For years many professionals in finance have been touting the benefits of diversification, however there remains a long-standing debate about how best to pursue the objective—should investors have a concentrated yet diverse portfolio, a diverse portfolio, or should they just index and buy as broadly as possible? 2011 was a year in which we saw all kinds of problems with the pursuit of diversification and it highlights exactly how and why we like to approach this problem a little differently. First, the important math. The theoretical reason behind diversification is to eliminate single-stock risk to the point where a problem with one portfolio company or holding does not overly infect the entirety of the portfolio. Joel Greenblatt, a Columbia University professor, hedge fund manager and author did some interesting work to quantify this factor of risk. Owning two stocks eliminates 46% of the risk associated with individual stocks, eight stocks eliminates 81% of the risk, sixteen stocks eliminates 93%, and thirty-two stocks eliminates 96% of the risk. In order to mitigate 99% of the single-stock risk one most own 500 holdings. The clear point here is that mathematically speaking, the benefits to diversification continually diminish when the portfolio holds more than sixteen total stocks.
Why do we bring this up now? Well in 2011, correlations were as high as they ever were. In other words, every stock, and just about every asset class moved in the exact same direction. This implies that what was a risk to one stock, was also a risk to another. While in aggregate that statement appears to be true, it’s not entirely true. Directionally, stocks moved consistently in tandem, but in terms of magnitude of the move, there were vast differences. Plenty of stocks finished the year up nicely, while others finished the year down badly. Many diversified portfolios took substantial hits and the key factor lies in how diversification had been pursued.
Simply diversifying holdings is not enough. Importantly, 2011 highlighted the fact that people need to diversify their correlations. That means that investors must expose their portfolios to as many different catalysts as possible. It means buying quality companies in a variety of sectors, with a variety of strengths and weaknesses, which should overlap as little as possible in aggregate. Some might say, well why not just index then and achieve ultimate diversification? And that question, while a legitimate one, further confirms the initial point. An overly broad, diversified portfolio is exposed to nothing other than just economic growth. Without economic growth it’s nearly impossible for the market in aggregate to move higher, but that’s not entirely true for a well-selected, diversified basket of stocks, bonds and cash that are rebalanced tactically with layered and stratified correlations.