Bridging the Gap between Value Relevance and Information Content

Three main approaches for examining the effect of accounting information in financial markets have emerged in the last three decades. We formally define information content, valuation relevance, and value relevance. Respectively, these approaches are based on Beaver (1968), Ball and Brown (1968), and tests of association between market values and accounting numbers. We systematically compare and contrast these paradigms to highlight their relative strengths and weaknesses, and to identify possible reasons why one method provides different results from another. Implementing these research methods using data from the last three decades, we show that information content has remained constant while valuation relevance and value relevance have both declined. We find that our conclusions are robust with respect to return volatility, non-linearities of the valuation model used, earnings composition, and the earnings expectation model used. We conclude that the unrecognized disclosures released concurrently with earning sare likely to have contributed to this divergence