Accruals quality and analyst forecast accuracy: evidence from the property-casualty insurance industry

This paper examines the association between property-casualty insurer accruals quality and analysts' earnings forecasts (i.e., accuracy and dispersion of forecasts). Using insurer-specific accruals, loss reserves, we calculate accruals quality which can be decomposed into its innate and discretionary components. Our results provide evidence that higher accruals quality - as measured by lower standard deviation of loss reserve errors - is positively associated with analysts' forecast accuracy. In other words, our results suggest that analysts provide less accurate forecasts for firms with higher reserve error volatility. Also, we show that lower accruals quality is associated with higher forecast dispersion indicating more disagreement among analysts. Our results hold consistent with decomposed components of accruals, innate and discretionary, and conclude that both managerial discretion and basic operations of firms affect insurers analysts' earnings forecasts.