The Predictable Cost of Earnings Manipulation

51 Pages Posted: 17 Aug 2007  

Messod Daniel Beneish

Indiana University - Kelley School of Business - Department of Accounting

D. Craig Nichols

Cornell University - Samuel Curtis Johnson Graduate School of Management

Date Written: August 13, 2007

Abstract

Although financial reporting fraud generates considerable losses, we find that investors do not fully exploit publicly available information relevant for detecting fraud. We show that firms with a high probability of overstated earnings have lower future earnings, less persistent income-increasing accruals, and lower future returns. The trading strategy based on the probability of manipulation ranks subsumes the relation between accruals and future performance and yields a hedge return of 13.9%, mostly arising from the short position. Although this suggests a limits-to-arbitrage explanation, we show that institutional investors actually increase their holdings in firms with a high probability of manipulation, and that hedge returns remain large for firms with market capitalization in excess of $1 billion. Thus, the returns concentrated on the short side of the strategy appear to arise not from asymmetric arbitrage costs, but from asymmetric errors in the expectations of (even sophisticated) market participants.

Keywords: Accrual Mispricing, Earnings Management, Asset pricing

JEL Classification: M41, M43, G11, G14

Suggested Citation

Beneish, Messod Daniel and Nichols, D. Craig, The Predictable Cost of Earnings Manipulation (August 13, 2007). Available at SSRN: https://ssrn.com/abstract=1006840 or http://dx.doi.org/10.2139/ssrn.1006840

Messod Daniel Beneish (Contact Author)

Indiana University - Kelley School of Business - Department of Accounting ( email )

1309 E. 10th Street
Bloomington, IN 47405
United States
812-855-2628 (Phone)
812-855-4985 (Fax)

D. Craig Nichols

Cornell University - Samuel Curtis Johnson Graduate School of Management ( email )

Ithaca, NY 14853
United States

Paper statistics

Downloads
2,439
Rank
3,887
Abstract Views
6,167