility and the same correlation with the S&P 500 Index. Day-traders think of themselves as risk seekers, but they are better described as thrill seekers. There is a biological basis to thrill seeking. It is correlated with MAO, a substance that affects the brain. On average, older men are less thrill seeking than younger men and women are less thrill seeking than men. Barber and Odean (1998) found that, on average, women trade less frequently than men and they translate their commission savings into higher returns.

Returns Risk has many facets, and so do returns. Why does Bill Gates want $100 billion? Is it not possible to own jets and build fancy houses with only $80 billion? You cannot understand the drive for returns without insights from evolutionary psychology. We want more than high returns from our portfolios; we want to be #1. Financial advisors are frustrated by the desires of their clients to win. Some clients are not happy even if they beat the S&P 500 and most of their neighbors. They want to beat them all. Is the "competition" module in our brains useful? For a perspective, think about McDonald's restaurants. Evolutionary psychologists describe McDonald's as a temple to our hunter/ gatherer brains. We have a craving for sugar, protein, and fat, and McDonald's supplies all three in vast quantities. The craving for sugar, protein, and fat conferred evolutionary advantages because these nutrients were rare in the hunter/ gatherer environment. Those who gorged on these nutrients when they could had a better chance to survive and leave descendants. But what was a blessing in the hunter/ gatherer environment is a curse in today's environment. Now, we struggle against the urges that take us to McDonald's fare. Is the urge to be the richest man or woman in the world analogous to the McDonald's urge? Is the urge for riches rational? Is it smart? Whatever it is, the urge is surely a scourge in the work of financial advisors.

Conclusion

Risk and Thrill Seeking. Trading activity is not necessarily related to risk; in the absence of transac-

There was a time when we thought risk consisted only of variance. But we know now that risk is more. Many factors playa role in investment choices, and risk is a box of factors, not a single one. We discussed here a few of the factors: loss, regret, familiarity, and thrill seeking.

24

©Association for Investment Management and Research

InvestmentCounseling for Private Clients

Question and Answer Session Meir Statman Question: Are you saying that when a conflict occurs between financial theory and behavioral practice or the behavior of investors, the first responsibility of advisors is to try to reframe the issue in a way that moves people toward the rationality of financial theory?

there other ways to read clients' risk tolerance, such as looking at their portfolios? Statman: Many risk questionnaires are out there. Fisher and I (1997) studied the ones created by mutual fund companies. I am not impressed by the questionnaires I have seen for two reasons. First, risk comprises many factors, such as the ones I discussed earlier-loss, regret, familiarity, and thrill seeking. There is a need for questions that address these factors. Second, the biggest benefit of questionnaires is not in the questions but in the conversations they spark between investors and advisors. Again, think about advisors as financial physicians. Good questionnaires provide diagnoses, but they do not provide education and care.

Statman: The responsibility of advisors is to help normal investors act as smart investors act. It is important not to move too fast to recommendations because we first have to find out what investors want and how we can help them frame their options. Imagine that you are advising an investor who has most of his money in bonds. You recommend a shift of some money into stocks, but the investor resists. He needs income, he says, and the dividend yield of stocks is lower than the rate of interest. "Why not sell a few shares of stock from time to time to generate the income that you need?" you ask. "I could," the investor replies, "but my mom said don't ever dip into capital." This investor frames income and capital in two separate mental accounts, as if money in the income account were different in color from money in the capital account. The investor may also fear that lack of self-control might lead him to sell more than the prudent number of shares, endangering his financial future. You can help this investor reframe all money, income and capital, into one mental account and, if necessary, create a structure that will prevent him from dipping too deeply into capital. (Some advisors place investors on budgets so that the investors must come to the advisors for approval of substantial expenditures.)

Statman: Experience surely changes people's perceptions of the market and themselves. People tend to attribute both high returns and low returns to their own skills rather than to the luck of bull or bear markets. So, they become overconfident in bull markets and dispirited in bear markets. Thorley and I (1999) showed that investor confidence affects the volume of trading. There was a major decline in volume of trading following the crash of 1987. In contrast, volume in today's bull market is very high.

Question: Are there psychological tests that profile the risk aversion of individuals, or are

Can you comment on the idea that the nature of risk changes in the presence of taxes?

26

Question: Have researchers studied how an individual's utility function changes with experience, and if so, how does it change, how much does it change, and does it change differently on upside versus downside performance?

Question:

Statman: The nature of risk changes because the IRS shares in both gains and losses. When the IRSis ignored, gains and losses will appear to be higher than when the share going to the IRS is taken into consideration. It is also important to help clients frame losses smartly. Clients are reluctant to realize losses because realization brings the pain of regret. But tax deductions that accompany loss realization provide a silver lining. Question: Defining risk in terms of volatility and variance is convenient because it enables us to deal with risk with mathematical rigor. How do we deal as rigorously with the behavioral aspects of risk? Statman: Just deal with it. Rigor has become associated with math. But math is no more than a language, and language is useful only when it helps us communicate and reach better decisions. I do not care for the use of math to impress and look "scientific." Think about mean-variance optimization. You put perfectly reasonable estimates of expected returns, variances, and covariances into an optimizer. The scientific answer from the computer is: Put 90 percent in Japanese stocks. Do you follow the scientific answer? Of course not. You add a constraint to the optimizer: Put no more than 5 percent in Japan. The next optimizer result says: Put 17 percent in Malaysia. Again, you add a constraint to the optimizer: Put no more than 1 percent in Malaysia. When you are done, you get the portfolio you wanted in the first place, but now it is Nobel Prize rigorous. Use math to clarify ideas and use math to get numerical solutions to properly framed problems. But remember that math is a servant, not a master.

©Association for Investment Management and Research

The Psychology of Risk and Taxes Think about the rigor of diagnoses by physicians. Good physicians ask many questions about lifestyle, stress, sleeping habits, and more. They measure blood pressure, pulse, hearing, and more. Physicians do not try to squeeze it all into one factor. Again, risk is a box of many factors. You gain nothing by pretending that risk is variance. Identify the factors and communicate with your clients about them. Question: Going to the other extreme, how much time should you spend educating uneducated clients? When do you just give in? Statman: I am an educator, so I never give up. All people can be educated, but education takes time and teachers need patience. Think again about physicians. Your doctor notes that you are overweight, you smoke, and you never exercise. A doctor who commands that you quit smoking today, eat no more than 500 calories a day, and exercise three

hours a day does you no good. Smart physicians know that they can push, but not too far and only gradually. "Why don't you begin with 15 minutes on the treadmill every other day?" they say. It is not that 15 minutes of exercise are sufficient, but these 15 minutes are 15 minutes more than zero. They might push you to 30 minutes next month. Now, think about investors who say they are comfortable investing in Europe because they are familiar with Europe but they are not comfortable with investing in Asia. You want them to diversify to Asia as well as Europe, but today, your education fails. Let them diversify into Europe today and come back to the subject of diversification into Asia next quarter. Don't let the perfect be the enemy of the good. Question: Empirical studies have shown that investors do, in fact, behave irrationally. What strategies have been created to exploit this irrational behavior?

Statman: It might be possible to take advantage of the irrationality of others in the market. The most popular example is value investing. People tend to think that good stocks are the stocks of good companies, which drives up the prices of the stocks of good companies and sets the people up for disappointment. This phenomenon explains the lower returns of growth stocks relative to value stocks over the long run. But as anyone who invested in value stocks in the last few years knows, you can suffer a lot of pain over the short run. Similarly, you might be able to exploit the tendency of investors to extrapolate trends in the stock market. Fisher and I (1999a) showed that the sentiment of individual investors and the sentiment of Wall Street strategists are good contrary indicators. But be careful in trying to exploit the irrationalities of other people. You have plenty on your hands with the irrationality of your own investors.

References Barber, B., and T. Odean, 1998. "Boys Will Be Boys: Gender, Overconfidence, and Common Stock Investment." Working paper. University of California at Davis. Chernow, Ron. 1998. "Hard Charging Bulls and Red Flags." New YorkTimes (September 2):A31. Colas, Emily. 1998. Just Checking: Scenesfrom the Life of an Obsessive-Compulsive. New York: Pocket Books. Cosmides, Leda. 1985. "Deduction or Darwinian Algorithms? An Explanation of the 'Elusive' Content Effect on the Wason Selection Task." PhD. dissertation. Harvard University, University Microfilms. de Pontenay. P. 1996. "Long-Run Risk in Stocks." Financial Analysts Journal (Marchi April):73. Fisher, Kenneth, and Meir Statman. 1997. "Investment Advice from Mutual Fund Companies." Journal of Portfolio Management (Fall):9-25.

- - . 1999a. "The Sentiment of Investors, Large and Small." Working paper. Santa Clara University. - - . 199%. "A Behavioral Framework for Time Diversification." Financial Analysts Journal (MayIJune):88-97. Heath, Chip, and Amos Tversky. 1991. "Preferences and Beliefs: Ambiguity and Competence in Choice under Uncertainty." Journal ofRiskalldUncertainty Oanuary):5-28. Huberman, Gur. 1997. "Familiarity Breeds Investment." Working paper. Columbia University. Jeffrey, Robert. 1984. "A New Paradigm for Risk." Journal of Portfolio Management (Fall):33-40. Kahnernan, Daniel, and Amos Tversky. 1979. "Prospect Theory: An Analysis of Decision Making under Risk." Economei-

Lowenstein, Roger. 1997. "'97 Moral: Drop Global-Investing Bunk." Wall StreetJournal. (December 18):Cl. Samuelson, Paul. 1994. "The Long-Term Case for Equities: And How It Can Be Oversold." Journal of Portfolio Management (Fall):15-24. Statrnan, Meir. 1995. "Behavioral Framework for Dollar-Cost-Averaging." Journal of Portfolio Management (Fall):70-78. - _ . 1999. "Foreign Stocks in Behavioral Portfolios." Financial Analysts Journal (MarchiApril):12-16. Statman, Meir, and Steven Thorley. 1999. "Investor Overconfidence and Trading Volume." Working paper. Santa Clara University. Tversky, Amos, and Daniel Kahneman. 1992. "Advances in Prospect Theory: Cumulative Representation of Uncertainty." JournalofRiskand Llnccrtainiu (October):297-323.

rica:263-291.

©Association for Investment Management and Research

27

After-Tax Asset Allocation Nancy L. Jacob

President and Managing Partner Windermere Investment Associates

Investment managers for taxable clients have a fantastic tool-mean-variance optimization. After making some adjustments to this tool, managers can use it to add value in the asset allocation decision on an after-tax basis. The key is to look at the asset allocation decision along three dimensions-risk, return, and taxes. Taxable clients are now demanding this approach, and to remain competitive, managers must respond.

ne of the most important services a manager can provide individual clients is to deal with their asset allocation in a customized, after-tax setting. Academic and practitioner research has pointed out the value of proper asset allocation. However, most of the analytical models developed for asset allocation have been focused exclusively on tax-exempt portfolios. Few commercial optimizers permit customized asset allocation for people who pay taxes. That emphasis is about to change. Clients are demanding after-tax allocation. Taxable clients want their managers to minimize their taxes and to suggest asset allocation strategies that work well within estate-planning structures. Taxes are private clients' biggest expense: Taxes far exceed what they pay their managers, their consultants, or their custodians. Taxes must be considered in asset allocation for private clients, but taxes do not drive asset allocation strategy. This presentation addresses the consideration of taxes and other aspects of taxable investors' needs in the context of mean-variance optimization.

The application of mean-variance optimization to private clients' assets is different from its application for tax-free investors. Asset allocation by meanvariance optimization has been the approach of choice for optimal portfolio construction since the work of Harry Markowitz.' Mean-variance optimization is not used as much as it should be, however, for a number of reasons. First, many quantitative analysts object to using mean-variance optimization

because, even after enhancements, optimizers are still primarily applied in a single-period framework and assume away market inefficiencies (e.g., transaction costs and taxes). In addition, some academics say mean-variance optimization is not consistent with "pure" utility maximization, except in the extremely narrow case of quadratic utility functions. Mean-variance optimization does have limitations. The two main ones, according to Michaud, are statistical limitations and unstable inputs. 2 Both emanate from the statistical estimation process, not from the mean-variance optimization algorithm itself. Estimating means, correlations, and standard deviations is difficult, and optimization procedures tend to magnify errors in the estimates and create portfolios that are skewed in peculiar ways. Also, a certain ambiguity surrounds optimal portfolios; many portfolios can be found that are potentially close to the efficient frontier. Again, keep in mind that these limitations are inherent in any portfolio optimization process relying on statistical estimates, not in meanvariance optimization alone. Despite the problems, if managers have estimated or forecasted the parameters carefully, adjusted for statistical biases, and tested the meanvariance-efficient frontier to make sure it is robust, mean-variance optimization is a wonderful tool that can be used to great advantage for private clients. My interest in the practical aspect of applying mean-variance optimization to the taxable side of the investment management business came about when I was in consulting practice nearly 10 years ago and

IHarry Markowitz, Portfolio Selection: Efficient Diversification of Investments (New York:John Wiley & Sons, 1959).

