- Shopping Bag ( 0 items )
Strategies continue to be explored, and tactics can change almost as quickly as the markets. What's the story behind Peter Bernstein’s challenge to a fixed-asset-allocation mix? Did the financial-planning community take a wrong ideological turn in espousing it? What can behavioral finance tell you about serving your clients? What choices can you make to ensure tax efficiency in your clients' portfolios? Downside risk measures have come a long way since Markowitz brought them so much attention. But when's the last time you checked into your reward-to-semivariability ratios? How current is your understanding of the core-and-satellite approach to portfolio design? And how much do you know about putting one in place for your client?
To get some answers to these and other questions, financial advisers Harold Evensky and Deena B. Katz invited some of the best minds in investment management to share their best thinking. The result is a gathering of eagles that will challenge your beliefs, reinforce your convictions, pique your curiosity, and maybe even improve some of those tried-and-true practices you put in place too long ago.
So sit in on this remarkable think tank. Treat yourself to a compelling array of ideas—from the doggedly practical to the delightfully abstract—that will inform and stimulate your own thinking and reawaken the reasons you came to investment management in the first place.
JEAN L. P. BRUNEL
Portfolio managers typically construct a diversified portfolio by following a sequential process in which they seek the best possible strategies in targeted market segments and assemble them in a way that makes for optimal risk/return trade-offs. This approach requires regular rebalancing, and although it may achieve some of the manager's key objectives, tax efficiency is not one of them. The reasons are twofold. First, regular portfolio rebalancing is costly. In an article that appeared in the Journal of Private Portfolio Management in 1998, I discussed the costs associated with rebalancing the portfolio to a long-term strategic mix. In fact, the costs associated with the regular rebalancing of the portfolio across these individual building blocks may indeed be higher than the potential loss of value added that arises from entrusting the management of the portfolio to a generalist rather than to a number of specialists. Second, each specialist will tend to operate in a relative vacuum, preventing them from taking maximum advantage of opportunities to raise the overall portfolio's tax-efficiency.
Tax efficiency requires investors to change the way they construct adiversified portfolio. This chapter investigates overall portfolio construction, explores why achieving tax efficiency naturally leads away from the traditional portfolio-construction model, and then proposes a different model-core and satellites-that can achieve greater tax efficiency.
Value Added and Portfolio Activity
Tax efficiency is tied to the relationship between portfolio activity and expected value added. In this context, we'll define "value added" as the return earned over and above the relevant market benchmark as a result of decisions that move the portfolio away from that index. To create a level playing field, assume that we keep manager insight constant: that is, the manager expects to generate that value added in a way that's proportional to the level of activity he or she undertakes. For example, every time the manager executes a transaction, he or she expects to add 0.05 percent to the return on the portfolio. It stands to reason that there would be a linear relationship between the manager's activity level and the portfolio's expected value added. FIGURE 1.1 represents that relationship as a straight line. Note that we're not talking about the observed, actual value added, which depends on the manager's success. We're depicting only the value added that the manager expects to generate. Note also that we don't need to limit ourselves to a return-based definition of value added. We can adjust it for risk, assuming that the manager also has a tracking-error expectation associated with the value added.
Tax Efficiency and Portfolio Activity
In their 1993 article in the Journal of Portfolio Management, Robert D. Arnott and Robert H. Jeffrey eloquently demonstrated that the relationship between tax efficiency and portfolio turnover is nonlinear. Rather than defining portfolio activity in terms of portfolio turnover, one might consider a slight variation on the theme. Indeed, one can argue that portfolio turnover is not a decision variable: in short, there is both good and bad turnover. Good turnover occurs when you execute an "alpha-enabling" transaction, that is, a loss-harvesting trade that enables the execution of another trade that was hitherto precluded by the likelihood of too large an unrealized gain in the position intended to be sold. Although an alpha-enabling transaction raises portfolio turnover, it tends to improve tax efficiency. Consequently, portfolio turnover is not the best measure of portfolio activity as it relates to tax efficiency. Rather, it seems best to think in terms of portfolio activity defined as net capital gain realization rate.
The observed tax efficiency of a portfolio is therefore a function of two distinct factors: first, the degree to which the manager's portfolio activity generates gross capital gains and, second, the degree to which the manager's tax-sensitivity will allow some of these gains to be sheltered by realized capital losses. FIGURE 1.2 illustrates the basic truism that tax efficiency falls rapidly as portfolio activity rises. A tax-sensitive investment process can defer some of the inevitable, but it cannot eliminate it. This relationship simply reflects the fact that certain strategies require a high level of portfolio activity and do not naturally lend themselves to tax efficiency. Consider an interest rate arbitrage between a Treasury bond and a corporate bond. Assume that you expect the difference in the yields between the two bonds to decrease. You'll typically buy a corporate bond (or a portfolio of such bonds to diversify away the bond-specific risk if the move you are trying to capture is generic to the corporate bond market rather than to an individual bond) and sell short an equivalent notional amount of same-maturity Treasury obligations. Though you may have some time frame in mind when the transaction is initiated, you will surely close the trade-repurchase the Treasury bond and sell the corporate bond-when you feel that the gap, or spread, between the two yields is back to normal, or even expensive. You may have to realize a short-term gain, but that would be preferable to waiting and not booking any gain, or worse, booking a loss.
