基于统计套利理论的股指期货跨期套利研究

发布时间：2018-01-18 07:02

本文关键词：基于统计套利理论的股指期货跨期套利研究　出处：《东华大学》2012年硕士论文　论文类型：学位论文

【摘要】：股票指数期货是现代资本市场发展的产物。20世纪70年代,西方各国受石油危机的影响,经济发展十分不稳定,利率波动剧烈导致股票市场价格大幅波动,股票投资者迫切需要一种能够有效规避风险、实现资产保值的金融工具。于是,股票指数期货应运而生。发展到今天,股指期货已经成为世界投资最为活跃的期货交易品种。股指期货市场的套利交易在促使市场价格趋于理性、增加市场的活跃程度方面起着非常重要的作用,是期货市场功能能够得到有效发挥的重要保障。本文以我国推出沪深300股指期货市场为研究背景,基于统计套利的思想研究了沪深300股指期货市场推出初期的跨期套利机会。文章选取沪深300股指期货当月连续合约IFL0与下月连续合约IFL1的5分钟高频数据为研究对象,首先对两合约间的关系进行了协整关系检验。接着检验了持有成本理论下无风险利率和股息率这两个变量与合约间价差波动的关系。结果显示,其对价差波动的解释能力为35%,同时格兰杰因果关系检验表明,其确实是价差波动的格兰杰原因。合约间价差的统计套利建立在价差均值回归的前提条件下,但由于外部变量的变化导致价差均值回复的中枢也会随之变化,文章选取加权移动均值(WMA)来对价差均值回复的中枢进行刻画。同时和一般金融时间序列一样,价差的波动表现为广义自回归条件异方差(GARCH)的特点,文章分别选用GARCH(1,1)与EWMA模型来刻画条件异方差。在确定价差的均值与方差后,文章选用正态分布N(μ1,σ12)来刻画每一时刻价差的分布状态。最后在正态分布的基础上结合Vidyamurthy (2004)的交易机制给出套利交易的开仓时点与平仓时点,建立套利交易策略。文章最后分别应用样本内数据与样本外数据实证检验了套利模型的交易效果。实证套利结果表明,在不考虑股指期货杠杆交易提高资金使用效率的条件下,基于GARCH所刻画的条件异方差,样本内数据累积年化收益率为11.79%,样本外数据累积年化收益率为15.89%；基于EWMA所刻画的条件异方差,样本内数据累积年化收益率为12.30%,样本外数据累积年化收益率为21.33%。总之,取得了较好的收益水平：
[Abstract]:The stock index futures is the product of the development of the modern capital market. In 70s, the western countries were affected by the oil crisis, the economic development was very unstable, the fluctuation of interest rate caused the stock market price to fluctuate sharply. Stock investors are in urgent need of a financial tool that can effectively avoid risks and maintain the value of assets. Therefore, stock index futures emerge as the times require. Stock index futures have become the most active futures trading species in the world. Arbitrage trading in the stock index futures market plays a very important role in promoting rational market prices and increasing the active degree of the market. It is an important guarantee that the function of futures market can be effectively played. This paper takes the introduction of Shanghai and Shenzhen 300 stock index futures market as the research background. Based on the idea of statistical arbitrage, this paper studies the intertemporal arbitrage opportunities in the initial stage of Shanghai and Shenzhen 300 stock index futures market. This paper selects the 5-minute high frequency data of Shanghai and Shenzhen 300 stock index futures' continuous contract IFL0 and next month's continuous contract IFL1 as the research object. Firstly, the cointegration relationship between the two contracts is tested, and then the relationship between the risk-free interest rate and the dividend yield and the fluctuation of the price difference between the contracts under the holding cost theory is tested. Its ability to explain the fluctuation of the spread is 35 and the Granger causality test shows that it is the Granger cause of the fluctuation of the spread. The statistical arbitrage of the price difference between contracts is based on the premise of the price difference mean regression, but because of the change of external variables, the center of the average price difference return will also change. In this paper, the weighted moving mean (WMA) is chosen to describe the center of the average return of the spread. At the same time, it is the same as the general financial time series. The fluctuation of price difference is characterized by generalized autoregressive conditional heteroscedasticity (GARCH(1). 1) describe conditional heteroscedasticity with EWMA model. After determining the mean and variance of the spread, the normal distribution N (渭 1) is selected. 蟽 12) is used to describe the distribution of the price difference at each moment. Finally, based on the normal distribution, the distribution is combined with Vidyamurthy / 2004). The trading mechanism gives the opening point and closing point of arbitrage trade. Establish arbitrage trading strategy. At the end of the paper, we use the data inside the sample and the data outside the sample to test the effect of the arbitrage model. The empirical arbitrage results show that, without considering the leverage trading of stock index futures to improve the efficiency of the use of funds. Based on the conditional heteroscedasticity described by GARCH, the cumulative annualized rate of return of data in the sample is 11.79 and that of the data outside the sample is 15.89; Based on the conditional heteroscedasticity described by EWMA, the cumulative annualized rate of return of data in the sample is 12.30 and that of the data outside the sample is 21.33.
【学位授予单位】：东华大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F832.5;F224

【参考文献】