Wealth Management

Overview Christopher Wiese, CFA Publications Manager CFA Institute According to the latest World Wealth Report released by Merrill Lynch and Capgemini, the number of millionaires (or equivalent in U.S. dollars) worldwide grew to 8.3 million in 2004, a net increase of about 600,000 over the previous year.1 The combined wealth of this high-net-worth group came in near $31 trillion in 2004. As this private wealth population continues its brisk pace of growth, so do the opportunities for investment professionals to add value. The private wealth market is finding itself with ever more complex and sophisticated needs, which require the constant development and refinement of complex tools and sophisticated products to serve these clients well. Given this context, one should not be surprised that the CFA Institute Wealth Management 2005 conference saw record attendance. It is exciting to watch private wealth evolve toward intensive study and the rigorous application of investment theory to bear down on increasingly complex investment problems. To help in this regard, we have a great lineup of presentations within this proceedings from the Wealth Management 2005 conference. The topics cut a broad swath through some of the most pressing issues. In the first presentation, Derek Sasveld, CFA, walks through the practical application of some of the philosophical issues surrounding investment policy and asset allocation in a taxable environment. The discussion then evolves to asset location with Martin Silfen pointing out the ways in which a systematic approach to allocating funds among taxable and tax-deferred accounts can significantly improve total after-tax performance and terminal wealth. 1 Merrill

Lynch and Capgemini, World Wealth Report (2005). This report can be accessed at www.ml.com/media/48237.pdf.

©2005, CFA Institute • cfa pubs .org

From there, Jeffrey Horvitz turns the discussion toward tax management and how taxes affect the way advisors look at portfolios. His presentation delves into the value of tax deferrals and offers an analysis of tax-loss harvesting. Robert Gordon then discusses tax efficiency through appropriately structured transactions and through the use of derivatives. Although the application of these principles can introduce increased complexity in the portfolio management process, and they may need additional explaining to clients, the potential for these strategies to mitigate tax liabilities makes them well worth the effort and study. Mr. Gordon is followed by Jarrod Wilcox, CFA, who shifts the discussion toward a framework for adding value through risk management. He advocates a focus on discretionary wealth as a means by which to determine appropriate risk aversion and ultimately to add value. Much of the high-net-worth market is composed of entrepreneurs and privately held businesses. Clients in this group have an additional layer of complexity in that a lot of their wealth may be tied to the business. In addition, they need a savvy succession plan as a component of their holistic wealth management. Sean Sebold, CFA, discusses the needs of this group of clients and how to integrate those needs into a holistic wealth management plan. Lastly, Joseph Maxwell discusses another, usually overlooked, way to add significant service and value: client reporting. The private wealth market is demanding not only more-sophisticated investment options but also the kind of performance reporting that used to be reserved for institutional investors. He discusses the keys to a good client reporting strategy that build client confidence and trust.

CFA Institute Conference Proceedings • 1

Asset Allocation to Meet Client Goals Derek A. Sasveld, CFA Executive Director and Senior Asset Allocation Strategist UBS Global Asset Management Chicago

To set an appropriate investment policy that considers the impact of taxes, simulation provides a better starting point than optimization because it provides a perspective on illiquidity. The simulations provide general guidelines for allocation rather than a single optimal solution. Fortunately, an appropriate allocation in a taxable environment is only modestly inferior to the corresponding pretax mix. Although tax minimization can be challenging, managers can use tax-free investments and a number of techniques to balance tax efficiency with investment efficiency.

his examination of asset allocation and the appropriateness of certain asset allocation strategies for individual investors begins with brief discussions of allocation philosophy and capital market assumptions—particularly as related to risks and returns—that underpin the rest of the ideas. Following those discussions, I address why using optimal tax-adjusted assumptions does not generally result in much better portfolios after tax than using pretax assumptions with tax restrictions imposed. I then discuss selecting an appropriate pretax investment policy, applying modifications of certain asset class assumptions after taxes, and finally, analyzing some considerations for carrying out active strategies in a taxable environment.

T

Asset Allocation Philosophy Asset allocation in the typical case of a long-only investment client begins with declaring an investment policy or benchmark and making active investment decisions to modify it. Generally, the investment policy is global and the “market” is the global investable capital market. Asset classes and their corresponding weights are blended to achieve unique portfolio characteristics. At UBS Global Asset Management, we begin with a neutral asset allocation policy around which we actively manage. A team determines active asset and currency weights in these portfolios; in a taxable environment, the active asset allocation decision becomes more complicated. We also actively manage currencies and expect security selection within asset classes ©2005, CFA Institute • cfa pubs .org

to be an important part of value added in investment portfolios. The result of asset class choices, asset class weights, currency exposure weights, and security selection is global performance. We believe that when market prices of various assets, securities, currencies, industries, or countries move away from intrinsic value, that is the time to take positions in portfolios. Long-only asset allocation at UBS is influenced by the familiar finding that approximately ninetenths of the variability of investment performance over time is best explained by investment policy.1 Although the proportion of variability in our own asset allocation portfolios has been less than ninetenths since 1981, the policy decision is still critical in determining expectations for total portfolio risk. As a result, I will discuss our views on the appropriate asset allocation policy for both nontaxable and taxable investors.

Capital Market Assumptions The investment policies we determine are based on an estimate of the total global investable market (GIM). Figure 1 shows the estimated composition of the GIM denominated in U.S. dollars at the end of 2004. We make this estimate each year because it represents the foundation market estimate that we use to assess the expected risks and returns of different markets and asset classes around the world. The same estimates are used in identifying the total—and 1

Gary P. Brinson, Brian D. Singer, and Gilbert L. Beebower, “Determinants of Portfolio Performance II: An Update,” Financial Analysts Journal (May/June 1991):40–48.

CFA Institute Conference Proceedings • 3

Wealth Management

Figure 1.

Total GIM, 31 December 2004 Emerging Market Equities 1.1%

U.S. Equity Private Markets 0.3% 17.0% Cash Equivalent 3.9% U.S. Real Estate 5.1%

All Other Equities 17.6%

High-Yield Bonds 1.1%

Emerging Market Debt 3.1%

U.S. Dollar Bonds 25.6%

All Other Bonds 25.1% Note: Data are preliminary.

relative—risk and return implications of active asset class weights in portfolios. As of December 2004, the market was roughly $87 trillion in size, and as Figure 1 shows, approximately 40 percent of the market was composed of equities, 50 percent of fixed-income classes, and the remainder of private (equity) markets and real estate. In forming return expectations for these various asset classes, we take a risk premium approach. The starting point is the real risk-free rate of return, which represents what an investor can expect in compensation simply for deferring consumption and holding the risk-free asset for a year. To the risk-free rate, we add an inflation premium. The inflation premium for various countries incorporates monetary policy, fiscal policy, and the general impact of global inflation. The third component is then a risk premium, which recognizes the additional compensation investors expect for accepting greater risk relative to the GIM, consistent with volatility characteristics, covariance or correlation with other asset classes, and illiquidity. The result is equilibrium asset return estimates for various asset classes. Table 1 shows the equilibrium asset returns and the building blocks on which

Table 1.

they are based for six asset classes. The real risk-free rate is the same for each class, as is the inflation premium (from the perspective of U.S. investors); only the risk premiums differ. Combining each of the building blocks geometrically produces the equilibrium asset return estimate. For example, for U.S. equities, we take (1 + 0.024)(1 + 0.023)(1.035) – 1.0 = 0.084 or 8.4%,

which represents approximately what an investor might expect in equilibrium (over the next 20 or 30 years, for instance) for aggregate U.S. equities, proxied by the Russell 3000 Index or the Wilshire 5000 Index. To develop the risk premiums for each asset class, we begin by assigning the GIM a beta of 1.0. We then apply a variety of econometric analyses, some backward looking and some current and forward looking, to derive an expected beta for each asset class. Figure 2 shows seven asset classes, U.S. equities, and the GIM plotted by equilibrium return and beta. With an expected beta coefficient and an expectation of the GIM’s Sharpe ratio, we can build risk premiums that are commensurate with different asset classes.

Components of Equilibrium Asset Returns U.S. Equities

Global (ex-U.S.) Equities

Emerging Market Equities

U.S. Bonds

Global (ex-U.S.) Bonds

High-Yield Bonds

Equilibrium asset return

8.4%

8.1%

11.4%

5.7%

5.6%

6.8%

Real risk-free rate

2.4

2.4

2.4

2.4

2.4

2.4

Risk premium

3.5

3.2

6.4

0.9

0.9

2.0

Inflation premium

2.3

2.3

2.3

2.3

2.3

2.3

Component

4 • CFA Institute Conference Proceedings

©2005, CFA Institute • cfa pubs .org

Asset Allocation to Meet Client Goals

Figure 2.

Equilibrium Return/Beta Profile

Equilibrium Return (%) 15

Private Equity Hedge Funds

10 Timber Real Estate 5

GIM U.S. Bonds

Global (ex-U.S.) Bonds

U.S. Equities Global (ex-U.S.) Equity

0 0

1

2

3

4

Beta

Generally, if the manager believes the world is integrated (as opposed to segmented)—that is, if the manager believes all assets have risks in common and all countries are fully available to all investors worldwide—then the linear relationship depicted by the straight line in Figure 2 represents the manager’s view. In reality, some assets, such as hedge funds, private equity, timber, and real estate, will finish above the line, as shown. They tend to provide greater compensation for market risk because of different degrees of illiquidity and segmentation from the GIM. Like return estimates, risk and correlation estimates by asset class should also be forward looking. Through the integration of our work in building risk premium estimates, we are able to extend out to estimates for standard deviation and ultimately correlations.

Selecting Appropriate Policy In building an appropriate policy portfolio, we focus on simulations rather than on optimization. The reason is that optimization assumes a single-year holding period and fails to properly account for differences in liquidity between asset classes. Simulation, however, involves modeling cash flows over multiple periods and captures expectations and the properties of return and risk in a probabilistic way. It captures both the benefits and costs of illiquidity. Our parameters for a simulation of policy are risk tolerance and liquidity tolerance. At times, higherrisk or illiquid investments might be more appealing, but some investors may not be comfortable with, or be able to take on, the additional risk. Therefore, we search for the policy allocations resulting from the simulations that are “appropriate,” rather than those that are “optimal.” In practice, rather than resulting from a single appropriate frontier, portfolios result from a band of ©2005, CFA Institute • cfa pubs .org

appropriate approaches. From these possibilities, we select a policy and then establish a range around that policy to allow for maximum and minimum asset weights from an active asset allocation. A risk budget is then allocated across varying dimensions relevant to client circumstances: information ratio or Sharpe ratio targets or, more generally, asset class exposures versus market allocation benefits. Rather than always using a specific risk metric, we typically think about risks in different ways for different types of portfolios. In some cases, we use explicit return and variance estimates; in other cases, we use those same estimates in a calculation like value at risk. When we are actively managing asset allocation and other dimensions of relative return portfolios, we tend to focus on relative return and relative risk. For this purpose, we might focus on alpha or active beta and tracking error versus the opportunity set. Figure 3 shows a sample moderate-risk, pretax, global policy portfolio that includes both liquid and illiquid investments. Global public equities and bonds are the major components and are recommended as global optimal approaches free of home bias. Alternative assets are led by the allocation to direct investments in real estate properties, diversified by region and property type.2

The Impact of Taxes What changes and adjustments does the reality of taxes demand?

Figure 3.

Moderate-Risk, Pretax, Global Policy Portfolio

Hedge Funds 3% Natural Resources 2%

Private Equity 5%

Real Estate 10% Global Equity 52%

Global Bonds 28%

2

Kevin Terhaar, Renato Staub, and Brian Singer, “Appropriate Policy Allocation for Alternative Investments,” Journal of Portfolio Management (Spring 2003):101–110.

CFA Institute Conference Proceedings • 5

Wealth Management

Adjustments to Capital Market Assumptions. When making adjustments to expected returns and risks, we consider income at the highest marginal rate. One important consideration for equity investments is the impact of the expected dividend payout ratio from aggregate earnings. In addition, interest rate paths should be considered because of their impact on price fluctuation of fixed-income securities. We expect the long-term Treasury yield of 4.6 percent to migrate to about 5.2 percent by 2008. At the shorter end of the yield curve, today’s cash rate— a little more than 2 percent—should rise to just above 4 percent by 2015 and stay there in equilibrium. The resulting difference between interest rates today and interest rates in the future produces a certain amount of implied capital loss on longer-dated bonds in addition to implications for interest rate revenue. In evaluating the impact of capital gains, we consider both short-term and long-term gains, of course in light of anticipated turnover rates. Later, I will present some results from this kind of analysis; in it, we may be penalizing equities a bit heavily because we include a relatively high turnover assumption but Figure 4.

do not include the implied added value from the investment decisions responsible for that turnover. In other words, we look at the asset class only as an index, but we include a turnover rate that is higher than what the index would suggest. Some disagree with this approach, but it is appropriate for after-tax policy discussions because if one assumes no benefit from manager selection, the average active equity manager can be expected to provide index-like performance. We also must look at the risk implications of a tax regime. In this area, we acknowledge that tax-loss carryforwards dampen expected volatility for most asset classes. This effect is more pronounced with the riskier and less-liquid asset classes because they are more volatile. This benefit will become apparent when we consider efficient appropriate allocations. Figure 4 shows the equilibrium expected returns on a pretax and after-tax basis for 11 asset classes. Clearly, investments such as municipal bonds look much better than some other classes when taxes are a factor. Note that developed market equity returns are not expected to be significantly higher than municipal bond returns after taxes are considered.

Equilibrium Expected Asset Class Returns Asset Class U.S. Equity

Global (ex-U.S.) Equity Emerging Market Equity U.S. Bonds Global (ex-U.S.) Bonds Real Estate Private Equity High-Yield Bonds Emerging Market Debt Cash Equivalents Municipal Bonds 0

2

4

6

8

10

12

14

Equilibrium Return (%) Pretax

6 • CFA Institute Conference Proceedings

After-Tax

©2005, CFA Institute • cfa pubs .org

Asset Allocation to Meet Client Goals

The cause is the relatively high turnover statistic that is implied; for conservatism, we do not take into account the potential added value of that turnover. Actually, we typically expect the alpha or added value from equities at our firm to be 200 bps, which is also, in itself, a modest estimate versus what we have been able to deliver in the past. What happens to risk in an after-tax world is shown in Figure 5. Risks fall quite a bit more for some asset classes, such as emerging market equity and private equity, than for others; these types of investments can be tax advantaged because of substantial potential loss carryforwards. The impact is much less dramatic on the relatively safer assets, such as fixedincome investments. Figure 6 shows pretax and after-tax efficient allocations: Circles represent efficient allocations based on pretax returns and risks, and squares represent aftertax efficient allocations. To develop these comparisons, appropriate portfolios with the highest Sharpe ratios were determined by using after-tax returns and risks. The triangles depict the situation differently by

Figure 5.

showing what the pretax portfolio would look like on an after-tax basis without re-evaluation. Notice that the after-tax appropriate allocations are not significantly different in terms of their risk and return expectations, especially at higher risk levels. For less-risky portfolios, municipal bonds are explicit substitutes and the tax advantage for loss carryforwards becomes less important, so the difference between after-tax allocations is slightly greater. Figure 6 allows us to draw a number of conclusions. First, after-tax efficient mixes are only modestly inferior to pretax efficient mixes after tax adjustment. In other words, if one appropriate policy is determined with the use of pretax data and another is determined with the use of after-tax data, the aftertax expected return differential between these two sets of mixes is, on average, only about 30 bps lower for the pretax mixes after tax adjustment. Of course, 30 bps can be meaningful over a long horizon, but it is not as meaningful as might be expected given the imposition of a tax regime.

Equilibrium Expected Asset Class Risks Asset Class U.S. Equity

Global (ex-U.S.) Equity Emerging Market Equity U.S. Bonds Global (ex-U.S.) Bonds Real Estate Private Equity High-Yield Bonds Emerging Market Debt Cash Equivalents Municipal Bonds 0

5

10

15

20

25

30

Equilibrium Standard Deviation (%) Pretax

©2005, CFA Institute • cfa pubs .org

After-Tax

CFA Institute Conference Proceedings • 7

Wealth Management

Figure 6.

Equilibrium Return/Risk Profiles of Appropriate Allocations

Equilibrium Return (%) 11

8

5

2 0

5

10

15

20

Risk (standard deviation, %) Pretax Efficient Allocations After-Tax Efficient Allocations Pretax Efficient Allocations after Tax Adjustment

Second, tax-exempt bonds are highly favored in after-tax mixes that are judged to be appropriate. Municipal bonds look attractive when compared with conventional fixed income and, on a riskadjusted basis, when compared with all asset classes. Because of their appropriateness and a strong demand from investors, UBS Global Asset Management is beginning to manage active asset allocation strategies with investment policies and valuation models for municipal bonds, as is already the case for taxable asset classes. Third, risky assets are generally favored when the impact of taxes is included in the analysis. Greater volatility implies greater possibility for taxloss carryforwards. The final conclusion is that “growth-oriented” assets, those that accumulate without large distributions of income, also tend to be favored. Clearly, illiquid assets have a significant benefit in this area, not in terms of carryforward of tax benefits but because they would be sold only when exiting the strategy. Asset Class Modifications for Tax Effects. I will explore some asset class modifications to incorporate tax effects by examining a couple of examples: tax-exempt bonds and so-called alternative assets. ■ Municipal bonds. The favorable after-tax return on municipal bonds makes them popular; more than 10 percent of the assets managed for individual investors in the United States by UBS Global Asset Management are municipal bond mandates. Municipal bonds reflect some regional differences, but the 8 • CFA Institute Conference Proceedings

differences are of little importance in the context of generalized asset allocation advice and management. Despite their benefits, municipal bonds should not entirely replace taxable bonds in allocations for several reasons. First, the expected added value in actively managed municipal portfolios is typically more modest than the expected added value available from other actively managed fixed-income asset classes. Managers have fewer dimensions across which to add value in municipals, per unit of active risk, than in other types of fixed-income instruments that can benefit from greater sector, country, and issue selection. A second reason is that the sources of market risk and added value in a municipal bond can be very different from the sources in taxable bonds. For example, from day to day, the yield curve looks substantially different for municipals from the yield curve for taxable bonds. The analysis and sources of credit risk are also substantially different for municipals versus many kinds of taxable bonds. Third, municipal bond mandates typically offer little flexibility to pursue negative, or short, duration positions. An increasing number of fixed-income managers are managing absolute-return strategies that have proven especially popular in periods (such as at present) when interest rates are expected to rise. The flexibility to go negative on duration can be used by active managers to add value and can also improve the risk and return profile of a diversified portfolio. Another example of why municipal bonds should not entirely replace taxable bond allocations is shown in Table 2, which presents expected correlations between municipals and various traditional fixed-income asset classes according to UBS Global Asset Management’s capital market expectations as detailed previously. The low correlations with highyield and emerging market bonds are to be expected; these types of bonds also behave substantially differently from conventional taxable fixed income. But the expected correlation of municipals is low even with the Citigroup Broad Investment Grade Bond Index; Table 2.

Municipal Bond Correlations with Other Asset Classes

Asset Class

Correlation

Citigroup Broad Investment Grade Bond Index

0.79

Citigroup World Government Bond Index (Non-U.S.)

0.30

Merrill Lynch High Yield Bond Index

0.37

JPMorgan Emerging Markets Bond Index

0.22

3-Month Eurodollar

0.00

Source: Correlations are from Constellation, the UBS Global Asset Management forward-looking risk management system.

©2005, CFA Institute • cfa pubs .org

Asset Allocation to Meet Client Goals

a correlation of 0.79 is low enough to imply that meaningful differences exist between investmentgrade taxable bonds and municipal bonds. The yield curves suggest how opportunities may differ, as shown in Figure 7. Yields on the taxables are higher, although after adjustment for taxes they will typically offer less compensation than municipals. The slope of the yield curve can be substantially different between different maturities in either type of instrument: Taxables have a lower slope at intermediateterm maturities—that is, between, say, three and seven years—and are demonstrably flatter the longer the maturity. So, a fixed-income strategy would be different for the two types of assets, which suggests that fixed-income exposure other than munis may be appropriate for some individual investors. The justification for including taxable bonds is not just that they may allow active managers to add more value but that the markets have different term structures. Another aspect to keep in mind is that the expansion and contraction of taxable spreads can make taxable products more or less enticing than municipal products over time. Figure 8 shows corporate spreads by credit rating in the U.S. market over time. Note that the additional yield provided by a corporate bond, depending on its bond rating, is currently much lower than it has been historically. So, the differences in return will likely be more modest in the future than they have been. This circumstance makes municipals relatively more attractive at the moment, but this is clearly not always the case.

Figure 7.

National Tax-Exempt Bonds vs. U.S. Treasuries, 31 December 2004

Yield to Maturity (%) 5 4 3 2 1 0

5

10 15 20 Maturity (years)

U.S. Treasuries

25

30

National Tax Exempt

Note: Based on the 31 December 2004 yield-to-maturity curve of the AAA municipal bond market. Sources: Based on data from Bloomberg, UBS Global Asset Management, and Money Fund Report Average.

Differences can also be observed—but to a greater degree—between high-yield and emerging market debt and conventional yields, as Figure 9 shows. Yield spreads were roughly 10 percent for emerging market debt as late as 2001 and for high yield as late as 2003. By February 2005, both had dropped to near 3 percent. These assets may be delivering less-robust returns in the future, but they are nevertheless different from municipal bonds, which means that they may provide periodic benefits and, again, suggests that not all taxable bonds in a portfolio should be replaced by tax-exempt bonds.

Figure 8. Corporate Fixed-Income Yield Spreads by Credit Rating, 1999–2004 Yield Spread (%) 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 4/99

10/99

4/00

10/00

U.S. BBB

©2005, CFA Institute • cfa pubs .org

4/01

10/01 U.S. A

4/02

10/02 U.S. AA

4/03

10/03

4/04

10/04

U.S. AAA

CFA Institute Conference Proceedings • 9

Wealth Management

Figure 9. Emerging Market Debt Stripped Spread vs. U.S. High-Yield Spread, 1 January 1991 to 28 February 2005 Spread (%) 18 16 Emerging Market Debt 14 12 10 8 6 4

U.S. High Yield

2 91

92

93

94

95

96

97

98

99

00

01

02

03

04

05

06

Notes: Emerging market debt spread data adjust the JPMorgan Emerging Markets Bond Index for defaulted securities. Stripped yields account for strip-out enhancements such as credit guarantees. The remaining “stripped yield” is a truer measure of a sovereign’s implied credit risk. Source: Based on U.S. high-yield spread data from Merrill Lynch.

■ Alternative and illiquid investments. In the area of alternative or illiquid investments, hedge funds are certainly of great interest to investors. After fees, however, hedge funds might produce some return disappointments. Many high-net-worth individuals have a lower risk tolerance for illiquidity than other investors and expect an appropriate premium for tolerating it. Unfortunately, not everyone can tolerate illiquidity, and all taxable investors face the challenge of maintaining enough working capital or short-term income to pay tax liabilities generated by illiquid strategies, which is particularly an issue with investments in many hedge funds. Tax implications for a number of illiquid strategies have recently been mitigated by the use of derivatives. Complex swaps and other derivative agreements are increasingly being used to change a package of income and short-term capital gains into a package that favors long-term capital gains. Tax authorities may decide in the future to alter the rules and disallow these specific tax-avoidance strategies. Lack of transparency in hedge funds is an important problem for many investors who are trying to plan appropriately for taxable events in their portfolios, although perhaps less so for high-net-worth investors with sufficient resources beyond their illiquid allocations. Whether hedge funds are appropriate depends somewhat on the degree to which wash sales

10 • CFA Institute Conference Proceedings

are generated and how much knowledge the fund managers have of taxable events for individual investors. In general, the amount of investment in illiquid assets should be constrained for taxable investors because these investments do not provide income or distributions to help fund tax liabilities. In other words, investments in real estate, natural resources, or private equity may be very attractive before the impact of taxes, but they are a problem if the client does not have the wherewithal to pay tax liabilities as they arise. Applying Modifications. When modifying any taxable portfolio, we change the target allocations as designed for the pretax moderate-risk investment policy (as given in Figure 3) to resemble the allocations given in Figure 10. The allocation to alternatives is still 20 percent, but natural resources drop out in favor of a slightly increased exposure to hedge funds. In addition, tax-exempt bonds play a significant role, and we can design flexibility to allocate some funds to lower-income equities and single-stock positions (labeled “Custom Equity”). Within bond allocations, we carve out a healthy allocation to tax-exempt instruments; almost two-thirds of the fixed-income piece is exempt from taxes. Aggregate global bonds and equities, including these new categories, are weighted as before.

©2005, CFA Institute • cfa pubs .org

Asset Allocation to Meet Client Goals

Figure 10.

