Comment on "tests of scaling and universality of the distributions of trade size and share volume: evidence from three distinct markets".

In a recent publication, Plerou and Stanley [Phys. Rev. E 76, 046109 (2007)] use the Meerschaert-Scheffler estimator to verify the "inverse half-cubic law" of trade size distributions. We show that this procedure systematically underestimates these tail exponents.