Detecting Fraud Using Newly Developed Potential Fraud Indicators

• In Event: Breakfast with Research Round Table Forum
  In Workshop: Table 3

Abstract

The Association of Certified Fraud Examiners (ACFE) estimates that organizations lose five percent of annual revenues to fraud which translates to a potential global fraud loss of $3.5 trillion dollars (ACFE 2012). In July of 2010 alone the SEC charged or settled five enforcement cases involving well-known companies including KBR, Goldman Sachs, Dell, General Electric, and Citigroup (Verschoor 2010). Moreover, fines and settlements continue to increase along with tougher laws and more aggressive enforcement. One instance involved sanctions to a small group of companies of $1.28 billion dollars (Verschoor 2010). Fines and settlements often top $100 million when they were only a few million dollars a few years ago (Schoen 2006). Clearly, there is a great need for more accurate prediction of fraudulent activity, and a need for those predictions to be timely enough to substantially mitigate the resulting financial fallout. In an effort to help minimize such losses, this proposal seeks to develop a new fraud detection model using potential fraud indicators. I will develop these indicators with the help of former prominent financial accounting executives; hereafter referred to as perpetrators. The perpetrators have served in the role of CEO or CFO, been convicted of white-collar crimes, served prison time, and are willing to share how they personally manipulated financial statements. Together, we will scour annual and quarterly financial reports and proxy statements filed with the Securities and Exchange Commission (SEC). Observing and documenting how these perpetrators manipulated financial statements without detection for years will allow me to identify and develop potential fraud indicators which will lead to building stronger fraud detection models.

This will be the first study, to my knowledge, to: (1) document how perpetrators manipulated financial statement accounts and footnotes to hide multiple years of fraud from investors, auditors, and the SEC and (2) apply a method called content analysis to the footnotes of the financial statements, in conjunction with other methods, to build a fraud detection model (prior studies (e.g. Bretton and Taffler 2001; Churyk, et al. 2008, 2009; Lee, et al. forthcoming; Previts et al. 1994; Rogers and Grant 1997) used content analysis to analyze Management Discussion and Analysis (MDA), the Chairman’s letter, or sell-side analyst reports). Because this study will build a model using newly developed potential fraud indicators as identified by perpetrators, it will provide a valuable tool to detect potentially falsified financial statements before catastrophic damages result.

Author

• Natalie T Churyk, Northern Illinois University