Optimum Distribution-Free Tests and Further Evidence of Heteroscedasticity in the Market Model

In this paper several powerful distribution-free tests for heteroscedasticity are introduced and are used to test the hypothesis of constant variance in the market model. These tests are noted for their flexibility in specifying alternative hypotheses. It is found that the assumption of homoscedasticity is untenable for the majority of stocks analyzed. The implications of this finding for the efficient estimation of the parameters of the market model are also discussed.