Scholarlink

This paper presents a preliminary analysis of the yen-dollar exchange rate from 2000 to 2024, using advanced time series techniques to uncover the complex dynamics driving currency fluctuations. Traditional models, such as ARIMA and GARCH, offer limited insight into non-linear behaviors. To address this, we apply Recurrence Quantification Analysis (RQA), Orthogonal Matching Pursuit (OMP), and Kernel Change Detection (KCD) to explore the hidden structures, recurrences, and anomalies in the time series. RQA reveals key metrics such as recurrence rate, determinism, and laminarity, suggesting that the yen-dollar exchange rate, though chaotic, exhibits recurring patterns and periods of stability. OMP decomposes the time series into successive approximations, isolating long-term trends and high-frequency fluctuations, while KCD identifies significant ruptures corresponding to major economic events like the 2008 financial crisis and the COVID-19 pandemic. These methods reveal that the yen-dollar exchange rate is influenced by both recurring patterns and external shocks. Our findings have implications for traders, policymakers, and researchers seeking to understand financial markets, enabling better decision-making and risk management.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Analyzing the Complexity of Yen-Dollar Exchange Rates: Recurrence Quantification, Orthogonal Matching Pursuit, and Anomaly Detection

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Yuwen Zhao,
  • Amir Assadi

摘要

This paper presents a preliminary analysis of the yen-dollar exchange rate from 2000 to 2024, using advanced time series techniques to uncover the complex dynamics driving currency fluctuations. Traditional models, such as ARIMA and GARCH, offer limited insight into non-linear behaviors. To address this, we apply Recurrence Quantification Analysis (RQA), Orthogonal Matching Pursuit (OMP), and Kernel Change Detection (KCD) to explore the hidden structures, recurrences, and anomalies in the time series. RQA reveals key metrics such as recurrence rate, determinism, and laminarity, suggesting that the yen-dollar exchange rate, though chaotic, exhibits recurring patterns and periods of stability. OMP decomposes the time series into successive approximations, isolating long-term trends and high-frequency fluctuations, while KCD identifies significant ruptures corresponding to major economic events like the 2008 financial crisis and the COVID-19 pandemic. These methods reveal that the yen-dollar exchange rate is influenced by both recurring patterns and external shocks. Our findings have implications for traders, policymakers, and researchers seeking to understand financial markets, enabling better decision-making and risk management.