Data Mining and Predictive Modeling

Paper 080-31

Fraud Detection – A Primer for SAS® Programmers
Thomas J. Winn Jr., State Auditor’s Office, Austin, Texas
Prevention of fraud and abuse has become a major concern of many organizations. The SAS System includes powerful tools for data management, data analysis, and reporting, which can be used in battling against fraud and abuse. This expository paper will describe some basic methods for detecting possibly fraudulent activities, using Base SAS software. Additional remarks will mention using other SAS software components for fraud detection.

Recent highly publicized corporate scandals have resulted in strict new laws, policies, and standards that affect most companies, as well as many other organizations. The Sarbanes-Oxley Act of 2002 specifies new requirements regarding the financial management of publicly traded companies. It requires companies to implement antifraud programs and controls. Also in 2002, the American Institute of Certified Public Accountants (AICPA) issued Statement on Auditing Standards No. 99, which specifies requirements for auditors concerning fraud considerations in financial statement audits. Those rules affect more than just auditors, they also impact the work of employees in most operational components of many companies. Efforts aimed at preventing or detecting fraud and abuse are becoming widespread features of business entities around the world. The Association of Certified Fraud Examiners estimated that, in 2004, occupational fraud and abuse cost the U.S. economy about $660 billion, which represents a loss of about 6 percent of revenues. Fraud is a significant and growing financial risk, which threatens the profitability and reputation of companies around the world. Fraud can occur anywhere, and it has become a major concern of many organizations. Consideration of the possibility of fraud and abuse now is a part of every audit performed by the...