基于CVaR模型的商业银行汇率风险管理研究

发布时间：2018-01-01 21:33

本文关键词：基于CVaR模型的商业银行汇率风险管理研究　出处：《东华大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着人民币汇率市场化进程的不断深入及金融体制改革的日渐深化,中国经济与全球经济整体之间的联系逐渐加强,汇率风险日益凸显。在此背景下,中国商业银行的外汇业务量逐渐增多,丰富的外汇业务种类及整个市场联动性的加强,使得中国商业银行较之一般企业面临更加严峻的汇率风险。为了更好地管理汇率风险,许多商业银行都开始探索先进的风险管理工具,与VaR模型相比,CVaR模型以其数理统计上的优良特性受到越来越多管理者的青睐。此外,由于商业银行持有的外汇币种众多,在风险测算时考虑多币种汇率之间的动态相关性(本文称之为联动性)就显得尤为重要。基于此,研究如何将考虑了联动性的CVaR模型应用于中国商业银行汇率风险管理具有重要的理论和现实意义。本研究从相关文献出发,就汇率风险管理和CVaR方法两方面对前人的研究进行概括综述;随后对商业银行汇率风险管理的度量方法、管理方法和管理工具进行了基本理论阐述,进一步地,结合中国宏观经济与商业银行的实际情况分析了中国商业银行汇率风险的三条传导路径,探讨了汇率风险管理的现状,进而揭示了中国商业银行汇率风险管理的不足,其中最主要的问题就是风险计量方法的落后,显示引入国际上先进计量方法cvar的必要性;为此,引入了考虑多币种汇率动态相关性的cvar模型从而构建联动cvar模型。利用多元garch模型测算联动cvar,其中常用的bekk-garch和dcc-garch模型都具有模型稳定性差且假设条件不符合实际等缺陷,因此,尝试引入新颖的ic-egarh模型并与上述两种模型进行对比分析,为后文的测算模型选择和测算设计做铺垫。出于对中国商业银行外汇币种结构优化的考虑,本研究将投资组合的思想与联动cvar模型相结合,从而构建均值-cvar模型。本文选取美元、欧元、港币及日元兑人民币汇率自2005年7月21日至2016年3月4日之间工作日的每日中间牌价日频数据,通过对数处理得到四组汇率收益率时间序列。在测算模型选择上,将ic-egarch模型与bekk-garch、dcc-garch模型做对比分析,结果显示其不仅弥补了其模型稳定性差、假设条件不符合实际的缺陷,具有显著降低模型维度的优势及良好的测算效果,还能反映各组汇率收益率序列之间的波动溢出,遂选择ic-egarch模型估计时得到的标准差序列来计算单位币种头寸汇率收益率的联动cvar,利用该结果与四家商业银行近十年各年末美元头寸数据相结合,计算美元头寸的联动cvar,进而分析各银行面临的汇率风险状况。运用均值-cvar模型测算美元头寸的最优权重,依据权重结果对中国商业银行外汇头寸币种结构调整提出建议。最后,本文结合联动CVaR在数理统计及汇率风险管理上的优势,对中国商业银行汇率风险管理提出了几点政策建议并总结了本文的研究结论。
[Abstract]:With the deepening of RMB exchange rate marketization process and the deepening of financial system reform, the relationship between China's economy and the global economy is gradually strengthened, and the exchange rate risk is becoming increasingly prominent. The volume of foreign exchange business of Chinese commercial banks is increasing gradually, the rich foreign exchange business types and the whole market linkage are strengthened. In order to better manage the exchange rate risk, many commercial banks have begun to explore advanced risk management tools, compared with the VaR model. The CVaR model is favored by more and more managers because of its excellent mathematical and statistical characteristics. In addition, commercial banks hold a large number of foreign exchange currencies. It is very important to consider the dynamic correlation between multi-currency exchange rate (this paper is called linkage) when measuring the risk. It is of great theoretical and practical significance to study how to apply the CVaR model considering linkage to the exchange rate risk management of Chinese commercial banks. This paper summarizes the previous studies on exchange rate risk management and CVaR methods. Then the measurement methods, management methods and management tools of the exchange rate risk management of commercial banks are described in the basic theory, and further. This paper analyzes the three transmission paths of exchange rate risk of Chinese commercial banks and discusses the current situation of exchange rate risk management in combination with the actual situation of China's macro economy and commercial banks. Furthermore, it reveals the shortage of exchange rate risk management in Chinese commercial banks, among which the most important problem is the backwardness of risk measurement methods, which shows the necessity of introducing international advanced measurement method cvar. Therefore, this paper introduces the cvar model considering the dynamic correlation of multi-currency exchange rate to construct the linkage cvar model, and uses the multiple garch model to calculate the linkage cvar. The commonly used bekk-garch and dcc-garch models have some defects, such as poor stability of the model and the assumption that the conditions do not conform to the actual situation. This paper attempts to introduce a novel ic-egarh model and compare it with the above two models. In order to optimize the foreign exchange currency structure of Chinese commercial banks, this study combines the idea of portfolio with the linkage cvar model. In order to construct the mean value -cvar model, this paper selects the dollar and euro. Daily median daily frequency data for the working days between July 21st 2005 and March 4th 2016 at the exchange rates of the Hong Kong dollar and the yen against the renminbi. Through logarithmic processing, four groups of exchange rate return time series are obtained. In the selection of calculation model, the ic-egarch model and the bekk-archarchdcc-garch model are compared and analyzed. The results show that it not only makes up for the poor stability of the model, but also the hypothetical condition does not accord with the actual defects, and has the advantage of significantly reducing the model dimension and the good measuring effect. It can also reflect the volatility spillover between each group of exchange rate return series, and then select the standard deviation series obtained in the estimation of ic-egarch model to calculate the linked cvar of exchange rate return rate of unit currency position. By combining the results with the dollar position data of four commercial banks at the end of each year, the cvar of dollar position is calculated. Then it analyzes the exchange rate risk situation faced by banks and calculates the optimal weight of US dollar positions by using the mean value-cvar model. According to the weight of the results of foreign exchange positions of Chinese commercial banks currency structure adjustment suggestions. Finally, this paper combined with the linkage of CVaR in mathematical statistics and exchange rate risk management advantages. This paper puts forward some policy suggestions on exchange rate risk management of Chinese commercial banks and summarizes the conclusions of this paper.
【学位授予单位】：东华大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F832.6

【相似文献】