[Paper Review] Exploring the Interconnectedness of Cryptocurrencies using Correlation Networks

Andrew Burnie|arXiv (Cornell University)|Jun 18, 2018

Complex Systems and Time Series AnalysisEconomics, Econometrics and Finance19 citations

TL;DR

This study uses correlation networks to analyze price co-movements among 14 major cryptocurrencies from September 2016 to March 2018, revealing statistically significant positive correlations, especially among forked coins like Bitcoin and Bitcoin Cash, and identifying distinct clusters linked to Ethereum and Cardano. The findings suggest that factors beyond speculation—such as technical lineage and network effects—significantly influence price dynamics.

ABSTRACT

Correlation networks were used to detect characteristics which, although fixed over time, have an important influence on the evolution of prices over time. Potentially important features were identified using the websites and whitepapers of cryptocurrencies with the largest userbases. These were assessed using two datasets to enhance robustness: one with fourteen cryptocurrencies beginning from 9 November 2017, and a subset with nine cryptocurrencies starting 9 September 2016, both ending 6 March 2018. Separately analysing the subset of cryptocurrencies raised the number of data points from 115 to 537, and improved robustness to changes in relationships over time. Excluding USD Tether, the results showed a positive association between different cryptocurrencies that was statistically significant. Robust, strong positive associations were observed for six cryptocurrencies where one was a fork of the other; Bitcoin / Bitcoin Cash was an exception. There was evidence for the existence of a group of cryptocurrencies particularly associated with Cardano, and a separate group correlated with Ethereum. The data was not consistent with a token's functionality or creation mechanism being the dominant determinants of the evolution of prices over time but did suggest that factors other than speculation contributed to the price.

Motivation & Objective

• To investigate the extent and structure of price co-movement among major cryptocurrencies using correlation networks.
• To assess whether a cryptocurrency's functionality or creation mechanism is the primary driver of price evolution.
• To evaluate the robustness of observed correlations across different time periods and cryptocurrency subsets.
• To identify structural clusters or groups of cryptocurrencies with shared price behavior.
• To determine whether network effects or technical lineage (e.g., forks) play a more significant role than speculative behavior in shaping price dynamics.

Proposed method

• Constructed correlation networks using daily price returns of 14 major cryptocurrencies from November 2017 to March 2018.
• Applied a subset of nine cryptocurrencies from September 2016 to increase data points and improve temporal robustness.
• Used statistical methods to test the significance of observed correlations, focusing on positive associations.
• Identified clusters within the correlation network using network analysis techniques to detect groups of highly correlated assets.
• Examined the role of technical lineage (e.g., forks) by comparing correlation patterns between parent and child cryptocurrencies.
• Evaluated the influence of external factors such as token functionality and whitepaper design by cross-referencing with project documentation.

Experimental results

Research questions

• RQ1 Are there statistically significant positive correlations among major cryptocurrencies over time?
• RQ2 Do cryptocurrencies that are technical forks of one another exhibit stronger price co-movement than non-forked pairs?
• RQ3 Do distinct clusters of correlated cryptocurrencies emerge, and are they associated with specific projects like Ethereum or Cardano?
• RQ4 To what extent do factors other than speculation—such as network effects or technical lineage—influence price evolution?
• RQ5 How does the inclusion of earlier data (starting from 2016) affect the robustness and reliability of correlation-based findings?

Key findings

• Excluding USD Tether, a statistically significant positive correlation was found across the majority of the 14 cryptocurrencies studied.
• Six cryptocurrency pairs involving a fork relationship showed robust, strong positive correlations, with Bitcoin and Bitcoin Cash being a notable exception.
• A distinct cluster of cryptocurrencies was identified that showed strong intercorrelation, primarily associated with Cardano.
• Another separate cluster emerged, predominantly linked to Ethereum, indicating strong price co-movement among its ecosystem tokens.
• The data did not support the dominance of a token’s functionality or creation mechanism as the primary driver of price evolution.
• The results suggest that non-speculative factors such as technical lineage and network effects significantly contribute to price dynamics in cryptocurrency markets.