On-chain analysis and cryptocurrency price forecasting using on-chain metrics

Abstract

Cryptocurrencies have emerged as a new type of financial asset offering investors an alternative to traditional investments such as stocks, bonds, and commodities. In recent years, the crypto market has experienced a significant boom which can be attributed to various factors, including the growing interest of investors in cryptocurrencies, the increasing adoption of blockchain technology, and the rise of decentralised finance (DeFi) applications. One of the key features of cryptocurrencies is that they are decentralised and operate on a blockchain, a distributed ledger that records all transactions in a transparent and secure manner. Even though blockchain as a technology has unlimited potential, the data stored on these blockchains is no less than gold dust because blockchain data or on-chain metrics play a vital role in determining the value of cryptocurrencies.

This project presents an investigation into the use of on-chain metrics for cryptocurrency price forecasting. The project examines the effectiveness of on-chain metrics of 4 cryptocurrencies namely - Bitcoin, Ethereum, Dash and Dogecoin in predicting the price movements of cryptocurrencies using various forecasting models such as GARCH+SARIMAX, Bidirectional LSTM and Prophet model. The research findings reveal that on-chain metrics are useful in forecasting cryptocurrency prices, with the Prophet model achieving high levels of accuracy but only for short-term forecasts. Furthermore, the study highlights specific on-chain metrics having the highest predictive power for each cryptocurrency, such as Miner revenue, daily transaction fee and active addresses for Bitcoin. Various trading strategies are also implemented on the predicted prices in order to test if our models' forecast is accurate enough to make profits. However, due to the extremely volatile nature of cryptocurrencies with prices fluctuating rapidly and often unpredictably, any long-term forecasts resulted in inaccurate predictions. This volatility is due to several factors, including the lack of regulation in the cryptocurrency market, the speculative nature of investments in cryptocurrencies, and the vulnerability of the cryptocurrencies to external events such as whale dumping and social media.

Despite the high volatility, cryptocurrencies continue to attract investors who are willing to take on the risks associated with this new asset class. As blockchain technology continues to evolve, it is likely that the use cases for cryptocurrencies will expand, and the demand for these digital assets will continue to grow. Overall, this paper contributes to the growing body of research on the use of on-chain analysis in cryptocurrency price forecasting and provides insights for investors and traders interested in using on-chain metrics for decision-making purposes.