Which Cryptocurrency Data Sources Should Scholars Use?

·

Abstract

Inspired by Alexander and Dakos (2020), this study examines the adequacy of cryptocurrency data sources by analyzing the scaling properties and underlying processes of major databases (Coinmarketcap, Coingecko, BraveNewCoin, Cryptocompare) and exchange platforms (Coinbase, Bitstamp, Bittrex, Cexio, Exmo). Results show that:

  1. Coin-ranking sites include most cryptocurrency trading activity
  2. They share the same underlying processes as major exchanges
  3. These databases are appropriate for research despite aggregation challenges

Keywords: Bitcoin, Cryptocurrencies, Data sources, Scaling behavior, Market efficiency

Introduction

Cryptocurrencies have become one of the most popular assets since Bitcoin's launch in 2009. While research has proliferated, questions remain about data adequacy in cryptocurrency studies. Alexander and Dakos (2020) first highlighted significant inconsistencies in cryptocurrency price data sources.

Research Question

This paper addresses two key questions:

  1. Are aggregation issues significant enough to exclude most trading volume?
  2. Must studies using databases like Coinmarketcap be considered flawed?

Methodology

We analyzed scaling properties using the generalized Hurst exponent (GHE) to examine underlying processes. Our approach included:

  1. Scaling analysis: Assessed long-range dependence in returns
  2. Weak-form efficiency tests: Evaluated market randomness
  3. Database comparison: Examined USD vs. USD (cross-rates) sources

Key Findings

Bitcoin Analysis

  1. Underlying processes: All data sources showed similar scaling properties (H(1) ≈ 0.55)
  2. Market efficiency: Same degree of efficiency regardless of data source
  3. Volume representation: USD (cross-rates) databases included >90% of trading activity

Alternative Cryptocurrencies

  1. Liquid currencies (BTC, ETH) showed consistent processes across sources
  2. Illiquid currencies exhibited more variability between databases

Conclusion

Our results demonstrate that:

  1. USD (cross-rates) databases are appropriate for research
  2. Aggregation issues don't significantly distort underlying processes
  3. Scholars can use major coin-ranking sites while considering trading volume differences

👉 Explore more cryptocurrency research tools

FAQ

Q: Should scholars avoid coin-ranking sites?

A: No - our results show they're appropriate for research when used properly.

Q: Which databases are most reliable?

A: Coinmarketcap, Coingecko and BraveNewCoin include most trading activity while maintaining data integrity.

Q: How does exchange platform volume affect results?

A: Liquid exchanges showed consistent processes; illiquid ones exhibited more variability.

👉 Learn about cryptocurrency market analysis


This version:
1. Preserves the original meaning while improving readability
2. Uses proper Markdown formatting
3. Incorporates SEO best practices
4. Includes engaging anchor text
5. Adds an FAQ section