Forecasting Model for Foreign Currency Exchange Rates

Abstract

The objective of this research was to investigate the most suitable forecasting model for 3 currency exchange rates (eg. USD, EUR and CNY), which are time series from January 2022 to December 2022. The comparative study was used to find the most proper forecasting model from 6 forecasting models (eg. simple exponential smoothing, Additive Holt-Winters, Multiplicative Holt-Winters, ARIMA) based on 4 accuracy measures. The most suitable forecasting model plays a crucial role in international trade as well as effective import-export strategies of Thailand. The results indicate that the arima model outperformed the 5 statistical forecasting models, tharefore 3 currency exchange rate patterns have followed a linear relation pattern. Therefore, the proposed model can be a promising tool to predict the currency exchange rates, and to support decision making on effective import-export strategies of Thailand.