A Common-Sense Approach to Asset Allocation

Did diversification fail during the 2008-09 bear market? Is the traditional, buy-and-hold approach to asset allocation a vestige of a bygone era?

Investors began to ask these and other important questions in the wake of the bear market, when many asset classes they had relied on for diversification moved in a single direction: down.

To get an expert perspective on these topics, we recently sat down with Harold Evensky, president of Evensky & Katz Wealth Management. Morningstar managing director Don Phillips once called Evensky "the Dean of Financial Planning," and for good reason: Evensky marries the wisdom of a topnotch practitioner with a focus on cutting-edge research.

Part I of our conversation, focusing on asset allocation, appears today, and Part II, about planning for retirement, will appear on Morningstar.com next week. 

Following 2008, a lot of people began to feel very skeptical about the viability of long-term strategic asset allocation and suggested that perhaps a more tactical approach was in order. What's your thought on that topic?
Hope springs eternal. Everyone's always looking for the pot of gold at the end of the rainbow. The idea that it would be nice to be at the right place at the right time is not new. It didn't come with the 2008-09 crash. Every time there's a major market correction, we hear the same things.

If we're talking to a client and they say "Well, gee, so and so successfully timed the market, why can't you do it?" I'll say, "Name the top 10 artists of all time, the top 10 musicians. Or pick baseball players, anything you want." And we'll end up arguing about who should be on the list.

And then if I ask clients to name the top 10 market-timers of all time, everyone looks at me blankly. If it were possible, there would be a whole bunch of names we'd be arguing about. So my skepticism is pretty strong.

Asset allocation and modern portfolio theory didn't fail. We may have failed in picking the right things in which to diversify. Had I known to be in long-term government bonds, I'd have looked a lot better.

Were long-term government bonds typically a part of your client portfolios?
Nope. We do not invest in long-term bonds. However, we were traditionally diversified, meaning we invested in bonds and stocks, and certainly the bond allocations helped a lot. But in the stock or the equity allocations, we had large-cap, small-cap, value, and international. Everything went to a correlation of one in a short period of time.

The reality is a large universe of the "alternative investments" that were supposed to be absolute-return ended up being highly correlated with the markets. But that's really not a criticism of modern portfolio theory. It was the fact that we simply picked some of the wrong things and hadn't anticipated some of the consequences.

The objections I've heard to the concept of strategic asset allocation are that it's based solely on historical returns, which is nonsense. Anyone who ever read [Harry] Markowitz knows that he said that's not appropriate; you need to use forward-looking expectations. Some critics say that strategic asset allocation doesn't include valuation, which is also nonsense because any forward-looking expectations certainly integrate the concept of valuation.

The assertion that strategic asset allocation is something that you fix and forget forever is also nonsense. We revisit our strategic allocations at least on an annual basis.

So the question is what, if anything, did change or can we do differently? The answer is fairly limited amounts. We still believe diversification makes sense. To believe that we're going to accurately predict the right place to be in the next cycle is kind of ludicrous. No one's ever successfully done it. So, overall, I think investors need to intelligently diversify between bonds and stocks and between large, small, and international.

Now, having said that, many years ago we went to a core and satellite approach, and we carved out of the equity allocation 20% for what we call our tactical satellite, meaning we do use a tactical overlay on a small portion of the equity allocation.

So how do you distinguish between tactical and market-timing?
I interpret market-timing as the Ouija board, being able to look at chart patterns. Tactical is based on a belief of long-term relationships between various parts of the market. Over time, we expect small companies to provide greater returns than large companies. They're riskier, and you're being rewarded.

For example, during the tech boom when large companies were doing great and small companies were being killed, we believed that was not sustainable and there would be a regression back to a more historical norm.

So the tactical overlay is one way you've departed from a traditional buy-and-hold, strategic asset-allocation program.
Yes, and one other thing that is fairly new for us is that we've been considering some types of alternative investments. And by alternative I mean the hedge fund universe or the ability to go long and short. Intellectually, some of those strategies have always made sense to us. But the reason we've not been in that market is that what's been available in the retail market, has, for the most part, been expensive, opaque, and second-rate. So we've stayed out of it.

What's happened in the last couple of years is that some new products have been developed by well-established, intellectually very sound institutional managers that are reasonably priced for the retail market. So we've recently taken 3% from our fixed-income allocations and 3% from our equity to form a new category, which we call alternative.