2Richard Michaud, "The Markowitz Optimization Enigma: Is 'Optimized' Optimal?" Financial AnalystsJournal (january/February 1989):31-42.

28

©Association for Investment Management and Research

O

Mean-Variance Efficiency

After-Tax Asset Allocation our clients, who were rightly concerned about the enormous taxes they had to pay, would say, "This asset allocation stuff is great, you know, but the turnover is huge, and the checks we are writing to the Internal Revenue Service are enormous. At the end of the day, after taxes, our returns are like 'l-bill returns. Isn't there something you can do? How come you can't simply take an asset allocation model and do something after tax?" It is hard to argue with a client who is paying your salary. So, colleagues and I began developing an after-tax asset allocation model.l The most eye-opening aspects of after-tax optimization are, first, that portfolios that are optimal after tax are very different from portfolios that are optimal in a pretax world and, second, after-tax mean-variance optimization can actually add value after taxesan aspect that is very important to the client. Drawbacks. The drawbacks to mean-variance

optimization for private clients, however, are significant. First, implementing efficient portfolios triggers capital gains taxes. The portfolio generated by the optimizer may be a wonderful portfolio, but how does the client get there? The main problem with optimizers is that they do not take into account embedded gains (i.e., the difference between a portfolio's tax cost basis and its current market value). For instance, if a client has a portfolio with a large block of low-cost-basis stock, a manager cannot normally get rid of it without triggering the payment of huge capital gains taxes. Second, the efficient portfolios that most optimizers generate are tax inefficient and undesirable on an ongoing basis. They will include such classes as high-yield debt or a lot of low-volatility hedge funds whose after-tax returns are about the same as T-bills, Third, mean-variance optimization generally ignores the tax consequences of rebalancing. For a pension plan, a static mix is optimal. Given long-term forecasts for the various asset classes, the model will indicate the optimal asset mix. To a pension plan, managing this mix over time is not a problem. If the original mix is still optimal, the pension plan simply rebalances the portfolio back to the optimal mix. Private clients cannot do this. In addition, traditional optimizers cannot help evaluate whether a client should diversify out of some asset. For that advice, some other methodology is needed. Finally, conventional mean-variance analysis is single period and indifferent to time horizon. Unlike pension plans, which for the most part have an infi3Dr. Marc Moulton is the systems programmer who deserves the credit for writing the first known after-tax mean-varianceoptimization program. He authored PORTAX.

©Association for Investment Management and Research

nite time horizon, private clients have finite lives. Single-period analyses are inappropriate for private clients because if a client has, for example, a large block of low-cost-basis stock and wants to know whether diversifying out of it is optimal, the answer depends on whether the client's expected life span is a year or 40 years. If the expected life span is a year, the client can pass along the stock to his or her heirs with a step-up in basis and avoid ever paying that embedded gain. If the client is at an age where she or he can expect to live another 40 years, the client may be able to sen the stock and, over time, recoup the taxes paid. None of these factors are considered in typical asset allocation models. For private clients, a different approach from the current method of mean-variance optimization is needed. Complexities. In addition to taxes, private clients have characteristics that complicate optimization in various ways. No two clients are alike, so optimization must be customized for private clients to take into account their unique needs and circumstances. The structures used in estate planning can dictate the way assets are taxed, the applicable tax rates, or who pays the taxes. For example, under charitable remainder unit trusts (CRUTs), distributions are made to the grantor of the trust and the grantor is taxed on the amount of ordinary income generated in the CRUT, but the CRUT itself does not pay any taxes. In this case, the client can invest in taxinefficient strategies and build up the corpus (principal) but whenever it is distributed, the grantor pays taxes on that distribution. The effect is to introduce multilayer tax optimization and multilevel optimization of the client's portfolio. Private clients often have unique assets-the family farm, stock in privately held companies, or stock options. These kinds of assets are illiquid, and in many cases, they cannot be sold. They are "legacy assets," which some clients say they will never selL The manager must be able to indicate what kind of income these assets are expected to generate (or not generate) over time. Private clients may also be involved in private investment partnerships, such as hedge funds, which are difficult to model in a classical asset allocation approach. An after-tax optimizer will not help deal with these partnerships, but they must be considered when mean-variance optimization is used for private clients.

Before-Tax versus After-Tax Allocation For background to discussion of after-tax asset allo29

InvestmentCounseling for Private Clients cation, consider the differences between a traditional, "tax-ignorant" efficient frontier and the frontier for a typical private client's portfolio. Frontier 1 in Figure 1 is a traditional pretax efficient frontier that allows the sale of low-cost-basis assets. Frontier 2 is the frontier for portfolios that must hold the low-costbasis assets to avoid taxes. A client's current portfolio, which includes a large block of low-cost-basis stock, is shown on Frontier 2 and is the starting point in this analysis for considering how to improve taxable clients' returns. If the manager uses a normal optimizer and allows the optimizer free rein over what assets to buy and sell, the result is Frontier 1, which allows the sale of the low-cost-basis asset. Although the client gets much better portfolios than the current portfolio, the problem is that the client cannot get to Frontier 1 because doing so would require paying a lot of taxes. The next step might be to force the optimizer to hold the low-cost-basis asset and then model the efficient frontier, which reproduces Frontier 2, but why would the client change anything when the current portfolio is already on that efficient frontier? What the manager really wants to do is get to the portfolios between Frontier 1 and Frontier 2, where some amount of selling and reinvesting in other assets is optimal. The after-tax asset allocation methodology incorporates one-time gains, ongoing tax impacts, and the client's unique tax status. The after-tax asset allocation model has to recognize the one-time tax consequences associated with restructuring a portfolio. To incur the tax, the portfolio does not have to contain a large block of low-cost-basis stock; any existing port-

folio may contain embedded gains. If a manager wants a client's business, the manager has to be able to recommend a plan that improves upon the current portfolio's performance net of the realization of those embedded gains. Otherwise, the existing manager is probably going to keep the portfolio. So, the recommended plan must deal with the ongoing tax impact of each asset manager's style and the portfolio's characteristics. Suppose you are advising a client and investing the client's money in a series of mutual funds. The manager of each mutual fund has some amount of turnover, because of the manager's particular style, that will be either tax efficient (will generate low taxes) or tax inefficient (will generate high taxes). The advisor has to evaluate those aspects and embed the effects of styles in the asset allocation. The advisor also has to take into account the ongoing tax impact of periodically rebalancing the mix. Rarely will a static mix apply to a client over the long term, particularly if the strategy is to work down a large block of low-cost-basis stock. The advisor may want to move the portfolio toward a target mix, and reaching that target may take five years. Finally, each private client has a unique tax status. Each has his or her own marginal tax rate, loss carryforwards, and tax deferrals. Some clients are subject to the alternative minimum tax (AMT). The advisor must customize each allocation to deal with every aspect. Optimization. PORTAX is an after-tax personalcomputer-based mean-variance optimization

Figure 1. Tax-Ignorant Efficient Frontier and Frontier with Low-Cost-Basis Asset 16 14

.e : .. , • Frontier 2

:J:

"---

S 12 E iii

Frontier 1 .111

~rent Portfolio

~

-e 2u

10

iii

0..

x

~

>
$55, Buys Stock

Stock < $55, Exercises Put

~

~

Broker/ Dealer

--------

Receives Stock

G 75

Investment Counseling for Private Clients Strategy 1 is to harvest stocks if the marginal benefit exceeds the marginal cost of transacting. Strategy 1 is applicable if the marginal tax benefit for realized losses is 25 percent of the realized loss and the marginal cost is 1 percent of the notional value of the stock sold (including round-trip commission costs and the market impact of trading). Strategy 2 involves minimizing transaction costs versus the value of losses realized by harvesting stocks with the largest percentage loss first until the desired amount of losses is harvested or a target level of turnover is reached. This strategy can apply whether or not the investor uses put options. Strategy 3 involves harvesting stocks with the largest dollar loss first until the desired amount of losses is harvested or a target level of turnover is reached. Strategy 3 is less efficient than Strategies 1 or 2. As the comparison in Table 1 shows, if 95 percent or less of the available losses are harvested, Strategy 2 provides the most efficient means of harvesting. In Strategy 2, an investor can harvest 85 percent of the losses by selling 53 percent of the value of stocks with losses. Thus, the investor will realize the highest percentage of losses among the smallest percentage of stocks. Figure 3 shows the relationship between the percentage of realized losses and the percentage of stocks with harvested losses for the three strategies. Unless the percentage of available losses exceeds 95 percent, Strategy 2 remains the best tax-

Table 1. Percentage of Stocks with Losses Harvested Strategy 1 (MB> MC)a

Percent of Losses

Strategy 2 (by % loss)

(by $ loss)

Strategy 3

80

57.7

46.9

64.2

85

63.1

90

67.5

95 99

87.3

52.9 60.3 70.6 87.2

70.0 75.4 80.9 90.5

73.9

"Marginal benefit exceeds marginal cost of transacting.

loss-harvesting strategy. Managing Exposure during the 31-Day Period. At least three approaches exist to mitigate the exposure during the 31-day waiting period needed to avoid a wash sale: holding cash, buying SPDRs, and selling a put option on the stocks to be repurchased in 31 days. Each strategy has distinct advantages and disadvantages. Holding cash. This approach is quite simple for investors; it involves no transaction costs and no loss of principal if the market declines. But investors will experience a drag on performance if the market rebounds. Holding cash will also earn investors less than selling put options. BuyingSPDRs. This method reduces tracking risk relative to a benchmark and provides market

Figure 3. Performance of Strategies for Harvesting Tax Losses 110 ~-----------------------------,

100

/,

80

Strategy 3 .'

Strategy 1

.

20

o w:.... o

-'

20

~J.

40

.

60

80

100

Percent of Stocks with Losses Harvested

Note: Indicative pricing as of October 7, 1998.

76

©Association for Investment Management and Research

Equity Derivative Strategies exposure if the market rebounds. Drawbacks include the transaction costs to buy and sell SPDR shares and the condition that buying SPDRs provides no offset to stock-repurchase costs. Selling puts. An investor who follows this approach will obtain an up-front option premium, which lowers the investor's cost basis. Transferring stock to a dealer (as shown in Figure 2) reduces the market impact of selling stock. Complexity and performance lags if the market rises rapidly are major disadvantages of selling puts. In addition, the effective purchase price of reestablishing stock exposure may be above the current market price if the market declines sharply. Figure 4 demonstrates the payoff from selling puts for a client who wants to sell a put option on a $50 million basket of 300 stocks with a dividend yield of 1.95 percent to be repurchased in 31 days. If the stock falls significantly during the 31-day blackout period, the investor is better off simply liquidating the position. With the current basket price at $100, the client sells a put with a $107strike price for a premium of $8. Table 2 indicates that as long as the basket remains below $107, the effective cost of reestablishing the basket of stocks is $99. At prices above $107, the put option expires worthless but the initial premium helps offset the cost to repurchase the stocks.

Table 2. Example Strategy of Selling Puts Basket Price $ 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110

Put Value

Purchase Price

Effective Price

$15 14 13 12

$107 107 107 107 107 107 107 107 107 107 107 107 107 107 107 107 108 109 110

$ 99 99 99 99

11

10 9 8 7 6 5 4 3 2 1 0 0 0 0

99

99 99 99 99 99 99

99 99 99 99 99 100 101 102

Note: Indicative pricing as of October 7, 1998.

Conclusion Equity derivatives are useful instruments for maximizing the after-tax returns of a taxable investor's

Figure 4. Payoff from Seiling Puts 115

110

Purchase Price if Put Not Sold

-,

t

Effective Price if Put Sold 90

85 L-_---l._ _--L..._-.J_ _ _ _..l...__ 92 94 96 98 100 102 ~

_..l.._ __ ' __

104

106

_ _ '_ __'___

108

110

__'

112

Market Price of Basket ($)

©Association for Investment Management and Research

n

Investment Counseling for PrivateClients portfolio because they offer alternatives to highdividend stocks, provide a way to implement asset allocation shifts, or allow the creation of synthetic index exposure. Equity derivative strategies can also be used to efficiently convert short-term capital gains into long-term capital gains, reduce the implementation risk as a tax-lass-harvesting strategy by helping investors manage market risk during the 31day lock-up period associated with avoiding wash sales, and maximize the benefits of tax loss harvesting.

