We examine whether management earnings forecast errors exhibit serial correlation and how analysts understand the serial correlation property of management forecast errors (MFEs). MFEs should not exhibit serial correlation if managers efficiently process information in prior forecast errors and truthfully convey their earnings expectations through management forecasts. However, for long‐horizon management forecasts of annual earnings, we find significantly positive serial correlation in MFEs, and sample self‐selection does not seem to drive this phenomenon. Further analyses suggest that managers’ unintentional information processing bias contributes to this positive serial correlation. Analysts anticipate the intertemporal persistence of MFEs but underestimate the persistence level when reacting to management forecasts. Our findings have implications for market participants who rely on management forecasts to form earnings expectations, and also shed light on the efficiency of managerial decision making.
Purpose The purpose of this paper is to introduce analyst estimates and option pricing-based variables in modeling material accounting misstatements. Design/methodology/approach The paper uses a logistic regression model to analyze a comprehensive sample of AAER and non-AAER firms listed in the USA. Findings By applying a cross-sectional, sequence of time-series logistic regression models, the authors find better identifiers of ex ante risk of fraud than prediction models based on an inspection of abnormal accruals. These identifiers include the managed earnings (ME) component of a firm and the change in a firm’s option contracts’ implied volatility (IV) prior to an earnings announcement. Practical implications The empirical findings contribute to an understanding of earnings manipulation (fraud) and should be of value to auditors and regulatory bodies interested in identifying financial statement fraud, particularly the Securities and Exchange Commission, which has been improving its accounting quality model (AQM or Robocop) fraud detection tool for many years. The results contribute substantially to enhancing the current accounting literature by introducing two non-accrual-based measures that significantly enhance the predictive power of an accrual-based accounting misstatement prediction model. Originality/value This paper radically departs from relying on the assumption that the clearest and easiest pathway to detect fraud reporting ex ante is through an examination of accruals. Instead, the authors use a richer source of information about the possibility of a firm’s misstatement of its financial accounting numbers, namely, analyst estimates of ex post earnings and the IV from exchange-traded option contracts.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.