The Research On The Detection Indicators Of Auto Insurance Fraud In Company A

With the rapid increase in auto ownership, auto insurance business also is growing fast. But the auto insurance business has been plagued by high loss ratio. Although it has the characteristics of a high frequency of the dangerous condition, the insurance fraud is a major cause of the high loss ratio. According to the expert and regulatory agencies, nearly 20%-30% of the total payment amount of auto insurance is fraud. Auto insurance fraud has been a serious threat to the survival and development of property insurance company. What is more, it is damage to the road traffic safety, the property of other road users and personal security, seriously disrupting social order. Therefore, the research of auto insurance fraud detection indicator has important theoretical and practical significance.In this thesis, we focus on the detection indicators of auto insurance fraud in A company. Firstly, outlining the concept of auto insurance fraud and characteristic, introducing advantages and disadvantages of the AAG, PROBIT, RIDIT, EXPERT SYSTEM and BP neural networks model. Then we analyze A typical company’s auto insurance fraud cases. A company proposed auto insurance fraud management problems, method leads to the establishment of insurance fraud by identifying indicators to control. Finally, Author selects data of A company as samples and refines 13 auto insurance fraud detection indexes to do empirical analysis. Author finds eight significant recognition indexes, excluding three invalid identification indicators.The empirical results show that the driver whether immediate family members was insured, whether to provide public security organs of the traffic accident liability Confirmation issued by the department of the claim, provide maintenance services of vehicle maintenance business type, the useful life of the insured vehicle, report time; the insured vehicle and its policy is in force incurred time intervals, report time interval dangerous condition of the insured vehicle and the insured has dangerous condition 8 recognition index was significantly associated with the number of fraudulent claims and other policy year within the vehicle.