78

Investors and portfolio managers need to understand the general tax issues surrounding equity derivatives before considering using them. Important considerations are the rules involving straddles, qualified covered calls, and wash sales. Also vital to effective use of equity derivatives is an understanding of what each instrument or strategy can and cannot do; the benefits, risks, and costs of each strategy; and most importantly, the tax implications of alternative strategies.

©Association for Investment Management and Research

Equity Derivative Strategies

Question and Answer Session Joanne M. Hill Carmen Greco 2 Question: Are SPDRs considered Section 1256 contracts? Hill: No. Section 1256 contracts include futures or options that trade on a listed exchange. Section 1256 contracts were initially created to apply to commodities futures contracts. When index options started trading on the Chicago Board Options Exchange, the CBOE wanted to make them comparable in terms of tax treatment to futures index options that trade on the Chicago Mercantile Exchange. Initially, the Section 1256 contracts were brought in to include index options, but for tax purposes, SPDRs are treated like the purchase or sale of a stock. The primary disadvantage of SPDRs is that you pay a commission to purchase one, but if you hold the position a long time, the commission is amortized over a long period of time. When an investor buys a SPDR, the investor does not receive exactly the same return as the S&P 500. For example, if you mark the SPDR at the end of a month and calculate the tracking error of that SPDR position to the S&P SOD return, the error will be 1O()-.140 basis points. In other words, a portfolio performance report at the end of any month for SPDR positions will not be marked where the S&P 500 closes that month but where the SPDR closes that month. If you accumulate enough SPDR shares, the shares can be converted into units of the trust and you can actually take physical delivery of the 2Carmen Greco, who works with Ms. Hill in the area of equity derivative strategies at Goldman Sachs, joined Ms. Hill for this question and answer session.

trust holdings. Large institutions often arbitrage SPDRs by accumulating a number of units and converting them. You have to take tracking risk to hedge the market risk of a portfolio that isn't a perfect replica of the benchmark index. So, SPDRs provide the tax advantage but take away the tracking benefit. SPDRs trade at different prices from the index because they track futures, in the following sense: If the market goes down sharply on the last day of the month, futures will go out cheap and end up selling below where they should. Why? Because traders are in the market hedging since they cannot easily short a stock portfolio. There is a plus tick rule, but investors can sell futures, so futures are trading cheap. The market maker of the SPDRs, who is a specialist on the floor of the exchange, will now adjust the SPDR price based on where the market maker can hedge his or her position in the futures market. So, the market maker is making a market based on the hedging instrument, which is the futures contract. SPDRs will go out a little bit below the index on a big down day or above the index on a big up day. On a big up day, everybody comes in and buys SPDRs because they want to capture that move. So, a potential marking risk is present at the end of a, say, quarterly performance period that introduces some tracking error. That error washes out, however, over an annual period. Greco: An advantage of SPDRs is that you don't need a plus tick to sell them short. So, when managers are not allowed to use derivative products, many of

©Association for Investment Management and Research

them use SPDRs for hedging because they don't want any extra paperwork and their charters allow the use of SPDRs. They sen SPDRs short for asset allocation and hedging purposes. Question: Is shorting a SPDR against a diversified portfolio considered a straddle? Hill: If that portfolio contains more than 70 percent of the market cap of the S&P 500, then shorting the SPDR constitutes a straddle. The 70 percent rule also applies if you go on the other side-long a SPDR and short a diversified portfolio.

Question: In selling call options to convert short-term capital gains to long-term term capital gains, if you write a $52-strike call and the stock immediately goes through the strike price, are the capital gains on the stock still long term? Hill: Going through the strike price has no effect as long as the option holder does not exercise the option. But if the option gets deep enough in the money that someone exercises it prior to expiration (12 months), then the other side will be forced to sell the stock. Exercising an option early is irrational (it'snever optimal to exercise options early from an option-pricing standpoint), but such things do happen, especially before a dividend payment.

Greco: A physically settled option entails no tax on the option. Options take on the holdingperiod characteristics of the stock position, so if you enter into a physically settled option, you actually deliver your stock. If the

19

Investment Counselingjor Private Clients stock is held for 12 months and a day, it will trigger a long-term capital gain that is equal to the exit strike price plus a premium. For example, if you sell the $50-strike call for $5 and you have a zero cost basis, you're facing a $55long-term capital gain, but you can convert long-term gains to short-term gains if you settle the option for cash. To remain flexible, however, you need to be leery of physical settlement. A customized option allows you to stay flexible as long as you don't have restricted stock. Hedging with put options and short calls (collars) to minimize concentration risk requires liquidity in the security. What other strategies are feasible for relatively illiquid securities? Question:

Stock borrowing and liquidity are the first things we look at in terms of engaging in a strategy because we do not take a directional view on the stock. We hedge our positions, and that practice is pretty consistent throughout the investment community. If someone can't borrow the stock and the underlying stock has no liquidity, people who take on a counterparty position will not be able to maintain an economically neutral position throughout the life of the trade, so they will not engage in the trade in the first place. Greco:

What are some of the risks that one should be wary of in entering into these kinds of contracts? Question:

Hill: First, there is documentation risk. You should make sure that your client is educated about all the features of these derivatives and that your investment advisory agreement dearly authorizes you to engage in options and/ or futures transactions. It is one thing if the market falls 50 percent and the client owns a stock portfolio

80

and is not happy. There is not much the client can do about it. But if a client has a loss on a derivative contract, the client will probably look for ways to get out of realizing that loss. And the client might say that you were not authorized to trade derivatives. Marked-to-market risk tips the scales in favor of trading listed or exchange-traded instruments. Listed markets broadly disseminate closing prices, of course. The possibility always exists that the option market will be disrupted in an emergency situation, but recent revisions in the circuit breakers reduce the likelihood that a closing price cannot be established at the end of each trading day. Growth in stock option volume has been 2(}-30 percent in the past three years. Stock options are regularly used as part of corporate buyback programs, hedging singlestock risk, and covered call writing. A good balance exists between buyers and sellers in single-stock options, so options don't contain much hidden risk-s-as long as you are using options in an unleveraged way. Some kind of leverage test is always a good idea, which does not mean that you should never use leverage. If you are using leverage, however, you have to make sure you are authorized to do so and are not doing it surreptitiously through a derivative. Question: What approach to recognize which strategies are legitimate would you suggest for investment advisors who are not well practiced in the use of options and derivatives?

Hill: First, try to involve a broad group of people in the organization, induding operations, legal, and custody staff. Not everyone needs to know all the nuances of derivatives, but you should identify at least two people in each area of the organization to become

knowledgeable about derivatives and then have them go to brokers, exchanges, industry educational organizations, or experts for education. AIMR, the Futures Industry Institute, and the Options Industry Institute regularly run educational programs on derivative strategies. Second, one or two in-house staffon the trading desk and in research-should specialize in equity or fixed-income derivatives to be a resource to the portfolio managers in helping them with strategies. Another area where you need a derivatives expert is in the risk management group. Start-up costs are involved, but you need people in your organization who spend at least 50 percent of their time on risk management. Greco: One of the best books on options is the reference book Options asa Strategic Investment. 3 It is long on strategies and short on math, and it does a good job of explaining how strategies work, how to implement them, how to get out of them when things go wrong, and (because it is written for the high-net-worth individual) how to optimize returns from a tax standpoint. Question: Are investment advisory firms or trust companies using derivative strategies to control fee revenue?

Hill: Yes, somewhat, but the practice is not widespread. We have received a number of inquiries about hedging fee income, but fewer than 10 percent of institutions use derivative strategies to control fee revenue. The logic of such hedging is dear: The fees that many investment management organizations 3Lawrence G. McMillan, Optionsas a StrategicInvestment: A Comparative Analysis of ListedOptionsStrategies, 3rd ed. (New York: New York Institute of Finance, 1993).

©Association for Investment Management and Research

EquityDerivative Strategies earn are based on the portfolios they manage. Their fees are sensitive to the gains and losses on equities (and on fixed-income instruments, to some extent, if they are fixed-income managers). And if a firm has sensitivity to the equity market, it can hedge the risk with an index option. The reason for not using derivative strategies to hedge fee income is that shareholders are buying the

stock of public securities management firms precisely for their sensitivity to certain markets. For example, investors buy T. Rowe Price because it is an equity-related firm. If you hedge away the market sensitivity, you change the nature of the firm and investors are no longer buying an equity-related firm. The transactions that we've done tend to be for financial institutions that are subsidiaries of

©Association for Investment Management and Research

larger entities, publidy traded entities, or non-publicly-traded entities with budget goals. Most of the cases we have been involved in have been situations in which a firm is midway through the year, is wen in excess of its budgeted fee income, and is wining to pay a little amount of the excess of the fee budget or target to make sure those gains will be realized or to hold those gains through the year.

81

Estate and Tax Planning William R. Levy Executive Vice President Brown Brothers Harriman Trust Company of Pennsylvania

Combining the investment plan with the estate plan is critical to maximIzmg performance for taxable u.s. clients. A variety of trusts can reduce or defer tax liabilities and create new funds for investment. Lifetime gifting is another easy way to transfer wealth and keep growth in assets beyond the reach of the long arm of the U.S. Internal Revenue Service. Guidance from experts possessing a thorough understanding of estate and tax planning tools is necessary to derive the maximum benefit from their use and to avoid potential pitfalls.

tional dollar of appreciation or accumulated income at the time of one's death. Investment advisors who recognize this reality up front will understand how some of the techniques and ideas discussed in this presentation can significantly enhance the value of what they are doing for their clients.

state and tax planning should be a subject of great interest to the wealthy clients of investment advisors. In 1998,a couple with a joint net worth in the range of $1,250,000 could pass their combined assets to the next generation at death (or through lifetime gifts) free of federal estate and gift taxes through the proper use of their individual $625,000 unified credit exemptions. This unified credit threshold is increasing progressively each year, and by 2006, the credit equivalency will be $1 million per individual. Thus, a couple will be able to move $2 million to the next generation in the year 2006 with no diminishment resulting from federal estate and gift taxes. Unfortunately, many clients whose net worth exceeds the unified credit exemption are looking at a federal estate tax bill of 50 percent or more. Most clients simply cannot believe that the U'S. Internal Revenue Service (IRS) will take such a significant portion of their assets before the money passes down to their heirs. Estate planners and investment advisors serve the same clients and are equally concerned about their financial health, but they approach things from different perspectives. Keep in mind that preserving the augmented wealth that investment managers hope to generate for their clients is directly related to the estate and tax planning done in preparation for the inevitable taxes coming down the road. When one begins estate planning, a reasonable assumption is that the federal estate tax rate will be 50 percent (once the estate exceeds the unified credit exemption amount) and that, from today on, the government will receive 50 percent of every addi-

Annual Exclusion Gifts. In a lifetime gifting program, every individual is permitted to give $10,000 annually to as many other individuals as she or he chooses. For married clients, the threshold is $20,000 per person per year. The money can come from his, her, or their assets. If it all comes from one spouse, that spouse can use the other spouse's annual exclusion and apply it to the gift. It may make sense for the husband, say, to give the money to the wife first and then have them each cut a $10,000 check or to put the money in a joint account and make the gifts from that account. This approach will avoid the filing of a gift tax return in which one spouse joins in the other spouse's gift. For example, suppose a mother and father want to provide a down payment for a married child's

82

©Association for Investment Management and Research

E

Lifetime Gifting One of the easiest and most effective ways to reduce death taxes is to enter into a lifetime gifting program to transfer wealth before death at a significantly diminished tax cost and to separate future appreciation of the assets from the reach of the tax authorities. All clients who have reasonably substantial net worth should be carrying out such gifting now and doing it as aggressively as possible.