So now our portfolios have the core, the satellite, and the alternative. We believe the alternatives will have a fairly modest return, somewhere between stocks and bonds, but with a correlation expectation of 0.3 or less with the general market. The inclusion of alternatives doesn't change our return expectations, but it will also provide some volatility damper should everything go wrong. Examples would be Cliff Asness; he's got a number of funds but AQR Diversified Arbitrage Fund (ADAIX) is the first one that we've used. And Andy Lo of MIT has a hedge fund-replication strategy,  Natixis ASG Global Alternatives (GAFAX). What ultimately sold us on these particular alternative investments were the quality of individuals and intellectual capital behind them as well as the pretty significant transparency of what they're doing.

But the core of our approach remains what we consider intelligent diversification. And the fact of the matter is if you have an asset-allocation policy and you have a rebalancing policy, there is an inherent tactical overlay to the portfolio.

For example, during the tech boom we sold a lot of large-cap growth and technology and bought bonds, small-cap value, and international. They kept getting clobbered, but when the market turned around, in hindsight we'd taken a lot of profits off the table and bought a lot of things that came back more quickly. More recently, when the market collapsed, we sold bonds, we bought equities, and in hindsight we bought a fair amount of equities at a fairly low point in the market.

So that provided a significant positive impact longer term, and not because we're brilliant, not because we predicted the market, it's just a very fundamentally sound investment policy.

Backing up to discuss strategic asset allocation, how would you recommend someone arrive at what is an appropriate long-term strategic asset-allocation plan?
It's a lot of science and a lot of art. The process we use is a classic Markowitz mean-variance optimization process. We develop our own forward-looking expectations, meaning what we think various asset classes and styles are going to earn. How confident are we of those guesstimates? It's basically the standard deviation. Then, how are they going to correlate with the other investments we're using? The biggest change that we've made in the last couple of years is the assumption that investments will be more volatile and that they'll be more highly correlated. Then we go through an optimization process.

If you are not doing this for yourself, then you need to put your efforts into finding an institution which has made that effort that you can leverage off of and that you've got some confidence in the firm's conclusions.

It's important to understand what's driving those conclusions. All too often people and even advisors are buying a package and haven't got a clue where those allocations came from or why they're like that. At a minimum, you need to understand the logic, theory, philosophy, and assumptions behind those conclusions, and make sure they're consistent with what you believe.

Do you incorporate the whole idea of human capital into your asset-allocation analysis?
I understand the logic of the human capital, and it makes sense. But I'm not at all sure it makes sense to integrate in an asset-allocation policy for a number of reasons.

Human capital is simply one element of someone's capital-needs analysis. In other words, what's coming in, then you have to factor in taxes, the goals, the expenses, and what's coming out of it.

There's also a significant behavioral aspect to deal with. One analogy is Social Security's role in a portfolio. Academics will argue, and I think appropriately, that Social Security is in an inflation-protected bond. If you're doing an asset-allocation model and you factor in Social Security as a bond, then you can have a much larger equity allocation in the financial assets. Intellectually, that makes all the sense in the world. The problem is when the market is down to tell someone, "Oh don't worry about it. Your bond is doing fine." And they say, "What? I don't own any bonds." If your response is "Oh, Social Security," it just doesn't resonate. It doesn't feel like they aren't doing so badly.

And the same kind of thing is true of human capital. It's not something that individuals can feel or relate to as an investment, per se. So, when your financial investments are performing in a manner that makes you uncomfortable, to say, "Don't worry about it your other investments, your human capital, is doing fine," that just doesn't resonate. It just won't work.

To me, the problem with a lot of those discussions is that it's good math and it's good theory. But I'm a big believer in the whole field of behavioral economics and behavioral finance. If a theory doesn't reflect how people feel, then ultimately, feelings, not theory, will win.

How do you come up with the forward-looking expectations that you arrive at for various asset classes?
I've concluded that it makes sense to make projections for broad categories and then do an overlay in terms of how much we're going to overweight a particular style.

So I'll look at domestic market, the S&P 500, the S&P 600, the EAFE, and some broad categories. I will look at returns for the last one, three, five, and 10 years and [examine] the direction. So that's one part of it, just to give me a general feeling.

The next question we ask is what is a safe rate of return likely to be? Depending on your time frame, a zero-coupon Treasury bond is a pretty good benchmark for that because that takes into account both the current and the reinvestment risk. Then we ask what premium would you expect looking forward? We usually use roughly a five-year economic cycle in terms of extra returns for taking the risk of going to the [equity] market.

Another way of looking at it is what are the various components of equity return? Well, stocks' prices in general are not likely to grow much faster than the economy. And this really is Jack Bogle's very simple but very logical approach. So what do you think the economy is going to grow at? You say, well, 6% is reasonable. What are the dividends, 3%? Now I'm up to 9%.

We'd also consider any leverage. If the market's price/earnings ratio is 20, and you think that's going to be moving back up to 30, then you can leverage your expected equity rate of return of 9% up to 12%. Our conclusion is there will be no leverage, so that brings us back to a 9% assumption.

So it's looking at all those different pieces. And the last thing I'll do is simply go out and look at what all the gurus are saying, the Goldmans, the Merrills, the J.P. Morgans, put all that together and reach a conclusion. That's why I say there's a great deal of art in the process.

The standard deviation, or the risk, is a little more quantifiable. There again we look at one-, three-, five- and 10-year data to see what the direction is. Clearly the direction has been toward higher risk, higher volatility.

I'll look at what happened during some of the stress points. Like during the tech crash and during the collapse in the last couple of years. I won't use the worst case, but I'll certainly factor in those stress periods.

I tend to be biased toward the high end. I use the assumption that if the markets have been moving in a more volatile direction, which they have been, then I have concluded that it's likely to stay that way. I'm not necessarily assuming things will get more volatile but that they'll stay relatively volatile.

And I'll take a similar approach to correlations. We've certainly seen a trend toward higher correlations [among asset classes] during stress points, so we'll expect to see less benefit from diversification than we'd have seen five years ago.

So can you share what your current expectations are for various asset classes?
Very generally, we're operating with a real equity return of about 6%, and a nominal return of 9%, with the assumption of inflation around 3%. For bonds we're using the same inflation assumption, so the nominal return we're using is about 5.5% and a real return of about 2.5%.

Off the top of my head, we do assume a slightly higher rate for small-cap and value--1 percentage point or so higher. We believe returns from developed international markets, basically EAFE, will be comparable to domestic stock returns. We don't distinguish between domestic and international.

How about for emerging markets? 
It would be a couple of percentage points over the domestic and international assumptions, with substantially greater volatility but less correlation. Although now the correlations are, like everything else, a lot higher than in the past.

What is the most effective way to add inflation protection to a portfolio? I know that you've long been a user of Treasury Inflation-Protected Securities.
Right. We think that inflation is likely to be the single biggest risk in the next decade to our clients' portfolios. So TIPS are the single easiest, biggest, most obvious inflation hedge. The problem is, they're not a big hedge, and they're also extraordinarily tax inefficient.

J.P. Morgan now has a municipal inflation-protected bond fund that we've been using, JPMorgan Tax Aware (TXRIX). So we now have a TIPS or a TIPS-type of allocation in all of our portfolios.

And that's a strategic allocation so it's not a function of "Gee, we think TIPS are a little pricey today and a little more attractive tomorrow." It is in effect a permanent allocation in our fixed-income category.

In terms of the equities, we believe that in reasonably high-inflation environments--single-digit--equities are a reasonable inflation hedge. If we get into hyperinflation, that's more problematic. That's not something we're anticipating or planning for, but if we get to that point then we would be looking more to commodities and something of that nature.

So what about commodities and the long-only commodities investments and some of the problems they've run into with contango and so forth? What's your thought on commodities as a slice of the inflation-fighting part of one's portfolio?
We've vacillated over the years, but still consider commodities a credible and appropriate investment, a good inflation hedge for high or very high inflation. The contango has been a particular problem for some of the index commodities; they have been more hurt than some of the actively managed commodities. So we've learned to be very careful about understanding the strategy of the underlying investment.

Is it designed for long-term commodity exposure? Or is it designed for commodity-trading exposure? There's a big difference, and it's not a good or a bad. There are just different kinds of products out there.

See More Articles by Christine Benz

Become a Morningstar Content Contributor Today! Contributing content to Morningstar allows you to increase your firm's exposure to a premium audience of financial advisors and individual investors. Please click here to learn more about the Morningstar Content Submission Platform. For additional guidance in using the Content Submission Platform, please view this video.
		New! 30-Minute Money Solutions
	Need help picking up the pieces in this turbulent market? 30-Minute Money Solutions by Morningstar director of personal finance Christine Benz simplifies the daunting task of getting your financial house in order. Written for novice and experienced investors alike, this book offers manageable, step-by-step solutions for tackling money challenges and building a comprehensive financial plan in simple 30-minute increments. Learn more.
	Order Your Copy Today--$16.95