基于LASSO-VAR模型的外汇市场风险溢出网络研究

Research on Risk Spillover Network of Foreign Exchange Market Based on the LASSO-VAR Model

随着外汇市场越来越受到投资者的青睐,其风险分析研究也变得十分重要且非常必要。因此,本文探讨了2005年9月至2023年2月期间的全球外汇传递冲击溢出效应和网络连通性。我们使用波动率溢出指数和最小绝对收缩和选择算子向量自回归(LASSO-VAR)的方法来构建35种全球货币的网络连通性。总体而言,该研究发现全球外汇市场之间存在显著的溢出关联性(达到78.54%),其中SGD (新加坡元)和CHF (瑞士法郎)是冲击的主要净传递者,连通性分别为96.99%和75.29%。相比之下,ARS (阿根廷比索)和INR (印度卢比)是冲击的主要净接收者,连通性分别为94.96%和79.15%。从动态上看,总溢出连通性对国际经济变化和危机的反应不同。这项研究的结果可以对全球外汇市场的风险分析做出额外贡献,这也将有助于对冲在危机期间的相关风险。

Widespread interest in foreign exchange markets among investors makes risk analysis study important and well-needed. Therefore, this paper explores the global foreign exchange return shock spillovers and network connectedness between September 2005 and February 2023. We use the volatility spillover index and the Least Absolute Shrinkage and Selection Operator-Vector Auto-regression (LASSO-VAR) approach to construct network connectedness of 35 global currencies. Overall, the study found a significant spillover connectedness among the global foreign exchange markets (78.54%), with SGD (Singapore Dollar) and CHF (Swiss Franc) as the major net transmitters, accounting for 96.99% and 75.29%, respectively. In contrast, ARS (Argentine Peso) and INR (Indian Rupee) are the major net receivers of shocks, with 94.96% and 79.15%, respectively. Dynamically, total spillover connectedness reacts differently to international economic changes and crises. The findings of this study can be an additional contribution to the risk analysis in the global foreign exchange markets, which will also be useful to hedge the risk during crisis periods.