Anomalies in CAPM : A Panel Data Analysis Under Indian Conditions

Despite more than a decade of reforms, Indian stock markets have not been able to mobilize and allocate capital effectively. While Indian corporate sector depends predominantly on debt to raise finance, the Indian household depends primarily on banks to invest its savings. A possible explanation for such a phenomenon can be inaccurate pricing of assets in the stock market. One of the widely used asset-pricing models is the CAPM. The present paper tries to analyse the relevance of factors other than beta that affect asset returns in the Indian stock markets. We improve upon the methodology of Fama and MacBeth (FM) cross-sectional regression, generally followed in CAPM anomalies literature. One reason why the FM method is widely applied is because it allows for time variation in estimated coefficients. However, due to the averaging of coefficients across time, the FM method will accept the null hypothesis of no relation between beta and stock returns even when the underlying model is true. To overcome this limitation we use LSDV technique. The LSDV method, while incorporating time variation also helps us overcome the limitation of averaging of coefficients. In conformity with prior literature, our results also indicate presence of size and value effect