Benchmarking Low-Volatility Strategies

Abstract

In this paper we discuss the benchmarking of low-volatility investment strategies, which are designed to benefit from the empirical result that low-risk stocks tend to earn high risk-adjusted returns. Although the minimum-variance portfolio of Markowitz is the ultimate low-volatility portfolio, we argue that it is not a suitable benchmark, as it can only be determined with hindsight. This problem is overcome by investable minimum-variance strategies, but because various approaches are equally effective at minimizing volatility it is ambiguous to elevate the status of any one particular approach to benchmark. As an example we discuss the recently introduced MSCI Minimum Volatility indices and conclude that these essentially resemble active low-volatility investment strategies themselves, rather than a natural benchmark for such strategies. In order to avoid these issues, we recommend to simply benchmark low-volatility managers against the capitalization-weighted market portfolio, using risk-adjusted performance metrics such as Sharpe ratio or Jensen’s alpha.

Keywords: benchmarking, low-volatility, volatility effect, minimum variance, minimum volatility, alpha