What does the Beneish M-Score measure?

It combines financial statement variables into an accounting-risk screening score. It can guide further review but cannot determine whether a company committed fraud.

Which formula does this page use?

The eight-variable implementation in Beneish, Lee and Nichols (2013), with a 4.679 accrual coefficient and a -0.327 leverage coefficient. Its cash-flow-based accrual definition differs from the original balance-sheet estimate.

Does an M-Score below -1.78 mean normal accounting?

No. It means the company is not flagged by that cutoff and implementation. The model can miss manipulation, and data or business-model differences can affect its usefulness.

Can I calculate it when a field is missing?

Not reliably by inserting zero or deleting the term. Identify the missing filing input or report the calculation as unavailable. A different model requires its own definitions, coefficients and validation.

Can it be applied to banks and insurers?

The cited research excludes financial services firms. Their balance sheets and revenue structures require other analytical methods.

Financial Statement Analysis

Beneish M-Score: Eight Variables, Formula and Screening Limits

The Beneish M-Score combines financial statement ratios to flag companies for further accounting review. This page uses the eight-variable implementation published by Beneish, Lee and Nichols in 2013, including its cash-flow-based accrual measure. A score above -1.78 crosses that study’s screening cutoff; it does not prove manipulation or assign a reliable present-day fraud probability.

See the decision guide

Eight inputs and their definitions

Let t be the current fiscal year and p the prior year. CA means current assets, PPE means net property, plant and equipment, and TA means total assets. Use reported receivables and the same sales scope in both periods. These definitions follow the 2013 paper’s Appendix A; some online calculators use different accrual definitions or coefficients.

Input	Calculation	Input check
DSRI	(Receivables_t / Sales_t) / (Receivables_p / Sales_p)	Review allowances, factoring and changes in customer terms.
GMI	Gross margin_p / Gross margin_t; gross margin = (Sales - COGS) / Sales	Use a margin fraction, not gross profit divided by sales twice.
AQI	[1 - (CA_t + PPE_t) / TA_t] / [1 - (CA_p + PPE_p) / TA_p]	Do not add a separate securities deduction to this version.
SGI	Sales_t / Sales_p	Match reporting periods and distinguish acquisitions from organic growth.
DEPI	Depreciation rate_p / Depreciation rate_t; rate = Depreciation / (Depreciation + net PPE)	Do not silently substitute combined depreciation and amortization when depreciation is unavailable.
SGAI	(SG&A_t / Sales_t) / (SG&A_p / Sales_p)	Check expense classification and restructuring effects.
LVGI	[(Current liabilities_t + Long-term debt_t) / TA_t] / [(Current liabilities_p + Long-term debt_p) / TA_p]	This definition includes current liabilities; it is not simply interest-bearing debt.
Accruals	(Income before extraordinary items_t - CFO_t) / TA_t	Use the paper’s income scope and year-end assets; document how modern reported income maps to it.

Formula and a reproducible calculation

The 2013 paper reports: M = -4.84 + 0.920 × DSRI + 0.528 × GMI + 0.404 × AQI + 0.892 × SGI + 0.115 × DEPI - 0.172 × SGAI + 4.679 × Accruals - 0.327 × LVGI. The accrual coefficient here is 4.679 and the leverage term is negative. Do not mix these with another calculator’s coefficients.

Hypothetical arithmetic check: set all seven year-over-year indices to 1.00 and Accruals to 0.05. The weighted index terms total 2.360 and the accrual contribution is 0.23395, so M = -4.84 + 2.360 + 0.23395 = -2.24605.
Keeping every other hypothetical input unchanged but setting DSRI to 1.60 adds 0.920 × 0.60 = 0.552. M becomes -1.69405, above the -1.78 cutoff. This illustrates sensitivity, not a finding about a real company.
For a spreadsheet, store each source filing, reporting date, line item and unit alongside its input. Calculate the ratios in separate cells, then sum the weighted terms. Retain unrounded values until display.
If a denominator is zero or inputs cannot be mapped consistently, report the score as unavailable and show the unresolved input. A five-variable model is a separately specified model, not permission to omit missing variables from this equation.

How to interpret the -1.78 cutoff

The cutoff reflects a tradeoff between missed cases and false alarms in the cited research. It is not a universal legal, audit or investment standard. A raw M-Score is not a percentage, and a score below the cutoff does not certify accurate accounts.

Result	What it supports	Next step
M > -1.78	A flag under this particular implementation	Trace the largest changes to filings and footnotes; test business explanations.
M ≤ -1.78	No flag under this cutoff	Continue normal financial analysis; do not label the company clean or fraud-free.
Missing, unstable or incomparable inputs	No reliable comparable score	Resolve data definitions, very small denominators and reporting changes first.

Review the business explanation before drawing conclusions

Growing receivables can reflect legitimate credit sales; an acquisition can change assets, debt and margins. Neither observation establishes misconduct. Read revenue policies, receivable aging, reserves, acquisition notes and depreciation estimates before interpreting a changed score.

The cited study excludes financial services firms. Do not apply industrial-company ratios mechanically to banks or insurers.
Losses, negative gross margins and very small denominators can produce mathematically finite but hard-to-interpret indices. Disclose them rather than presenting a normal risk label.
The research used sample-level outlier treatment. A raw single-company calculation without that preprocessing is not a complete reproduction of its empirical tests.
Negative SGAI and LVGI coefficients are conditional model weights; they do not mean rising overhead or debt makes a business safer. Coefficient size alone also cannot rank variable importance.
Use filings available at the analysis date. Later restatements must not be inserted into a historical investment test without disclosure. A screening result does not establish a profitable trading rule.

Where Pineify fits

AI Finance Agent

Retrieve available financial statements and organize the inputs for review. Verify filing definitions and any missing fields yourself; this page does not provide an automatic Beneish calculator or an audit opinion.

Research financial statement inputs

Related resources

Earnings Quality Accrual Ratio Operating Cash Flow Ratio Net Profit Margin Financial Ratios Tool AI Finance Agent

Frequently asked questions

Educational financial analysis, not an audit opinion or investment recommendation. Verify the model version and filing inputs; no score establishes or rules out fraud.

Sources and verification

Last verified: 2026-10-20