Empirical Research On Stock Index Future' Margin Setting Of China Based On Extreme Value Theory

The existence of margin money guarantees the safeness and high availability of futures bargain. Value at Risk (VaR) is now one of the most popular methods which are used to manage financial risk in the world. Many corporations and scholars take great efforts to study this project. However, in the most methods of VaR calculation, normal distribution of returns is unquestionably chose to serve as the elementary hypothesis. Many empirical studies indicate that the real distribution of the percentage price change is not normal distribution for it has palpable fatter tails and a thinner waist. So the VaR on the basis of normal distribution often leads to underestimation of the real risk. In this paper a new algorithm based on the Extreme Value Theory are applied to conquer the shortcoming of normal distribution hypothesis in order to increase the accuracy of VaR estimation.This paper examines the proper margin level of the future contract of Hushen 300 index and Shanghai 50 index, applies generalized Pareto distribution inherited from extreme value theory to examine the prudent margin policy for price extreme movements. This paper suggests selecting expected shortfall under GPD as basis of margin level setting, and considering setting different margin level for long and short positions. We also improve the model at the last part, which is associating extreme value theory with g-h distribution to examine the margin level.