A Suitable Forecasting Model for Exchange Rates of the Top 10 Foreign Currencies Most Preferred by Thai Tourists Compared to the Thai Baht

Abstract

This research aims to find a suitable forecasting model for predicting the exchange rates of ten foreign currencies most commonly used by Thai people in comparison to the THB exchange rates. The currencies considered are Japan, South Korea, Hong Kong, Singapore, the United States of America, Taiwan, China, the United Kingdom, Vietnam, and Australia. Four time series analysis methods were employed: the decomposition method, smoothing method, Box and Jenkins methods, and combined forecasting methods. The exchange rate information was collected from the website Investing.com, and the time series data covered the period from January 2000 to September 2022. Each exchange rate used a different number of data points based on the availability of data. The data were separated into two sets: a training data set for constructing models and a test data set for verifying the accuracy of the models. The forecasting models were evaluated using the mean absolute percentage error (MAPE), and data analysis tools such as Microsoft Excel 2019, Minitab version 19, and R Software version 4.2.2 were utilized. The results showed that the combined forecasting method consisting of the Regression Analysis Method (REG), the Min Mean Absolute Error Method (mMAE), the Whale Optimization Algorithm Method (WOA), and the Inverse of Mean Squares Error Method (INV), was the most suitable model for the majority of the foreign currencies.