What Is the Low Volatility Anomaly?
The low volatility anomaly is the empirical finding that stocks with lower risk, measured by volatility or beta, have historically delivered returns that were similar to, or in some studies higher than, those of riskier stocks. That cuts against a basic textbook expectation that higher risk should be rewarded with higher return. Researchers call it an anomaly because standard models do not predict it, not because it is guaranteed to appear in every market and period.
The textbook expectation
The Capital Asset Pricing Model (CAPM), developed in the 1960s, says that a stock’s expected return should rise in a straight line with its beta, its sensitivity to market moves. A stock with a beta of 1.5 should, on average, earn more than the market, and one with a beta of 0.6 should earn less. Plotted on a chart, that relationship is called the security market line.
What the data have shown
Several landmark studies found that the line in real data is flatter than the model predicts:
- Black, Jensen and Scholes (1972) tested CAPM on US stocks and found that low-beta portfolios earned more than the model implied and high-beta portfolios earned less.
- Haugen and Heins (1975) examined US stock returns and reported that, over the period studied, lower-risk stocks did not earn less than higher-risk ones.
- Ang, Hodrick, Xing and Zhang (2006) studied the relationship between stocks’ idiosyncratic volatility and later returns and found that stocks with high past volatility tended to have low subsequent returns.
- Later work has looked at many countries and periods and has often, though not always, found a similar pattern.
None of these results guarantees what happens next. Sample periods, methods and definitions matter, and some researchers argue that parts of the effect shrink after accounting for factors such as size, valuation or trading costs.
Why might it exist? Competing explanations
Researchers disagree on the cause, and several explanations may operate together.
Leverage constraints
Many investors cannot or will not borrow to increase market exposure. Instead, they buy higher-beta stocks to chase greater returns, which may push those prices up and future returns down. Frazzini and Pedersen (2014) developed a model along these lines.
Benchmarks and career risk
Fund managers are often judged against an index. Buying a low-risk stock that might lag a rally carries tracking-error risk, so some managers favour riskier stocks. Baker, Bradley and Wurgler (2011) argued that benchmark-based mandates can leave low-risk stocks relatively underpriced.
Preference for large payoffs
Some investors prefer assets with a small chance of a very large gain, even at a poor average return. If enough investors do so, those stocks may become overpriced.
Disagreement and limits on short selling
If investors disagree about a stock and optimists can buy more easily than pessimists can sell short, prices of contested, volatile stocks may reflect the optimistic view. Miller (1977) proposed an argument of this kind.
How researchers test for it
A typical study follows a simple recipe. At the end of each month, every stock in a universe is ranked by a risk measure, for example its volatility over the past year or its estimated beta. The stocks are divided into groups, often five (quintiles). Each group is held for the next month, returns are recorded, and the process repeats for decades. The researcher then compares the average return and risk of the lowest-risk group with the highest-risk group.
Variations matter. Equal-weighted groups give small stocks a bigger say than value-weighted ones. Excluding the smallest or least-traded stocks changes results. Adjusting for other known factors, such as value, momentum, size or profitability, asks whether low volatility offers something distinct. Each choice can raise or lower the apparent strength of the anomaly, which is part of why the academic debate continues.
A note on the evidence base
Most of the long-run data come from the United States, where records are deepest, although later studies have extended the work to Europe, Japan and emerging markets. Data before the 1960s are thinner, and the number of independent decades is smaller than the number of monthly observations suggests. That is one reason careful researchers describe the finding as a robust pattern in the samples studied rather than a law of markets.
Caveats worth knowing
- Period dependence. Low-volatility portfolios have lagged strongly rising markets and have held up better in falls, but outcomes differ by decade.
- Sector tilts. Portfolios of calmer stocks often hold more utilities, consumer staples and health care and fewer technology or cyclical firms.
- Interest-rate sensitivity. Because defensive stocks are sometimes treated as bond substitutes, they may react to changes in interest rates.
- Valuation and crowding. If many investors move into the same stocks, prices can rise and expected future returns fall.
- Measurement. Studies use different risk measures (total volatility, beta, idiosyncratic volatility) and different holding periods, so results are not interchangeable.
What it does and does not tell us
The anomaly is a statement about historical averages across many stocks. It says nothing certain about any individual company and does not promise future outperformance. It does, however, challenge the idea that taking more risk always brings more return, and it has influenced the design of index methodologies and funds. The site’s summary of research on low-risk stocks and its overview of academic papers on low volatility give further reading.
Frequently asked questions
Is the low volatility anomaly proven?
It has been documented in many data sets, but explanations and size of the effect remain debated.
Does it mean low volatility stocks always beat the market?
No. They have outperformed in some periods and lagged in others, particularly in strong rallies.
Is it the same as the low beta effect?
They are related. Low beta refers to market sensitivity; low volatility refers to total variability. The two overlap but are not identical.
Can I use it as an investment strategy?
That is a personal financial decision. Costs, taxes, goals and risk tolerance matter, and a licensed adviser can help.
Not financial advice
This article is general education and does not recommend buying or selling any security or strategy. Historical research findings do not guarantee future results, and investing involves risk, including loss of principal. Check primary sources, such as the original journal articles, for full details.