Consider the impact of US interest rate policy on the RMB exchange rate forecast research against the US dollar

Global economies are interconnected, with each country’s monetary policy spilling over to affect others’ goods and capital, especially under full capital mobility. The U.S., faced with the twin challenges of rising inflation and unemployment due to the pandemic, the Russia-Ukraine conflict, and trade tensions with China, has started raising interest rates. This has intensified fluctuations in the RMB exchange rate, highlighting the need for accurate prediction. To address this, a deep learning-based ARIMA-EEMD-IAWT-LSTM model is proposed, which enhances time series denoising by combining improved adaptive wavelet thresholding (IAWT) and ensemble empirical mode decomposition (EEMD). This model, integrated with LSTM, effectively predicts the nonlinear aspects of the time series, reducing risks associated with exchange rate volatility. The model’s prediction accuracy reaches 98.4822%, underscoring its practical significance in exchange rate forecasting.