Analysts’ GAAP Earnings Forecasts and Their Implications for Accounting Research
52 Pages Posted: 25 May 2014 Last revised: 18 Jan 2018
Date Written: December 1, 2017
We examine analysts’ GAAP earnings forecasts and illustrate their usefulness in two prominent research settings. First, we find that the availability of GAAP forecasts has increased dramatically since 2003, and they are now available for most I/B/E/S-covered firms. Next, we utilize GAAP forecasts to solve important measurement error problems in prior research that attempts to examine GAAP forecast errors without an explicit GAAP forecast. We begin with research exploring investors’ preferences for GAAP versus non-GAAP earnings. We find that traditionally-identified GAAP forecast errors are subject to 37% measurement error. Nevertheless, in contrast to the pervasive caveats in the non-GAAP literature, evidence of an investor preference for non-GAAP earnings relative to GAAP earnings is robust after correcting for this measurement error. Second, we revisit the literature identifying firms that use non-GAAP exclusions to meet or beat analysts’ forecasts when GAAP earnings fall short of expectations. We find that 34% of the traditionally-identified meet-or-beat firms are misidentified due to measurement error and this misidentification masks the inference that firms more frequently exclude transitory expenses rather than recurring expenses for meet-or-beat purposes.
Keywords: Measurement error; Earnings expectations; Analysts’ forecasts; GAAP forecasts; Non-GAAP earnings; Street earnings; Earnings surprise; Meet-or-beat
JEL Classification: G14; M40; M41
Suggested Citation: Suggested Citation