Ethereum Value Forecasting Model using Autoregressive Integrated Moving Average (ARIMA)

Abstract

The purpose of this study is to test the ability of the ARIMA model to predict the value of Ethereum, especially during economic shocks such as the current COVID-19 pandemic. The population in this study is Ethereum value weekly data for the period January 2017 to December 2020, so there are 208 samples in this study. The results showed that the use of the ARIMA method in predicting the value of Ethereum got poor results, where the forecast value was very much different from the actual value. This is evidenced from the results of the accuracy test using MAPE which got a result of 51.94%. On the other hand, the economic conditions that are experiencing uncertainty due to the COVID-19 pandemic and the emergence of deficit (decentralized finance) in early 2021 have pushed up a very significant increase in the value of Ethereum so that the error standard is higher and reduces the ability of the ARIMA model to predict the value of Ethereum. Further research is recommended to use a more advanced model such as the Autoregressive Fractionally Integrated Moving Average (AFRIMA) in order to obtain a better forecast value.