Modelling and Forecasting Volatility of Bitcoin Cryptocurrency: A Comparative Study of Conditional Heteroscedastic Model

Abstract

This study details three cryptocurrencies: Bitcoin, Ethereum, and Tether. These cryptocurrencies provide a high return but also pose a high degree of risk to investors. The changes in the prices of the currencies impact their return volatility. This report presents the performance of two models employed in forecasting the volatility in the times series data. These models are the ARCH and GARCH. The research presented in this report employs a GARCH suite of models: ARCH (1), GARCH (1,1), as well as GARCH-M and PARCH. The three models are described based on the literature review and their roles. Using the data collected based on these models, prices, and changes in investments are determined. A methodology based on the models was applied to test the data and measure the stationarity and structural breakpoints and the ARCH effects. From the collected data and methods employed to evaluate it, this study concludes that the ARCH (1) model has the best performance for all three cryptocurrencies when it comes to return and risk.