Modifications to the Pretax Policy Portfolio Private Equity 5%

Hedge Funds 5% Real Estate 10%

Global Equity 52%

Global Bonds 28%

Tax Exempt 19%

Custom Equity 24%

Modifications to Investment Strategies We prefer to use managed accounts as a vehicle for tax-efficient management instead of mutual funds. Mutual funds offer less control for tax management as investors enter and exit the fund at their discretion. With managed accounts, we can control security-bysecurity trading, and entry and exit from the strategy do not negatively affect other investors. The investment process is crucial. We tend to look at a tax adjustment as an overlay, not necessarily as a fundamental change to how we identify opportunities and manage risks in the underlying investments. When we use an overlay, we can establish more taxefficient control of rebalancing as well as buy and sell security and asset decisions. In conducting simulations of various tax strategies, we have found that some work better than others. For example, a “firstin, first-out” (FIFO) strategy—even with perfect strategy implementation through rebalancing—does not produce a sound tax management strategy. Aggregated short-term gains are volatile, and gains may prove to be costly because frequent rebalancing results in more transactions. Of course, the benefits to the FIFO approach are that the portfolio is always “on strategy” and the tracking error is low relative to the underlying strategy. But it results in taxable events and represents the worst possible outcome. As an alternative, specific identification of tax lots with tolerated drift (minimal rebalancing) produces much better results. In this approach, the aim is to maximize long-term capital gains and minimize short-term capital gains through matching lot sales to previous purchases in individual securities. In other words, we do not sell the most recent purchases; we use an optimization function. The goal is to achieve zero aggregated short-term gains over time; the result is their minimization. ©2005, CFA Institute • cfa pubs .org

Using specific identification with tolerated drift (ID) requires that we tolerate a certain amount of tracking error relative to the underlying investment strategy. This element is part of the tension that all investment managers face: How much should I focus on taxes versus focusing on the underlying investments? Sometimes what is best for taxes is exactly opposite to what is best by investing standards. The simulation results for these different strategies clearly show that the ID method is preferable. We carried out simulations for FIFO with and without drift and the ID method with and without drift. The results are as follows: • Under a FIFO strategy with drift, the result for a diversified portfolio was high taxes but about a 40 bp advantage relative to FIFO with no drift in terms of the annual change in the net asset value (net of taxes) of the portfolio. • For ID with low tolerance for drift (heavy rebalancing), the results were slightly lower taxes but only about a 15 bp benefit to the annual after-tax net asset value. • When we engaged in ID with a trade program to maximize long-term capital gains (selling the oldest lots of individual stocks and bonds while holding the newest lots), the results were only slightly higher taxes than would be the case for a no-drift policy, but the extra performance after tax was approximately 100 bps. Therefore, we recommend this plan of action for many taxable investors: At least pursue a simple trading rule to maximize long-term gains and minimize short-term gains and allow a certain amount of tracking error or drift from the desired asset allocation strategy. But remember that this is not as feasible in a mutual fund framework.

Active Management with Taxes The last challenge in active management in light of taxes is how to carry out active investment strategies in asset allocation, currency choices, and security selection. Generally, an active manager is looking for significant mispricing in equities and fixed income, and the additional tax liabilities brought about by active management suggest that mispricing must be greater for active managers to transact on behalf of taxable clients. Unfortunately, an active asset allocation approach may offer fewer opportunities to add value for taxable investors now in early 2005 because large price-to-value discrepancies are not as visible in stock and bond markets. Equities appear to be undervalued to some degree, and fixed income appears to be slightly overvalued, but the opportunities for active asset allocation are less significant when viewed on an after-tax basis. CFA Institute Conference Proceedings • 11

Wealth Management

But lack of asset allocation opportunities may not always be the case. As Figure 11 illustrates, valuations can drift quite far from their “intrinsic” or “fair” values. For the several years prior to 2002, our diagnosis of a technology bubble produced large overvaluations in equity markets and commensurate opportunities to add value in active asset allocation by avoiding these classes. Similarly, late 2002 and 2003 brought significant undervaluations across equity markets. In currencies, plenty of opportunities have been sizable enough to justify investigation on an after-tax basis, such as overvaluation of the yen in the mid1990s or the misvaluation between the euro and the U.S. dollar in the early 2000s. Today, relatively modest valuation divergences are visible in some cases— for example, overvaluation of the British pound against the dollar and attractive opportunities in Asian currencies. Security selection is substantially different from the two other dimensions of active management mentioned earlier because significant mispricing is more frequent and widespread. In pursuing these opportunities, however, the manager must ensure that turnover does not become excessive and must maintain a strong buy-and-sell discipline. The valuation concept is unchanged, but expected excess performance by

Figure 11.

individual securities should be tempered by the tax treatment of cash flows to investors. In essence, a mispriced security’s attractiveness after tax is greater for ones where the proportion of total return realized through long-term capital gains is higher. As long as the gain upon sale will result in a taxable event and the environment is one in which differential tax treatment exists for dividends and short-term capital gains versus long-term capital gains, then all else being equal, a security intended to be a long-term strategic position would be an ideal candidate for emphasis in taxable strategies. Similarly, a security whose price moves toward intrinsic value gradually over longer periods might be considered superior after tax compared with one just as mispriced that might more rapidly approach intrinsic value. Or a security with strong dividend income characteristics in its future cash flows might be judged to be superior compared with one with more potential price appreciation of its identical expected total return.

Conclusion As fiduciaries of our clients’ assets, we strive to make our investment strategies more tax efficient. Although identifying a long-term investment policy is important, identifying a policy that is efficient after

Asset Class Over- and Undervaluation, 31 December 1995 to 31 March 2005

Price/Value Discrepancy (%) 60 Overpriced

50 40 30 20 10

Fair Value Range 0

Underpriced

96

97

98

99

00

01

MSCI World Ex-U.S. Index

02

03

04

05

Russell 3000

MSCI Emerging Markets Index Source: Based on data from UBS Global Asset Management proprietary valuation models.

12 • CFA Institute Conference Proceedings

©2005, CFA Institute • cfa pubs .org

Asset Allocation to Meet Client Goals

tax effects is not as value-added as it would originally seem. By applying a few clear points of emphasis, financial advisors can quickly improve the tax efficiency of their asset allocation strategies: • Use municipal bonds as a significant proportion of fixed-income assets. • Hold single-stock positions if liquidation triggers tax liabilities, and move gradually to morediversified equity approaches.

©2005, CFA Institute • cfa pubs .org

•

Use a simple trading rule that aims to maximize long-term gains and minimize short-term gains. • Ensure that active managers consider the tax implications of their decisions when they transact. The desired outcome is client satisfaction with after-tax return and risk, as well as avoidance of repeating the unfortunate historical record of mutual funds’ after-tax, after-fee, and after-inflation performance.

CFA Institute Conference Proceedings • 13

Wealth Management

Question and Answer Session Derek A. Sasveld, CFA Question: Do you make asset allocation decisions differently for U.S.-based investors and European or Far Eastern investors? Sasveld: Ideally, we would follow a pure global, balanced approach for most of our investors as reflected in global investment policies and active global asset allocation decisions. The latter are, in fact, made on a global basis, but investment policies differ because of different local market regulations, investor preferences, and equilibrium conditions that favor domestic equities or fixed income. Investors in smaller countries, for example, are typically open to approaches that fully encompass global capital markets. U.K. and U.S. investors tend to prefer heavier equity weightings in moderate-risk portfolios, whereas European investors have a higher tolerance for fixed-income solutions. So, global asset allocation solutions that are similar for many types of investors and reflect only different risk tolerances would be ideal. As it is, active asset allocation exposures at UBS Global Asset Management are consistent across all strategies, but with different regulations, different tax treatments, and with segmentation present in various markets, the ideal approach is not currently possible. Canada provides an interesting example of a market looking to reduce regulations in the pursuit of truly global portfolio management. Most investors were restricted to investing no more than 30 percent in non-Canadian securities prior to February 2005. The speed of legal dissolution of that Foreign Property Rule and its adoption by both institutional and individual investors in tax-exempt accounts bear watching.

Question: Is the principle of outperforming inflation in any jurisdiction still meaningful with highnet-worth clients, or is the measure now absolute outperformance or relative returns? Sasveld: Inflation-adjusted investing or absolute-return investing, not relative to markets, is actually increasing in importance. We believe that over the next 5–10 years, equity and fixedincome market returns will be more modest than they have been in the past 20 years. So, investors who were relying on long-only allocations to traditional asset classes may be disappointed with absolute performance. As a result, we are managing more asset allocation, equity, and fixed-income absolute-return strategies that are less constrained, particularly with respect to short-selling decisions. Many taxable investors appreciate this greater flexibility but are somewhat concerned about the greater trading activity required by this type of investing. Depending on how much trading takes place, long–short absolute-return strategies can be less tax appropriate for taxable investors than many long-only strategies. Question: Should manager alpha be added to the inputs in the asset allocation model? Sasveld: The difficulty with adding manager alphas is that— especially for the average individual investor, who may be accessing from a universe of mutual funds, where manager selection is primitive or where financial advisors may be more motivated by sales charges than by appropriate investment strategies—we think it’s appropriate for investment policy to consider the asset classes

14 • CFA Institute Conference Proceedings

alone. In the business of estimating future alphas, equities would tend to be favored over bonds and highturnover strategies would be favored. So, even though focusing on alpha means better returns, on average, for most investors, the jury is still out on the appropriateness of some approaches for taxable investors. Question: What did you mean by there being little opportunity for negative duration flexibility with regard to munis? Sasveld: Speaking from our own experience, we don’t go meaningfully short in duration in core municipal and conventional bond strategies; duration is managed against a benchmark, and the typical duration for, say, a municipal bond benchmark would be something like five or six years (versus four or even fewer years for conventional bond indices). We can reduce duration well below the benchmark in municipal strategies—to the extent of the risk budget—but we cannot go short or net short. I assume the same is true for many municipal bond managers; they can’t go net short and take a negative view on interest rates, but they can move up and down the tax-exempt yield curve, within certain limits. Question: What is the role of exchange-traded funds (ETFs) in taxable accounts, and how do they compare with mutual fund holdings in terms of risk and return? Sasveld: In the managed account complex that we manage, which is more tax efficient than mutual funds, we occasionally use ETFs for certain types of allocations, including strategies that have higher turnover, more ©2005, CFA Institute • cfa pubs .org

Q&A: Sasveld

numerous security holdings, or capacity constraints. In emerging market equities, for instance, and in some small-capitalization strategies, which have capacity constraints, we use ETFs, but investors should be cognizant of the higher fees for some specialized ETFs. In core equity and core bond strategies and in broad asset alloca-

©2005, CFA Institute • cfa pubs .org

tion strategies, however, we prefer our own active strategies that pursue reasonable turnover and hold slimmed-down security lists in some cases for optimal tax efficiency. Our concerns regarding a more modest market return environment also play a role in the recommendation of approaches that offer passive market returns less

fees. ETFs are certainly appropriate for taxable investors and reduce the tax burden, but it is also our duty to consider the interests of our investors from an active standpoint as well—in terms of risk management and return management.

CFA Institute Conference Proceedings • 15

A Systematic Approach to Asset Location Martin Silfen Senior Vice President Brown Brothers Harriman Trust Company New York City

Asset location, not to be confused with asset allocation, is the placement of assets among numerous taxable and tax-deferred accounts so that a client’s stated goals are honored while terminal wealth is maximized. Because terminal wealth is influenced by a client’s personal circumstances, tax rates, choice of taxable accounts, and capital market expectations, the naive strategy of identical allocations in all accounts is not optimal. Three hypothetical (although fairly typical) examples demonstrate that after-tax performance can be improved significantly by solving for the optimal asset location through easy-to-build spreadsheet projections.

nvestment advisors spend countless hours determining the optimal allocation of assets within their clients’ portfolios—and rightly so because the effective apportioning of asset types, such as equities and bonds, is essential to the success of a portfolio strategy. But once the optimal asset allocation is determined, the next step is to find the optimal location for those types of assets because optimal asset location can significantly increase after-tax returns. Asset location is the process of placing a client’s assets in optimal tax accounts so that the projected return characteristics of selected asset classes are most effectively matched by the tax characteristics of the client’s accounts. Dozens of such accounts are available, and high-net-worth clients tend to collect them like prizes. Typically, however, a high-net-worth client has at least one taxable account and one account intended to optimize long-term tax consequences. To simplify my analysis, I have limited my examples to four familiar types of tax-qualified accounts: • credit shelter trusts, • marital trusts, • generation-skipping tax trusts (GST trusts), and • retirement accounts, including IRAs. To further simplify the discussion, I will use the term “IRA” to refer to any tax-qualified retirement account. When addressing the needs of high-net-worth clients, investment advisors are typically dealing with family situations (as do all three of the examples that I present in this discussion). The term “family,” however, describes a variety of conditions. Some-

I

16 • CFA Institute Conference Proceedings

times “family” means husband and wife; sometimes “family” indicates two or even three generations of a family; and sometimes “family” includes charitable interests. While taking the family’s particular circumstances into account, investment advisors strive to match the return characteristics of different asset classes to the tax characteristics of the family’s accounts. Some of the factors that advisors must account for are regulatory, legal, and practical constraints. Among these are applicable taxes, which include the following: • federal income tax, • state income tax, • federal gift and estate tax, • state gift and estate tax, and • unrelated business income tax.

Forecasting Methodology Any sound investment strategy is designed to meet the client’s stated investment goals. The job of investment advisors, therefore, is to analyze the likely outcomes of different asset location strategies and recommend the one that best meets those goals. The analysis conducted by my own firm uses the following steps: 1.

Analyze a client’s long-term total wealth over an investment horizon and consider the ways in which different asset location strategies might affect that total wealth.

©2005, CFA Institute • cfa pubs .org

A Systematic Approach to Asset Location

2.

Choose the location strategy that generates the highest terminal wealth in the client’s portfolio.

Table 1.

3.

Quantify the additional returns that a client can potentially earn. (This step reassures the client, adds value, and helps justify our professional judgment.)

Initial value of QTIP trust

$9,000,000.00

Initial value of credit shelter trust

$1,000,000.00

For the purposes of this discussion, I investigate asset location strategies for three hypothetical families, with each family able to choose between two different tax accounts. For all three families, I assume that asset allocation has already occurred and that it is fixed for each family over the entire forecast horizon. This assumption enables me to isolate the effects of the location decision. Typically, a family will have exposure to many asset classes, perhaps as many as 15, but limiting the analysis to equities and bonds (the two classes that comprise the largest number of assets in a typical portfolio) simplifies the projections. Finally, all projections are calculated with nothing more complicated than a spreadsheet.

Family One: Credit Shelter Trust vs. Marital Trust Family One is the Didwell family.1 Herbert Didwell has died, and he has left his wife, Martha, $10 million in trust; $1 million of that is in a credit shelter trust, and $9 million is in a marital trust, also called a QTIP trust (Qualified Terminable Interest in Property trust). Under the terms of both trusts, Martha is entitled to all the income and principal if the trustees determine she needs them. Therefore, Martha has no particular preference regarding the location of equities and bonds within the two trusts. Furthermore, the Didwells’ only son, George, will receive the remaining assets of the two trusts when Martha dies, which makes the tax consequences of the two trusts a key issue in asset location because the marital trust is a taxable part of Martha’s estate and the credit shelter trust is designed not to be taxable. Thus, the logical next step for an investment advisor is to maximize the value of the credit shelter trust and minimize the value of the marital trust. Table 1 presents the list of assumptions used to test asset location strategies (for example, Martha Didwell has a life expectancy of 15 years and the desired allocation of assets is 45 percent fixed income and 55 percent equity). Certainly, other assumptions could be used, but I consider these assumptions to be a reasonable starting point for projecting terminal wealth. For the Didwell family, I tested two strategies: (1) a naive strategy in which the same asset allocation— 1 All names and circumstances throughout this discussion are fictional.

©2005, CFA Institute • cfa pubs .org

Assumptions for Family One

Martha Didwell’s life expectancy

15 Years

Desired fixed-income allocation

45%

Desired equity allocation

55%

Fixed-income yield

2.5%

Dividend yield on equities

1.7%

Annual appreciation on equities

5.3%

Annual turnover in equities

25%

Trusts’ capital gain tax rate

15%

Estate tax rate

45%

55 percent equities and 45 percent fixed income—is used for both trusts; (2) a tax-conscious strategy in which the equities are located in the credit shelter trust because the equities are expected to appreciate more than the fixed income, thus allowing the credit shelter trust to grow at the expense of the marital trust. Based on the spreadsheet projections, the taxconscious strategy is the better choice of the two. After 16 years, when Martha is projected to have died and the estate tax is projected to have been collected, $264,000 more will be left to George Didwell than if the same proportion of asset types had been located in both trusts. The better result equates to an incremental after-tax return of 18 bps a year.

Family Two: Generation-Skipping Tax-Exempt Trust vs. Personal Account The scenario for Family Two is similar to that for Family One but a bit more complex because the estate plan provides for another generation. Allen Gotmore has died and left his daughter, Michele, $10 million, the same figure as in Family One. But Gotmore is leaving Michele $9 million outright and $1 million in a generation-skipping tax-exempt (GST) trust. When Michele herself dies, any money in the GST trust will not be subject to transfer taxes, whether they are gift taxes, estate taxes, or generation-skipping taxes. For the other $9 million, Michele wants to eventually pass as much of that wealth as possible to her children. Another difference in this situation is that the trustees of the GST trust have the discretion to withhold distributions or to share them among Michele, her children, and her grandchildren. Table 2 presents the assumptions for Family Two. Michele has a 40-year life expectancy (being younger at her father’s death than Martha Didwell was at her husband’s death), so the opportunity for wealth accumulation is 25 years longer for Family Two than for Family One. The asset allocation strategy CFA Institute Conference Proceedings • 17

Wealth Management

Table 2.

Assumptions for Family Two

Michele Gotmore’s life expectancy

40 Years

Initial value of personal assets

$9,000,000.00

Initial value of generation-skipping taxexempt trust

$1,000,000.00

Desired fixed-income allocation

45%

Desired equity allocation

55%

Fixed-income yield

5%

Dividend yield on equities

2%

Annual appreciation on equities Annual turnover in equities

7% 25%

Estate tax rate

45%

Michele’s interest income tax rate

40%

Michele’s dividend income tax rate

20%

Michele’s capital gain tax rate

20%

Trust’s interest income tax rate

35%

Trust’s dividend income tax rate

15%

Trust’s capital gain tax rate

15%

Annual spending as a percentage of overall wealth Maximum size of trust as a percentage of overall wealth

4% 80%

of 55 percent equity and 45 percent fixed income remains the same, but the capital market assumptions are slightly different, projecting higher appreciation for equities and higher yields for both equities and fixed income. The assumptions also indicate that Michele pays state income tax, so she is at a 40 percent combined ordinary income tax rate and a 20 percent tax rate for dividends and capital gains. For the trust, I have assumed a 35 percent ordinary income tax rate and a 15 percent rate for capital gains and dividends. (In many cases, an investment advisor can structure a trust so that it is not taxable in any state; New York residents can do this easily.2) In addition to the assumptions listed in Table 2, Michele and the trustees of the GST trust have made two important decisions. First, Michele wants all of the $10 million to last her entire life, so she will withdraw no more than 4 percent of her wealth each year from both accounts, assuming that 4 percent will be enough to satisfy her needs. Second, in concert with the trustees of the GST trust, Michele has decided that she will not ask for distributions from the GST trust for as long as she can afford it because she wants the GST trust to grow while her other personal assets shrink. If her personal assets drop below 20 percent of total assets, however, she will ask the trustees for 2

If in the GST trust there are no New York trustees, then the state of New York does not tax the income of the trust. Thus, if Michele lives in New York, then the trust does not pay any income tax on its accumulated earnings to the state. The trust, of course, pays federal tax. This tax advantage would also apply to residents of Massachusetts.

18 • CFA Institute Conference Proceedings

distributions from the GST trust. The investment advisor, therefore, has two overriding concerns: (1) maximize the growth of the GST trust and (2) ensure that Michele’s personal assets remain adequate to cover her annual spending budget. Given these assumptions, I tested three different location strategies. The base strategy, as for Family One, is to locate the same proportion of equities and bonds in both the GST trust and the personal account; the second strategy locates a preponderance of equities in the trust; and the third strategy locates a preponderance of bonds in the trust. Not surprisingly, locating the equities more heavily in the trust is the winning strategy. Projecting out for 41 years, Michele realizes an astounding $3.1 million more by pursuing that strategy than if she pursued the base strategy— an incremental 48 bps a year in after-tax annual return. Figure 1 shows the three location strategies and assumes that capital markets remain fixed. Notice that a tremendous difference in the strategies occurs at Michele’s death, when estate taxes are paid. The estate tax is a liability that is not apparent at the beginning of the plan but becomes all too obvious at the end. Figure 2 shows a similar end game, but the lines are no longer straight because they better represent inherent variability within capital markets. Over a 41year period, I use the S&P 500 Index as a proxy for equity returns and the intermediate-term Treasury index for fixed-income returns. These variable assumptions generated 37 different 41-year periods to back test. The earliest period started in 1926 and ended in 1962; the latest period started in 1966 and ended in 2002. Then I repeated the spreadsheet analysis for all three location strategies. Again, assigning a preponderance of equities to the GST trust continues to be the best for all 37 different periods. The benefit ranged from an incremental 41 bps in annual return to 67 bps in annual return, with a median of 56 bps. Figure 3 shows what happens to Michele’s wealth as she spends from her personal account and allows the GST trust to grow through the use of equity investment. The trust goes from being a small portion of the estate to the majority of it, thus providing much greater tax protection upon her death. Note too that the dividing line between the trust and the personal account straightens out around the 35th year of the plan because Michele’s annual spending budget requires that she ask the trustees to distribute money from the GST trust.

Family Three: Taxable Account vs. Tax-Deferred Account The final family scenario represents a typical client for most investment advisors: a family with a ©2005, CFA Institute • cfa pubs .org

A Systematic Approach to Asset Location

Figure 1.

Fixed Assumptions for Capital Markets, 2004–2045

Aggregate Wealth ($ millions) 20 19 18 17 16 15 14 13 12 11 10 04

07

10

13

16

19

Equities to Trust

Figure 2.

22

25

28

31

34

Fixed Income to Trust

37

40

43

45

Proportional

Variable Assumptions for Capital Markets, 2004–2045

Aggregate Wealth ($ millions) 40 35 30 25 20 15 10 04

07

10

13

16

Equities to Trust

19

22

28

31

Fixed Income to Trust

taxable account and a tax-deferred retirement account. For the sake of simplicity, I refer to the taxdeferred account as an IRA, but it could be one of many types of individual retirement accounts, such as a 401(k) account. In this scenario, Sam Sunbelt has retired early at age 55. He and his wife, Sally, expect to live off their $6 million of investments, of which $5 million is in a taxable account and $1 million is in an IRA—a typical arrangement for many high-net-worth clients. Sam and Sally have a 20-year-old son, Edward, and anything remaining at their death will pass to him. ©2005, CFA Institute • cfa pubs .org

25

34

37

40

43

45

Proportional

Table 3 shows the assumptions for Family Three. Among other things, the table indicates that more money is invested in equities for Family Three than for the previous two families—75 percent allocated to equities and 25 percent allocated to fixed income— which reflects Sam’s preferences. Sam and Sally expect their after-tax spending to be $200,000 a year, with the assumption that Sally will continue to spend at this rate (with cost of living adjustments) after Sam’s death and that Edward will do the same once he inherits the estate. To make Family Three’s circumstances more realistic, I have built in some unrealized appreciation in their taxable assets. CFA Institute Conference Proceedings • 19

Wealth Management

Figure 3.

Aggregate Wealth with Equities Located in Trust, 2004–2046

Value of Accounts ($ millions) 20 18 16 14 12 Trust

10 8 6

Personal Account

4 2 0 04

Table 3.

07

10

13

16

19

22

55 Years

Sally Sunbelt’s current age (primary beneficiary)

50 Years

Edward Sunbelt’s current age (contingent beneficiary)

20 Years

Sam’s age at death

90 Years 90 Years

Initial value of taxable account

$5,000,000.00

Initial value of IRA

$1,000,000.00

Current unrealized appreciation in taxable account

$500,000.00

Annual after-tax spending

$200,000.00

Cost-of-living adjustment in spending

3%

Desired fixed-income allocation

25%

Desired equity allocation

75%

Fixed-income yield

5%

Tax-free fixed-income yield

3%

Dividend yield on equities

2%

Annual appreciation on equities

7%

Annual turnover in equities

25%

Estate tax rate

45%

Sam’s, Sally’s, and Edward’s interest income tax rate

42%

Sam’s, Sally’s, and Edward’s dividend income tax rate

20%

Sam’s, Sally’s, and Edward’s capital gain tax rate

20%

Deduction tax saving rate

40%

After establishing the assumptions listed, I tested for four location strategies: 1. Base case: assets located proportionately in both accounts. 2. Equities located preponderantly in the IRA; bonds located preponderantly in the taxable account. 3. Bonds located preponderantly in the IRA; equities located preponderantly in the taxable account.

