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

The article presents a commentary on the paper "Optimal Distribution-Free Tests and Further Evidence of Heteroscedasticity in the Market Model," by Carmelo Giacotto and Mukhtar Ali. The authors point out a flaw in the analysis' statistical reasoning. They explain that uncorrelated random variables that are not normally distributed are not necessarily independent. This observation renders the paper's claims involving the distribution-free nature of their tests incorrect. The authors contend that it is not clear whether any of the test statistics possess well-defined distributions in large samples when based on recursive residuals.