Multivariate testing of the capital asset pricing model in the Hong Kong stock market

The paper provides a multivariate testing of the Sharpe-Lintner and Black Capital Asset Pricing Model (CAPM) for the Hong Kong stock market. Adopting the multivariate approach has many advantages, such as elimination of the error in variables problem encountered in the univariate approach as well as provision of an efficiency gain in the parameter estimates. The estimation results reject both versions of CAPM. The source of rejection seems to come from the firm size effect as it is found that small firms (which also have small betas) have earned returns higher than the CAPM has predicted. Furthermore, the estimated return of the zero beta portfolio in the Black CAPM is found to be statistically insignificant.