Property Insurance Underwriting Plan Risk Early Warning Model Research

With the development of science and technology, and the progress of society, insurance company underwrites more and more diversified and complicated risk, which requires high level of insurance company' risk management. Risk control is an important component of risk management for insurance company. The improvement of risk control technology can not only reduce the loss ratio of insurance company, but also plays a positive role in controlling the accidents of enterprise.In this paper, risk early warning theory was used in property insurance risk control. According to the acceptable risk level of underwriting plan, and the underlying own risk condition, with the application of fault tree analysis and neural network methods, the early warning model was established to control the change of underwriting risk. First of all, by using RAC method, the high risk accident was identified; Secondly, through the method of fault tree and signal-to-noise, the risk factors which effect the fluctuation of underwriting risk were determined, and risk early warning surveillance index were identified; Finally, with the application of neural network method, indicators were real-time monitored. When the risk factor's variation exceeds the underwriting risk threshold, the risk warning system alarms automatically. According to the result of early warning system, insurance company would take corresponding risk control measures to control insured risk, and achieve the goal of scientific risk control. Finally, making an example of a market, proved the practicability of the early warning model.