Modelling cryptocurrency price volatility through the GARCH and EWMA model / Andree Irawan and Wiwik Utam

The number of cryptocurrency investors has grown rapidly compared to conventional financial asset investors. This condition needs attention considering the high price volatility of cryptocurrency without any underlying transactions. This research aimed to provide empirical evidence for the best price volatility prediction model. The research selected two cryptocurrencies, namely Bitcoin and Ethereum, because they have the largest capitalization. The data used was the daily price of cryptocurrency from January 1, 2020 to June 30, 2023. Data from 1 January 2020 to 31 December 2022 was used to create a prediction model, and data from 1 January 2023 to June 30, 2023 was used to test the accuracy of the prediction model. Tests were carried out to determine which volatility model provided the best validity and smallest error between GARCH and EWMA. The result showed that EGARCH (1,1) model was proven to have the smallest error value compared to the GARCH (1,1) and EWMA model. The research results are useful for investors who have a preference for carrying out technical analysis to minimize risk by using EGARCH (1,1). Further research should carry out cryptocurrency portfolios as each cryptocurrency has different price volatility.