SEC Eyes Data Mining to Boost Fraud Detection

  |  July 16, 2013

Newly formed Securities and Exchange Commission task forces will explore the use of the SEC’s increased data-mining capability as a way to detect financial-reporting fraud at corporations, the two leaders of one of the units told CFO.

Formed earlier this month, the SEC’s Financial Reporting and Audit Task Force will try to bring “fresh thinking” to the detection of financial and accounting fraud, says David Woodcock, the director of the Fort Worth, Texas, regional office of the SEC. Woodcock will be the task force’s chairman.

To be sure, the “games being played” in financial reporting, especially related to revenue recognition, asset, and liability valuation, reserving, accrual accounting and the metrics existing outside of Generally Accepted Accounting Principles, have been around for a long time, according to Woodcock.

What’s changed is the “drastically reduced cost” and the greatly increased power of electronic data analysis, he said. Via data-mining, SEC investigators can explore how “the many threads” of complex frauds converge, Woodcock, says, noting that the coordination of such efforts across different SEC units will help make fraud detection more efficient.

On July 2, the SEC announced three new initiatives aimed at buttressing its Division of Enforcement’s already existing efforts to bring fraudsters to heel. Besides the Financial Reporting and Audit Task Force, the commission launched The Center for Risk and Quantitative Analytics and The Microcap Fraud Task Force.

Woodcock said that his group would be working with The Center for Risk and Quantitative Analytics on data-mining efforts. The center will use “quantitative data and analysis to profile high-risk behaviors and transactions and support initiatives to detect misconduct, increasing the [Enforcement] Division’s ability to investigate and prevent conduct that harms investors,” according to an SEC press release.

For its part, the Microcap task force will target “abusive trading and fraudulent conduct in securities issued by microcap companies, especially those that do not regularly publicly report their financial results,” according to the release.

Aberrant Behavior

What uses of data will Woodcock’s group, which will include about eight to 10 accountants and lawyers, focus on?

A good comparison is the SEC’s effort to detect “aberrational performance” among hedge funds as a way of unearthing fraud, according to Margaret McGuire, senior counsel to the SEC’s department directors, who will serve as vice-chair of the reporting and auditing task force.

The hedge fund investigation, which resulted in December 2011 enforcement actions against three advisory firms and six people associated with hedge funds, aims to identify hedge-fund performance that’s inconsistent with fund strategies, benchmarks or both.

The Aberrational Performance Inquiry, a joint effort among SEC staff in its Enforcement Division, Office of Compliance, Inspections, and Examinations, and Division of Risk, Strategy and Financial Innovation, uses its own risk analytics. The point is to locate, in the words of former SEC Chair and current Commissioner Elisse B. Walter, a hedge-fund performance that’s “too good to be true.” Among such aberrations are consistently better performance than market indexes and consistent achievement of positive performance with low volatility.

In a similar way, the new accounting and financial reporting task force might, for example, look at ways to spot outliers in the reporting of revenue recognition, according to McGuire. How to recognize aberrations in financial-statement and debt-covenant ratios could be another focus of the group, she added.

The task force will also explore increasing the use of technology-based tools like the Accounting Quality Model (AQM), a set of quantitative analytics that can be used across the SEC to assess how unusual a company’s financials seem. In combination with publicly available data services, SEC investigators can use the AQM “like a dashboard” to detect the likelihood that fraud has occurred at a given company.