基于跳-扩散模型的DC型养老金时间一致最优投资策略的研究

将养老金投资过程分成财富积累阶段和财富给付阶段,建立了DC型养老金在退休前和退休后个人账户积累额变动的连续时间随机模型.该模型考虑了工资的随机风险因素,并用跳扩散模型刻画风险资产.以均值方差准则作为优化目标,运用推广的HJB方程分别得到了退休前和退休后的时间一致最优风险资产投资最优解.最后通过算例及敏感性分析研究了各个因素对风险资产投资的影响.在这些因素中缴费比例、死亡力对风险资产投资比例均有负向影响.

Time-Consistent Optimal Investment Strategy for DC Pension Plan Under Jump-Diffusion Model

The pension investment process is divided into two stages: the wealth accumulation stage and the wealth payment stage. Continuous time stochastic models of the amount of DC pension accumulation in individual accounts before and after retirement are established. The models consider the stochastic risk factors of wages and jump-diffusion model is used to portray risky assets. Taking the mean-variance criterion as the optimization objective, the optimal solution of venture capital investment before and after retirement was obtained by using the generalized HJB equation. Finally, through numerical examples and sensitivity analysis, the influences of various factors on venture capital investment are analyzed. Among these factors, contribution ratio and death force have negative influences on the proportion of venture capital investment.