The Fama-French Five-Factor Model Explained

The Fama-French model is the most important asset pricing framework in modern finance. Developed by Eugene Fama (Nobel Prize in Economics, 2013) and Kenneth French, it evolved from a single-factor world (CAPM) through the three-factor model (1993) to the current five-factor specification (2015). This article explains the mathematics, the factor construction methodology, key empirical findings, criticisms, and how to use the model for performance attribution.

The Authors: Fama and French

Eugene Fama is a professor at the University of Chicago Booth School of Business and is widely regarded as the father of modern empirical finance. His work on the efficient market hypothesis in the 1960s and 1970s laid the intellectual foundation for index investing. In 2013, Fama was awarded the Nobel Memorial Prize in Economic Sciences (shared with Lars Peter Hansen and Robert Shiller) for “empirical analysis of asset prices.”

Kenneth French is the Roth Family Distinguished Professor of Finance at the Tuck School of Business at Dartmouth College. Beyond his research contributions, French has performed an invaluable service to the academic community by maintaining a comprehensive, freely accessible data library at mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html. This dataset provides daily and monthly factor returns, portfolio returns sorted by various characteristics, and industry portfolios, all updated regularly. It is the standard reference dataset for virtually all asset pricing research.

The Evolution: CAPM to Five Factors

Stage 1: The Capital Asset Pricing Model (1960s)

The Capital Asset Pricing Model, developed by William Sharpe (1964) and John Lintner (1965), proposed that a stock’s expected excess return is determined solely by its sensitivity to the overall market — its beta. The CAPM equation is:

E(Ri) - Rf = βi * [E(Rm) - Rf]

Where E(Ri) is the expected return of asset i, Rf is the risk-free rate (typically the one-month Treasury bill rate), βi is the asset’s market beta (its covariance with the market divided by the market’s variance), and E(Rm) - Rf is the expected market risk premium. Under the CAPM, the only way to earn higher expected returns is to take on more market risk (higher beta).

The CAPM was theoretically elegant but empirically inadequate. By the 1980s, researchers had documented numerous patterns in stock returns that the CAPM could not explain: small stocks outperformed large stocks even after adjusting for beta (Banz, 1981), value stocks outperformed growth stocks (Rosenberg, Reid & Lanstein, 1985), and the relationship between beta and expected returns was much flatter than the CAPM predicted (Fama & French, 1992). These anomalies demanded a richer model.

Stage 2: The Three-Factor Model (1993)

In their landmark 1993 paper “Common Risk Factors in the Returns on Stocks and Bonds,” published in the Journal of Financial Economics, Vol. 33, Fama and French proposed a three-factor model that added size and value factors to the market:

Ri - Rf = αi + βi(Rm - Rf) + si(SMB) + hi(HML) + εi

The three factors are:

• Rm - Rf (Market) — The excess return of the value-weighted market portfolio over the risk-free rate. This captures the broad equity risk premium.
• SMB (Small Minus Big) — The return difference between portfolios of small-capitalization and large-capitalization stocks. Small is defined as below the NYSE median market cap; big is above the NYSE median. The factor captures the size premium.
• HML (High Minus Low) — The return difference between portfolios of high book-to-market (value) stocks and low book-to-market (growth) stocks. High and low are defined by the NYSE 30th and 70th percentiles of B/M ratio. The factor captures the value premium.

The coefficients βi, si, and hi are the asset’s factor loadings — its sensitivities to each factor. The intercept αi is the asset’s alpha: the portion of its return not explained by the three factors. If the model is correct and complete, alpha should be zero for all assets.

The three-factor model was a major advance. It explained most of the cross-sectional anomalies that had plagued the CAPM, including the size and value effects. Portfolios that appeared to generate alpha under the CAPM turned out to have significant SMB and HML loadings — their excess returns were compensation for size and value risk exposure, not genuine skill.

Stage 3: The Five-Factor Model (2015)

In 2015, Fama and French published “A Five-Factor Asset Pricing Model” in the Journal of Financial Economics, Vol. 116. Motivated by the theoretical work of Novy-Marx (2013) on profitability and the empirical evidence on the asset growth anomaly, they extended the model with two additional factors:

Ri - Rf = αi + βi(Rm - Rf) + si(SMB) + hi(HML) + ri(RMW) + ci(CMA) + εi

The two new factors are:

• RMW (Robust Minus Weak) — The return difference between portfolios of stocks with robust (high) operating profitability and stocks with weak (low) operating profitability. Operating profitability is defined as annual revenue minus cost of goods sold, minus selling, general, and administrative expenses, minus interest expense, all divided by book equity.
• CMA (Conservative Minus Aggressive) — The return difference between portfolios of stocks with conservative (low) total asset growth and stocks with aggressive (high) total asset growth. Asset growth is measured as the year-over-year percentage change in total assets.

Factor Construction: The 2x3 Sort Methodology

Understanding exactly how the factors are constructed is critical for anyone implementing or evaluating Fama-French factors. The process follows a precise, well-documented methodology.

