Introduction
Among all of the market anomalies discovered, the low-volatility anomaly stands out. It refers to the tendency of low-risk assets to outperform high-risk assets on a risk-adjusted basis, and, in some cases, even on an absolute return basis. This effect has been consistently observed across various asset classes, time periods, and countries, fundamentally challenging the traditional belief that higher risk is always rewarded with higher returns.
In this blog post, I explore the low-volatility effect in stocks, summarize key findings from the literature, test a variety of low-volatility signals, and examine how it can be combined with other types of signals to enhance portfolio performance.
Table of Contents
Background
Low-Volatility Signals
Empirical Testing
Data
Low-Volatility Results
Further Improvements
Combining Low Volatility with Momentum and Value
Investor Takeaways
References
Background
The low-volatility effect is one of the most well-documented anomalies in financial markets, directly challenging the risk-return relationship proposed by the Capital Asset Pricing Model (CAPM). First observed in the early 1970s, this phenomenon shows that low-risk stocks tend to generate higher risk-adjusted returns—and often higher absolute returns—than high-risk stocks. As early as 1975, Haugen and Heins observed that stock portfolios with low volatility outperformed their riskier counterparts, contradicting the traditional belief that higher risk should be rewarded with higher returns.
To illustrate this effect, I construct 10 equal-weighted decile portfolios by sorting stocks monthly based on their historical volatility, measured over the past year using daily returns. I then compute the market beta for each portfolio and plot the Sharpe ratio of each decile portfolio against its market beta:
In contrast to traditional finance theory, the figure illustrates that portfolios of low-risk stocks achieve significantly higher Sharpe ratios than those of high-risk stocks. Numerous studies have documented this pattern.