Experimental Tests of the Separation Theorem and the Capital Asset Pricing Model

Abstract

A computer-controlled portfolio selection task with three risky assets and either with or without a riskless asset was devised to test experimentally assumptions underlying the separation theorem and the capital asset pricing model. Two differently paid groups of subjects completed individually up to 300 portfolio selection problems. Although most of the subjects diversified among the risky assets, the introduction of a riskless asset did not have the effect predicted by the separation theorem, nor were the subjects affected by systematic changes in the variance-covariance matrix governing the risky returns. However, performance improved as the reward was increased tenfold. Copyright 1988 by American Economic Association.