Monday, January 4, 2010

Year-end Thoughts on Correlation

I suppose I’m one of the most vocal critics of Modern Portfolio Theory (MPT), the body of academic work that lays the theoretical groundwork for today’s status quo method of portfolio construction, otherwise known as Strategic Asset Allocation. MPT is based on Harry Markowitz’s Nobel Prize winning paper, Portfolio Selection, in which he shows us how to craft “efficient portfolios” with the best or optimal mix of asset classes to earn the highest returns with the lowest amount of risk. I believe the investment industry has distorted Markowitz’s work over the years and the practical application of how it is used in portfolio construction is a scandal.

The secret sauce of Markowitz’s MPT formula is that by combining the correlation of returns with the standard deviation of returns, otherwise known as how asset class returns zig and zag with each other as well as how much asset class returns zig and zag by themselves, you can increase the returns of the portfolio at the same time that you reduce the risk of the portfolio. Of course, this all depends on a good forecast of zigging and zagging. As we close the books on 2009 Pinnacle investment performance, I’m struck by how Pinnacle analysts are forced to deal with MPT in the real, practical playing field of trying to deliver benchmark-beating performance. Because stocks and bonds are the only two asset classes in our benchmark, and because the S&P 500 Index (our stock proxy) is having an excellent year, we are forced to balance two important goals. One is to remain diversified in our portfolio construction as a matter of good risk management, and the other is to have enough risk or “juice” in the portfolio to outperform the S&P as our risk proxy in a bull market.

The problem, of course, is that the two goals can be mutually exclusive in the short-term. If we succeed in providing effective diversification, then the high octane securities we own to beat the S&P may zig and zag at precisely the wrong time, creating a situation where the portfolio has less volatility overall, but delivers lower performance…in the short-term. As we watch the performance of commodities, gold, the dollar, and international stock positions as we approach year-end, we only know that they are volatile enough by themselves to outperform the stock market. We can’t know if they will move up or down on the same days as the U.S. stock market, an unfortunate state of affairs if we are only focused on relative returns. At worst, we should end the year with about the same returns as our benchmarks but we will have done so with significantly less portfolio volatility.