Factor Investing Beyond Momentum: What Still Works in Value, Quality, Size, and Low Volatility

Momentum gets the headlines because it is intuitive, visible, and hard to ignore when it works. But if you stop there, you miss the larger story in equity factor investing: the market has rewarded several distinct return drivers over long stretches, and the ones that survive best are usually the ones investors can explain, diversify, and actually hold through ugly periods.

The classic academic map starts with Fama and French’s work on size and value, then expands into profitability and investment, while the practitioner literature from AQR and others has argued for momentum, quality, and low volatility as economically sensible premiums rather than one-off anomalies.^[1][3][4] The catch is that publication changes behavior. Once a factor becomes widely known, capital flows in, spreads compress, and the easy version of the trade often gets arbitraged away.^[5][6]

That is why the useful question is not “which factor is best?” It is “which factor still earns its keep after costs, crowding, and regime changes?” For investors building multi-factor portfolios, that distinction matters more than any single Sharpe ratio. If you want the mechanics behind momentum specifically, see Momentum Premium; if you want the implementation pitfalls, Backtest Checklist is the right companion read.

1) The five-factor map: what each premium is trying to capture

Factor investing is not a single strategy. It is a family of return tilts that try to harvest persistent differences in expected returns across stocks. The five most discussed equity factors in public markets are value, momentum, quality, size, and low volatility. Each has a different economic story and a different failure mode.

Fama and French’s original size and value work remains foundational, but later updates showed that profitability and investment help explain cross-sectional returns better than size alone.^[1] Momentum, meanwhile, has one of the strongest and most persistent academic records, documented across countries and asset classes.^[2] Low volatility is more controversial in theory, but the empirical pattern is hard to dismiss: lower-risk stocks have often delivered surprisingly competitive risk-adjusted returns, especially when leverage constraints and behavioral preferences are in play.^[7]

Common mistake: investors often treat factors as interchangeable “alphas.” They are not. Value and momentum can complement each other because they tend to behave differently across market regimes, while size and low volatility often bring more implementation friction than the headline backtest suggests.

2) What the literature actually says: durable premiums versus post-publication decay

The cleanest way to think about factor evidence is to separate three layers: pre-publication discovery, post-publication validation, and post-publication decay. A factor can be real in the first layer and disappointing in the third. That is not a contradiction; it is the market doing what markets do.

Momentum is the best-known example of a premium that survived broad scrutiny. Jegadeesh and Titman’s original work showed that buying recent winners and selling recent losers produced abnormal returns in U.S. equities, and later research extended the effect internationally.^[2] But momentum is also a factor where implementation details matter enormously. Turnover is high, crashes happen, and the premium can be eaten by costs if the portfolio is not managed carefully.^[2][8]

Value has a longer and messier history. Fama and French documented that cheap stocks outperformed expensive ones over long samples, and the value effect became a cornerstone of factor investing.^[1] Yet the post-2007 period has been a reminder that a premium can go dormant for a long time. The recent “value is dead” narrative is overstated, but the live experience for investors has been uneven, especially when growth stocks dominated and rates were low.^[5][6]

Quality is the newer institutional favorite because it is easier to explain to clients than “buy junk that is statistically cheap.” Novy-Marx’s gross profitability work and later AQR-style quality definitions suggest that profitable, conservatively financed firms have tended to outperform lower-quality peers.^[3][4] Still, quality is not magic. It can become a crowded defensive trade, and if you define it too narrowly, you may simply be buying expensive large-cap compounders.

Size is the awkward one. The original small-cap premium was powerful in early samples, but later evidence has been less convincing once you control for other factors and realistic trading costs.^[1][5] In practice, the size premium is often the first factor investors overestimate and the first one they under-diversify into. Small stocks can be illiquid, expensive to trade, and highly sensitive to microstructure frictions.^[9]

Low volatility has a different problem: it is often dismissed because it sounds too defensive to be a “premium.” Yet the empirical record shows that low-risk stocks have often delivered better risk-adjusted returns than the market, even if their raw returns are not always spectacular.^[7] The tradeoff is obvious: you may give up upside in roaring bull markets, and the factor can become crowded when investors chase yield-like equity substitutes.

