Fraudulent financial reporting is a growing problem, says John Sutter. Sutter: Auditors are required to apply quantitative analytic procedures to financial data. Fraud risk detection is a natural candidate for software-based automated assistance, he says. The American Institute of Certified Public Accountants issued Statement of Accounting Standards (SAS) No. 99.
In the first category, most of the well known analytical tests for detecting fraud risks have applied ratio analysis to the consolidated financial statements. In the second category, analytics and associated procedures may be applied to detailed financial data to identify individual transactions at risk of being fraudulent. Scanning analytics may be used to identify outliers, or ratio and trend analysis normally applied to aggregated financial statements may also be applied.
Fraud risk detection analytics should operate at the level of financial transactions, say PricewaterhouseCoopers. Such analytics would need to work on large amounts of data (anywhere from 1 to 50 gigabytes per client per year), they say. The research and development challenge is to create such analytics that are cost-effective to apply on a large scale.
Auto116
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.83.2039&rep=rep1&type=pdf