可控损失的股指期货套利研究

发布时间：2018-05-31 19:08

本文选题：股指期货 + 可控损失　；参考：《华南理工大学》2013年硕士论文

【摘要】：我国沪深300股指期货自推出后近三年来，市场逐渐成熟，无风险套利机会与利润空间已变得十分狭窄。在此背景下，本文提出了可控损失的股指期货套利策略新模式，建立了较为完整的一套股指期货套利量化模型，并运用期现货市场高频数据进行了严谨的实证分析，为套利者在成熟市场上运用该新的交易模式提供了初步的理论框架。首先,本文以期现套利为研究对象，用股指期货当月连续合约和沪深300指数1分钟数据对股指期货自运行以来的32个合约进行了分析。结论表明，股指期货推出当年存在大量期现套利机会且收益可观。但此后两年，无风险套利机会已大为减少，收益显著降低，期现套利不再是良好投资选择。在此现实条件下，本文提出了可控损失的股指期货套利新模式并对其主要内涵进行了详细阐述。第一，，模型允许套利者在不能确保不亏损的点位进场开仓，通过优化平仓策略以使夏普比率最大化。第二，不同的进场点位下套利者交易最大损失是不同的，但它是可控的，可以事先计算最大可能亏损值。每一个进场点位就会有其唯一对应的交易最大损失和最大夏普比率，该夏普比率与交易最大损失构成了模型交易策略。第三，对模型下所有有效交易策略进行组合就可以得到模型投资组合策略及有效边界，以供具有不同风险偏好的套利者决策。随后，本文依据金融时间序列理论对该模型进行了数理论证。本文分别假定资产或资产收益率序列满足白噪声过程、单位根过程，ARMA模型、条件异方差模型以及随机波动模型，对模型下的开平仓信号出现概率进行了数理分析与计算，讨论了不同进场策略下的交易最大损失、最大夏普比率、投资有效边界，保证了模型的理论严谨性。最后，本文对模型进行了历史数据回测。结果表明，套利者只需在较低的交易最大风险下就可获得可观的投资收益。因此，模型有相当的应用价值，为套利者在成熟市场条件下提供了新的选择。
[Abstract]:Since the launch of CSI 300 stock index futures in recent three years, the market has gradually matured, and risk-free arbitrage opportunities and profit margins have become very narrow. Under this background, this paper puts forward a new arbitrage strategy model of stock index futures with controllable loss, establishes a complete set of quantitative model of stock index futures arbitrage, and makes a rigorous empirical analysis using high-frequency data of spot market in the future. It provides a preliminary theoretical framework for arbitrage to use the new trading model in mature markets. Firstly, with the aim of arbitrage, this paper analyzes 32 contracts of stock index futures since its operation by using the continuous contract of stock index futures that month and the 1-minute data of Shanghai and Shenzhen 300 index. The conclusion shows that there are a lot of arbitrage opportunities and considerable returns in the year of stock index futures launch. But over the next two years, risk-free arbitrage opportunities have been significantly reduced, earnings significantly reduced, and current arbitrage is no longer a good investment option. Under this realistic condition, this paper puts forward a new arbitrage model of stock index futures with controllable loss and expounds its main connotation in detail. First, the model allows arbitrage to open at points where no loss can be ensured, and optimizes the liquidation strategy to maximize Sharp ratio. Second, the maximum loss of arbitrage under different entry points is different, but it is controllable and can calculate the maximum possible loss in advance. Each entry point has its unique corresponding maximum loss and maximum Sharp ratio, which constitutes the model trading strategy. Thirdly, the portfolio strategy and the efficient boundary can be obtained by combining all the effective trading strategies under the model for arbitrageurs with different risk preferences. Then, according to the theory of financial time series, this paper makes mathematical proof of the model. In this paper, assuming that the asset or asset return sequence satisfies the white noise process, the unit root process ARMA model, the conditional heteroscedasticity model and the stochastic volatility model, the probability of the open position signal under the model is numerically analyzed and calculated. The maximum transaction loss, maximum Sharpe ratio and investment efficient boundary under different approach strategies are discussed, and the theoretical rigor of the model is ensured. Finally, the historical data of the model are measured back in this paper. The results show that the arbitrage can gain considerable investment returns only at the lowest maximum risk of the transaction. Therefore, the model has considerable application value and provides a new choice for arbitrage under mature market conditions.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F832.51;F224

【参考文献】