The existence of heteroscedasticity and its effect on estimates of the market model parameters

A number of recent studies have found significant evidence of heteroscedasticity in the market model. These studies have only documented the existence of heteroscedasticity and have not determined the actual impact of this econometric aberration in the estimation process. This study develops a more general specification to determine the effects of heteroscedasticity on the market model parameters. The consequences of this econometric problem in relation to efficiency‐bias, temporal stability of betas, and estimates of systematic risk are revealed in the estimation of a maximum likelihood specification.