GMM-Based Tests of Efficient Market Learning and an Application to Testing for a Small Firm Effect in Equity Pricing

Abstract

In this paper, we extend Bossaerts’ (2004) analysis of the implications of the efficient learning market hypothesis (ELM) for asset prices by reformulating it in a GMM setting. Our representation is more amenable to widespread application and allows the econometrician, in testing ELM, to make use of the full range of specification tests that have been developed by the empirical literature in the context of tests of the more restrictive Efficient Market Hypothesis (EMH). We apply this framework to test for efficient learning in the pricing of small capitalization stocks. We find evidence of mispricing of small stocks but we cannot rule out that, in spite of possibly incorrect priors about the future payoffs of small firms, the market efficiently processes information as it becomes available over time. That is, our evidence contradicts the Efficient Market Hypothesis (EMH) but it is not incompatible with efficient learning in the manner of Bossaerts (2004).