Measuring Value Relevance in a (Possibly) Inefficient Market

Abstract

An interesting question in assessing value relevance of accounting numbers is whether measures of value relevance are materially affected by market inefficiencies. We explore this question in the case of price level regressions in two parts: First, we derive a procedure for estimating value relevance coefficients from a model of an inefficient stock market in which stock price contains measurement error that may be correlated with accounting variables. Second, we apply this procedure to estimate value relevance coefficients of earnings and book values. A longitudinal comparison of value relevance coefficient estimates from conventional price level regressions with estimates from our procedure finds statistically significant differences, though not large enough in magnitude to alter conclusions of earlier studies that have examined the pattern of changes in value relevance of earnings and book value over a similar time frame. Further results speak to an interdependency between market efficiency and value relevance not previously reported. Overall, the empirical findings accord well with the intuition that market inefficiencies with respect to earnings and book values must affect the measurement of value relevance, but that arbitrage mitigates the magnitudes of those effects.

JEL Classification: M41, G12, G14