Montier C-Score

The Montier C-Score is a six factor earnings manipulation screen developed by James Montier, the long time behavioral finance and value investing author, while he was working at Societe Generale and later at GMO. The "C" stands for cheating. Montier introduced the score in his 2008 book "Value Investing: Tools and Techniques for Intelligent Investment" and has updated it periodically since. The model was designed as a complement to the Beneish M-Score. Where Beneish looks at eight continuous variables and reports a probability style number, Montier picks six binary questions and counts the yes answers. The score ranges from zero (passes every test) to six (fails every test). A score of four or higher is Montier's flag for heightened manipulation risk.

The design philosophy behind the C-Score is important. Montier wanted a screen that could be computed quickly from standard financials, that was robust to sector and geography, and that picked up patterns that Beneish missed. The six tests are deliberately blunt. Each one asks whether the company is exhibiting a pattern that manipulators commonly show but honest companies rarely do. Individually, any single test is easy to fail for innocent reasons. Collectively, a company that fails four or more is showing a suspicious pattern that deserves real scrutiny. Montier explicitly described the C-Score as a sorting tool, not a conviction signal, and warned against using it as a standalone trading rule.

The academic validation of the C-Score is less rigorous than Beneish's, because Montier was writing for practitioners rather than journals. But the qualitative pattern has held up in the subsequent decade. Companies scoring four or higher on the C-Score have historically underperformed their sector by a meaningful margin in the year after the score is computed, and the underperformance comes disproportionately from large negative tail events (earnings restatements, short seller attacks, accounting investigations) rather than from broad based mean reversion. This is the kind of pattern you would expect from a real forensic signal, where the average effect is driven by a minority of disaster cases rather than by uniform underperformance across the screen.

TickerXray runs the C-Score on every stock in the coverage universe and reports the overall score together with the individual test pass or fail results, so you can see exactly which of the six patterns are triggering the flag.

The single most important use of the C-Score is to look at the combination of which tests are failing rather than the score alone. A company flagging on tests 1 and 2 (net income above cash flow, rising DSO) is likely stretching revenue recognition. A company flagging on tests 5 and 6 (slowing depreciation, rapid asset growth) is likely capitalizing expenses aggressively. The pattern diagnoses the specific manipulation style.

As a second opinion on Beneish: run both the M-Score and the C-Score on every forensic candidate. When they agree (both flag or both clear), the signal is much stronger than either alone. When they disagree, the reason is usually instructive: Beneish is picking up something the six binary tests miss, or the C-Score is picking up a pattern that is averaged out in Beneish's continuous model.

As a short candidate filter: companies that flag on the C-Score (score of 4 or higher) have historically produced a disproportionate share of short squeezes, restatements, and fraud investigations. Public short sellers often cite combinations of C-Score style flags in their published theses.

As a portfolio screen: run the C-Score quarterly on every holding. A position whose C-Score rises from 1 to 4 over two or three quarters is almost always showing an underlying deterioration in earnings quality, regardless of what the stock price has done. That is often the earliest quantitative signal a portfolio manager will get.

As a teaching tool: the six tests are simple enough that analysts can learn to compute them by hand from a 10 K, which builds the forensic intuition that pure model outputs cannot provide. Many forensic accounting courses use the C-Score as the introductory framework before moving to Beneish and more sophisticated models.

The C-Score is binary at every test, which means it loses information. A company with DSO that has risen 0.5 days and one with DSO that has risen 30 days both fail test 2 with the same penalty. Beneish handles this more gracefully because each variable is continuous. The C-Score is therefore best read alongside the M-Score rather than on its own.

Several of the tests are designed to flag growth. Rapid asset growth and rising DSO often appear together at healthy high growth companies, which is why rapidly growing technology firms routinely score 2 or 3 without any underlying manipulation. Read the C-Score in the context of the company's growth rate and sector norms.

The C-Score does not adjust for industry. A retailer's DSO behavior is structurally different from a software company's, and the test thresholds are not calibrated to sector. TickerXray supplements the raw score with sector normalized versions of each test so you can see whether the flag is sector specific or company specific.

Finally, the C-Score is a sign of manipulation risk, not of manipulation. Many companies with a score of 4 are not manipulators. They are companies with accounting patterns that merit investigation. The investigation is the job; the C-Score just tells you where to start.

James Montier is a British value investor, behavioral finance author, and long time chronicler of the limits of efficient markets. He published the C-Score in 2008 while at Societe Generale, and expanded on it in his 2008 book "Value Investing: Tools and Techniques for Intelligent Investment." Montier was influenced by his own work on the behavioral biases of sell side analysts and by Richard Sloan's academic research on accruals. The C-Score's design is explicit about its goal, to find the kinds of accounting patterns that honest companies do not usually display but manipulators routinely do, and to keep the screen simple enough that any investor could apply it. Montier moved from Societe Generale to GMO in 2009, where he has continued to publish on market psychology and forensic value investing. The C-Score remains a standard teaching tool in the forensic accounting curriculum and a routine input in professional short selling research.