Several Models Of Evaluating Claim Reserving

Claim reserves, one of insurance reserves, refers the capital reserve in order that insurance company have sufficient capital to compensate for indemnify or pay insurance losses during fiscal account period.In other words, insurance company extract parts of premium from the total premium of the current year to prepare for the next account year in order to insure insurance company bear the future compensation responsibilities.The purpose of extracting reserves is to protect the benefit of the insured and beneficiaries. Claim reserves is not the operating income, but the liability of the insurance company.Since natural and man-made catastrophe losses occurred frequently and with increasing competitive market, insurance company must increase their competitiveness.on the one hand, this inflected in the product design and sales on the other hand reflected in the approach that reserves are extracted and investment strategy.Extract to much reserves will reduce the company’s prof-itability, but to little will not enable insurance company to give full claim, this may lead to financial distress or bankruptcy. So it is important to extract reserves reasonably, calculated by reasonable method will greatly increase the adequacy and accuracy of the assessment of non-life insurance reserves, there-by enhancing the ability of the company to carry out the compensation that insurance company promise.Therefor,evaluating claim reserves is not only the main regulatory goal of the CIRC, but also an important issue of theoretical study.Classic methods that evaluate non-life claim reserves, that is determinis-tic methods, such as chain ladder method, payments per claim incurred,B-F method, PCE method and so on, because of their simplicity, they are widely accepted by insurance companies. But deterministic methods can only get the mean estimate of claim reserves, they are unable to measure the volatility of claim reserves. At the same time, the random variable methods can measure the estimate’s uncertainty, and provide important theoretical basis in deter-mining the liability of the estimate, the best estimation and interval estimation of claim reserves.Firstly, we introduce commonly used methods of claim reserves, and give three stochastic model.In the next section, we demonstrate the reasonableness of that three methods that are used to evaluate claim reserves by some ex-amples.we do empirical analysis for various claim reserves model using SAS, Matlab, VBA software to do data processing, and compare the advantages of the latest methods. In this paper, I use bayesian normality model solved by M-CMC, Bootstrap method, generalized linear model to evaluate claim reserves, then compare.The innovation of this paper is to use WinBUGS software, com-pared with previous literature [1], error is smaller, the parameter is more closer to the results that chain ladder method get.