关键词: 风险对冲, 核估计, 在险价值

Abstract: In this paper, we research the risk hedging strategies based on Value-at-Risk under normal distribution and arbitrary distribution settings respectively. Under the assumption of normal distribution, this paper discusses the properties of the hedge ratio of minimum variance strategy and minimum VaR strategy thoroughly and find that the mean and variance of portfolio return induced by minimum VaR strategy are smaller than minimum variance strategy. Without any distribution assumption, this paper constructs a new VaR hedging model which firstly estimates VaR by nonparametric kernel estimation method and then embeds the VaR kernel estimator into minimum VaR fomula. By this way, we achieve the goal that risk estimation and risk hedging are implemented simultaneously. The empirical results are very robust to show that the kernel estimation method can reduce risk more effectively than minimum variance strategy and minimum VaR strategy under normal distribution setting.

Key words: risk hedging, kernel estimation, value-at-risk