On yer bike

The ABI reckons that at least one in ten UK motor insurance claims involves some element of fraud. My own company's experience, providing a specialist outsourced claims service to motorcycle insurers and schemes brokers, suggest that bike claims fraud easily matches this ratio. The following offers a brief overview of how patterns of fraud in the motorcycle insurance market differ from those in other market sectors and what can be done to identify and prevent fraudulent claims.

In a variety of different ways - some of them obvious, some of them less so - bikes are not like cars. Under the former heading, it is pretty obvious that bikes generally have half as many wheels, fewer passengers, take up less space on the road, and are more readily transportable.

But claims handlers with significant experience of bikes will understand subtler differences such as the distinctive ways in which bike accidents occur and what types of injury go with what types of damage to the machine. This kind of knowledge can prove invaluable in allowing experienced individuals to pick up on possible instances of fraud.

As far as it goes, that is all well and good. But to get properly to grips with fraud, you cannot simply rely on the insight and instincts of particular individuals who may or may not get sight of fraudulent claims when they arise. What is needed is a more systematic approach.

My own company has been working with fraud detection specialist Conversant Data and systems provider CSC on a bike-specific fraud model that will allow us to automatically check every claim that comes in against a set of 180 specific rules we have put together to highlight instances of potential fraud - as well as referencing all the details logged against both internal and external data sources.

Internally, the system checks for any details that correspond with previous claims received - highlighting, for example, claimants, witnesses, repairers or addresses that crop up repeatedly. Externally, the system checks against databases maintained by the UK's fraud prevention service CIFAS, the DVLA and the Central Underwriting Exchange (CUE) as well as the Motor Insurance Database and the police national computer.

This helps show up the involvement of any names, addresses or other details that should set alarm bells ringing. This all takes place online in a matter of seconds, so that any concerns can be flagged up immediately and acted upon as appropriate. The system generates an overall score based on an evaluation of all relevant factors, allowing efficient prioritisation of those claims that most deserve further investigation and rapid processing of very low-scoring claims.

The rules we have built into the system are based on long experience of motorcycle claims and claims fraud. They are designed to show up the presence of any factor or combination of factors that is strongly suggestive of a possible fraud attempt.

At the most generic level, claims made within a week or two of a policy being taken out may actually relate to damage or theft that has occurred prior to the purchase of insurance.

To take a more specific example, we have programmed in the geographical coordinates of all known racing circuits around the UK and the rules flag up any accidents reported within one mile of these. This can prove a significant detail because experience tells us that bikes damaged while racing (not covered by insurance) are sometimes scooped up and moved to a nearby roadside. Another warning sign, for example, might be the use of unusual accommodation addresses such as pubs or guesthouses.

The rules are extensive and relatively complex, to reflect the many different patterns we have learned to associate with fraud attempts. A good example of this is what is known as the 'eleventh month theft claim'. Manufacturers sometimes offer free - or more often heavily subsidised - insurance with a new bike. New owners can get a bit of a shock when they receive a renewal notice telling them that their second year's insurance is going to cost several thousand pounds instead of the £99 they paid for the first year.

This leads to a suspiciously high incidence of apparent thefts during the final weeks of the original policy. So theft claims at this stage in the policy automatically alert us to the fact that the claim merits a closer look. Beyond that, this type of fraud can effectively be prevented if the insurance provider clearly states that all total losses (at least during the first year) will be settled with the provision of a replacement bike rather than a cash sum which the claimant could then go out and spend on a new bike with a new £99 policy.

Bikers themselves, of course, are not the only perpetrators of bike fraud. Another classic example of fraud is repairer parts fraud, where the cost of parts is subtly - or often not so subtly - inflated to somewhere just inside the write-off ratio. For example, if a bike is worth around £2,000 the repairer will know that the insurer would write off at anything over £1,400, and will therefore add about £100 to the stated cost of each part replaced to bring the cost of the job up to around £1,300.

Unscrupulous repairers know they can get away with this because - in contrast to the four-wheel motor market where insurers' engineers can easily check the replacement cost of parts - finding out the cost of bike parts involves making individual enquiries from dealers and very often does not happen. Again, estimates just inside the write-off ratio may be worth a second look, but this type of fraud can be eliminated by using only trusted specialist repairers who know they will be audited periodically, and that is the approach my own company takes with all service providers.

Bike fraud spans the full spectrum from the relatively trivial to the very serious. At the trivial end, bike-proud owners may have no qualms about taking the opportunity to clean up previous damage when claiming for an accident-damaged bike. An experienced practitioner, however, can spot the difference between collision damage and a paintwork scratch caused over time by the zip on a pillion passenger's leathers! At the other end of the scale, a sophisticated fraud detection methodology, such as that described above, can help uncover systematic organised fraud by gangs involved in complex theft or staged accident claims.

Bike claims fraud has languished for too long as a neglected corner of the motor market. It is high time for detailed specialist knowledge to be applied in rooting out and controlling the dishonest behaviour of the few who cost insurers many thousands of pounds in false claims and inflate the cost of insurance for the honest majority. IT

Dickon Tysoe is director of Bankstone