09 Aug Addendum to the Earnings Example
To better understand if the improvements are consistent over the 60-year period, let’s consider a moving average of a few risk-adjusted performance ratios: Sharpe, Sortino, and Calmar. The table shows that the low volatility portfolios have better, consistent performance for the equal- and value-weighted portfolios . To compare the portfolios, I calculate the monthly differences of the 36-mo. average of the ratios (i.e., low volatility portfolio minus the nee=0 portfolio) and show how often and by how much the low volatility portfolio exceeds the nee=0 portfolio.
| Equal-Weighted | Value-weighted | |
| Sharpe Ratio | ||
| Average Spread | 0.71 | 0.54 |
| Frequency (in %) | 88.72 | 85.56 |
| Sortino Ratio | ||
| Average Spread | 1.21 | 0.74 |
| Frequency (in %) | 87.62 | 81.29 |
| Calmar Ratio | ||
| Average Spread | 0.83 | 0.78 |
| Frequency (in %) | 86.93 | 82.39 |
The Sortino Ratio’s minimum acceptable return is the 3-mo. t-bill rate.
The graphs below illustrate the point. When the line is above the horizontal axis, the low volatility risk-adjusted performance exceeds the performance when nee=0. The Sortino Ratio considers performance below the minimum acceptable return to be risky. As a result, it does not penalize upside volatility or performance above the target return. This explains the larger average spread for the Sortino ratio, especially for the equal-weighted version.
In the previous post, I pointed out that the low volatility version had less dramatic drawdown events. The graphs of the Calmar ratio comparisons below illustrate the point. As a reminder, the Calmar ratio for a month is the annualized performance divided by the maximum drawdown over the trailing 36-month period.
