极值理论在股指期货保证金设定中的应用

发布时间：2018-03-21 16:40

本文选题：股指期货　切入点：保证金水平　出处：《北方工业大学》2013年硕士论文　论文类型：学位论文

【摘要】：在股指期货市场中,保证金的设定对整个市场起到关键性的作用,保证金设置太低,将会造成期货市场投资过度,从而对市场造成危险；保证金设置过高,从而使得市场交易量下降。合理的保证金水平在股指期货交易市场中起到关键作用,使股指期货交易能够高效率和安全进行。传统的正态分布不能很好地对数据尾部加以描述,从而造成了对保证金设定的不合理,使得现实中发生严重亏损的概率远远大于在正态分布下的估计。极值理论对股指期货数据尾部的风险可以很好的进行模拟,从而得到在险价值和期望损失。期望损失相较之在险价值更能够涵盖价格波动,我们选择期望损失作为确定期货保证金水平的依据。本文以沪深300股指期货为研究对象,首先对全样本数据进行正态性检验,通过参数及图形发现不服从正态分布,具有尖峰性和厚尾性,而极值理论中的广义帕累托分布对收益率数据的尾部能够很好的拟合,在广义帕累托参数估计中,对于样本数据采用了bootstrap重抽样的方法克服样本数据量过小的这一不足。对于极值理论计算的保证金水平和重抽样后所得到的保证金水平进行比较,发现后者能够更好的涵盖股指期货的价格波动。在保证金水平的设定过程中,我们对交易双方代表的不同头寸,计算不同的保证金数量,为我国在股指期货交易中的发展提供一些参考。本文认为基于bootstrap的极值理论在保证金的设定中,起作用更加有效。对我国的股指期货市场有一定的指导作用。
[Abstract]:In the stock index futures market, the setting of margin plays a key role in the whole market. If the margin setting is too low, it will cause excessive investment in the futures market, thus causing danger to the market. The reasonable margin level plays a key role in the stock index futures trading market and enables the stock index futures trading to be efficient and safe. The traditional normal distribution can not describe the tail of the data very well. As a result, the margin setting is unreasonable and the probability of serious loss in reality is far greater than the estimate under normal distribution. The extreme value theory can simulate the tail risk of stock index futures data very well. The expected loss can cover the price fluctuation more than the risk value. We choose the expected loss as the basis to determine the futures margin level. This paper takes the CSI 300 stock index futures as the research object. First of all, the whole sample data is tested for normality. Through the parameters and graphs, it is found that there is a spike and a thick tail in the normal distribution, and the generalized Pareto distribution in the extreme value theory can fit the tail of the yield data very well. In the generalized Pareto parameter estimation, the method of bootstrap resampling is used to overcome the shortage of too small sample data. The margin level calculated by extreme value theory is compared with the margin level obtained after resampling. It is found that the latter can better cover the price fluctuation of stock index futures. In the process of setting margin levels, we calculate different amounts of margin for different positions represented by both sides of the transaction. This paper holds that the extreme value theory based on bootstrap plays a more effective role in the setting of margin and has a certain guiding effect on the stock index futures market in China.
【学位授予单位】：北方工业大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F830.91;O211.4;F224

【参考文献】