Module 3

The Cross-Section of Stock Returns

Overview

The cross-section of stock returns asks why some stocks earn higher average returns than others. The CAPM gives a one-dimensional answer — exposure to the market portfolio — but decades of empirical research have documented that market beta alone cannot explain the variation in returns across stocks. This module builds a richer picture by introducing the multi-factor models that have become the standard tools for risk adjustment, return attribution, and performance evaluation in both academia and the asset management industry.

We begin with the Fama-French three-factor and five-factor models, developing the theoretical motivation, the portfolio construction mechanics, and a connection to the mean-variance frontier through the Intertemporal CAPM. We then study the momentum anomaly — one of the most robust and challenging findings in empirical finance — and examine how it has evolved over time and across international markets. The module closes with an application to mutual fund performance evaluation, showing how factor models can be used to disentangle a fund manager’s exposure to systematic risk factors from genuine value-added through active selection.

Topics

Fama-French Factors

  • The limitations of the CAPM as a single-factor model
  • The three-factor model: market, size (SMB), and value (HML); the five-factor model adds profitability (RMW) and investment (CMA)
  • Portfolio construction via the 2×3 double sort; theoretical interpretation via the Intertemporal CAPM and the mean-variance frontier
  • Downloading and visualizing factor data using getfactormodels
  • Rolling factor premiums and their time-varying behavior

The Momentum Factor

  • The momentum anomaly and the Carhart four-factor model
  • Construction of the MOM (momentum) factor
  • The momentum crash of 2009 and conditions unfavorable for trend-following
  • International comparison: US, Europe, and Emerging Markets momentum using getfactormodels

Mutual Fund Performance

  • The augmented five-factor model with momentum as a benchmark
  • Interpreting factor loadings: size tilt, value tilt, profitability, investment style, and momentum exposure
  • Estimating alpha as a measure of risk-adjusted outperformance
  • Survivorship bias and its implications for interpreting mutual fund alpha
  • Application: analyzing the Fidelity Value Fund (FDVLX) using OLS regression