Stock price discreteness and clustering: decimals and ordered probit model.

Purpose – The purpose of this paper is to revisit the ordered probit model of Hausman et al. after the NYSE decimalization. Design/methodology/approach –The changed ordered probit model. Findings - The model can somewhat capture the different impact of trading-related “explanatory” variables on price changes among three different decimals but does not explain much about price discreteness and irregular transaction intervals among the existing models of stock price discreteness. Overall 1/16th and 1/24th range of the dependent variable is better explained by trading-related explanatory variables than 1/8th range of the dependent variable for small firms and there is not much difference in large firms among three decimals. The results imply that finer specification in decimalization and smaller firm size matters in trading after the decimalization project. Originality/value - First paper to revisit the ordered probit model of Hausman et al. after the NYSE decimalization.