Rationalizing Firm Forecasts
We conduct a five-year, ten-wave panel survey in collaboration with a large U.S. payments company, collecting incentivized sales forecasts from over 6,000 firms. The sample is representative of the size distribution of U.S. employer firms. We find that firms exhibit remarkably poor forecasting accuracy, with bias, predictable errors, and over-precision, particularly in smaller firms. To explore the sources of forecast inaccuracy we implement four randomized controlled trials: improving data use, increasing incentives, providing training, and encouraging contingent thinking. While enhanced data use and incentives lead to modest improvements in contemporaneous forecast accuracy, training and contingent thinking yield no measurable effects. None of the interventions improve forecasting accuracy in subsequent waves. Surveying economists, we find they predicted about half the inaccuracy and bias we found, highlighting the surprisingly large deviation of firms from rational expectations.
-
-
Copy CitationNicholas Bloom, Mihai A. Codreanu, and Robert A. Fletcher, "Rationalizing Firm Forecasts," NBER Working Paper 33384 (2025), https://doi.org/10.3386/w33384.Download Citation
-
-
Related
Topics
Programs
More from the NBER
In addition to working papers, the NBER disseminates affiliates’ latest findings through a range of free periodicals — the NBER Reporter, the NBER Digest, the Bulletin on Health, and the Bulletin on Entrepreneurship — as well as online conference reports, video lectures, and interviews.
