코스닥시장에서의 사전적 효율성 및 CAPM 검증

In this study, we test the standard CAPM with the return data of the firms listed in KOSDAQ which has recently shown high volatility. The likelihood ratio test is employed to test the null hypothesis that the value-weighted index(VWI) or the equally-weighted index(EWI) is efficient ex ante. We also employ the nonlinear seemingly unrelated regression model to find out whether the standard CAPM explains stock returns properly as it theoretically implies. Our sample includes 138 monthly return data in the period of January 1999 to June 2010. One hundred ninety sample firms are classified into one of ten portfolios by industry, firm size, and beta. Important empirical findings of this study can be summarized as follows: (1)When portfolios are classified by industry, the null hypothesis that VWI or EWI is efficient cannot be rejected. However, for portfolios grouped by either firm size or beta, only EWI has efficiency. (2)The test of the CAPM shows that EWI is preferred to VWI as a proxy for the market portfolio because the former results in better parameter estimates and higher explanatory power. (3)In addition to LRT, GMM is also employed for robustness test and the results based on GMM also indicate that EWI is efficient while VWI is inefficient. In sum, this study suggest the use of EWI as a proper proxy for the market portfolio when testing or applying the CAPM.