The Most Influential Quant Factors, Ranked by Evidence

Five factors carry most of the weight in the quant literature: momentum, value, quality, size, and low volatility. Momentum has the strongest evidence of the five. It shows up in over a century of data, across countries and asset classes, and it survived post-publication scrutiny better than its peers. Size has the weakest case, with much of the original small-cap premium fading after the papers that documented it appeared.

Key takeaways

• Momentum, the tendency of recent winners to keep winning over the next 3 to 12 months, is the most consistently documented return factor in equities.
• The value premium, measured as book-to-market in Fama and French's early-1990s work, was real over long horizons but went through a roughly decade-long drawdown in the 2010s.
• Quality, formalized by Novy-Marx in 2013 using gross profitability, predicted returns about as well as classic value measures and often worked when value didn't.
• The size premium is the most contested of the classic factors, and much of it weakened or disappeared in the decades after publication.
• McLean and Pontiff found that anomaly returns decayed by roughly a third to a half once the research was published, which makes implementation the deciding variable.

What is a factor in investing?

A factor is a measurable stock characteristic that has historically explained return differences across large groups of stocks. The idea came out of academic asset pricing. The original CAPM said one thing should explain returns: a stock's sensitivity to the market, its beta. Decades of data said otherwise. Certain characteristics, like how cheap a stock was or how it had performed recently, predicted returns in ways beta couldn't account for. Researchers packaged those characteristics into long-short portfolios and called them factors. The CAPM and factor models lesson walks through that progression step by step.

Every factor comes with a fight over why it works. The risk-premium camp says the extra return is compensation for bearing some real risk. The behavioral camp says investors make systematic mistakes and the factor harvests them. The answer matters, because a risk premium should persist after everyone knows about it, while a behavioral mispricing can get arbitraged away. Keep that fight in mind as you read each section below. (For a broader primer on the field itself, see what is quant finance.)

Momentum: the strongest evidence in the literature

Momentum is the tendency of stocks that outperformed over the past 3 to 12 months to keep outperforming over the next 3 to 12 months. Jegadeesh and Titman documented it in 1993: portfolios that bought recent winners and sold recent losers earned excess returns on the order of 1% a month over the following year. Carhart made it canonical in 1997 by adding momentum as a fourth factor to the Fama-French three-factor model, after finding that it explained most of the apparent persistence in mutual fund performance.

The standard signal is the 12-1 month return: total return over the trailing twelve months, skipping the most recent month. The skip matters. Returns over the latest few weeks tend to reverse, so including them dilutes the signal.

Why does it work? The honest answer is that nobody has produced a convincing risk story. The behavioral explanations are more persuasive: investors underreact to news at first, then pile in late. Momentum is not free, though. It crashes occasionally, most famously in 2009, when the prior year's losers ripped higher and a winners-minus-losers portfolio got destroyed. The premium has been compensation for living with that tail.

Momentum also has the rare distinction of working out of sample in both directions: in data predating the study and in markets the original authors never touched.

Value: the oldest premium

Value is the tendency of stocks priced cheaply relative to fundamentals to outperform expensive ones, classically measured by book-to-market. Fama and French put it at the center of asset pricing in 1992 and 1993, showing that high book-to-market stocks outperformed low ones by more than their market betas could explain, and building it into their three-factor model as HML.

The explanation war is sharpest here. Fama and French argued value stocks are riskier, often distressed businesses whose cheapness is fair payment for the chance they fail. Lakonishok, Shleifer, and Vishny countered in 1994 that investors simply extrapolate past growth too far, overpaying for glamour stocks and abandoning boring ones.

The honest caveat: the 2010s were brutal for book-to-market. Value lagged growth for the better part of a decade, and a live debate continues about whether book value still captures cheapness in an economy where the biggest companies' assets are mostly intangible. The premium revived after 2020, but anyone selling value as a smooth ride is selling something.

Quality: profitability as a factor

Quality is the tendency of highly profitable companies to earn higher subsequent returns than unprofitable ones, even at similar valuations. Novy-Marx made the cleanest case in 2013 with gross profitability, revenues minus cost of goods sold, scaled by assets. His striking claim was that this simple measure predicted returns about as well as book-to-market itself. Fama and French effectively conceded the point in their five-factor model, adding profitability and investment factors to the original three.

Quality earns its place partly through what it does for value. A cheap stock might be a bargain or it might be cheap for good reason, and profitability is the screen that separates the two. Pairing them has historically worked better than either alone.

