PTA期货市场有效性及交易策略实证研究

发布时间：2017-12-27 08:29

本文关键词：PTA期货市场有效性及交易策略实证研究　出处：《山东大学》2017年硕士论文　论文类型：学位论文

【摘要】：经过多年发展,现在国内期货市场交易品种繁多交易规模庞大,尤其是就商品期货而言。各个期货品种的价格会受到来自国内和国际上政治变动、经济变化和自然灾害突发等诸多方面的影响。在这种形势下,期货类产品的管理过程中,决定是否交易某一品种和选择合适理性的交易策略都是值得关注的问题。本论文尝试在验证PTA期货市场的半强式有效性的基础上,利用宏观指标和行业指标等公开信息预测PTA期货价格未来市场行情,为投资者们选择交易品种和制定交易策略提供一些参考。首先,本论文探讨PTA期货市场流动性特点,依据主力合约生成连续价格序列以供后续分析。实证结果表明PTA(精对苯二甲酸)期货主力合约对应的月份主要是一月、五月和九月,而且它们持续作为主力合约的时间长度并不固定,时多时少,没有规律。但是三个月份合约成为主力合约后平均持续时间都在80个交易日,也就是4个月左右。然后,选取单位根检验法验证PTA期货市场是否已经是弱势有效的。结果显示:ADF统计量的数值是-0.776812,明显大于1%置信度下的临界值-2.566325。从而说明PTA期货合约对数价格序列是一个单位根过程,服从随机游走假设,PTA(精对苯二甲酸)期货市场已经是弱势有效的。接下来,运用事件研究法验证PTA期货市场是否达到了半强式有效。本论文选取"英国脱欧"作为目标事件,估计期间包括122个交易日,事件期间包括13个交易日。在此基础上,计算出正常收益率和异常收益率,再检验累积异常收益率与0之间的差异性。检验结果显示:在事件日2016年6月24日以后的T统计量数值都明显大于1%置信度下的临界值。这说明累计收益率均值与0有显著差别,"英国脱欧"事件对PTA期货价格走势有显著影响。从而证明PTA(精对苯二甲酸)期货市场不具备半强式有效性,理论上来讲可以利用事件研究策略获利。最后,利用二元逻辑回归模型预测PTA期货未来市场行情。根据PTA产业链情况,选取货币供应量实际值和预测值、生产物价指数预测值和全部工业品当月同比4个宏观指标,PPIRM化工原料类当月同比、PPIRM纺织原料类当月同比、布伦特原油价格变化率和聚酯切片市场均价变化率4个行业指标作为自变量,运用二元逻辑回归模型预测PTA期货市场的未来行情。实证结果显示:引入了交互效应之后的模型在拟合优度上优于原始模型。引入了交互效应之后的逻辑回归模型整体的伴随概率是0.00(小于显著性水平5%),预测正确率是86.9%。由此得出结论:在PTA期货市场可以依赖逻辑回归模型预测PTA期货未来市场行情,并据此制定相应交易策略。
[Abstract]:After years of development, there are a wide variety of transactions in the domestic futures market, especially for commodity futures. The prices of various futures varieties will be affected by many aspects, such as political changes at home and abroad, economic changes and natural disasters. In this situation, it is a matter of concern to decide whether to deal with a certain variety and choose a reasonable and rational trading strategy in the process of the management of futures products. This paper attempts to verify the semi strong efficiency of PTA futures market, and predict future price of PTA futures market based on public information such as macro indicators and industry indicators, so as to provide some references for investors to choose trading varieties and formulate trading strategies. First, this paper discusses the liquidity characteristics of the PTA futures market, and generates a continuous price sequence based on the main contract for the follow-up analysis. The empirical results show that the corresponding month of PTA main contract is mainly in January, May and September, and their duration as main contract is not fixed. But the average duration of the three month contract as the main contract is on 80 trading days, that is, about 4 months. Then, it selects the unit root test to verify whether the PTA futures market is weak and effective. The results show that the value of ADF statistics is -0.776812, which is obviously greater than the critical value of the 1% confidence level of -2.566325. It shows that the logarithmic price sequence of PTA futures contract is a unit root process. It obeys the hypothesis of random walk, and the PTA (PTA) futures market is already weak and effective. Next, we use the event study method to verify whether the PTA futures market has reached a semi strong effect. This paper selects "Britain to take off Europe" as the target event, which is estimated to include 122 trading days, and the event includes 13 trading days. On this basis, the normal rate of return and the abnormal return are calculated, and then the difference between the cumulative abnormal return and 0 is tested. The test results show that the value of the T statistics after the June 24, 2016 event is obviously larger than the critical value under the 1% confidence level. This shows that the mean value of cumulative return is significantly different from that of 0. The "British deeurope" event has a significant impact on the price trend of PTA futures. It is proved that the futures market of PTA (fine terephthalic acid) does not have a semi strong validity. In theory, it can make profit by using the event research strategy. Finally, the two yuan logical regression model is used to predict the future market market of PTA futures. According to the PTA industry chain, choose money supply actual value and predictive value, production price index forecast value and the total industrial products month up 4 macroeconomic indicators, PPIRM chemical raw materials, textile raw materials PPIRM month year month year, Brent crude oil price change rate and market price change rate of polyester slice 4 industry index as the use of two independent variables, binary logistic regression model to predict the PTA futures market in the future market. The empirical results show that the model introduced after the interaction effect is superior to the original model in the goodness of fit. The overall adjoint probability of the logistic regression model after the introduction of the interaction effect is 0 (less than the level of saliency 5%), and the prediction accuracy is 86.9%. It comes to the conclusion that the PTA futures market can rely on the logistic regression model to predict the future market market of PTA futures and formulate corresponding trading strategies accordingly.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F767;F724.5

【相似文献】