20 • CFA Institute Conference Proceedings

28

31

34

37

40

43

46

4.

Assumptions for Family Three

Sam Sunbelt’s current age

Sally’s age at death

25

Equities located preponderantly in the IRA for the first few years then relocated to the taxable account; bonds located preponderantly in the taxable account for the first few years then relocated to the IRA. Strategy 4 proves to be optimal, although it wins by only a slight margin over Strategy 3. The spreadsheet projections indicate that by using Strategy 4 over the full projected time frame of 67 years, Family Three’s wealth increases by 466 percent, representing an annual incremental return of 13 bps over the base case. At the time the family scenario begins, Sam Sunbelt is 55 years old, and the projections for Strategy 4 indicate that he should relocate his assets when he is 61. Using spreadsheet calculations like those I have used, an investment advisor can identify an optimal switching age for any client, but such forecasts should be recalculated each year to take into account the client’s changing circumstances, variations in capital market expectations, and changes in the tax code. Regarding the 67-year projection for Family Three, I have two good reasons for doing that. First, a dollar in an IRA and a dollar in a taxable account are not comparable over the short term; they become comparable only after all the distributions have come out of the IRA and have been turned into after-tax dollars. Second, the IRA’s tax deferral continues after the death of both Sam and Sally. The children/ beneficiaries of an IRA can keep deferring and taking distributions throughout their lives. Therefore, I projected all the way to the expected end of Edward’s life.

Stress Testing the Assumptions for Family Three Outcomes depend on assumptions, and all of my analysis has been deterministic. That is, none of the assumptions are allowed to vary simultaneously

©2005, CFA Institute • cfa pubs .org

A Systematic Approach to Asset Location

across time. A more sophisticated approach would involve Monte Carlo simulation techniques in which all inputs can vary simultaneously, thereby offering a range of outputs, not just a single outcome as I have presented here.3 Nevertheless, I can test for the optimal age at which Sam’s asset relocation should occur by varying my assumptions one at a time. For example, all else being equal, • if the tax rate on dividends increases from 20 percent to 30 percent, then the switchover age increases from 61 to 64, which makes intuitive sense because the investment advisor will want to keep equities in the IRA for a longer period; • if the appreciation of equities increases from 7 percent to 8 percent, then the switchover age increases from 61 to 64 for the same reason given above; • if Edward’s age is increased from 20 to 50 (which clearly is not possible if Sam retires at age 55, but the illustration is informative), then Sam’s switchover age decreases slightly from 61 to 60; • if the family’s annual spending decreases, then the switchover age increases; • if the annual turnover of equities increases, then the switchover age increases. One change that surprised me and that required some in-depth analysis was that a decrease in Sam’s and Sally’s age at death, from 90 to 85, increases the switchover age. I had expected a lower switchover age because the step-up in basis of appreciated securities that are outside the IRA is closer. But a hidden factor negates that expectation: When Mr. and Mrs. Sunbelt die, the IRA’s legally required distribution period increases because the distribution period is now based on the son’s life expectancy, not the parents’. Thus, the longer the life expectancy, the greater the tax-deferral benefit within the IRA and the longer an advisor will want to keep equities in the IRA.

a young client, aged 21 years, who can choose whether to have $1,000 of appreciation occur inside or outside the IRA. Certainly, the benefit of having the appreciation occur inside the IRA is obvious—the $1,000 will compound tax-free, whereas investing the $1,000 outside the IRA will provide an after-tax return of perhaps 80 percent or 90 percent of the pretax return. Obviously, then, compounding at a full 100 percent of the return seems better than compounding at 80 or 90 percent of the return. But taxes will eventually be paid on the IRA’s return. The choice of asset location, therefore, depends in part on the rate of tax that must be paid on the higher terminal wealth generated by the $1,000 in the IRA. Therefore, the advisor must ask the following question: Which provides the client with the greater benefit, paying taxes year by year or deferring taxes that will eventually need to be paid? Spreadsheet analysis and stress testing can help answer such a question. Sometimes the results are close, and when they are, the advisor must rely more heavily on nonquantifiable factors. One factor that favors equities in the IRA is that the IRA is a better investment environment: The advisor and the client can make buy and sell decisions without having to consider the capital gains tax, which is not the case for equities traded in a taxable account. Another factor that favors the location of equities in the IRA is that when the client decides to switch the location of assets, it is easier to go from equities in the IRA to fixed income in the IRA than vice versa because equity appreciation within the IRA is not subject to the capital gains tax. Going the other way, however, and having to reduce equities in the taxable account will almost certainly incur some taxable capital gain. Intuitively, therefore, an advisor should see that it will be harder to change strategies in one direction than the other.

General Observations

Conclusion

Investing within a nontaxable account, such as a traditional IRA, seems obviously preferable to investing within a taxable account. But conditions are not always as simple as they seem. For example, consider

Investment advisors should place as much value on asset location as they do on asset allocation because optimal asset location can significantly increase aftertax returns, especially for high-net-worth clients with a long investment horizon. Investment advisors can add value to their services by adopting this systematic procedure (which requires nothing more complicated than a spreadsheet for stress testing of basic assumptions) and providing clients with plausible numbers that help them make more-informed choices.

3 See

Gobind Daryanani and Chris Cordaro, “Asset Location: A Generic Framework for Maximizing After-Tax Wealth,” Journal of Financial Planning (January 2005). These authors use a more complex generalized modeling technique across 10 asset classes. Compared with the naive strategy of equal asset allocation in every account, the optimal locations generated an incremental 20 bps a year.

©2005, CFA Institute • cfa pubs .org

CFA Institute Conference Proceedings • 21

Wealth Management

Question and Answer Session Martin Silfen Question: The extra after-tax performance is enticing, but could the additional costs of each location and the costs of rebalancing erode these benefits? Silfen: The tax costs of rebalancing and relocating asset classes have been factored into the analysis, although other transaction costs, such as brokerage commissions, have not. Question: How do you rebalance to the original asset allocation when your assets are in multiple locations? Silfen: I will tell you the methodology in these spreadsheets, but how you do it in practice is probably a bit more difficult. Assume that the optimal allocation is 45 percent in fixed income and 55 percent in equities. I know my strategy is to make sure that the GST trust is heavily invested in equities. I add up the money in the GST trust and the personal account. I compare the 55 percent optimal allocation to equities with the size of the GST trust. If my overall equity exposure exceeds the assets of the GST trust, no rebalancing is required. Everything in the trust is equities. Therefore, the rebalancing will occur in the personal account. Question: If the family had a family limited partnership as the vehicle for its taxable assets, would it ever make sense to put growth assets in the IRA? Silfen: I have not included this scenario in any of my models, but my guess is that you would probably favor putting the equities into the family limited partnership. Question: How could the modeling techniques incorporate more than two asset classes?

Silfen: The amount of work would increase. For example, if you had three asset classes, you would need to evaluate for six different outcomes. If you had four asset classes, you would evaluate for 12 outcomes. The number of outcomes to be analyzed would be N × (N – 1). Question: How does your analysis include the risk of these location strategies? Silfen: Because my analysis is deterministic, there is only one final outcome for each family. My spreadsheets ignore the risk of the final terminal wealth values. To include risk, you would need to calculate the variability (standard deviation) of final outcomes, and that would require Monte Carlo simulation techniques. Question: Because the IRA is a poor wealth transfer vehicle between generations, would you consume the IRA first before consuming the taxable assets? Silfen: No, you should not normally consume the IRA first; you should consume the taxable assets first. Although conventional wisdom says that the IRA is a poor wealth transfer vehicle, I believe that attitude is not entirely accurate. If the client has sufficient taxable assets, the next generation still ends up with some cash outside the IRA after estate taxes are paid. If we are talking about a significant amount of cash, then the IRA turns out to be a very good wealth transfer vehicle, not a poor one. If, in contrast to the case of Sam Sunbelt, there is $4 million in an IRA and $1 million in a taxable account, then the IRA becomes a poor wealth transfer vehicle, and it is wiser to deplete the IRA first.

22 • CFA Institute Conference Proceedings

Question: In that case, would the change in withdrawal strategy change the optimal asset location decision? Silfen: Yes, absolutely. Changing the withdrawal strategy changes the amount of tax-deferral benefit the IRA gives you, and that changes the asset location decision. In our analysis of Sam Sunbelt, we assumed ongoing real spending of $200,000, and for many years, taxable assets support that spending. Toward the end of Sam’s and Sally’s lives, they may have to dip into their IRA to support their spending. But the more you dip into the IRA earlier than required by law, the earlier you should shift toward putting fixed income into the IRA. Question: As the child in Family Two gets older, suppose she wants income from the GST trust and you have committed to an all-equity strategy. How do you build the portfolio to generate enough income to support the child’s needs without generating enormous capital gains? Silfen: Let me stress two points here. First, the investment advisor is not committed absolutely to anything. The advisor reanalyzes the optimal strategy year after year and changes strategies whenever appropriate. Second, the need for income is not that great because people do not spend income. They spend cash. And the trust could generate a distribution of cash either from income or the sale of assets. Cash is the be-all-and-end-all of the trust’s need to meet the distribution requirements of its beneficiaries, which is recognized by the proliferation of total return statutes across the United States. They recognize that income is not as critical as cash, whatever the source may be. ©2005, CFA Institute • cfa pubs .org

Q&A: Silfen

Question: With respect to Family Three, is there any difference if the deferred account is a Roth IRA rather than a traditional IRA? Silfen: The strategy of locating equities in a tax-deferred account and then switching to fixed income is more pronounced with a Roth IRA than it is with a traditional IRA. For example, if the switchover age using a traditional IRA is 61, then the switchover age for the Roth IRA will be much higher, perhaps 71. The tax-deferral benefit of the Roth IRA exceeds the tax-deferral benefit of a traditional IRA. Question: Again with respect to Family Three, does the switchover age of 61 change if the IRA holder decides to take distributions immediately rather than taking the required mandatory distribution at age 70? Silfen: Yes. I assumed that Sam Sunbelt was taking only the required minimum distribution. But if you assume that he will take more, then that will decrease the

©2005, CFA Institute • cfa pubs .org

switchover age. In fact, if the switchover from equities into bonds is lower than 55, then fixedincome assets will be located in the IRA from the start. Question: Would you be subject to additional fiduciary risk by having the GST trust invested 100 percent in equities during adverse market conditions? Silfen: The strategy is risky if you think of the children solely as trust beneficiaries, but it is not so risky for the children if you look at the broader picture and think of them as both trust beneficiaries and family members. They will get their wealth from two different places: the GST trust and their parents’ personal accounts when they die. But if the children were solely remainder beneficiaries of the trust—that is, if the parents were to change their wills and leave everything in the personal accounts to charity— then the children would be at risk for adverse market conditions.

Question: If a client with a distribution period to his own death has a nonqualified account and a qualified account, is it better to invest equities in the nonqualified account? Silfen: No. Generally speaking, equities in the IRA will still beat equities in the nonqualified account because nonqualified plans do not typically give as much long-term tax deferral as IRAs. When clients reach retirement, they can pick a distribution period that might be, say, 15 years, although it probably won’t be as long as the 60 years in my analysis. Question: What are your thoughts about making a distribution from the IRA to pay estate taxes? Silfen: That is a bad situation, and it occurs when the IRA contains a disproportionately large amount of the family’s liquid wealth, which would give grounds for depleting the IRA before death.

CFA Institute Conference Proceedings • 23

Tax Deferral and Tax-Loss Harvesting Jeffrey E. Horvitz Vice Chairman Moreland Management Company Beverly Farms, Massachusetts

By not realizing capital gains, equity investors can postpone or even fully avoid future tax payments to the government. The value of the tax deferral increases at an increasing rate as the holding period lengthens. For investors who actively trade, however, tax-loss harvesting can be used. Investors then maximize losses that can offset gains in other securities. For tax-loss harvesting to work, however, the present value of uncertain tax savings must more than offset higher transaction costs.

quity investors who are subject to tax can realize substantial tax savings, but they must be disciplined and patient in order to reap these savings.1 In this presentation, I will explain two basic but powerful strategies that investors and their advisors can follow. And although my focus is on equity investments, the principles apply to bond investments as well. I will also discuss the implications for the active versus passive management debate. I will not, however, discuss the myriad of tax shelters, such as insurance wrappers, offshore accounts, and retirement accounts. The advice here is general rather than particular to an investor’s unique tax circumstance, so the conclusions can be broadly applied. Keep in mind that my findings apply to U.S. investors, but the general conclusions apply to taxable investors in other jurisdictions. The key insight about the taxation of investments, particularly stock investments, is as follows: Unlike what most people think, the U.S. federal tax code operates nearly identically to that of a carried interest in a private equity or venture capital partnership. That is, the federal government has an ongoing share of the profits in a portfolio of securities, but the investor decides when the government will receive its share of taxes. This situation can be thought of as an option owned by the investor. So, the investor should exercise the option when it is most suitable for him or her, rather than for the government. I will compare this carried interest analogy with conventional wisdom later.

E

1

By “tax,” I mean the U.S. federal capital gains tax, but I urge investors and advisors to take into consideration state and, in some instances, local taxes.

24 • CFA Institute Conference Proceedings

Short-Term vs. Long-Term Capital Gains Taxes Why do active investment managers trade taxable accounts so often? Because they seek positive alpha for their clients. But the catch, from a tax perspective, is that frequent trades in mutual funds, private accounts, and broker-advised accounts can trigger recognized short-term gains—defined as those that occur within 12 months. This one-year horizon is of utmost importance because it means the difference between paying taxes at the same rate as ordinary income, which could be as high as 35 percent, versus paying the 15 percent long-term federal capital gains tax. With taxes, timing means everything; even one day can make a dramatic difference. So, how much more return do investors need in order to justify paying the higher short-term capital gains tax? If one assumes that the short-term capital gains tax rate is 35 percent and the long-term capital gains tax rate is 15 percent, then the ratio of after-tax returns is (1.0 – 0.15)/(1.0 – 0.35) = 0.85/0.65 = 1.31. So, all else being equal, active investment managers who make only short-term trades need to earn a return that is 31 percent higher than that of managers who make only long-term trades. This hurdle rate is staggering. I believe that few managers can break that hurdle in investment returns and certainly not with any consistency for their taxable clients. In addition, this hurdle rate makes no allowance for leakage caused by transaction costs, which would effectively increase the hurdle rate because managers who trade more often pay more in transaction costs. ©2005, CFA Institute • cfa pubs .org

Tax Deferral and Tax-Loss Harvesting

And I have even more bad news: Short-term capital losses have to be used first to offset short-term capital gains, and then they can be used only to offset 15 percent long-term capital gains, and only then on up to $3,000 of ordinary income. So, apart from dividends, the worst possible form of income for most people is short-term capital gains. Yet, many investment managers generate lots of it. The question then becomes whether investment managers are able to justify their trading activity. Table 1 is from an article by Robert Arnott, Andrew Berkin, and Jia Ye, who showed the percentage of mutual funds that beat the Vanguard S&P 500 Index Fund.2 Even taking into account the costs associated with the Vanguard fund, one can see that it is not a big number. In fact, shockingly few managers can beat this benchmark. Table 1.

Percentage of Mutual Funds That Beat the Vanguard S&P 500 Index Fund 10 Years 1989–1998

15 Years 1984–1998

20 Years 1979–1998

Before tax

14%

5%

22%

After tax including final liquidation

12

5

16

Pure tax savings over average fund

0.13

0.23

0.25

Source: Arnott, Berkin, and Ye (2000).

To some extent, the amount of alpha, what Arnott et al. label as the “pure tax savings,” looks puny here. And it is, but part of the reason for that small alpha is the turnover in the S&P 500, which itself triggers plenty of tax events. One could actually create passive portfolios that are much more tax efficient than the S&P 500. Another reason for the small “pure tax savings” is that the dividend component historically has been significant. Of course, dividends are paid periodically, and the size and timing of dividends are out of the control of investment managers. Aside from the anomaly of the 15-year period, which is probably a reflection of market conditions over that time, I believe that only about 15 percent of active managers can beat the Vanguard S&P 500 Index Fund on an after-tax basis, including all the tax at final liquidation. What is the likelihood of an investor picking one of those managers? What is the probability of picking three of those managers? It is low in both cases. 2

See Robert D. Arnott, Andrew L. Berkin, and Jia Ye, “How Well Have Taxable Investors Been Served in the 1980s and 1990s?” Journal of Portfolio Management (Summer 2000):84–93.

©2005, CFA Institute • cfa pubs .org

Algebra of Deferred Taxes The longer investors can go without realizing gains, the more valuable the gains are to investors because the compounding works for the benefit of investors and not the U.S. Treasury. Investors themselves determine when to liquidate an unrealized gain. In this context, taxation can be thought of in the same way as the carried interest of a limited partnership (such as private equity or venture capital). That is, the investor and the government will earn their share of the gains and taxes on the gains, respectively, at the end of the holding period, allowing the gains to grow tax free during the holding period. This concept is essential for understanding the simple mathematical mechanism of how tax deferral adds value. The following example will help explain how tax deferral works. Imagine a steady state rate of return, denoted by r. So, in each period, an investor’s portfolio grows by 1 + r. An investor’s terminal wealth is the product of 1 + r for each compounding period times the investor’s initial investment:

(1 + r1) × (1 + r2 ) × ... × (1 + rn ) . First, consider the case where an investor recognizes gains annually and must pay the corresponding capital gains tax each year, denoted by τ. The preceding equation can be modified to recognize the investor’s after-tax return:

⎡⎣1 + r1 (1 − τ )⎤⎦ × ⎡⎣1 + r2 (1 − τ )⎤⎦ × ... × ⎡⎣1 + rn (1 − τ )⎤⎦ . This equation, of course, assumes that the investor pays long-term capital gains tax at a constant rate over the entire holding period. Effectively, the investment is compounding at a lower rate. Now, consider the extreme case of 100 percent tax deferral. (I refer only to capital gains tax deferral because dividends cannot be deferred; they are taxed when they are paid.) At the end of the investor’s holding period is a final-period tax payment that can be thought of as a negative single-period return, denoted by rtax:

(1 + r1) × (1 + r2 ) × ... × (1 + rtax ) . It really does not matter where that negative return occurs from a purely mathematical perspective; but for this example, assume that it will always occur at the end of the holding period. A variant is a Roth IRA, where tax is paid only at the beginning. Remember that the government owns a profit interest in the returns, and the best that an investor can do is to defer the payment of taxes for as long as possible so that the compounding effect works to the investor’s full advantage over the holding period. Therefore, the compounding effect can work to its full potential over

CFA Institute Conference Proceedings • 25

Wealth Management

the holding period and reduces the taxes payable to the government, as shown by the following equation:

(1 + r1) × (1 + r2 ) × ... × (1 + rn ) × (1 − τ ). I will now put these simple equations to work. Assume that investments grow at an annual rate of 6 percent and that the applicable tax rate is the longterm capital gains tax rate of 15 percent. Figure 1, which shows the growing divergence between terminal wealth when taxes are paid annually and terminal wealth under a buy-and-hold scenario where taxes are paid at liquidation, demonstrates the power of tax deferrals, particularly if liquidation can be deferred until at least 10 years. Before 10 years, the difference between the annual payment of taxes and deferral is not material. As the length of the holding period increases, however, the power of nonlinear compounding and deferred taxes becomes stronger, as can be seen by the spacing between the dotted line that denotes the annual tax payment and the solid line that denotes the tax deferral. Notice that the difference between the deferral and annual tax lines increases as the holding period lengthens. The spread illustrates the dilutive effect of stalling the government’s share of investment gains for as long as possible. Horvitz and Wilcox demonstrated this effect in Figure 1.

a recent article.3 It is not something for which there are any exceptions, or at least important exceptions. This “carried interest” analogy will always work.

Short Deferral Periods Are Not Worthwhile Interestingly, many of those who have commented on tax deferral, those who sell tax-deferred accounts, and those who offer tax-deferred advice talk about tax deferrals over time periods shorter than 10 years. Figure 1 shows that material deferral benefits do not start until after 10 years. So, when advisors are talking to clients about tax deferral, and if the clients are not prepared to have serious deferral, meaning somewhere around the 10-year mark and beyond, then it really is not worthwhile. Table 2 examines the data from Figure 1 as a percentage of the tax-free terminal wealth. It uses the same return and tax assumptions and investigates, again, capital gains only, not dividends. Notice that at five years to liquidation, only a 40 bp difference exists between paying annual tax and deferring the tax. Think about how many products are being sold with 3

Jeffrey E. Horvitz and Jarrod W. Wilcox, “Tax Management of Stock Portfolios,” Journal of Investing (Spring 2005):83–89.

Terminal Wealth at Liquidation Amount ($) 5,500 5,000 4,500 Deferred Tax 4,000 3,500 Annual Tax 3,000 2,500 2,000 1,500 1,000 0

1

2

3

4

5

6

7

8

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Years to Liquidation

Notes: Price change only; dividends excluded. Annual rate of return of 6.0 percent, and capital gains tax of 15.0 percent.

26 • CFA Institute Conference Proceedings

©2005, CFA Institute • cfa pubs .org

Tax Deferral and Tax-Loss Harvesting

Table 2.

Terminal Wealth as a Percentage of Tax-Exempt Investing

Years to Liquidation

No Tax

Tax Deferred

Annual Tax

5

100.0%

96.2%

95.8%

0.4%

10

100.0

93.4

91.8

1.5

15

100.0

91.3

88.0

3.3

20

100.0

89.7

84.3

5.4

25

100.0

88.5

80.8

7.7

30

100.0

87.6

77.4

10.2

Spread

Note: Price change only; rate of return = 6 percent.

a three- to five-year tax deferral for which clients are paying a material fee to earn a negligible benefit. (I assume that the deferral is complete and that the withdrawal occurs all at once.) These short-term tax deferrals are simply not worth doing. The spread becomes material starting after 10 years. At the extreme in this illustration, the patient investor is rewarded with about 10.2 percent more wealth after 30 years. Incidentally, I have also investigated the spread assuming modest dividend payments and corresponding income tax each year, and the spreads are not overwhelmingly different. For 1.5 percent paid-in dividends each year, the corresponding taxes paid at the rate of ordinary income, a 9.2 percent spread still exists after 30 years. Tax deferral over long holding periods is worthwhile, but the benefit over shorter periods greater than one year is negligible. So, if somehow investors believe in active management, or they believe they have the hot stock, or they believe they have an edge, they should not give that up just to get tax deferral unless they can get the deferral for more than 10 years. And this result also has a lot of bearing for clients who have highly concentrated portfolios. If the expectation is that their highly concentrated position, for one reason or another, will be sold and capital gains recognized, meaning for tax purposes in five years or just a few years longer, then they should not spend a whole lot of effort trying to defer that tax. It is just not worth it in that case.

Mitigating Estate Taxes Estate taxes are important, but not a lot is written about them. Table 3 shows proposed estate tax changes under current law. President Bush in 2001 signed into law the Economic Growth and Tax Relief Reconciliation Act of 2001, which gradually reduces the maximum tax rate and increases the annual exclusion amount (exemption) up until 2009. ©2005, CFA Institute • cfa pubs .org

Interestingly, the act fully repeals any estate tax beginning on 1 January 2010. But absent any U.S Congressional intervention, this repeal will “sunset” or expire on 1 January 2011, returning to the old tax system, which includes the 55 percent top tax rate. Of course, no one knows for sure what will happen beyond 2010. Not a single trust and estate attorney believes that the old tax system will be put back in place in 2011. And it would be unreasonable, I believe, to present clients with projections based on a reversion to a 55 percent maximum tax rate. At the same time, I would not be surprised to see the repeal expire with a new tax rate in place for 2011 and beyond or to see the estate tax schedule changed by Congress before 2010. Table 3.

Estate Tax Changes, 2005–2011

Year

Maximum Rate

Exclusion Amount

2005

47%

$1,500,000

2006

46

2,000,000

2007

45

2,000,000

2008

45

2,000,000

2009

45

3,500,000

2010

(Estate tax repealed)

2011

55

1,000,000

My axiom—deferring gain recognition until death is always better—holds true even if the estate is going to be subject to estate tax. Table 4 has some simple examples that illustrate the power of estate taxes. This simple matrix can be prepared for any client. In this instance, I assume the cost basis is $1,000, the future market value is $1,500, the longterm capital gains rate is 15 percent, and the estate tax rate is 45 percent. If the estate is tax exempt and if the tax deferral lasts until death, then a taxable investor is converted into a tax-exempt investor with zero taxes payable. If the estate is tax exempt but the capital gain is recognized before death, then the investor pays $75 in capital gains tax [($1,500 – $1,000) × 0.15]. If the estate is not tax exempt but the capital gain is not recognized before death, then the investor pays $675 in estate tax [$1,500 × 0.45]. If the estate is not tax exempt and the capital gain is recognized before death, then the investor pays $75 in capital gains tax and $641.25 in estate tax [($1,500 – $75) × 0.45] for a total of $716.25. Table 4.

