Machine Learning Improves Accounting Estimates
39 Pages Posted: 12 Oct 2018 Last revised: 20 May 2019
Date Written: May 16, 2019
Managerial estimates are ubiquitous in accounting: most balance sheet and income statement items are based on estimates; some, such as the pension and employee stock options expenses, on multiple estimates. These estimates are affected by estimation errors as well as by managerial manipulation, thereby adversely affecting the reliability and relevance of financial reports. We show in this study that machine learning can improve managerial estimates. Specifically, using insurance companies’ data on loss reserves (future claims) estimates and realizations, we document that the estimates generated by machine learning were superior to managerial estimates reported in financial statements in four out of five insurance lines of business examined. Our evidence suggests that machine learning techniques can be highly useful to managers and auditors in improving accounting estimates, thereby enhancing the usefulness of financial information to investors.
Keywords: machine learning, accounting estimates
Suggested Citation: Suggested Citation