Sins of Omission in Value Relevance Empirical Studies

We contribute to the value relevance literature by investigating critical methodological deficiencies emerged in past and current empirical research. Using Monte Carlo simulations calibrated on the basis of the statistical properties of market and accounting data for a large sample of European listed companies, we are the first to document and quantify the effects of neglecting the lag of stock price as an explanatory variable in the conventional approach for estimating price level regressions. We demonstrate that for European listed companies this is an important source of omitted variable bias and the extent of such bias increases as the autocorrelation coefficient for stock price and the explanatory variables increases. We show that using alternative specifications which deflate the accounting variables by the lag of stock price, commonly employed in the accounting literature, can lead to high over-rejection rates. Our findings are relevant for the interpretation of most of the empirical studies on the impact of IFRS on value relevance in Europe.