Forecasting Risk in Earnings

Abstract:     
Conventional measures of risk in earnings based on standard deviation are inadequate when the distribution of earnings deviates from normality. We propose a methodology based on current fundamentals and quantile regression to forecast risk reflected in the shape of the distribution of future earnings. We derive measures of downside risk, upside risk and uncertainty in future earnings. Our analysis shows that the level of current earnings is a significant determinant of the upper and lower quantiles of the distribution of future earnings. In-sample and out-of-sample forecasting performance improves significantly when we: (i) decompose earnings into accruals and cash flow; and (ii) allow forecasting coefficients to vary across profit and loss firms. We provide evidence that out-of-sample forecasts of quantile-based risk measures are related to a range of commonly used equity risk proxies and explain incrementally analysts’ forecast accuracy and analysts’ risk ratings. Our study provides insights into the relations between earnings components and risk in future earnings. It also proposes new risk measures that will be useful in developing understanding of the links between financial statement information and fundamental risk.

Number of Pages in PDF File: 36

Keywords: Earnings, accruals, fundamentals-based risk forecasts, quantile regression

JEL Classification: M41, C13

working papers series