基于尾部风险溢出网络的全球外汇市场关联性研究

doi:10.16381/j.cnki.issn1003-207x.2021.0389

摘要/Abstract

摘要：

全球金融自由化和一体化在降低外汇市场运转成本的同时，也促进了风险在全球外汇市场上的传染。本文采用全球外汇市场33个主要货币的汇率数据，通过LASSO-CoVaR模型构建尾部风险溢出网络，以测度不同货币间汇率风险关联性，并重点考察欧洲债务危机与中美贸易摩擦期间全球外汇市场网络关联性结构，从系统、区域、个体三个层面对动态网络关联性展开实证分析。实证结果表明：（1）危机期间全球外汇市场总体关联性处于较高水平，并存在明显的货币集群效应；（2）中东和南美地区多表现为风险承担角色，货币稳定性较弱，而亚洲和北美地区风险输入和输出强度多处于持平状态，货币稳定性较强；（3）美元和欧元的风险输入和输出强度较低，说明全球外汇市场波动并不一定来源于世界主导货币，日元和韩元在2018年后受中美贸易摩擦带来的政治经济影响，风险输出强度不断增加，而人民币在纳入SDR后受到其他货币的风险溢出影响显著增强。

关键词: 复杂金融网络, 外汇市场, 尾部风险溢出, 关联性, 拓扑特征

Abstract:

With the advancement of financial liberalization and financial integration， the transaction cost of global foreign exchange （FX） markets has been greatly reduced， and the cross-border investment has become more convenient. But at the same time， it has also promoted the risk contagion in the global FX network. Therefore， it is of great significance to study the connectedness change of the FX markets and the network topology evolution at different periods. Using the data of global FX rates of 33 currencies and seven macroeconomic variables from 2011 to 2021 on the pacific exchange rate service website， tail risk spillover network is constructed based on the LASSO‒CoVaR model to measure the currency risk interconnectedness. The network interconnectedness structure of global FX markets during the European debt crisis and the Sino‒US trade friction is emphatically studied， and the dynamic network connectedness is empirically analyzed from the system‒wide， sector‒conditional and individual‒level measures. The empirical results show that：（i） The total connectedness of global FX markets is at a high level during the crisis， and there is an obvious currency clustering effect；（ii） The currencies from Middle East and South America mostly acted as the risk‒recipients and their stability is weak， while the incoming risk strength （in‒strength） and outgoing risk strength （out‒strength） of currencies from Asia and North America are mostly at the same level and their stability is relatively strong； and （iii） The in‒strength and out‒strength of the US dollar and euro are relatively low， suggesting that the fluctuations of global FX markets do not necessarily come from the world’s dominant currencies. The Japanese yen and South Korean won are affected by the political and economic impact of the Sino‒US trade friction after 2018， and their risk spillover strength persistently increased. The risk it received from other currencies has enhanced remarkably， after the Renminbi joined the special drawing rights. Through a comprehensive analysis of the tail risk spillover network in global FX markets， the investment strategies are suggested that investors should be more cautious when investing in currencies from the Middle East and Latin America， and consider more about the impact of the connectedness between the currencies of the United States， Japan， and Europe when constructing an investment portfolio.

Key words: complex financial network, foreign exchange markets, tail risk spillover, interconnectedness, topological features

中图分类号:

F830.91

王纲金, 马欣宇, 谢赤. 基于尾部风险溢出网络的全球外汇市场关联性研究[J]. 中国管理科学, 2025, 33(3): 13-23.

Gangjin Wang, Xinyu Ma, Chi Xie. Measuring Interconnectedness of Global Foreign Exchange Markets Using Tail Risk Spillover Network[J]. Chinese Journal of Management Science, 2025, 33(3): 13-23.

