Navaratnas

Equity Strategy · 6 min read · 2026-02-27

The 9 reasons the low-vol anomaly persists.

Low-volatility stocks shouldn't outperform on a risk-adjusted basis. They do. Nine structural reasons explain why.

By the Navaratnas methodology team

The 9 Reasons the Low-Volatility Anomaly Persists — Navaratnas blog cover

A factor that violates CAPM and persists for decades.

Sharpe 1.5×
Low-vol vs broad market on risk-adjusted basis

Low-volatility stocks have produced market-equivalent or slightly higher returns at meaningfully lower volatility, producing risk-adjusted returns 50% higher than the broad market. The anomaly contradicts CAPM. Nine structural reasons explain its persistence.

The nine indicators

The nine reasons the anomaly persists.

Each is a structural feature of how investors and institutions actually behave. Together they explain why the factor has not been arbitraged away.

01

Leverage aversion in retail

Pattern: cap on borrowing

Most retail cannot or will not lever low-vol stocks to match broad-market exposure. Demand for high-beta stocks instead pushes them to overpriced levels.

02

Benchmarking constraints in institutions

Pattern: tracking-error penalties

Active managers benchmarked to the broad index avoid low-vol overweights because they produce tracking error in bull markets.

03

Lottery preferences for high-beta names

Pattern: convex demand

Retail buyers prefer high-beta stocks for upside potential, ignoring downside symmetry. The preference inflates high-beta valuations.

04

Compensation structure incentivizing high beta

Pattern: option-like payoffs

Hedge fund compensation structures (asymmetric upside) encourage high-beta strategies. The aggregate flow tilts toward high-beta and away from low-beta.

05

Lower analyst coverage of low-vol names

Pattern: fewer eyes

Low-vol stocks tend to be mature, less exciting companies with less analyst coverage. The information asymmetry creates inefficiency.

06

Behavioral neglect of compounding

Pattern: mathematical asymmetry

Compounding favors low-volatility return streams. A 50% loss requires 100% gain to recover. Most investors underweight this math.

07

Sector concentration in defensive industries

Pattern: utilities, staples, healthcare

Low-vol overlaps with defensive sectors. The structural growth of defensive sectors (e.g., aging population for healthcare) supports the factor.

08

Short-selling friction limits arbitrage

Pattern: high-beta hard to short cheaply

Arbitrageurs cannot easily short overvalued high-beta names; the borrow is expensive and risky. The anomaly persists because the short side is constrained.

09

Investor flow concentration in benchmarks

Pattern: index investing

Passive flow into market-cap-weighted indices does not preferentially fund low-vol stocks. The factor remains under-purchased.

What CAPM predicts and what reality shows

Capital Asset Pricing Model predicts that expected return is proportional to systematic risk (beta). High-beta stocks should earn higher expected returns; low-beta stocks should earn lower expected returns; risk-adjusted returns (Sharpe ratio) should be similar across the beta spectrum. The model is the foundation of modern portfolio theory.

Empirically, the model fails. Low-beta stocks have historically delivered returns similar to or higher than high-beta stocks while bearing materially lower volatility. The Sharpe ratio of low-volatility portfolios exceeds that of high-volatility portfolios consistently across decades and across markets. The pattern violates CAPM but persists in the data.

Frazzini-Pedersen and the leverage explanation

Andrea Frazzini and Lasse Pedersen's 'Betting Against Beta' (2014) provided the most influential explanation. Investors are leverage-constrained — they cannot borrow at the risk-free rate to scale a low-beta portfolio to broad-market risk. Because of this, demand for high-beta stocks (which provide built-in leverage) is inflated, pushing high-beta valuations to unsustainable levels and depressing forward returns.

The explanation is consistent with empirical evidence. The factor persists most strongly in periods and markets where leverage constraints are most binding. It fades when leverage is cheaply available.

Implementation considerations

Low-volatility ETFs (USMV, SPLV) provide retail access to the factor at low cost. Implementation choices vary — minimum variance optimization (USMV) versus simple low-volatility screening (SPLV) produce different sector and stock exposures. Both have delivered the factor's risk-adjusted advantage over the past decade.

The factor underperforms during sharp market rallies, where high-beta stocks dominate. Holders need patience through these periods. The long-term Sharpe advantage is durable; the short-term tracking error is meaningful.

Will the factor decay?

Smart-beta funds targeting low volatility have grown substantially. Some research suggests partial decay in the factor's premium since 2015 as flows have been substantial. However, the structural reasons for the anomaly — leverage constraints, benchmarking, lottery preferences — remain in place. Pure arbitrage of the factor away would require institutional change that has not occurred.

Get the nine, every Monday.

Free weekly digest. The only U.S. equity letter that publishes a name only when nine independent signals align.

Common questions

Questions.

Should I tilt my portfolio toward low-vol?

Modest tilt (10–20% overweight) captures the factor without excessive concentration. Low-vol ETFs are cheap; the tilt is cost-effective.

Does this work for individual stocks?

Yes — disciplined low-volatility stock selection has historical alpha. Implementation requires more work than ETF tilting.

What about high-yield as low-vol?

Different exposure. High-yield bonds are credit-sensitive in ways that low-vol stocks are not. Not interchangeable.

How does low-vol perform in bear markets?

Better than broad market. Low-vol stocks typically lose less in drawdowns, producing the persistent risk-adjusted advantage.

Is min-vol or simple low-vol better?

Min-vol portfolios use full covariance matrix optimization and achieve lower realized vol. Simple low-vol screens are less precise but cheaper to implement.

Does low-vol work internationally?

Yes, with similar magnitudes. The factor has been documented in most developed and emerging markets.