Size: the shakiest of the canon

The size factor is the historical tendency of small-cap stocks to outperform large-cap stocks. Banz first documented the small-firm effect in 1981, and Fama and French built it into their model as SMB. It belongs on this list for influence, not for durability.

The caveats are serious. Much of the measured premium sat in the very smallest, least liquid stocks, where trading costs are highest and the returns hardest to capture. In US data after publication, the raw premium was roughly flat for long stretches. Asness and co-authors argued the premium reappears once you control for quality, since small caps include a heavy dose of junk. That may rescue the factor academically. As a standalone reason to buy small stocks, the evidence is the weakest of the five.

Low volatility: the anomaly CAPM said couldn't exist

The low-volatility factor is the finding that low-beta stocks earned higher risk-adjusted returns than high-beta stocks, the opposite of what the CAPM predicts. The lineage runs back to Black, Jensen, and Scholes in 1972, who found the relationship between beta and return far flatter than theory required. Frazzini and Pedersen sharpened it in 2014 with "betting against beta," a factor that bought low-beta stocks levered up and shorted high-beta stocks.

Their explanation is structural. Many investors who want more risk can't or won't borrow, so they reach for high-beta stocks instead, overpricing them. Add a retail preference for lottery-like names and you get a persistent overpricing of the exciting end of the market. One nuance worth keeping: this is mostly a risk-adjusted anomaly. Low-volatility portfolios earned similar returns with less risk, not spectacular returns outright.

The factors at a glance

Factor	Definition	Canonical study	Typical signal
Momentum	Recent winners keep winning over 3-12 months	Jegadeesh & Titman (1993); Carhart (1997)	12-1 month return
Value	Cheap stocks beat expensive ones	Fama & French (1992, 1993)	Book-to-market
Quality	Profitable firms beat unprofitable ones	Novy-Marx (2013); Fama & French five-factor	Gross profits / assets
Size	Small caps beat large caps	Banz (1981); Fama & French (1993)	Market capitalization
Low volatility	Low-beta stocks beat on a risk-adjusted basis	Black, Jensen & Scholes (1972); Frazzini & Pedersen (2014)	Beta or trailing volatility

Do factors still work after publication?

Mostly yes, but smaller. McLean and Pontiff tested nearly a hundred published anomalies and found their returns decayed by roughly a third to a half after the research went public. Capital chases published signals, and the easy part of the mispricing gets competed away.

The decay was not uniform. Size faded badly. Momentum kept showing up in fresh data and foreign markets. Value kept paying over full cycles despite its long droughts. The pattern fits the theory: factors backed by durable behavioral forces or structural constraints decayed less than fragile statistical artifacts. Publication is a stress test, and the canon above is what passed it.

Implementation is the real ranking

After decay, the question stops being "which factor has the best backtest" and becomes "which factor survives contact with real trading." Three things decide that: universe, rebalance schedule, and trading costs.

Momentum is the clearest example. It is a high-turnover factor, and in small caps the costs of all that trading can consume the premium. Run the same signal on large, liquid stocks at a monthly cadence and the costs become manageable while most of the signal survives. That trade-off is why Quant GT's model applies momentum and relative strength to a universe above $10B in market cap, rebalanced monthly into five picks; over its 8-year history the model returned an average of roughly 58% a year. The current model portfolio shows what that selection looked like this month, and why Quant GT's returns have been so high breaks down the construction choices behind those historical results.

The deeper lesson applies to every factor on this list. A factor is not a return you collect. It is a tendency you implement, and the implementation, not the paper, determines what you keep. The factors that survive aren't the ones with the cleverest backtests. They're the ones still cheap to trade after everyone has read the study, and by that standard, large-cap momentum has aged better than almost anything else in the literature.

FAQ

Which quant factor has the strongest evidence?

Momentum. Stocks that outperformed over the past 3 to 12 months kept outperforming over the next 3 to 12 months in over a century of data, across countries and asset classes, and the effect held up after publication better than the other classic factors.

How is the momentum factor measured?

The standard signal is the 12-1 month return: a stock's total return over the past twelve months, excluding the most recent month. The latest month is skipped because very recent returns tend to reverse.

Do quant factors stop working after they're published?

They weaken but rarely vanish. McLean and Pontiff found that published anomaly returns decayed by roughly a third to a half after publication, which is why trading costs and implementation now matter as much as the signal itself.

Is the small-cap size premium still real?

It is the most contested of the classic factors. Much of the original premium sat in the smallest, hardest-to-trade stocks and weakened after publication, though some research finds it reappears once you control for quality.