Total Tax Paid

Gain Recognized before Death?

Exempt Estate

Nonexempt Estate

Yes

$75.00

$716.25

No

0

675.00

CFA Institute Conference Proceedings • 27

Wealth Management

Of course, the ultimate value of deferring taxes until death will depend on the status of the estate. Following are four possibilities: The estate is below the exemption, the estate is left to charity, the estate is subject to tax, and the estate is left to a spouse. Estate Is Below the Exemption. When the estate is below the exemption, no tax is paid. There will be a step-up in basis, and the effect is to compound the deferrals just like for a tax-exempt investor. So, for older or ill clients who do not intend to spend all of their investment gains before they die, deferring the sale until after death is a way to make powerful changes in the residual estate, the amount of the estate that is left. Estate Is Left to Charity. For an estate left to charity, the effect is the same as for the estate being below its exemption. For those investors who have charitable bequests in mind, tax deferral is quite valuable. Just think of the effects of long-term compounding on the value of the estate passed to the charity. The investor has been converted to tax-exempt status without ever having to consider a private foundation. For some reason, this bequeathing to charity is not typically recognized in the investment community. Estate Is Subject to Tax. When the estate is subject to tax, there are large tax savings if an investor does not make inter vivos liquidations. The result is to reduce the effective long-term capital gains tax. The savings is equal to the long-term capital gains rate times the estate tax rate. So, if the estate tax rate is roughly 50 percent, then the investor cuts his or her effective long-term capital gains tax in half from 15 percent to roughly 7.5 percent. Estate Is Left to a Spouse. If an investor leaves an estate to a spouse and the spouse takes the basis, then the investor can perpetuate the tax deferral. Think of it in terms of estimating the life span of this as a last-to-die insurance policy. The actuarial expectancy could be quite long and may provide a powerful way to elongate the long-term tax deferral. So, for a married couple (one not in divorce proceedings) where one spouse is expected to die before the other, which is almost certainly the case, taking advantage of the extension of tax deferral can be very beneficial.

Tax-Loss Harvesting Some investors believe in active management’s ability to create positive alpha. For these investors who recognize capital gains through trading, capital gains taxes can be mitigated. 28 • CFA Institute Conference Proceedings

Tax-loss harvesting means taking voluntary losses for the sole purpose of creating a current tax deduction to offset other gains. Of course, investors have other reasons to sell securities, but the sole motivation for tax-loss harvesting is to trigger tax savings through realized losses now to offset gains elsewhere in the portfolio. For tax-loss harvesting to work, the present value of the tax deduction must meaningfully exceed the total transaction cost. This is the theory, but it is more difficult to quantify in practice. Furthermore, the amount of loss taken on each trade should be maximized. Other more complicated rules also apply—for example, in a highly volatile portfolio having multiple tax lots with multiple basis—but the basic principle holds. The investor needs to keep a record for gains as well as for losses. Without this disciplined accounting, the investor cannot be successful at tax-loss harvesting. Therefore, one of the essential tools that clients must have for tax-loss harvesting is adequate record keeping.4 Normally, the transaction costs are round-trip costs because, using my definition of tax-loss harvesting (i.e., for the sole purpose of creating realized losses), the investor is not really trying to dispose of the security. Rather, the investor is trying to keep the security. So, the investor somehow has to get it back. Getting the security back would be easy except for the Internal Revenue Service’s wash-sale rules. The wash-sale rules state that if an investor sells a security for a loss, the investor cannot buy the same security back within 30 days of the transaction or the loss will be disallowed for tax purposes. To further complicate matters, the investor also cannot have purchased the same security, or essentially the same security, for 30 days prior to the transaction. So, the investor has a 61-calendar-day window because the day of the trade itself is not counted. That is a long waiting period. The problem with the purchase of the substitute security is that it is purchased at a lower tax basis due to the loss, and this lower basis could be recaptured through a higher realized capital gain in the future. This point is very important to understand because a lot of the products that engage in tax-loss harvesting ignore this gain, unless, of course, the investor can defer realized gains until after death, as I have already discussed. 4

Berkin and Ye discuss the highest in, first out (HIFO) method, whereby the shares with the highest cost basis are sold first. This approach minimizes the capital gains tax. See Andrew L. Berkin and Jia Ye, “Tax Management, Loss Harvesting, and HIFO Accounting,” Financial Analysts Journal (July/August 2003):91–102.

©2005, CFA Institute • cfa pubs .org

Tax Deferral and Tax-Loss Harvesting

The present value of the tax deduction is difficult to calculate. But the deduction effectively puts extra money in the investor’s pocket, and so in this context only, tax-loss harvesting is like an interest-free loan from the government. (Do not confuse this with tax deferral, where the tax deferral is a future profits participation between the investor and the Treasury.) That is, investors can effectively spend the money derived from recognizing the tax loss and taking the tax deduction. When this recapture occurs and how much it will be are hard to tell. There is no algorithm that yields an exact solution, but it can be modeled based on probability-adjusted recaptures that extend into the future. Then, the investor can apply a term structure of discount rates to each future recapture. This process gives a crude estimate of the present value of the recaptures. Going further, the investor can calculate the net present value (NPV) as the difference between the positive present value of the recaptures and today’s transaction costs, which would include brokerage costs, commissions, and market impact, if the investor has a model for it. Transaction costs will be the most predictable in the NPV equation, but keep in mind that they can vary with investor size. That is, transaction costs are low for small traders and people who invest in mutual funds, they are really low for big institutions, and they are really high for everybody in the middle. Before taking the tax-harvest loss, there should be a significant positive NPV because of the high uncertainty of the recapture. So, the investor does not want to just hit the breakeven point by a dollar. The calculated NPV must be large because its true value is not known. If the investor is wrong and the NPV is, in fact, negative, then the investor will lose money and he or she will never know it. Thus, the operational problems of tax-loss harvesting are (1) how does the investor calculate the NPV of the tax deduction and the transaction costs when the recapture is uncertain and (2) how does the investor know when to trigger a loss recognition? That

©2005, CFA Institute • cfa pubs .org

is, how does the investor know when the share is trading low enough? With perfect insight and hindsight, the investor would pick the bottom of the market for that particular stock, sell it, rebuy it, and then write it up. Of course, investors would do that anyway if they could, almost irrespective of the tax-loss harvesting. So, obviously, if it were that easy to do, everybody would be doing it for other reasons. Some people have suggested creating a rule of thumb, such as taking losses that are perhaps 30 percent of face value. And some researchers have simulated possible solutions, but I am interested in an algorithmic solution.

Conclusions Investors subject to tax can realize substantial tax savings if they act and plan accordingly: • Avoid short-term capital gains. Investors should wait for long-term capital gains. This should probably be the first rule of taxable investment management. • Federal taxation is like a carried interest in profits. The investor and the government own the profits, and the share of the capital gains will be distributed to the investor and the government at the end of the holding period. • Tax deferral becomes more valuable as the holding period lengthens. The deferral is not valuable for at least 8–12 years, under normal ranges of market performance. (Note that earlier I said that the minimum period was 10 years, but of course, that was based on my assumption of a 6 percent compounded return.) • The investor’s estate tax situation is critical. Tax deferral is particularly valuable when no estate tax will be due. • Tax-loss harvesting has to solve two key problems. Tax-loss harvesting must address how to create a positive NPV out of the current tax savings when the potential future recapture is uncertain and how to maximize the loss for a given single transaction cost.

CFA Institute Conference Proceedings • 29

Wealth Management

Question and Answer Session Jeffrey E. Horvitz Question: How does volatility affect the tax-loss harvesting strategy? Horvitz: The more volatile the stocks, the more up and down you can get. So, you have more opportunity for tax-loss harvesting. Question: What is the best way to measure an investor’s after-tax performance? Horvitz: If you want to know how well you are doing, create the unrecognized gains for purposes of performance measurement as if liquidated and taxed at whatever period you are looking at. Question: What if tax rates are 70 percent when we withdraw? Shouldn’t we give value to the “devil” we know versus deferring to unknown future tax rules? Horvitz: Remember, it doesn’t matter what the tax rate is, just as long as you can defer this all the way to a tax-exempt estate. Therefore, the problem of tax rates is irrelevant. In terms of the mathematics, we know that the deferral spreads will increase as the tax rate increases; it pays even more to defer as tax rates increase. Tax rates right now are really about the lowest they have been since World War II. About 20 years ago, the long-term capital gains rates were either 20 percent or 28 percent, jumping more or less back and forth every few years. Today, the long-term capital gains tax rate is 15 percent. There are reasonable

prospects that one day we will see higher rates. Question: If an estate is left to the spouse, then half the amount is supposed to be updated with the date-of-death value. Then, there’s a yes or no answer for this deferral to work. Shouldn’t you gift all the assets to the second-to-die spouse while he or she is alive, if you can predict it? Horvitz: To the extent that you have exemption left in the estate, then it is really an estate-planning question, which is pretty straightforward: Take the step-up in basis, and in all events, use up all of your exemption first. The remainder, then, goes to the spouse with the old basis, which you could also do directly or in a qualified terminal interest property trust. To the extent that there are grandchildren involved, you want to use up your complete generation-skipping transfer tax exemption. The ideal estate-planning device is what’s called a dynasty trust, in which you actually put the money in trust for all future generations. In the right jurisdictions, you can effectively defer taxes forever. There are enough jurisdictions to do this in, but the major drawback is the legal cost. Question: Should tax-gain harvesting be used in the context of a tax-loss carryforward? Horvitz: No, because normally there is a significant amount of

30 • CFA Institute Conference Proceedings

gain that is involuntary. If you could control all your gains and losses, which you can’t, the game would be quite different. Tax-loss harvesting is the voluntary loss recognition for purposes of offsetting gains. Do those gains have to occur right away? No, not particularly. I think it is more valuable to use the voluntary losses in most all cases to shelter the involuntary future gains. Question: If you apply the taxdeferral advice to an investor who owns a concentrated position in IBM with a $2 cost basis, how do you deal with the trade-off of diversifying the investor’s portfolio versus holding the deferral taxes as long as possible? Horvitz: Irrespective of tax deferral, I would just sell the IBM stock. The amount of concentration risk is so extreme and the benefits of diversification so great that I can’t imagine rationally maintaining this position. In general, if you already own a well-diversified portfolio of stocks, then I don’t think there is a trade-off. If you invest in a diversified index fund, then you can mitigate the risk of having too much concentration in any particular stock. To be somewhat cynical, an active manager has only a 15 percent chance of beating the S&P 500, which is even more compelling evidence to invest passively; you would avoid the diversification issue entirely.

©2005, CFA Institute • cfa pubs .org

Investing with a Tax-Efficient Eye Robert Gordon President Twenty-First Securities Corporation New York City

Investors can gain exposure to a bond or stock in many ways—some of which are tax efficient and some of which are not. For instance, bond investors can realize long-term gains today and pay less tax overall tomorrow. And depending on the length of their investment horizon, equity investors can follow one of three tax-efficient strategies. Anomalies in the tax code also allow derivatives, especially options and swaps, to mitigate tax liability without altering investment returns. And finally, investors in collective investments, such as mutual funds and hedge funds, can use tax-efficient strategies as well.

n this presentation, I discuss a number of transactions, some of which may be unfamiliar. By including the unfamiliar, I hope to open advisors’ minds to new ways of managing taxable clients’ portfolios through the use of innovative transactions to improve after-tax performance. In my experience, portfolio managers tend to resist complex transactions, especially those involving derivatives and that require time and patience not only to explain to clients but also to properly construct. But I believe that advisors owe it to their clients to explore ways to improve the tax efficiency of their portfolios, whether their holdings are individual securities or collective investments, such as mutual funds and hedge funds. Fortunately, anomalies within the U.S. tax code enable savvy advisors to improve their clients’ after-tax performance.

I

Academic Findings Groundbreaking research by Arnott, Berkin, and Ye found that investors should care about taxes and that they ought to have a disciplined approach to harvesting tax losses.1 But do individual investors care about taxes? Two studies of tax efficiency of individual investors’ portfolios shed some light. Barber and

Odean used data from 1994 and 1998 to investigate the tax awareness of individual investors.2 For part of their study, they analyzed stock trades at a large discount brokerage house of tens of thousands of individual investors, many of whom had both taxable and tax-deferred accounts, such as individual retirement accounts and Keogh plans. They found that most individuals could improve their trading efficiency. For example, taxable investors took their gains more frequently than their losses. Interestingly, the researchers further observed that this behavior was almost double for married couples compared with the behavior of singles. Ivkovic´ , Poterba, and Weisbenner studied whether tax incentives influence how investors realize capital gains and losses. They used the same trading data as Barber and Odean for a large discount brokerage house but for a six-year sample period from 1991 to 1996. 3 Ivkovic´ , Poterba, and Weisbenner compared the portfolio transactions made by the same person in two different accounts: the regular, taxable account and the tax-deferred account. They found stronger evidence than Barber and Odean did that investors care about tax. For example, they found a higher tendency for capital gains to be unrealized 2

1

Robert D. Arnott, Andrew L. Berkin, and Jia Ye, “How Well Have Taxable Investors Been Served in the 1980s and 1990s?” Journal of Portfolio Management (Summer 2000):84–94. See also Jeffrey Horvitz’s presentation in this proceedings, where this research is summarized.

©2005, CFA Institute • cfa pubs .org

Brad M. Barber and Terrance Odean, “Are Individual Investors Tax Savvy? Evidence from Retail and Discount Brokerage Accounts,” Journal of Public Economics (January 2004):419–442. 3 Zoran Ivkovic´, James Poterba, and Scott Weisbenner, “TaxMotivated Trading by Individual Investors,” NBER Working Paper No. 10275 (February 2004).

CFA Institute Conference Proceedings • 31

Wealth Management

in taxable accounts than in tax-deferred accounts. This tendency was stronger for larger transactions and intensified as the investor’s holding period lengthened. Individual investors were somewhat efficient in that tax-loss selling occurred throughout the year, although it was most pronounced in December, especially if an investor had realized capital gains earlier in the year. Collectively, this analysis of the same dataset implies that investors trade in ways that are tax efficient. Nonetheless, Ivkovic´ , Poterba, and Weisbenner were disappointed: At times, they found tax-exempt municipal bonds in the IRA account, meaning that asset allocation between taxable and tax-deferred accounts was suboptimal. So, if these studies suggest that investors care about tax, then so should brokerage firms. In fact, given that clients could be trading in a suboptimal fashion, advisors perhaps have a duty to educate them. That is, advisors can find ways to implement tax-efficient trading strategies without distorting the investor’s performance. Savvy practitioners on Wall Street have devised ways to replicate positions and synthesize investments. Pretax returns for the various positions and instruments may be the same, but the after-tax returns can be quite different. Thus, some investors ought to re-evaluate their positions and instruments to maximize their returns.

U.S. Bond Strategies Most brokerage firms, individual brokers, and investment advisors are well aware of bond swaps— the simultaneous sale of a bond with capital losses and purchase of another, similar bond. Near the end of the tax year, investors may hear: “Get out of your muni bond and switch into another bond and take the loss.” Sometimes that transaction is genuinely in the client’s best interest. But without being too cynical about the brokerage industry, of which I am a part, I would like to point out that getting out of one bond and into another provides a decent commission. Nevertheless, when interest rates are very low, investors should take their long-term gains on taxable bonds and then immediately repurchase those same bonds; there is no wash-sale rule for gains. Keep in mind that a brokerage house cannot charge an investor too much to sell a bond and to repurchase it immediately. Although I am talking about trying to be tax efficient, advisors do not want to distort an investor’s results. For example, say that several years ago an investor bought a U.S. Treasury note at par ($100) when interest rates were at 5 percent. Today, interest rates are at 2 percent and the note matures in one year. Theoretically, that bond will trade at $103. 32 • CFA Institute Conference Proceedings

If the investor does nothing, then in one year’s time the $5 coupon will be taxed as ordinary investment income at rates as high as 35 percent. But if the investor sells the bond today at $103, and because it has been held for more than 12 months, the long-term capital gains tax of 15 percent applies to that $3. The wash-sale rules do not prevent an investor from realizing a capital gain. If an investor sells something and buys it back within the statutory 31 days and if the investor has a capital gain, the U.S. government is only too pleased to tax the investor on the gain. So, it is all right to sell the bond and buy it back a second later if the investor has a gain.4 When the investor repurchases this bond at $103, it trades at a premium because it will mature one year from today at $100. Under Section 171 of the IRS code, the federal government allows the investor to amortize the bond premium of $3. When the investor pays tax on investment income, the amount payable will be based on a gross 5 percent coupon, but the investor can deduct $3 of amortization against the $5. In essence, the investor is paying tax on only $2 of net interest rather than the full gross $5. So, the investor faces two choices: • Do nothing and pay a 35 percent ordinary investment income tax on $5, making the total tax payable $1.75, or • Sell the bond and buy it back immediately, pay 15 percent long-term capital gains tax on $3, and pay a 35 percent tax on only $2, making the total tax payable $1.15. Obviously, the investor is better off voluntarily paying the 15 percent long-term capital gains tax today in order to pay lower total tax in the future. Solely by selling and repurchasing this hypothetical government or corporate bond, this investor’s tax bill is more than one-third lower than if he or she had done nothing. Clearly, this decision will influence the bond’s after-tax performance. If the bond in question were to mature 10 years from now, the investor could end up paying a large amount of tax today with the benefit spread out over the next 10 years. Therefore, the correct analysis would include the calculation of an internal rate of return (IRR) to derive the estimated benefit. Our analysis at Twenty-First Securities shows that if an investor lives in a high-tax state, such as New York (remember that the investor will pay state tax as well as federal tax), the after-tax IRR on a government bond due in 10 years is 11.72 percent.5 This figure 4 If the bond swap is done for tax purposes to realize a capital loss, the investor must avoid swapping into what the IRS deems to be the same security or the wash-sale rules will apply. 5 Our website has an online calculator that can be used to conduct a similar analysis: www.twenty-first.com/bond/index.htm.

©2005, CFA Institute • cfa pubs .org

Investing with a Tax-Efficient Eye

includes an eighth of a point for friction costs of getting in and out. If the bond were due in only one year, the after-tax IRR would become a breathtaking 74.96 percent—for doing nothing more than resetting the holding period. The IRR analysis and sell/repurchase decision apply only to government bonds and corporate bonds, not to tax-free municipal bonds. One universal tax decision rule holds with respect to municipal bonds: Do not engage in the sale and repurchase of the bond prior to maturity. Keep in mind an important caveat when analyzing a government bond: Effectively, the investor elects to pay a capital gains tax in the state where that interest income would have been tax-free later. Some nuances make the analysis more difficult than one might think.

S&P 500 Index Strategies A broad basket of large-cap equities, which can be proxied by the S&P 500 Index, is another basic building block that investors use in their portfolios. Please note that my discussion in this presentation is not about how to manage the individual securities in investors’ portfolios more efficiently but, rather, how to choose the investment vehicles that are more efficient from a tax perspective. Investors in recent years have embraced Standard & Poor’s Depositary Receipts (SPDRs, or “Spiders”). SPDRs and the newer exchange-traded funds (ETFs), such as iShares, are a cost-effective way to quickly build a diversified portfolio. They are generally tax efficient but only for long-term investors. As Jeffrey Horvitz explained in his presentation, if an investor defers

Figure 1.

capital gains until death, then (as the tax laws read today) capital gains taxes can be forgiven upon the death of the investor.6 So, if an investor can buy SPDRs or an ETF or even a mutual fund that minimizes distributions each year, then all the investor’s unrecognized gains continue to grow and possibly can be forgiven at death. It is estimated, however, that between 80 and 85 percent of all trading in SPDRs and iShares is done for fewer than 90 days. Therefore, profitable investors would be taxed at 35 percent, the rate for shortterm capital gains. The implication is that investors are drawn to these popular vehicles not by the prospect of tax efficiency but, rather, by aggressive marketing campaigns. To get any tax efficiency, investors must commit to an investment horizon greater than one year. Indeed, at the extreme, the tax efficiency is maximized when an investor purchases a SPDR or ETF and holds it forever. Still, advisors can find taxefficient vehicles for those taxable investors with short time horizons. Time Horizon Impact on Strategies. Investors can own the S&P 500 in many ways, so at TwentyFirst Securities, we determine the most tax-efficient investment vehicle by time horizon, as shown in Figure 1. ■ Horizon less than one year. If an investor will be in and out of the S&P 500 in less than one year, then as I have explained, the SPDR will be taxed at 35 percent. But if the investor seeks exposure through exchange-traded index options, realized gains will be taxed at a blended capital gains rate of 6

See Jeffrey Horvitz’s presentation in this proceedings.

Effect of Time Horizon on Choice of Investment Vehicle Long Position What time horizon does the investor have?

Less than 1 year

1−5 years

More than 5 years

Futures or BroadBased Options

Equity Swap

Outright Ownership

All gains and losses are treated as 60 percent long term and 40 percent short term.

Cost to carry is deductible. Losses are ordinary. Gains are long term.

Interest expense is deductible to the extent of investment income. Gains and losses are not recognized until realized.

©2005, CFA Institute • cfa pubs .org

CFA Institute Conference Proceedings • 33

Wealth Management

only 23 percent.7 So, if an investor will be invested for less than a year, index options make a lot of sense. Compared with a mutual fund, options do not distribute any gain, so even if a stock leaves the S&P 500, there will not be a recognized gain on an index option; a mutual fund will have some gains to pay as it removes that stock from its portfolio. The return of the S&P 500 can be replicated, more or less, with a deep in-the-money index option (options that are in the money behave more like the underlying than those that are out of the money). Although the pretax returns of a SPDR and an index option are the same, the after-tax returns will be different because of the quirkiness of U.S. federal tax law. So, with index options, the investor chooses to be taxed at 23 percent if he or she makes money. If the investor loses money, however, he or she is burdened because only 40 percent of the capital loss will be calculated at the higher short-term tax rate of 35 percent and 60 percent of the loss will be calculated at the lower long-term tax rate of 15 percent, again yielding a blended tax rate of 23 percent. My discussion so far has been on bullish sentiment—when one thinks the market will go higher. I will now discuss tax efficiency using options for bearish strategies. Investors consider short sales to be un-American (and so does the IRS). If an investor makes a short sale, no matter how long the holding period until a gain is realized, it will always be taxed as a short-term gain. When the investor owns a listed put from an exchange, the capital gains will be taxed at the now-familiar blended rate of 23 percent. But if the investor had bought a put from a dealer, then any short-term gains would have been taxed at 35 percent and the long-term gains would have been taxed at 15 percent, the only difference being the issuer of the contract. ■ Horizon between one and five years. Equity swaps are available only to investors with a high commitment, such as $1 million or more; they are not available to everybody. But the tax features of an equity swap are nonetheless interesting. What is an equity swap? It is a private contractual agreement to exchange cash flows at intervals over a predefined period. A swap has two counterparties, and each agrees to make the contractual payments to the other over the life of the swap. In this way, an equity swap, indeed all swaps, involves credit or counterparty risk. Every swap involves an unknown or floating payment. In an equity swap, that payment could be the total return (dividends and 7 For options that trade on an exchange, all gains or losses are considered 60 percent long term and 40 percent short term, which yields a blended 23 percent tax rate. The same OTC option purchased from a broker does not get this tax advantage.

34 • CFA Institute Conference Proceedings

capital gains) on the S&P 500. In exchange for receiving the total return on the S&P 500, the other counterparty pays a fixed return. In the eyes of the IRS, if an investor makes or takes contractual payments, the payments are taxed as ordinary income or loss. But if the investor terminates an equity swap before it expires, then the swap is taxed like a capital asset. So, the investor can decide how he or she would like to be taxed after knowing the outcome of the investment. For example, if an investor is two years into an equity swap and the market has declined, the investor should not terminate the swap early. He or she should continue to make the contractual payments because the losses will accrue as ordinary losses, rather than as capital losses constrained by the limit of $3,000 per tax year. In contrast, if the investor had a profit over a holding period longer than 12 months, then he or she should terminate the swap sometime before its expiration. Again, it becomes a capital asset, and the investor recognizes a long-term gain. I know of no other security for which investors can bank ordinary losses when they lose money, recognize a long-term capital gain when they make money, and have the luxury of deciding when to recognize each of these tax-mitigating strategies. ■ Horizon more than five years. If the investor is going to hold an investment for more than, say, five years—maybe forever—then he or she should own the asset itself, not a derivative, because most derivative dealers will not write an equity swap for more than five years. Other Rationales for Various Strategies. Table 1 summarizes the features of some of the alternatives I have been discussing. Notice from the table that economic reasons, as opposed to tax-efficiency reasons, might suggest the use of index options over ETFs. For example, ETFs have management and custodian fees built into them. And ETFs have a slight amount of tracking error against the index that they are designed to mimic; index options do not, provided they are held to expiration. Therefore, based on these reasons, using the listed or the unlisted index options is usually better than using ETFs. An investor may trade an ETF anytime during the day, but the cost of that trade is the bid–ask spread associated with the cost of buying and selling, as well as a standard stock commission for a roundtrip transaction. But if an investor purchases an open-end fund, it does not have that bid–ask spread; the open-end fund is bought and sold at the net asset value, and generally, an investor would not pay a commission. With options, investors will definitely pay a commission. ©2005, CFA Institute • cfa pubs .org

Investing with a Tax-Efficient Eye

Table 1.

