Fraudulent claims account for a small but significant portion of all claims received by insurers, costing billions of dollars annually. Insurance frauds are diverse, occurring in all areas of insurance. Their severity ranges from slightly exaggerated claims to deliberately caused accidents or damage.

Such is the case with auto insurance fraud, where individuals use a variety of methods to claim more money than they are entitled. Every insurer should realize that one in three auto insurance claims are potentially fraudulent.

The enormity and frequency of fraud occurrence is greater at the claims stage than at the application stage. Fraud may be committed at different points in the claims transaction and involve various parties-the insured, third-party claimants and professionals providing services to claimants. Common fraud includes "padding" or inflating actual claims; misrepresenting of facts on the claim application; submitting claims for injuries or damages that never occurred and "staging" accidents.

The claims department is the first line of defense for insurers to combat fraud. However, the manual detection processes most companies utilize often prove ineffective, since most fraud slips through undetected. Based on the claims processing experiences of insurers, following represents some practical tips to watch for from the insured:

* loss value not justifiable by location, time and nature of accident involved

* injuries (payable under PIP) claimed to be from a certain accident not being in tune with the accident's severity

* considerable delay between the date of accident and the date of the claim filing

* repeat claims from a specific group or from individuals filed at regular intervals

TECHNOLOGY'S ROLE

Technology is the most reliable way to combat fraud. Some emerging technologies insurers should keep on their radar include the following:

Transaction-based Profiling. Credit card vendors effectively employ this method to track customers' spending behavior. This involves collecting, analyzing and transforming client data into a set of features describing the insured's behavior. When mapped to financial profiles, a policyholder's insurance history (including number and type of policies held, number of claims made, etc.) presents a clear picture regarding intent and is essential to tracking customer profiles.

Neural Networks. These models replicate human brain functionality and stimulate situational customer behavior based on certain preset "stress factors" or claim parameters. The output is usually scored from 0 to 1,000, with higher scores indicating a greater likelihood of fraudulent intent. This model is particularly effective since systems can adjust to various sizes of investigative units. Scoring methods help create a range by which potential perpetrators can be scientifically measured.

Data Mining. This requires first mining all data sources for information and then correlating results with known fraudulent claims to indicate relationships to monitor for future claims or new business enrollments. The volume of data to be searched depends on a predefined set of parameters and the reason for the search. As in transaction-based profiling, access to large amounts of data is required to identify necessary information.

Insurers have tried to lessen customer inconvenience through direct repair programs (DRPs), which eliminate a lot of work for the insureds. DRPs aim to improve customer satisfaction, reduce repair time, manage costs and reduce fraud. Through networks of designated repair shops, insurers seek control of repair standards and pricing to provide a speedy, seamless collision claims process. DRPs often act as the claims adjuster, reducing bickering and delays.

Use of generic auto parts in repairs has a long history. During the 1970s, original equipment manufacturers (OEM) were the sole source of crash parts. As after-market parts became more prevalent, OEMs aggressively sought to respond to the threat to their monopoly. After denial of help in Congress, OEMs turned to states.

In 1999, the auto industry targeted 23 states for a range of restrictive legislation and was unsuccessful. To promote competition and keep prices down, some states actually require that insurers use non-OEM parts. This ensures that the burden on insurers (by way of repair bills) are minimized and the purpose of having DRPs in the first place is not defeated.

There is no single profile of fraudulent auto insurance claimants: anyone from professional criminals to ordinary citizens can commit fraud. Risk-mitigating strategies for the insurer should be reliable and effective and be capable of detecting fraud in real-time, when such activities occur. Such strategies should involve effective risk identification, data analysis and reporting, data validation processes, data mining capabilities, visualization techniques and reporting tools to identify questionable behavior before claim payment.

Rajiv Juneja is a business analyst with NIIT Technologies' insurance practice, and David Shaw is the SVP and practice leader for the North American insurance vertical at NIIT Technologies, New Delhi.

Register or login for access to this item and much more

All Digital Insurance content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access

Corrected February 1, 2011 at 1:50PM: yes