Predicting Market Betas

Abstract

This research, using firm-level data on NSE listed stocks for the period of 2008–2015, seeks empirical evidence on the significance of company characteristics in prediction of market betas in Indian context. The empirical methodology involves two-stage statistical procedure where an exploratory factor framework determines the historical relationship between firms’ fundamentals and systematic risk at first, and a generalized least square (GLS) technique of panel regression modelling estimates their predictive power thereafter. The search process discovers debt ratios and returns on investment predict market betas in Indian stock market. The study strongly suggests that the use of 1-year lagged historical betas along with the measures of financial leverage and overall profitability produces better predictions than the benchmark predictions solely based on historical betas. However, the analysts need to consider the industry differences in systematic risk to enhance the predictive power of the model.