Factor Model

Factor models are statistical frameworks that decompose the returns of financial assets into a systematic component explained by common factors and an idiosyncratic component unique to each asset. The Capital Asset Pricing Model (CAPM) is the simplest factor model: it attributes all systematic risk to a single factor, market beta. Modern extensions — the Fama-French three-factor model, the Carhart four-factor model, the q-factor model — add size, value, momentum, and investment factors to capture patterns that CAPM cannot explain.

The mathematical structure is seductive: if asset returns are driven by a small number of observable factors, then diversification across many assets eliminates idiosyncratic risk, leaving only factor exposure. But this assumes the factors are stable, the relationships are linear, and the world does not change its regime. In practice, factor premiums decay as capital flows into strategies designed to harvest them. Value investing was profitable until it became popular; momentum was robust until it was published. Factor models are elegant snapshots of historical regularities that provide no insurance against structural change.

The quant crisis of August 2007 demonstrated the danger. Quantitative funds using similar factor models simultaneously deleveraged, creating a liquidity spiral that erased years of returns in days. The models had not failed individually; they had failed collectively because their shared structure made them vulnerable to the same shock. This is the network effect of quantitative finance: when too many participants use the same model, the model becomes the market, and the market becomes the model's blind spot.

The search for robust statistical methods that survive regime change — that maintain their validity when correlations shift, volatilities spike, and liquidity evaporates — remains the central unsolved problem in quantitative finance.