Imam Uddin, Qurat Ul Ain Siddiq, Mosab I. Tabash, Arsalan Qayyum, Muhammad Asadullah
The goal of this study is to forecast the Malaysian Ringgit exchange rate against the US Dollar using individual and combined models. There are three univariate models and one multivariate model among the four individual models. ARIMA, Nave, and Exponential Smoothing are the three univariate models, and NARDL is the multivariate model. The monthly data has been taken over the period from 2011 January to 2020 December. The observations from January 2020 to December 2020 are held back for the in-sample forecasting. The findings show that the NARDL plays an important role in exchange rate forecasting and outperforms all individual and combined by the least MAPE value of 0.2066. It is also clear that the var-cor combination of NARDL and Nave outperforms three individual and other combined models with a MAPE value of 0.352. It supports Poon and Granger's (2003) argument that combining models can produce better forecasting than individual models. The findings are useful for policymakers, FOREX markets, investors, traders, tourists, and hedgers in adjusting their policies accordingly.