Estimation of Bitcoin volatility: GARCH implementation

As bitcoin has been a topic of high interest for academic and professional life over recent years, a number of literature has examined its price movements, volatility, and predictions. Bitcoin is the first and perhaps the most popular cryptocurrency with a high volatility pattern compared to the other cryptocurrencies. This paper examines the models that explain the volatility of Bitcoin prices. The daily data for the Bitcoin prices are used through a period of July 31, 2017, to April 3, 2019, with a total number of observations of 484. Initially, unit root tests are implemented. Then, the heteroskedasticity problem is tested among variables. Based on the results of the heteroskedasticity test, it is decided to use ARCH models. Then, ARCH, GARCH, TGARCH, and EGARCH results are tested to find out the best fit model that explains the bitcoin price movements.