RELATIVE PRICES, REAL ACTIVITY AND THE VALUATION OF SHARES

PETER JOHN ELMER, Purdue University

Abstract

A variety of empirical studies have documented both a negative correlation between real equity returns and inflation and a significant decline in Tobin's "q" during the past decade and a half. These empirical findings have elicited explanations emphasizing the role of taxes, valuation errors and real activity. This study extends the share price debate in a three-stage process. First, the relation between one index of real activity, industrial production, and real equity returns is explored. We find that industrial production explains the cyclical movement of real equity returns, but not the trend. We discuss why cyclical factors may be viewed as exogenous with respect to other fundamental determinants of real equity returns. Given the exogeneity of cyclical factors, a need arises to specify forms of real activity that can be used to explain real equity return trend. Second, a link is proposed between one form of real activity, relative factor prices, and share values. Microeconomic principles of the theory of the firm are used to show that the value of outstanding (vintage) capital declines when relative factor prices change unexpectedly. Thus, share prices are shown to be a negative function of the extent to which current relative factor price expectations differ from those of previous periods. A relative factor price variability index is estimated to determine the empirical content of the proposed "relative price" hypothesis. Finally, the relative merits of alternative share price hypotheses are tested. Initially, a rational valuation model, based on the index of factor price variability, is shown to explain share prices better than models that assume either of two valuation errors. Alternative hypotheses are tested by adding representative variables to the initial rational specification. Variables representing the impact of personal taxes, risk and industrial production are included in the test. Test results suggest that a four factor share price model is most consistent with prior expectations. The factors include a profit variable, a real discount, a personal tax term and an index of relative factor price variability. Poor share performance is largely explained by increased factor price variability, and a higher real discount.

Degree

Ph.D.

Subject Area

Finance

COinS