Industry fights back against ‘untenable’ bogus commercial claims

Industry fights back against ‘untenable’ bogus commercial claims

Industry fights back against ‘untenable’ bogus commercial claims The insurance industry is fighting back against the old habit of paying bogus claims to avoid costly investigations, by implementing predictive analytics to forecast and prevent fraud in commercial lines.

Paul Bermingham, executive director of claims at Xchanging Group, said the industry cannot keep paying claims to avoid costly investigations, recommending instead they must start pooling information from various sources to fight fraud.

Evidence suggests the industry is beginning to do just that by implementing predictive analytics (using methods such as data mining) to make predictions about future unknown events.

“Insurers will sometimes tolerate paying what is suspected to be an inflated claim because they want to avoid the costs associated with a prolonged investigation. This situation has become untenable due to increasing levels of fraud stemming from economic uncertainty and the financial pressures that come with it.

“Predictive analytics can help by identifying policies that carry a high risk of yielding fraudulent claims, giving insurers the option of declining to write those policies.”

Bermingham said predictive analytics provide insurers and brokers with the necessary information to build accurate predictive models that can be embedded into business processes and avoid fraudulent settlement payouts.

Information used in predictive analytics can be retrieved from multiple sources including the crime bureaux, which includes the NSW Bureau of Crime Statistics and Research, the National Crime Authority and the Australian Institute of Criminology. Companies can also make use of internal databases and company websites – all of which can be used to build customer profiles and detect potential fraudulent behaviour.

 “Wherever brokers and insurers have access to a dataset of historic information, it is now possible to analyse and predict activity based on that history," he said.

“Using effective data stored at the point of underwriting and working through fraud triaging at first notification of loss, can dramatically improve rates of detection and prevent expensive litigation later in the claims process.”