基于VaR与多尺度分析的保险业风险值估计

Risk Estimation of Insurance Based on the Value-at-Risk and Multi-scale Analysis

风险值(Value-at-Risk,VaR)是现今衡量风险的标准。本文利用风险值(VaR)方法来为风险基础资本估计风险,使其能够准确地呈现保险人本身所面临风险的状况,并利于监督机关建立适当的监督预警措施,来保障全体保险人权益并维持金融秩序的稳定。考虑到多尺度变换对估计报酬率风险型态模型无需作假设的优点,且小波变换是一种重要的多尺度分析工具,本文引入小波变换来对非线性的保险数据序列中提取频率域的高频信息,利用多尺度分解的系数得到模型参数,从而实现更加准确的风险值估计。

Value at Risk （VAR） is a common measure for risk estimation. In this paper, we used the value at risk method to estimate the risk for the risk-based capital, so as to make it indicate the risk level of insurers. This method can also help the regulator to establish an appropriate supervision and early warning method to safeguard the inter- ests of all insurers and to maintain the stability of the financial order. Considering that the multi-scale transformation of the estimated rate of return on the risk of type model does not need to make assumptions for the model, and that wavelet transform is an important multi-scale analytic tool, we use the discrete wavelet transform to extract the high- frequency information of the nonlinear reward sequence. And by using the multi-scale coefficients of the sequence, we can obtain relevant model parameters, and then make an accurate estimation of the value at risk.