中文摘要：

首先通过采用基于各个国债之间相关性的距离测度方法,对修正后的国债日收益率时间序列进行了聚类分析,将上海证券交易所的20只国债分成了5类,分类结果显示我国国债市场中的各个国债日收益率之间的相关性程度是很明显地依赖于它们的到期时间的。同时,研究了每一个国债修正后日收益率时间序列的价格波动结构,结果表明中国的国债市场也表现出非常明显的多标度分形特征,还发现属于同一类国债的多标度特征具有很强的自相似性,而不同类型国债之间的多标度特征则具有一定程度上的差异性。

英文摘要：

The paper first makes clustering analysis for 20 treasury bonds＇modified daily return time series by using a suitable distance metrical approach based on the correlation among the bonds. Using the Matlab software, the paper classifies the bonds of Shanghai Stock Exchange into five groups and the results show that the correlation among the bonds evidently depend their time to maturity. Then, this paper also studies the price fluctuation structure of each bond＇s modified daily return series well, and the results also suggest that there is evident multi-scaling character for China treasury bonds market. In the meanwhile, the paper finds that there is strong self-similarity among bonds belonging to the same group, while the multi-scaling characters among different bonds are different.