Strategic asset allocation is a large part of our business and has rightly earned a prominent place in the firmament of High Money Management. One core assumption behind diversification across asset classes is that those asset classes will demonstrated relatively low levels of correlation (or, to be more precise, covariance). When A is down, the theory goes, B will be up (or down less, or in any case somehow independent of A). Rooted in modern portfolio theory, the point here is not only to reduce portfolio volatility and risk, but to achieve the maximum possible return for a given level of volatility and risk.
But like everyone else on and around the Street, we've noticed an apparent rise is correlation among several major asset classes in recent months and years, with the financial markets' Summer squall demonstrating this development in stark terms.
In the chart at right, you can see year-to-date performance for six major asset classes: U.S. equities (blue line); international equities (red line); emerging market equities (black dashes); U.S. bonds (green dashes); U.S. REITs (gold line); and commodities (pink line). Note that you can click on the chart to see a larger image.
The obvious exceptions to the relatively tight correlations on display in this chart are domestic fixed income (which held relatively steady, with Treasuries outperforming corporates) and REITs (which underperformed other asset classes after their January spike and still took a big hit in late July along with nearly everything else).
Look at the YTD numbers for U.S. and international, stocks, commodities, and fixed income: the range runs from 2.35 to 6.51% (not including dividend and interest payments, which would tighten the range even more).
Now take a look at the mid-July to mid-August period of maximum market stress. With the clear exception of fixed income and the partial exception of commodities, asset values deteriorated in near-lockstep, falling 10 to 15% from peak to trough.
We see three primary implications of all this.
First, it's not at all obvious that correlations will remain elevated, so the point is definitely not to abandon strategic asset allocation. And even if correlations have risen, they're still well short of one, so there's still value in diversification across asset classes.
Second, the inclusion of historically unorthodox (but increasingly widely-held) asset classes such as commodities has probably become more important for individual investors. Which isn't to say commodities will definitely outperform, just that they're marginally more likely to perform differently.
Third, we think it makes more sense than ever to seek out non-correlated strategies as a complement to increasingly correlated asset classes. Whether it's through hedge funds, non-traditional mutual funds, separate account managers, or some other mechanism, the risk-return optimization of modern portfolio theory increasingly calls for diversification across strategies as well as asset classes.
If you're interested in reading more on the dilemmas created by rising correlations, we recommend James Picerno's thoughtful piece in the latest issue of Wealth Manager.
Source
James Picerno, "Decisions, Decisions...," Wealth Manager, September, 2007