Scholarlink

<p>In this paper, we study the forecasting model and economic factor analysis of the RMB against the USD (RMB/USD) exchange rate. We regard the time series of the RMB/USD exchange rate as a highly nonlinear signal and use the complementary ensemble empirical mode decomposition (CEEMD) algorithm of signal processing to decompose the time series of the RMB/USD exchange rate into several intrinsic modal components, and then filter out the components that have the least influence on the original signal through correlation analysis, which complete the filtering and noise reduction of the original signal. Then different methods are used to predict the remaining components after filtering, and the results of the prediction of each component are summed up, so as to construct the forecasting model of the RMB/USD exchange rate in this paper. The results of empirical analysis show that the average relative error of the prediction model built in this paper is 0.2124%, which can give a more accurate prediction of the RMB/USD exchange rate within two months. Subsequently, we carry on the correlation analysis of economic factors to the prediction algorithm constructed in this paper. First of all, we carefully analyze the economic indicators that affect the exchange rate, and we find that there are six factors that have a large impact on the RMB/USD exchange rate, which are foreign exchange reserves, import and export balance, current year-on-year growth rate of industrial added value GI, money supply, consumer price index CPI, and inflation rate. Secondly, using vector autoregression (VAR), we establish the economic correlation between these six factors and each intrinsic modal component obtained by CEEMD decomposition. Finally, by analyzing the changes of each intrinsic modal component, we can obtain the economic factors that have the greatest influence on the RMB/USD exchange rate. In the empirical analysis, since China's foreign exchange administration bureau put forward the foreign exchange management policy on January 4th, 2023, we chose the exchange rate data from January to March 2023 as the test set for empirical analysis, and the empirical analysis results show that in January–February 2023, the main economic factor affecting the change of RMB/USD exchange rate is the amount of foreign exchange reserves, which is completely consistent with China's foreign exchange administration bureau putting forward the foreign exchange management policy on January 4th, 2023. As an application of the RMB/USD exchange rate forecasting model and the economics factor analysis model established in this paper, we present the forecast data of the RMB/USD exchange rate for the coming two month from November 1st to December 31st 2024 and point out that the money supply and foreign exchange reserves are the main factors affecting the exchange rate fluctuation in the near future and give the corresponding suggestions.</p>

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

RMB Exchange Rate Forecasting Model and Economic Factor Analysis Based on CEEMD Decomposition

  • 期刊论文

  • 发表日期：

    2025年7月17日

查看全文
  • Ruoxi Gu,
  • Shaohua Guo,
  • Zhizheng Xu,
  • Kaisu Wu

摘要

In this paper, we study the forecasting model and economic factor analysis of the RMB against the USD (RMB/USD) exchange rate. We regard the time series of the RMB/USD exchange rate as a highly nonlinear signal and use the complementary ensemble empirical mode decomposition (CEEMD) algorithm of signal processing to decompose the time series of the RMB/USD exchange rate into several intrinsic modal components, and then filter out the components that have the least influence on the original signal through correlation analysis, which complete the filtering and noise reduction of the original signal. Then different methods are used to predict the remaining components after filtering, and the results of the prediction of each component are summed up, so as to construct the forecasting model of the RMB/USD exchange rate in this paper. The results of empirical analysis show that the average relative error of the prediction model built in this paper is 0.2124%, which can give a more accurate prediction of the RMB/USD exchange rate within two months. Subsequently, we carry on the correlation analysis of economic factors to the prediction algorithm constructed in this paper. First of all, we carefully analyze the economic indicators that affect the exchange rate, and we find that there are six factors that have a large impact on the RMB/USD exchange rate, which are foreign exchange reserves, import and export balance, current year-on-year growth rate of industrial added value GI, money supply, consumer price index CPI, and inflation rate. Secondly, using vector autoregression (VAR), we establish the economic correlation between these six factors and each intrinsic modal component obtained by CEEMD decomposition. Finally, by analyzing the changes of each intrinsic modal component, we can obtain the economic factors that have the greatest influence on the RMB/USD exchange rate. In the empirical analysis, since China's foreign exchange administration bureau put forward the foreign exchange management policy on January 4th, 2023, we chose the exchange rate data from January to March 2023 as the test set for empirical analysis, and the empirical analysis results show that in January–February 2023, the main economic factor affecting the change of RMB/USD exchange rate is the amount of foreign exchange reserves, which is completely consistent with China's foreign exchange administration bureau putting forward the foreign exchange management policy on January 4th, 2023. As an application of the RMB/USD exchange rate forecasting model and the economics factor analysis model established in this paper, we present the forecast data of the RMB/USD exchange rate for the coming two month from November 1st to December 31st 2024 and point out that the money supply and foreign exchange reserves are the main factors affecting the exchange rate fluctuation in the near future and give the corresponding suggestions.