Published online by Cambridge University Press: 05 October 2018
Using novel earnings calendar data, we show that firms’ advanced scheduling of earnings announcement dates foreshadows their earnings news. Firms that schedule later-than-expected announcement dates subsequently announce worse news than those scheduling earlier-than-expected announcement dates. Despite scheduling disclosures being observable weeks ahead of earnings announcements, we show that equity markets fail to reflect the information in these disclosures until the announcement itself. By also showing that option markets respond efficiently to volatility-timing information embedded in the same scheduling disclosures, we provide novel evidence that markets fail to react to information about future earnings despite investors immediately trading on the underlying signal.
We thank Jarrad Harford (the editor) and Jacob Thornock (the referee) for helpful feedback. We also thank Malcolm Baker, John Campbell, Wesley Chan, Asher Curtis, Ed deHaan, Lily Fang, Sam Hartzmark (Miami Behavioral Finance (MBF) Conference discussant), John Mahler, Dawn Matsumoto, Chris Noe, Mort Pincus, Rodrigo Verdi, Joseph Weber, and seminar participants at MIT, the 2014 Citigroup Quant Research Conference, NASDAQ Economic Research, Arrowstreet Capital, Ohio State University, Arcadian Asset Management, the University of Washington, the 2014 MBF Conference, and the 2015 Power of Events Conference for helpful comments and suggestions. We also thank Wall Street Horizon for providing daily earnings calendar data and Xiang (Nicole) Liu for excellent research assistance.