Fama and French (1993)
Fama, E.F. and French, K.R. (1993), “Common risk factors in the returns on stocks and bonds.” Journal of Financial Economics, 1993.
Most academic papers analyzing equity portfolio returns over time reference Fama & French. Their model extends CAPM by adding Book-to-Price and Size variables to account for longstanding anomalies observd in the market. Note that the model is structured so that if there were no valuation effects or size effects, the associated terms would zero out and we would be left with CAPM again.
Journalist Justin Fox comments ("What the Great Fama-Shiller Debate Has Taught Us" - link):
Fama and French chose to try to save [the Efficient Market Hypothesis] by jettisoning CAPM and replacing it with their own multi-factor risk model in which the small-stock effect and the value stock effect are said to “explain” price behavior (others have since added other factors). But it’s just as valid to hold on to CAPM and label these effects market inefficiencies.
According to Calma (2017), this is the most-cited paper of all time across ten leading financial journals. For a brief but very good summary of the historical threads leading up to it, see Davis (2001).
But note also that multifactor models had been used extensively since the 1970s, particularly by institutional investors. In a responsible investment context, see especially Rudd (1979). Luck and PIlotte (1993) used a commercial multifactor model to analyze the Domini Social Index in same year that this paper was published.
See also Grossman and Sharpe (1986).