Using Analytics to Detect Possible Fraud: Tools and Techniques

Chapter 4

The Beneish M-Score Model

The Beneish M-Score model (the Model), developed in 1999 by Messod D. Beneish, Ph.D., professor of accounting in the Kelley School of Business at Indiana University—Bloomington, consists of eight indices capturing financial statement anomalies that can result from earnings manipulation or other types of fraudulent activity. Actual data in the financial statements builds the calculations of the indices that create the overall M-Score describing the degree of possible earnings manipulation or possible other fraudulent activity, such as concealing embezzlement activity. In his study, Beneish found that he could correctly identify 76% of the earnings manipulators and incorrectly identify 17.5% as non-manipulators.¹ In other words, Beneish found that 17.5% of the companies whose financial statements he thought were free from earnings manipulation re-filed financial statements later due to earnings manipulation. From the financial forensic examiner's perspective, the percentage of correct identification provides reassurance that the calculations deliver reliable information concerning the examination of the financial information, thus allowing the investigative work to be more effective and efficient.

Beneish also determined that if the calculation of the Model is greater than a negative (−) 2.22, the calculation suggests a higher probability of financial statement manipulation. To understand exactly which numbers are greater than a −2.22 requires ...