Engineering returns did a piece on whether low risk outperforms high risk and his results confirmed that it indeed does. He measure risk as being the historical volatility. In this piece, I’d like to do it with another measure and apply it to three different indices of stocks to test for robustness.
In the following study, I will be using beta as the measure for volatility. My universe of stocks is separated into 3 portfolios: SP500, custom Mid-Cap, and custom Small Cap. My custom mid-cap portfolio consists of 950 stocks with market cap between 1B-5B while my custom small-cap portfolio consists on 1100 stocks with market cap between 300M-1B. All portfolios and data are from Thompson Reuters point-in-time database adjusted for survivorship bias.
To test, each week I will be buying ether the top 50 stocks with the highest beta or buying the bottom 50 stocks with the lowest beta. The rationale behind this is that stocks that are low in beta are less volatile while vice versa for stocks that are high in beta. Rinse and repeat each week. Test starts from 2001 to 2012.
The above chart represents the strategy applied to the SP500 stock index assuming 100 dollars initially invested. One can clearly identify the out-performance of the strategy which also has a smoother equity curve
The above table represents the corresponding return and performance measures against each respective benchmarks. As you can see low volatility (Bottom 50) stock have outperformed in all instances with lower drawdown and standard deviation of returns. Note that mid-cap low volatility strategy underperformed its respective benchmark. I do not have an answer but it will require further testing to confirm if size affects the performance of the low volatility anomaly.