From the July 2003 issue of Investment Advisor • Subscribe!

July 1, 2003

Master Harold's New Portfolios

In light of the shrinking equity risk premium, Har

Being the nation's most quoted financial planner makes Harold Evensky an easy target for those who envy his visibility and influence. But there's a good reason why Harold gets so much more attention than the thousands of other financial planners out there. It's the same reason why a mutual fund executive was once quoted in Fortune saying "Harold is a god" and why a guy who is so nerdy that people wonder aloud if he sleeps in his bow tie is so damn popular. It's simply this: when it comes to financial planning, Harold Evensky is a great thinker. This guy lives and breathes planning. His wife, Deena Katz, is also one of the nation's leading planners, and she runs their Coral Gables, Florida, business, Evensky, Brown & Katz, which frees Harold to think.

So when Harold says that advisors need to reevaluate the way they design portfolios, you'd think advisors would listen. Well, think again. Despite the fact that we have been in a bear market for three years and that many of the thought leaders in financial planning and in academia are saying that we have entered a long period of low returns in the markets, I hear almost no discussion among advisors about whether they should change their method of managing client portfolios.

Even though Harold submitted a masterly paper last year to the Journal of Financial Planning about designing portfolios differently in a low-return era, the planning profession seems to be doing business just as it did before the bear market started and before a string of academics wrote papers saying the equity premium had shrunk. While we have had a 26% recovery in the last three months from the March 11 low in the Russell 3000, what Harold is saying is important to advisors and their clients. Here is what he is saying, and what he is doing for his clients.

About a year ago, you wrote an article arguing that advisors should design portfolios differently than they have been. But I haven't seen many advisors lining up to follow you. Why hasn't it happened? Does a change still need to be made? To the extent I can tell, there has been little or no changing of portfolio design in the profession, and I am not sure why. One obvious answer is that people disagree with me. But I have not heard much disagreement. I think it's a perception of, "If it ain't broke, don't fix it." I have to say, however, that right now I cannot understand why advisors don't think it is broken. Our job as planners is to look ahead, and I think it will be broken in the future. I'd say it is complacency, but I have more respect for my peers than that. I think it just hasn't hit their radar screens. We've all been overwhelmed in dealing with client feelings and concerns, and our first priorities are those issues. I think advisors are dealing with fundamental planning issues and have not had the luxury to step back and think about the broader issues, about [portfolio] design and implementation strategies. My best guess is that most advisors simply have not had time to reflect. What's frustrating is not so much that they're not doing it, but that they're not talking about it. We need to get this out on the table for discussion. It is extraordinarily important to look forward.

Could it be that the simplified portfolio design you are advocating poses challenges to the way an advisor runs his business? It is a more difficult story to sell if you want to get paid on the basis of assets under management. There's not much excitement to telling someone you'll put half their equities in a single ETF. When you think about the types of discussions we've traditionally had with clients, it's focused on this and that manager, on hiring one manager and firing another. At one point, we had three large-cap growth managers, one core equity manager, one large-value manager, a small-cap growth manager, two small-cap value managers. In international we had core, value, and small-cap value managers. Whether someone adds value or not, it's easy to create a perception that there is significant value because we're monitoring all these styles and strategies, as opposed to saying, "I'll take the bulk of your money and put it in a few investments." Also, it's a lot more exciting to say, "Pay me, and I will make you some money," as opposed to saying, "Pay me and I will save you expenses on investing."

The change in portfolio design that you've implemented is largely driven by your long-term outlook for stocks. Please explain that outlook. Over the next decade, we believe equity returns will be significantly lower than the last couple of decades and somewhat less than the long-term 70-plus year market history. No knowledgeable ob-servers disagree with that expectation. I'm actually on the optimistic end, with a 5% to 6% real return expectation. You've got Rob Arnott and others who say the equity risk premium is perhaps zero. Robert Shiller and Clifford Asness are not very different. I don't know anyone who has not accepted [the notion] that returns will be lower in coming years. I am familiar with people arguing about things I say. But no one has challenged my conclusion that equity returns will be in the range I am talking about--certainly not in professional circles. If you do not accept this assumption, I think you have the burden of proof to show why returns should not be much more modest. And, if you do accept that returns will be lower, then the challenge is to ask yourself what it means in terms of the way you implement client portfolios. It is difficult, if not impossible, to defend what I call the multiasset-class, multiasset-style strategy that we have traditionally used. The numbers are so simple and compelling to me that it is not debatable.

