Many insurers are redesigning their strategies to deliver systems that forestall attempted fraud, adapt to changing fraud techniques, and are applied at all points of interaction between a policyholder and an insurer. This change is driven by the alarming increase in professional fraud networks in the last five years and the realization that an effective fraud strategy can provide a competitive advantage in the currently very difficult market conditions.
A number of technologies have now matured to a level where, when used in combination, they offer insurers highly effective means of detecting and preventing even the most sophisticated fraud schemes. As a result, these technologies—which include predictive analytics, link analysis, text mining, in-memory database, and cloud services—will become significant areas of investment for insurers during the next 12–24 months.
A recent Ovum report, Tackling Insurance Fraud, discusses the reasons behind the increased focus on insurance fraud and explains how a range of technologies can be applied to implement a comprehensive and effective insurance fraud system.
Professional fraudsters use sophisticated schemes—such as staged or induced auto accidents, life insurance owned by strangers, or false hit-and-run claims—to illegally obtain significant sums from insurers. Professional fraud schemes often involve complex networks of criminal players within medical services providers, auto repair centers, hospitals, insurance agents, or even insurance companies. The volume of organized professional fraud is low in comparison with that of opportunist fraud by amateurs, but the sums being claimed illegally by an individual group can amount to many tens of thousands of dollars and, in some cases, millions.
The vast majority of insurers have already invested, to some degree, in fraud technology. Although this investment has delivered benefits, it has tended to take a piecemeal approach. Investment is usually focused only on the claims phase. Technology used today is generally not sufficient to address the growing problem of professional insurance fraud.
The continued fragility of many developed economies means that insurance markets will remain extremely competitive for at least the next 36 months, with only muted premium growth, limited room for rate increases, and investment returns that will remain volatile. So, insurers are urgently seeking ways to significantly and sustainably reduce both administration and claims costs. As claims payments typically account for 80% of an insurer's costs (excluding administration costs), reducing them by decreasing the level of fraudulent payouts can have a significant impact on a carrier’s cost base and margins.
With the Association of British Insurers (ABI) estimating that fraud currently adds approximately £50 to each auto policy, an effective fraud strategy can drive significant competitive advantage in a tough market by allowing insurers to reduce premium levels. An effective fraud strategy can also bring additional service benefits. In particular, simplifying and expediting the processing of legitimate claims increases the likelihood that a policy-holder will renew a policy.
There is no single “silver bullet” technology that can fully address the issue of complex insurance fraud. However, technology areas such as predictive analytics, text mining, link analysis, in-memory databases, and cloud services, together with fraud technologies commonly used today (such as rules-based systems and anomaly detection) can be combined to create highly effective systems that detect even the most sophisticated and complex fraud schemes. These systems are able to detect and adapt even to previously unknown fraud techniques that may be employed by professional fraudsters.
To date, most insurers have focused their fraud strategies on the claims process. While this is a critical point at which to detect potential fraud, the effectiveness of a fraud strategy, particularly in avoiding organized criminal fraud, can be significantly enhanced by using technology across the entire insurance product lifecycle. It is now possible to apply technology in real time, at multiple points at which insurers and policyholders interact. For example, the use of link analysis can stop fraud by identifying applicants with connections to others that have either committed fraud or are suspected of doing so.
Professional fraudsters can be very sophisticated. Insurers need to keep rearming themselves if they are to win the battle and help themselves thrive in a tough market.