基于动态Copula-ARMA-GJR模型的汇率间相依性研究

发布时间：2017-12-30 20:14

本文关键词：基于动态Copula-ARMA-GJR模型的汇率间相依性研究　出处：《湖南大学》2014年硕士论文　论文类型：学位论文

【摘要】：研究金融市场间的相依特征对于资产定价、投资组合及风险管理都具有重要意义。过去对金融市场相依性的考察较多集中在证券市场，而对外汇市场涉及较少。2005年汇率制度改革启动后，人民币开始走上开放化进程，汇率波动的不确定性增大，导致风险加剧。同时，，近年来国际金融危机频繁爆发，各国汇率出现大幅波动，外汇风险进一步凸显。另外，一些国家为缓解危机后本国经济恢复与增长的压力，不断要求人民币升值。在此形势下，加强外汇风险管理势在必行，考虑到不同汇率序列间相依性的准确刻画在风险管理中的关键作用，因此对其进行深入考察显得十分重要。人民币汇率受到央行在市场上的操作，这些噪音的存在使得市场不是真正的自由出清，所以在人民币汇率研究过程中计量方法的选取极富艺术性，单一一种动态Copula模型很难准确刻画人民币汇率间的相依特征。因此，本文根据金融资产表现出来的尖峰厚尾及非对称效应等特性，构建多种动态Copula-ARMA-GJR模型对人民币兑美元、欧元、日元和英镑的4种汇率间的相依结构进行考察。研究发现：人民币兑美元与兑欧元、兑日元、兑英镑汇率间存在显著的负相依性，其中人民币兑美元与兑欧元汇率间的负相依性最为突出；人民币兑欧元与兑英镑汇率间呈现正相依性；人民币兑日元与兑欧元、兑英镑汇率间的相依性则时正时负。在极端事件下，各汇率间相依性较正常时期发生很大变化，但不是像证券市场那样呈现增强的正相依性。另外，尾部相依性显示，人民币兑美元与兑欧元、兑日元、兑英镑汇率的上、下尾部相依性基本为零，说明它们基本不存在同时大涨或大跌的可能性；人民币兑欧元与兑日元、兑英镑汇率有着波动较大的上、下尾部相依性，说明两者存在汇率风险传染关系。
[Abstract]:The study of dependence characteristics of financial markets is of great significance for asset pricing, portfolio and risk management. In the past, the study of dependence of financial markets focused on the securities market. In 2005, the reform of the exchange rate system started, the RMB began to open up process, the uncertainty of exchange rate fluctuations increased, resulting in increased risk. At the same time. In recent years, the frequent outbreak of the international financial crisis, countries exchange rate fluctuations, foreign exchange risk further highlighted. In addition, some countries to ease the post-crisis domestic economic recovery and growth pressure. Under this situation, it is imperative to strengthen the management of foreign exchange risk, considering the key role of accurately portraying the dependence of different exchange rate sequences in risk management. Therefore, it is very important to conduct a thorough investigation. The RMB exchange rate is operated by the central bank in the market, and the existence of these noises makes the market not really free to clear, so the choice of measurement method in the research process of RMB exchange rate is very artistic. A single dynamic Copula model is difficult to accurately describe the characteristics of RMB exchange rate dependence. Therefore, according to the characteristics of financial assets such as peak, thick tail and asymmetric effect. A variety of dynamic Copula-ARMA-GJR models are constructed to examine the structure of the dependence of the RMB against the dollar, the euro, the yen and the pound. It is found that there is a significant negative dependence between RMB and euro, yen and sterling, among which the negative dependence of RMB against US dollar and euro is the most prominent. The exchange rate of RMB against euro and sterling is positively dependent; The dependence of RMB against yen, euro and sterling is positive and negative. In extreme events, the dependence of each exchange rate is much different than that of normal period. In addition, the tail dependency shows that the renminbi is basically zero against the dollar, against the euro, against the yen and against the pound. It shows that there is basically no possibility that they will rise or drop simultaneously; Against the euro and the yen, the yuan has a volatile upper and lower tail against sterling, indicating a risk contagion relationship between the two.
【学位授予单位】：湖南大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F832.6;F224

【参考文献】