Zero-Value Company Returns, Thin Trading and the Use of State Asset Pricing Models in Event Study Research

Abstract

The calculation of expected returns is a necessary ingredient in data processing for an event study. The method most commonly used, the market model, often fails to meet the OLS requirement of normally distributed residuals, and tends to furnish regression output (low R2, and insignificant t- and F-statistics) that, in other contexts, one would not rely on. A family of state asset pricing models may offer improved performance in this respect. This issue becomes important when a listed company’s stocks are thinly traded and missing data is proxied by zero-value returns whose rate of occurrence impacts on the ability of the market model to cope. A 3-state asset pricing model has superior performance characteristics when applied to thinly-traded data sets.