With the development of China's enterprise annuity, the growing accumulation of fund scale in operation increased the risk. Risk measurement has become a focus which people care about. This article is designed to fit the linear time series model of enterprise annuity index yield to predict yield, to estimate the bounds of loss probability.In chapter 1, we introduce the concepts of enterprise annuity and the calculation of investment yield, as well as with the time series related to prior knowledge.In chapter 2, we create a linear time series model( we call it model 1) for the yield of enterprise annuity index by analyzing history data, and use it to forecast future values.In chapter 3, because Model 1 has a poor fit effect, we need to find ways to improve it( with model 2): the use of China's bond yields and the enterprise annuity yields to establish a cross-linear model. We analyze the cross correlativity which is between the enterprise annuity index yield and China's bond index yield. Facts have proven that using China's bond index yield to predict the enterprise annuity index yield indeed has a better effect.In chapter 4, we derive the exponential upper bounds of loss probability under some different conditions.In this paper, we only study the discrete time series, and create linear models for them. |