Odd-Lot Trading in the Stock Market and Its Market Impact: A Reply

Professor Stevenson in his comment of my recent paper on odd-lot trading introduces some interesting additional odd-lot data for both the NYSE and the ASE. His criticisms, however, in my opinion are without merit. First, in my paper I state clearly that due to limited financial resources the study concentrates only on the NYSE. Therefore, the odd-lot participation rate on the ASE is not calculated. Although Stevenson's ASE data are interesting, I fail to see how these ASE data raise serious doubts as to whether my statistical significant results between NYSE odd-lot purchases and sales and ASE odd-lot purchases and sales have any meaningful interpretation. Secondly, Stevenson quarrels with my observation in regard to odd-lot short sales that “odd-lotters have become somewhat more speculative and perhaps more sophisticated in the 1960's.†Again I fail to see why a shifting of marketplaces by small investors, which he mentions but does not prove, makes my statement less valid. Thirdly, because of the nature of the distributed lag model, I do not see why I cannot conclude that “results suggest that odd-lot and round-lot volume did affect each other.†Of course his interpretation that odd-lot and round-lot volume tend to move together is also valid. Finally, my data are through 1967, and the two odd-lot dealer firms did not merge until 1970. In short, as Stevenson's criticisms are all minor, he is making a mountain out of a molehill.