S&P 500 Investment Alternatives

Item

ETFs

Open-End Index Funds

Listed Index Options

Unlisted Index Options

Management fee

Yes

Yes

No

No

Tracking error

Yes

Yes

No

No

Commissions on purchase/sale

Yes

No

Yes

Yes

Short-term gains taxed at

35%

35%

23%

35%

Long-term gains taxed at

15%

15%

23%

15%

Marked to market at year-end

No

No

Yes

No

Distribution of gains

Yes

Yes

No

No

Possible mispricing

Yes

No

Yes

Yes

Trades intraday

Yes

No

Yes

Yes

Taxable dividends

Yes

Yes

No

No

At a conference in 2002 at New York University, academic researchers demonstrated that low-cost open-end index funds outperformed SPDRs every year since their inception—both pretax and after tax.8 I was surprised at the pretax performance, but keep in mind that open-end funds can reinvest their dividends and SPDRs cannot. The after-tax outperformance also surprised me; open-end index funds did pay some distributions. Theoretically, SPDRs are more tax efficient than the open-end funds, but in this study, the distributions were just not big enough to make a difference. Mutual Funds with Built-In Losses. Another strategy worth discussing was pointed out by Mark Hurley of JPMorgan Asset Management’s Undiscovered Managers fund family. If a mutual fund takes capital gains, then these gains must be distributed by the end of the tax year, but if a mutual fund makes capital losses, then these losses can be carried for five years and do not flow through to investors. So, some funds could have a sizable bank of capital losses, which can shelter future gains. I certainly would not suggest screening Morningstar for funds that have the largest amount of losses because it also could suggest that the fund’s manager is poor. But if an investor wanted to invest in an Internet fund, I would certainly suggest buying one that is seasoned with a bank of capital losses rather than a brand-new one without that benefit. If an investor is going to buy an S&P 500 fund that contractually has to be invested in the constituents of the S&P 500, why would the investor not choose one that had built-in losses? Obviously, these 8 Edwin J. Elton, Martin J. Gruber, et al. show this finding on a pretax basis in “Spiders: Where Are the Bugs?” and James M. Poterba and John B. Shoven draw after-tax conclusions in “Exchange Traded Funds: A New Investment Option for Taxable Investors.” Links to these papers can be found at www.twentyfirst.com/newsletter/newsletter_summer2002-4.htm.

©2005, CFA Institute • cfa pubs .org

accumulated losses swing from year to year, depending on whether the market is in a bull or bear phase. Purely in terms of tax efficiency, these funds and similar ones would be attractive to taxable investors because the gains would be tax free for quite some time. These opportunities can be explained by behavioral finance: Investors run into these funds at the high, and they run out at the low; the losses are realized, and they become a valuable bank of losses to offset future gains. Remember that a fund can carry forward the losses for only five years, just like any other corporation. So, take advantage of the window of opportunity while it remains open. The Wash-Sale Rules. Earlier in the conference, Jeffrey Horvitz stated that he believes that no meaningfully statistical method exists for harvesting tax losses without distorting investment returns. But I disagree. I cite research conducted by David Schizer, dean of the Columbia Law School, who explains that it is possible to construct “perfect end runs.” A perfect end run means that the economic position of the investment has changed but the investment return has not.9 No doubt, getting around the wash-sale rules is tricky. For example, if I own a stock, sell it for a loss, and then buy a call option, that series of transactions will trigger the wash-sale rules, which makes sense if the option is a replacement security. But note that if I bought a stock at $30 and it is down at $20 and I buy any call option, it triggers the wash-sale rules. If I bought a deep out-of-the-money call with a strike price of, say, $100, which certainly is not the same thing as stock ownership, it would trigger the washsale rules—just as if I had bought a deep in-themoney call with a strike price of, say, $1, which would be similar to owning the stock. 9 David M. Schizer, “Scrubbing the Wash Sale Rules,” Taxes — The Tax Magazine, vol. 82, no. 3 (March 2004).

CFA Institute Conference Proceedings • 35

Wealth Management

In contrast, if I have a short sale that I lost money on, I can cover my short and buy a put. Symmetry would indicate that this transaction should not be allowed under the wash-sale rules because selling a long and buying a call is no different from covering a short and buying a put. Yet again, this example demonstrates the anomalies within the IRS tax code. So, until the law is changed, there are tax-efficient strategies that can be exploited. Other ways do exist to work with the wash-sale rules. An obvious one is not to trigger it in the first place: Sell the stock and be out of it for 31 days. But if during those 31 days the share price skyrockets, the investor will probably regret being penny wise and pound foolish, and unlucky at speculative trading. Some tax practitioners suggest doubling up. If an investor owned a thousand shares trading at a deep loss, the investor could buy another thousand shares of this loser stock and hold two thousand shares for 31 days. At the end of 31 days, the investor could sell a thousand of the older shares, not the ones purchased 31 days ago. This way, the investor can realize his or her loss within the wash-sale rules by owning twice as much of it. I believe that this is a scary position because of the overexposure for those 31 days, and the investment performance will certainly be distorted. Nothing exists in the tax law, however, that says investors cannot totally remove the risk of owning that second lot of stock. If an investor buys a stock and then sells a call and buys a put that has the same strike price as the call, the investor can eliminate his or her performance risk. To illustrate, if you sell a call with a strike price of $30 and buy a put with a strike price of $30, then stocks trading above $30 at expiration will be taken away from you as the call writer. Alternatively, if the stock trades below $30 at expiration, you will put it to the put writer. Buying a stock, selling a call, and buying a put with the same strike price is called a forward conversion. It is a riskfree transaction that, in theory, should give a T-bill rate of return.10 The investor is still doubling up but without enduring the performance risk. And the investor earns an interest-like return on the forward conversion. At the end of the 31 days, one of the options will be exercised and the investor will deliver the highest basis shares against that option and thereby realize his or her loss. This strategy must be done before 28 November, however, if the investor wants the loss in 10

Put–call parity on a European option (a put or a call that can be exercised only at expiry) says that, in theory, a risk-free investment can be created synthetically by selling a call, buying a put, and owning the underlying security. The risk-free investment will, naturally, earn the risk-free rate of return.

36 • CFA Institute Conference Proceedings

the current tax year. Remember that the loss will not be realized until after 31 days and either one of the options has been exercised. Unfortunately, most investors harvest their losses in mid-December, which is too late for this strategy. Interestingly, if investors miss the 28 November deadline, they can still come close to creating a perfect end run by voluntarily triggering the wash-sale rules. I will explain: The investor owns a stock bought at $50, and it has gone down to $20. The investor then sells the stock at $20 and buys a call option with a strike price of $30 for $1, purposely triggering the wash-sale rules. Why? The cost basis on that option will go from $1 up to $31. Immediately after buying the call, the investor repurchases the stock at $20. Remember, the wash-sale rules can be triggered only once. The timing here is crucial: 1. Sell stock. 2. Buy call. 3. Buy stock. For some period of time, the investor will have more exposure to that stock because he or she also holds a call option. I recommend that investors hold that call option for a day or two. Obviously, this strategy is not a perfect end run because the investor is forced to own both the option and the stock for a few days. But if the investor delays recognizing losses until December, this approach is an alternative, although suboptimal, strategy. In addition, because a significant 20 percentage point difference exists between what long-term losses are worth and what short-term losses are worth, the investor’s holding period could be altered so that a long-term capital loss is transformed into a short-term capital loss, resulting in a larger bank of short-term losses that can offset short-term gains elsewhere in the portfolio. Here is how it works: 1. Sell the stock that has a long-term capital loss. 2. Buy a call on that stock. 3. Exercise the call. 4. Sell the stock acquired through the call exercise. This exercise of the call started a new holding period, and investors must buy the call within 31 days after they have taken the long-term loss.

Tax Inefficiencies of Hedge Funds Hedge funds are tax inefficient: They recognize a lot of short-term capital gains, triggering taxable events. I recently wrote about another, more insidious tax problem with hedge funds.11 When an investor receives his or her Schedule K-1 from a hedge fund, 11

Robert N. Gordon, “Is Your Hedge Fund a Trader or an Investor?” Journal of Wealth Management (Summer 2005):54–57.

©2005, CFA Institute • cfa pubs .org

Investing with a Tax-Efficient Eye

many times the accompanying cover letter assumes that the hedge fund is a trader. A trader is allowed to net its management fees against its gross profit; an investor is not. If the fees flow through (rather than being netted), they are treated as miscellaneous itemized deductions, which many high-net-worth investors cannot use. Suppose that an investor owns a fund of funds, and a fund of funds clearly is not a trader. Assume that the fund of funds earned 11 percent in its underlying hedge fund portfolios and that it charged a 1 percent management fee and a 10 percent performance fee, leaving a net return of 9 percent. What actually goes on the investor’s IRS Schedule K-1 (if completed properly) is that the investor is taxed on a gross rate of return of 11 percent, even though the investor earned a net return of only 9 percent. In other words, 11 percent is taxable income and 2 percent is miscellaneous itemized deductions that the investor probably cannot use, creating a tax on phantom income. This issue is more important today than in the past because when investors were making 20 percent and 30 percent returns on their hedge funds, they really did not care much about taxes. Now that returns have come down, investors are more aware of what they can do around the edges to improve their returns. In addition, the IRS hedge fund audit manual prioritizes the challenging of trader status, and I believe that investors will have to amend their past tax filings to their disadvantage. From a tax perspective, investors should invest in hedge funds so that they are taxed on a net basis, not a gross basis. One solution is to use offshore life insurance, in which investors pay on their profits only when they take the money out on a net basis. I am not enamored with these vehicles because no one knows the future tax rate that will apply when the money is taken out. (When I started working, tax rates on investment income were 70 percent and had just come down from 80 percent. We are now in a lull with rates at 35 percent. I am not sure that paying tax at 35 percent is so bad versus deferring into the future, when rates could be much higher than they are today.) But if the investor is going to leave the money offshore until he or she dies and then leave the money to his or her children, then the investor’s estate will not pay any tax, provided that it is held as an insurance trust. But for shorter time periods, insurance does not make a lot of sense. Another solution is to invest in an offshore hedge fund entity; this vehicle is typically used by taxexempt investors and non-U.S. citizens. Of course, nothing is stopping a U.S. investor from entering into an offshore hedge fund entity. Because it is an off©2005, CFA Institute • cfa pubs .org

shore corporation, it can be bought and sold at its net asset value. Therefore, investors pay tax only on the profits that are made on a net basis. But there is a downside. Because the entity is deemed a passive foreign investment company (PFIC), investors are allowed to defer and compound their taxable gains in that entity.12 But when the money is repatriated, the IRS will tax the gains at the highest rate, which today is 35 percent, in addition to a 5 percent interest penalty per annum. Therefore, this tactic would be less effective for hedge funds that might realize income that would have been taxed at 15 percent. Although the U.S. government enforces the PFIC rules, many states do not have PFIC rules. Interestingly, if an investor remains invested in a fund for seven years, only the profit made in the seventh year will be taxed in those states; the other six years will escape state taxation forever. Shrewd investors could stay in the fund for just one month into the seventh year and pay tax only on that one month’s worth of profits; the other six years will escape state taxation. Our analysis at Twenty-First Securities shows that investors in most high-tax states would be better off in an offshore fund than in an onshore fund. If an investor has a domestic-based hedge fund with a lot of long-term gains or qualifying dividend income, an offshore PFIC investment is not recommended because the gains in the U.S. hedge fund would be taxed at 15 percent, compared with 35 percent plus the 5 percent interest charge on the PFIC. I am enthusiastic about mutual funds that engage in hedge fund strategies, of which there are roughly 61 long–short funds and 4 risk arbitrage funds. All these funds possess economic and tax advantages. Mutual funds typically do not engage in excessive leverage, and they are less inclined to suffer from the challenge of pricing illiquid securities. Both hedge funds and mutual funds have management fees, but mutual funds do not have any performance fees. Although these kinds of mutual funds are more expensive than the average fund, they are much cheaper than a hedge fund. From a tax perspective, investors in mutual funds pay tax only on net profits. Investors may not be excited about going to the next cocktail party and talking about their “mutual fund,” but that may be the biggest impediment to investors realizing this golden opportunity. Derivatives, such as equity swaps, are available on hedge funds. By itself, a derivative will not help investors get the long-term capital gains tax rate, but it will make sure that investors pay tax only on how much they make. The constructive ownership rule 12 See Philip S. Gross, “Tax Planning for Offshore Hedge Funds: The Potential Benefits of Investing in a PFIC,” Journal of Taxation of Investments (Winter 2004):187–195.

CFA Institute Conference Proceedings • 37

Wealth Management

says that if an investor owns an equity swap on a hedge fund and it is held for more than a year and a day, that long-term gain reverts to ordinary income. But at Twenty-First Securities, we do not think that rule applies if an investor uses an equity swap on a hedge fund index. The law contains language about a derivative on a flow-through vehicle, like a limited partnership or a limited liability company, but it did not change the taxation of derivatives on an index. So, hedge fund indices now exist, such as the S&P Hedge Fund Index and the S&P Risk Arbitrage Index. Therefore, by investing in those indices through a derivative, investors will not pay tax until they want to, and then only at long-term capital gains tax rates. A hypothetical comparison of the returns between a fund of funds and a hedge fund indexlinked note is shown in Table 2. The assumed fund of funds has a management fee of 1 percent and a performance fee of 10 percent; the indices have a built-in management fee but no performance fee. A fund-of-funds manager may be able to outperform the index, but he or she would need a 14.82 percent return to equal the after-tax performance of a 10.00 percent return in the index. This 482 bps of alpha may be possible but quite a challenge. The source of the after-tax improvement can be decomposed into its constituents: extra fees, tax deferral and conversion to long-term gains, compounding, and deductibility of the fees, as shown in Table 3. For example, if the fees of the fund of funds are tax deductible to the investor, then the fund-of-funds manager needs to generate only 328 bps of alpha over an index—still a sizable amount—before the investor is indifferent between the two strategies.

Constructive-Sale Rules The Taxpayer Relief Act of 1997 forced investors to recognize capital gains on a constructive sale of a stock but not a debt instrument. So, if an investor has a bond with a big profit, the investor can lock in that profit by selling short the bonds that he or she already owns, known as a short against the box. In a short against the box, investors can get 99 percent of the money out of a bond and not pay the capital gains tax until they sell the long bonds. But if an investor owns appreciated stocks, then he or she has to navigate around the constructive-sale rules. Any hedging strategy must contain a risk for further appreciation or losses; otherwise, the constructive sale could trigger a taxable event. One approach to protect gains is to create an optionsbased collar by simultaneously buying a put and selling a call; we believe investors need about a 15 percent band between the put and the call.

38 • CFA Institute Conference Proceedings

Table 2.

Fund-of-Funds Alpha Required for After-Tax Breakeven

Index Annual Return Tax Rate on Fund-of-Funds Income

28.00%

30.00%

35.00%

5.00%

1.06%

1.26%

1.83%

7.50%

2.07

2.39

3.30

10.00%

3.13

3.57

4.82

12.50%

4.24

4.81

6.39

15.00%

5.39

6.08

8.01

Term of Investment (years)

5

6

7

5.00%

1.81%

1.82%

1.83%

7.50%

3.23

3.27

3.30

10.00%

4.70

4.76

4.82

12.50%

6.20

6.30

6.39

15.00%

7.74

7.88

8.01

Fund-of-Funds Fees Deductible

No

Yes

Difference

5.00%

1.83%

0.80%

1.03%

7.50%

3.30

2.01

1.28

10.00%

4.82

3.28

1.54

12.50%

6.39

4.60

1.80

15.00%

8.01

5.96

2.06

One of the problems is that this strategy could be a straddle, meaning any carrying costs are not deductible now at 35 percent but capitalized into the stock, thus causing a higher cost basis in the shares. And only if the investor actually disposes of the stock will he or she ever get the benefit of the deduction for those costs—and even then, it would be only at a 15 percent benefit. So, if the investor holds the stock until he or she dies, he or she will never get the benefit of the costs. Beware: A collar will usually destroy the holding period for the dividends, and then the investor will pay a 35 percent tax on them. A nonrecourse loan is one way to avoid a straddle. Following is an example. If you want to purchase real estate worth $10 million, you could borrow $9 million with recourse only to the real estate itself and not against you as the borrower. And if the real estate appreciates, you can sell the real estate and pay off your loan, or roll over the loan. But if the value of the real estate goes down, you can walk away from the loan and keep the $9 million. In the eyes of the IRS, you have then, and only then, sold the real estate for $9 million. The benefit is that it does not create a straddle, and it does not affect the dividend holding period. Therefore, you would be taxed at only 15 percent on future dividends.

©2005, CFA Institute • cfa pubs .org

Investing with a Tax-Efficient Eye

Table 3. Return Analysis of a Hedge Fund Index-Linked Note and a Fund of Funds Assumptions

Amount invested

$10,000,000

Term of investment (years)

7

Hedge fund index-linked note Annual index rate of return

10.00%

Annual management fee

1.75%

Final adjustment factor

88.37%

Tax rate on sale

15.00%

Fund of funds Annual rate of return

10.00%

Management fee

1.00%

Incentive fee

10.00%

Tax rate on income

35.00%

Fees deductible (yes/no)

No

Analysis

Value after Seven Years Index-linked note net of fees and taxes Fund of funds net of fees and taxes

$16,138,468 13,563,105

Cumulative Return

Annualized Return

61.38%

7.08%

35.63

4.45

Fund-of-funds pretax return needed for after-tax breakeven

14.82

Less pretax index return

10.00

Fund-of-funds alpha required for after-tax breakeven

Now, imagine that you have owned an appreciated stock for 10 months and you want to eliminate it from your portfolio. You have a choice. First, you could sell it now and pay short-term capital gains tax at 35 percent. Second, you could wait for two months and a day, betting that the stock does not nose-dive in the meantime, and then have a long-term holding period. Or, you could borrow through a nonrecourse loan starting at 10 months, and your holding period will not reset. Therefore, you should borrow for 2 months and a day, taking you into the 12-month long-term tax rate, and if the stock is up, sell it for a long-term gain and pay off your loan. But if the stock is down, walk away from the loan; again, it is the day you walk away from a loan that you trigger a gain for tax purposes. Naturally, the higher the loan-tovalue ratio, the higher the interest rate applicable to nonrecourse loans.

©2005, CFA Institute • cfa pubs .org

4.82%

Conclusion Investors ought to care about reducing their tax bills, and with various strategies, they can avoid paying the short-term capital gains tax of 35 percent to the federal government. Fortunately, the tax code contains anomalies, so ways exist to replicate an economic position without jeopardizing after-tax returns. And tax savings can be found almost everywhere: individual bonds and stocks, passive indices, mutual funds, and hedge funds. Many of the taxefficient strategies include the use of derivatives, although the expenses and complications for some of them preclude their use for all but high-net-worth individuals. For those wealthy individuals, they have plenty of flexibility within the current tax code to make the payment of high taxes voluntary rather than statutory.

CFA Institute Conference Proceedings • 39

Wealth Management

Question and Answer Session Robert Gordon Question: What size transaction do you need to make these transactions cost-effective? In other words, how big does the loss need to be? Gordon: The loss has to be about 15 percent of the price of the stock for options and swap agreements obtained from a dealer, and you probably have to have $1 million of stock to get the dealer to transact with you. If the stock has listed options, then there is no minimum; a standard contract is for 100 shares. For positions of $1 million or 10,000 shares, whichever is less, you can use what are called flex options on the exchanges, which allow investors to pick their own expiration date and strike price just like with OTC options. Question: What is the risk that the IRS will move quickly to stamp out some of these efficiency anomalies? Gordon: This is always a risk with these structures, especially some of the more aggressive ones. But the United States has a tax code that makes prospective rather than retroactive changes. So, we should always have some time to figure out innovative ways to unwind unallowable positions into allowable ones. And as David Schizer explains, the wash-sale rules have many frailties so that wealthy investors can create those “perfect end runs.” Question: If I sell a stock to recognize a loss, can I re-buy it immediately in my IRA without violating the wash-sale rules? Gordon: No. The tax law does not allow you to avoid the wash-sale rules by purchasing that stock for your IRA. In fact, this would apply to any related party—whether it is a business, a partnership, or a trust owned by a member of your family.

Question: What is the best investment approach for a taxdeferred IRA? Gordon: Growth stocks are not optimal in an IRA because if you made money on a growth stock, it probably would have been a longterm gain taxed at 15 percent. If you lose money in a growth stock, then you have a capital loss that you can’t use. I use growth stock as the substitute for a volatile asset, and I don’t think a volatile asset belongs in an IRA. So, bonds are the most appropriate assets to hold in an IRA because they pay a dependable stream of income. Question: Is the quality of managers available through mutual fund hedge funds equal to the quality of managers available through direct investing? Gordon: Some mutual fund hedge funds employ the same managers who have hedge funds, so obviously, the quality will be the same because the managers are the same. Management quality issues are more relevant for indices because they will only give money to a hedge fund manager who allows total transparency in daily liquidity—not necessarily the manager with the best skill. The Center for International Securities and Derivatives Markets (CISDM) has investigated skill differences between risk arbitrage and convertible arbitrage managers, and most of their returns are similar because they are all limited by the same subset of investments.1 Only so many risk arbitrage deals are going on, and only so many convertible arbitrage opportunities are coming out of the new-issue calendar. 1

The CISDM website can be found at cisdm.som.umass.edu/.

40 • CFA Institute Conference Proceedings

The performance of long–short managers is all over the place, as is that of macro managers. Question: If you’re going to use a derivatives strategy to get exposure to hedge funds, how do you deal with such issues as survivorship bias in the indices? Gordon: Much of the academic work you are alluding to is by Malkiel, who investigated hedge fund indexes that went back about 20 years. He found that they excluded those managers who performed poorly and went out of business. So, overcoming survivorship bias there will be tough. The investable indices have been around for only two years, and everyone that was put into them is still there. So, they may not have the same distortions. If we believe that a fund of funds must generate 482 bps of alpha to overcome tax inefficiencies, we certainly need to distinguish good managers from poor ones. But it is difficult to find managers who can make up the tax difference. Furthermore, that 482 bps won’t be reached by those following convertible arbitrage or risk arbitrage strategies, where they’re all stuck doing the same thing. Question: With hedge funds originally being the province of private investors, why aren’t they more tax efficient in the way they manage their assets? Gordon: A lot of money is held in offshore funds, especially the stakes owned by the manager, which benefit from deferred compounding. So, why should they—and their advisors—bother with tax issues, especially because some of the onshore strategies are complex and expensive to enter into? They are not paid or incentivized to be tax friendly; it is only when investors start demanding it that tax efficiency will come to hedge funds. ©2005, CFA Institute • cfa pubs .org

Risk Management for Private Wealth Jarrod W. Wilcox, CFA President Wilcox Investment, Inc. Newton, Massachusetts

Although enhancing return is one way to add value for high-net-worth investors, often overlooked is the value added from risk management. What is needed is a paradigm shift— a focus on growing discretionary wealth. In addition, advisors can add value through risk management by making sure that taxable investors take appropriate risk and by using complementary funds to offer better diversification for concentrated risk positions.

lthough this presentation is titled “Risk Management for Private Wealth,” in essence, it is about how to have a better relationship with clients through risk management. In his poem “If,” Rudyard Kipling wrote:

A

If you can keep your head when all about you Are losing theirs and blaming it on you, If you can trust yourself when all men doubt you But make allowance for their doubting too . . .

These lines of poetry, to my mind, describe what makes a good investment advisor in the private wealth area, and I use this passage as a reminder of what I should be doing. Some of us have learned the hard way that picking the best stocks or calling the bottom of the dollar or the top of oil prices is not the most reliable way to keep a client. Risk management is also value added, and it is much more reliable than trying to pick the best stocks or the best time to shift asset classes. My philosophy is that the client knows a lot more than we advisors do about what he or she wants, but we should know more than he or she does about how to get it. My assertion is that client-appropriate risk management is a critical part, maybe the most important part, of our value-added activities. In this presentation, I will first address the importance of focusing on growing discretionary wealth. This topic is really a paradigm shift, moving from an academic view of good investing through mean–variance optimization to something that is still mean–variance optimization but is more practical to use with clients. After this discussion on growing discretionary wealth, I will cover how to use risk management to make sure that taxable investors take an appropriate level of risk. Last, I will address a problem that many wealthy families have—better ©2005, CFA Institute • cfa pubs .org

diversification of concentrated risk positions. The client may be an executive with a lot of stock options, or the client may be retired and have a concentrated risk in long-term bonds. By using complementary funds, advisors can offer better diversification and essentially more value through risk management.

