Analysis of approaches and their features to research the dynamics of cryptocurrencies

Abstract

The purpose of the study is to provide a comparative analysis of approaches to analyzing the trends in cryptocurrency dynamics. The paper analyzes the trends in cryptocurrency development, which has shown an increase in the influence of cryptocurrency on the structure of the financial market. It has been determined that from 2013 to 2023, the capitalization of cryptocurrency market increased almost 1 000 times. However, in recent years, the number of "fake" cryptocurrencies has also increased, so the total number of cryptocurrencies has almost not changed in the last two years. The works of researchers on the analysis of the trends in cryptocurrency exchange rates have been studied, and three main approaches to the analysis have been formed. The main components of the first approach have been investigated, the influence of miners, mining costs, blockchains, and the interaction of mining participants on the formation of exchange rates has been determined. It is found that the necessity to analyze the cryptocurrency market in conjunction with other elements of the financial market is the key aspect of the second approach to the study of trends in the exchange rate of cryptocurrencies. In this approach, cryptocurrency acts as an alternative to centralized components of the financial market and as an element of financial freedom. The study of the works of the third approach allows to identify the main methods and models for analyzing the dynamics of exchange rates, among which the main place is occupied by: models of time series analysis taking into account sentiments (Sentiment-Enriched Time Series Forecasting – SETS models), deep learning models for forecasting of processes with long and short-term memory, recurrent neural networks, and gated recurrent unit models.