面向高頻量化交易的滬深300股指期貨跨期套利研究
發(fā)布時(shí)間:2018-10-29 15:06
【摘要】:國(guó)外的股指期貨已有近30年的歷史,而我國(guó)大陸的第一支也是唯一一支股指期貨——滬深300股指期貨是2010年4月16日才發(fā)行的,歷史很短,,但一經(jīng)推出就成為研究熱點(diǎn)。目前比較熱門(mén)的研究集中在利用股指期貨進(jìn)行期現(xiàn)套利和跨期套利兩方面。 本文主要研究滬深300股指期貨的跨期套利,選取近期交易的實(shí)際數(shù)據(jù)進(jìn)行建模分析。本文選取滬深300股指期貨日內(nèi)每五分鐘收盤(pán)價(jià)這一高頻數(shù)據(jù)作為研究對(duì)象,選擇兩種目前領(lǐng)域內(nèi)最為熱門(mén)的兩種模型——持有成本模型和協(xié)整模型建立模型進(jìn)行實(shí)證分析。對(duì)模型中涉及的參數(shù)結(jié)合目前國(guó)內(nèi)市場(chǎng)的實(shí)際情況加以設(shè)定,得出了比較好的實(shí)驗(yàn)結(jié)果,發(fā)現(xiàn)了一些套利的機(jī)會(huì)。在協(xié)整模型中還根據(jù)風(fēng)險(xiǎn)因素對(duì)套利機(jī)會(huì)加以過(guò)濾,發(fā)現(xiàn)了比較穩(wěn)定的套利機(jī)會(huì),對(duì)模型有一定的改進(jìn)和完善。 在持有成本模型方面,根據(jù)國(guó)內(nèi)市場(chǎng)的特點(diǎn)對(duì)一些參數(shù)進(jìn)行了設(shè)定,在對(duì)數(shù)據(jù)進(jìn)行跨期套利時(shí)得到了比較好的結(jié)果,發(fā)現(xiàn)了一些正向和反向的套利機(jī)會(huì)。在協(xié)整模型方面,除了對(duì)兩組期貨合約的價(jià)格指數(shù)進(jìn)行協(xié)整分析,還對(duì)均衡誤差項(xiàng)即為價(jià)差建立了顯著的ARMA模型,使用這一模型的靜態(tài)預(yù)測(cè)功能能夠在很大程度上預(yù)測(cè)出下一時(shí)間點(diǎn)價(jià)差的走勢(shì)。利用ARMA模型的預(yù)測(cè)功能能夠從協(xié)整模型發(fā)現(xiàn)的大量套利機(jī)會(huì)找到較為穩(wěn)定的套利機(jī)會(huì),對(duì)模型進(jìn)行了改進(jìn) 通過(guò)對(duì)兩模型的理論和實(shí)證分析,對(duì)二者發(fā)現(xiàn)的套利機(jī)會(huì)進(jìn)行了對(duì)比,發(fā)現(xiàn)針對(duì)本文所選擇的數(shù)據(jù),兩種模型都是有效的。但是協(xié)整模型更加準(zhǔn)確和敏感能夠發(fā)現(xiàn)更細(xì)微的套利機(jī)會(huì)。
[Abstract]:The stock index futures of foreign countries have a history of nearly 30 years, and the first stock index futures of mainland China is the only stock index futures, the Shanghai and Shenzhen 300 stock index futures was issued on April 16, 2010, the history is very short, but once it is launched, it has become a research hotspot. At present, the popular research focuses on using stock index futures to carry out current arbitrage and cross-term arbitrage. This paper mainly studies the intertemporal arbitrage of Shanghai and Shenzhen 300 stock index futures, and selects the actual data of recent trading to model and analyze. This paper selects the high frequency data of Shanghai and Shenzhen 300 stock index futures every five minutes closing price as the research object, and chooses two most popular models in the field at present, holding cost model and co-integration model, to make empirical analysis. The parameters involved in the model are set according to the actual situation of the domestic market at present, and some good experimental results are obtained, and some opportunities of arbitrage are found. In the co-integration model, the chance of arbitrage is filtered according to the risk factors, and the stable arbitrage opportunity is found, which improves and perfects the model to some extent. In the aspect of holding cost model, some parameters are set according to the characteristics of domestic market, and some positive and reverse arbitrage opportunities are found when the data is arbitraged over time. In terms of cointegration model, in addition to the cointegration analysis of the price indices of two groups of futures contracts, a significant ARMA model is also established for the equilibrium error term, that is, the price difference between the two groups of futures contracts. The static prediction function of this model can predict the trend of the next time point spread to a great extent. By using the prediction function of ARMA model, we can find stable arbitrage opportunities from a large number of arbitrage opportunities found by cointegration model, and improve the model through theoretical and empirical analysis of the two models. The paper compares the arbitrage opportunities found by the two models and finds that both models are valid for the data selected in this paper. But cointegration models are more accurate and sensitive to detect more subtle arbitrage opportunities.