Growing Discretionary Wealth This discussion on growing discretionary wealth has several important parts. The first part looks at constructing an expanded investor balance sheet that includes both explicit and implicit assets and liabilities. The second part is using that analysis to derive discretionary wealth and to grow discretionary wealth using mean–variance optimization and, along the way, coming up with the client’s risk aversion or risk tolerance. I will also discuss bigger problems where the advisor cannot take the existing level of assets as a given because the client may borrow; I will note the kinds of cases where mean–variance optimization is not appropriate. And last, I will briefly discuss performance measurement within this context. Asset/Liability Framework. The asset/liability framework can be thought of as a slightly expanded balance sheet. Figure 1 shows this framework for a wealthy family. The left side is a list of diversified assets that would be appropriate for maintaining and slowly growing wealth for a long period. The right side, in contrast, contains the family’s liabilities and equity. In this case, the family’s liability is mainly the capitalized spending pattern that maintains the family’s lifestyle. What is left is discretionary wealth, which can be thought of as being similar to equity on a traditional balance sheet. CFA Institute Conference Proceedings • 41

Wealth Management

Figure 1.

Asset/Liability Framework: Established Wealth

Portion (%) 100

Implied Liabilities

Cash and Short-Term Instruments Inflation-Protected Bonds

80 Tax-Exempt Bonds 60

Complementary Funds Core U.S. Stocks

40 Discretionary Wealth International Stocks 20

Leverage = 2.6

Private Capital, Partnerships

0 Assets

Liabilities and “Family‘s Equity“

Note: Implied Leverage = 100% Assets/38.5% Discretionary Wealth = 2.6.

I am describing a big paradigm change. I am asserting that the advisor should move beyond a focus on managing the assets on the left-hand side of the balance sheet and should instead be watching and managing the discretionary wealth on the right-hand side. The goal should be to keep discretionary wealth from going to zero and to grow it according to a particular objective that I will describe. By changing the frame of reference in this way, one can see that a little volatility on the left side means a lot more volatility when compared with that smaller discretionary wealth base. One can also see that the fraction of total wealth represented by discretionary wealth has a huge impact on how risky, or how risk averse, the client should be with respect to the assets on the left. One can also think in life-cycle terms. Over a life cycle, the fraction of the balance sheet represented by discretionary wealth will change for various reasons: The client is employed; the client is not employed; the client is retired; the client’s life is drawing to a close; the client has run out of money; and so on. These changes are the basis for customizing results rather than just using textbook examples that say an investor should have a gradually increasing fraction in stocks until the middle of life and then more bonds at the end of life. An appropriate mix of stocks and bonds really depends on what is happening to discretionary wealth. 42 • CFA Institute Conference Proceedings

Notice that Figure 1 identifies a degree of leverage. In this instance, leverage refers to the ratio of assets on the left-hand side of the balance sheet to the discretionary wealth, which appears on the right. People may not be used to thinking about that as leverage, but it is just like borrowing money from reserves. Interest is not paid on it, but it has the same effects in terms of increasing risk. The higher the degree of leverage, the larger the impact on discretionary wealth based on changes in the values of the assets. Figure 2 shows an example investor balance sheet as it appears for a young executive working at a technology company who is just beginning to enter into the high-net-worth area. The house is a significant asset. This individual also has some securities, such as stocks from other companies, some company stock, and vested company stock options. These are often very long-term options that may range from three to seven years. The figure also shows a large area defined as net employment capital, which, in this case, represents unvested stock options, unvested restricted stock, and the capitalized value of this individual’s ability to save from his or her employment income every year—all discounted every year by the probability that the individual will leave the employment of this company and lose these benefits. So, now I have expanded the picture to include implied assets. And the nature of those implied assets and their large size will determine most of what should go on in this investor’s financial future. ©2005, CFA Institute • cfa pubs .org

Risk Management for Private Wealth

Figure 2.

Asset/Liability Framework: New, High-Net-Worth Employee

Portion (%) 100

Mortgage

House

College 80

Other Stock Company Stock

Retirement

Company Stock Options 60 Net Employment Capital

40 Leverage = 2.8

Discretionary Wealth

20

0 Assets

Liabilities and “Family‘s Equity“

Note: Implied Leverage = 100% Assets/35.7% Discretionary Wealth = 2.8.

On the right-hand side, I have made the implied liabilities a little more explicit, a little more objective, by building them up from the bottom (i.e., in a conversation with the investor). The resulting conversation would likely touch on a fund to cover the mortgage on the house, money needed for college for the kids, and a capitalized or net present value retirement fund. What is left after identifying those needs is discretionary wealth. A Basis for Risk Aversion. The primary objective for an advisor should be to maintain discretionary wealth at a level above zero. The secondary objective should then be to grow discretionary wealth and to derive the best median, or 50th percentile, outcome. To illustrate, think about these two cases. The first case is a sure-thing investment that makes 25 percent a year. If it runs for 10 years, what is the median outcome? The answer is 25 percent a year. The second case is a 50/50 chance of either doubling the investment each year or cutting it in half. The mean return every year is 25 percent: 50 percent times 100 percent plus 50 percent times –50 percent—the same average as for the first case. But what is the median outcome at the end of 10 years? If the number of up years equals the number of down years, the investor ends up right where he or she started; the median return is zero percent. Thus,

©2005, CFA Institute • cfa pubs .org

too much risk can interfere with compound growth on a median basis. Adapting the Mean–Variance Framework. I cannot go through all the math here, so I will just have to assert that the best median compounding outcomes over many periods are the result of maximizing the expected log return each period. A log return, closely related to a geometric return, is determined by taking the natural logarithm of the ratio of terminal wealth to beginning wealth. An approximation is helpful. For diversified portfolios, which do not have big downside risks compared with upside risks and do not have fattailed returns, an excellent approximation is E – V/2, where E is the average single-period return and V is the statistical variance of the return over a single period, or the standard deviation squared. ■ Discovering appropriate risk aversion. Although the academic literature points out that maximizing median outcomes can be a risky approach, the academic literature pertains to using the approach for assets. Remember that the goal here is not to maximize median outcomes for assets. Rather, it is to maximize median outcomes for discretionary wealth. The ratio between the two determines how risky total assets should be.

CFA Institute Conference Proceedings • 43

Wealth Management

By mapping the total return of the assets problem onto the return on just the discretionary wealth, one can use the implied leverage that I spoke of earlier. Leverage, L, multiplies the expected return, but it also increases the risk. Consequently, expected return for discretionary wealth becomes L × E, and expected risk becomes L2 × V/2. The standard deviation goes up with L, so the variance goes up as L2. To optimize the existing assets (i.e., without borrowing or paying off debt), one can divide (L × E) – (L2 × V/2) through by L and get E – LV/2 as the objective to maximize. Putting E – LV/2 into a mean–variance optimizer will then maximize median outcomes for discretionary wealth. A conventional mean–variance optimizer has an objective of mean return less some risk aversion parameter times the variance. Now, that risk aversion parameter has been derived! Consequently, the client’s risk tolerance for a single period will reflect appropriate long-term goals, which are having the best 50th percentile expected outcome and never having a shortfall. If the advisor gets most of the emotion onto the balance sheet, these goals turn out to be perfectly acceptable. Implied leverage, which takes into account emotional needs (implying liabilities), has thus been used to measure the investor’s objectively appropriate risk aversion given these emotional needs. ■ What happens when leverage can change? My discussion until now has been based on not changing leverage. But by optimizing the approximation for expected log return without dividing through by leverage, one can expand the problem to consider taking on more debt. Figure 3 shows the results for changing leverage. It shows the results of a Monte Carlo simulation run at different levels of leverage to assess the impact on discretionary wealth. In this simple example with lognormal returns, the mean stock return in excess of cash is 6 percent and the standard deviation is 20 percent, which implies that the variance is 0.04. The different levels of leverage tested are 0.75, 1.5, and 3.0. For this figure, I ran the Monte Carlo simulation 100,000 times for 10 years each at all three leverage levels. The figure shows the distribution of outcomes. The steepest line is for high leverage (3.0), and the flattest line is for the lowest leverage (0.75). For example, with a leverage of 3.0, about 20 percent of the time the client loses 90 percent of his or her money in 10 years. Of course, about 20 percent of the time the client multiplies his or her money by five, but many people would not be happy with that risk. The median result in wealth is slightly above one, which implies a quite low annual return. In contrast, the flattest line, the low 0.75 leverage, does not show much chance of a loss, and it has a 44 • CFA Institute Conference Proceedings

Figure 3.

Monte Carlo Simulation of the Impact of Various Leverage Levels on Discretionary Wealth

Ending Wealth after 10 Years 10.0 Leverage 3.0 5.0 2.0

Leverage 1.5

Leverage 0.75

1.0 0.5 0.2 0.1

10

20

30

40

50

60

70

80

90

Percentile of 100,000 Trials

better median return than the high leverage of 3.0. But it is not as good as the middle case, a leverage of 1.5, which shows a median 10-year final wealth of about 1.6, which equates to about a 60 percent cumulative return. By graphing median returns for many different leverage levels, one would discover an upside-down U-shaped function for the final wealth. But it is possible to take a shortcut by way of calculus to optimize E – LV/2 to determine the most advantageous level of leverage, here setting L equal to E/V. For this simulation, that optimal level of leverage turned out to be 1.5 times. ■ When mean and variance are not enough. In some cases, however, mean–variance optimization is not enough because it is not a good approximation of expected log return. (In these cases, one needs to take account of the return distribution’s higher statistical moments.) An obvious case is when clients have investments with very skewed or fat-tailed return distributions, such as for some hedge funds or junk bonds. But the less obvious cases are when the asset returns are very volatile, when there are long periods between rebalancing (most people do not know that when there are long periods between rebalancing, variance becomes less and less adequate as a measure of risk), and when the leverage on discretionary wealth is high. High-leveraged discretionary wealth can come about in several ways. Of course, buying securities on margin is one. Another is through losses on the ©2005, CFA Institute • cfa pubs .org

Risk Management for Private Wealth

asset side of the investor’s balance sheet. That is, if the implied liabilities are fixed, discretionary wealth will experience a larger percentage decline than the decline in assets, which results in higher leverage. Here is a very different example: Suppose you are a retiree and that you have very limited means—just barely enough to support yourself. Even a tiny chance of a 15 percent loss is a catastrophe for you. You are thus the type of person who will not be happy with the outcome of just pure mean–variance optimization. Advisors can do some things if they run into one of these situations. For example, they can run extensive Monte Carlo simulations, which will work but is a lot of trouble, or (and the easiest solution) shorten the period between re-auditing the client’s portfolio, which may or may not result in a rebalancing. So, do a one-year plan for a client instead of a five-year plan. Or, if the client is investing in a lot of exotic instruments or has a lot of risk tied up in a company, the review might need to be done more often than yearly, maybe even quarterly. Performance Measurement. The final topic in this section is performance measurement. The best measure in this framework would be risk-adjusted performance that looks like the objective (i.e., return less the client’s own risk parameter times the variance divided by two). Although that measure is easy to compute mathematically, few people use it, so advisors might have a hard time selling that one to the client. But the formula remains useful in that it points out that high-implied-leverage investors gain a lot from good diversification; it shows advisors where to focus their efforts. Advisors may be measured by using Sharpe ratios, which are calculated as the portfolio return in excess of the risk-free return divided by the standard deviation of the portfolio return. The biggest problem with Sharpe ratios is that they do not differentiate between general volatility and harmful volatility. For a very needy client, downside risk is not well measured by variance. The worst of all performance measures, although beloved by consultants, is the information ratio, which measures skill by looking at excess return over the benchmark divided by tracking error. This measure penalizes managers and advisors for creating tracking error even in cases where the client’s risk is substantially reduced.

Incorporating Taxes into Risk Management Both taxable and tax-exempt assets, or tax-advantaged assets, can be put in the same mean–variance optimization. But it is important to ensure that they are all ©2005, CFA Institute • cfa pubs .org

put on a level playing field by converting them to after-tax mean returns and after-tax risks. For those advisors using optimization software, it is important that the software be flexible enough to make those adjustments. The advisor needs to know the effective tax rate, which is not always obvious. The highest tax rate is a statutory tax rate, but the effective rate can be a lot less than that. So, for example, consider amortizing tax rates over a long holding period. If a client holds a stock for 20 years, the stock increases by 10 percent annually, and the client pays 15 percent tax on the gain at the end, what is the effective tax rate? The annual tax rate that would produce the same aftertax balance after 10 years is only 7.5 percent.1 Consequently, long holding times imply lower effective capital gains tax rates than short holding times. This is true even when the statutory rate is the same, which it may not be if the holding period is less than a year. This tax impact is even more important if the client is wealthy enough and poorly legally advised enough that he or she is subject to the estate tax. When taxes are incorporated into mean–variance analysis, something odd happens. The effective tax rate not only affects expected earnings; it also has an impact on risk. In fact, taxes reduce the risk in mean– variance optimization more than they reduce the mean return. Whereas the expected single-period return is multiplied by 1 minus the tax rate to determine the after-tax expected return, so too is the aftertax standard deviation of return. But that means that the after-tax expected return variance is calculated by multiplying the variance by the square of 1 minus the tax rate.2 Consequently, taxes have a bigger effect on risk and on its negative impact on compounding than on expected return, which implies (because the government is sharing risk with investors) that highly taxed individuals can afford to take a little more risk than nontaxable institutions. This information is important for clients who are facing a high capital gains tax rate or who are facing an estate tax. And it is particularly important information for private wealth advisors whose background is in the institutional area and who formerly managed tax-exempt funds. A further inclination to risk taking arises because of something called tax-loss harvesting. Because the tax law does not make investors pay taxes until some triggering event (i.e., not when the gains are really earned), investors have a choice about when they will 1The effective tax rate can be calculated by setting [1 + r (1 – T*)]n =

( 1 + r )n × ( 1 – T) + T and then solving for T*, where r = rate of return, n = number of years, T = tax rate, and T* = effective tax rate. 2

The variance is equal to standard deviation squared, so the impact on variance is the square of 1 minus the tax rate.

CFA Institute Conference Proceedings • 45

Wealth Management

pay the taxes, and that choice can be used to their advantage. By taking losses preferentially, investors can essentially build up their unrealized capital gains. This choice ultimately results in more working capital and the ability to make money faster. This compounding can, of course, have a dramatic impact on discretionary wealth. So, the tax-timing option is important, and the odd thing about it is that advisors are used to reducing risk, but in this case, advisors may want to increase risk on individual securities for their clients. Like any option, the tax option is more valuable if it deals with more dispersion. So, advisors should want greater dispersion within the portfolio of securities and for tax lots within securities measured in terms of the ratio of the current market price to the basis cost. This can be achieved in two ways. The first approach is to buy smaller stocks that are volatile. The second approach is to use dollar cost averaging. Why? Because the investor can buy the same stock at different times and get different ratios. Although this is a complicated subject, it does not take any math to know that the wealthier the client, the more he or she should be tilted toward taking specific risk and ideally diversifying those risks away at the portfolio level. For tax-loss harvesting, clients do not need risk at the portfolio level; they only need it for individual tax lots.

Using Complementary Funds Complementary funds are designed to be complementary to some concentrated risk in the portfolio. For example, recall from Figure 2—the balance sheet for the young, new, high-net-worth employee—that most of this person’s wealth is tied up in one company. Even the house’s value is tied up with how well the company does in that community. So, that concentrated risk is well worth diversifying, and as advi-

46 • CFA Institute Conference Proceedings

sors, we can add value by going beyond just using the building blocks that we are normally given. Statistical analysis should be used to provide the most rigorous examination, but a good advisor can come up with several alternatives based on intuition. For example, stock holdings for a highnet-worth Microsoft Corporation employee could be in such industries as medical services (other than drugs and biotech), utilities, or precious metals, whose returns are especially lightly correlated with those of Microsoft. As another example for complementary funds, consider bond-heavy clients. For these clients, I would include within their residual stock holdings a higher proportion of very-small-cap stocks. I actually manage a couple of accounts this way, and I have found that the realized return correlation for these accounts versus bonds is more highly negative than the correlation of the S&P 500 Index versus bonds.

Conclusion Risk management is about improving client relationships. It is about getting clients to look at the big picture and about adding value. I believe in the importance of risk management as a way to add significant value for the client. Shifting the focus from the traditional portfolio view to one that emphasizes the growth of discretionary wealth helps advisors and clients quantify the big picture and understand important changes. It also makes the subjective more objective and reduces valid client objections to mean– variance output. Risk management, in addition, reinforces the advisor’s ability to add value in detail by enabling better understanding of tax impacts and highlighting diversification, both by security and by tax lot, thus providing a value-added opportunity at the security level.

©2005, CFA Institute • cfa pubs .org

Q&A: Wilcox

Question and Answer Session Jarrod W. Wilcox, CFA Question: When you said that the optimal leverage equals 1.5 times, how did the government risk-sharing aspect of taxes fit in? Wilcox: In that hypothetical example, the tax rate didn’t come into play. But if it had, say, a tax rate of 0.15, the correct best leverage would have been 1.5/(1 – 0.15), giving 1.76. Question: If a client has enough financial assets to pay for college and retirement, are any additional discretionary assets needed? Wilcox: For some people, five houses are absolutely essential. So, that’s not part of discretionary wealth. I take the view that I know more about the framework that will get clients what they want, but they’re going to have to tell me what they want. Question: On the left-hand side of the balance sheet under net employment capital, does that include only stock options and the more complex things, or does it include earnings power? Wilcox: It includes such items as unvested stock options and unvested restricted stock. Additionally, it includes savings power. I use this rather than total earnings because then I’d have to balance it out with consumption spending. I netted consumption spending and

©2005, CFA Institute • cfa pubs .org

just took how much the client is able to save and capitalized that. Question: How do you calculate the value of net employment capital (Figure 2)? Wilcox: Estimating the value of net employment capital requires going through, in some detail, the value of any options discounted not only by time but also by the probability that the client will still be employed. The probability of continued employment is, of course, difficult to estimate, but the client can typically provide a ballpark estimate. For the nonsecurity parts of employment value, I look at the savings rate. In other words, I don’t look at the total earnings and then have to worry about spending. I just look at net savings and capitalize that as the value. And again, that was discounted for the chance that the client might leave and move to some other job where he or she cannot save. Question: Please explain your statement that taxes reduce the risk in mean–variance optimization more than they reduce the mean return. Wilcox: Let’s say I had a 50 percent tax. My mean return would be reduced by 50 percent. That’s easy. The standard deviation of my return would also be reduced by

50 percent. The variance is standard deviation squared, so I’m going to get a bigger impact on reducing variance. Question: Does your risk management approach differ when a client starts spending from his or her portfolio, particularly in light of nonlinear markets? Wilcox: The big challenge, of course, is when the client’s spending starts creeping up and you begin worrying that maybe you won’t be able to grow median discretionary wealth. I’m not sure that there’s a foolproof way to deal with that other than to keep pointing out that this spending is bad. If necessary, you could do a Monte Carlo simulation and show that this spending pattern will cause the discretionary wealth to run out. In terms of nonlinear markets, when we put assumptions about mean and risk into mean– variance analysis, they’re just assumptions. We could be wrong. If the markets for the next 10 years look like they’re not going to make as much money as in the past 20 or 30, then we’re going to have to put in lower assumptions about expected return. Nonlinear markets could also mean that we are thinking about downside risk, and again, the most practical thing to do about that is to review or audit more frequently.

CFA Institute Conference Proceedings • 47

Succession Planning for Business Owners Sean Sebold, CFA President Sebold Capital Management Naperville, Illinois

The importance of succession planning arises from two major needs: (1) business continuity and (2) obtaining and transferring value. Succession planning can be accomplished with either a transactional approach or a wealth management approach. Although transactional planning typically begins with a timely need, such as the owner retiring or receiving an offer to sell his or her business, wealth management succession planning is a more holistic approach focused on the owner and intergenerational wealth transfer.

n this presentation, I will concentrate on a relatively small segment of the private wealth market, the business owner. Specifically, I will address succession planning. Although I will focus on the concerns that arise in this context, keep in mind that succession planning is only part of a holistic wealth management approach. That is, some of the points I will discuss are intertwined with other aspects of a client’s life and financial circumstances. Done correctly, it is nearly impossible to separate wealth management from succession planning. The two distinctive approaches I will discuss are transactional planning and wealth management planning. Neither of these techniques is inherently superior, but as advisors, it is incumbent upon us to clearly explain to clients the approach we are taking and the services we are providing.

I

Importance of Succession Planning The importance of succession planning arises from two major needs: (1) business continuity and (2) obtaining and transferring value. Despite the importance of business continuity in the context of wealth management, the frequency at which it is ignored in the marketplace is rather staggering. Owners of restaurants, consulting firms, distributorships, and others seem to have an inferiority complex as far as the value they have created in building their businesses over their lifetimes. All too often at Sebold Capital Management, we have seen opportunities disappear as a result of an owner closing a business for what we consider to be invalid reasons and not moving through the succession planning process. 48 • CFA Institute Conference Proceedings

Obtaining and transferring value, conversely, is what most advisors tend to think about in succession planning. Business owners have spent a tremendous amount of time creating value and a business. At some point, it is time for them to leave the business and receive their rewards. In addition, they may wish to transfer some of the value they have created to someone they believe to be deserving, either a key employee or family member. I would argue that the end game or exit strategy for any business should not be left to the end or to the middle. The exit strategy should be thought out well in advance, during the building stages of the business. In the end, an exit strategy in the form of a transaction should be an almost insignificant event because it has been planned for and prepared for long before its execution. It is in this planning and preparation that we, as advisors, can create true value. The succession planning process may begin for the business owner for several reasons. The most common reason, by far, is retirement. Therefore, I will work from a common definition of retirement. Certainly, a lot of different definitions of retirement exist. Some of the definitions I hear are “not working” and “perpetual golf.” Those definitions certainly describe what most people think about as retirement. At my firm, however, we define retirement a little bit differently. We see retirement merely as a change in priorities. Ultimately, retirement is a change from an economic priority to a life priority. This distinction is important to us for the simple fact that by our definition, most of our clients are already retired. Does this mean they no longer work? ©2005, CFA Institute • cfa pubs .org

Succession Planning for Business Owners

No, it does not. What it means is the motivation behind what our clients do on a daily basis is different. The business is no longer just a tool that can create economic sustenance for them to survive. Instead, it is a tool that can provide greater things to them. It can provide a greater lifestyle. It can provide a means to change an industry. It can even provide a means to change society. Some might say that this is a matter of semantics, but the distinction is important. As behavioral finance shows, the way that we look at an issue can affect the outcome. Just thinking about retirement from this alternative perspective can change how we achieve the end results. Using our definition, we see that business owners have a different set of goals, aside from money, that drive them through their business lives. If their economic goals have been completed, their life goals are the driving considerations for their wealth. Buyout offers are another common situation that triggers succession planning. These offers are usually unsolicited and are what I call “offers that cannot be refused.” From a transaction perspective, buyout offers are typically the best option for a few reasons. First, the owner of the business does not have to go out and look for someone to buy the business; the buyer is coming to the owner. Second, the buyer usually names a price. It might not be the right or appropriate value, but at least it is a starting point. Third, a buyout offer frequently includes a premium within the price to encourage the business owner to act quickly. From a transaction perspective, a buyout offer is a great scenario for business owners: They created the business; now it is time to get out. An offer comes in the door, it is more than what would have been expected, they take the money, and they are done. Although financially this is seemingly a great outcome, we see several personal challenges to this scenario. Imagine this is your business. For the past several years, you have been focused on building it. The vast majority of your time has been spent in deciding how you can create value in the business for the products and services that you offer to customers. Everything you do is focused on growing the business and creating value. One day someone walks in your door, wants to write you a check in acknowledgment for everything you have done in the past several years, and then asks you to step aside. In this situation, you have to ask yourself a few critical questions, the answers to which could affect the rest of your life. First, is it a fair value? Second, is it enough? Third, are you willing to give it up? And fourth, what do you do then? ©2005, CFA Institute • cfa pubs .org

Yet another trigger for succession planning is when the business landscape changes. A change in landscape comes about when developments are made in technologies, products, and services. As a result, the owner’s skill set may fall out of sync with what is required for success in the new environment. For example, my grandfather was a stockbroker on the NYSE in 1930. He had one set of skills, which involved telling people about stocks and making transactions happen. If we were to suddenly drop him into today’s environment where he would be a trusted advisor, he would fail miserably without significant retraining. The landscape has changed, and the required skill set is different. If a business owner is unwilling or unable to make this transition, succession is a viable exit option. An additional scenario that brings succession planning to the forefront involves the desire to create opportunities for others. This area encapsulates two different scenarios. The first involves creating opportunities to give greater management and ownership responsibilities to family members. Many companies are created solely for this reason, and at some point, the “letting go of the reins” must occur. The second involves transferring value to a key stakeholder, such as an employee. For example, often a business owner has one or more key employees who were instrumental in building the business. The owner may wish to give these employees the opportunity to “show their stuff,” create equity, and build value as they see fit. The final reason for turning one’s attention to the succession plan simply is that the owner is not having “fun.” Many business owners come into my office and tell me they are sick of the business and sick of the problems. They want out. This situation can pose some value and succession challenges. If the business owner is not having fun, will sales and profitability be up or down? Does the business have the best people on staff, or is the staff a little bit challenged? Will the future prospects of the business be great or small? Generally speaking, business owners who want out because they are not having fun will probably not get the maximum value from the businesses they are exiting. It is thus critical in creating value to sell “at the top” rather than in a depressed state.