Annual Rebalancing

Factor portfolios are reformed at the end of each June. The timing is deliberate: it ensures that fiscal year-end accounting data (most US firms have a December fiscal year-end) is available by the time portfolios are formed, avoiding look-ahead bias. Book equity values from the previous fiscal year-end (typically December) are used with market equity values from the end of the prior December to compute book-to-market ratios. Market equity (size) is measured at the end of June.

The Sorting Procedure

For each factor pair (size/B/M, size/profitability, size/investment), stocks are independently sorted into groups:

• Size breakpoint: The NYSE median market capitalization. Stocks below the median are “Small”; stocks above are “Big.” Importantly, only NYSE stocks determine the breakpoint, but stocks from NYSE, AMEX, and NASDAQ are all assigned to groups based on that breakpoint.
• Characteristic breakpoints: The NYSE 30th and 70th percentiles of the sorting variable (B/M, profitability, or investment). Stocks below the 30th percentile are in the bottom group, stocks above the 70th percentile are in the top group, and stocks in between are in the neutral group.

This independent 2x3 sort creates six value-weighted portfolios for each characteristic. For the B/M sort, the six portfolios are:

Computing the Factors

HML is computed as the average return of the two value portfolios minus the average return of the two growth portfolios:

HML = (SV + BV) / 2 - (SG + BG) / 2

This construction isolates the value effect from the size effect, because both the long and short sides contain a mix of small and big stocks.

RMW and CMA are constructed identically, substituting profitability and investment for B/M:

RMW = (Small Robust + Big Robust) / 2 - (Small Weak + Big Weak) / 2
CMA = (Small Conservative + Big Conservative) / 2 - (Small Aggressive + Big Aggressive) / 2

SMB in the five-factor model is the average of three SMB factors — one from each of the three 2x3 sorts (B/M, profitability, investment):

SMB = (SMB_BM + SMB_OP + SMB_Inv) / 3

Where each component SMB is the average return of the three small portfolios minus the average return of the three big portfolios from its respective sort. This ensures that the size factor is not contaminated by value, profitability, or investment effects.

Key Empirical Findings

HML Becomes Redundant

One of the most striking findings of the 2015 paper is that HML becomes redundant when RMW and CMA are included in the model. The value factor’s alpha relative to the other four factors is close to zero, and its explanatory power is largely absorbed by the profitability and investment factors. Fama and French interpreted this as evidence that the value premium is not an independent source of return but rather a manifestation of the profitability and investment effects.

This finding has profound implications. It suggests that what investors have traditionally called “value” — buying cheap stocks — works not because cheapness itself is rewarded, but because cheap stocks tend to be profitable firms with conservative investment policies. The deep driver is profitability and capital discipline, not low price ratios per se.

What the Model Does Not Explain

Despite its improvements over the three-factor model, the five-factor model struggles with certain patterns. It does not fully explain the returns of small stocks with low profitability and aggressive investment — these stocks earn returns that are much lower than the model predicts. The model also fails to capture momentum, which remains a significant anomaly even after controlling for all five factors.

The Missing Factor: Momentum

The most important criticism of the Fama-French five-factor model is the absence of momentum. The momentum effect — that stocks with high recent returns continue to outperform and stocks with low recent returns continue to underperform over 3–12 month horizons — was documented by Jegadeesh and Titman (1993) and formalized as a factor by Carhart (1997).

Momentum remains statistically and economically significant even after controlling for all five Fama-French factors. Adding Carhart’s UMD (Up Minus Down) factor to the five-factor model produces a six-factor specification that explains a broader set of cross-sectional return patterns. Many practitioners use this six-factor version as their standard model.

Fama and French have acknowledged that momentum is a challenge for the five-factor model. They excluded it partly because it is difficult to explain within a risk-based framework — it is hard to argue that stocks that have gone up recently are fundamentally riskier than stocks that have gone down. Behavioral explanations (underreaction to news, herding, the disposition effect) are more naturally suited to explaining momentum, which sits uncomfortably in Fama’s efficient-markets worldview.

Practical Use: Performance Attribution

The most common practical application of the Fama-French model is performance attribution — decomposing a portfolio’s returns into the contributions from each factor and a residual alpha. This is how institutional investors evaluate whether a fund manager is truly skilled or simply harvesting well-known factor premiums.

The Regression

To perform a Fama-French attribution, you regress your portfolio’s excess returns (returns minus the risk-free rate) on the five factors using ordinary least squares (OLS):

Rp - Rf = α + β(Rm - Rf) + s(SMB) + h(HML) + r(RMW) + c(CMA) + ε

The output tells you:

• α (alpha) — The intercept. This is the portfolio’s risk-adjusted excess return after accounting for all five factor exposures. A positive, statistically significant alpha indicates genuine skill (or at least a source of return not captured by the five factors).
• β, s, h, r, c (factor loadings) — These tell you how much of the portfolio’s return came from each factor. A positive s means the portfolio tilts toward small stocks; a negative h means it tilts toward growth stocks.
• R-squared — The fraction of the portfolio’s return variance explained by the five factors. A high R-squared (e.g., 0.95) means the portfolio is essentially a factor portfolio with little idiosyncratic return.

