Indonesian Rupiah Exchange Rate in Facing COVID-19 (A Time Series-Machine Learning Approach)

Solichatus Zahroh, ResaSeptiani Pontoh, Yuyun Hidayat, Nabila Mahardika Jiwani, Enny Supartini, Sukono

Nowadays, the world is currently facing COVID-19 health issues. In addition to the risk of health problems, this pandemic has also disrupted the global economy. The exchange rate of Rupiah also devalued during the COVID 19 outbreak. The interventions from the government to reduce the spread of COVID-19 have impacted the Rupiah exchange rate too. This study tries to use the most appropriate forecasting method to predict further values of the Rupiah exchange rate. The aim of this research is to look for the most useful method of predicting the Rupiah exchange rate, which can be used to capture various data patterns caused by the COVID-19 pandemic. The methods that will be involved in this research are Long Short Term Memory (LSTM), NNAR (Neural Network Auto-Regressive), Extreme Learning Machine, and Support Vector Machine. The result is that the LSTM model is the best prediction method for predicting an exchange rate from Rupiah that can allow long-term dependency. This model shows that although there were no COVID-19 events based on the model formed, it was predicted that the rupiah would decline in value against the US dollar. The COVID-19 case made the depreciation of the rupiah currency far greater than predicted. The declining growth of Covid-19 cases in Indonesia, accompanied by the Indonesian government's policy on large-scale social restrictions (Bahasa: PSBB), appears to have managed to increase the value of the rupiah against the US dollar.

Volume 12 | Issue 6

Pages: 862-872

DOI: 10.5373/JARDCS/V12I6/S20201103