基于局部相似性分析的股票市场投资策略研究

发布时间：2017-12-30 20:22

本文关键词：基于局部相似性分析的股票市场投资策略研究　出处：《山东大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着经济的发展,人民生活水平的提高,越来越多的人开始投资理财,将自已的闲钱投资于股票、债券、基金等金融产品。其中,股票市场无疑是前几年最火热的一个地方。无论是散户还是机构投资者,都迫切地希望自已能够在股市中能够大赚一笔。我们发现股票市场经常出现一种同变现象,股票之间的关联度非常高。股票市场联动性的研究不仅能够帮助投资者分析股票信息、掌握市场动态,还能够协助投资者作出相对正确的投资策略,从而实现收益增加。就目前来说,股票联动性的研究很少涉及到个股之间,反而研究比较多的是地域之间、板块之间的研究。为了探索股票市场个股之间的相关关系,本文运用局部相似分析方法来研究个股之间的相似区间以及位置关系。最后基于局部相似分析方法建立了投资策略模型。本文主要介绍了一种具有时间延迟的局部相似性方法,并将其运用到股票市场对个股进行相似性分析,从而得到两支股票的局部相似区间以及时间延迟D。我们的数据集为A股市场深圳交易所以及上海交易所上市的90支股票2013年1月4号到2016年12月30号的收盘价,在4005个样本中,我们首先选出D≠0的样本,再在D≠0的样本中选出时间延迟4≤|D|≤20,并且相似区间长度占总长度达到80%以上的数据作为投资策略研究样本。本文制定的投资策略的核心思想是根据两只股票的同涨同跌性质,来建立交易模型。本文中涉及到的参数取值方法为将参数在测试集得到最优取值,到最后通过实证分析来检验我们模型的效果。本文得到的结果如下所示:一:股票市场中的个股之间存在很强的相关性,个股之间通常有同涨同跌的现象。二:运用局部相似方法得到两支股票的局部相似区间I =[i,i+l-1]和J =[j,j + l-1]以及时间延迟D后,我们发现在D≠0的情况下,深市样本中相似区间占总序列长度的80%以上的有490对(占比62.26%)其中相似区间占总序列长度90%以上的达到344对(占比43.71%)。沪市样本中相似区间占总序列长度的80%以上的有712对(占比91.04%)其中相似区间占总序列长度90%以上的达到440对(占比56.27%)。三:运用基于局部相似分析方法建立的投资策略模型实现了非常好的收益,在深市样本中2016年平均收益率达到12.78%,而在2016年深证成指下跌了 12.46%。在沪市样本中2016年的股票数据平均收益率达到4.93%,而大盘指数在2016年变化率为-5.84%。在2016年的测试集里面我们的策略平均收益每天都是跑赢大盘的。
[Abstract]:With the development of economy, the improvement of people's living standard, more and more people began to invest their own money to invest in stocks, bonds, funds and other financial products. Among them, the stock market is undoubtedly a place a few years ago the most fiery. Whether retail or institutional investors, are eager to own to be able to earn a lot of money in the stock market. We found that the stock market often appear with a variable phenomenon, correlation between stocks is very high. Research on the linkage of the stock market can not only help investors to analyze stock information, grasp the market dynamic, also can help investors make relatively correct investment strategy, so as to realize the income increase. At present, the research of stock linkage rarely involved between stocks, but more research is research between regions, between plates. In order to explore stock market. The relation, using the local similarity analysis method to study the stocks of the similarity between interval and the position relations. Finally the local similarity analysis method based on the investment strategy of the model. This paper mainly introduces a kind of partial similarity method of time delay, and applied it to the stock market on the stock of similarity analysis and get two shares of local similarity interval and time delay D. our data set for the A shares listed on the Shanghai stock exchange and Shenzhen stock exchange market 90 stocks from January 4, 2013 to December 30, 2016 closing price, in 4005 samples, we first select the D = 0 sample, and then in the D = 0 sample selected time delay of 4 = |D| = 20, and the total length of similar interval length of more than 80% of the data as the study sample. The investment strategy of the core idea of this paper is to develop investment strategies two Stocks rise or fall together in nature, to establish trading model. The parameters involved in this method for the optimal parameters obtained in the test set, finally to test our model results by empirical analysis. The results of this paper are as follows: A: there is a strong correlation between stock market in, usually with the rise and fall of the phenomenon between stocks. Two: two stocks using the local similarity method of local similarity of interval I =[i, i+l-1] J and =[j, j + l-1] and time delay of D, we found that in D = 0, Shenzhen samples of similar interval accounted for more than the length of the sequence was 490 to 80% (accounting for 62.26%) which is similar to the total length of interval of more than 90% to 344 of the Shanghai stock market (accounting for 43.71%). Samples of similar interval accounted for more than 80% of the sequence length of 712 (accounting for 91.04%) which is similar to the total interval order The column length of more than 90% to reach 440 (accounting for 56.27%). Three: the use of local similarity analysis method to establish the model of investment strategy to achieve a very good income based on the sample in Shenzhen in 2016 the average yield reached 12.78%, and in 2016 the Shenzhen stock index fell 12.46%. stock data in Shanghai City in 2016 the average income of the sample the rate reached 4.93%, while the stock market index in 2016 was -5.84%. change in the 2016 test set inside our strategy of average income of every day is to beat the market.

【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】：F832.51

【相似文献】