If the Sharpe-Lintner-Black CAPM is a model of expected returns, why even test it using realized returns? Are there ways to estimate ex ante returns?

Replacing the vector of expected returns with realized returns [historical] may be convenient but often results in the SML slope being flat or even negative. Roll (1977) and Fama and French (1992)...

22 October 2013 1,277 14 View