节点文献

人民币汇率波动性特征的实证分析

An Empirical Study on the Rmb Exchange Rate Volatility Characteristics

【作者】 温跃峰

【导师】 刘纪显;

【作者基本信息】 广东商学院 , 金融学, 2010, 硕士

【摘要】 本文以人民币/美元、人民币/欧元、人民币/日元和人民币/英镑汇率为研究样本,对人民币汇率的收益率序列与绝对值收益率序列的基本统计特征、非对称性和长记忆性展开实证研究,得出如下结论:对人民币汇率收益率与绝对值收益率序列均不服从独立同分布,不服从标准正态分布,具有典型的尖峰厚尾特征;收益率序列不存在或只存在极为微弱的短期自相关,而绝对值收益率序列都具有较强的短期自相关性;对人民币汇率收益率序列进行EGARCH-M模型估计,结果表明除了人民币/日元收益率有较强的非对称性,其余的三个序列的非对称效应都很微弱。信息冲击曲线也很清楚地表明了这一结论;使用R/S方法分别对人民币汇率收益率序列和绝对值收益率序列进行长记忆性检验与Hurst参数估计,结果表明除了人民币/美元收益率序列,其他收益率序列均不具备长记忆性,而绝对值收益率序列都具有较强的长记忆特征。

【Abstract】 Taking RMB exchange rate against U.S. dollar, Euro, Yen and Pound as the sample, the paper analyzes basic statistics, non-symmetry and long memory in volatilities of the RMB exchange rate’s returns and its’absolute value of returns, drawing a follow conclusion:Both of the RMB exchange rate of return and absolute value of returns are not subject to independent and identical distribution, do not obey the standard normal distribution, and have the typical characteristics of fat tails. Return series do not exist or exists only very weak short-term autocorrelation, while the absolute value series have strong short-term autocorrelation. The estimation results of EGARCH-M model indicate the returns of RMB exchange rate against Yen has a strong non-symmetry, but that of the other three are weak. It’s obviously to see that from their Information shock curves. We using the R / S method for the return series and the absolute value series to test the long memory and estimate the Hurst parameter, and the results indicating that except the return series of RMB exchange rate against U.S. dollar, all the return series have no long memory, while all the absolute value series have strong long memory characteristics.

【关键词】 人民币汇率波动性非对称性长记忆性GARCH模型
【Key words】 VolatilitiesAsymmetryLong memoryGARCH model
  • 【网络出版投稿人】 广东商学院
  • 【网络出版年期】2011年 03期
  • 【分类号】F224;F832.6
  • 【下载频次】277
节点文献中: