By Christopher M. Wright

Name: Roger Gibson, CFA, CFP Title: President, Gibson Capital Management Born: 1951 Experience: Gibson is a nationally recognized expert in asset allocation and portfolio design. He has been named by Money and Worth magazines as one of the top financial advisors in America. He has been quoted in numerous publications including the Wall Street Journal, Forbes, Money, Fortune, The New York Times, and U.S. News and World Report. His own writing has won awards from Financial Planning and Investment Advisor magazines. He serves in various advisory capacities to Greycourt and Co., Inc., the Journal of Retirement Planning, the Foundation for Financial Planning, and SunGard Online Investment Systems. His firm, Gibson Capital Management, located in Pittsburgh, has been providing investment advisory services to institutions and high net worth individuals since 1989.

Chagrined investors rediscovered Roger Gibson's classic text “Asset Allocation: Balancing Financial Risk” after the tech bubble burst in 2000. Gibson's strategic approach to investment management emphasizes diversification across asset classes to reduce risk and enhance returns. Portfolio recently sat down with Gibson and asked him, among other things, where REITs fit into his clients' portfolios.

Portfolio: Your book indicates that you diversify broadly across several asset classes. What advantage do you gain by doing that?
Gibson: In the short run, there is always an asset class in first place that is delivering better performance than the other parts of the portfolio. The outperformance of this asset class, however, is not something that is predictable in advance. The rotation of leadership in the portfolio from one asset class to another acts to improve the longer-term, risk-adjusted performance of the portfolio as a whole. The dissimilarity in patterns of returns for the various asset classes is the underlying source of this diversification benefit.

We use four asset classes to diversify the equity side of client portfolios—U.S. stocks, non-U.S. stocks, real estate securities and commodities. We have reliable data on these asset classes beginning in 1972. Since then, real estate securities have delivered the best long-term returns and they have done so with the least amount of volatility.

Suppose you knew in 1972 that real estate securities would outperform. Armed with this perfect foresight, most investors would concentrate all of their equity money in real estate securities, right? Why would anyone choose to invest in asset classes that have both lower returns and more volatility? Yet, an equally allocated, annually rebalanced portfolio comprising all four asset classes had a higher return than a portfolio concentrated only in real estate securities, and it earned that higher return with considerably less volatility. As you move from single asset class portfolios to two-asset class portfolios, then three-asset classes, and finally to a four-asset class structure, volatility drops and long-term returns improve.

I have a chart that shows that the best portfolios are all multiple-asset class structures. Now suppose you knew ahead of time in 1972 that commodity-linked securities would have the highest risk of the four asset classes and you decided not to include them in your portfolio for that reason. By doing that, you would have cut out all the highest returning portfolios as well.

People really are surprised by this. Everybody's heard you should diversify to reduce risk but most people are surprised to learn that diversification, even with a more risky asset, also boosts returns. The whole moves faster than any of its parts. The greater the dissimilarity in patterns of returns, then the lower the correlations are and the bigger the diversification payoff. For example, commodities are negatively correlated to the other asset classes. They tend to behave in a countercyclical fashion.

Portfolio: What does it mean to rebalance a portfolio using strategic asset allocation?
Gibson: The portfolio structure strays from its initial design as a result of the divergent performance of the asset classes. With strategic asset allocation, you periodically rebalance the portfolio back to its target asset allocation. For example, a client may have a target allocation of 35 percent, plus or minus 5 percent, in U.S. stocks. If unusually strong U.S. stock performance causes the client's allocation to reach 40 percent of his portfolio, we trim the position back to the 35 percent target and redirect the proceeds to whatever part of the portfolio is below its target allocation.

The strategy is inherently “buy low, sell high” in implementation. We're a net buyer when an asset class underperforms and a net seller when a market outperforms. You keep coming back to the portfolio's target structure so you get the risk-return characteristics of that structure over time. Sometimes it's hard to convince clients that they need to be selling what's in the news and buying what other people are shunning.