Estate and Tax Planning house. They need to think through the plan and carry it out in such a way as to maximize the tax benefit. Suppose the purchase date is February and we are in December. The mother and father could give $20,000 to the son and $20,000 to the daughter-in-law in December and do it again in January. These gifts would allow the son and daughter-in-law to have an $80,000tax-free gift to apply toward the purchase of their house. Now, suppose the children in the same scenario need a little extra to purchase the home. If the parents have other children, they can provide the extra by maximizing the gifts to the other children, who would then turn around and make gifts to their sibling and his wife that year; the idea would be to reverse the process in later years, but the siblings cannot be required to make the gifts. The gift tax law allows a person to give up to $10,000a year to as many individuals, including minors, as that person wants. The earlier a gift program is initiated, the better, because lifetime gifting may also remove future appreciation from one's estate. If parents give away multiple $10,000 and $20,000 gifts to their family members, any appreciation of those assets that occurs from the time of the gift until the time the donors die has, therefore, avoided the 50 percent estate tax. So, lifetime gifting should begin as soon as the donors are comfortable enough to start parting with their hard-earned assets. The $10,000gift is one of the easiest things to do because it entails no reporting requirements, no gift tax obligation for the donor, and no tax obligation for the recipient. Tuition and Health Care Gifts. Tuition and health care expenses provide another way to reduce estate and gift taxes. For example, a grandparent may pay the tuition for a private school, college, or any type of educational service for any individual whom the grandparent does not have the responsibility to support, provided those payments are made directly to the third-party provider. That is, the grandparent's checks should be made payable directly to the school and the gift should not pass through the parents. It is an unlimited, tax-free gift, so if the grandparent pays three $20,000 college tuitions, that money is considered a tax-free transfer out of the estate of the donor and has no other tax implications. A client can also pay for health care expenses of an adult child or a grandchild. If your clients' adult children are paying premiums for health care insurance, for example, a grandparent or other relative can step in and pay those insurance expenses, and the transfer is free of gift tax. Again, the payments must be made directly to the doctor, institution, or insurance carrier.

©Association for Investment Management and Research

Tuition and health care gifts can pass free of gift tax in addition to the $10,000and $20,000annual gifts, so this process is a wonderful opportunity to pass additional property down to the next generation with no tax. Split-Interest Gifts. Investment advisors may be familiar with certain terms that refer to splitinterest gifts-grantor-retained income trusts (GRITs), grantor-retained annuity trusts (GRATs), and grantor-retained unit trusts (GRUTs). For example, GRUTs work as follows: A parent transfers a piece of real estate into a trust and has the opportunity to live on the property for seven years. After that seven-year period, the property passes to the children. This gift is taxable, but the value of the gift for tax purposes is only the value of the remainder interest. Thus, the GRUT can reduce the taxable value of a $1 million property transferred to the trust by the actuarial value of a seven-year right to live in it (possibly a 30-40 percent discount). The process allows someone to reduce the value of the gift that is subject to tax when gifting the property down to the family; the heirs simply do not have the right to receive it for seven years. Although this concept is fairly sophisticated, it can significantly reduce the ultimate transfer tax for which a client will be responsible. Another sophisticated way to leverage the gift tax is with a family limited partnership, which moves property to the next generation or generations at a significantly discounted value which, in tum, ultimately reduces the transfer tax.

Taxable Gifts Clients can save additional transfer taxes in the long run by taking the tax hit on a lifetime gift. Clients may be reluctant to make wealth transfers that require paying a gift tax now, but advisors can explain the importance of recognizing that taxable giving provides another opportunity to preserve wealth. Clients who have substantial net worth need to recognize the ultimate tax hit that occurs when the property passes on to their heirs at death. For example, if a client has exhausted her unified credit during her life and wants to leave $1 million to her heirs after the estate tax is assessed, her estate will need to be at least $2 million. The reason is that the estate will be paying a tax on the money used to pay the IRS as well as on the money left to the heirs, which seems almost unconscionable. On the other hand, a gift tax paid on assets that the client has conveyed during her lifetime is not subject to further taxes. Thus, from a tax perspective, for the client to make taxable gifts while alive is significantly cheaper than paying estate taxes. 83

Investment Counseling for Private Clients In summary, to pass $1 million to heirs at the client's death requires $2 million ($1 million for the heirs, $1 million for the IRS); to pass $1 million to the client's family during the client's life requires $1.5 million ($1 million for the heirs, $500,000 for the IRS). When clients recognize the tax benefit of lifetime versus testamentary gifting, they should be motivated to begin getting assets out of their estates now.

Spousal Assets When a husband and wife have very disparate values in their taxable estates, the estate should be divided up so that each has in his or her own name an amount at least equal to the unified credit equivalency$625,000in 1998and working its way up to $1 million by 2006. If most of the assets are in the husband's name and the wife dies first without an individual net worth equal to the unified credit equivalency in her estate, the couple has wasted a substantial tax shelter. The IRS will give a taxpayer the benefit of that $625,000exemption only if the assets are solely in the name of the decedent. Assets owned by a husband and wife jointly will not provide any benefit because, although they pass to the surviving spouse without tax, they are subject to tax at the survivor's death. So, assuming the marriage is reasonably solid, the couple should arrange that each spouse have at least the threshold amount in her own name and his own name and should maintain that amount until it reaches $1 million in 2006. From a tax perspective, how they want to hold the balance of their assets does not matter. So long as each spouse has $625,000 worth of assets titled in his or her name alone and a properly drawn will or deed of trust for transferring the assets at death, the unified credit will not be wasted.

Dotting i's/Crossing t's Wealthy individuals are sometimes reluctant to pay qualified advisors, lawyers, and managers to assist with their estate planning, but families and heirs suffer if the older generation has not received the appropriate professional advice in preparing documents and carrying out suggested procedures correctly.

trust may have a provision that the trustees can use trust property for a child or a grandchild in the event of serious need. Suddenly, the whole marital deduction is blown. If the couple had professionally drawn documents, chances are they would be able to benefit from the generous tax deferral that the IRS offers for property passing between spouses. Generation-Skipping Transfers. Another important opportunity for tax avoidance is in the area of generation-skipping transfers. It applies, for instance, in a family of three generations in which the first generation is dying off and wants to move property to the third generation without it being taxed at the death of the second generation. That is, the family wants to eliminate the 50 percent tax and allow the money to pass tax free down to the grandchildren. The provisions in such trusts and the actual language in the will must be very specific because it is often carefully scrutinized by the IRS. In generation-skipping trusts, spending the extra time and money to have advisors who really know what they are doing is extremely important. Failure to do so may result in the imposition of confiscatory taxes. Irrevocable Life Insurance Trust. If the trust is properly structured, the IRSdoes not tax the proceeds of life insurance held by an irrevocable trust for the benefit of the decedent's heirs. If the trust is not properly drawn or not carefully administered, clients face potential estate tax issues and possible gift tax issues. Gift Tax Returns. The failure to properly prepare and file gift tax returns may open the door to serious problems. If a client makes a gift during life and pays taxes on it, the valuation of that gift and the proper filing of that gift tax return is what triggers the statute of limitations. If a client has given away a closely held business interest or a piece of real estate and filed the gift tax in a timely manner, with the proper valuation and with the appropriate substantiation regarding the valuation, and the client survives for at least three years without an audit, the IRSis stuck with that valuation. If the client decides not to worry about properly filing a gift tax return, the IRS can always question the value of the gifted property. Hopefully, the heirs can point out that a gift tax valuation was reported to the IRS and the IRS never questioned it within the statue of limitations. In this area, an expert accountant may be equally as important as a lawyer in making sure these steps are carried out properly.

Marital Deductions. A failure to seek and follow professional advice is particularly dangerous in the establishment of the marital deduction trust. The difference between doing it wrong and doing it right is the difference between paying the fullest measure of tax or no tax. A key point is to make sure provision is made in the marital trust that no one other than the surviving spouse can benefit from property that qualifies for the marital deduction. Such a provision seems simple, but errors do occur. For example, the

The charitable deduction has the potential to be the most effective tax-saving device for many people. The deduction may be leveraged-in some cases,

84

©Association for Investment Management and Research

Using Charitable Deductions

Estate and Tax Planning from both an estate tax standpoint and an income tax standpoint-in such a way that clients feel as if they are actually benefiting financially from the transaction as well as enjoying the feeling of being a philanthropist. In the charitable remainder trust, a donor puts property, usually appreciated assets, into a trust to provide income for a period of time (life of the client or a term of years) and then the remainder interest is passed to charity. With a charitable lead trust, the donor puts money into a trust and provides for a charity to receive an annuity or a distribution amount for a certain period; at the end of the period, the money passes on to the donor's heirs. A private foundation allows clients to make gifts to an entity that they continue to control, but the beneficiaries of that new entity must be public charities. In many situations, a client is happy to set up a private foundation, and this mechanism is even more effective when it is funded with retirement benefits, as is more fully discussed in the next section.

Protecting Retirement Plan Assets The assets that are potentially subject to the most severe tax at one's death are retirement benefitsindividual retirement accounts (IRAs), 401(k) plans, and qualified pension and profit-sharing interests in which the individual's plan assets will be passing to a designated beneficiary. The IRS gives an employee significant income tax breaks during life and during the retirement years of the employee and the employee's spouse, when the individual need pay income taxes only on assets taken out of these plans. At the same time, the balance of the funds accumulate and earn income tax free. When the retiree and spouse die, however, the IRS hits the estate of the surviving plan beneficiary with the full force of income and estate taxes. In many cases, the estate beneficiaries receive all the taxable income in one year, which can push the heirs from, perhaps, the middle income tax brackets up to the top brackets. When an individual has substantial other assets and a significant retirement plan, what makes the most sense is for the client to make charitable gifts at death with those retirement plan assets. In fact, retirement monies should be the first assets selected for use in charitable giving at death. In addition to avoiding a 50 percent estate tax, charitable giving also avoids any income taxes. This mechanism can avoid taxes of up to 85percent of the value of the retirement benefits. By passing 100 percent of the retirement assets into a private foundation, the family gives up as little as 15 percent of the value of a retirement plan and has 100 percent of the retirement funds to use in a private ©Association for Investment Managementand Research

foundation. The family completely controls that foundation in perpetuity and can even receive some fees for acting as trustees. This concept provides a number of benefits to the donor's family, and only the IRS is a loser. Two case studies may help investment advisors relate the trust ideas discussed here to some of the clients with whom they work. Case StUdy A. Mr. Generous is a 65-year-old executive who has recently retired from his position in XYZ Corporation with a substantial block of XYZ stock. His cost basis is less than $1 a share, and his stock, which has appreciated significantly over the years, is now valued at $50a share. The stock has never paid a dividend of more than 1 percent annually. Mr. and Mrs. Generous's net worth is more than $4 million, but that amount includes their home, from which they receive no income, and investments, artwork, and tangibles that do not pay them any income. They are concerned about generating enough income to retire comfortably. If they sell their investments, particularly their XYZ stock, they will pay a significant capital gains tax and will not have as much to reinvest in fixed-income investments. They have one son, Thomas, who is an attorney. He is married, has one child, and is doing well. The Generous family is also interested in preserving the environment, and they spend their summers camping throughout the United States. Planning suggestion. Mr. Generous should consider a charitable remainder trust, in which he would transfer $1 million worth of XYZ stock to a trustee that would provide distributions to him and his wife during their lives at an annual rate of 7 percent. At the death of the surviving Generous spouse, the remainder would pass to the Sierra Club. Benefits. The stock could be sold by the trustee and generate $1 million in proceeds to be reinvested in a diversified portfolio of stocks and bonds. None of the proceeds would be subject to capital gains taxes because the ultimate beneficiary is a charity. The Cenerouses would have the full $1 million to reinvest, and its full earning power would be available to Mr. and Mrs. Generous for the rest of their lives. Exhibit 1 shows the breakdown of this plan for an older couple with a 7 percent annual payout-the plan for Mr. and Mrs. Generous. The key pieces of information needed for this kind of planning are the amount being contributed, the percentage annual distribution, and the ages of the beneficiaries. The calculations in Exhibit 1 were run in September 1998,so the $236,030 deduction may not be entirely precise; the remainder factor that the IRS uses to determine these numbers changes every month. When the remainder factor was taken into account for this example, the calculations produced a little more than $236,000 as the charitable deduction in the year of the gift. 85

InvestmentCounseling for Private Clients Exhibit 1. Charitable Remainder Unit Trust for Mr. and Mrs. Generous Type of calculation Transfer date Section 7520rate Fair market value of trust Rate of annuity Payment periods in year Number of months valuation date precedes first payout Ages Payout frequency factor Adjusted payout rate Remainder factor Present value of remainder interest = $1 million x 0.23603 Donor's deduction Donor's deduction as percent of amount transferred

