Home / Articles / Early fraud detection: How the SEC uses data mining

Early fraud detection: How the SEC uses data mining

June 21, 2012

Share:

The most effective strategy for combating fraud is a combination of prevention and early detection. Once a fraud has been completed, the chances of fully recovering losses are relatively low — around 16%, according to a survey by the Association of Certified Fraud Examiners. Increasingly, organizations are turning to risk analytics, such as data mining, to uncover fraudulent activity.

The power of technology

Data mining harnesses the power of technology to sift through massive amounts of information to reveal trends. One reason fraud is difficult to detect is that individual transactions often appear to be legitimate. Data mining makes it possible to examine large volumes of data to uncover relationships, patterns and trends that may signal fraudulent activity.

The Securities and Exchange Commission (SEC) has had great success using data mining to detect fraudulent hedge fund activity. Under an initiative it calls Aberrational Performance Inquiry (API), the SEC uses proprietary risk analytics to examine hedge fund returns and identify performance that “appears inconsistent with a fund’s investment strategy or other benchmarks.” Once the SEC flags a firm that’s generating abnormally high returns, it investigates the firm’s activities more closely.

API in action

Recently, API has led to fraud charges against several hedge funds and individuals. In one case, the SEC’s complaint alleged that a fund’s former portfolio manager schemed with brokers to inflate the fund’s reported monthly returns and net asset value by manipulating its supposedly independent valuation process.

In another case, the SEC charged a hedge fund firm and its sole managing director with engaging in a pattern of deceptive conduct. Among other things, they materially overstated a fund’s performance, gave investors the false impression that the fund’s returns were consistently positive and minimally volatile, and inflated the firm’s assets.

According to SEC personnel, “The extraordinary returns reported by these advisers and portfolio managers were, in most cases, too good to be true. In other cases, outlier returns were a telltale sign that something else was amiss.” Robert Khuzami, Director of the SEC’s Division of Enforcement, commented that “this approach — especially in the absence of a tip or complaint — minimizes both the number of victims and the amount of loss while increasing the chance of recovering the funds and charging the perpetrators.”

No substitute

Data mining can be an effective fraud detection tool in a variety of contexts — particularly when used in connection with traditional fraud detection methods. Although no substitute for a thorough investigation, it can point you in the right direction by identifying suspicious patterns.

All content provided in this article is for informational purposes only. Matters discussed in this article are subject to change. For up-to-date information on this subject please contact a Clark Schaefer Hackett professional. Clark Schaefer Hackett will not be held responsible for any claim, loss, damage or inconvenience caused as a result of any information within these pages or any information accessed through this site.

Guidance

Related Articles

Article

2 Min Read

The Details on GASB 97

Article

2 Min Read

Lease Accounting Standard ASC 842 Impacts on Debt Covenants and Capital Requirements

Article

2 Min Read

Federal Audit Clearinghouse Provider Changing from Census to GSA

Article

2 Min Read

Consequences of Not Being Proactive on Lease Accounting Standard ASC 842

Article

2 Min Read

Infographic: 4 Steps to Implementing the New Lease Accounting Standard

Article

2 Min Read

Maximize Your Tax Filing Preparedness & Awareness

Get in Touch.

What service are you looking for? We'll match you with an experienced advisor, who will help you find an effective and sustainable solution.
  • Hidden
  • This field is for validation purposes and should be left unchanged.