Factor Premiums: Risk-Based vs. Mispricing Explanations

The biggest unresolved question in factor investing is why factors work. Do they deliver returns because they compensate investors for bearing systematic risk—like beta compensates for market risk—or do they exploit persistent mispricings created by behavioral psychology and market frictions? The answer shapes how you expect factors to behave in crises, how long premiums persist, and whether they belong in a permanent portfolio or fade as markets evolve.

The Risk-Based Camp: Factors as Compensated Risk

The risk-based explanation draws from the Capital Asset Pricing Model (CAPM): investors require higher expected returns for bearing risk. If beta alone (sensitivity to the overall market) is insufficient to explain all systematic risk, then other risks—like value exposure or low-volatility exposure—deserve their own risk premiums.

Under this view:

• Value stocks are riskier: they are cheap because their future is uncertain, and investors demand a premium for bearing that distress risk. Value stocks underperform in recessions, so owning them is like buying insurance against growth—the premium is compensation.
• Small-cap stocks are riskier: they have less access to capital, weaker balance sheets, and higher bankruptcy risk. The size premium is the price of that risk.
• Low-volatility stocks are perceived as safer: they should offer lower returns, not higher. But if they do outperform on a risk-adjusted basis, perhaps they are bearing a hidden risk (e.g., duration risk, negative skew) that markets underprice initially.
• Momentum is harder to justify as risk, but some argue it proxies for macroeconomic momentum risk: stocks rising with improving conditions are less risky than those falling despite good conditions.

The appeal of risk-based reasoning: if factors compensate for risk, then their premiums should be stable and long-lived. You can reliably count on value beating growth over the long run, just as you count on equities beating bonds. You can hold the factor permanently as a core portfolio holding.

However, risk-based theory struggles with some puzzles: Why does low-volatility outperform high-volatility by so much if low-vol is safer? Why does momentum work across different asset classes and time periods if it is truly a risk adjustment?

The Mispricing Camp: Behavioral Anomalies

The mispricing explanation flips the script: factor premiums exist not because they compensate for risk but because investors are irrational and make systematic errors.

Under this view:

• Value outperformance is not a risk premium but a reward for buying stocks that investors have anchored too pessimistically on. When sentiment shifts, the undervalued stock rebounds not because risk decreased but because the bias corrected.
• Momentum wins because investors extrapolate recent trends (representativeness bias), pushing prices too far in one direction. Momentum riders profit from the trend before it mean-reverts. This has nothing to do with risk.
• Low-volatility outperformance reflects overconfidence bias: investors pay too much for exciting (high-growth, volatile) stocks and neglect boring, stable ones. As that bias fades (or as investors with different biases enter), the low-vol discount vanishes.
• Size premium comes from the fact that small stocks are harder to value and less liquid; investors avoid them, creating a mispricing that skilled investors exploit.

The appeal of mispricing reasoning: it explains why factors go in and out of favor (when the specific bias waxes or wanes) and why factors sometimes stop working for extended periods (when the market corrects the mispricing).

However, the mispricing view has a liability: if everyone now knows about the factor, won’t arbitrageurs and passive investing eliminate it? The answer (somewhat circular) is that some biases are so deeply rooted in human psychology that they never fully disappear, even when widely publicized.

How the Two Theories Make Different Predictions

On persistence: Risk-based → premiums are stable; mispricing → premiums fade post-discovery.

Look at the value premium. It was discovered academically in the 1980s. Risk-based theorists expect it to remain. Mispricing theorists expect it to shrink as passive money flows into value indices. The data from 2007–2022 is ambiguous: value underperformed for 15 years (supporting mispricing), then rebounded sharply (supporting risk). Both sides claim vindication.

On crisis behavior: Risk-based → factors that are risk should crash when that risk materializes (value crashes in panics because growth is safer; low-vol should outperform). Mispricing → behavior is context-dependent (panic can trigger herding into any asset, regardless of logic).

In March 2020, low-volatility stocks initially fell as hard as the broad market (a surprise for risk theory). But within weeks, they steadied while growth careened, suggesting the risk-based view eventually held. Mixed evidence again.

On factor correlation: Risk-based → factors are independent risks, so they should diversify (one factor crashes when its risk materializes; another is safe). Mispricing → factors tend to crash together when the bias affecting them all (e.g., “pay for growth”) reverses suddenly.

Data shows factors are somewhat diversifying but not perfectly—again, mixed.

The Practical Investor’s Dilemma

An investor building a portfolio must decide: do I count on factors to work forever (risk-based view) or to decay over time (mispricing view)?

If risk-based, a permanent 30% value tilt and 20% low-vol tilt are reasonable. If mispricing, you might use factors tactically, rotating them as the market corrects past biases, or avoid them altogether once they are widely known.

Most practitioners adopt a hybrid stance: assume some factors (like value, size) are risk-based and stable; assume others (like some momentum anomalies) are mispricing and decay; and treat low-volatility as unclear but defensible for retirement portfolios regardless of theory.

The strength of the empirical premium also matters: larger, more consistent premiums (value, size) are more likely to be risk-based; tiny, disappearing premiums (some earnings surprises) are likely mispricing. Momentum sits in the middle: large and persistent, but hard to justify as risk, making it the most theoretically contested factor.

Evidence and Unresolved Disputes

Academic research has produced:

• Supportive evidence for risk-based: Factors do covary with plausible economic risks (e.g., value stocks fall in bad times); some factors deliver lower returns in high-risk periods, consistent with risk compensation.
• Supportive evidence for mispricing: Factors outperform on a risk-adjusted basis by large margins that seem excessive for a pure risk premium; behavioral proxies (like dispersion of analyst forecasts) predict factor performance in ways consistent with behavioral psychology.
• Evidence of decay: Several factors (especially small-cap and momentum in the U.S.) have underperformed or disappeared post-discovery, suggesting mispricing.
• Evidence of persistence: Value and size premiums remain globally and historically, even in countries with advanced markets, suggesting they are not fleeting anomalies.

The honest answer: we don’t know for certain, and the answer may differ by factor.

Implications for Strategy

If you believe factors are risk-based:

• Hold factors permanently; they are core portfolio tilts.
• Expect performance to regress to historical averages over 10+ year periods.
• Don’t abandon factors after a bad stretch; the premium will eventually compensate.

If you believe factors are mispricing:

• Use factors tactically; rotate them as the market corrects past biases.
• Abandon factors if valuations become crowded (too much money chasing the factor).
• Expect premiums to shrink as knowledge spreads; use factors before they are too popular.

If you believe both (the rational middle ground):

• Use a core factor allocation (e.g., value, size) as a permanent holding, expecting a modest, stable premium.
• Overlay with tactical rotations in smaller factors (momentum, quality) based on valuation and crowding.
• Rebalance annually to reduce the influence of near-term behavioral swings.

See also