Study On The Stochastic Programming Model Of Asset & Liability Management Of Social Insurance Fund

With the gradual increase of the elder population, more and more problems appear during the reformation of social insurance fund in China. The most critical one comes from the pressure of payment on schedule. It requires that our government should improve the level of asset & liability management, increase the yield of pension fund, and give a stable social insurance to all of the retired.Based on the analyses of different models of asset & liability management, this paper applies the stochastic programming which has achieved lots of successful applications in the world. Some improvements on the subject constraints of this model were made by the actual information of our country, and then a stochastic programming suited to the domestic circumstances was developed. After the generation of two-stage scenarios tree by the method of the combination of VAR and clustering analysis, which includes interest rate, yield of treasury, increase of stock index, we used the theory and methods of actuary to estimate the debt and payout of pension fund in the two-stage, after these, we applied the stochastic programming model to the asset & liability management, then we obtained the optimization solution to the investment of pension fund in the two-stage, and it also indicated that it's possible to increase the yield of pension fund, to decrease the expenditure rate of people who joined in the social insurance, which would relieve the economic pressure partly of enterprises & residents in our country, to extend the range of social insurance, and make the pension fund to a normal circle.