基于跳跃行为的汇率风险测算研究

论文总字数：32056字

摘 要

作为世界上最大的金融市场，外汇市场24小时不间断运作，具有庞大的交易规模和良好的流动性。1973年，二战后统治世界经济近三十年的布雷顿森林体系的崩溃使得浮动汇率制度登上历史舞台，汇率变化的不确定性使得全球外汇市场交易更加活跃，也导致了汇率波动和跳跃行为的发生。近年来，人民币国际化趋势越发明朗，中国的金融市场进一步开放，日益暴露的汇率风险给从事外汇交易的机构和个人等相关主体都带来了潜在的负面影响。因此，研究汇率的跳跃行为有着积极意义。

本文选取人民币与美元、欧元、日元、港元、英镑共计五种主要货币对的日频数据为样本，对其走势和变化规律进行研究，分析根据GARCH-JUMP模型估计的跳跃特征及其影响，并识别出具体的跳跃日期。实证结果表明：（1）跳跃行为对汇率波动整体上存在负向影响；（2）美元/人民币和港元/人民币的跳跃行为之间有较强相关性；（3）收益率存在尖峰厚尾的分布特征，且汇率波动存在不对称性，美元/人民币和港元/人民币存在右偏分布，英镑/人民币、日元/人民币和欧元/人民币存在左偏分布，且前者分布的不对称性大于后者；（4）汇率的历史跳跃强度对当期跳跃强度有正向影响；（5）历史条件波动率对当期跳跃强度存在影响，影响方向取决于具体货币对；（6）跳跃预期偏差对当期跳跃强度存在影响，影响方向取决于具体货币对。

本文的实证结果不仅为外汇投资者有效度量和规避汇率风险以及相关政策制定者制定相关决策提供了理论依据，也为后续研究外汇期权等金融衍生工具的资产定价和风险管理问题奠定了基础。

关键词：汇率；跳跃行为；风险测算；GARCH-JUMP模型

ABSTRACT

As the largest financial market in the world, the foreign exchange market operates 24 hours a day, with huge trading scale and good liquidity. After the collapse of the Bretton Woods System in 1973, currencies of many countries began to adopt the floating exchange rate system. The uncertainty of the capital market and financial market made the global foreign exchange market more active, and also aggravated the exchange rate volatility and jumping behavior. With the advance of RMB internationalization and the further opening of China's financial market in recent years, the increasingly exposed exchange rate risks have brought potential negative impacts on institutions and individuals engaged in foreign exchange transactions. Therefore, it is indispensable for us to identify the jumping behavior of exchange rate and measuring the jumping probability or risk.

This article selects the data of five major currency pairs, including USD/CNY, EUR/CNY, JPY/CNY, HKD/CNY, GBP/CNY, to study the distribution of log return and descriptive statistics characteristics, and analyzes different jump characteristics according to the GARCH-JUMP model. Then, it applies the Lee-Mykland jump recognition algorithm to identify the specific jump dates. The empirical results show that: (1) jump could affect exchange rate volatility, and impact is negative; (2) there is a strong correlation between the jumping behavior of USD/CNY and HKD/CNY. (3) exchange rate return represents Fat-tail Distribution, and the volatility of it is asymmetrical, distribution of USD/CNY and HKD/CNY are right-skewed, while the distribution of GBP/CNY, JPY/CNY and EUR/CNY are left-skewed, and the asymmetry of the former distribution is greater than that of the latter. (4) historical jump intensity has a positive impact on the jump intensity in the current period. (5) conditional volatility has impact on the jump intensity in the current period, and how conditional volatility affect jump intensity depends on the specific currency pair. (6) expected jump deviation has impact on the jump intensity in the current period, and how the former affect the latter depends on the specific currency pair.

The empirical results of this paper not only help foreign exchange investors effectively measure and avoid exchange rate risks and provide a theoretical basis for policy makers to make relevant decisions, but also lay a foundation for the further study on asset pricing and risk management of financial derivatives such as foreign exchange options.

KEY WORDS: exchange rate；jump behavior；risk measure；GARCH-JUMP model

目 录

第一章 引言 1

1.1研究背景 1

1.2研究意义 1

1.3论文结构与研究创新点及不足 2

第二章 文献综述 3

2.1汇率决定理论相关研究 3

2.2汇率波动相关研究 4

2.3跳跃行为建模及检验的相关研究 5

2.4文献评述 6

第三章 外汇市场现状分析与样本选取 8

3.1外汇市场现状分析 8

3.2样本选取 8

第四章 基于GARCH-JUMP模型的跳跃特征与影响分析 11

4.1 GARCH-JUMP模型构建 11

4.2 GARCH-JUMP模型参数估计与分析 13

第五章 跳跃的识别与风险分析 15

5.1 Lee-Mykland跳辨识算法 15

5.2 跳跃日期识别与分析 16

第六章 结论与政策性建议 21

参考文献 22

附录 GARCH-JUMP模型极大似然估计程序 26

致谢 28

剩余内容已隐藏，请支付后下载全文，论文总字数：32056字