Forecasting Taxes: New Evidence from Analysts

• In Event: Buy-side and Sell-side analysts forecasts

Abstract

We provide new evidence about how analysts incorporate and improve upon management forecasts, and how accounting standards limit the usefulness of certain management forecasts for predicting future earnings. Quarterly ETR reporting under the integral method provides a setting with mandatory point-estimate forecasts by management, but firms must record certain “discrete” tax items fully in the quarter they occur, which pollutes these forecasts. Recently available I/B/E/S reports of disaggregated analyst forecasts permit specific comparisons of the management estimate to an implied analyst ETR forecast. Using 133,989 analyst forecasts and managements’ interim ETR estimates for 39,751 firm-quarters from 2003 through 2012, we predict and find the following: 1) discrete items and complexity impair the accuracy of management’s ETR estimates; 2) analysts are more likely to mimic management’s ETR estimate when management is more accurate, there are no discrete items, they have lower ability, and they face less complexity; 3) analysts who form their own ETR estimate are more accurate when management is more accurate, there are no discrete items, and complexity is lower; 4) analysts’ ETR and EPS forecasts are less disperse when management’s estimate is more accurate and excludes discrete items. Together these results indicate that analysts exert effort to try to understand complex tax settings, in contrast with prior researchers’ conclusions that analysts ignored or were too confused by tax matters. Further, we suggest that standard-setters re-examine the discrete item exception to the integral method, based on our evidence that discrete item reporting impairs the predictive information in quarterly ETRs.

Authors

• Brian Bratten, University of Kentucky

• Cristi A Gleason, University of Iowa

• Lillian F Mills, University of Texas at Austin

• Stephannie A Larocque, University of Notre Dame