【學(xué)位授予單位】:哈爾濱工業(yè)大學(xué)
【學(xué)位級(jí)別】:碩士
【學(xué)位授予年份】:2012
【分類(lèi)號(hào)】:F832.51;F224
本文編號(hào):2298031
[Abstract]:The stock index futures of foreign countries have a history of nearly 30 years, and the first stock index futures of mainland China is the only stock index futures, the Shanghai and Shenzhen 300 stock index futures was issued on April 16, 2010, the history is very short, but once it is launched, it has become a research hotspot. At present, the popular research focuses on using stock index futures to carry out current arbitrage and cross-term arbitrage. This paper mainly studies the intertemporal arbitrage of Shanghai and Shenzhen 300 stock index futures, and selects the actual data of recent trading to model and analyze. This paper selects the high frequency data of Shanghai and Shenzhen 300 stock index futures every five minutes closing price as the research object, and chooses two most popular models in the field at present, holding cost model and co-integration model, to make empirical analysis. The parameters involved in the model are set according to the actual situation of the domestic market at present, and some good experimental results are obtained, and some opportunities of arbitrage are found. In the co-integration model, the chance of arbitrage is filtered according to the risk factors, and the stable arbitrage opportunity is found, which improves and perfects the model to some extent. In the aspect of holding cost model, some parameters are set according to the characteristics of domestic market, and some positive and reverse arbitrage opportunities are found when the data is arbitraged over time. In terms of cointegration model, in addition to the cointegration analysis of the price indices of two groups of futures contracts, a significant ARMA model is also established for the equilibrium error term, that is, the price difference between the two groups of futures contracts. The static prediction function of this model can predict the trend of the next time point spread to a great extent. By using the prediction function of ARMA model, we can find stable arbitrage opportunities from a large number of arbitrage opportunities found by cointegration model, and improve the model through theoretical and empirical analysis of the two models. The paper compares the arbitrage opportunities found by the two models and finds that both models are valid for the data selected in this paper. But cointegration models are more accurate and sensitive to detect more subtle arbitrage opportunities.
【學(xué)位授予單位】:哈爾濱工業(yè)大學(xué)
【學(xué)位級(jí)別】:碩士
【學(xué)位授予年份】:2012
【分類(lèi)號(hào)】:F832.51;F224
【參考文獻(xiàn)】
相關(guān)期刊論文 前5條
1 仇中群;程希駿;;基于協(xié)整的股指期貨跨期套利策略模型[J];系統(tǒng)工程;2008年12期
2 吳ng超;股指期貨標(biāo)的指數(shù)設(shè)計(jì)的國(guó)際比較與借鑒[J];北方經(jīng)貿(mào);2003年11期
3 趙聰;俞熹;;滬深300股指期貨跨期套利價(jià)差的R/S分析[J];科學(xué)技術(shù)與工程;2010年33期
4 王春峰;張龍斌;房振明;;股指期貨對(duì)沖比率和對(duì)沖期限關(guān)系的多尺度研究[J];系統(tǒng)工程理論與實(shí)踐;2009年01期
5 李世偉;;基于協(xié)整理論的滬深300股指期貨跨期套利研究[J];中國(guó)計(jì)量學(xué)院學(xué)報(bào);2011年02期
本文編號(hào):2298031
本文鏈接:http://sikaile.net/guanlilunwen/zhqtouz/2298031.html
最近更新
教材專(zhuān)著