Nudging Towards Better Earnings Forecasts

Abstract

We examine whether “nudging” forecasters to reduce their herding behavior without altering their economic incentives or limiting their choices improves the consensus earnings forecast. Specifically, we examine how introducing a social-norm nudge that promotes exerting one’s best forecasting effort on the Estimize.com forecasting platform impacts the quality of its consensus earnings forecast. Consistent with behavioral economic theory’s predictions, we show nudging reduces forecast herding and leads to a consensus earnings forecast that is less biased, more accurate, and more representative of investors’ earnings expectations. We show that the nudge confers net benefits to the consensus forecast via changes in forecaster behavior.

Keywords: nudging, social norms, earnings forecasts, analysts, consensus forecast, market reaction, crowdsourcing

JEL Classification: D91, D26, G41, G14, M41