October 10, 2011

As insurers continue to expand into the direct channel they are encountering new fraud challenges. The anonymity of direct underwriting via the Web gives fraudsters a quick and easy way to obtain policies without the troubles of in-person conversations. But this channel also presents new opportunities for analytics-based fraud detection.

Staged Accident Rings

Claims fraud has been a challenge for auto insurers for years. When organized staged accident rings target an insurer, they must first secure policies in order to stage accidents and file claims. Historically, that meant finding an agent that didn't ask too many questions. But with the advent of direct Internet business, fraudsters have a much easier time obtaining policies for their illicit activities. The bad guys don't waste any time in exploiting this opportunity. Typically, insurers see exposure to organized staged accident insurance fraud rings within 60 days of launching a direct Internet channel. And with a policy in hand, the accident claims and medical bills start rolling in.

Though this channel, there is an opportunity to exploit loopholes in the application and underwriting process. And fraudsters can test system thresholds by filing many different applications to see which ones are flagged for further underwriting and which ones sail through with ease.

Rate Evasion

The direct Internet channel is also a boon to rate-evaders. It is harder to lie to another human being than it is to type some falsehoods into an online application. Don't like the quote you received because your teenage son is driving up your rates? No problem, just erase him from your application and resubmit for a much better deal. Other would-be rate-evaders are more creative. A traditional rate evasion technique is to misrepresent a garaging location. Instead of an expensive urban address, applicants use a rural one to keep their rate low. By testing the system repeatedly, fraudsters have discovered that some rating engines rely solely on zip code for this calculation. On their application they provide a genuine city and state but a false (more rural) zip code. In this case, the applicant was "honest" about their garaging city but simply exploited a loophole in the underwriting process that was easily discovered via the direct Internet channel.

Analytics to the rescue

With no agent or service associate verifying information, insurers are more reliant than ever on technology to aid them in screening business for risk. Real time analytics applications are able to screen new business and renewal applications for various types of fraud risk. Predictive analytics can cull through millions of transactions to help insurers flag the files that have clear and significant risk, those that might require further review, and even those that are low risk and can be pushed straight through.

By leveraging a robust analytical platform, insurers can take a holistic approach to fraud risk, combining intelligence from all parts of the organization, including sales, product, underwriting, claims, and audit. Many organizations currently lack an automated feedback loop from claims to underwriting. Does your claims SIU routinely inform underwriting about the fraud risk associated with particular individuals or businesses so they can be screened on new business? An enterprise fraud framework can support this level of business process integration, helping the organization become more proactive in preventing fraud risk rather than investigating it after the fact. As insurers continue to expand into the direct Internet business, this capability is no longer a luxury but a necessity.

About the Author:James Ruotolo is an insurance fraud technologist, thought leader and the principal for Insurance Fraud Solutions at SAS (Cary, N.C.). Connect with him on Twitter or LinkedIn.