Impact of Asymmetric Volatility and Extreme Shocks on USD/PEN and EUR/PEN Exchange Rates (2015-2024): A Comparative Analysis Using GARCH Models

The data presented consist of daily time series of USD/PEN and EUR/PEN exchange rates for the period from January 2015 to June 2024. These data were collected from official sources such as the Superintendencia de Banca, Seguros y AFP del Perú (SBS) . The purpose of the analysis is to evaluate asymmetric volatility in these currency pairs in response to extreme events like the COVID-19 pandemic, the war in Ukraine, and political instability in Peru.

Research Hypothesis The study proposes that asymmetric econometric models like TGARCH, EGARCH, and PGARCH are more effective in capturing and predicting volatility generated by negative news and extreme shocks compared to symmetric models like GARCH.

Key Findings

There is a strong relationship between global geopolitical and economic events and fluctuations in exchange rates.
Asymmetric models more accurately capture the sensitivity of currencies to negative news and extreme shocks.
Exchange rate fluctuations directly affect investment decisions and hedging strategies in emerging markets.
Data Collection The data were transformed using logarithms and differences to stabilize variance and ensure stationarity, which was validated through unit root tests (ADF). Additionally, a wavelet-based outlier detection method was implemented to ensure the accuracy of the models.

Interpretation and Use of the Data These data are useful for:

Predicting volatility in emerging markets.
Designing hedging strategies in the face of financial uncertainty.
Formulating public policies that mitigate the effects of exchange rate volatility.

本数据集包含2015年1月至2024年6月期间美元/秘鲁新索尔（USD/PEN）与欧元/秘鲁新索尔（EUR/PEN）的每日时间序列汇率数据。这些数据采集自秘鲁银行、保险与养老基金监管局（Superintendencia de Banca, Seguros y AFP del Perú，简称SBS）等官方渠道。本分析旨在评估上述货币对针对新冠疫情、乌克兰冲突及秘鲁国内政治动荡等极端事件的非对称波动特征。

研究假设
本研究提出，相较于广义自回归条件异方差模型（GARCH, Generalized Autoregressive Conditional Heteroskedasticity）等对称计量模型，门限广义自回归条件异方差模型（TGARCH, Threshold GARCH）、指数广义自回归条件异方差模型（EGARCH, Exponential GARCH）及幂广义自回归条件异方差模型（PGARCH, Power GARCH）等非对称计量模型，能更有效地捕捉并预测由负面消息与极端冲击引发的汇率波动。

主要研究发现
1. 全球地缘政治与经济事件与汇率波动存在显著关联；
2. 非对称模型能够更精准地捕捉货币对负面消息及极端冲击的敏感性；
3. 汇率波动会直接影响新兴市场的投资决策与对冲策略。

数据采集与预处理
本数据集经对数变换与差分处理以稳定方差并确保序列平稳性，该平稳性通过增广迪基-富勒单位根检验（ADF, Augmented Dickey-Fuller Test）得到验证。此外，本研究还采用基于小波变换的异常值检测方法，以保障模型构建的准确性。

数据解读与应用场景
本数据集可用于以下场景：
1. 预测新兴市场的汇率波动；
2. 制定应对金融不确定性的对冲策略；
3. 出台缓解汇率波动负面影响的公共政策。