1995 Reports

The CAPM with human capital: Evidence from Japan

Jagannathan, Ravi; Kubota, Keiichi; Takehara, Hitoshi

We find that the CAPM beta has little ability to explain the cross section of average returns on Japanese stocks when the TOPIX return is used as the proxy for the return on the aggregate wealth portfolio. The relation between average return and beta is flat, with an R-Square of less than 3%. The empirical performance of the CAPM improves substantially if a measure of the return on human capital is also included when measuring the return on the aggregate wealth portfolio -- the R-Square rises to 75%. There is little evidence against the CAPM specification with this modification. It performs almost as well as the linear factor model which, following Fama and French, uses the TOPIX return and the payoffs on portfolios that capture the size and book to price effects found in the data as the three factors.