Glossary
Beneish M-score
A statistical model that uses eight ratios from the financial statements to estimate how likely a company is to be manipulating its earnings.
The Beneish M-score was published by Messod Beneish, an accounting professor, in a 1999 paper. He compared companies that had been caught manipulating their earnings with similar companies that had not, and found eight measures that, combined, separated the two groups better than chance.
The eight inputs
Each compares this year with last year:
- Days sales in receivables index: receivables growing faster than sales.
- Gross margin index: a falling gross margin, which creates pressure to flatter results.
- Asset quality index: more assets that are neither cash-like nor physical, which may be capitalized costs.
- Sales growth index: rapid growth, which brings both pressure and opportunity.
- Depreciation index: a slowing rate of depreciation.
- Selling, general and administrative expense index.
- Leverage index: rising debt.
- Total accruals to total assets: profit not backed by cash. See earnings quality.
The inputs are combined with fixed weights into one score.
Reading the score
Higher scores, meaning less negative ones, suggest a greater chance of manipulation. Beneish's paper used a cutoff of −1.78, and he still uses it. Other cutoffs circulate on investing websites without support in his published work. A score above the cutoff flags a company for closer study. It is not proof.
Track record
The model is often credited with flagging Enron before its collapse; a group of Cornell students used it on Enron's accounts in 1998. It has also been tested many times since, with useful but imperfect results.
Limits
- False alarms are common among fast-growing companies.
- Banks and insurers do not fit the model.
- It finds only some types of manipulation, and it relies on the very statements that may be wrong.
