Using Mixed Copula to investigate the structure of dependence in NTD、JPY、SGD、AUD

• Main Authors: Wen-Jeng Wang, 王文正
• Other Authors: 黃文光
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/59612223737539066892

碩士 === 淡江大學 === 財務金融學系碩士班 === 96 === The Dependence among the Financial Markets has been widespread studied and the focus of economical analysis was on the degree of dependence;but not the structure of dependence,another significant point of view but neglected generally。 This research targets the Currency Exchange Markets in Taiwan, Japan, Singapore, and Australia, based on the report of daily exchange rate, to analyze the frequency from 2000 to 2007; however, the actual data extracted from the report don’t match the normal allocation。 Accordingly, this research applies quantitative method such as Unit root test、Arch test。 Further, this research studies structure of dependence of rate of foreign exchange between these four countries through GARCH(1,1) model、the method of copula、genetic algorithms、Goodness of Fit Test。 This research found the following results: 1. Exchange Markets exhibit Frank Copula U Model & Gumbel Copula J Mode。2. From the most of investors’ point of view, all markets will drop when Exchange Rate is under Bear Market; however, they will be conservative and take prospective response to the Bull Market; meanwhile,the appreciation range is rather slight because investors will regard it as a temporary bullish phenomenon。