December 11, 2023

A Simple Calculation Does Not Imply a Simple Conclusion

By Elizabeth Surabian, Senior, Advisory Services

A Simple Calculation Does Not Imply a Simple Conclusion Valuation

The Beneish Model is a tool for evaluating the quality of earnings. It is a mathematical model that utilizes ratios derived from a company’s financial statements in the current year and the prior year. The factors used in the formula are as follows (Y1 refers to the current year, and Y0, the preceding year):

  1. Day’s Sales in Receivables Index (DSRI):
    (Receivables Y1 / Sales Y1) / (Receivables Y0 / Sales Y0)
  2. Gross Margin Index (GMI):
    ((Sales Y0 – Cost of Sales Y0) / Sales Y0) / ((Sales Y1 – Cost of Sales Y1) / Sales Y1)
  3. Asset Quality Index (AQI):
    [1 – (Current Assets Y1 + PP&E Y1 + Securities Y1) / Total Assets Y1] / [1 – (Current Assets Y0 + PP&E Y0 + Securities Y0) / Total Assets Y0)]
  4. Sales Growth Index (SGI):
    Sales Y1 / Sales Y0
  5. Depreciation Index (DEPI):
    (Depreciation Y0 / (PP&E Y0 + Depreciation Y0)) / (Depreciation Y1 / (PP&E Y1 + Depreciation Y1))
  6. Selling General and Administrative Expenses Index (SGAI):
    (SG&A Expense Y1 / Sales Y1) / (SG&A Expense Y0 / Sales Y0)
  7. Total Accruals to Total Assets (TATA):
    (Income from Continuing Operations Y1 – Cash Flows from Operations Y1) / Total Assets Y1
  8. Leverage Index (LVGI): [(Current Liabilities Y1 + Total Long-Term Debt Y1) / Total Assets Y1] / [(Current Liabilities Y0 + Total Long-Term Debt Y0) / Total Assets Y0]

The Beneish M-Score formula is as follows:

M = -4.84 + 0.92*DSRI + 0.528*GMI + 0.404*AQI + 0.892*SGI + 0.115*DEPI – 0.172*SGAI + 4.679*TATA – 0.327*LVGI

If the M-Score is greater than -1.78, it suggests likely manipulation.

Although the M-Score is a relatively straightforward calculation, it does not lead to a straightforward conclusion. This M-Score is a probabilistic model. Thus, it does not prove whether earnings are manipulated or not. Instead, an M-Score greater than -1.78 suggests the need for further investigation. It is essential to consider other variables that may not be included in the model but could indicate earnings manipulation or lack thereof.

Distortion of Financial Statement Data

Distortion of financial statement data may not be a result of earnings manipulation. In Mr. Beneish’s paper “The Detection of Earnings Manipulation,” he states, “…one must recognize that such distortions can have an alternative origin.” He lists examples of potential financial distortions that are unrelated to earnings manipulation, including:

  • “A material acquisition during the period examined,”
  • “A material shift in the firm’s value-maximizing strategy;” or
  • “A significant change in the firm’s economic environment.”

The distortion of financial statement data emphasizes the need to investigate an M-Score further.

Management Can Game the Model

If a company’s management knows this model will be used to potentially detect earnings manipulation, it is less likely to manipulate financials included in this model. Additionally, a manager can make sure the “manipulation” is essentially hidden (i.e., not detected in the model). As a result, there could still be a need to investigate a company even if the M-Score suggests earnings manipulation is unlikely.

Earnings Manipulation and the Beneish Model

Earnings manipulation in the Beneish Model involves overstating earnings. The model does not consider, and therefore cannot be reliably used for, companies manipulating earnings downward. A company might artificially adjust earnings downward to meet an earnings or growth target for a subsequent quarter.

When Mr. Beneish created the M-Score model, he utilized public companies and excluded financial firms. Thus, caution should be used when applying this model to other types of companies.

Overall, the M-Score can be viewed as a probabilistic indicator forensic accountants can use to dive deeper into a company’s earnings quality. However, without further analysis of relevant factors that impact the company and the Beneish Model, relying solely on the M-Score generated from the Beneish Model would be inappropriate.


  • The Detection of Earnings Manipulation by Messod D. Beneish, June 1999.