The Problem with Asset Allocation
There is a widespread consensus that the most important decision investors make is asset allocation. The argument is that the mix of stocks, bonds, and other assets explains much more of the variance in investor returns than which exact stocks, bonds, or assets you choose.
But asset allocation guru Meb Faber compared a host of asset allocation strategies and found that the spread between the worst performing asset allocation and the best was only 1.8% per year—and that was based on backtests that likely overstate the returns of the best asset allocation strategy (which had a heavy weight on emerging markets and private equity starting in the 1970s). All of the complex models he tested had roughly the same Sharpe ratios of between 0.4 and 0.6.
Figure 1: Real Returns of Asset Allocation Strategies
Source: Meb Faber
Asset allocation may be the biggest driver of variance in portfolio returns, but that doesn’t mean that it’s easy to pick the best asset allocation strategy or that this is necessarily a great source of investor edge. In fact, researchers from London Business School and University of Texas put 14 asset allocation models to the test and found that the best performing was a naïve equal-weight allocation. They argue that the data on these asset classes does not go back far enough to make reliable statistical predictions—certainly not far enough to justify the use of the mean variance models so popular among consultants.
But we do know a few things about asset allocation.
We know that stocks outperform all other asset classes in the long run and do so with remarkable stability. In his book Stocks for the Long Run, Jeremy Siegel notes that from 1802 to 1977, “stocks have yielded between 6.6 and 7.2 percent per year after inflation” in all major subperiods, through massive economic, social, and political changes.
Figure 2: Total Real Return Indexes by Asset Class, 1802–1997
Source: Siegel, Stocks for the Long Run
And we know that as soon as we start talking about asset allocation and mixing in different asset classes, we are talking about moving away from equities and thus lowering returns. Below are the nominal returns and Sharpe ratios of various asset allocations compared to the stock market from 1972 to 2013. All the portfolios outperform stocks by Sharpe ratio, but fall behind on absolute returns, with the exception of the El-Erian portfolio, which included a large weight in private equity and a large weight in emerging markets starting in the early 1970s and thus benefits from hindsight bias.
Figure 3: Portfolio Nominal Returns, 1973–2013
Source: Meb Faber
Investors seeking to maximize long-term returns, therefore, do not have a good alternative to a 100% equity portfolio. There are strategies within equities for improving long-term returns, most notably the long-term premium for buying small value stocks. And there are strategies for reducing volatility in equities without giving up returns, like international diversification. But that’s about the limit of what we really know with a high degree of confidence.
But a 100% equity portfolio can experience significant volatility and long stretches of poor performance (as investors paying attention to the market this month can attest). Ray Dalio, famous for having predicted “5 of the last 3 recessions,” warned in a recent piece about his fear of paradigm shifts where what’s worked well recently (equities) might not work well in the future. Dalio advocates diversification as the best defense against paradigm shifts (and buying gold, but we won’t touch that one). “Any single approach to investing—e.g., investing in any asset class, investing via any investment style (such as value, growth, distressed), investing in anything—will experience a time when it performs so terribly that it can ruin you,” he warns.
So is there a way to diversify out of equities without sacrificing return? Is there a way to avoid some of the stress of volatile markets without paying a high price in terms of performance?
In a famous paper entitled “Why Not 100% Equities?” AQR founder Cliff Asness examined Richard Thaler’s argument that endowments should be 100% in equities. He argued that while diversification out of equities reduced returns, investors willing to lever up lower-risk assets could achieve better returns than 100% equity exposure.
Figure 4: Effect of Leverage
Source: Asness, "Why Not 100% Equities?"
Asness’s research suggests that investors willing to lever up bonds and other lower-risk diversifying assets can indeed do better than 100% equities, a thesis he shares with Ray Dalio.
But investors unwilling to use leverage don’t have any better options than a 100% equity portfolio if long-term returns are the primary goal. Other asset classes can outperform equities for a time and so tactical asset allocation through market timing might be a possibility, but market timing is notoriously difficult. So investors are left with stocks and, for those willing to use leverage, leveraged bonds.