“The key to identifying fraud lies in the ability to comprehend what lies beneath.”
The statement above is the basic purpose of forensic accounting, and the role that a forensic accountant plays in the analysis of a company’s financial statements. Forensic accounting, defined as “a form of investigative accounting which examines financial records in order to find evidence for a lawsuit or criminal prosecution,” may seem as though it has only been around for a few short years, but in reality the prevention and detection of fraud has been found to date back as far as 3300-3500BC. While the term “forensic accounting” has not always been used, archaeological findings tell of the use of these practices in the ancient Egyptian time (Shields, 2009).
The term financial accounting was coined much more recently. As the gap between accountancy and legal professions closed during the 1800s, it became more common for accountants to testify as expert financial witnesses during court cases involving financial fraud (Shields, 2009). One of the most well known cases involved the work of Al Capone (pictured to the right) in the 1930s. For nearly two decades, Capone was able to commit many acts of racketeering and tax evasion, evading some of the best FBI agents and detectives. Finally, it was the work of accountant Frank J. Wilson, later called a forensic accountant, which allowed Capone to be taken down and incarcerated for his tax evasion.
However, the forensic accounting of that day and age is quite different when compared to the forensic accounting that we see the need for today. Forensic accounting services, according to a study done by the AICPA, “generally involve the application of specialized knowledge and investigative skills possessed by CPAs to collect, analyze, and evaluate evidential matter and to interpret communication findings in the courtroom, boardroom or other legal/administrative venue” (AICPA, 2014). In this study conducted in 2014 however, the AICPA found that with the increase of technology and the use of Internet source, electronic data analysis, also known as big data, led the pack of the top issues in the field of forensic accounting. As seen on the chart below titled Top Five Issues Facing Forensic and Valuation Professionals Over the Next Two to Five Years, while technology accounted for only 9% of the concerns of Forensic Professionals in the year 2011, the use of electronic data analysis had jumped to the top spot at 25% in the year 2014 (AICPA, 2014).
With the rapid growth of technology over the recent years and the increase in companies engaging in new technologies that can support these large volumes of data, we must look at the impact this will have on the field of forensic accounting, and the new skills and expectations that will be required of our future forensic accountants.
The Evolution of Big Data
In order to understand the importance of data analytics in regards to forensic accounting, we must first look into “big data” and how the evolution of big data has changed the course of the business world.
Big data, while seemingly new and scary, is something that has been around for many years. While the collection and storage of big data may have changed over the recent periods, the information remains the same as that which has always been used by marketers in the business world. Big data is defined by Investopedia.com as “the growth in the volume of structured data and unstructured data, the speed at which it is created and collected, and the scope of how many data points are covered. Big data often comes from multiple sources, and arrives in multiple formats” (Investopedia.com). While today big data is most often assumed to come from digital inputs (web browsing, social media, etc.), it is important to remember that big data can also come from traditional sources as well, like financial records, product transactions, call centers, etc.
In order to understand big data, we must also understand its components, structured and unstructured data. Structured data refers to data that is highly organized, for example data that is input into a relational database with multiple columns and rows that make the identification of and searching through the information easy and quick. Unstructured data is essentially the opposite of structured data. As the name suggests, this type of data has no identifiable internal structure; it is chaotic, time-consuming to search through, and just about impossible to organize. Things like emails, social media posts, and video or audio files are considered unstructured data.
Today, IBM data scientists define big data using four dimensions, commonly called “the four V’s:” Volume, Variety, Velocity, and Veracity. Using the info graphic below, we see a rundown of this breakdown.
Volume refers to the massive scale of data that is generated each day, using machines, networks, and human sources. To understand the true enormity that is the volume of big data think of it this way; Facebook stores photos as a means of big data. This may not sound like a lot, but the average amount of photos per profile is approximately 100, and the number of profiles registered to Facebook is greater than the number of people that populate China. Solely looking at the photos that are being stored (near 250 billion images), the volume of big data relating to Facebook is already immense. The second “v” when defining big data is variety. Variety is known as the different forms of data (for example, structured and unstructured mentioned above). Velocity indicates the pace at which data is generated from the many different sources in the world. Lastly, and challenging in the world of big data, has been the veracity. This, according to an article by Kevin Normandeau called Beyond Volume, Variety, and Velocity is the Issue of Big Data Veracity, is what Normandeau feels IT professions should be most concerned with. Veracity, according to Normandeau, “refers to the biases, noise, and abnormality in data.” These uncertainties can make it hard to understand and store data in a meaningful way.
The Importance of Data Analytics
“It’s important to remember that the primary value from big data comes not from the data in its raw form, but from the processing and analysis of it and the insights, products, and services that emerge from analysis. The sweeping changes in big data technologies and management approaches need to be accompanied by similarly dramatic shifts in how data supports decisions and product/service innovation.”