Life September 19, 1998 6.8% $1 million 7% Annual

o 65 and 64 1 7% 0.23603 $236,030 $236,030 23.603%

Reducing the distribution from 7 percent to 5 percent would significantly increase the deduction from the $236,030 amount given in Exhibit 1; higher interest rates would reduce the value of the deduction. If the couple were much older, that would increase the amount of the deduction. An advisor to this couple would have to evaluate the numbers so the client could understand the benefits of carrying out this plan. For a 64-year-old individual in a situation where the plan is going to run for the life of the survivor, a fair number of years is involved, which explains why the deduction may be smaller than people might assume. Nevertheless, it is still a significant tax deduction and benefit for the donor. As Exhibit 1 indicates, Mr. and Mrs. Generous would receive annual distributions equal in the first year to $70,000, or 7 percent of $1 million, and in subsequent years equal to 7 percent of whatever the trust was worth each year. If the proceeds from the stock sale were invested in a diversified portfolio that achieved 10 percent annual growth, the Generouses would each year have 10 percent more income. That amount would be significantly greater than the $10,000a year they are receiving now from their XYZ stock, and they would have the full value of the proceeds and its appreciation to earn income for them for the rest of their lives. Mr. and Mrs. Generous would receive an income tax deduction of approximately $236,000in the year they made the gift, or funded the trust, which could be used in that year if they had sufficient income or could be used over five subsequent years (because there are limitations on how much of a charitable deduction can be taken in anyone year). For example, if they put $1 million of the XYZ stock in the charitable remainder trust in Year 1 and received this nice deduction, they could then sell, say, another $250,000 of their XYZ stock. Even though they would have a big capital gain that year, they would have a big tax shelter against which to offset that gain. In addition

Case Study B. Mrs. Bigbucks has accumulated a significant investment portfolio; her net worth now is more than $10 million. She is a widow in her early 60s with three children and eight grandchildren. Her greatest concern is that, upon her death, about 50 percent of the estate will have to be paid to the IRS.She is not concerned abouther ownfinancial security, butshe is dedicated to leaving as much as possible to her family after her death. She has already committed to a program of annual exclusion gifts to her children, her inlaws, and her grandchildren. She is also paying tuition and some health care expenses on their behalf. She has always been charitably inclined and has established a

86

©Association for Investment Management and Research

to getting the income tax deduction for the creation of the charitable trust, they would have several alternatives for using the benefits of sheltering gains from the sale of assets they are going to keep. The gifted stock that was put into the charitable remainder trust, together with any appreciation in that charitable trust, would be exempt from any estate taxes at the death of the surviving Generous spouse. (This approach can be suggested in this scenario because Thomas, who is doing fine on his own, will not need every nickel of his inheritance.) The only downside is that Mr. and Mrs. Generous have taken some money away from Thomas and his family, but not a significant amount considering that they have increased their income dramatically for their lives so they will not have to use other assets to support themselves. In this plan, the Sierra Club would receive a substantial contribution at the death of Mr. and Mrs. Generous. (The Generouses could gear that contribution even more specifically to their particular interests-for example, interest in a particular national park for which they wanted to do a particular project.) The family, the Sierra Club, and those who value preservation of the environment would be the ultimate beneficiaries.

Estate and Tax Planning family foundation that she has begun to fund. Her primary concern remains the needs of her family. Planning suggestion. Mrs. Bigbucks has an opportunity to save significant federal, state, and generation-skipping taxes by funding a charitable lead unit trust with $1million to provide a 7 percent annual distribution to her family foundation for a period of 10 years, with the remainder payable to a generationskipping trust for her children and grandchildren. Benefits. Exhibit 2 shows exactly how the plan would work. The family would save at least $250,000 in federal estate and gift taxes in exchange for a possible deferral of their inheritance for up to 10 years. Mrs. Bigbucks will probably live those 10 years anyway, so the inheritance may not be deferred at all. If she lives 10 years, the plan will have no negative impact on her family, although she has basically separated herself from that $1 million for 10 years and receives no benefits from it. In this plan, the family would save $250,000 in taxes immediately because, normally, if Mrs. Bigbucks left the $1 million to her family, it would be taxed at 50 percent. The family foundation would receive at least $700,000in contributions over a lO-year period, which would enable the entire family to work together to carry out Mrs. Bigbucks's charitable inclinations. She has already set up the foundation and has begun to fund it, but the terms of the trust suggested here provide that, each year, 7 percent of the value will be paid as additional contributions to the foundation, which the family would control and from which it would distribute to whatever charities the family members support. Clients love the idea that they will be getting together once or twice a year to see each other and make some collective family decisions about founda-

tion distributions. From the perspective of a grandparent, this aspect is often the key factor in the whole plan because it is a great way to keep everybody together, do a wonderful thing for worthy charities, and get a significant tax savings for the family. If Mrs. Bigbucks lives for the lO-year period, she will have avoided any estate tax on any unspent income and the appreciation on the $1 million. If she holds on to the $1 million and lives 10 years, that $1 million could be worth $2 million to $3 million, and it could have generated several hundred thousand dollars in income, all of which would be subject to a 50 percent tax at the time of her death. We now have sheltered all those assets from estate tax. If a client wants to save taxes in the long run, this mechanism, with these multiple benefits, is a wonderful way to do it. Mrs. Bigbucks also has a $1 million generationskipping tax exemption. Under current estate tax law, she can give $1 million to her grandchildren that will not be subject to tax when her three children die. But she can leverage the generation-skipping tax exemption to almost $2 million by applying it to the charitable lead trust contribution: Suppose she is able to give $2 million to this charitable lead unit trust; 50 percent of that amount is subject to the charitable deduction, which means she can technically give $2 million to the trust and have it qualify for the $1 million generationskipping tax exemption. When this trust terminates 10 years hence and goes to her grandchildren, it may then be worth $4 million if the $2 million has appreciated at a reasonable rate. That $4 million is not subject to generation-skipping taxes and passes to the grandchildren with no additional taxation at her three children's deaths.

Exhibit 2. Charitable lead AnnUity Trust Type of calculation Transfer date Section 7520 rate Fair market value of trust Rate of annuity Payment periods in year Payments made at beginning or end of period Term of years Annual payout Annuity factor Payout frequency factor Present value of annuity = Annual payout x Factors Remainder interest = Fair market value of trust - Present value of annuity Charitable deduction for income interest Donor's deduction as percent of amount transferred

©Association for Investment Management and Research

Term September 19, 1998 6.6% $1 million 7%

Annual End 10 $70,000 7.1553 1 $500,871

$499,129 $500,871 50.087%

87

Investment Counseling for Private Clients If Mrs. Bigbucks carries out this plan in her lifetime, she also obtains the benefit of paying gift taxes versus paying estate taxes, which provides an additional discount of at least 25 percent from death taxes. An advisor would not propose this idea to a client with $2 million, but a client with $10 million can afford to do something like this and achieve wonderful benefits down the road.

Conclusion Wealthy families have a legitimate concern that they will share a significant portion of their net worth with the federal government before it passes down to the next generation or generations. To reduce the tax burden, families should do the following: • Initiate a lifetime gifting program-the $10,000 and $20,000 annual gifts, tuition gifts, and health care gifts; consider the use of split-interest trusts and family limited partnerships.

88

•

•

•

•

•

Use the unified credit. Each spouse should own at least as much as the unified credit equivalency each year so that no matter who dies first, the credit is not wasted. Consider the benefits of making taxable gifts rather than holding money until death and leaving heirs with a higher transfer tax burden. Consider using the charitable deduction and the leveraging techniques to help reduce the family's ultimate tax. Recognize the long-term savings potential from the generation-skipping tax exemption and incorporate it into the plan if a couple has grandchildren to whom they want to pass property; consider leveraging the savings through a charitable lead trust. Evaluate the net benefit of directing retirement plans to a charity rather than to the family to avoid a significant portion of those retirement benefits going to the government instead of the heirs.

©Association for Investment Management and Research

Estate and Tax Planning

Question and Answer Session William R. Levy Question: Are preschool tuition and braces acceptable under the tuition and health care expense gifts that are not subject to tax? Levy: Braces are normally considered cosmetic, but they can also be characterized as necessary for a child's long-term health. If the dentist or orthodontist will give you a statement that the need for braces is clearly health related rather than cosmetic, you could likely benefit from the exception. Preschool tuition is acceptable as long as it is paid to an educational institution. Question: Are these tools open to people other than grandparents? Levy: If the one making the payment is legally responsible for the care and education of the child, then there is no benefit. But these tools can be used by an aunt, uncle, friend, and so on. Question: Can income from a charitable remainder trust be used to buy life insurance that would then pay the children at death, which would allow them to effectively receive the estate on a tax-efficient basis? Levy: Many financial planners also sell life insurance and are raising the concept of the dynasty trust, in which you create a charitable remainder trust to move assets out of a client's estate, receive all the benefits I discussed, and also increase the client's income from those assets from, say, 1 percent to 6-7 percent. Because you have taken that charitable remainder away from the client's family, financial planners then suggest buying some insurance in

an irrevocable life insurance trust, which is considered a wealth replacement trust, and using the additional income earned on the charitable remainder trust to pay the premiums. It works and is a viable concept for people who want to use it. In essence, the plan replaces the wealth that the family has lost to the trust by creating a tax-free pool of funds that is available for the family when the insured dies. Clients or advisors must be very careful, however, that they are not dealing with someone who simply wants to sell some life insurance. The premiums to pay the life insurance can be set up in several ways. Instead of it being owned by an irrevocable life insurance trust, the insurance is owned by the children. Say,three or four children are involved and the premium is $40,000-$80,000. The husband/ wife can make a gift of $80,000 with no tax, and then the children pay the premium and they are the owners and the beneficiaries of the life insurance. The money-the $80,000 or a portion of it-that's needed to pay the premiums comes from the additional income generated by the charitable remainder trust. When the insured or donor gives the money to the children or to the trust for their benefit, that money is considered a gift and may reduce the donor's ability to make other tax-free gifts directly to the individual beneficiaries. Question: For families with a great deal of wealth who want to leave as much as they can to children but most of that wealth happens to be in IRAs or taxdeferred assets, what are the best strategies?

©Association for Investment Management and Research

Levy: It's a problem situation. For years, we have heard about the benefits of building up our retirement plans because they will continue to appreciate and accumulate money tax free. Unfortunately, when the primary beneficiaries of those retirement plans die, the IRS takes a huge bite out of all those benefits. One option is to draw down funds from the retirement plan before you die. Once you reach age 59 1/2, you can start taking out up to 100 percent of the money and pay a 40 percent income tax. Then, you can start planning and building that money for the future, and that money will basubject only to the straight estate tax. Some people argue that one is better off leaving the money in the IRA in the long run because the assets taken out and reinvested are subject to capital gains taxes. You must crunch the numbers to see what makes the most sense for you. Question: Can you give large retirement plan balances to a foundation or a charity during your lifetime? Levy: To give them up during your lifetime requires that you first be deemed to have received them, so they are subject to income tax before you turn them around. If you are considering giving up a retirement plan, take the minimum amount during your lifetime and then have the remaining interest pass to a charitable institution at death. In that way, you can get an up to 85 percent subsidy from the IRS. Question: Whatis the minimum dollar amount that a private foundation must distribute each year, the maximum amount a

89

InvestmentCounseling for Private Clients family can take for managing a family foundation, and would paying travel expenses for the trustees be allowable?

levy: As to disbursements, legally, at the beginning of each 12month period, the foundation must give at least 5 percent of the value of the foundation at the end of the subsequent 12-month period to public charities. For example, if you have a $1 million foundation, you must give $50,000away over a 12-month period. If income has accumulated at the end of the next 12-month period, the amount that must be given away may be more than that. Most of the time, there is no limit on the amount you can give away. You can give away 100 percent anytime you want, provided the document establishing the trust doesn't require the trustees to maintain the foundation forever. If the family is no longer interested in continuing the administration and cost of running a foundation, for example, they can find one charity and make a significant gift to it, and the foundation disappears. There is no hard-and-fast rule about how much you can pay the trustees, but a family foundation is not a good place to try to get away with something. Keep the fees reasonable. If a foundation has trustees all over the United States and holds an annual meeting in, say, Chicago, paying for them to fly in for the meeting is a reasonable expense. The IRS looks at how big a foundation is and how big its expenses are. If a huge foundation gives a lot of money to charity, the IRS is not going to begrudge reasonable trustee expenses to attend meetings. However, if the trustees start taking huge fees out, the IRS is more likely to scrutinize the return. Question: Does a donor get any continuing tax benefits from establishing the foundation?