What investors get wrong: they often assume a factor’s academic existence guarantees a tradable edge. The literature says otherwise. The investable premium is usually smaller than the paper premium, and sometimes much smaller once the strategy is scaled.^[5][8][9]

3) A decade-by-decade comparison: the regime story matters more than the average

Long-run averages hide the real experience. Factors do not pay evenly. They cluster by decade, macro regime, and valuation starting point. That is why a decade-by-decade view is more useful than a single full-sample number.

Footnote: This table is an illustrative synthesis, not actual performance data. It is designed to show the regime pattern commonly discussed in the literature: value tends to be cyclical, momentum more persistent but crash-prone, quality more defensive, size less reliable in recent decades, and low volatility more stable but sensitive to crowded positioning. Assumptions: U.S. large- and mid-cap equities, long-only factor tilts, annualized decade-level directionality rather than exact return figures, no transaction costs, no taxes, and no leverage. For actual factor series, use the cited academic and data-library sources below.

The point of the table is not to pretend precision where we do not have it. It is to remind you that factor investing is regime investing in disguise. If you only look at the average, you miss the sequence of droughts and floods that determines whether a strategy is actually holdable.

For readers who want a broader portfolio lens, the regime question connects directly to Regime Detection and to the practical tradeoff between drawdown control and upside capture discussed in Sharpe vs. Calmar.

4) Correlation is the hidden variable: factors are not independent bets

One of the biggest mistakes in multi-factor investing is assuming that owning five factors means owning five independent return streams. In reality, factor correlations can be unstable and sometimes surprisingly high, especially during stress periods. Momentum and value often move differently over long horizons, but they can both suffer when market leadership changes abruptly.^[7][8]

Methodology note: the correlation matrix below is an illustrative planning tool, not AIBROKER backtest output. It is shown to help readers think about portfolio construction. If you want to understand how AIBROKER handles factor research, backtest hygiene, and data provenance, see /learn/methodology.

Footnote: Illustrative matrix for portfolio construction only. Assumptions: monthly factor returns, U.S. equities, 10-year rolling sample concept, no transaction costs, no leverage, and correlations rounded for readability. This is not AIBROKER backtest output and should not be interpreted as actual historical correlation estimates.

Even as a simplified matrix, the lesson is useful. Quality often overlaps with low volatility because profitable, stable firms tend to have lower drawdowns. Value can be negatively related to momentum because cheap stocks are often yesterday’s losers. Size may add diversification in theory, but in practice its correlation benefit can be overwhelmed by liquidity costs and factor overlap.

Practical takeaway: the best multi-factor portfolio is not the one with the most factors. It is the one with the least redundant exposure after costs. That is why a factor sleeve should be evaluated like a portfolio, not like a shopping list.

5) What investors get wrong about factor decay

The most common error is to treat factor decay as proof that the factor was never real. That is too simplistic. Decay usually reflects one or more of four things: crowding, implementation costs, changing market structure, or a bad starting valuation regime.^[5][6][9]

First, crowding. Once a premium becomes widely known, more capital chases it. That can compress expected returns even if the underlying anomaly still exists. Second, costs. A factor with high turnover can look excellent on paper and mediocre after spreads, slippage, and taxes. Third, market structure. The rise of passive investing, faster information diffusion, and tighter arbitrage can change how anomalies express themselves. Fourth, valuation. A factor can be “right” and still underperform for years if the entry price is poor.

This is where survivorship bias sneaks in. Investors often compare today’s surviving factor ETFs or model portfolios with dead strategies that disappeared from the record. That comparison flatters the survivors. If you want a refresher on the mechanics, read Survivorship Bias. The same logic applies to factor research: the published literature is not a random sample of all ideas ever tested.

