On Bankruptcy Indicators From Analysts’ Earnings Forecasts

Abstract

Research has documented that forecasts of future earnings made by financial analysts capture aspects of risk and expected return for firms, and reflect information not reflected in firms’ accounting reports. This study investigates the usefulness of measures developed from analysts’ earnings forecasts as indicators of impending bankruptcy. Results indicate that healthy and failing firms differ significantly with respect to four properties of earnings forecasts: forecast level, forecast error, forecast bias, and forecast dispersion. When combined in a model, measures related to these four forecast properties are able to predict bankruptcy in the year prior to failure as accurately as more traditional approaches using accounting ratios.