Heaping on Dividends : The Role of Dividend Size and Information Uncertainty

In this paper we contribute to a line of literature that examines behavioral biases that impact important corporate decisions. Our paper builds on prior articles that examine heaping or rounding of EPS forecasts. Herrmann and Thomas (2005), Bamber, Hui and Yeung (2010) and Dechow and You (2012) observe that firm managers and analysts systematically round EPS forecasts. They also find more rounding for EPS forecasts with larger magnitude, and when management or analysts face higher levels of information uncertainty. In this study we examine whether dividend policy is also affected by rounding and the heaping heuristic. The previous literature that examines heaping related to corporate earnings focuses on heaping in earnings forecasts, however our study focuses on heaping in actual distributed dividends that directly impact key firm characteristics such as firm capital structure, investor returns and overall firm performance. From 1926 through 2015 we report 51.40% of dividends are heaped. In the earlier years, dividends are almost always heaped, but heaping declines through time. In some years 100% of quarterly dividends are heaped. We find heaping is significantly related to dividend size, and the level of information uncertainty faced by firm management. Of the uncertainty metrics, inverted firm size has the strongest relation with the likelihood of heaping. We also observe that heaped dividends are altered less frequently, and when there is a dividend change to a previously heaped dividend it is more often an increase or decrease to another heaped amount