Stock Price Associations with Expected and Unexpected Earnings

Abstract

Prior research fails to explain why theoretically predicted magnitudes of earnings response coefficients (ERCs) and price-to-earnings ratios (P/E ratios) differ from their empirical counterparts. Therefore, we (re)investigate the associations between earnings innovations, persistence of expected earnings, investors’ expectation revisions, and the magnitudes of P/E ratios and ERCs, respectively. In doing so, we compare the traditional price model with our extended price model, which is based on a new model of investors’ expectation formation, simultaneously incorporating the following assumptions: (i) expected earnings are less than purely persistent; (ii) only the value-relevant fraction of an earnings innovation triggers future expectation revisions; (iii) expected earnings’ time series is nonstationary but integrated of order 1; and (iv) the persistence of expected earnings is negatively correlated with the value-relevant magnitude of earnings innovations, i.e. high persistence in expected earnings requires a low level of expectation revisions. We first show on a theoretical level that under these assumptions: (a) P/E ratios and ERCs differ in their magnitudes and it thus becomes necessary to distinguish between these two constructs in modelling price-earnings relations; (b) ERC magnitudes are negatively associated with the persistence of expected earnings, contradicting both theoretical implications of the traditional price model which predicts a positive association as well as empirical evidence; (c) expected magnitudes of P/E ratios (ERCs) range between 10-60 (0-7). We second test our model empirically. The empirical results are consistent with our predictions, closing the gap between theoretically predicted and empirically estimated P/E ratio and ERC magnitudes, respectively.