图/表 12

表1

关联网络拓扑特征度量指标"

指标	表达式	描述
总体关联性 (total connectedness, TC)	$T C = ∑ j = 1 N ∑ i = 1, i ≠ j N β ̂ q, V a R q, t i j \| X ˜ t (j)$	从系统层面测度整个网络总体的尾部风险溢出强度，量化整个网络的互联性
区域输入强度 (region in-strength, RIS)	$R I S m = ∑ j ∈ V m ∑ i = 1 N β ̂ q, V a R q, t i j \| X ˜ t (j)$	从区域层面测度其他地区所有货币对地区m所有货币的尾部风险溢出强度，量化地区m的尾部风险输入强度
区域输出强度 (region out-strength, ROS)	$R O S m = ∑ j ∈ V m ∑ i = 1 N β ̂ q, V a R q, t j i \| X ˜ t (i)$	从区域层面测度地区m所有货币对其他地区所有货币的尾部风险溢出强度，量化地区m的尾部风险输出强度
货币输入强度 (currency in-strength, CIS)	$C I S j = ∑ i = 1, i ≠ j N β ̂ q, V a R q, t i j \| X ˜ t (j)$	从个体层面测度其他所有货币对货币j的尾部风险溢出强度，量化货币j的尾部风险输入强度
货币输出强度(currency out-strength, COS)	$C O S j = ∑ i = 1, i ≠ j N β ̂ q, V a R q, t j i \| X ˜ t (i)$	从个体层面测度货币j对其他所有货币的尾部风险溢出强度，量化货币j的尾部风险输出强度

表1

表2

图1

图2

表3

图3

图4

表4

图5

图6

图7

图8

参考文献 18

1	Adrian T， Brunnermeier M K. CoVaR［J］. America Economic Review， 2016， 106（7）： 1705-1741.
2	Hautsch N， Schaumburg J， Schienle M. Forecasting systemic impact in financial networks［J］. International Journal of Forecasting， 2014， 30： 781-794.
3	Hautsch N， Schaumburg J， Schienle M. Financial network systemic risk contributions［J］. Review of Finance， 2015， 19： 685-738.
4	Fang L， Sun B， Li H， et al. Systemic risk network of Chinese financial institutions［J］. Emerging Markets Review， 2018， 35： 190-206.
5	蒋海，张锦意. 商业银行尾部风险网络关联性与系统性风险——基于中国上市银行的实证检验［J］. 财贸经济， 2018， 39（8）： 50‒65.
	Jiang H， Zhang J Y. Tail risk network correlation and systemic risk of commercial banks： An empirical test based on China’s listed banks［J］. Finance and Trade Economics， 2018， 39（8）： 50-65.
6	Xu Q F， Li M T， Jiang C X， et al. Interconnectedness and systemic risk network of Chinese financial institutions： A LASSO-CoVaR approach［J］. Physica A， 2019， 534： 122173.
7	Härdle W K， Wang W， Yu L. TENET： Tail-Event driven NETwork risk［J］. Journal of Econometrics， 2016， 192： 499-513.
8	Wang G， Jiang Z， Lin M， et al. Interconnectedness and systemic risk of China’s financial institutions［J］. Emerging Markets Review， 2018， 35： 1-18.
9	Wang G， Feng Y， Xiao Y， et al. Connectedness and systemic risk of the banking industry along the belt and road［J］. Journal of Management Science and Engineering， 2022， 7（2）： 303-329.
10	Billio M， Lo A W， Getmansky M， et al. Econometric measures of connectedness and systemic risk in the finance and insurance sectors［J］. Journal of Financial Economics， 2012， 104： 535-559.
11	Wang G J， Xie C， He K， et al. Extreme risk spillover network： Application to financial institutions［J］. Quantitative Finance， 2017， 12（9）： 1417-1433.
12	Matesanz D， Ortega G J. Network analysis of exchange data： Interdependence drives crisis contagion［J］. Quality & Quantity， 2014， 48： 1835-1851.
13	Keskin M， Deviren B， Kocakaplan Y. Topology of the correlation networks among major currencies using hierarchical structure methods［J］. Physica A， 2010， 390： 719-730.
14	Kwapień J， Gworek S， Drożdż S， et al. Analysis of a network structure of the foreign currency exchange market［J］. Journal of Economic Interaction and Coordination， 2009， 4（1）： 55-72.
15	Wang G， Xie C. Tail dependence structure of the foreign exchange market： A network view［J］. Expert Systems with Applications， 2016， 46： 164-179.
16	Wen T， Wang G. Volatility connectedness in global foreign exchange markets［J］. Journal of Multinational Financial Management， 2020， 54： 100617.
17	Tibshirani R. Regression shrinkage and selection via the LASSO［J］. Journal of the Royal Statistical Society： Series B （Methodological）， 1996， 58（1）： 267-288.
18	Belloni A， Chernozhukov V. l₁-penalized quantile regression in high-dimensional sparse models［J］. The Annals of Statistics， 2011， 39（1）： 82-130.