There is also a behavioral trap. Investors tend to buy factors after a strong run and abandon them after a weak one. That is exactly backward if the premium is cyclical. The better question is whether the factor still has a plausible economic rationale and whether the implementation is disciplined enough to survive the bad years.

6) A worked example: building a simple multi-factor sleeve

Suppose an investor wants to tilt a core U.S. equity portfolio toward factors without turning it into a trading project. A sensible starting point is not “maximize factor exposure.” It is “choose a small set of complementary tilts with tolerable turnover.”

This is not a recommendation; it is a design pattern. The point is to combine factors that behave differently, then keep the implementation simple enough that you can actually stick with it. If you want a process lens, compare this with Systematic vs. Discretionary and the execution realities in Bid-Ask Spread.

Checklist: before you buy a factor ETF or build a custom sleeve

• Do I know the exact index methodology and rebalancing schedule?
• What is the turnover, and what does that imply for trading costs?
• Is the factor definition broad enough to avoid one-stock concentration?
• How does the factor behave in inflation shocks, rate shocks, and recessions?
• Am I comparing gross backtest returns to net investable returns?
• Do I understand whether the strategy is long-only, long-short, or market-neutral?

7) Decision tree: which factor deserves attention first?

Use this simple decision tree when you are deciding where to start. It is not a ranking of “best” factors. It is a filter for fit.

Worksheet: score each factor from 1 to 5 on three dimensions before allocating capital: evidence durability, implementation cost, and portfolio fit. A factor that scores 5 on evidence but 1 on cost may still be a poor choice for a taxable retail account. A factor that scores 4 on fit and 4 on cost can be more useful than a factor with a prettier backtest.

8) The honest assessment: which factors have aged best?

If you force a ranking based on the broad literature, momentum and quality have aged best from an investability standpoint, value remains credible but cyclical, low volatility is useful but often misunderstood, and size is the most fragile of the classic premiums.^{[1][2][3][4][7]}

That does not mean you should ignore size forever. It means you should be skeptical of claims that small caps are automatically superior. The original evidence was real, but the investable edge has been diluted by costs, changing market structure, and the fact that many small stocks are simply harder to own efficiently.^[9] Likewise, low volatility is not a substitute for a risk-free asset; it is an equity factor with its own sector and style biases.

Value deserves special nuance. It is the factor most likely to look broken right before it rebounds, which is why it attracts both believers and skeptics. The practical tradeoff is patience: if you cannot tolerate multi-year underperformance, you do not own value, you rent it. That is a very different proposition.

Momentum remains the most elegant anomaly and the most emotionally difficult to hold. It works because trends persist, but it can reverse violently. Investors who understand that tradeoff tend to size it more carefully and rebalance more systematically. Those who do not usually discover the crash risk the hard way.^[2][8]

Quality is the quiet workhorse. It rarely makes the loudest claim, but it often improves the portfolio’s behavior. For many intermediate investors, that makes it more useful than a more dramatic but less stable factor. If you are building around a core equity allocation, quality is often the factor that helps you stay invested.

For a broader framework on how to think about risk-adjusted outcomes rather than raw returns, see Three Numbers That Matter and Risk Measurement.

9) A timeline for evaluating factor claims

When a new factor paper or ETF launch crosses your screen, use a simple timeline rather than a gut reaction. The goal is to separate signal from story.

This timeline is the simplest way to avoid being seduced by a beautiful chart. It also aligns with the discipline behind Backtest Checklist and the risk framing in Risk Measurement.

So what: factor investing works best when you treat it as a portfolio engineering problem, not a prediction contest. The durable edge comes from combining sensible premiums, controlling costs, and accepting that some factors will spend long stretches looking wrong before they look right again.

That is the real lesson beyond momentum. The market still rewards certain styles, but it rarely does so in a straight line. Investors who survive the cycle are usually the ones who respect the evidence, distrust the easy backtest, and keep their implementation boring enough to endure.

Closing thought: if you want a factor portfolio that lasts, build for the years when your favorite factor is out of favor. That is where the difference between a theory and an investable strategy is usually decided.