Factor Models

CAPM says one number, beta, explains every asset’s expected return — and by the early 1990s that story was broken: small stocks, cheap stocks, and recent winners all earned far more than their betas could justify. Factor models are the fix, admitting there’s more than one kind of priced risk, and the brutal lesson is that once you measure honestly, most “alpha” evaporates — it was hidden factor beta all along.

Here’s what you’ll build, from the machinery up to the skepticism:

• From CAPM to factors — which anomalies broke the one-factor world, and what a “factor” even is.
• The Fama–French three-factor model — adds size and value, and how SMB and HML come from a 2×3 sort of every stock on the exchange.
• Momentum and the five-factor model — the strongest, scariest factor (momentum, with its rare catastrophic crashes), plus profitability and investment, and the q-factor challenger.
• Estimating factor exposures — the econometrics: time-series regressions for an asset’s loadings and alpha, plus cross-sectional and Fama–MacBeth two-pass tests of whether a factor is priced at all.
• Alpha vs factor exposure — decompose a real fund’s return into factor contributions plus a residual, and see why “beating the S&P” usually isn’t skill.
• Smart beta and the factor zoo — the 300+ published factors, why most are statistical mirages, and how the real ones lose a third to a half of their premium once everyone piles in.

By the end you’ll read a fund’s factor fingerprint, run a factor regression, tell genuine skill from a clever tilt, and give any shiny new “factor” exactly the suspicion it deserves — the toolkit professional allocators actually use to decide whether a manager is worth paying.