A Model for Stock Return Distribution

The Laplace mixture distribution for stock share returns is derived from conditional N(0,x2) distribution. The conditioning variable, x2, is assumed to be an exponentially distributed random variable. This offers a natural stochastic interpretation of the risk involved with the stock share. Maximum likelihood estimates for returns of the twenty most traded shares and the aggregate index of the Helsinki stock market in late 1980's do not reject the Laplace distriburion model. The results extend to returns over longer periods than one day.