Skip to main content
Earnings QualityM-Score

What is the Beneish M-Score?

The Beneish M-Score is an earnings manipulation detection model developed by Messod Beneish in 1999. It uses 8 accounting variables to produce a probability score. M-Score above -1.78 suggests possible earnings manipulation. Famous for identifying Enron (M-Score = -0.47 in 2000, far above the -1.78 threshold) before its collapse.

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

The TATA Variable: Why Accruals Are the Strongest Signal

Among the eight M-Score variables, TATA (Total Accruals to Total Assets) receives the highest absolute coefficient in the model (4.679), making it by far the most impactful single variable. TATA is calculated as the change in working capital minus depreciation, divided by total assets -- essentially the proportion of net income coming from non-cash accruals rather than real cash generation. High TATA means a company's reported profits are substantially derived from accounting estimates and timing decisions rather than collected cash, which is precisely the mechanism used in most earnings manipulation schemes.

The connection between the M-Score and the Accrual Ratio is direct: both measure the same underlying phenomenon (earnings backed by accruals rather than cash) from slightly different angles. Sloan's 1996 accruals anomaly research and Beneish's 1999 manipulation model both identified high accruals as a key predictor of future disappointment -- whether through legitimate earnings reversion or outright fraud. The combination of a low accrual ratio (indicating good earnings quality) and a low M-Score (indicating low manipulation probability) is one of the most powerful forensic accounting screens available to individual investors.

Understand the Accrual Ratio

The Accrual Ratio is closely related to the M-Score's TATA variable. Use both together to build a complete picture of earnings quality and manipulation risk.

Learn About Accrual Ratio →

Frequently Asked Questions

What is the Beneish M-Score and what does it detect?+
The Beneish M-Score is a statistical model developed by Indiana University professor Messod Beneish in 1999 to detect the probability of earnings manipulation in financial statements. It uses eight accounting variables derived from publicly available financial data. The eight variables are: DSRI (Days Sales Receivable Index -- rising accounts receivable relative to sales may signal premature revenue recognition), GMI (Gross Margin Index -- deteriorating margins may indicate manipulation pressure), AQI (Asset Quality Index -- increasing non-current assets relative to total assets), SGI (Sales Growth Index -- high revenue growth companies are more likely to manipulate), DEPI (Depreciation Index -- slowing depreciation may signal asset life extension), SGAI (Selling, General & Administrative Expenses Index), TATA (Total Accruals to Total Assets -- high accruals relative to assets is the key manipulation signal), and LVGI (Leverage Index -- increasing leverage increases manipulation incentive).
How do I interpret the M-Score threshold?+
The critical threshold is -1.78. M-Score above -1.78 suggests the company is likely a manipulator (or at minimum, exhibits financial characteristics similar to known manipulators). M-Score below -2.22 suggests the company is probably not manipulating. The range between -2.22 and -1.78 is ambiguous -- neither a clear red flag nor a clean bill of health. In practice, value investors use -1.78 as the hard screen: any company above this threshold gets flagged for deeper investigation of footnotes, revenue recognition policies, and auditor comments. The TATA variable (Total Accruals to Total Assets) receives the highest absolute weight (4.679) -- confirming that accruals-to-assets is the most powerful single indicator of manipulation risk.
What are the famous M-Score predictions?+
The Beneish M-Score's most famous prediction is Enron. A Cornell University analysis later showed that Enron's M-Score in 2000 was -0.47 -- far above the -1.78 threshold -- indicating extremely high manipulation probability. This was a full year before Enron's bankruptcy filing in December 2001. Worldcom (M-Score significantly above threshold before its 2002 fraud revelation) and Satyam Computer Services (Indian outsourcing company that admitted to accounting fraud in 2009) also showed elevated M-Scores before their collapses. The model has been validated in multiple academic studies showing it correctly flags a significant percentage of companies later found to have engaged in earnings manipulation.
What are the limitations of the M-Score?+
The M-Score has several important limitations to keep in mind. First, it generates false positives: some companies score above -1.78 due to rapid legitimate growth (fast revenue growth and rising receivables can trigger high DSRI and SGI scores even without manipulation). Second, the model was developed primarily on U.S. manufacturing companies and works less reliably for financial companies (banks, insurance) whose balance sheets are structured differently and whose accruals patterns are inherently different. Third, it captures patterns but cannot identify fraud with certainty -- it is a probabilistic flag, not a definitive judgment. Best practice: use the M-Score as one layer of a multi-signal forensic accounting screen, combined with accrual ratio analysis, auditor changes, related-party transaction scrutiny, and management incentive analysis.

Related Terms

Accrual Ratio

Gap between reported earnings and cash flow

Piotroski F-Score

9-point financial strength scoring system

Altman Z-Score

Multi-factor bankruptcy prediction model

Free Cash Flow

Operating cash flow minus capital expenditures

Explore More

Investing Tools

Free Calculators

Learn

← Back to Glossary

Weekly Stock Analysis - Free

5 undervalued stocks, fully modeled. Every Monday. No spam.

Cookie Preferences

We use cookies to analyze site usage and improve your experience. You can accept all, reject all, or customize your preferences.