Example Interpretation

Suppose you run a Fama-French regression on a fund and find:

α = 0.15% per month (t-stat = 2.3)
β = 1.05, s = 0.35, h = 0.42, r = 0.18, c = 0.10
R² = 0.92

This tells you the fund has a statistically significant alpha of about 1.8% per year (0.15% x 12). It has slightly above-market beta, a significant small-cap tilt (s = 0.35), a strong value tilt (h = 0.42), a modest profitability tilt, and a slight conservative-investment tilt. The five factors explain 92% of its return variance.

Without the Fama-French regression, this fund might appear to have much higher alpha — say 6% per year under the CAPM. But most of that apparent alpha is actually compensation for size and value risk, which could be obtained cheaply through index funds. The true skill-based alpha is only 1.8%.

Using the French Data Library

Kenneth French’s data library is an extraordinary resource. It provides:

• Factor returns — Daily and monthly returns for the five factors plus the risk-free rate and momentum, from July 1963 to the present.
• Sorted portfolio returns — Returns on portfolios sorted by size, B/M, profitability, investment, momentum, and other characteristics. These are essential for replicating academic studies.
• Industry portfolios — Returns on portfolios of stocks grouped by SIC code into 5, 10, 12, 17, 30, 38, 48, or 49 industry groups.
• International data — Factor returns for developed markets (Europe, Japan, Asia Pacific) and global portfolios.

All data files are available as CSV downloads at no cost. The data is updated monthly, typically within the first two weeks of each month. The documentation is minimal but the file formats are consistent and well-understood by the research community.

Criticisms and Limitations

Momentum Is Missing

As discussed above, the omission of momentum is the most widely cited limitation. The six-factor model (adding UMD) is preferred by most practitioners.

Low-Volatility Anomaly

The five-factor model does not explain the low-volatility anomaly — the empirical finding that low-beta and low-volatility stocks have historically earned higher risk-adjusted returns than high-beta and high-volatility stocks. This anomaly contradicts the fundamental CAPM prediction that higher risk (beta) should be compensated with higher returns.

Factor Timing and Macro Regimes

The Fama-French model assumes constant factor premiums. In reality, factor premiums vary over time with macroeconomic conditions. Value tends to outperform during economic recoveries; momentum tends to crash during sharp reversals. The model does not capture these dynamics.

International Evidence

While the factors have been documented internationally, the magnitudes and statistical significance vary across regions. The value premium, for example, has historically been stronger in Europe and Japan than in the US. The five-factor model was primarily developed and tested on US data, and its applicability to other markets should not be taken for granted.

The Bigger Picture: Risk vs. Mispricing

The deepest debate in asset pricing is whether factor premiums represent compensation for risk or mispricing due to behavioral biases. Fama, a staunch advocate of market efficiency, interprets the factors as risk premiums: investors who hold small, value, profitable, conservative stocks are bearing risks that the average investor wants to avoid, and they are compensated for doing so.

The behavioral camp, led by researchers like Robert Shiller, Werner DeBondt, and Richard Thaler, argues that factors reflect systematic mistakes by investors. Value stocks are cheap because investors overreact to bad news and extrapolate recent poor performance too far into the future. Momentum exists because investors underreact to new information, causing prices to adjust too slowly.

The truth likely lies somewhere in between. Some factors (like the equity risk premium) are almost certainly compensation for risk. Others (like short-term reversals) are almost certainly behavioral. Most fall in a gray area where both explanations have some validity. For practical purposes, the distinction matters less than the persistence and reliability of the premiums — whether a factor premium exists because of risk or mispricing, an investor who harvests it still earns excess returns.

References

1. Banz, R. W. (1981). “The Relationship Between Return and Market Value of Common Stocks.” Journal of Financial Economics, 9(1), 3–18.
2. Carhart, M. M. (1997). “On Persistence in Mutual Fund Performance.” The Journal of Finance, 52(1), 57–82.
3. Fama, E. F. & French, K. R. (1992). “The Cross-Section of Expected Stock Returns.” The Journal of Finance, 47(2), 427–465.
4. Fama, E. F. & French, K. R. (1993). “Common Risk Factors in the Returns on Stocks and Bonds.” Journal of Financial Economics, 33(1), 3–56.
5. Fama, E. F. & French, K. R. (2015). “A Five-Factor Asset Pricing Model.” Journal of Financial Economics, 116(1), 1–22.
6. Jegadeesh, N. & Titman, S. (1993). “Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency.” The Journal of Finance, 48(1), 65–91.
7. Lintner, J. (1965). “The Valuation of Risk Assets and the Selection of Risky Investments in Stock Portfolios and Capital Budgets.” The Review of Economics and Statistics, 47(1), 13–37.
8. Novy-Marx, R. (2013). “The Other Side of Value: The Gross Profitability Premium.” Journal of Financial Economics, 108(1), 1–28.
9. Rosenberg, B., Reid, K. & Lanstein, R. (1985). “Persuasive Evidence of Market Inefficiency.” The Journal of Portfolio Management, 11(3), 9–16.
10. Sharpe, W. F. (1964). “Capital Asset Prices: A Theory of Market Equilibrium under Conditions of Risk.” The Journal of Finance, 19(3), 425–442.