The Beneish M-Score
Eight ratios, one cut-off, and a number that is far more useful when you can see what moved it.
What it measures
The M-Score combines eight ratios, each comparing the current year against the prior year. Together they describe whether a company’s reported earnings are behaving the way manipulated earnings tend to behave.
The eight are: days sales in receivables (DSRI), gross margin (GMI), asset quality (AQI), sales growth (SGI), depreciation (DEPI), SG&A (SGAI), leverage (LVGI) and total accruals to total assets (TATA).
It was built by Messod Beneish in 1999 by comparing companies later found to have manipulated earnings against those that had not. It is a screen trained on a historical sample, not a proof of anything.
FilingEye computes all eight from the filing’s own XBRL and shows each one against its own threshold, because a single composite number hides which factor is doing the work.
How to read it
The conventional cut-off is −1.78. A score above it means the accounts resemble the manipulator group in Beneish’s sample more than the non-manipulator group.
| Range | What it suggests |
|---|---|
| Below −2.22 | Comfortably in the range of ordinary accounts |
| −2.22 to −1.78 | Unremarkable, but worth reading the factors |
| Above −1.78 | Flagged — the model’s classification threshold |
What it does not tell you
The part most tools leave out. A score you over-trust is worse than no score.
- It does not detect fraud. It measures resemblance to a statistical profile. Most companies above the cut-off have done nothing wrong, and a determined fraud can sit below it.
- It is unreliable for financial firms, REITs and insurers, whose balance sheets do not have the structure the model assumes.
- Rapid legitimate growth pushes several factors up at once. A company that genuinely doubled its sales will look worse on SGI and DSRI than a flat one.
- It needs two comparable years. A company that changed its fiscal year, restated, or completed a large acquisition will produce factors that compare two different businesses.