-Thomas H. Davenport
With big data comes big data analytics, or the study of numbers that are manually examined to uncover insights and tendencies in a business. The use of big data has been used to derive significant value from meaningless data with speed and efficiency. An article on SAS.com says, “Whereas a few years ago business would have gathered information, run analytics and unearthed information that could be used for future decisions, today that business can identify insights for immediate decisions. The ability to work faster – and stay agile – gives organizations a competitive edge they didn’t have before.”
So, data analytics is important in today’s business world, because it helps companies take these heaps and heaps of pointless information and turn it into opportunity. Through data analytics companies can do three things. First, data analytics has helps immensely with cost reduction in businesses. Not only does the use of big data and data mining allow for significant cost advantages, through the use of storage technologies like “the Cloud,” but the results of data analytics can help a company to ascertain more efficient business tactics and strategies. Second the efficiency of big data also helps in the decision making process within a company. Businesses are able to make faster and smarter decisions with the ease of access and analysis of their sources of data. Lastly, data analytics helps companies to gauge what their clientele base want and need, and as a result create new products or offer new services to better suit their market.
As we can see, big data and data analytics is complex. With these voluminous and growing amounts of data, there comes the associated increased risk of potential fraud to be buried within.
What is Forensic Analytics?
One of the largest risks to accompany the upsurge in big data is fraud. With the intricacy and massiveness of big data, fraud becomes “hard to spot…harder to trace…harder to prevent” according to an article written by the Florida Atlantic University College of Business. This is where forensic analytics comes into play. Forensic analytics is a special type of investigative process that helps to completely examine the depths of immeasurable sums of digital material. While forensics has typically been used to uncover fraud that has occurred in the past in order to be used for legal purposes, forensic analytics is slightly different. Forensic analytics are being used as a preventative measure, in an attempt to identify the possibility of fraud before the fraud actually occurs. According to a statement released by Ernst and Young:
Forensic Data Analytics is a science used to proactively seek opportunities to prevent and detect fraud, waste, and abuse by leveraging information in corporate data assets. It enables identification of meaningful patterns and correlations in existing historic data to predict future events and assess the reasons for various fraudulent activities. Such insightful predictive information is generally “invisible,” but provides a platform on which organizations can take business decisions related to fraud, disputes, and misconduct.
Unlike traditional forensics, which was used after a fraud had occurred as a method to provide an expert opinion and proof said fraud, forensic analytics is now being used prior to the occurrence of fraud. Accounting firms like Ernst and Young, KPMG, Deloitte, and others have adopted techniques to perform these analytics. Ernst and Young uses a three step technique, pictured below:
First, a professional manages the data, which means they sift through all data collected by a company in order to leverage the pertinent data. Next they perform the analytics on the data. Ernst and Young lists three different types of analytics: prescriptive, predictive, and descriptive analytics (Demystifying “Big Data” Analytics, 2013). Lastly, EY uses the results from their analytics to drive decisions regarding the company, i.e. ways to grow revenues, reduce costs, improve internal controls, etc.
One specific example that Ernst and Young also uses to prevent fraud within a company is something they call “Fraud Triangle Analytics” (Curtis, 2012). In an article titled Ernst and Young: Using Forensic Analytics to Prevent Fraud, EY’s forensic IT team director, Rashmi Joshi, says EY has come up with a method to analyze unstructured data (employee emails) in order to prevent fraud. The fraud triangle theory states that in order for a fraud to occur within a company, three things must be present: opportunity, pressure, and rationalization. With the help of the FBI, EY was able to pull together 3,000 words that are typically associated with these factors. Any email that is flagged with at least one of the three factors mentioned above is considered part of the fraud, and emails that are associated with all three factors typically lead to the person/people ultimately responsible.
Another example of forensic analytics in action took place in 2014, when KPMG realized that the refunds being issued by a national call center were not accurately reflecting Benford’s Law. Also referred to as the first-digit law, Benford’s Law states that any number occurring naturally in a situation is like to be lead by a smaller digit. In other words, a number that begins with a 1 has a probability of 30%, and this percentage decreases as the first digit increases, until 9, which has a probability of less than 5%. In the case with the call center, operators were allowed to issue refunds up to $50 without a manager’s approval. Based on the rules of Benford’s Law, KPMG’s forensic accountant team found that the number four was occurring far too frequently in the data they were analyzing. Through the forensic analytics of the data, KPMG was able to check the validity of the data, and found that a small handful of the call-center operators were issuing these small refunds to themselves, and their family and friends. Seeing as the refunds did not reach the $50 threshold, these operators were able to steal several hundred thousand dollars without warning the managers. However, using forensic analytical techniques mixed with mathematical calculations and procedures, forensic teams were able to detect these abnormalities and put a stop to the fraud.