Peramalan Volatilitas Return Cryptocurrency Menggunakan Model Asymmetric Generalized Autoregressive Heteroscedasticity (Garch) Dilengkapi Gui-R

Cryptocurrency is a digital asset that has high level of risk and high rate of return so that it attracts many investors to invest. Risk is described as volatility, forecasting volatility can be a consideration for investors in viewing risks and making decisions in investment activities. High volatility raises the problem of non-constant residual variance (heteroscedasticity) so that the ARIMA model can’t be used to predict volatility. ARCH/GARCH models provide a solution for predicting volatility, but when applied to financial data, there is an asymmetric effect involving different positive and negative residuals with financial characteristics. Therefore, this research aims to anticipate the variation of returns for the next 7 days using asymmetric ARCH/GARCH models (EGARCH, TGARCH, APARCH). Volatility is represented using daily BNB-USD data from March 2nd, 2020 to February 4th, 2023. The research results show that the EGARCH (2,2) model is the best-performing model as it has the lowest RMSE and MAE values. SMAPE value obtained is 17,72%, which is in the range between 10% and 20%. It means model has good forecasting performance. Data analysis equipped with GUI-R can make it easier for investors to obtain forecasts of return volatility.