90

levy: The foundation is a taxpaying entity that is subject to a 2 percent excise tax on its earnings and capital gains. That tax can be reduced by the fees paid to investment advisors or to the trustees, but the tax is so small that the tax benefit is nominal. The donor may be one of the trustees, but the donor receives no more tax benefits than the full benefit obtained by making the gift. Question: Are these foundations at risk of being shut down by the IRS?

levy: No, the IRS encourages family foundations because the foundations support charitable organizations. If the charitable deduction and private foundations were eliminated, all these wonderful tax-exempt organizations doing great work-for schools, churches, medical causes, and so on-would have to be funded by the taxpayers. From an administrative standpoint, Congress would have to decide which charitable organizations to fund and to what degree. Congress does not want to make those decisions. The government would rather let us do that and reduce the taxes we have to pay in order to fund it. Question: What is your opinion on using loans to make gifts greater than $10,000 a year and then forgiving the loans in later years and recognizing the gifts at that time?

levy: Loans properly structured and set up as business transactions are not a problem. They can be appropriate, and the process can work smoothly, just as in making gifts to third parties when you understand the person to whom you give the money will tum around and give it to somebody else. There's nothing illegal about that transaction as long as the person you gave it to is not under an

obligation to give it to the person you want to receive it. If you structure a legitimate loan with no interest due, the IRS may impute interest on it because it views it as a business transaction. So, if you lend your son $100,000 and he doesn't have to pay you any interest, you (the lender) are responsible for paying income tax on the interest that should have been paid. In essence, you're giving him the benefit of that $100,000interestfree loan even though you may pay some additional income tax. There also may be a gift tax assessed, if the interest forgiven exceeds the annual exclusion. Question: How should one view the asset allocation of a marital trust, family trust, or generation-skipping trust from an investment perspective?

levy: A marital trust is subject to federal estate taxes when the surviving spouse dies. Keep in mind that any appreciation in the trust will be subject to a 50 percent tax, so you may not necessarily want to make that trust totally growth oriented. The marital trust is generally not subject to significant capital gains taxes because the cost basis will be stepped up to date-of-death values at the first spouse's death. A residuary trust, which can grow and will not be subject to estate tax when the surviving spouse dies, is subject to capital gains taxes from the time of the original funding because it receives no step-up in the cost basis when the surviving spouse dies and the trust passes on to the remaining beneficiaries. So, the two forms involve a trade-off between capital gains (taxed at 20 percent) and estate taxes (taxed at 50 percent). Growth-oriented securities should thus be steered into the residuary trust and the generation-skipping trust. If any

©Association for Investment Management and Research

Estate and Tax Planning of the trusts is to be oriented toward income, I recommend that it be the marital trust. An added benefit of that arrangement is that the spouse who will benefit from that income is usually interested in it being at a reasonable level. Are there any limits to the number of methods that might be used by a single client? Question:

levy: Investment advisors are working to increase clients' wealth through their investment expertise, so when you see that a lot of that wealth is going to go to taxes, you should suggest some plans the clients can consider. Investment advisors need not try to be the experts; they can bring in qualified professionals to meet with clients. Bring in someone with whom you are comfortable and who you know is competent. Then, you will look great to your clients because you are bringing the idea to the table and recommending somebody to carry out the idea. You should let that professional determine which techniques make the most sense for particular clients;

then, you both function as members of the team for the client. Question: Are there reference sources that you would recommend?

levy: Estates and Trusts magazine generally publishes good ideas about these topics in an understandable manner. I would also recommend, as an educational process, sitting in meetings between your clients and professional tax and estate advisors to see what concepts the advisors recommend to your clients. Question: In Exhibit 1 and Case Study A can the $236,000 be used to offset future capital gains or only to offset income? Must the gain be spread over the six years?

levy: The income tax deduction can be used to offset ordinary income or capital gains. There are limits on the portion of the deduction that can be used in anyone year based on a taxpayer's income and the nature of the assets transferred. Taxpayers have a total of six

©Association for Investment Management and Research

years to use up the deduction, and they must use it before their death or they lose the balance. Question: In Case Study A, wouldn't the trust's remainder interest need to reach the 25percent threshold for the trust to qualify?

levy: The requirement is that the charitable interest must be at least 10percent, which could create a problem when the income beneficiary is rather young or, in particular, when the payouts run for the joint lives of two relatively young beneficiaries. Question: Is there a step-up in basis with the $10,000 per donor annual exclusion gift?

levy: No. Property gifted during life, whether or not it is subject to gift tax, passes to the donee at the donor's tax cost. The only time one receives a step-up is at death. As a result, it is not beneficial that gifts to individual donees be made with highly appreciated assets. Those assets should be used for gifts to charity.

91

After-Tax Performance Evaluation James M. Poterba Mitsui Professor of Economics Massachusetts Institute of Technology

Focusing on after-tax returns is a great way to add value and gain competitive advantage in the investment management business. Managers need to understand the factors that affect tax efficiency, to realize that a "one size fits all" performance measure and tax strategy will not work, and to integrate portfolio management with income tax and estate tax planning. Algorithms, such as the "accrual equivalent" tax rate, can help managers educate clients about the various implications of taxes for their portfolios.

he after-tax return associated with a given pretax return may vary considerably among individual tax-paying clients. Portfolio returns from the perspective of pretax reporting or performance management for tax-exempt clients can look very different from returns from the perspective of tax-paying individual investors. This presentation focuses on the general issue of measuring after-tax performance-how taxes interact with performance management. This focus includes the AIMR Performance Presentation StandardssM (AIMR-ppsSM) and, more generally, the design of broad-based performance evaluation standards. This presentation examines whether an algorithm can be designed and used to tell a particular client the likely after-tax consequences of various portfolio strategies and to compare the after-tax performance of various managers. The presentation also provides an overview of the U.S. income tax environment for high-networth households, with particular emphasis on capital gains tax issues. Finally, the presentation reviews three important estate-planning tools that individuals can use to effectively manage estate taxes.

T

Why the After-Tax Focus A focus on after-tax returns is worthwhile for several reasons. One is that most managers who work with individual clients know that increasingly sophisticated individual investors are demanding analysis of how their taxes are influenced by manager behavior, portfolio selection, and asset allocation. Therefore, managers need to report results in a way that puts taxes into the broader picture.

92

Another reason is that reducing the tax drag on a portfolio may be an easier way of increasing after-tax returns than searching for additional pretax riskadjusted returns (alpha). For example, when a manager is liquidating a position, selling highest-cost-basis shares rather than shares with an acquisition price close to the average basis is straightforward and enhances after-tax returns. A manager does not need to be a rocket scientist or need to accurately predict future earnings to follow that strategy. Therefore, a focus on after-tax returns allows recognition that there is some "low-hanging fruit" that many managers can harvest. Finally, being able to make a coherent presentation on the tax consequences of various investment strategies and management styles can gain a manager an advantage in the competition for client money.

Factors Affecting Tax Efficiency Certain key factors influence any portfolio's tax efficiency-that is, the difference between the portfolio's pretax return and its after-tax return for a taxable investor: • Portfolio turnover. Portfolio turnover is dearly one of the key factors affecting tax efficiency, but the notion that high turnover equals bad tax efficiency is a myth. Turnover is like cholesterol: There is good, and there is bad. Good turnover is the harvesting of losses and the early realization of positions that have losses. Bad turnover is the selling of gains and the early triggering of capital gains tax liability. Tax-efficient portfolios should have more of the good and less of the bad turnover.

©Association for Investment Management and Research

After-Tax Performance Evaluation •

Inflows versus withdrawals. A second key factor affecting tax efficiency is the pattern of asset inflows and withdrawals. The importance of this factor is particularly clear when examining the after-tax return performance of mutual funds. Consider two funds that are holding identical portfolios at the beginning of a year, and assume that these portfolios have substantial embedded, unrealized capital gains. One of the funds experiences large redemptions; the other experiences inflows during the year. The redeeming fund will have, on average, a higher tax burden for the year. A central issue that the AIMR-PPS Implementation Committee has wrestled with is the extent to which the manager is burdened with the taxes that are realized as a result of withdrawals that are beyond his or her control. • Dividend yield versus capital gains. This factor, as most managers of taxable assets know, is critical. Because dividends are taxed more heavily than realized capital gains for most taxable investors and because unrealized capital gains are taxed even more favorably, a portfolio that generates a high fraction of its returns in the form of dividends will face a higher tax burden than a fund that generates primarily capital gains. The factors driving tax efficiency are fairly simple, so clients might expect most managers to behave in a tax-efficient manner. But many managers do not. Several hypothetical examples will show how these factors can increase the difference between pretax and after-tax returns. Mutual Fund Example. Suppose a hypothetical portfolio has the following characteristics: Beginning-of-period market value $10.00 Realized long-term capital gains 1.75 Realized short-term capital gains 0.25 Dividend income 0.50 Unrealized capital gains 0.50 Total pretax earnings 3.00 Pretax returns for this portfolio are an impressive 30 percent for the period. Unfortunately, after-tax returns are significantly less. After-tax returns on this fund, for an individual investor in the top U.S. federal marginal income tax bracket, can be calculated as follows: $_ _1.7_5...:.(_1.00_-_0_.2_0;....)+_(;....$_0._25_+--,-$0_.5_0.:.;)(...:.1._00_-_o._39_6..:...)_+..:...$0_.5_0 _ 23 .5%. $10.00

The terms (1.00 - 0.20) and (1.00 - 0.396) correspond, respectively, to the after-tax value of one dollar of realized long-term capital gains (taxed at 20 percent) and one dollar of realized short-term gains and dividend income, which is taxed at the federal marginal tax rate of 39.6 percent. Published estimates of pretax and after-tax returns on various mutual funds, such as those reported each year in BusinessWeek,

©Association for investment Management and Research

show just how much of the pretax return can be consumed by taxes. There is typically enormous variation among mutual fund managers in the amount of taxable income they generate for a given amount of pretax return. This variation reflects differences in each of the portfolio attributes described previously in this section. Given the range of effective tax burdens among mutual funds and the diversity of effective tax rates among private clients, taxable investors should consider after-tax returns when selecting mutual funds.

Client Portfolio with Cash Withdrawals. Cash inflows and withdrawals also influence after-tax portfolio returns. The following example illustrates the measurement of after-tax performance for a client portfolio that has cash withdrawals: $1,000 Initial portfolio value 500 Initial unrealized gains 1,150 Final prewithdrawal portfolio value Final postwithdrawal portfolio value 1,100 150 Total capital appreciation 50 Client cash withdrawal 140 Realized long-term gains 510 Final unrealized gains 30 Cash dividends (distributed) The pretax return on this portfolio is given by Dividends Pretax return ==

+ Change in portfolio value + Cash withdrawal Initial value

In this example, the pretax return is 18 percent. Suppose the general AIMR-PPS algorithm for after-tax performance measurement is used to measure the after-tax returns. This algorithm is as follows: Realized long-term gains (l - 'co) + (Realized short-term gains

After-tax return ==

+ Dividends) (I - tDlV) + Unrealized gains + Tax-free income + Client withdrawal adjustment factor Starting asset value

,

where tCG stands for the marginal tax rate on longterm capital gains realizations and tDIV stands for the ordinary income tax rate that applies to dividend income. These tax rates are set at 20 percent and 39.6 percent, respectively. In this algorithm, the client withdrawal adjustment factor is t CG (Net withdrawal)(Realized + Unrealized gains) (Final portfolio value + Net withdrawal)

When the AIMR-PPS algorithm is used, the aftertax return is significantly less-14.58 percent-than the pretax return of 18 percent. The client withdrawal factor raises the reported after-tax return by 6 basis points.