If you accept the view that stock returns going forward will be lower, then what are the implications for advisors and their clients? For the 15 years through June 2000--prior to the big drop in stocks--the S&P 500 returned 17.6% annually. That was when most of us developed our approach to portfolio design. If you're earning 17.6% and take off 1% for expenses for trading and the costs associated with a custodian, and if taxes are about 20% of the gross, which eats up about 3.5% of the return, and inflation is 3%, then your net after tax return is about 10.1%. That's your real net return after taxes and after expenses.

Now let's assume you earn a 9% gross return in the years ahead instead of the 17.6%. Take away 1% for expenses, and 1.8% of the return, or 20%, goes to taxes, and 3% is eaten by inflation. You end up with a 3.2% net return. The returns I am generating for clients are cut by over two-thirds. I end up with 3.2% annually instead of 10.1%. And if you have portfolios that are half in equities and half in bonds, the net after-tax and after-expense return will be even lower: you'll end up with 2.5% net. In this kind of environment, if I can save 50 basis points on expenses, I just increased the return by 20%. There's just not much left to play with. The problem is the expenses are going to be fixed and they are independent of the gross return, whatever that number is.

Does lowering your expectations mean completely rethinking the way you design portfolios and changing the way you make financial plans? Yes and no. Nothing has changed in our fundamental beliefs. We believe stocks will earn more than bonds and the world economy will be on an upward trend. We believe in diversification. We believe that there is a small-cap and value factor. What has changed is that we expect the real returns on stocks will be lower over the next decade. As a consequence, the strategy we used to implement, we now believe, will not be cost-effective in the future. We think the strategy needs to change and not the fundamentals of planning.

Explain the strategy change. The traditional institutional approach is still credible and valuable. But the problems arise from using that strategy in the retail world. So in looking at alternatives, what I have evolved our approach into is based on a "core-and- satellite strategy." That term has been around since the early '90s. It means you have some part of the portfolio in a core, where you are looking to capture market returns, and you are taking some risk around the perimeter, which is what is referred to as a satellite. We do the traditional work to determine the client's proper stock-versus-bond allocation using mean variance optimization. And then we split the equity portion by putting 80% into the core holdings and 20% into satellites.

What do you do in the core? The core is the allocation in equities that captures the stock market's return. In our case, the way we achieve equity market returns is with the Barclay Russell 3000 I Shares, which is pretty much the total market. Eugene Fama and Kenneth French [academics who have done extensive research on this issue] say you can get additional returns above the market from small-cap and value stocks. They call this the "three-factor model." Because we believe in their work, we also have as part of the core an allocation to large, mid-, and small caps. These are not opportunistic investments. These are long-term factors that are inherent in market returns.

Specifically, how do you capture the market factor in your core stock position? We use Dodge & Cox for large- and mid-cap value, and a DFA fund for small-cap value. And we may use DFA for large value again--we have in the past. We're using Dodge & Cox because when the government was hauling off CEOs and CFOs in handcuffs some months ago, we realized passive management is likely to own and, in fact, did own, significant allocations in many of those scandal-scarred companies. While [this approach] might be intellectually sound over extended periods, in the short term it was worrisome to our clients and to us to simply say, "Don't worry, it will all come out in the wash." So we made a decision to go with an active value manager at least for some period of time, and we think still there is some exposure and potential shoes to drop. We were looking for active management not so much to predict the future, but because when they hauled someone off in handcuffs we had reason to believe we'd be out of the position shortly. Our bias is toward passive management in the core positions, particularly in value, and we literally review that decision on a monthly basis to use an active manager.

Why do you have a bias toward passive management in the core? This is especially curious with value stocks, where I would think you'd want a stock picker. All the research we've done suggests that with core, in general, and value, in particular, active managers have not been successful in beating the indices. Value stocks are an expensive area to trade because they tend to have bigger spreads. And that's a tough hurdle for an active manager to overcome. Also, Fama refers to a value portfolio as a kennel of dogs. An active value manager is essentially saying he will throw out the dead dogs. But the dead dogs that come back alive may be a significant part of the value factor. So active management may, in fact, be counterproductive.

