基于协整理论的DFT-GARCH模型的统计套利研究
|
摘要
现有的统计套利策略大多建立在协整理论和GARCH模型的基础上.离散Fourier变换(DFT)的思想可以挖掘价差序列周期性、非线性的特征,保证其在拟合和预测中的精确度.利用沪铜期货合约的收盘价数据进行实证分析,研究结果表明:在高频数据下,新模型对数据的拟合和预测效果要明显优于传统的套利模型,在相同的交易规则下,新模型的套利成功率和收益率都高于传统的统计套利模型.
Most of the existing statistical arbitrage strategies are based on the cointegration theory and the GARCH model.The idea of discrete Fourier transform(DFT)can exploit the periodic and nonlinear characteristics of the spread sequence to ensure its accuracy in fitting and prediction.Using the closing price data of copper futures contract in the Shanghai futures exchanges for empirical analysis,the results show that under the high-frequency data,the new model fits and predicts the data better than the traditional arbitrage model.Under the same trading rules,the arbitrage success rate and yield of the new model are higher than the traditional one.
Most of the existing statistical arbitrage strategies are based on the cointegration theory and the GARCH model.The idea of discrete Fourier transform(DFT)can exploit the periodic and nonlinear characteristics of the spread sequence to ensure its accuracy in fitting and prediction.Using the closing price data of copper futures contract in the Shanghai futures exchanges for empirical analysis,the results show that under the high-frequency data,the new model fits and predicts the data better than the traditional arbitrage model.Under the same trading rules,the arbitrage success rate and yield of the new model are higher than the traditional one.
引文
[1]朱丽蓉,苏辛,周勇.基于我国期货市场的统计套利研究[J].数理统计与管理,2015,34(4):730-740.
[2]HANSON T A,HALL J.Statistical Arbitrage Trading Strategies and High Frequency Trading[J].Social Science Electronic Publishing,2012,49(2):177-202.
[3]KRAUSS C,DO X A,HUCK N.Deep neural networks,gradient-boosted trees,random forests:Statistical arbitrage on the S&P 500[J].European Journal of Operational Research,2016,259(2):689-702.
[4]仇中群,程希骏.基于协整的股指期货跨期套利策略模型[J].系统工程,2008,26(12):26-29.
[5]常宗琪.白糖统计套利理论模式研究及实例分析[J].经济师,2008,2008(11):30-31.
[6]韩广哲,陈守东.统计套利模型研究——基于上证50指数成份股的检验[J].数理统计与管理,2007,26(5):908-916.
[7]孙维.对深圳股票市场有效性的实证研究[J].经济研究导刊,2009(18):73-73.
[8]赵莉.基于GARCH模型的沪深300指数收益率波动性分析[D].成都:成都理工大学管理科学学院,2012.
[9]何树红,张月秋,张文.基于GARCH模型的股指期货协整跨期套利实证研究[J].数学的实践与认识,2013,43(20):274-279.
[10]梁斌,陈敏,缪柏其,黄意球,陈钊.基于LARS-Lasso的指数跟踪及其在股指期货套利策略中的应用[J].数理统计与管理,2011,30(6):1104-1113.
[11]刘阳,李艳丽,陆贵斌.基于信息更新NN-GARCH模型的统计套利研究[J].统计与决策,2016(2):169-171.
[12]张波,刘晓倩.基于EGARCH-M模型的沪深300股指期货跨期套利研究——一种修正的协整关系[J].统计与信息论坛,2017,32(4):34-40.
[13]雷井生,林莎.基于高频数据的统计套利策略及实证研究[J].科研管理,2013,34(6):138-145.
[14]覃良文,唐国强,林静.基于协整-GARCH模型最优阈值统计套利研究[J].桂林理工大学学报,2016,36(3):625-631.
[2]HANSON T A,HALL J.Statistical Arbitrage Trading Strategies and High Frequency Trading[J].Social Science Electronic Publishing,2012,49(2):177-202.
[3]KRAUSS C,DO X A,HUCK N.Deep neural networks,gradient-boosted trees,random forests:Statistical arbitrage on the S&P 500[J].European Journal of Operational Research,2016,259(2):689-702.
[4]仇中群,程希骏.基于协整的股指期货跨期套利策略模型[J].系统工程,2008,26(12):26-29.
[5]常宗琪.白糖统计套利理论模式研究及实例分析[J].经济师,2008,2008(11):30-31.
[6]韩广哲,陈守东.统计套利模型研究——基于上证50指数成份股的检验[J].数理统计与管理,2007,26(5):908-916.
[7]孙维.对深圳股票市场有效性的实证研究[J].经济研究导刊,2009(18):73-73.
[8]赵莉.基于GARCH模型的沪深300指数收益率波动性分析[D].成都:成都理工大学管理科学学院,2012.
[9]何树红,张月秋,张文.基于GARCH模型的股指期货协整跨期套利实证研究[J].数学的实践与认识,2013,43(20):274-279.
[10]梁斌,陈敏,缪柏其,黄意球,陈钊.基于LARS-Lasso的指数跟踪及其在股指期货套利策略中的应用[J].数理统计与管理,2011,30(6):1104-1113.
[11]刘阳,李艳丽,陆贵斌.基于信息更新NN-GARCH模型的统计套利研究[J].统计与决策,2016(2):169-171.
[12]张波,刘晓倩.基于EGARCH-M模型的沪深300股指期货跨期套利研究——一种修正的协整关系[J].统计与信息论坛,2017,32(4):34-40.
[13]雷井生,林莎.基于高频数据的统计套利策略及实证研究[J].科研管理,2013,34(6):138-145.
[14]覃良文,唐国强,林静.基于协整-GARCH模型最优阈值统计套利研究[J].桂林理工大学学报,2016,36(3):625-631.