The Constant Status of Beta Essay -- Business, Stock Returns, Investme

Beta being an efficient measure of riskiness of a security is an important financial instrument in investment decisions regarding estimation of market models, development of investment portfolios, estimation of cost of capital and emerging derivative markets. Since 1960s, the practical implication of CAPM has been in vague until recently challenged by some researchers Fama and French, (1992) for beta being insufficient in estimating the future returns of stocks based on historical data. The constant status of beta in estimating stock returns is questionable by the academic researchers for the high-low variation in its parameters. As systematic risk is time variant in nature, thus it is necessary to consider beta as a time series process withholding within it the stochastic behavior. Faff, Hillier and Hillier (2000) have demonstrated three modeling techniques to estimate time-varying beta namely multivariate generalized ARCH model (M-GARCH), another time varying heteroskedastic market models identified as EGARCH, TARCH and Kalman Filter approach. The dataset comprises of daily returns of 32 UK industry portfolios for the time period of January 1969 to April 1998. All the three models explain the time variability in systematic risk in the stocks. Of all the three models, Kalman Filter approach along with random-walk parameterization out-performs in efficiently explaining the stochastic behavior of betas in the daily stock returns of UK. The Kalman filter approach can be summed up as an optimal recursive computation of the least-squares algorithm. It is a subset of a Bayesian filter where the assumptions of a Gaussian distribution and that the current state is linearly dependant on the previous state are imposed. In oth... ...el errors while estimating these models using a Kalman filter algorithm has been considered and taken into account. The magnitude of reduction in the variance of market model errors that can be achieved if betas are allowed to vary is also measured. This gives a useful technique for comparing time-varying beta models with the constant beta model. This work also suggests that time-variation in beta is present in case of developing markets as in case of developed markets. The overall results have important implications for portfolio diversification and hedging strategies. Another implication of this academic research is that a dynamic hedging strategy, in which the hedge ratios are frequently adjusted in the light of the new information, will perform better compared to a static strategy where the hedge ratio is chosen at the beginning of the investment horizon.