Carhart Four-Factor Model

The Carhart Four-Factor Model is an asset pricing model developed by Mark Carhart in 1997. It expands upon the Fama-French Three-Factor Model by incorporating a fourth dimension: price momentum.

While the Fama-French model successfully captured the risk premiums associated with company size and valuation metrics, it struggled to explain persistence in short-term returns.

By formalizing momentum as a systematic risk factor, the Carhart model provides a robust framework for mutual fund evaluation, performance attribution, and quantitative asset management.

Mathematical Formulation

The expected excess return of an asset or portfolio under the Carhart model is expressed as:

$R_{it} - R_{ft} = \alpha_i + \beta_{i1}(R_{mt} - R_{ft}) + \beta_{i2}SMB_t + \beta_{i3}HML_t + \beta_{i4}WML_t + \epsilon_{it}$

$R_{it}$ : Total return of asset or portfolio $i$ at time $t$ .
$R_{ft}$ : Risk-free rate of return.
$R_{mt} - R_{ft}$ : Market Risk Premium.
$SMB_t$ (Small Minus Big): Size Factor Premium.
$HML_t$ (High Minus Low): Value Factor Premium.
$WML_t$ (Winners Minus Losers): Momentum Factor Premium — also denoted as $MOM$ , this captures the return differential between a portfolio of past short-term outperforming stocks and past short-term underperforming stocks.
$\beta_{i1}, \beta_{i2}, \beta_{i3}, \beta_{i4}$ : Factor loadings indicating the portfolio’s sensitivity to each respective risk factor.
$\alpha_i$ : Carhart’s Alpha — the excess return that remains unexplained by the four systematic factors.

The Momentum Factor ( $WML$ )

The inclusion of $WML$ addresses an empirical anomaly identified by researchers Narasimhan Jegadeesh and Sheridan Titman in 1993: assets that perform well over a 3- to 12-month horizon tend to continue outperforming over the subsequent few months, while poor performers continue to underperform.

Attribute	Detail
Economic Rationale	Driven by behavioral biases such as investor underreaction to new information, herd behavior, and institutional “window dressing” (buying winning stocks at quarter-end to show clients).
Portfolio Construction	Typically constructed by ranking equities by their cumulative return over the past 11 months (excluding the most recent month to avoid short-term reversal effects). The factor is the average return of the top 30% (“Winners”) minus the average return of the bottom 30% (“Losers”).
Risk Profile	Momentum is highly cyclical and prone to sudden, severe “momentum crashes” during sharp market reversals, when historical losers abruptly outperform historical winners.

Institutional Applications and Global Examples

The Carhart Four-Factor Model is an institutional standard for isolating skill from systematic factor exposure, particularly in the evaluation of active fund managers:

AQR Capital Management (Global): Founded by Clifford Asness, a pioneer in momentum research, AQR heavily integrates the Carhart momentum factor across its alternative and long-only quantitative strategies. AQR implements systematic momentum overlays across global equities, fixed income, and currency markets.
Russell Investments (Global): Utilizes the Carhart framework within its manager research process. By analyzing a sub-advisor’s factor loading on $WML$ , Russell can determine whether a manager is generating genuine idiosyncratic alpha or simply riding market momentum.
Sovereign Wealth Funds (e.g., GIC Private Limited, Singapore): Use multi-factor models incorporating momentum to benchmark active public equity portfolios and manage total portfolio risk against structural style drifts.
Academic and Performance Benchmarking Data: The Center for Research in Security Prices (CRSP) and the Kenneth French Data Library continuously track and publish the $WML$ factor data alongside the original Fama-French factors for institutional research and performance baseline verification.

Comparison of Factor Frameworks

Model	Factors	Explanatory Power (R2)	Primary Use Case
CAPM	Market	~70%	Baseline cost of capital calculation.
Fama-French 3	Market, Size, Value	~90%	Structural asset allocation and style tilt evaluation.
Carhart 4	Market, Size, Value, Momentum	~93-95%	Performance attribution of active equity managers and mutual funds.

Mathematical Formulation

The Momentum Factor ()

Institutional Applications and Global Examples

Comparison of Factor Frameworks

The Momentum Factor ( $WML$ )