Penerapan Fuzzy Time Series Markov Chain dalam Meramalkan Nilai Tukar Rupiah terhadap Yuan, Dollar Amerika dan Dollar Singapura

Abstract

Forecasting currency exchange rates is an important aspect of economic analysis as exchange rate fluctuations can affect various sectors of a country's economy, including international trade, investment, and monetary policy. Indonesia, as an open economy, is highly vulnerable to the dynamics of the global economy. Therefore, understanding and projecting currency exchange rate movements is crucial for better economic planning and decision-making. This study aims to project the purchase price of the Rupiah exchange rate against Yuan, US Dollar, and Singapore Dollar in January-February 2024 using the Fuzzy Time Series Markov Chain method with historical data in 2023. This method uses two membership functions, namely sigmoid and gauss, to handle data complexity, because of its ability to represent membership levels and capture changes more subtly between low and high membership. The results show good performance with low prediction error (0-10%). The Mean Absolute Percentage Error (MAPE) results with the FTS-Markov Chain method for the three exchange rates namely Yuan, US Dollar and Singapore Dollar with sigmoid membership function 1.127% while gauss membership function 2.435%. Meanwhile, the time series method, namely ARIMA, obtained a MAPE of 17.383%. The Rupiah exchange rate against Yuan fluctuates, against the US Dollar is predicted to appreciate, and against the Singapore Dollar experiences an initial decline but increases again.