More Evidence on the Nature of the Distribution of Security Returns

Abstract

finance has assumed the existence of the second moment. It is, therefore, impor? tant to determine whether or not the security return distribution actually has a finite variance. In this paper, we develop a simple, yet interesting, methodology for testing finite versus infinite population variance. The results obtained indicate that the likely cause of the empirically-observed fat tails of security return distribu? tions is not an infinite variance but a finite variance that changes in a complex fashion over time. One implication of this result is that more effort should be directed towards understanding those economic factors that influence security re? turn variability. The rest of the paper is structured as follows: Section II contains a brief description of previous work; Section III describes the methodology which we employ; Section IV reports the simulation results for this methodology; Section V describes the data used; Section VI reports and discusses the results; and Sec? tion VII is a brief summary.