Estimating and testing beta pricing models on industries

This paper uses the betas of book-to-market portfolios as proxies for systematic risks of industries instead of the individual betas computed from individual time-series regressions. Our empirical specification improves both the precision of the beta estimates and the cost of equity estimates. Estimating and testing beta pricing models via the proposed method highlights that consumption growth, liquidity risk, market excess returns, and the value factor explain the cross-sectional differences in expected industry returns, while there are no significant risk factors using the traditional approach. The fact that consumption risk is priced with monthly data is an interesting result, as the financial literature has struggled to prove that the consumption capital asset pricing model explains monthly returns.