Forecasting the Exchange Rate for the Thai Baht Against the Chinese Yuan by Using a Genetic Algorithm-Based Subset Autoregressive Integrated Moving Average Model

Accurate forecasting of foreign exchange rates plays a crucial role in future global financial market investment, international business decision-making, and travel planning. This paper proposes a model for forecasting the daily exchange rate for the Thai baht (THB) against the Chinese yuan (CNY) during the Novel Coronavirus 2019 (COVID-19) pandemic by comparing a genetic algorithm (GA)-based subset autoregressive integrated moving average (ARIMA) model to the classical ARIMA model. Data was gathered from April 1, 2020 to April 14, 2022. Forecast accuracy was measured by mean absolute percentage error (MAPE), root mean squared error (RMSE) and mean absolute error (MAE). A GARI program was developed using non-seasonal time series prediction, with the best model ARIMA( \(4, 1, \{1, 5\}\) ) forecasting daily CYN/THB exchange rate attaining a nadir of MAPE, RMSE and MAE at \(1.2180\%, 0.066674\) and 0.064061, respectively. These findings indicate that the GA-based subset ARIMA model via GARI program outperformed the classical ARIMA model in the auto ARIMA with Python. This program may be applicable for predicting other foreign exchange rates and non-seasonal time series data.