Portfolio: Do you rebalance on the basis of how well markets are expected to perform?
Gibson: The portfolio structure we establish for a client is meant to be an “all-weather” strategy. As long as there are not any significant changes in the client's risk tolerance or time horizon, we maintain the same client-specific asset allocation year after year through widely varying market conditions. We buy and sell, not because we think we know what will happen next, but rather because the change in the value of a position causes the portfolio's asset allocation to stray from its target, triggering a need for rebalancing. We don't try to time markets and we don't abandon asset classes when they take their turn in the doghouse.

For example, real estate securities delivered back-to-back negative returns in 1998 and 1999. Those losses occurred against the backdrop of the last two years of the spectacular bull market for U.S. stocks. Some investors sold real estate securities after those two disappointing years and many used the proceeds to double up their bets on U.S. stocks—just in time to see the U.S. market lose half its value during the 2000 to 2002 bear market, while real estate securities better than doubled.

Lack of discipline can lead to whipsaw results and major portfolio damage that can be avoided. So we stick to our discipline and don't try to outguess the markets in the short run.

Portfolio: Your whole approach hinges on correlations—having asset classes that don't all move in the same direction at the same time. But correlations shift over time. Do you restructure portfolios to account for the fact that correlations have changed?
Gibson: Not usually. I don't know anyone who can accurately predict what correlations will be among asset classes over the next five years. Correlations may be fluid, but there generally is enough dissimilarity among the patterns of returns of different asset classes to generate significant diversification benefits. We could restructure target portfolios if we saw a dramatic change in correlations, if the reason for the change was evident, and if we believed the change represented a long-term structural shift that would characterize the relationship between asset classes going forward.

Portfolio: Why do you think that correlations between the S&P 500 and international stocks have increased in recent years while correlations between the S&P and REITs have decreased?
Gibson: This is just a guess, but I think the dominance of large-cap growth and technology stocks throughout the 1990s drove the correlation higher between U.S. and non-U.S. stock markets.

If we think back to the speculative tail end of the bull market, large-cap growth and technology stocks were soaring both here and abroad. When the bear market arrived, those high-flying, large-cap growth and technology stocks were hit the hardest. Again, this was true for both domestic and international stocks. They rose together, and they fell together. We may find that, once the excesses of the 1990s are fully washed out of the markets, that the U.S. and non-U.S. stock correlation drops a bit, while the correlation between U.S. stocks and real estate increases somewhat. Time will tell.

Portfolio: You have advocated the use of REITs as an asset class for as long as your firm has been in existence. Where do REIT stocks fit into your clients' portfolios, and what benefits do they provide investors?
Gibson: We give a decent allocation to real estate equity securities—generally 3 percent to 14 percent depending on client goals and risk tolerance. We include equity REITs in order to diversify the portfolio structure. It all goes back to what I was saying earlier about a diversified portfolio winning in the end. We do not have a separate allocation for mortgage REITs.

Portfolio: What do you think about some of the newer wrinkles in REIT investing like exchange-traded funds (ETFs) and international REITs?
Gibson: I'm happy with the asset class. It's hard to beat REITs for long-term total return, a major portion of which is distributable cash flow. To the extent that people are finding more efficient, lower cost ways to access the asset class, I applaud that. So I like the index funds and exchange-traded funds. I believe that the use of real estate as an asset class will continue to grow and become increasingly securitized in the U.S. and internationally.

Portfolio: You wrote in your book in 2000 that REITs were something of a “no-man's land” in the investment world. Are they stocks? Are they real estate? But you wrote that those attitudes were beginning to change. How do your institutional and high net worth clients view REIT investing today? Is it still a tough sell?
Gibson: There's a lot more interest in REITs. People are aware of how well REITs performed during the U.S. stock bear market. Fortunately, REITs moved completely counter-cyclically, going up three years in a row while the broader U.S. stock market was cut in half. A meaningful allocation to real estate securities during that painful time made a huge difference in client portfolios.

As an asset class, the real estate security markets have expanded and matured considerably over the last 10 years. There's more trading volume and more depth in the REIT market now. REIT shares are becoming increasingly mainstream.