Nonlinear Fluctuations in the Cryptocurrency Market: The Modern Approaches to Analysis and Forecasting
Kochorba V. Y.

Kochorba, Valeriia Yu. (2025) “Nonlinear Fluctuations in the Cryptocurrency Market: The Modern Approaches to Analysis and Forecasting.” The Problems of Economy 2:166–175.
https://doi.org/10.32983/2222-0712-2025-2-166-175

Section: Finance and banking

UDC 330.4:519.8

Abstract:
This article presents a comprehensive study of modern methods for analyzing and forecasting cryptocurrency market dynamics. The author examines the evolution of the cryptocurrency market from a niche technological innovation to a significant segment of the global financial system, characterized by high volatility and sensitivity to a wide range of external factors. The research is based on an analysis of historical data from leading cryptocurrencies for the period 2015-2025 and systematization of recent scientific publications. The historical price dynamics of leading cryptocurrencies (Bitcoin, Ethereum and Litecoin) were analyzed for the period 2015–2025, confirming exponential market growth alongside extremely high volatility and cyclical fluctuations. Special attention is paid to identifying cyclical patterns and factors determining price dynamics. A SWOT analysis of the market was conducted, which allowed systematizing internal strengths and weaknesses, as well as external opportunities and threats. The main emphasis is placed on comparative analysis of forecasting models from classical statistical methods to modern machine learning and deep learning approaches. The author demonstrates evolution from simple linear models to complex neural networks that better account for the nonlinear nature of the market. The main classes of forecasting model are categorized, ranging from classical statistical approaches (e.g. ARIMA and GARCH) to modern machine learning and deep learning methods (e.g. LSTM, GRU and Transformers). The study reveals that deep learning models demonstrate higher accuracy on short- and medium-term forecasting horizons, however long-term forecasting remains a challenging task due to the influence of fundamental factors. The work has practical significance for investors, traders and analysts, providing structured information about market functioning features and tools for its analysis. The research contributes to understanding the complex dynamics of cryptocurrency markets and provides guidance for selecting appropriate analytical methods based on specific forecasting objectives and time horizons.

Keywords: cryptocurrency, Bitcoin, Ethereum, price forecasting, volatility, machine learning, LSTM, ARIMA, GARCH, market dynamics, financial markets, blockchain, digital assets, investment risks, time series, artificial intelligence, financial modeling, SWOT analys

Fig.: 4. Tabl.: 3. Bibl.: 12.

Kochorba Valeriia Yu. – Candidate of Sciences (Economics), Associate Professor, Deputy Director, Educational and Scientific Institute «Karazin Banking Institute» of V. N. Karazin Kharkiv National University (55 Peremohy Ave., Kharkiv, 61174, Ukraine)
Email: V.y.kochorba@karazin.ua

List of references in article

Ahmed, W. M. A. “Robust drivers of Bitcoin price movements: An extreme bounds analysis“. The North American Journal of Economics and Finance, art. 101728, vol. 62 (C) (2022). DOI: 10.1016/j.najef.2022.101728
Akyildirim, E. et al. Forecasting Mid-price Movement of Bitcoin Futures Using Machine Learning (Working Papers 20-2020). Copenhagen Business School, Department of Economics, 2020.
Almeida, A., and Goncalves, T. C. “Cryptocurrency price prediction: A systematic literature review“. Journal of Open Innovation: Technology, Market, and Complexity, art. 100097, vol. 9 (3) (2023).
Chen, Z., Li, C., and Sun, W. “Bitcoin price prediction using machine learning: An approach to sample dimension engineering“. Journal of Computational and Applied Mathematics, art. 112395, vol. 365 (2020).
“CoinMarketCap. Today's Cryptocurrency Prices by Market Cap“. https://coinmarketcap.com/
Guo, Q. et al. MRC-LSTM: A Hybrid Approach of Multi-scale Residual CNN and LSTM to Predict Bitcoin Price. arXiv preprint arXiv:2105.00707, 2021.
Katsiampa, P. “An empirical investigation of volatility dynamics in the cryptocurrency market“. Research in International Business and Finance, vol. 50 (2019): 336-356.
Livieris, I. E. et al. “Investigating the problem of cryptocurrency price prediction: A deep learning approach“. Computation, art. 43, vol. 8 (2) (2020).
Nguyen, D. K., and Vo, X. V. “A combined ARIMA-GARCH model for predicting cryptocurrency prices“. Mathematics, art. 234, vol. 9 (2) (2021).
Patel, R., and Thakur, M. “Utilizing ARIMA and GARCH models for predictive analytics in cryptocurrencies“. International Journal of Computer Applications, vol. 186 (62) (2024).
Pintelas, E. et al. “Forecasting cryptocurrency prices with deep learning“. Algorithms, art. 140, vol. 13 (6) (2020).
Ye, Z. et al. “Bitcoin price prediction using machine learning: A comprehensive review“. Journal of Industrial Information Integration, art. 100358, vol. 28 (2022).