Optimal Distribution-Free Tests and Further Evidence of Heteroscedasticity in the Market Model: A Reply