【作者】 夏庆；

【导师】 童恒庆；

【作者基本信息】 武汉理工大学 ， 应用数学， 2005， 硕士

【摘要】 2001年2月19日,上海B股市场对国内居民正式开放。大量国内投资者涌入B股市场,对B股市场的风险结构产生了不可忽略的影响。为了研究这种影响,本文首先对所选取的样本——1997年7月21日到2003年10月21日的上证A股指数、上证B股指数、上证综合指数收益率数据,以2001年2月19日为界线分为开放前和开放后两期进行深入的分析,发现其收益率分布具有明显的尖峰厚尾的特性;第一期收益率序列具有明显的自相关性;所有序列都是平稳的且具有ARCH效应。因此我们使用GARCH模型来研究上海股票市场是否具有杠杆效应和B股市场的风险结构变化是合适的。本文得出的结论如下: 一、上证A股、上证B股和上证综合指数收益率序列明显具有尖峰厚尾的性态。学生氏t分布比正态分布更能准确的描绘条件误差分布。 二、上证A股、上证B股和上证综合指数第一期收益率序列明显存在自相关,并且3阶自相关后的残差序列中不再具有自相关性。 三、上证A股、上证B股和上证综合指数收益率序列明显具有集聚性现象。 四、上证A股,上证综合指数和上证B股第二期的收益率序列具有非对称性和杠杆效应。上证B股第一期的收益率序列不具有非对称性和杠杆效应,这可能是由于B股市场对国内居民正式开放之前,每天市场交易的数量不是很大,而且上市公司发行的B股不能超过其发行总量的25%。B股市场的利坏消息对公司的负面影响一般不会导致上市公司陷入财务危机。所以,上市公司对利坏消息的反应是如此的不敏感,以至于我们不能有效的察觉。 五、由于沪市存在杠杆效应,EGARCH模型和GJR GARCH模型更能准确地描述股市收益和波动性之间的关系。 六、信息冲击对外国投资者的影响在B股市场对国内居民正式开放之后有变弱的趋势。这极有可能是因为外国投资者在第一期除了承担股市风险以外,还要承担汇率风险——任何与汇率有关的信息都对外国投资者影响重大。更多还原

【Abstract】 On February 19, 2001, the B-share market of Shanghai was open to domestic residents legally. A large number of domestic investors pour into the B-share market, has produced the influence that can’t be ignored on the risk structure of the B-share market. In order to study the influence, this paper divided the sample chosenfirstly-the return of A-share index of Shanghai Stock Exchange , the B-shareindex of Shanghai Stock Exchange, the composite index of Shanghai Stock Exchange ranging from July 21, 1997 to October 21, 2003 into two stages, regarding February 19, 2001 as the boundary, carried on deep analysis. We find that the index return is distributed with obvious excess kurtosis, fatted tail; data in period I are autocorrelated; all series are stationary and have ARCH effect. So we use GARCH model to study whether Shanghai stock market have lever effect or not and the change for the risk structure of B-share market is proper. This study finds the following results:1. A-share index of Shanghai Stock Exchange, the B-share index of Shanghai Stock Exchange and Shanghai Stock Exchange composite index return series obviously are fat tailed, with excess kurtosis. Student t distribution fits the conditional error distribution better than normal distribution.2. A-share index of Shanghai Stock Exchange in Period I, the B-share index of Shanghai Stock Exchange in Period I and Shanghai Stock Exchange composite index in Period I return series was autocorrelated, and the autocorrelation in the residual for the third lag disappeared.3. A-share index of Shanghai Stock Exchange, the B-share index of Shanghai Stock Exchange and Shanghai Stock Exchange composite index return data was clustering.4. A-share index of Shanghai Stock Exchange, Shanghai Stock Exchange composite index and Shanghai Stock Exchange B-share in Period II have asymmetry nature and lever effect, the B-share index of Shanghai Stock Exchange in Period I did not have asymmetry nature and lever effect. It is possible for the reason that before the B-share market is open to domestic residents legally, the quantity for market trading each day is not very large, and the B-share which the listed company issued can’t exceed 25% of total issued. The influence coming from the bad news in the B-share market can’t lead the company to fall into the financial crisis. Thus, list companies were not so sensitive to the bad news that we can’t find it effectively.5. For stock markets of Shanghai have lever effect, EGARCH model and GJR GARCH model describe the relation between the stock return and the volatility.6. The influence to the foreign investor coming from news impact was weaken once B-share markers was also open to domestic residents. One plausible interpretation of this finding could be news, such as news related to exchange risk, had important impact on foreign owner in Period I.更多还原