Beneish M-Score: Detecting Earnings Manipulation
Beneish M-Score (Earnings Manipulation Detection) · Also known as: Beneish Model, M-Score Model, Earnings Manipulation Score, Beneish M-Skoru
The Beneish M-Score is a statistical model developed by Messod Beneish in 1999 to identify whether a company has manipulated its reported earnings. The model combines eight financial-statement ratios into a single composite score using coefficients estimated from a probit regression on a sample of detected earnings manipulators. A score above −2.22 indicates a heightened probability of manipulation, making the M-Score a widely used tool in forensic accounting and investment due-diligence.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
The Beneish M-Score is appropriate when an analyst needs a rapid, quantitative screen for earnings manipulation risk using publicly available financial statements. It assumes the firm prepares GAAP-compliant annual reports and that two consecutive periods of data are available. The model was estimated on US firms and may require recalibration for other regulatory environments. It performs best as a triage screen rather than a definitive judgment; flagged firms warrant deeper forensic investigation. Alternatives include the Altman Z-Score (for distress) and the Dechow F-Score.
Strengths & limitations
- Requires only publicly available financial-statement data, making it practical and low-cost to implement.
- Provides a single composite score that can rank firms by manipulation risk, facilitating portfolio-wide screening.
- Empirically validated: Beneish (1999) reports roughly 76% accuracy in identifying manipulators in-sample.
- Interpretable coefficients allow analysts to identify which specific accounting dimensions drive the elevated score.
- Calibrated on US GAAP firms from the 1980s–1990s; external validity to IFRS jurisdictions or more recent periods is uncertain.
- High false-positive rate in practice: most firms flagged as likely manipulators are not actually manipulating earnings.
- Cannot identify the mechanism or magnitude of manipulation — only an aggregate risk signal.
- Backward-looking by design; detected only after at least one full fiscal year of manipulated data has been reported.
Frequently asked
What does an M-Score of −2.22 mean exactly?
It is the probit-model threshold below which a firm is classified as a non-manipulator. The value was chosen by Beneish to balance Type I errors (falsely accusing clean firms) and Type II errors (missing actual manipulators) in his estimation sample. Scores above −2.22 indicate higher manipulation probability, but the threshold can be adjusted based on the analyst's cost tolerance for each error type.
Can the Beneish M-Score be applied to non-US companies?
The model was estimated on US GAAP firms, and applying it to IFRS-reporting or emerging-market firms requires caution. Differences in accounting standards, depreciation conventions, and accrual measurement can systematically shift index values. Some researchers have re-estimated the coefficients on local samples with reasonable results, but any cross-jurisdiction application should be treated as exploratory until locally validated.
How is the Beneish M-Score different from the Altman Z-Score?
The two models address distinct risks. The Altman Z-Score predicts financial distress and bankruptcy using leverage, liquidity, and profitability ratios. The Beneish M-Score detects earnings manipulation using accrual and ratio-change signals. A firm can score poorly on both, but the underlying phenomena — insolvency risk versus reporting fraud — are conceptually separate and require different analytical responses.
Sources
- Beneish, M. D. (1999). The detection of earnings manipulation. Financial Analysts Journal, 55(5), 24–36. DOI: 10.2469/faj.v55.n5.2296 ↗
How to cite this page
ScholarGate. (2026, June 2). Beneish M-Score (Earnings Manipulation Detection). ScholarGate. https://scholargate.app/en/finance/beneish-m-score
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Altman Z-ScoreFinance↔ compare
- DuPont AnalysisFinance↔ compare
- Logistic RegressionResearch Statistics↔ compare