So you have those three positions--the Barclay's index funds, the DFA, and Dodge & Cox funds--representing 100% of your domestic core stock portfolios. What else are you doing? In large taxable portfolios, in lieu of the Barclays index fund, we use an active quantitative tax manager--Parametric Portfolio Associates. With Parametric, we select an index, such as the S&P 500 or Russell 3000, and Parametric's goal is to track that index with little error and manage for tax efficiency. Parametric can add about 1% of return, which in this environment we feel can be meaningful. As a separate account manager, Parametric imposes a $500,000 per account minimum.

The last part of our core is international stocks. We use an actively managed mutual fund there: Julius Baer Inter-national. Those four positions, which are dominated by passive investments, represent 80% of our equity allocation. It is very tax- and cost-efficient. We set our rebalancing parameters so that we will not be rebalancing very often--unless the market is very volatile, every three years. So we're minimizing hidden transaction costs and the tax costs of rebalancing. In the past, our portfolios had much more movement because we were constantly rebalancing and changing managers.

We have not talked about the bond portion of the typical portfolio. We also made some changes with the bond portfolio. We previously used two short-term, two short-intermediate-term, and two intermediate-term bond funds diversified among managers and maturities. Now, we've reduced the number of managers. We no longer have multiple short-term or multiple intermediate-term bond funds. We break the bonds into two or three managers now. So we cut to three funds from five or six. They're all active with one exception because we think the bond market is much less efficient than the stock market.

Okay, so that's your core. Now tell us about the satellite part of your portfolios. The satellite portion of the portfolio is where the real art comes into play. In the past, we had a certain risk budget that we applied to an equity portfolio. Beyond market returns, we were prepared to take extra risk by picking managers and slicing and dicing for extra returns. So in that core portfolio--where 80% of the assets are--we won't take risk beyond the market's risk. We saved our entire risk budget for 20% of the equity portion of the portfolio. So in the satellite part of a portfolio, we can take far more risk. The only criterion for a satellite position is that it is a fundamentally sound investment; beyond that anything is fair game. It doesn't have to balance the portfolio. It is likely to have a low correlation to the core but that is not required. It can be high-yield bonds, a concentrated position manager, hedge funds, or some other alternative investment. It may be expensive or tax-inefficient, but the key criterion is that we believe it is a fundamentally sound investment that, net those costs, will perform better than the core.

Give me some specific examples of satellite positions. We started off by splitting the satellites into at least three positions: emerging markets and commodities (both of which were part of our portfolios before), with the third being a new one: micro-cap growth. We felt each was volatile and risky, but each also had fairly significant upsides. We apply to the satellites what we call a tactical strategy, which is not market timing. Market timing is based on chart patterns and the belief that the market tells all. You're in or out of the market. Tactical asset allocation is based on a belief that there are certain long-term relationships between various market returns, and that when those relationships vary significantly, there will be a regression back to the historic norm. Back in the late 1990s, large-caps blew small-caps away. That made no sense to us. Small companies are riskier. We expected a reversion.

However, we were not prepared to use a tactical overlay in our former portfolio design. Rebalancing partly did that because we did buy more small caps, for instance, when they were less expensive. So rebalancing always had a certain element of tactical strategy built into it. In the satellites, we say we are now prepared to take more risk in this portion of the portfolio.

How have your satellites performed? The commodities did extraordinarily well through early 2003. We set up a bunch of triggers [indicating that] the investment committee needed to examine the situation. So we looked at commodities, which were largely natural resources-based, and looked at the fundamentals and saw that future prices of natural resources were lower. This was right at the beginning of the war, but it was at that stage where the threat of the massive destruction of oil fields had subsided. So we concluded there was a substantial oil premium that was likely to disappear. We thought the upside was low and the downside was significant going forward. We have a client who is a trader in the field, and he confirmed that we had it right. So we said, let's sell it. We never did that before. If we had a natural resources position, we were in it for good. In the satellite, we can get out, and we did. We used the Oppenheimer Real Assets Fund, which we liked a lot and could use again at some point.