The Murky Middle
The typical portfolio manager occupies the murky middle of portfolio activity-a place somewhere between the extremes represented, on the one hand, by the totally passive manager who aims to replicate the index and, on the other, by the very active trader. That the typical manager falls in that slot should not surprise for two simple reasons. First, the average investor likes to strike a balance between no activity and a lot of activity. Second, for the manager-particularly a large organization or one associated with an organization susceptible to reputational risk-targeting a middle-of-the-pack value added and tracking error is the smart risk-management strategy. Indeed, the further the manager deviates from the average, the greater the risk that his or her performance will deviate from that of peers and thus potentially disrupt client relationships.
Although such positioning makes a great deal of sense in a pretax, or tax-oblivious, business model, it's potentially quite damaging for someone who needs to worry about taxes. Figure 1.2 tells us that by the time we're in the murky middle, we have more than likely reached almost minimal tax efficiency and have stopped substantially short of maximum potential value added. We pay taxes on total return and not solely on value added. Therefore, assuming average market returns, it would not be unusual for a murky-middle manager to produce positive pretax value added and negative after-tax value added at the same time. Assume, for example, that the equity market has returned 10 percent and that our murky-middle manager has produced a pretax value added of 2 percent. Further assume that the portfolio's tax efficiency is around 70 percent. The after-tax return on that portfolio is 8.4 percent, since we will have paid 30 percent of 12 percent in taxes, or 3.6 percent. Assuming that the index has a tax efficiency of around 98 percent, which is probably unkind to the index, the after-tax return on the index would be 9.8 percent. Our murky-middle manager has done a great pretax job and cost us 1.4 percent relative to the index-even before we tack on the difference in management fees between the active murky-middle manager and the manager of an index fund.
An Alternative Design
Suppose we think instead in terms of "barbelling" the portfolio. Rather than placing all our eggs in the basket located in the middle of the portfolio-activity spectrum, we divide our portfolio into two strategic baskets, which, as we'll see later, don't need to be the same size. Indeed, the relative sizes of these two "extremes" of the portfolio will help determine the degree of overall average portfolio activity sought by the investor. The first subportfolio aims to produce the highest possible tax efficiency, and its design therefore includes the possibility that no value added will be generated at all. The second subportfolio aims to produce the highest reasonable value added and is therefore designed to be very tax inefficient. FIGURE 1.3 illustrates that alternative design, which resembles a barbell.
A numerical example helps illustrate the difference such a design can make. As in our earlier example, assume that an index fund generates 9.8 percent after-tax returns and that a satellite strategy will have a 60 percent tax-efficiency. (Sixty percent is a good approximation of average minimum tax efficiency for individuals living in states with a 5 percent or so state tax rate; indeed, adding that 5 percent to the maximum 35 percent federal tax rate yields an expected 40 percent tax rate on returns made up of interest and realized short-term capital gains.) It follows that any satellite able to generate 16.3 percent pretax returns, or a 6.3 percent value added, will match the after-tax performance of the index. This reinforces two important points. First, you need a lot of pretax value added to beat an index after tax, but we already knew that. Second, combining a tax-efficient core and a tax-inefficient satellite might produce a higher risk-adjusted after-tax return than a murky-middle alternative would. Although that satellite would have a higher risk than the index, it would offer potential tax-efficiency benefits over time because that volatility would produce opportunities to use realized capital losses to offset some of the gains.
The core-and-satellite design further illustrates the importance of thinking in terms of the total portfolio. Imagine that instead of using a standard passive index fund, we invest in an active tax-managed strategy, such as the one described in 1999 by David M. Stein and Premkumar Narasimhan in the Journal of Private Portfolio Management. Such a strategy potentially has a tax efficiency greater than 100 percent, since it generates net realized capital losses while matching the performance of the index. Even after accounting for dividend income, it's not unreasonable to aim for a 101 percent tax-efficiency ratio, given the approximate 2-3 percent net realized capital losses that Stein and Narasimhan-and later Arnott in his 2001 article in the Journal of Wealth Management-reported in their experiments, at least in the earlier years, on average. If we aim only to beat the 9.8 percent after-tax return of the index, we can afford to combine a "hyper tax-efficient" passive structured strategy with some less-efficient subportfolio, with a high expected value added. FIGURE 1.4 illustrates the different combinations of expected satellite value added and satellite portfolio exposure needed to match the after-tax performance of the index. Note that the hyper tax efficiency of the passive structured approach allows the satellite strategy to offer minimal alpha at very low levels of portfolio exposure to that strategy. Yet, at higher levels of satellite portfolio exposure, the required alpha quickly becomes very imposing.
