as it involves an employee of the organization. Corruption schemes are difficult to detect as the gain to the employee is paid by an outside party and not recorded in the organization’s records. Corruption schemes include bribery, tender or bid rigging, kickbacks, illegal gratuities, extortion, and conflict of interest. Data analytics can expose transactions that may be associated with corruptions schemes. One of the most powerful tests that can be applied to data sets is a relationship-linking test, as demonstrated in this chapter.
• Chapter 15, “Money Laundering.” This chapter introduces the reader to money-laundering schemes and how to detect them. Anti-money-laundering organizations are listed to provide resource information on this subject matter. Techniques used to convert illicit funds to what appear to be legitimate funds in the placement, layering, and integration stages are discussed. Nontraditional banking systems and new payment methods are included in this introduction. Data analytical tests to detect money laundering are outlined.
• Chapter 16, “Zapper Fraud.” Zappers, or electronic suppression-of-sale devices, are being used to delete sales transactions from point-of-sales systems. This is a worldwide problem that is of great concern to taxation authorities and commercial landlords that base rental revenue on a percentage of sales. This chapter provides some background information on POS systems and zappers. Step-by-step instructions are shown on how to prepare POS files for analysis and on using various analytic techniques to analyze, detect, evaluate, and quantify deleted sales in IDEA.
• Chapter 17, “Automation and IDEAScript.” This chapter introduces the automation tools of IDEAScript, Visual Script, and custom functions. Automating procedures, especially complex ones, can save much time. Considerations for automation are discussed. The benefits of IDEAScript are compared with those of Visual Script. Creating Visual Script and IDEAScript from the recording of procedures manually done in IDEA are shown. In addition, a look into how IDEAScript can create a script from the history log file is included in this chapter. IDEAScript resources are outlined, including useful completed scripts that are available from the companion website.
• Chapter 18, “Conclusion.” This chapter explains why using data analytics to initially detect financial-statement fraud is not appropriate. Features, equations, and functions demonstrated throughout the book are summarized here. The project overview feature within IDEA is briefly discussed, showing how it brings together all the databases used in the project and how they relate to one another. There are some final words regarding data analytic challenges and future needs.
Acknowledgments
I WOULD LIKE TO THANK Athena Mailloux, who developed my interest in occupational fraud and fraud detection. Without her insistence that I join her in teaching a financial-fraud program that she developed for a local college, I would not have found the enthusiasm for fraud detection.
Alain Soublière and all the wonderful people at CaseWare IDEA Inc. deserve much credit for their assistance, support, and professionalism. CaseWare IDEA allowed me to combine fraud detection with data analytics.
A special thanks to Brian Element, an IDEAScript master. Without Brian, the automated IDEAScripts offered in the companion website would not exist. I met Brian, a chartered professional accountant, at an advanced IDEA course that he presented in 2011. We started chatting and it came out the Brian wanted to learn IDEAScripting. Brian said that he would have to come up with a scenario for his self-learning IDEAScript project. At the time, I had given much thought to the detection of anomalies through the use of data analytics, so I said to him, “Have I got a deal for you!” Since that time we have been working together on ideas for scripts for fraud detection that Brian programs. From self-learning, Brian has become the world’s leading expert in IDEAScripts and he has been conferred the certified IDEAScript expert (CISE) designation by CaseWare IDEA. At the time of writing, Brian is the sole holder of this designation.
Most important, I would like to thank my family, friends, and associates for all their support and encouragement to be able to complete this book.
CHAPTER 1
Introduction
ORGANIZATIONS GENERATE AND RETAIN more information stored in electronic format than ever before, yet even though there is more analysis performed with the available data, fraud persists. With such vast amounts of data, abusive scheme transactions are hidden and are difficult to detect by traditional means. Data analytics can assist in uncovering signs of potential fraud with the aid of software to sort through large amounts of data to highlight anomalies.
This book will help you understand fraud and the different types of occupational fraud schemes. Specific data analytical tests are demonstrated along with suggested tests on how to uncover these frauds through the use of data analytics.
A short definition of fraud is outlined in Black’s Law Dictionary:
An act of intentional deception or dishonesty perpetrated by one or more individuals, generally for financial gain.1
This simple definition mandates a number of elements that must be addressed in order to prove fraud:
• The statement must be false and material.
• The individual must know that the statement is untrue.
• The intent to deceive the victim.
• The victim relied on the statement.
• The victim is injured financially or otherwise.
The false statement must substantially impact the victim’s decision to proceed with the transaction and that perpetrator must know the statement is false. A simple error or mistake is not fraudulent when it is not made to mislead the victim. The victim reasonably relied on the statement that caused injury to the victim or placed him or her at a disadvantage.
It is intentional deception that induces the victim to take a course of action that results in a loss that distinguishes the theft act.
In addition to the employer suffering a financial or other loss, occupational fraud involves an employee violating the trust associated with the job and hiding the fraud. The employee takes action to conceal the fraud and hopes it will not be discovered at all or until it is too late.
The word abuse is employed when the elements for defining fraud do not explicitly exist. In terms of occupational abuse, common examples include actions of employees:
• Accessing Internet sites such as Facebook and eBay for personal reasons.
• Taking a sick day when not sick.
• Making personal phone calls.
• Deliberately underperforming.
• Taking office supplies for personal use.
• Not earning the day’s pay while working offsite or telecommuting.
There is an endless list that can fall under the term abuse, but no reasonable employer would use this word to describe any employee unless the actions were excessive. Organizations may have policies in place for some of these items, such as an Acceptable Internet Use Policy, but most would be considered on a case-by-case basis, as the issue is a matter of degree that can be highly subjective. There would unlikely be any legal actions taken against an employee who participated in a mild form of abuse.
In the data analysis process, “Detecting a fraud is like finding the proverbial needle in the haystack.”2 Typically, fraudulent transactions in electronic records are few in relation to the large amount of records in data sets. Fraudulent transactions are not the norm. Other anomalies, such as accounting records anomalies, are due to inadequate procedures or other internal control weaknesses. These weaknesses would be repetitive and will occur frequently in the data set. Sometimes, they would regularly and