Predicting Crypto Trending by Using Markov Models
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Cryptocurrencies have become integral to daily transactions and income generation. It’s a most important source to get income. This study is tried to predict the crypto trending movement based on the globally leading digital currencies value changes by employing Markov model. Thus, the primarily aim of this study is to determine the future tendency of increasing and decreasing the value of cryptocurrencies. The study focuses on Bitcoin, Ethereum and Monero, utilizing daily open values collected from Yahoo spanning 1795 days for Bitcoin and 1787 days for Ethereum and Monero. The dataset was Analyzed to determine the classification of price fluctuations across two successive days. Two Markov models were applied, one with two states (gain and loss) and the other with six states (large-gain, moderate-gain, small-gain, small-loss, moderate-loss, and large-loss). Acceptance of the Markov property and stationary assumption paves the way for further analysis. The steady state probability vector π is determined to understand the long-term patterns of crypto values, and expected recurrence times are computed. Finally, the future range for cryptocurrencies has been established using WMC (Weighted Markov Chain model). Cryptocurrency prices fluctuate daily, while the predicted values from the model fall within fixed intervals for several days. Therefore, accurately determining the exact value of a cryptocurrency on a specific day is very difficult. The findings offer valuable insights for investors and traders, aiding them in making informed decisions about future investments in the dynamic world of cryptocurrencies.
Fields of Study & Subjects
Stock Market Forecasting Methods
Blockchain Technology Applications and Security
Financial Markets and Investment Strategies
Publication details
- Journal / source
- Sri Lankan Journal of Applied Statistics
- Volume
- 27
- Issue
- 1
- Pages
- 158–180
- Publication date
- 31 March 2026
- Language
- EN