【作者】 王敏艳；

【导师】 李金昌；

【作者基本信息】 浙江工商大学 ， 统计学， 2010， 硕士

【摘要】 众所周知,价格和成交量构成了对交易最为直观和及时的两方面描述,量价关系的研究对于认识和理解股票本身、股票交易以及股票市场而言都是最为基础也是最为重要的一个切入点。对于相对历史较短且在特殊国情下产生、发展的中国股票市场而言,为了对其进行全面深刻地认识和理解,也必定离不开对量价关系的研究,并应该在借鉴国外研究的思想和结论的基础上,注意联系本国市场的特性,创造性地去发现本国市场上可能存在的客观规律。本文以中国证券市场为研究对象,以分位数回归为主要研究方法,分别从国内和国际两个视角来观察其成交量与股价变动之间的关系,深入剖析了国内证券市场的量价关系特征。全文共分五章。第一章为绪论,总结国外学者关于量价关系的研究,归纳得到比较成熟的量价关系研究成果。第二章具体阐述分位数回归方法的基本思想、应用及其实现方式。第三章对国内大、中、小盘分别进行量价分析,并分阶段对各板块股票进行量价分析。第四章对中美两国股市的量价关系分别进行研究,并在此基础上研究美国股市对中国股市的动态影响。最后一章是全面的总结。从国内视角分析。首先,各板块比较,在25%分位点之后成交量的增加往往伴随着股票收益率上涨,出现“量利齐扬”现象；而成交量的萎缩则伴随着股票收益率下跌,出现“量缩利减”现象。而在左尾处,大中小盘股成交量与收益率的这种关系受到扭曲,此时成交量与收益率成负相关,成交量的增加反而导致股票收益率下降,出现“量利分离”现象。并且从所有分位点来看,大盘股在同分位点,成交量对收益率的影响小于中小盘股成交量对收益率的影响。在左尾处,小盘股成交量对收益率影响最大；在右尾处,中盘股成交量对收益率影响最大。说明,大中小盘股之间存在板块轮动,且总体来讲中小盘股波动较大,这里不排除存在人为操控的因素。其次,分阶段各板块比较,在平稳期,中小盘股β值变动基本一致；在下跌期,2007年10月16日到2008年11月7日,收益率对成交量变的更加“敏感”,其“敏感度”是上升期的2倍以上,这从侧面反映了在股价疯狂攀升的时候,20%的情况下,人们是怀疑成交量的,而在下跌期,接近90%的情况下,人们肯定成交量所带来的(多半负面)信息；在上升期2,2008年11月7日到2009年6月19日,在45%分位点之后无论是增值速度还是增值方向大盘股和中小盘股都表现出截然相反的态度。大盘股的成交量对收益率的影响继续延续之前的正相关关系并且其相关系数越往右越大,而中小盘股的相关系数在此之后逐渐回落,并分别在86%与97%分位点处下滑为负值。从国际视角分析。上证综指与深圳成指的β值在20%(包括20%分位点)以左小于零,其余皆大于零；纳斯达克指数、标准普尔500指数以及道琼斯指数的β值在50%分位点以左都小于零,其余皆大于零,且三个指数的β值走势基本相同。这表明,美国的股市通过多年的发展,已经形成了健康的运行机制,少有人为操纵的痕迹,投资者多数趋向于价值投资；而中国股市仍属于资金推动型,投资者多属于短线投机。其次,中美股市之间存在微弱且平稳的联动性。但是,上证综合指数和美国各项指数的波动主要还是受到其自身历史数据的影响,与之相比,两国股市的联动性显得微不足道。更多还原

【Abstract】 As usual, trading volume and prices are the basic information we can directly get from the market, researches about the relationships between these two variables are helpful for us to understand more about the stocks and the market.This paper studies the relationship between the trading volume and return of China’s stock market; it will use Quantile Regression as the main method and combine the situation of China market and the foreign market (USA stock market) to find more information about the relationships of trading volumes and return.The paper consists five chapters. The first chapter summarizes and generalizes the actual study developed by overseas scholars. The second chapter will introduce and describe the Quantile Regression way and its applications. In the third chapter, it will analyze the relationships of trading volumes and return of three different stock plates (that are plates of big, middle and small size companies) of China market. It will also analyze different stages to see more deeply. Then it will compare China and USA market to see where the differences are and why. The last chapter is the conclusion.Firstly for all these different plates, after 25% fractile, return rates and trading volume have positive correlation; return rates go up and down with the increase and decrease of trading volume. But at the left tail, the correlation is negative. And in all fractile, compared with other plates, volumes’ influence on prices for the plate of big size companies is smaller. Secondly, in the steady stage, plates of middle and small size companies have almost the same variation rate. In bear stage, return rate of prices are more sensitive to the change of volumes, it’s about two times as in the bull stage. Moreover, in bull stage, compared with other two plates, plate of big companies has different performance. It continues to show positive correlation, but for other two plates, the positive correlation is going weaker.From the international perspective,βof SSE composite index and Shenzhen index were under zero at the left of 20% fractile and above zero under other situation. The Nasdaq index, Standard & Poor’s index of 500 and Dow Jones Industrial Average were under zero at the left of 50% fractile and above zero under other situation. This shows that American stock market is under healthy operating mechanism and is not manipulated by people, investors mostly make value investment. But China market is still pushed by money, most of investors are speculators. Moreover, there is weak correlation between China and American market, the fluctuation of these two markets is mainly influenced by their own history data.更多还原