Price and volume in the Tokyo stock exchange : an exploratory study

stable law with the Student's t distribution.Both hypotheses were derived from the framework of subordinated stochastic processes.In particular, if the variance of a normal random variable follows an inverted gamma distribution, then the posterior distribution is a Student's t.The empirical results of Blattberg and Gonedes sup- ported the Student's t distribution in preference to the stable Paretian model.Recently Kon (1984) considered a discrete mixture of normal distribution (MND) as an explanation for the observed excess kurtosis and positive skewness of stock returns in the U.S. market.He estimated MND models for the daily returns of the 30 stocks of the Dow Jones Industrial Index and three market indices.By fitting MND models with orders up to five, he argued that the MND is considerably more descriptive of the data generating process than the simple normal model and the Student's t distribution.This finding is congruent with a time-varying nonstationary return process.Periodicity anoma- lies and varying financial and operating leverages were offered as