Transactional Planning Transactional planning usually starts when one of the aforementioned events occurs. So, the first step is to consider what started the transaction. Was it something that the business owner initiated, or did some external force cause the owner to act? This is an important consideration because the time when an owner sells a business has a great deal to do with the ultimate value achieved in the sale. CFA Institute Conference Proceedings • 49

Wealth Management

The second step is to evaluate the transaction: Does this transaction create maximum value? What are the current tax ramifications? What do we have to do to create maximum value? The remaining steps can be broken down into (3) planning, (4) preparing the company, and (5) executing. Planning. The planning begins with determining a few key issues. The first issue is whom to involve in this discussion. Depending on the size of the transaction and the type of business, various people might be needed to help complete the transaction: investment bankers, commercial bankers, lawyers, accountants, management consultants, or information technology consultants. Another important planning consideration is the type of strategy used to transfer value. Five common strategies that can be used in the transfer of the business are an asset or stock sale, employee stock option plans (ESOPs), stock recapitalizations, partnerships and gifts, and seller financing. ■ Asset or stock sale. In an asset sale, the assets of the company—buildings, equipment, inventory, and so on—are simply sold to the buyer. In a stock sale, the buyer acquires all assets as well as all liabilities. The choice of approach can result in different accounting, legal, and tax treatments. Of the two options, we tend to lean toward stock sales. Stock sales often result in more favorable tax treatment because the seller may be able to treat the entire gain at the more favorable long-term capital gains rate, whereas a portion of the gains may be taxed at the less favorable income tax rates in asset sales. Perhaps more important, though, is that a stock sale enables the seller to eliminate many of the liability issues that can come about well after the business has been transferred to the new owner. Although the tax and liability issues of a transaction can be critical, from a wealth management perspective, personal issues can also rear their head long after the transaction has been completed. For example, one of our clients ran a business that was in a highliability industry. He died suddenly before a lot of the planning work could be completed, so the business was left to his wife. She decided that the sale of this business would be to a key employee who spent a lot of time building the business with her husband. When we started talking about valuations and strategies, she was much less interested in the value she would obtain from the business than in making the transfer as easy as possible and on the best economic terms for the key employee. The key employee wanted an asset sale, which would stop all past liabilities and start the business afresh. Fortunately, we were able to purchase sufficient insurance to cover all past claims 50 • CFA Institute Conference Proceedings

against this business. Of course, this transaction did not generate the best value for the seller, but we acquiesced and moved forward on what we considered to be a suboptimal transaction. Three years after the transaction was done, a lawsuit was filed against the company. Economically, the original owner’s wife was covered through the insurance. But the emotional impact on her was tremendous. She thought her husband had done a wonderful job running the business and never expected an issue like this to come up. So, because of the way this deal was structured, something came back a few years down the road that she would have preferred not to have known about. ■ ESOP. An ESOP is another tool to transfer the value of a business. An ESOP is essentially a trust established by the business that makes employees partial owners of the business through contributions of company stock. According to the ESOP Association, more than two-thirds of all ESOPs are created for transferring closely held businesses to key employees. They are great tools for a lot of different reasons. For instance, they create liquidity, marketability, tax advantages, and business continuity. They can also add value to a company through lower turnover and ultimately more highly valued employees who have a stake in the company, which aligns their interests with the business owner’s interests. ■ Stock recapitalization. A stock recapitalization is a fairly straightforward strategy. The owner creates two separate classes of stock, one voting and one nonvoting. The owner then keeps the voting stock while giving or selling the nonvoting stock to others. This approach is really more of an estateplanning tool, but I have included it in this discussion because it is another method by which an owner can transfer value and ultimately ownership of a closely held company. ■ Partnership/gift. As with stock recapitalizations, granting partnerships and gifts is often thought of as an estate-planning tool for reducing the value of an estate. The creation of LLCs (limited liability companies) is useful for transferring portions of a company outside the owner’s estate and ultimately for implementing a succession plan. ■ Seller financing. Seller financing of a business sale often comes up in cases where the owner wishes to transfer a business to a key employee or family member but the buyer is unable to procure appropriate financing for one reason or another. In these cases, the seller can certainly finance the sale and affect the succession plan in that way. ■ Summary. These five strategies by no means constitute an exhaustive list. The key takeaway is that a lot of different tools and methods can be used in ©2005, CFA Institute • cfa pubs .org

Succession Planning for Business Owners

achieving a succession plan; we are not limited to a few. As a result, we can and should focus our efforts clearly on the desired results for the client, not necessarily on which strategy should be used. Preparing the Company. Preparing the company involves giving some thought to the operations of the company and the people involved so as to uncover possible issues, such as key management personnel moving out. We are also concerned with aligning the operations and the people with the exit strategy. For example, if a large company is taking over a smaller company, will their systems be integrated going forward? Or, will the smaller company be maintained on a stand-alone basis? Executing. The last step is the execution, with value transference taking place and operations management changing. If the planning work has been done well, execution should be straightforward: All of a sudden, the business owner has value, whether it is cash, stock, or a note; the operations and management have changed; and the deal is done. Example. An example can illustrate this entire process. Jim started a medical billing company from scratch with four employees. Within about five years of rapid growth, the company had 150 people on staff and tens of millions in revenue. Things were looking very good for this company. One day, Jim was approached by a publicly traded company that wanted to buy his company. This is Step 1 of the five steps I outlined earlier. That external force, the buyout offer, was what started the transaction. The value of that offer got Jim’s attention because it was more than he had ever expected to get out of this company. So, in Step 2, he decided that the offer made sense. Next is Step 3: Jim, along with his advisor, began the planning and decided on a strategy. This was going to be a stock-for-stock strategy. In evaluating their decisions, Jim and the buying company confirmed that the chosen strategy appeared to be the most economical from a value perspective while serving them well from a tax perspective. Preparation of the company was done (Step 4). And succession was completed, documents were signed, and value was transferred (Step 5). At the end of the transaction, Jim had stock in the new company. The stock that Jim acquired in the publicly traded company had a lockup period of 12 months, meaning that Jim was unable to sell his shares during that oneyear period. By the end of the 12 months, the stock had declined by about 60 percent. Jim, however, believed in the new management and in the company, so he held the stock for another 12 months. But the market did not like the shares of the combined company. The stock dropped another 20 percent. ©2005, CFA Institute • cfa pubs .org

After looking at this scenario, my question becomes: Did this transactional approach achieve the greatest value in wealth for Jim? For the answer, let me quote one of my accounting professors at the Kellogg School of Management. He said: “The answer for every accounting question is always, ‘It depends.’ ” I will come back to Jim at the end of this presentation, but suffice it to say that the correct answer right now is, “It depends.”

Wealth Management Succession Planning In contrast to transactional planning, wealth management succession planning is completely focused on the owner and intergenerational wealth transfer. Ideally, it begins with the founding of the business rather than with some event that suddenly triggers the need. The transactional approach can be a good solution for the short term, but we believe that wealth management succession planning can ultimately provide greater value for the client. As I outline the steps involved, keep in mind that I am discussing an ideal process. The actual terms of each client engagement will vary based on the client, the size of the company, and what the client hopes to achieve. Despite these differences, we strive for a consistent philosophy with each client. We embrace the ideal but implement the practical. In the remainder of this section, I will talk about three steps in our wealth management process. This is not an exhaustive list of our process, but these three steps most clearly identify the role of the business in the wealth management process. The first step involves setting goals, both personal and business. Second, we examine the role of investments in the context of the business owner’s situation. And third, we plan for what will happen after the business owner exits—the evaluation of estate-planning issues. Goals. The wealth management process for the business owner begins with our gaining an understanding of the owner’s short-term as well as longterm goals. But note that our definition of short-term goals is those goals sought during the client’s lifetime; long-term goals are those occurring after the client’s lifetime. We start this process with all our clients by asking a lot of questions. Many of the questions may seem odd at first, but they get to the heart of what we do. We begin with personal questions. We start by asking our client to describe an ideal day: “What does it look like? What does it feel like? What are you doing? How are you being challenged? Where are you? Are you traveling? Do you have responsibility? Do you have risk?” The answers to these questions give us insight into who the client is and what the client is striving for personally and emotionally. CFA Institute Conference Proceedings • 51

Wealth Management

Next question: “If you were given $25 million today, how would it change your life? Would you do anything differently?” It is a critical question for us because it gets back to the question I posed earlier about what retirement is. If retirement is a change from economic priorities to life priorities, then once that economic priority has been accomplished (in this case by receiving $25 million), what will drive this client for the rest of his or her life? We see a lot of issues arise when people suddenly come into money. They spend their entire life working for an economic goal, but they are not sure what to do once they have achieved it. This question helps clients determine what they will do with their life when they achieve all their economic goals. Another line of questioning helps establish what thoughts a client has regarding long-term (post-lifetime) goals: “How would you like to be remembered when you are gone? What is your legacy? What do you want to leave? What do you want to leave to your family? What is important to you from a charity perspective? What is important to you from a societal perspective?” We also try to establish how much a particular business owner has thought through the steps needed to accomplish the goals currently at his or her forefront with a different question: “What would you like to accomplish in the next five years, and what are the obstacles?” This question helps us identify how action oriented this client is and what role we will need to play in the wealth management process. A final line of questions attempts to establish the client’s priorities: “What does the phrase ‘make a difference’ mean to you?” The answer helps us from the perspective of having the client prioritize what is important to him or her. We can begin to determine how the client thinks, how money is involved, and how family and the community are involved in this goal. After the personal goals questions come the business questions: “Why did you start the business? You are going to get paid less, work more, have more headaches, and have more work than you can ever imagine. Why are you doing this?” We want to find out if the decision is economic, or is it based on wanting to change the industry or the way people look at things? The answer gives us an idea as to where the client is going with this business, and it certainly helps in deciding a lot of other issues, especially an investment policy. This line of questioning continues with: “How do you want to get out of the business?” We try to get this question answered up front because the answer drives our strategy going forward. Although many clients do not like thinking about the answer to this question, we really press hard for an answer because it is so important to establish the exit plan up front. 52 • CFA Institute Conference Proceedings

We then ask: “What is the role of the business? Is it purely economic? Is it to provide a family business that you can use to transfer wealth to future generations? How do you go about achieving your stated business goals?” Once we have come this far, we look at the answers for both the personal and business goals and we work to reconcile them. We look for consistency and congruency wherever possible. If a client says she wants to build her business and will do whatever it takes to get there but on a personal level she wants to spend more time with her family, we have to try to reconcile those potentially conflicting goals in advance. What good is wealth if we lose 50 percent to a divorce proceeding? Role of Investments. A client can easily have 80–90 percent of his or her net worth tied up in the business. Think about that situation for a moment. How many investment managers would write an investment policy statement with 80–90 percent invested in one stock? The heart of this question revolves around whether the wealth associated with the private company should be included in the investment policy statement. Some advisors think it should, and some think it should not. Even among those who think it should be included, many do not include it because of the difficulty managing it and implementing it within the policy statement in a practical way. At Sebold Capital Management, our stance is that the business asset absolutely must be in the investment policy statement. The largest asset that a business owner holds is usually the business. It is a high-return, high-risk investment that the client is holding over a very long period of time. This should be the highest rate of return investment that the client has. As a result, the client’s other investments should be supportive of this “asset.” They should not be based solely on return but on how they perform relative to the business. How are the returns of the investments correlated with the business? If the business falters, will there be money available from other investments to take up the slack? After-Exit Estate-Planning Issues. Typically, succession planning is completed from a transactional perspective. Once that transaction is completed, then a wealth manager’s job begins. Estate taxes are onerous and complicated. Tax rates can be as high as 55 percent, and every single one of our clients has estate tax issues. If these issues are not carefully considered, we could do a fantastic job of managing assets and creating a succession plan only to promptly lose 50 percent to taxes. Thus, a tremendous amount of value can be added in this area. ©2005, CFA Institute • cfa pubs .org

Succession Planning for Business Owners

We believe that the succession planning perspective should be started at the absolute earliest time possible. Essentially, the earlier we get involved in the succession plan, the more easily we can drive the business toward that succession plan. In addition, the earlier the succession plan is developed, the more easily we can integrate the plan into the estate plan. In this manner, we will not only be identifying and working toward our exit strategy, but we will also lower our overall estate-planning costs. A quick example of this is a simple strategy of gift shares of a rapidly growing company. It is possible to gift $11,000 a year tax free, or up to $22,000 as a split gift from each spouse. If a gifting program (of shares of the company) starts 10 years earlier, then 10 years of additional value transfer can be accomplished without any cost, except for the small cost of the actual transfer. Not only does this strategy lower overall estate taxes by removing a highly appreciating asset from the estate; it can also be coordinated with the succession plan. This strategy, and other strategies like it, must be planned for. We do not know how the market will behave, but we know the goal we are trying to achieve. We do not know how the company will do, but we can make reasoned expectations. We ultimately get our client where he or she needs to be by adapting: We adapt to tax laws. We adapt to changes in the investment process. And we adapt to changes in investment returns. Therefore, we think that the succession planning starts the minute we begin working with a client. The sooner we start that strategy, the lower the cost and the better the implementation going forward, from a short-term perspective during life and from a longterm perspective after life.

©2005, CFA Institute • cfa pubs .org

Conclusion The biggest difference between the transactional approach and the wealth management approach is that transactional planning is shorter term and in the moment, whereas the wealth management process is continual. It is a process, not a product. Going back to my early example of Jim and the medical billing company, remember that I posed the question: Did this transactional approach achieve the greatest value in wealth for Jim? The reason I said the answer is “it depends” is because I did not tell the whole story about Jim. I said he was in the medical billing business, because that is what he created. Actually, he is an entrepreneur who likes to create companies. He built the company, and five years later, he sold the company. With regard to his losing money on the transaction from a value perspective, he viewed it the same way a fund manager might see a bad stock trade. Since then, he has created two other companies. One of them he has already sold, and he is looking for an offer on the other. In that context, Jim’s stock-for-stock transaction is perfectly consistent and congruent with what he is trying to achieve. For other clients, the situation is rarely so clear-cut. Succession planning in the wealth management process is difficult to pull apart. The wealth management process is so closely tied to the client’s hopes, dreams, goals, and desires that it is almost impossible to tackle the issues relating to the business without simultaneously understanding their effects on the client’s wealth circumstance. It is for this reason that we do not try to separate them. We try to integrate them as a lifelong process that continues for generations. We believe that in helping business owners identify their own path and destination, we can ultimately create more value for them, their families, and their communities through this process.

CFA Institute Conference Proceedings • 53

Wealth Management

Question and Answer Session Sean Sebold, CFA Question: How do you get a business owner to start thinking about succession planning before he or she is ready? Sebold: We ask our clients to meet with us on a quarterly basis, and we press the issue at every opportunity. It is extremely important. If we create $50 million of wealth but half of it goes to taxes, does that make sense? Would it similarly make sense to create $50 million of wealth that ultimately goes to an 18-year-old kid after a death? There are a lot of very powerful reasons why these issues must be brought up and talked about. The consequences for inaction can be dire. So, if the client is not ready,

we continue to press until we make progress. Question: Does the client sometimes say: “This approach is not for me”? Sebold: Yes, some clients react that way. We are in the position that if we can’t help a client with his or her wealth management planning, then we have to think about the relationship we have because that is the value proposition that we provide. Being an independent firm certainly helps our position. It helps us really talk to the client in terms of what he or she has hired us to accomplish. Ultimately, if a client is not willing to talk about these wealth management issues, we

54 • CFA Institute Conference Proceedings

may have to help him or her look for somebody else because we are not in a position to help. Question: How do you charge fees? Sebold: With regard to the fee structure, for anything outside the business we have a basis points approach in terms of our management fees. We also have a retainer fee structure in place that varies depending on the complexity of the client engagement and the work we are doing for the client. Usually, we charge a larger fee up front for the first year because of all the initial work that must be done. In subsequent years, the fee is lower unless circumstances change.

©2005, CFA Institute • cfa pubs .org

Using Reporting to Anchor a Successful Client Communication Strategy Joseph E. Maxwell Chairman and CEO Investment Scorecard, Inc. Nashville, Tennessee

Client reporting is an area in need of much improvement. Individual investors are demanding the kind of sophisticated performance reporting that until now was exclusively available to the institutional market. Client reporting can be improved by focusing on the four “R’s”: manage, improve, and demonstrate returns; control risk; respond to regulatory requirements; and build better client relationships. Additionally, firms should build a strategy around the elements that will be reported, the predictable methodology the client will receive every time, and how the firm’s “story” will be communicated. Reporting that is transparent, objective, and simple instills confidence and trust, which ultimately serves to help cement client relationships.

lient reporting has historically been fragmented, and the quality of the reporting has varied with the size of the client. Therefore, in this presentation, I will discuss systematic client communication and the challenges of consolidated reporting. Figure 1 is a view of what I call the “whole picture” of the pressures currently facing wealth management firms. Compared with the past, clients have greater demands and are less trustful of their advisors. Also, firms are seeing an impending transfer of wealth from the Baby Boomer generation to the Generation Xers. And to top it off, regulation has increased. The wealth management firm is left sitting in the middle, having to synthesize numerous reports and statements for clients. When it comes to consolidated reporting, who would not want the ability to aggregate all multicustodial accounts from the outside providers, integrate the information into life and goals-based planning for not only the patriarch but also the family, and pull it together for Web-based, paper-based, and real-time viewing? Some vendors are currently selling that. But guess what? It is a myth. The reality is that consolidated reporting is hard. A lot of the systems that are driving the books and records of today were built 20–30 years ago. They have extremely structured datasets.

C

©2005, CFA Institute • cfa pubs .org

Technology has finally gotten to the point where advisors can create a robust consolidated statement inside their own firm. So, my advice is that firms should work first on what they manage or what they can directly control. That is, they should first concentrate on doing an excellent job of consolidating the statement inside the firm and then wrap on top of it a cohesive manufacturing strategy, whether that is a separately managed account overlay discipline or an investment policy overlay discipline. In that way, firms can track how well they have done against their initial objectives. Part of the reason why this consolidated reporting problem exists today is because when portfolio management systems were built and performance standards and calculation methodologies were first introduced, they were built around supporting the product, not the customer. In the 1980s and 1990s, mutual fund wraps and the various investment management fee-based products were the first to use retail performance reporting. But if a client had six accounts within a firm, he or she could not get a single report with all of his or her strategies in a consolidated framework. The customer was never in the middle of that supply-side equation from a logistics, data, or technology framework.

CFA Institute Conference Proceedings • 55

Wealth Management

Figure 1.

The Whole Picture Market Pressures Greater Client Demands

Less Client Trust

Impending Wealth Transfer

Reporting Requirements

Reporting Requirements Client Access

Increased Regulation

Client

Separately Managed Account

Policy Statement

Benchmarks

Open Architecture

Held Away Assets

Market Commentary

Wealth Management Firm

Equity Attribution Portfolio Appraisal

Outside Custodial Data

Fixed-Income Ratings

Disclosures/Disclaimers Consolidated Report

Flexible Classification

Peer Universe Comparisons Target vs. Actual Reporting

Client

Although there used to be an arrogant view that the institutional market was the only one that deserved performance reporting, the world has changed. At some point, high-net-worth families became “worthy” of performance reporting. And now, it has been pushed down to the individual. Today, the buzz is around holistic wealth management. Clients now have greater demands for reporting, transparency, objectivity, and simplicity. And the U.S. SEC is interested in individual client reporting and performance because of new and pending regulations. At the end of the day, the best practices for reporting, whether a scorecard is used or something else, involve an independent, third-party, unbiased, consolidated view of what a client owns. Fortunately, standards do exist for this kind of reporting, and they are simple. It is math, not wizardry. The provider has to get the data and then prepare, scrub, edit, validate, reconcile, and normalize the data. The provider then does a bunch of mathematical calculations, takes a bunch of market statistics, marries them to the math, and marries them to a policy statement. If humans are smart enough to go to the moon, we can certainly do this.

Example The following example shows why consolidated reporting can be so important. Recently, a top-five private wealth management firm came to us. Despite 56 • CFA Institute Conference Proceedings

Style/Manager Grouping

producing extremely sophisticated institutionalgrade analysis for its retail clients, the firm had a fragmented distribution platform. It had the most disciplined “manufacturing” process in place for asset management but a high degree of dissatisfaction in its reporting: The firm could not tell its own story. In the past, when the topic of reporting would come up at this firm, management would ignore it. This firm discovered that its clients hated the reporting, causing some to leave. And it was not the lack of analysis that drove clients away; it was the fragmented distribution platform, which came down to the field support. The operations group was getting the numbers out on the second or third day of the month, but nobody ever asked the people in the field what they were doing with the numbers. They were just patching them together. No systematic reporting existed. Clearly, the firm needed an integrated approach. It did not have an in-house core competency (the experience, skill, methodology, or knowledge to continuously upgrade and focus on best practices in client reporting), so it decided to outsource this role. The firm now delivers timeliness, accuracy, scale, and systematic reporting to its clients.

Improving Reporting Client reporting can be improved by focusing on the four “R’s”: manage, improve, and demonstrate returns; control risk; respond to regulatory requirements; and build better client relationships. But firms must first build an integrated platform in the back ©2005, CFA Institute • cfa pubs .org

Using Reporting to Anchor a Successful Client Communication Strategy

office, from a data perspective, that takes the transactional data and normalizes, edits, scrubs, and validates the data and recreates positional analysis from which the four R’s can be driven. If firms do not have an integrated system, they will spend a lot of money trying to patch together migrant systems. In addition, to improve reporting, firms should focus first on the intra-institutional reporting and not try to aggregate the world; the technology to systematically incorporate multicustodial data is not there yet on a low-cost and scalable outsourced basis. Reports should also have a consistent look and feel. Firms should furthermore choose simplicity over complexity and select scalability over flexibility. That is, firms should create a systematic approach for 90 percent of the accounts and do one-off processes on the others. Firms should also incorporate into the reports detailed comparisons with the industry and peers. And finally, firms should choose a “best-of-breed” reporting solution in order to differentiate their services from the competition. Fortunately, with the CFA Institute Global Investment Performance Standards (GIPS ®), a gold standard exists for presenting performance numbers.1 1 For

more information on the GIPS standards, go to www. cfainstitute.org/cfacentre/ips/GIPS-standards.html.

©2005, CFA Institute • cfa pubs .org

Strategy Firms need a strategy for their reporting, and it needs to be anchored to the firm’s core investment and relationship competencies. From there, firms need to ask themselves the following questions: • What elements do we need to report on? • What is the absolute predictable methodology a client will get every time? • How are we going to tell our story? Ultimately, standardization is conducive to gaining an investor’s confidence and trust. As I am sure everyone knows well, confidence and trust are paramount in this business. At Investment Scorecard, our clients are firms (not individual investors), and some of our most successful clients have embraced this view, with the return on their efforts being phenomenal. We have found that once a firm builds a framework to educate investors, what comes into that framework, in terms of future business and opportunity, is amazing. If a firm reports on an objective and unbiased basis and can tell its story clearly and cogently, investors will want to see their assets involved in that story. Investors are starting to demand a multicustodial strategy. If a firm is being asked to monitor assets that are outside its purview, the firm is doing a good job: The firm’s investors trust the firm.

CFA Institute Conference Proceedings • 57

Wealth Management

Question and Answer Session Joseph E. Maxwell Question: Have you thought about adding after-tax performance and after-tax benchmarks?

the story so that it is heard in a particular way, it will help cement the relationship.

Maxwell: Yes, but it is not the most screaming demand for a technology resource and a development resource in our shop right now. What’s coming in front of it? Our fee-based compliance modules tied with reporting, internal business metrics to try to aggregate the portfolios and understand what’s driving the profitability of their books, daily methodology as it relates to the custody engines, and the multicustodial framework, which is in high demand right now.

Question: What are your thoughts on restricting online access to client portfolios in order to keep the focus away from the day-to-day changes and on longterm strategy?

Question: How does reporting help build client relationships? Maxwell: A systematic approach to reporting instills confidence because you can point to the same story every quarter, every month, every six months, every year. If you’re willing to share the good, the bad, and the ugly in a consistent way, as opposed to changing

Maxwell: We don’t give people daily access. It is a monthly view. Only the institution has access to daily views. Question: Will there be a revolution in the investment management business once retail clients understand their performance? Maxwell: Probably not a revolution, but a change will absolutely occur. Think about some other industries, such as the mortgage business. It underwent a huge transition because it became commoditized. The investment management industry is a great business.