93

InvestmentCounseling for Private Clients

Measuring After-Tax Performance Differences in investors' income tax rates, and the interplay between investor characteristics and decisions about realizing capital gains, represent major challenges to measuring and evaluating after-tax performance. Therefore, managers need to understand how each investor's federal tax rates and potential estate taxes interact to influence portfolio management decisions. A good starting point for this analysis is the AIMR-PPS algorithm for measuring after-tax performance. AIMR-PPS Algorithm. According to the AIMRPPS standards, managers are to use the maximum federal statutory rate for each type of client. So, a manager would use 20 percent as the statutory longterm capital gains rate. But the tax rates clients face actually can vary, even among high-net-worth clients. Therefore, managers may want to customize this rate in the calculation of after-tax return. An investor with large capital loss carryforwards from a failed past investment may face a lower effective capital gains tax rate than someone without such a loss carryforward. The AIMR-PPS standards also tell managers to exclude state and local taxes. The argument for doing so is that people live in different states. Everyone lives in some state, however, and for people who live in high-tax-burden states-in California (where the top income tax rate is 11 percent), New York (where the rate is 7-8 percent at the state level and more for New York City), and Massachusetts (which used to have a 12 percent tax rate on interest and dividend income)-managers should include the state tax rate in the algorithm so they can report what those taxes do to their clients' returns. Few investors would try to compare the returns on taxable and tax-exempt bond portfolios without recognizing the role of federal and state taxes. Similarly, a manager needs to recognize the role of state and local taxes in after-tax equity portfolio performance.

wealthy communities. The result is a database that includes a nontrivial number of respondents from the highest- net-worth segment of the U'S, population. The data show that only about a fifth of the households that were in the top 1 percent of the income distribution in 1995 were also in the top 1 percent of the net-worth distribution. In 1995, a family needed an income of about $225,000to be in the top 1 percent of the income distribution. To be in the top 1 percent of the net-worth group, a family needed about $3 million in total household assets, including retirement accounts. Whether investment managers' clients have high net worth or high income or both may not be clear. Part of the reason is that many individuals in highnet-worth households are past their prime earning years; they may be retirees whose income is primarily capital income. At the same time, some high-earning younger households may not yet have had time to accumulate substantial assets. Variation in income andasset values. Taxes also vary because different assets generate different income profiles. People with substantial net worth in real estate or municipal bonds, for example, may have low taxable income relative to others in their networth category. Therefore, their tax status may be different from that of others with similar net worth. Special circumstances. Finally, individuals' taxes vary because of "specialized tax circumstances." The importance of particular tax conditions, such ~s tax-planning problems generated by low-cost-basis stock, the alternative minimum tax (AMT), or loss carryforwards, is hard to identify. Nevertheless, such factors generate substantial scope for variations in tax rates. These issues argue against using a one-size-fitsall performance measure. Managers need to ~hink about a particular client's circumstances. The differences can be accounted for in simple computer programs that allow one to plug in various marginal tax rates as well as underlying information about the realization of gains and the income components of a given pretax return. Once the manager understands the person's marginal federal and state tax rates and other tax complications, the manager can tailor the presentation to those conditions.

Individual Federal Tax Rates. Individual investors' federal tax rates vary for several reasons. Dependence of taxes on income, not wealth. Some investors have substantial accumulated net worth, but their incomes do not put them in the top end of the income distribution. Managers can get a more systematic handle on this information than from anecdotal client interviews from a database that is collected every three years by the U.S. Federal Reserve Board, which is summarized in the Survey of Consumer Finances. This survey is the best publicly available nonproprietary information on the highest-net-worth part of the U.S. population. In collecting these data, the Fed focuses on households in

Marginal Income Tax Rates. The notion of the "effective marginal tax rate" rather than the"average tax burden" should drive both managers' and clients' behavior. In 1998, the top federal marginal tax rates on interest and dividends were 39.6 percent starting at $271,050 of taxable income, 36 percent starting at $151,750, and 31 percent starting at $99,600. In addition, phase-out rules on deductions can propel topbracket taxpayers into a tax bracket with a marginal

94

©Association for Investment Management and Research

After-Tax Performance Evaluation tax rate as high as 41 percent. Managers can get a handle on how the rules on deductions affect a client's marginal tax rate by asking the client to add another $200 of interest income to his or her reported taxable income and then to recompute the taxes. In many cases, this additional income will not change the taxes by 39.6 percent times $200. It will change the taxes by more in some (most!) cases and by less in others. This exercise will give managers an idea of the true marginal tax rate the client faces. The actual dividend tax rate faced by households with different characteristics is revealing. Table 1 is based on a data set that the U.S. Internal Revenue Service (IRS) has released for researchers that includes actual (but anonymous) tax returns for tax year 1994. Note that for households that reported at least $50,000 of dividend and interest income, only about 40 percent were facing tax rates higher than 37 percent. Many people who reported that much in dividend and interest income fell in the 28 percent bracket. In other words, a significant number of households with substantial liquid assets face a marginal tax rate that is lower than the top rate. In the category of households receiving dividends and interest of more than $200,000,about 70 percent faced tax rates of more than 37 percent, whereas nearly 95 percent of those who have high wage and salary income are in the 37-41 percent tax range. There is more variation in the taxes on capital income than on earned income. Capital Gains Taxes. Taxes on interest and dividends are not a very exciting part of the after-tax portfolio problem. Capital gains are where the action is. Most managers know the tax rules on short-term versus long-term capital gains. The tax rate on shortterm capital gains-gains on securities held for less than one year-ean go as high as 39.6 percent. Today, the top rate on long-term capital gains is 20 percent; starting in 2005, that rate will come down to 18 percent on assets that have been held for at least five years. The difference between the rates on short-term and long-term gains will be even larger in the future

Marginal Tax Rate < 16 percent 16-29 percent 29-37 percent 37-41 percent > 41 percent

Number of taxpayers

than it is today. The tax-efficient strategy, therefore, when all other aspects of a portfolio are the same, is to realize losses and hold gains until they become long term. The advantages of generating long-term capital gains relative to other kinds of portfolio income are larger today than they have been in the past. Table 2 shows how the Taxpayer Relief Act of 1997 increased the advantages of tax-efficient investing. As recently as 1989, the top marginal tax rate on the highest dividend income and interest income recipients was 28 percent, which was also the top tax rate on capital gains. During the past decade, a zero tax rate differential has widened to a 20 percentage point differential. In the future, the difference between taxes on capital income and long-term capital gains could be as large as 23 percentage points.

Table 2. Top Federal Marginal Tax Rate overTime Category

1996

1998

2005 (estimate)

Dividends and interest Long-term capital gains

-41% 28

-41% 20

41% 18

This tax environment explains the benefits of taxefficient portfolio management, but it also raises the issue of tax code risk. The tax system can shift in ways that destroy the advantages of previous tax-efficient behavior. Should investors and managers try to hedge bets on the tax system? Should taxpayers hold some qualified money in a Roth Individual Retirement Account and some in a traditional IRA? The tax treatment of these accounts is different, and one (the Roth) is not affected by future changes in marginal income tax rates. The answer is probably yes, although deciding how to model tax code risk is an unexplored problem. Unrealized Capital Gains. One of the most difficult problems in measuring after-tax portfolio performance concerns handling unrealized capital

Wages + Salaries > $500,000

Dividends + Interest > $50,000

Dividends + Interest > $200,000

10.0% 28.4 22.7 37.7 1.3

10.5

1.0

13.8

3.5

4.1 70.2 1.6

0.3 94.7 0.6

402,800

51,000

76,300

Source: Calculations made using National Bureau of Economic Research TAXSIM model.

©Association for Investment Management and Research

95

InvestmentCounseling for Private Clients gains. The AIMR-PPS standards recommend ignoring potential future taxes on unrealized capital gains. I believe that some positive tax burden should be applied to unrealized gains because they are not untaxed but, rather, carry a contingent future tax liability. The precise tax burden on unrealized gains depends on client circumstances. First, is the client likely to face substantial "forced realizations"? Such realizations may result from a transaction that sells them out involuntarily from a position in a low-costbasis stock or from a substantial consumption demand on the client's part-e-demand for a new house, a new yacht, or so on. Second, is the client someone who, perhaps because of age, is in a position, from the standpoint of dynastic wealth accumulation, to take advantage of basis step-up at death? Third, is the client going to carry out a gifting strategy that will transfer asset basis to the next generation? If so, the manager should consider carryover basis rather than basis stepup as the likely scenario for assets received by the next generation. These issues can be explored with clients to determine what capital gains tax circumstances apply. They need to be covered in the client's investment plan, not left to the tax planners. One way to describe the capital gains burden on unrealized gains is with the "accrual equivalent" capital gains tax rate. This concept finds the accrual tax system that gives the client the same total after-tax portfolio value at the end of the "realization period" as the current realization-based system, assuming asset sale at the end of the realization period. The approach is similar to asking what tax rate in New Zealand (which uses an accrual system on capital gains) would give you the same wealth after tax at the end of a given number of years as the 20 percent realization-based tax rate in the United States. The concept is useful to managers in explaining that deferring taxes is not the same as never paying them. It is a way of quantifying the interest-free loan a client is getting from the IRS when the client defers realizing capital gains on a position. For example, suppose a client buys an asset in Year 0 for $100 and it generates capital gains at 10 percent a year. The asset pays no dividends. If the gains were realized each year and taxed as short-term gains (that is, every 364 days, the manager realized the gains on the position), then the after-tax return each year would be 10(1.00 - 0.396)

=6.04%

because the short-term capital gains tax rate is 39.6 percent. After 10 years, no additional taxes would be due and the value would be $100.00(1.064)10 = $185.96.

96

The value of the portfolio would be greater if the asset were held for 10 years and then the capital gains were realized. Assuming that the asset grew at 10 percent a year for 10 years and that 20 percent of the gains were taxed away in Year 10, the asset's aftertax value in 10 years would be (1.00 - 0.20)[$100.00(1.10)10 - $100.00] + $100.00 = $227.50.

The accrual equivalent tax rate depends on the rate of return the asset manager would have had to earn on an after-tax basis year after year to get to the value of $227.50 in 10 years. To find this key bit of information, simply solve the following expression for R, the rate of return: $100.00(1.00 + R)lO = $227.50.

The answer is R = 8.57%.Thus, the accrual equivalent tax rate is 10.00% - 8.57% 10.00%

= 0.143, or 14.30%.

Using the accrual equivalent rate is like comparing the accumulated assets in an IRA with assets that are invested in a taxable account. How can one compare the values of those two investments? One way would be to find the internal rate of return on assets in the IRA that provides the same return as that available, after tax, outside the IRA. The accrual equivalent tax rate is the tax rate that if it were charged to your account every year on the accruing gains, would give you the same after-tax portfolio value at the end of the planning horizon that you would have had if you had been working under the realization-based system in which you pay taxes only at the end. The 14.3 percent accrual equivalent tax rate can be compared with the other tax rates applicable to the client. Because clients are accustomed to thinking in terms of 39.6 percent as an interest tax burden and 20 percent as the statutory tax burden on long-term realized gains, the 14.3 percent is a comparable measure. It describes what the rate would have to be if every year the assets with unrealized gains were marked to market and taxed on their gains. The accrual equivalent tax rate provides a flexible tool for analyzing future tax burdens. If the client's tax rate is likely to change in the future, the manager can build that change into the calculation of an accrual equivalent tax rate. It can also be used to illustrate the value of deferring capital gains realizations over different horizon lengths. Table 3 shows how the accrual equivalent tax rate changes as the holding period increases. If capital gains were realized after one year, the after-tax return would be 8 percent, which implies a 20 percent effective tax burden relative to the pretax 10 percent return. By 20 years, the tax burden is down to about 10 percent. The best outcome, as far as taxes go, is if

©Association for Investment Management and Research

After-Tax Performance Evaluation Table 3. Accrual Equivalent Return by Holding Period Holding Period

Return

1 year 5 years 10 years 20 years Until death

8.00% 8.28 8.57 8.98 10.00

the asset is held until death because then the tax liability is extinguished. If this asset generates only capital gains, the accrual equivalent return moves up to the 10 percent pretax return because of basis stepup at death. The tax return data the IRS receives indicate that in the mid-1990s, the average holding period for corporate stock on which gains were subject to long-term capital gains tax treatment was about 6.5 years. Managers may be able to persuade their clients to hold assets longer by using a similar analysis to that of Table 3. This table can also illustrate the important benefits of a low-realization tax-management strategy. Estate Tax Issues. Table 3 shows the potentially substantial tax savings that can flow from holding appreciated assets until death. Is the scenario realistic? Managers need to make some assumptions about the future tax circumstances of clients and about their estate-planning advice to evaluate the chance of basis step-up. Life expectancy is one factor that applies in judging the likelihood of holding an asset until death. Table 4 shows that life expectancy in the United Sates is quite long, even for people already at advanced ages. These data are based on the entire U.S. population, and the wealthy clients with whom managers deal are likely to live longer than these averages. The exact cause of wealth-related differences in mortality experience is not known, but there is a stark disparity between the mortality rates faced by those in the Table 4. life Expectancy at Various Ages, 1994 Age/Sex

Years Expectancy

65-year old Man Woman

16.1 20.0

75-year old Man Woman

9.8 12.8

85-yearold Man Woman

5.3 6.9

Source: U.S. Social Security Administration, unpublished tables used in preparing the 1995Social Security Administration Trustee's Report.

