Recently, Bloomberg reported that Fiat Chrysler Automobiles is under investigation for potential securities fraud after being accused of inflating U.S. car sales. The company is said to have coordinated a scheme in which individual dealers were paid to create false sales reports, thereby exaggerating financial performance and deceiving investors. Unfortunately, instances of fraudulent financial reporting such as this one have become somewhat commonplace among corporations under pressure to meet earning expectations or to conceal declining financial performance.
Fraudulent financial reporting is any intentional misstatement of, or omission from, the financial statements of a company with the purpose of misleading the statement users. The complexity of financial statement fraud has garnered significant attention over the past decade. It has become clear that there are a variety of subtle methods used to manipulate expenses and revenues that are not easy to detect without knowledge of the necessary analytical tools. In 2002, following a series of major corporate financial fraud scandals, Congress passed the Sarbanes-Oxley Act which enacted strict reforms to financial reporting standards in the hopes of protecting investors from potential fraudulent accounting activities. In spite of these new standards, the number of fraud cases, as well as the dollar value of false financial reporting’s, has continued to rise for the past 14 years.
Most cases of financial statement fraud take the form of either improper revenue recognition, misstatement of assets, liabilities or expenses. It is nearly impossible to trace the source of every revenue stream, verify the existence of all reported assets, or find potential expenses that may have been missed. Forensic analytic methods, however, can offer insight into the detection of highly specific financial-reporting irregularities by identifying irregularities in reported values. A few of these methods include analyzing the reported digits, detecting mathematical biases, and evaluating risk based on inconsistent data.
Analyzing Digit Patterns
Truthful financial statements generally conform to Benford’s Law, which is a mathematical principle asserting that, in sets of large natural numbers, the leading digit is most often small. The law states that 1 is the most commonly occurring digit, appearing 30% of the time; 2 appears around 18% of the time, 3 around 13% of the time, 4 around 10% of the time, 5 around 8% of the time, 6 around 7% of the time, 7 around 6% of the time, 8 around 5% of the time, and 9 around 4% of the time. Groups of reported numbers that deviate from the predictions of Benford’s Law significantly more than the median absolute deviation are often red flags for either intentional or unintentional misstatement. The forensic accountant who discovers these deviations should perform further tests to check for biases in the reported numbers.
Benford’s Law can be used to identify numerical biases influenced by psychological thresholds. For example, a company may be inclined to round a number like 196,497 to something over 200,000 in order to make it seem like a significantly larger number while only marginally increasing the reported amount. This bias can be identified by an excess in second-digit “0’s” and relative lack of second-digit “9’s”.
Biases can also be identified statistically by graphing the data, using smoothing methods to identify any large discrepancies, and calculating statistics to determine the significance of the discontinuities. This allows the analyst to quickly single-out inconsistent data.
Various risk-scoring methodologies have been developed to detect fraud based on reported financials. These methods evaluate risk of fraud by identifying erratic behavior, fluctuation in account balances (high or low), and targeted behavior in sales, profits, expenses, etc., to name a few.
Data from these categories is applied to a series of formulas to calculate a risk-score for each category. Each of these scores is then weighted to most accurately reflect the total risk. Once risk is evaluated and the likelihood of fraud is made clear, the analyst is able to further investigate as needed.
Although many of these tools can be automated, there is often no substitute for the hands on approach of a gumshoe. The first step in analyzing the possibility of fraud is to prepare a side-by-side trend analysis of the subject company. By comparing its balance sheet and income statement year-by-year and category-by-category, one can identify possible areas of concern due to variances, omissions and other trends. Another method used by forensic accountants is to compare the subject company to its peer group. In this instance financial irregularities can be identified by changes in performance and position as compared to similarly sized businesses within the same industry. This analysis can even be performed by region and statement date.
Detecting financial fraud is not limited to testing of data found solely in financial statements and income tax returns. Forensic accountants are also engaged to examine accounting and operation systems that can identify potential fraudulent behavior.
Mark S. Gottlieb, CPA/ABV/CFF, CVA, CBA is a credentialed business valuation and forensic accounting expert. If you would like more information about this or any other topic regarding business valuation or forensic accounting please feel free to contact him at 646-661-3800 or by email at firstname.lastname@example.org.