58 • CFA Institute Conference Proceedings

Where else can you have a 93 percent gross margin before you open the door, sustain that margin, and not share those fees down to the client level? That’s where the revolution will be and where the regulation is headed—into the management fees. Ten years from now, I think we’re going to have a lot of pure fees, fees unrelated to investment management, like you see on your mortgage statement (e.g., processing fees). We’re going to have to break out the fees, and we’re not going to be tied to basis points. Question: Where can advisors get a scorecard? Maxwell: Investment Scorecard is an outsource company. We do not sell directly to investors. Those who have questions can visit our website at InvestmentScorecard.com or e-mail me at jmaxwell@ investmentscorecard.com.

©2005, CFA Institute • cfa pubs .org

Tax Deferral and Tax-Loss Harvesting Jeffrey E. Horvitz Vice Chairman Moreland Management Company Beverly Farms, Massachusetts

By not realizing capital gains, equity investors can postpone or even fully avoid future tax payments to the government. The value of the tax deferral increases at an increasing rate as the holding period lengthens. For investors who actively trade, however, tax-loss harvesting can be used. Investors then maximize losses that can offset gains in other securities. For tax-loss harvesting to work, however, the present value of uncertain tax savings must more than offset higher transaction costs.

quity investors who are subject to tax can realize substantial tax savings, but they must be disciplined and patient in order to reap these savings.1 In this presentation, I will explain two basic but powerful strategies that investors and their advisors can follow. And although my focus is on equity investments, the principles apply to bond investments as well. I will also discuss the implications for the active versus passive management debate. I will not, however, discuss the myriad of tax shelters, such as insurance wrappers, offshore accounts, and retirement accounts. The advice here is general rather than particular to an investor’s unique tax circumstance, so the conclusions can be broadly applied. Keep in mind that my findings apply to U.S. investors, but the general conclusions apply to taxable investors in other jurisdictions. The key insight about the taxation of investments, particularly stock investments, is as follows: Unlike what most people think, the U.S. federal tax code operates nearly identically to that of a carried interest in a private equity or venture capital partnership. That is, the federal government has an ongoing share of the profits in a portfolio of securities, but the investor decides when the government will receive its share of taxes. This situation can be thought of as an option owned by the investor. So, the investor should exercise the option when it is most suitable for him or her, rather than for the government. I will compare this carried interest analogy with conventional wisdom later.

E

1

By “tax,” I mean the U.S. federal capital gains tax, but I urge investors and advisors to take into consideration state and, in some instances, local taxes.

24 • CFA Institute Conference Proceedings

Short-Term vs. Long-Term Capital Gains Taxes Why do active investment managers trade taxable accounts so often? Because they seek positive alpha for their clients. But the catch, from a tax perspective, is that frequent trades in mutual funds, private accounts, and broker-advised accounts can trigger recognized short-term gains—defined as those that occur within 12 months. This one-year horizon is of utmost importance because it means the difference between paying taxes at the same rate as ordinary income, which could be as high as 35 percent, versus paying the 15 percent long-term federal capital gains tax. With taxes, timing means everything; even one day can make a dramatic difference. So, how much more return do investors need in order to justify paying the higher short-term capital gains tax? If one assumes that the short-term capital gains tax rate is 35 percent and the long-term capital gains tax rate is 15 percent, then the ratio of after-tax returns is (1.0 – 0.15)/(1.0 – 0.35) = 0.85/0.65 = 1.31. So, all else being equal, active investment managers who make only short-term trades need to earn a return that is 31 percent higher than that of managers who make only long-term trades. This hurdle rate is staggering. I believe that few managers can break that hurdle in investment returns and certainly not with any consistency for their taxable clients. In addition, this hurdle rate makes no allowance for leakage caused by transaction costs, which would effectively increase the hurdle rate because managers who trade more often pay more in transaction costs. ©2005, CFA Institute • cfa pubs .org

Tax Deferral and Tax-Loss Harvesting

And I have even more bad news: Short-term capital losses have to be used first to offset short-term capital gains, and then they can be used only to offset 15 percent long-term capital gains, and only then on up to $3,000 of ordinary income. So, apart from dividends, the worst possible form of income for most people is short-term capital gains. Yet, many investment managers generate lots of it. The question then becomes whether investment managers are able to justify their trading activity. Table 1 is from an article by Robert Arnott, Andrew Berkin, and Jia Ye, who showed the percentage of mutual funds that beat the Vanguard S&P 500 Index Fund.2 Even taking into account the costs associated with the Vanguard fund, one can see that it is not a big number. In fact, shockingly few managers can beat this benchmark. Table 1.

Percentage of Mutual Funds That Beat the Vanguard S&P 500 Index Fund 10 Years 1989–1998

15 Years 1984–1998

20 Years 1979–1998

Before tax

14%

5%

22%

After tax including final liquidation

12

5

16

Pure tax savings over average fund

0.13

0.23

0.25

Source: Arnott, Berkin, and Ye (2000).

To some extent, the amount of alpha, what Arnott et al. label as the “pure tax savings,” looks puny here. And it is, but part of the reason for that small alpha is the turnover in the S&P 500, which itself triggers plenty of tax events. One could actually create passive portfolios that are much more tax efficient than the S&P 500. Another reason for the small “pure tax savings” is that the dividend component historically has been significant. Of course, dividends are paid periodically, and the size and timing of dividends are out of the control of investment managers. Aside from the anomaly of the 15-year period, which is probably a reflection of market conditions over that time, I believe that only about 15 percent of active managers can beat the Vanguard S&P 500 Index Fund on an after-tax basis, including all the tax at final liquidation. What is the likelihood of an investor picking one of those managers? What is the probability of picking three of those managers? It is low in both cases. 2

See Robert D. Arnott, Andrew L. Berkin, and Jia Ye, “How Well Have Taxable Investors Been Served in the 1980s and 1990s?” Journal of Portfolio Management (Summer 2000):84–93.

©2005, CFA Institute • cfa pubs .org

Algebra of Deferred Taxes The longer investors can go without realizing gains, the more valuable the gains are to investors because the compounding works for the benefit of investors and not the U.S. Treasury. Investors themselves determine when to liquidate an unrealized gain. In this context, taxation can be thought of in the same way as the carried interest of a limited partnership (such as private equity or venture capital). That is, the investor and the government will earn their share of the gains and taxes on the gains, respectively, at the end of the holding period, allowing the gains to grow tax free during the holding period. This concept is essential for understanding the simple mathematical mechanism of how tax deferral adds value. The following example will help explain how tax deferral works. Imagine a steady state rate of return, denoted by r. So, in each period, an investor’s portfolio grows by 1 + r. An investor’s terminal wealth is the product of 1 + r for each compounding period times the investor’s initial investment:

(1 + r1) × (1 + r2 ) × ... × (1 + rn ) . First, consider the case where an investor recognizes gains annually and must pay the corresponding capital gains tax each year, denoted by τ. The preceding equation can be modified to recognize the investor’s after-tax return:

⎡⎣1 + r1 (1 − τ )⎤⎦ × ⎡⎣1 + r2 (1 − τ )⎤⎦ × ... × ⎡⎣1 + rn (1 − τ )⎤⎦ . This equation, of course, assumes that the investor pays long-term capital gains tax at a constant rate over the entire holding period. Effectively, the investment is compounding at a lower rate. Now, consider the extreme case of 100 percent tax deferral. (I refer only to capital gains tax deferral because dividends cannot be deferred; they are taxed when they are paid.) At the end of the investor’s holding period is a final-period tax payment that can be thought of as a negative single-period return, denoted by rtax:

(1 + r1) × (1 + r2 ) × ... × (1 + rtax ) . It really does not matter where that negative return occurs from a purely mathematical perspective; but for this example, assume that it will always occur at the end of the holding period. A variant is a Roth IRA, where tax is paid only at the beginning. Remember that the government owns a profit interest in the returns, and the best that an investor can do is to defer the payment of taxes for as long as possible so that the compounding effect works to the investor’s full advantage over the holding period. Therefore, the compounding effect can work to its full potential over

CFA Institute Conference Proceedings • 25

Wealth Management

the holding period and reduces the taxes payable to the government, as shown by the following equation:

(1 + r1) × (1 + r2 ) × ... × (1 + rn ) × (1 − τ ). I will now put these simple equations to work. Assume that investments grow at an annual rate of 6 percent and that the applicable tax rate is the longterm capital gains tax rate of 15 percent. Figure 1, which shows the growing divergence between terminal wealth when taxes are paid annually and terminal wealth under a buy-and-hold scenario where taxes are paid at liquidation, demonstrates the power of tax deferrals, particularly if liquidation can be deferred until at least 10 years. Before 10 years, the difference between the annual payment of taxes and deferral is not material. As the length of the holding period increases, however, the power of nonlinear compounding and deferred taxes becomes stronger, as can be seen by the spacing between the dotted line that denotes the annual tax payment and the solid line that denotes the tax deferral. Notice that the difference between the deferral and annual tax lines increases as the holding period lengthens. The spread illustrates the dilutive effect of stalling the government’s share of investment gains for as long as possible. Horvitz and Wilcox demonstrated this effect in Figure 1.

a recent article.3 It is not something for which there are any exceptions, or at least important exceptions. This “carried interest” analogy will always work.

Short Deferral Periods Are Not Worthwhile Interestingly, many of those who have commented on tax deferral, those who sell tax-deferred accounts, and those who offer tax-deferred advice talk about tax deferrals over time periods shorter than 10 years. Figure 1 shows that material deferral benefits do not start until after 10 years. So, when advisors are talking to clients about tax deferral, and if the clients are not prepared to have serious deferral, meaning somewhere around the 10-year mark and beyond, then it really is not worthwhile. Table 2 examines the data from Figure 1 as a percentage of the tax-free terminal wealth. It uses the same return and tax assumptions and investigates, again, capital gains only, not dividends. Notice that at five years to liquidation, only a 40 bp difference exists between paying annual tax and deferring the tax. Think about how many products are being sold with 3

Jeffrey E. Horvitz and Jarrod W. Wilcox, “Tax Management of Stock Portfolios,” Journal of Investing (Spring 2005):83–89.

Terminal Wealth at Liquidation Amount ($) 5,500 5,000 4,500 Deferred Tax 4,000 3,500 Annual Tax 3,000 2,500 2,000 1,500 1,000 0

1

2

3

4

5

6

7

8

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Years to Liquidation

Notes: Price change only; dividends excluded. Annual rate of return of 6.0 percent, and capital gains tax of 15.0 percent.

26 • CFA Institute Conference Proceedings

©2005, CFA Institute • cfa pubs .org

Tax Deferral and Tax-Loss Harvesting

Table 2.

Terminal Wealth as a Percentage of Tax-Exempt Investing

Years to Liquidation

No Tax

Tax Deferred

Annual Tax

5

100.0%

96.2%

95.8%

0.4%

10

100.0

93.4

91.8

1.5

15

100.0

91.3

88.0

3.3

20

100.0

89.7

84.3

5.4

25

100.0

88.5

80.8

7.7

30

100.0

87.6

77.4

10.2

Spread

Note: Price change only; rate of return = 6 percent.

a three- to five-year tax deferral for which clients are paying a material fee to earn a negligible benefit. (I assume that the deferral is complete and that the withdrawal occurs all at once.) These short-term tax deferrals are simply not worth doing. The spread becomes material starting after 10 years. At the extreme in this illustration, the patient investor is rewarded with about 10.2 percent more wealth after 30 years. Incidentally, I have also investigated the spread assuming modest dividend payments and corresponding income tax each year, and the spreads are not overwhelmingly different. For 1.5 percent paid-in dividends each year, the corresponding taxes paid at the rate of ordinary income, a 9.2 percent spread still exists after 30 years. Tax deferral over long holding periods is worthwhile, but the benefit over shorter periods greater than one year is negligible. So, if somehow investors believe in active management, or they believe they have the hot stock, or they believe they have an edge, they should not give that up just to get tax deferral unless they can get the deferral for more than 10 years. And this result also has a lot of bearing for clients who have highly concentrated portfolios. If the expectation is that their highly concentrated position, for one reason or another, will be sold and capital gains recognized, meaning for tax purposes in five years or just a few years longer, then they should not spend a whole lot of effort trying to defer that tax. It is just not worth it in that case.

Mitigating Estate Taxes Estate taxes are important, but not a lot is written about them. Table 3 shows proposed estate tax changes under current law. President Bush in 2001 signed into law the Economic Growth and Tax Relief Reconciliation Act of 2001, which gradually reduces the maximum tax rate and increases the annual exclusion amount (exemption) up until 2009. ©2005, CFA Institute • cfa pubs .org

Interestingly, the act fully repeals any estate tax beginning on 1 January 2010. But absent any U.S Congressional intervention, this repeal will “sunset” or expire on 1 January 2011, returning to the old tax system, which includes the 55 percent top tax rate. Of course, no one knows for sure what will happen beyond 2010. Not a single trust and estate attorney believes that the old tax system will be put back in place in 2011. And it would be unreasonable, I believe, to present clients with projections based on a reversion to a 55 percent maximum tax rate. At the same time, I would not be surprised to see the repeal expire with a new tax rate in place for 2011 and beyond or to see the estate tax schedule changed by Congress before 2010. Table 3.

Estate Tax Changes, 2005–2011

Year

Maximum Rate

Exclusion Amount

2005

47%

$1,500,000

2006

46

2,000,000

2007

45

2,000,000

2008

45

2,000,000

2009

45

3,500,000

2010

(Estate tax repealed)

2011

55

1,000,000

My axiom—deferring gain recognition until death is always better—holds true even if the estate is going to be subject to estate tax. Table 4 has some simple examples that illustrate the power of estate taxes. This simple matrix can be prepared for any client. In this instance, I assume the cost basis is $1,000, the future market value is $1,500, the longterm capital gains rate is 15 percent, and the estate tax rate is 45 percent. If the estate is tax exempt and if the tax deferral lasts until death, then a taxable investor is converted into a tax-exempt investor with zero taxes payable. If the estate is tax exempt but the capital gain is recognized before death, then the investor pays $75 in capital gains tax [($1,500 – $1,000) × 0.15]. If the estate is not tax exempt but the capital gain is not recognized before death, then the investor pays $675 in estate tax [$1,500 × 0.45]. If the estate is not tax exempt and the capital gain is recognized before death, then the investor pays $75 in capital gains tax and $641.25 in estate tax [($1,500 – $75) × 0.45] for a total of $716.25. Table 4.

Total Tax Paid

Gain Recognized before Death?

Exempt Estate

Nonexempt Estate

Yes

$75.00

$716.25

No

0

675.00

CFA Institute Conference Proceedings • 27

Wealth Management

Of course, the ultimate value of deferring taxes until death will depend on the status of the estate. Following are four possibilities: The estate is below the exemption, the estate is left to charity, the estate is subject to tax, and the estate is left to a spouse. Estate Is Below the Exemption. When the estate is below the exemption, no tax is paid. There will be a step-up in basis, and the effect is to compound the deferrals just like for a tax-exempt investor. So, for older or ill clients who do not intend to spend all of their investment gains before they die, deferring the sale until after death is a way to make powerful changes in the residual estate, the amount of the estate that is left. Estate Is Left to Charity. For an estate left to charity, the effect is the same as for the estate being below its exemption. For those investors who have charitable bequests in mind, tax deferral is quite valuable. Just think of the effects of long-term compounding on the value of the estate passed to the charity. The investor has been converted to tax-exempt status without ever having to consider a private foundation. For some reason, this bequeathing to charity is not typically recognized in the investment community. Estate Is Subject to Tax. When the estate is subject to tax, there are large tax savings if an investor does not make inter vivos liquidations. The result is to reduce the effective long-term capital gains tax. The savings is equal to the long-term capital gains rate times the estate tax rate. So, if the estate tax rate is roughly 50 percent, then the investor cuts his or her effective long-term capital gains tax in half from 15 percent to roughly 7.5 percent. Estate Is Left to a Spouse. If an investor leaves an estate to a spouse and the spouse takes the basis, then the investor can perpetuate the tax deferral. Think of it in terms of estimating the life span of this as a last-to-die insurance policy. The actuarial expectancy could be quite long and may provide a powerful way to elongate the long-term tax deferral. So, for a married couple (one not in divorce proceedings) where one spouse is expected to die before the other, which is almost certainly the case, taking advantage of the extension of tax deferral can be very beneficial.

Tax-Loss Harvesting Some investors believe in active management’s ability to create positive alpha. For these investors who recognize capital gains through trading, capital gains taxes can be mitigated. 28 • CFA Institute Conference Proceedings

Tax-loss harvesting means taking voluntary losses for the sole purpose of creating a current tax deduction to offset other gains. Of course, investors have other reasons to sell securities, but the sole motivation for tax-loss harvesting is to trigger tax savings through realized losses now to offset gains elsewhere in the portfolio. For tax-loss harvesting to work, the present value of the tax deduction must meaningfully exceed the total transaction cost. This is the theory, but it is more difficult to quantify in practice. Furthermore, the amount of loss taken on each trade should be maximized. Other more complicated rules also apply—for example, in a highly volatile portfolio having multiple tax lots with multiple basis—but the basic principle holds. The investor needs to keep a record for gains as well as for losses. Without this disciplined accounting, the investor cannot be successful at tax-loss harvesting. Therefore, one of the essential tools that clients must have for tax-loss harvesting is adequate record keeping.4 Normally, the transaction costs are round-trip costs because, using my definition of tax-loss harvesting (i.e., for the sole purpose of creating realized losses), the investor is not really trying to dispose of the security. Rather, the investor is trying to keep the security. So, the investor somehow has to get it back. Getting the security back would be easy except for the Internal Revenue Service’s wash-sale rules. The wash-sale rules state that if an investor sells a security for a loss, the investor cannot buy the same security back within 30 days of the transaction or the loss will be disallowed for tax purposes. To further complicate matters, the investor also cannot have purchased the same security, or essentially the same security, for 30 days prior to the transaction. So, the investor has a 61-calendar-day window because the day of the trade itself is not counted. That is a long waiting period. The problem with the purchase of the substitute security is that it is purchased at a lower tax basis due to the loss, and this lower basis could be recaptured through a higher realized capital gain in the future. This point is very important to understand because a lot of the products that engage in tax-loss harvesting ignore this gain, unless, of course, the investor can defer realized gains until after death, as I have already discussed. 4

Berkin and Ye discuss the highest in, first out (HIFO) method, whereby the shares with the highest cost basis are sold first. This approach minimizes the capital gains tax. See Andrew L. Berkin and Jia Ye, “Tax Management, Loss Harvesting, and HIFO Accounting,” Financial Analysts Journal (July/August 2003):91–102.

©2005, CFA Institute • cfa pubs .org

Tax Deferral and Tax-Loss Harvesting

The present value of the tax deduction is difficult to calculate. But the deduction effectively puts extra money in the investor’s pocket, and so in this context only, tax-loss harvesting is like an interest-free loan from the government. (Do not confuse this with tax deferral, where the tax deferral is a future profits participation between the investor and the Treasury.) That is, investors can effectively spend the money derived from recognizing the tax loss and taking the tax deduction. When this recapture occurs and how much it will be are hard to tell. There is no algorithm that yields an exact solution, but it can be modeled based on probability-adjusted recaptures that extend into the future. Then, the investor can apply a term structure of discount rates to each future recapture. This process gives a crude estimate of the present value of the recaptures. Going further, the investor can calculate the net present value (NPV) as the difference between the positive present value of the recaptures and today’s transaction costs, which would include brokerage costs, commissions, and market impact, if the investor has a model for it. Transaction costs will be the most predictable in the NPV equation, but keep in mind that they can vary with investor size. That is, transaction costs are low for small traders and people who invest in mutual funds, they are really low for big institutions, and they are really high for everybody in the middle. Before taking the tax-harvest loss, there should be a significant positive NPV because of the high uncertainty of the recapture. So, the investor does not want to just hit the breakeven point by a dollar. The calculated NPV must be large because its true value is not known. If the investor is wrong and the NPV is, in fact, negative, then the investor will lose money and he or she will never know it. Thus, the operational problems of tax-loss harvesting are (1) how does the investor calculate the NPV of the tax deduction and the transaction costs when the recapture is uncertain and (2) how does the investor know when to trigger a loss recognition? That

©2005, CFA Institute • cfa pubs .org

is, how does the investor know when the share is trading low enough? With perfect insight and hindsight, the investor would pick the bottom of the market for that particular stock, sell it, rebuy it, and then write it up. Of course, investors would do that anyway if they could, almost irrespective of the tax-loss harvesting. So, obviously, if it were that easy to do, everybody would be doing it for other reasons. Some people have suggested creating a rule of thumb, such as taking losses that are perhaps 30 percent of face value. And some researchers have simulated possible solutions, but I am interested in an algorithmic solution.

Conclusions Investors subject to tax can realize substantial tax savings if they act and plan accordingly: • Avoid short-term capital gains. Investors should wait for long-term capital gains. This should probably be the first rule of taxable investment management. • Federal taxation is like a carried interest in profits. The investor and the government own the profits, and the share of the capital gains will be distributed to the investor and the government at the end of the holding period. • Tax deferral becomes more valuable as the holding period lengthens. The deferral is not valuable for at least 8–12 years, under normal ranges of market performance. (Note that earlier I said that the minimum period was 10 years, but of course, that was based on my assumption of a 6 percent compounded return.) • The investor’s estate tax situation is critical. Tax deferral is particularly valuable when no estate tax will be due. • Tax-loss harvesting has to solve two key problems. Tax-loss harvesting must address how to create a positive NPV out of the current tax savings when the potential future recapture is uncertain and how to maximize the loss for a given single transaction cost.

CFA Institute Conference Proceedings • 29

Wealth Management

Question and Answer Session Jeffrey E. Horvitz Question: How does volatility affect the tax-loss harvesting strategy? Horvitz: The more volatile the stocks, the more up and down you can get. So, you have more opportunity for tax-loss harvesting. Question: What is the best way to measure an investor’s after-tax performance? Horvitz: If you want to know how well you are doing, create the unrecognized gains for purposes of performance measurement as if liquidated and taxed at whatever period you are looking at. Question: What if tax rates are 70 percent when we withdraw? Shouldn’t we give value to the “devil” we know versus deferring to unknown future tax rules? Horvitz: Remember, it doesn’t matter what the tax rate is, just as long as you can defer this all the way to a tax-exempt estate. Therefore, the problem of tax rates is irrelevant. In terms of the mathematics, we know that the deferral spreads will increase as the tax rate increases; it pays even more to defer as tax rates increase. Tax rates right now are really about the lowest they have been since World War II. About 20 years ago, the long-term capital gains rates were either 20 percent or 28 percent, jumping more or less back and forth every few years. Today, the long-term capital gains tax rate is 15 percent. There are reasonable

prospects that one day we will see higher rates. Question: If an estate is left to the spouse, then half the amount is supposed to be updated with the date-of-death value. Then, there’s a yes or no answer for this deferral to work. Shouldn’t you gift all the assets to the second-to-die spouse while he or she is alive, if you can predict it? Horvitz: To the extent that you have exemption left in the estate, then it is really an estate-planning question, which is pretty straightforward: Take the step-up in basis, and in all events, use up all of your exemption first. The remainder, then, goes to the spouse with the old basis, which you could also do directly or in a qualified terminal interest property trust. To the extent that there are grandchildren involved, you want to use up your complete generation-skipping transfer tax exemption. The ideal estate-planning device is what’s called a dynasty trust, in which you actually put the money in trust for all future generations. In the right jurisdictions, you can effectively defer taxes forever. There are enough jurisdictions to do this in, but the major drawback is the legal cost. Question: Should tax-gain harvesting be used in the context of a tax-loss carryforward? Horvitz: No, because normally there is a significant amount of

30 • CFA Institute Conference Proceedings

gain that is involuntary. If you could control all your gains and losses, which you can’t, the game would be quite different. Tax-loss harvesting is the voluntary loss recognition for purposes of offsetting gains. Do those gains have to occur right away? No, not particularly. I think it is more valuable to use the voluntary losses in most all cases to shelter the involuntary future gains. Question: If you apply the taxdeferral advice to an investor who owns a concentrated position in IBM with a $2 cost basis, how do you deal with the trade-off of diversifying the investor’s portfolio versus holding the deferral taxes as long as possible? Horvitz: Irrespective of tax deferral, I would just sell the IBM stock. The amount of concentration risk is so extreme and the benefits of diversification so great that I can’t imagine rationally maintaining this position. In general, if you already own a well-diversified portfolio of stocks, then I don’t think there is a trade-off. If you invest in a diversified index fund, then you can mitigate the risk of having too much concentration in any particular stock. To be somewhat cynical, an active manager has only a 15 percent chance of beating the S&P 500, which is even more compelling evidence to invest passively; you would avoid the diversification issue entirely.

©2005, CFA Institute • cfa pubs .org