Mean-Variance Portfolio Selection For DC Pension Plans Under Multivariate GARCH Schemes

China has gradually stepped into the aging society, resulting in the increasing prominence of people’s elderly insurance problems. China’s DC pension plans, including the so-called "Enterprise Annuity" and "Occupational Pension", are playing a more important role as the second pillar of endowment insurance system. Along with the rapid development of DC pension funds in scale, how to realize the maintenance and appreciation of pension funds has become a hot spot home and abroad.This paper investigates the investment portfolio problem for DC pension funds under the mean-variance models and Multivariate GARCH models for assets’return rates. First, we obtain the dynamic asset allocation of DC pension funds under the multi-period mean-variance model with stochastic income and mortality. After that we choose typical indices’ rates representing various assets’returns and then simulate and forecast the returns using GARCH group models. Finally, we conduct empirical analysis on investment portfolio of DC pension funds, obtain the expression of minimum variance and analyze the dynamic asset allocation for each year on various assets.