©Association for Investment Management and Research

bottom 20 percent of household income or wealth distribution and those at the top. In short, even if the client is an 85-year-old woman, the client's life expectancy may be 7 years, so whether a basis stepup is just around the corner is hard to judge. Managers, therefore, need to be mindful that the planning horizon, even for older clients, could be substantial. Most decedents with substantial estates die at quite advanced ages. Table 5 shows that in 1992, nearly two-thirds of taxable decedents were in their 80s when they died and more than three-quarters of the value of the estates reported were for decedents who were older than 70. Table 5. Age Distribution of 1992 Taxable Decedents Age at Death (years) < 50 5G--60 60-70 70-80 > 80

Percentage of Tax Returns

Percentage of Estate Value

1.9% 3.1 9.9 23.4 61.7

3.1% 6.2 16.0 26.7 47.9

Source: Martha B. Eller, "Federal Taxation of Wealth Transfers, 1992-1995," Statisticsof Income Bulletin (Winter 1996-97):8--63.

The estate tax is a large and important tax in relation to other taxes for the small set of taxpayers who are likely to face it. The estate tax will change in the future as a result of a sliding scale of tax thresholds that will phase in between now and 2006. At the moment, an estate must be more than $625,000 to incur federal estate taxes; by 2006, the threshold will be $1 million. Today, marginal estate tax rates start at 37 percent at that $625,000 level and go as high as 60 percent. For a client who is thinking about keeping the assets that he or she has built for the next generation, the estate tax is thus an important part of the tax environment. Traditional bequests. Assets can be passed from the wealth accumulator to the next generation in three ways, and the three methods have different tax consequences. The first method, and the one that receives the most attention, is traditional bequests. Assets are held until the accumulator, or the spouse of the accumulator, dies, and then the assets pass as a bequest to the next generation. Such assets are subject to the estate tax. They receive a basis step-up on any accrued unrealized capital gains, but the heirs face a variety of costs at the time of death, which may involve probate, valuation, and other issues. The costs are likely to be lower for publicly traded securities than for interests in privately held businesses or other, less liquid, assets. 97

Investment Counseling for Private Clients Tax-free inter vivos gifts. The second way to transfer assets is by carrying out a gifting strategy during the lifetime of the accumulator generation. The allowed annual gifts of $10,000 per recipient per donor avoid the estate tax but do not participate in any step-up in basis, so the capital gains tax liability is not extinguished. In general, high-net-worth clients can have either capital gains tax relief or estate tax relief but not both. So, for taxpayers who could shift a substantial share of the wealth to the next generation in the form of intervivos gifts, the tax gains from basis step-up are not as great as the simple analysis suggests. Capital gains taxes may be reduced at the cost of higher estate taxes. For anyone who focuses on tax planning, the data on participation in gifting is remarkable. Table 6 shows the percentage of households with householders 55 years old or older that gave assets of $10,000 in support of other households by net worth and by age. (The $10,000 is not $10,000 per child; it is simply giving $10,000 total.) The Survey of Consumer Finances, from which these data are drawn, asks donors about their gifts. It also asks the potential recipients, younger households typically, about the inter vivos gift amounts that they received. It turns out that the amount of inter vivos giving reported is about twice the amount of intervivosreceiving that is reported. No one understands why, but keep it in mind when analyzing these numbers. Table 6 indicates that only about a fifth of the households that are headed by somebody over the age of 75 with a net worth of more than $2.5 million are making these kinds of gifts. Whatever the reason, the high-net-worth band is leaving unnecessary tax burdens to their heirs by underusing the intervivos giving option. Table 6. Households Making Gifts of at least $10,000 a Year, 1995 Household Head's Age (years)

55-64 65-74 75+

Net Worth $1.2 Million to $2.4 Million

Net Worth Greater than $2.4 Million

12°/"

29%

20 26

30 22

bequest. Estates and gifts are taxed ostensibly under the same tax rules because we have a unified estate and gift tax in the United States. However, gifts are taxed on a net-of-tax basis, whereas bequests are taxed on a gross basis. The result is that if the estate tax rate facing a potential decedent is T, the tax rate on the gifts is T / (1 + T) instead of T. With a 50 percent statutory tax rate, the tax rate on the gifts (0.5/1.5) effectively becomes a 33 percent rate instead of a 50 percent rate. This consideration is very important, and it is widely underappreciated. Taxable gifting is greatly preferable to leaving assets to pass through a taxable estate. This is true because of the tax rate difference and because paying the gift tax avoids the later estate and gift tax liability on whatever subsequent appreciation occurs on the assets given away. Appreciation on the assets and income on the assets will accrue to the next generation instead of to the donor generation. The gifting strategy raises, once again, the income tax versus estate tax trade-off. Taxable gifts reduce estate tax burdens, but they preclude taking advantage of capital gains basis step-up at death. Only a tax lawyer can advise on the trade-offs among estate, gift, and capital gains taxes, but a starting point is that estate and gift tax rates tend to be much higher than the long-term capital gains rate. Estate and gift taxes start at 37 percent and can go as high as 60 percent. So, avoiding the estate tax and paying the capital gains tax at a rate of 18-20 percent along the way is often the preferable strategy. This issue crystallizes the importance of understanding client-specific circumstances in measuring after-tax results. For example, the trade-off between capital gains and estate/gift taxes is moot for those who are perfectly happy to leave their assets to a charitable foundation; they have a different strategy for reducing estate taxes, a strategy in which the capital gains tax liability is not important. For clients who wish to leave their assets to their children, however, reducing the combination of capital gains taxes and estate taxes is important.

Conclusion

Taxable g~fts. The third possibility, which is attractive only for very high-net-worth clients, is to make taxable gifts of more than the untaxed $10,000 per recipient a year. On an after-tax basis, pursuing a taxable gifting strategy is attractive for two reasons. The first is that the effective tax burden on a taxable gift is lower than the effective tax burden on a

Measurement of after-tax portfolio performance is a crucial undertaking for any manager with private clients, and the AIMR-PPS standards are an important systematic effort to bring the industry up to speed in presenting returns on an after-tax basis. Taxes are complicated, however, for high-networth clients. Managers may find that the AIMR··PPS standards are too broad for reporting performance for particular clients. Managers may need to customize the algorithm by client or client group to recognize particular tax circumstances. This customization

98

©Association for Investment Management and Research

Source: Federal Reserve Board, the Survey of Consumer Finances.

After-Tax Performance Evaluation may involve building in specific tax rates, future tax liabilities, and even intergenerational plans in terms of wealth accumulation and wealth transfers. Tools such as the accrual equivalent tax rate provide managers a straightforward way to explain to clients how important and how valuable deferring capital gains can be. Managers need to think about modeling the client's after-tax returns for various portfolios. Managers also need to consider the implications of gifting behavior. Apparently, most high-net-worth couples are likely to pass along some assets at the death of the second-to-die spouse, and the estate tax is likely to affect their intergenerational transfers. This practice raises the probability that some assets will face a zero capital gains tax rate (through the basis step-up) and underscores the need to integrate the investment manager into the tax-planning and tax-

©Association for Investment Management and Research

management picture. In some cases, wealth accumulation must be viewed from the family rather than the individual perspective. Gift giving complicates the problem of measuring marginal tax rates for afterdeath performance evaluation. A common misconception is that taxes and investments can be managed separately. But investment managers cannot expect to achieve the best possible after-tax returns if they handle only the portfolio side of a client's affairs and tum over other aspects to tax managers, accountants, or tax planners. These aspects are intertwined. Accountants cannot try to minimize taxes when they are given a pretax return stream as the outcome of what the portfolio manager has done. Two-way communication is necessary if the strategy is to leave the client the largest after-tax wealth possible.

99

lmiestment Counseling for PrivateClients

Question and Answer Session James M. Poterba Question: How do you benchmark after-tax returns when the accrual equivalent tax rate is used? Poterba: You have all the mechanics in the basic formula, where a zero tax rate is being applied to the unrealized component of capital gains. The only question is whether to use a positive tax rate. If you want to work with the accrual tax rate, the modification is to put in a 1 minus 10 percent or 1 minus 12 percent tax burden on the unrealized capital gains term. Operationally, after you've run through the calculation for a client and identified the holding period as 20 years, that calculation tells you that the effective accrual capital gains rate on a gain that accrues today is on the order of 10 percentage points. You would assign that rate when you calculated after-tax returns. Question: Is it reasonable to assume that a client who withdraws 10 percent a year from the assets will have different after-tax performance from a client who adds 10 percent a year to the assets? Poterba: Yes, the situation is similar to the cost basis of a mutual fund that is getting positive cash flows versus one with negative cash flows. The one with positive flows doesn't have to realize as many gains. A fund with a 10 percent negative cash flow will have to realize gains and will have lower after-tax returns. That outcome is as true for a separately managed account as it is for a mutual fund. I would use effective accrual rates to cast light on the effect of withdrawals versus infusions. For example, consider a client who plans to withdraw 10percent ofthe 100

portfolio each year. That information tells you that the time horizon or the effective duration for which the money is going to be under management is significantly shorter than for someone who plans to contribute 10percent of the portfolio value each year until he or she dies. What's driving the difference is that the person who is adding money year after year is getting much longer average interest-free loans on the unrealized capital gains. You would expect to see a substantial difference in the aftertax return performance of those two portfolios. Of course, the portfolio manager would be wise to recognize that the strategies and the investments that are optimal for someone who plans to add 10 percent a year will not be the same as the strategies that are optimal for someone who is planning to withdraw 10 percent a year. The trade-offs between the tax savings for lowdividend securities generating capital gains and the potential risks or other costs that arise from tilting the portfolio in that direction will be very different for those two clients. Question: Is after-tax performance useful or not, then, in measuring manager value? Poterba: What's the alternative? Is it best to ignore the fact that there were tax consequences associated with realizations and not try to make corrections? Some sort of measurement of after-tax performance makes sense; AIMR introduced the adjustment factor to account for such differences in strategies, but comparability is still not perfect. If you are trying to make comparisons of after-tax performance among managers,

what you'd really like to see is the kinds of accounts a manager is running that have withdrawal characteristics like the account you're bringing to the table. The ability to dissect a manager's accounts by client objectives is the sort of information that could be useful, but that sort of detailed performance analysis is virtually impossible. Question: How does one deal with realized losses that must be carried forward to a future tax year? Are any adjustments made? Poterba: One thing you can do, in the spirit of the effective accrual tax rate, if a client has a very simple portfolio position with a limited number of managers working with the client and the client realizes a loss that is too large to use up this year (so the losses carry forward) is to ask: What is the discounted present value today of being able to reduce taxable income by $3,000 a year this year, next year, and all along the way until the loss is used up? If the tax rate at which you're deducting those losses is 39.6 percent today and you have $6,000 worth oflosses, then you're getting 39.6 percent on the first $3,000 and 39.6 discounted for one period; so, it may be effectively 36 percent on next year's losses in today's dollars. Simple analysis, however, omits the fact that managers can also modify their behavior with respect to future gain realizations so as to accelerate the use of the loss as a device for sheltering other gains. Therefore, the real issue is how valuable it is to have losses on the books today, given that the client can use these losses to realize some gains tax free and thereby rebalance the portfolio.

©Association for Investment Management and Research

After- Tax Performance Evaluation What do you do about taking over a portfolio that has large embedded gains when you have been hired to diversify the portfolio (so, the after-tax returns are going to be biased downward)? Question:

Poterba: You need to make some sort of correction in reporting portfolio performance, perhaps group client portfolios that start from similar positions. The issue is reminiscent of the debate that sometimes goes on in the popular press about mutual funds with different amounts of embedded capital gains: If you started tracking two funds today-c-one a new fund and the other an old fund with substantial embedded capital gains-you would expect the fund with the embedded capital gains to generate more taxable realizations going forward. Comparing the

managers would be unfair because they start from different positions. Question: At what net worth should you suggest to someone that they gift? For example, is $1.2 million enough assets for a 70year-old couple that may face nursing home costs? Poterba: The most likely explanation for the low level of taxable gifting, at least among those with net worth below, say, $3 million, is the fear of substantial expenses sometime before the end of their lives. Nursing homes probably loom largest in those anxieties. In most case, those expenses do not, however, make a substantial dent in high-net-worth household assets. Something like 20percent of 70-year-olds will go into a nursing home at some point before deaththe percentage is higher for women

©Association for Investment Management and Research

than for men. Most stays in nursing homes are relatively short, although most of the dollars spent on nursing home care are spent for the small subset of very long-term stays. Most of these households retain substantial assets at the time of death of the surviving spouse. The reason I focused on the $2.5 million net-worth category is that in that range, potential nursing home bills will not draw down most of the accumulated assets. One could probably build a simple Monte Carlo simulation to get a handle on the risk of nursing home need versus tax savings. For example, you could find out the rough odds of various expenditures and then point out the tradeoff between, say, a 5 percent chance of needing to pay an expense and the saving of 40 percent in terms of the difference between the estate tax and the capital gains tax.

101