我国股指期货价格发现功能的分频和分波段研究

发布时间：2018-02-26 19:23

本文关键词： 股指期现货价格发现高频数据 Morlet小波时频分析向量误差修正模型价格发现贡献度　出处：《暨南大学》2013年硕士论文　论文类型：学位论文

【摘要】：本文从低频日交易数据和高频1分钟交易数据两个方面展开对股指期货价格发现功能的研究。在高频交易数据部分，还对股指不同波段的股指期货的价格发现功能进行了差异分析。在每部分的研究中，首先通过Morlet小波时频分析方法，从时域和频域两个方面，对股指期货和现货收盘价原始序列的动态关联性和价格引导关系进行分析研究。然后通过VEC模型更加量化的研究股指期货的价格发现功能。在时频分析中，使用小波量图和小波互量图对股指期现货时间序列进行分析研究。发现：（1）在日间，股指期现货相互之间的价格引导关系并不明显。（2）在高频数据的分析中，在短周期范围内，股指现货价格引导期货的向上箭头和期货引导现货的向下箭头的大量存在，表明两者存在显著的价格相互引导作用，股指期货的价格发现功能显著。但是在长周期范围内，向右箭头的几乎清一色的存在，也说明股指期现货长期内仍然是协同波动的。在VEC模型的分析中，首先运用E-G两步法对两个变量之间的协整关系进行检验，，在此基础上建立股指期现货之间的向量误差修正模型以判断两者之间价格的领先-滞后关系。最后使用在VECM基础上建立的信息份额模型（I-S模型）和长期短暂模型（P-T模型）对股指期现货各自在市场价格发现中的贡献度进行量化分析。通过分析表明，在股指期货的日间价格发现功能中，股指现货在价格发现所起的作用更大；而使用1分钟高频数据进行建模的结果表明，沪深300股指期货价格领先现货6分钟，且价格发现贡献度要显著大于现货市场。且通过快速上涨波段和下降波段的分析发现，在快速下跌波段，只存在股指期货对现货的单向价格引导，而在上涨阶段，这种价格引导关系是相互的。
[Abstract]:This paper studies the price discovery function of stock index futures from two aspects of low-frequency daily trading data and high-frequency 1-minute trading data. The price discovery function of stock index futures in different bands is also analyzed. In each part of the research, the time domain and frequency domain of stock index futures are analyzed by Morlet wavelet time-frequency analysis method. This paper analyzes the dynamic correlation between stock index futures and spot closing price and the relationship between price guidance and stock index futures, and then studies the price discovery function of stock index futures more quantitatively through VEC model. In time-frequency analysis, wavelet quantification graph and wavelet reciprocal graph are used to analyze the spot time series of stock index period. It is found that: 1) during the day, the price guidance relationship between stock index period spot and spot stock index period is not obvious in the analysis of high frequency data. In a short period, the stock index spot price leads to the existence of a large number of upward arrows of futures and a large number of downward arrows of futures leading to spot, which indicates that there is a significant interaction between the two prices. The price discovery function of stock index futures is remarkable, but in the long period, the existence of the right arrow is almost uniform, which also indicates that the spot stock index period is still fluctuating in the long run. In the analysis of VEC model, E-G two-step method is used to test the cointegration relationship between two variables. On this basis, the vector error correction model between stock index and spot is established to judge the lead-lag relationship between them. Finally, the information share model (I-S model based on VECM) and the long-term transient model are used. The contribution of stock index spot to market price discovery is analyzed quantitatively by using P-T model. In the daytime price discovery function of stock index futures, the spot stock index plays a more important role in price discovery. The results of modeling using 1 minute high frequency data show that the price of Shanghai and Shenzhen 300 stock index futures is 6 minutes ahead of spot prices. And the contribution of price discovery is significantly greater than that of spot market, and through the analysis of the fast rising and falling bands, it is found that in the fast falling band, there is only one way price guidance of stock index futures to spot, but in the rising stage, This price-guiding relationship is mutual.
【学位授予单位】：暨南大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F224;F832.51

【参考文献】