Forecast the currency exchange rate between the baht and the currencies of Thailand's important trading partners.

Abstract

This article aims to identify the best models for forecasting Hong Kong, Singapore, Japan, and Taiwan exchange rates by studying data from the Bank of Thailand. It is a daily data from 4 January 2022 to 30 December 2022 for a period of 241 days. Excluding Saturday and Sunday. The model used for forecasting consists of Simple exponential smoothing method model Exponential smoothing method, recalibrated twice. Holt-Winter smoothing method (positive), Holt-Winter smoothing method (Multiplicative), and the method of combining built-in regression with moving average. Then select the best model using the following aberration criteria. percentage absolute error (MAPE), mean absolute error (MAE), root mean square error (RMSE), and mean square error (MSE). The best currency is Singapore, ARIMA (3, 1, 0), MAPE equal to 0.259, MAE equal to 0.066, MSE equal to 0.007396, RMSE equal to 0.086000.