3 Semiparametric methods for asset pricing models

Abstract

This paper discusses semiparametric estimation procedures for asset pricing models within the generalized method of moments (GMM) framework. GMM is widely applied in the asset pricing context in its unconditional form but the conditional mean restrictions implied by asset pricing theory are seldom fully exploited. The purpose of this paper is to take some modest steps toward removing these impediments. The nature of efficient GMM estimation is cast in a language familiar to financial economists: the language of maximum correlation or optimal hedge portfolios. Similarly, a family of beta pricing models provides a natural setting for identifying the sources of efficiency gains in asset pricing applications. My hope is that this modest contribution will facilitate more routine exploitation of attainable efficiency gains.