The Impact of Earnings Management on the Performance of Earnings-Based Valuation Models

Abstract

The purpose of this study is to examine empirically how the presence of earnings management may affect firm valuation. We compare the performance of earnings-based (e.g., Residual Income Model, RIM) and non-earnings-based (e.g., Discounted Cash Flow, DCF) valuation models, measured by absolute percentage pricing errors and absolute percentage valuation errors, for two subsets of publicly traded US firms: Suspect firms that are likely to have engaged in earnings management and “Normal” firms matched on industry, year and size. When valuation models use only analysts' short-term earnings forecasts as model inputs, results indicate that earnings management adversely affects the RIM model’s ability to estimate a firm’s intrinsic value while leaving that of DCF unchanged. We contribute to the valuation literature by showing that the well-known superiority of the RIM model over DCF does not hold when earnings are managed. By comparison, if the valuation model also includes analysts' long-term target price forecasts, RIM does not enjoy any economically significant accuracy advantage over DCF, with or without the presence of earnings management. Over a longer forecast horizon, financial analysts appear to account for the impact of earnings management on firms’ future values by adjusting their target price forecasts. We extend the earnings management literature by demonstrating that the way analysts react to earnings management over short to long-term forecast horizons has different implications for the estimation ability of RIM vis-a-vis DCF models.