Investment Models and Expectations: Some Estimates for the Italian Industrial Sector

Abstract

The essential characteristic of investment decisions is that they are inherently forward-looking as they depend upon the expectations of future values of the relevant variables. In the specification of investment functions it is crucial, therefore, to start from a model of a firm's behavior that provides a rationale for the importance of future events. At the same time, in empirical work, one must pay attention to the choice of appropriate proxies for the expectational variables. In this paper we shall examine the effects on investment expenditure of expected output and factor prices starting from a putty-clay model of the firm and using aggregate data on the industrial sector for Italy. In Section 2 the implications for the specification of the investment function of the assumption of a putty-clay technology and of the existence of delivery lag will be summarized, following Nickell [1979]. The discussion will be extended here to the case of quadratic adjustment costs. In Section 3 the problems involved in obtaining a proxy for expected output are discussed. The method used here is based on the theory of optimal predictors/rational expectations within a multivariate time series framework. In Sections 4 and 5 the estimates of the invest