区域	货币	简称	区域	货币	简称
中东	以色列新谢尔克	ILS	欧洲	冰岛克朗	ISK
	沙特阿拉伯里亚尔	SAR		挪威克朗	NOK
亚洲	中国人民币	CNY		波兰兹罗提	PLN
	印度卢比	INR		罗马尼亚列伊	RON
	印度尼西亚盾	IDR		俄罗斯卢布	RUB
	日元	JPY		瑞典克朗	SEK
	韩元	KRW		土耳其里拉	TRY
	马来西亚林吉特	MYR	大洋洲	澳元	AUD
	菲律宾比索	PHP		新西兰元	NZD
	新加坡元	SGD	北美洲	加拿大元	CAD
	泰铢	THB		美元	USD
	中国台湾新台币	TWD	拉丁美洲	阿根廷比索	ARS
欧洲	瑞士法郎	CHF		巴西雷亚尔	BRL
	捷克克朗	CZK		智利比索	CLP
	欧元	EUR		哥伦比亚比索	COP
	英镑	GBP		秘鲁比索	PEN
	匈牙利福林	HUF

输入输出	日期	USD	CNY	EUR	JPY	KRW	日期	USD	CNY	EUR	JPY	KRW
USD	2011/1/4	-	0.256	0.831	0.309	0.015	2012/1/3	-	0.185	0.795	0.222	0
CNY		0.532	-	0	0.061	0.113		0.293	-	0.158	0	0
EUR		0.702	0	-	0.419	0		0.866	0.309	-	0.381	0
JPY		0.749	0	0.631	-	0.001		0.480	0	0.823	-	0
KRW		0	0	0	0.197	-		0	0	0.156	0.290	-

输入输出	日期	USD	CNY	EUR	JPY	KRW	日期	USD	CNY	EUR	JPY	KRW
USD	2018/1/2	-	0	0.043	0	0.039	2019/1/2	-	0.185	0.707	0	0
CNY		0	-	0	0	0.284		0	-	0.472	0	0
EUR		0	0	-	0.146	0.104		0.877	0.309	-	0.228	0
JPY		0	0	0	-	0.040		0	0.396	0	-	0.008
KRW		0	0	0	0.176	-		0	0	0	0.121	-

[1]	沈嘉贤, 陈浩智, 张卫国. 基于知识图谱网络特征的中国外汇市场系统性风险测度研究[J]. 中国管理科学, 2025, 33(3): 45-61.
[2]	吴文洋,蒋海,唐绅峰. 数字化转型、网络关联性与银行系统性风险[J]. 中国管理科学, 2024, 32(3): 9-19.
[3]	李守伟, 文世航, 王磊, 何建敏, 龚晨. 多层网络视角下金融机构关联性的演化特征研究[J]. 中国管理科学, 2020, 28(12): 35-43.
[4]	熊正德, 文慧, 熊一鹏. 我国外汇市场与股票市场间波动溢出效应实证研究——基于小波多分辨的多元BEKK-GARCH(1,1)模型分析[J]. 中国管理科学, 2015, 23(4): 30-38.
[5]	刘善存, 宋殿宇, 金华. 分数布朗运动下带违约风险的可转换债券定价[J]. 中国管理科学, 2011, 19(6): 25-30.