Approaches to estimating and forecasting cryptocurrency volatility: дипломная работа / Крысальная Ольга Андреевна; Белорусский государственный университет, Экономический факультет; Кафедра аналитической экономики и эконометрики; науч. рук. Васенкова Е.И.

Abstract

The study is based on the application of a class of parametric models. It allowed describing both stationary and non-stationary time series and on this basis to develop a system of prognostic estimates for the prospects of further development of the series under study. With the help of ARIMA-GARCH model a system of prognostic assessments for the short term is developed. It is proved that the application of such models with a high level of reliability predicts future adjustments in the market under study. It leads to a high level of prospects for their use in modelling future parameters of the cryptocurrency market development. This creates a basis for a business to develop adaptive mechanisms for to emerging price index adjustments of digital money. The research object is the international cryptocurrency market. The subject of the research is the dinamics of the cryptocurrency volatility. The purpose of the research is studying the methods of estimating volatility in the cryptoсurrency market, as well as short-term forecasting the prices using binding models of autoregression and Generalized Autoregressive Conditional Heteroskedasticity (ARMA-GARCH). Research methods: graphic analysis, ARDL analysis, ARMA-GARCH. Results of the work: the theoretical foundations of the process were studied, thesis’ goals were archived, the cryptocurrency market was analysed, reliable forecast is obtained. The author of the paper confirms that the calculated and analytical material presented in the diploma work correctly and objectively reflects the state of the process under investigation, and all theoretical, methodological and methodological positions and concepts borrowed from literary and other sources are accompanied by references to their authors.