When Diversification Fails

Correlation coefficients are one of the most complicated areas of the asset management world, but the idea behind them is pretty simple–or, at least, most of us thought it was until the 2008-2009 meltdown.  The basic idea is that you study the price movements of, say, the stocks of large companies (represented by the S&P 500), and then look at the price movements of, say, stocks in the NAREIT (real estate) index.  You find that, on average, they tend to march to different drummers; when one goes up, the other goes up less, or it may go down.  When the other goes down, the first asset may go up or stay the same.  They have, in the parlance of experts, a low correlation.

These correlations between various flavors of stocks and real estate, commodities, bonds and other assets are expressed mathematically, and are one of the factors that professional investment advisors take into account when they build portfolios.  Whenever one kind of asset is going down, ideally you want something else in the portfolio to be going up, responding to different influences.

But all of these carefully-crafted models and all the higher mathematics went seriously awry during the 2008-2009 downturn, when every risk asset–from commodities to real estate to stocks–went down in concert as if the correlation coefficients had suddenly decided to converge at exactly the wrong time.  How could this happen?

At a recent investment conference, the outlines of a possible explanation began to emerge.  It was noted that all of those risk assets had one thing in common: they were financed or owned by the same small number of investment banking and brokerage institutions.  When Lehman Brothers went bankrupt and Bear Stearns was essentially folded into J.P. Morgan, when Citigroup and Merrill Lynch and Goldman Sachs suddenly had to rebuild their balance sheets, they all needed to sell assets to raise money.  The result: the world’s largest owners of risk assets were all desperate sellers at the same time.  Suddenly, all those assets, no matter how different their underlying economics, were in the same boat: they had to be sold so that companies could meet their net capital requirements and stave off bankruptcy.

And, of course, this caused those assets to have something else in common: they were dropping in value so fast that the average investor was scared out of his wits.  Instead of a run on the banks, as we saw in the 1930s, there was a run on the markets, fueled by the same kind of panic: will I be able to get my money out before it disappears?

This explains how the normal historical correlations failed to protect even the best-diversified portfolios.  The discussion then turned to: is there anything we can do about this going forward?  The solutions under discussion ranged from buying expensive hedges (which, of course, become dramatically more expensive during a panic), to selling into the teeth of the storm (and locking in significant losses), and, in general, the answers weren’t very satisfying.  The consensus was twofold: first, these kinds of panics don’t happen very often.  Interestingly, the mathematics of modern portfolio theory suggest that a 2008-like downturn should happen every 65-80 years, and that happens to be just about how long it was between the Great Depression and the Great Recession.

Second: these panics seem, in retrospect, to be great times to buy risk-based securities.  When others are selling in a panic, you can almost name your price, and to the extent that you don’t believe that civilization is coming to an end, you trust that sooner or later the stocks you bought cheaply will, when the panic subsides, rediscover their true value.  The trouble, as one advisor put it, is: how are you going to tell your frightened clients, in the height of a storm, that this is a great time to put more money into the market?  Is anybody going to listen to that advice when the largest global investing organizations are trying to unload those same assets at any price they can get?

The bottom line here is that professional investors are finally getting a handle on why well-diversified portfolios didn’t protect against the 2008 downturn.  But the fact remains that the people who can control their panic seem to be the only ones who will be protected the next time there’s a panic run for the exits.  Until we invent a cure for the human tendency to flee with the herd, investment portfolios are likely to go down the next time we experience a serious market downturn.  Let’s hope we’ll have to wait 60-80 years.