Arguably, this design is even more compelling when considered in the context of a multi-asset class or strategy portfolio. There, examples of strategies one might find in the core portion of the portfolio would include passive portfolios, actively tax-managed portfolios, municipal bonds, or private equity funds. By contrast, the tax-inefficient satellite portion of the portfolio might comprise taxable bonds, concentrated portfolios, or hedge funds-directional or nondirectional.
Making the Core/Satellite Decision
How does an investor decide on the relative sizes of the two subportfolios? Two considerations can help-one is philosophical; the other relates to whether the investor prefers static or dynamic tax efficiency. An individual who doesn't believe that managers can add value through active portfolio management will tend to privilege tax efficiency at the expense of active strategies. For instance, that individual may decide that the only management activity reasonably likely to produce significant, risk-adjusted value added lies in hedge funds. Yet that person might also believe that the nonmarket risks associated with hedge funds-possibly their lack of liquidity-militates in favor of capping his or her total portfolio exposure to that sector to, say, 20 percent. Such an investor would end up with a portfolio 80 percent allocated to the tax-efficient core.
The question of philosophical focus on static or dynamic tax efficiency is somewhat subtler. Let's highlight the critical element of that debate, on which substantial additional research is needed to arrive at a meaningful conclusion. Static tax efficiency is focused on minimizing tax drag over the immediate term, accepting the potentially adverse consequences associated with the risk of "portfolio freezing," as discussed, for instance, in 1996 in the Journal of Portfolio Management by Roberto Apelfeld, Gordon Fowler, and James Gordon. Again, that risk involves either forfeiting the potential to keep earning after-tax value added as trades become increasingly difficult to justify-because the level of unrealized gain in each position makes it increasingly hard to justify any sale-or accepting higher portfolio tracking error. This situation arises because the portfolio, which by definition is different from the index, will keep drifting away from the index as its individual holdings generate different returns from those of the index. When the portfolio freezes, it becomes impossible to rebalance it closer to the index, and the risk inherent in that drift becomes increasingly significant over time. Though Robert Jeffrey might argue, as he did in the Journal of Wealth Management in 2001, that this is not an important issue, it must still be factored in, because it's a risk for which one is not compensated and can be very significant, particularly when considered in a multigenerational context. After all, the ultimate incarnation of this problem is known as low-basis portfolio concentration.
A dynamically tax-efficient process would look to balance tax efficiency, risk management, and the potential to earn value added over the long term. A dynamically tax-efficient investor would be concerned with the risk of too wide a gap developing between the portfolio's market value and its tax basis. He or she may thus be potentially more willing to incur some short-term tax drag, provided a fair after-tax and risk-adjusted compensation is expected to be earned over the long term.
Excerpted from The Investment Think Tank Copyright © 2004 by Harold Evensky and Deena B. Katz. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
Part One. The Portfolio.
1. The Tax-Efficient Portfolio (Jean L. P. Brunel).
2. Death to the Policy Portfolio (William W. Jahnke).
3. A Holistic Approach to Asset Allocation (William W. Jennings and William Reichenstein).
4. Professional Portfolio Design (Harold Evensky).
Part Two. Strategy.
5. Managing Concentrated Stock Positions (Tim Kochis).
6. Managing the Taxable Equity Portfolio (David M. Stein).
7. Tax-Efficient Investing (Thomas J. Boczar and Robert Gordon).
8. A Different Approach to Asset Location (Gobind Daryanani).
Part Three. Investments.
9. Reinventing the Investment Fund (Gary L. Gastineau and Craig J. Lazzara).
10. The Cost and Consequences of Insurance Wrappers (Ben G. Baldwin).
11. Alternative Investments (Mark Hurley).
Part Four. Practice and Theory.
12. Human Capital and Asset Allocation: Is That Client a Bond or a Stock?(Moshe A. Milevsky)
13. Downside Risk Measures: A Brief History (David N. Nawrocki).
14. Fundamental Fund Analysis (Don Phillips).
15. Controlling Longevity Risk in a Retirement Portfolio (Roger G. Ibbotson, Michael C. Henkel, and Peng Chen).
16. Monetary Policy and Investment Returns (John B. Brynjolfsson).
17. Defining Investment Advice (Stephen C. Winks).
Part Five. Clients.
18. Financial Gerontology and Employee Benefits (Neal E. Cutler).
19. Assessing Risk Tolerance: A Micro-Behavioral Finance Case Study (Geoff Davey).
20. The Why of Wealth Management (Ross Levin).
21. Lessons in Behavioral Finance (Meir Statman).
22. Missing Persons: Black Investors and the Stock Market (John W. Rogers Jr.).