How do you make these tactical decisions? We have an investment committee and we have a weekly and monthly monitoring system. We monitor everything, of course, between monthly investment committee meetings. For that month, for instance, in fixed income, we looked at high-yield bonds and saw the wide spread with Treasuries but felt that the big risk was the interest rate risk, which more than outweighed the attractiveness of high-yield bonds. We thought the best opportunity would take advantage of interest rates that would be up in the next year or two . First, we discussed shorting long Treasuries, but it was not practical. We found a couple of funds out there that do that, and we settled on the Rydex Juno fund. Morningstar rates the fund one star, but we didn't care. In effect, the fund shorts the Treasury bond and mirrors the opposite return of the long Treasury bond. Expenses are not as low as I'd like but there is not a lot of choice out there for doing this. A 1% drop in the long Treasury is about equivalent to a 16% return before expenses. The downside seems finite based on where rates are now, and the upside seems significant. Of the 20% satellite portion of our portfolios, Rydex Juno represents about 5%. Most of the satellite money, 10% of the 20%, is in the Deutsche Micro Cap Fund, an actively managed fund.

With the stock market being so much lower than it was three years ago, are you buying stocks and reallocating to the asset class that is now relatively more attractive? Theoretically, everyone should be increasing their stock exposure now. But portfolios are constrained by a client's risk tolerance. To tell a client who we've properly balanced for risk that they should own more stocks in a period when we are less sure about the efficacy of stocks versus bonds--while the client is nervous and the market is more volatile--does not work. Instead, we're telling clients to readjust expectations. Fact is, in most cases, it is not a significant issue. Before, we used a 2% real return assumption on bonds and 7% on stocks. So in a 40%-bonds/60%-stocks portfolio, that equated to a 5% real return expectation. Using our new numbers, it is a 4.8% real return because we're assuming a 3% real return on bonds and 6% on stocks. So it is not a huge change.

You Also like TIPS and ETFs these days. Why? In the past, our basic default had been to place bonds in personal accounts and stocks in tax-deferred accounts. The logic was that we can buy tax-free bonds but not tax-free stocks. Because of the extraordinary market volatility we've seen, the most aggressive investment has been sheltered and there has been big losses in them. So it has been frustrating not to be able to use those losses. If you held the stocks in a taxable account and can take the losses, that was also problematic: if you sold something, what did you do to replace that position in the portfolio for the 30 days in which the wash sale rule applied? Now, with advent of ETFs, it is inexpensive and easy to move to an ETF for 30 days, and you can maintain the character of a portfolio without triggering the wash sale rule by buying an ETF in a similar but not the same index. This is part of an overall strategy to be more proactive on ongoing tax management. In the past, we would do pretty much all of our tax-driven trading at the end of year. What we do now is to harvest losses continually. Because of the advent of ETFs, it is more practical. And in a low-return environment there is more bang for the buck in doing this. This has made us rethink where we would make the investment in a taxable versus tax-deferred account.

TIPS are inherently tax-inefficient. We wanted to set aside some or all of our tax-deferred account dollars to invest in TIPS. So we started doing some of most aggressive investing in taxable accounts and putting the TIPS in tax-deferred accounts.

How will you know if you are right? If the Dow Jones industrial average goes back up to 10,000 again in the next year, will you change your mind about your portfolio strategy shifts? How about at 11,000? At what point do you change strategy again? We're planners and we have to look ahead. To change from here, we have to have a significant change in expectations. If the market dropped 20% or 30% tomorrow, that could happen. The risk of being wrong is modest. On a typical balanced portfolio of 40% bonds and 60% stocks, we estimate the after-tax, after-inflation, after-expenses return is 2.6%, as opposed to roughly 7.2% for the 15-year period that ended in 2000. So if we are right and we can save 50 basis points through tax and expense efficiencies, I've increased my clients' net return by 20%. That's a big difference. If I am wrong and they really get 7.2% and I am losing 50 basis points because I am not tweaking these pieces of the portfolio, our clients should still get a 6.7% return instead of 7.2%, and I am not doing them much damage.

That's what finally triggered the decision to do this. I still participate in the equity market. I'm just not trying to eke out the extra bit by having many more positions and strategies in the portfolio. By the way, if we are right about this being a period of low returns, the advisors who are using multiple separate account managers will have some real explaining to do.

Reprints Discuss this story
This is where the comments go.