Low-Volatility Returns: Evidence and Market Limits
Summary
The document asks whether stocks with lower historical return standard deviation tend to earn higher future returns, and whether this finding is broadly established. It distinguishes this measure from idiosyncratic volatility and cites two studies that report an inverse relationship between historical volatility and subsequent stock returns. A cited replication uses beta as an alternate risk measure across small, mid, and large-cap stocks; its reported comparison supports the anomaly.
The responses also give a counterexample from U.S. government bonds: a low-volatility portfolio concentrated in Treasury bills would historically earn less than one holding longer-duration Treasury bonds. This highlights that the relationship may depend on the asset class and the return being examined. The discussion is brief and does not provide study designs, sample periods, statistical estimates, or enough evidence to establish how universal or robust the stock result is.
Key ideas
- Some cited research reports that stocks with lower historical return volatility have higher subsequent returns.
- A replication using beta across multiple stock-cap segments is described as supporting the low-volatility effect.
- The stock finding is distinct from the debated relationship between idiosyncratic volatility and returns.
- Treasury bills and longer-duration bonds illustrate that lower volatility does not imply higher returns across all asset classes.
Tags
Full text
# Why do low standard deviation stocks tend to have superior future returns? # Why do low standard deviation stocks tend to have superior future returns? I've recently stumbled on something that really surprised me. These papers (1, 2) find that past standard deviation of returns is inversely related to future returns. That is, portfolio of low historical variance constituents (stocks) $\to$ better returns. My question: Are there any studies that contradict these findings for any market(s)? Or is this pretty much the consensus view at the moment? Is this anomaly as well established as the size, momentum, B/M anomalies? (I'm talking specifically about historical return standard deviation, and not idiosyncratic volatility. I already know that the results for idiosyncratic volatility are ambiguous and there are findings either way.) 1 "Benchmarks as Limits to Arbitrage: Understanding the Low-Volatility Anomaly", Baker, Bradley & Wurgler, (2011), Financial Analysts Journal. 2 "The Volatility Effect" by Blitz and Vliet, (2007), Journal of Portfolio Management. ## Answer by user1234440 (score 1) https://quant.stackexchange.com/a/7102 A blog article titled "Low Volatility" tries to replicate the results of the low volalitily anomaly. This article uses an alternate measure of volatility (beta) and checks for the robustness of the low volatility anomaly. Its universe encompasses small cap, mid cap, large cap. The results are summarized in the table whereby each top 50 stock’s performance are compared with the bottom 50’s performance. Results confirm the anomaly. ## Answer by Bryce (score 0) https://quant.stackexchange.com/a/4828 Yes, there are contradictions in other markets. Think of U.S. government debt securities. Building a low volatility portfolio would result in holding T-Bills, earning a meager return and experiencing very little volatility. Building a high volatility portfolio would result in holding longer duration T-Bonds, and would deliver appreciably higher returns than the T-Bill portfolio over time.
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.