Source and method
M. D. Beneish, “The Detection of Earnings Manipulation,” Financial Analysts Journal 55(5), 1999, built the model by comparing firms that were found to have manipulated earnings with firms that were not. The eight indices compare the current year with the prior year for receivables, gross margin, asset quality, sales growth, depreciation, SG&A, leverage and accruals. Coefficients and the cutoff are those cited above.
When to use it
Run it in planning as an analytical procedure that feeds the fraud risk discussion for a client with unusual revenue growth or accrual build, or on a customer or investee you are evaluating. Always read the table: if one index is doing most of the work, go to the underlying account and ask what changed, such as a new customer segment, longer terms, or a capitalized cost.
Limits
This is a screening statistic. It does not identify the manipulation, does not replace the audit procedures on revenue recognition and management override, and says nothing about misappropriation of assets. TATA is computed here as income from continuing operations less operating cash flow, over total assets, a common simplification of the balance-sheet accrual measure in the paper. Use consistent definitions across both years and do not mix restated and unrestated figures.
Illustrative figures only. Check the result against your firm's methodology before relying on it.
Common questions
What is the Beneish M-score?
A probit model published by Messod Beneish in 1999 (Financial Analysts Journal 55(5)) that combines eight financial statement indices into a score for how much a company’s statements resemble those of earnings manipulators. A higher (less negative) score points to higher probability.
What is the formula?
M = −4.84 + 0.920 DSRI + 0.528 GMI + 0.404 AQI + 0.892 SGI + 0.115 DEPI − 0.172 SGAI + 4.679 TATA − 0.327 LVGI. DSRI is the receivables-to-sales index, GMI gross margin, AQI asset quality, SGI sales growth, DEPI depreciation, SGAI SG&A, TATA total accruals to total assets and LVGI leverage.
What cutoff should I use?
This tool uses −1.78, the figure commonly given with the eight-variable model; a score above it flags. Other cutoffs trade false positives against false negatives, and the right trade-off depends on how you use the score.
How does this help with the fraud risk assessment?
AU-C 240 requires the auditor to assess fraud risks, including a presumed risk of improper revenue recognition, and analytical procedures during planning are one input. A high DSRI or SGI with a high TATA points at revenue and receivables. The score tells you where to look; it is not an audit conclusion.
When is the model unreliable?
Financial institutions, start-ups with no prior-year base, companies with large acquisitions or a change in accounting policy between the two years, and shrinking or cash-basis businesses distort the indices. The coefficients were estimated in 1999 and are not updated. A high score on a rapidly growing, honest company is common.
Runs in your browser. The figures you type are not sent to CORAA or stored. Check the result against your firm's methodology before relying on it.