If the prosecution rate of white-collar crime is relatively low in this country, it’s not because of a lack of will on the part of law enforcement officials to get the bad guys. It’s because of a simple lack of sufficient IT resources to get the job done.
That’s the assessment of Tim Deehan, president ofActionable Intelligence Technologies (AIT), a financial investigation software provider in Dulles, Va. I had the opportunity to speak with Deehan last week, and I started off the conversation by asking him why organizations are failing to devote sufficient IT resources to the problem. He said on the commercial side, the issue is the overhead charge; on the government side, the problem is that it’s simply not their core area of expertise. And in any case, the massive amounts of data that have to be analyzed in a fraud investigation can be overwhelming:
Peter Thiel, in his book, “Zero to One,” said that “Big Data is dumb data.” The tendency is to get massive systems, and to get all of this information via the IT capabilities, but you still need that human who understands how to do a financial investigation. They tend to get buried in massive volumes of data, and they lose the focus on the investigation. The other problem is that law enforcement agencies generally are generating the investigations, but they’re limited by resources—how much they can devote to a specific target, how fast they can figure out if they really have the right target, and if they really have a case that they can win. So the problem is how you can become most efficient and leverage the technology to help you rapidly do what we call “early case assessment”—how fast you can figure out that you have the right guy, the right bank accounts, the right target, and that it’s significant enough of a problem to devote the required resources to it.
CFIS ingests all the financial records that would be associated with a certain criminal or target of an investigation—banks, brokerages, credit cards, invoices, phone records, all of this data. It ingests it at a rapid speed, but it does more than that—it automatically analyzes. Clients tell us that what had taken them several years to do, they can now do in a morning with CFIS. So where we used to have to go through all those records and enter that data into Excel spreadsheets or use some other method, now we can start analyzing to see different pieces of the behavior. CFIS is automatically capturing critical elements of the data—identifying source names, source institutions, organizations, businesses, account numbers—and it’s also seeing the destination. So it rapidly can tell you how much money came from or went to different entities. You’re doing that the day you start pushing the records through—not a year and a half later. That really attacks the resource problem that law enforcement and prosecutors have.
I asked Deehan whether CFIS does anything to prevent white-collar crime, or if it’s strictly an investigation tool that comes into play after the fact. He responded by referring to cases like Enron and WorldCom, and what enabled that level of fraud to come about:
Larry Crumbley, an LSU accounting professor, wrote a textbook on it, and he explained that in the 1980s, the volume of transactions increased dramatically, across industries and across people’s account activities. That forced the accounting firms to become so expensive. The normal procedure was they would test all the data, 100 percent—they would go in and inspect the books and look at every transaction. No one would do it, because it was too expensive, so the accounting firms came up with regression analysis and other types of modeling methodologies and sampling techniques to try to test data. But the problem was that over the years, the junior accountants on those staffs who had to provide all the records, learned the game. They knew how to provide the audit firm with a set of records that excluded the activity in question. So prevention is difficult, due to the volume of data and the fact that they can’t test everything. And if you have an insider who’s cooking the books, it’s impossible unless you test all the data. That’s where CFIS can be very helpful. Companies can spot check—a board of directors can initiate a test, and they have the technology that can rip all the data and look at everything. That’s when you can detect the unusual activity and the fraud.
Finally, I asked Deehan what AIT’s clients tend to cite as areas in which there’s room for improvement with CFIS. His response:
The nature of the investigative world is that it’s dynamic, in the sense that years ago when we started, you would see on a bank statement, “Credits/Debits.” You’d see almost nothing else, in terms of remarks. Now, with Check 21 and the progression ofACH and electronic funds transfer, you have so much more data in the remarks line. So as the transactional activities and the way banks process the records give you different types of data and different formats and layouts, now you’ve got to be able to rapidly adapt and ingest that for them. In any investigation, they’ll often get unique sets of records. So what we normally get [from clients] in terms of requests is figuring out how to address these unique sets of data. It’s responding to constant changes in media and formats.