Correlations, Data Frequencies & Expectations

With the return of volatility this year, we’ve noticed that investors, journalists, and fund rating services immediately put alternative funds under the microscope to determine if they lived up to expectations. I recently wrote an article for Investment Advisor’s May issue discussing how many quickly concluded—incorrectly I might add—that alternative funds failed. This was a result of confusing low correlation with negative correlation, and by conditioning on risk statistics calculated with monthly returns.

Let’s start with correlation. Many mistakenly assume that alternatives are negatively correlated to stocks. A negatively correlated asset would be expected to rise as stocks fall. However, as we’ve written about before, managed futures, as an example, cannot be counted on to exhibit negative correlations to stocks. Over time these strategies have experienced periods of both high positive correlations and high negative correlations with stocks, as can be seen in the chart below, but overall, they have been relatively uncorrelated. Investors should assume a low correlation over time, and be prepared for the fact that it could be materially positive at inconvenient moments.

The second issue that confuses investors during short term volatility spikes, is the disparity between risk expectations based on monthly return data, and risk as experienced on a day-to-day basis. We’ve entered an era where investors are obsessed with ever shorter time periods; time periods that aren’t remotely adequate to judge the efficacy of a strategy. Following fund performance daily, or ETF performance hourly, will undoubtedly lead to disappointment when comparing how investments have acted relative to risk statistics calculated with monthly returns (which is the industry standard).

Portfolios built using monthly return data to gauge correlations and risk characteristics need to be evaluated using monthly return data, ideally over years of performance, and most certainly not over the course of a week.

Read more in How to Measure Returns in Volatile Times >