Dynamic component detection in a multifactor model for stock returns

Summary In this paper the factorial structure of some asset returns quoted at the Milan stock exchange is analyzed in order to detect the presence of a dynamic component. The maximum likelihood estimates of the dynamic model, for which a space-state representation and the Kalman filter were used, are compared with the estimates of the static model via the information criteria. There is no evidence of dynamic factors underlying the analyzed samples of asset returns, while one static factor seems to be relevant. Invited paper at the Conference held in Bologna, Italy, 27–28 May 1993, on ?Statistical Tests: Methodology and Econometric Applications?.