Paper: arXiv 2506.08718

Authors: Juan Plazuelo Pascual, Carlos Tardon Rubio, Juan Toro Cebada, Angel Hernando Veciana

Abstract

This document analyzes price discovery in cryptocurrency markets by comparing centralized and decentralized exchanges, as well as spot and futures markets. The study focuses first on Ethereum (ETH) and then applies a similar approach to Bitcoin (BTC). Chapter 1 outlines the theoretical framework, emphasizing the structural differences between centralized exchanges and decentralized finance mechanisms, especially Automated Market Makers (AMMs). It also explains how to construct an order book from a liquidity pool in a decentralized setting for comparison with centralized exchanges. Chapter 2 describes the methodological tools used: Hasbrouck’s Information Share, Gonzalo and Granger’s Permanent-Transitory decomposition, and the Hayashi-Yoshida estimator. These are applied to explore lead-lag dynamics, cointegration, and price discovery across market types. Chapter 3 presents the empirical analysis. For ETH, it compares price dynamics on Binance and Uniswap v2 over a one-year period, focusing on five key events in 2024. For BTC, it analyzes the relationship between spot and futures prices on the CME. The study estimates lead-lag effects and cointegration in both cases. Results show that centralized markets typically lead in ETH price discovery. In futures markets, while they tend to lead overall, high-volatility periods produce mixed outcomes. The findings have key implications for traders and institutions regarding liquidity, arbitrage, and market efficiency. Various metrics are used to benchmark the performance of modified AMMs and to understand the interaction between decentralized and centralized structures.

Complexity vs Empirical Score

  • Math Complexity: 8.0/10
  • Empirical Rigor: 7.5/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper employs advanced econometric techniques like Hasbrouck Information Share, Gonzalo-Granger decomposition, and Hayashi-Yoshida estimators, indicating high mathematical complexity, while the empirical analysis uses real cryptocurrency data (ETH/BTC) across multiple exchanges and time periods with specific event-driven backtests, demonstrating solid empirical rigor.

Research Flowchart

  flowchart TD
  A["Research Goal: Price Discovery in Crypto Markets"] --> B["Data Collection<br>ETH: Binance & Uniswap v2<br>BTC: Spot & Futures"]
  B --> C["Methodology: Lead-Lag & Cointegration<br>Hasbrouck's IS, GG-PT, Hayashi-Yoshida"]
  C --> D["Computational Process<br>1. Construct Order Books<br>2. Estimate Lead-Lag Effects<br>3. Analyze Cointegration"]
  D --> E["Key Findings & Outcomes"]
  E --> F["Centralized Markets Lead ETH Price Discovery"]
  E --> G["Futures Lead in Normal Volatility<br>Mixed Outcomes in High Volatility"]
  E --> H["Implications for<br>Liquidity, Arbitrage, & Market Efficiency"]

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper proposes to analyze price discovery in cryptocurrency markets by comparing centralized exchanges with decentralized exchanges and spot markets with futures markets, using Hasbrouck's Information Share, Gonzalo‑Granger decomposition and the Hayashi‑Yoshida estimator.

Source sentence
This document analyzes price discovery in cryptocurrency markets by comparing centralized and decentralized exchanges, as well as spot and futures markets. Chapter 2 describes the methodological tools used: Hasbrouck's Information Share, Gonzalo and Granger's Permanent-Transitory decomposition, and the Hayashi-Yoshida estimator.

Assumes 3

  • The cointegration model uses only one lag and assumes white‑noise residuals. model
    Source sentence
    We use only one lag in our cointegrated model, and we need to confirm that it is well specified.
  • Gas fees are incurred twice per arbitrage trade (entry and exit). execution
    Source sentence
    Since any arbitrage opportunity involves both buying and selling, we must account for gas fees twice.
  • Price series are non‑stationary but have a long‑run equilibrium relationship. statistical
    Source sentence
    2.2.2 Cointegration in a VAR Model Intuitively, cointegration refers to a situation where multiple time series are individually non-stationary, but a linear combination of them is stationary. In other words, even though the individual series may wander widely over time, there exists a long-term equilibrium relationship between them.

Data

  • Binance price and volume data · Binance market-prices public
    one‑second intervals for selected dates · ETH and BTC spot prices
    Source sentence
    For the centralized market data, we obtained it from Binance [6], one of the largest and most liquid cryptocurrency exchanges globally. The dataset includes detailed price and volume information, recorded at one-second intervals.
  • Uniswap v2 on‑chain data · Etherscan crypto-onchain public
    from May 19, 2020 to 2024 events · ETH pool WETH/USDT and BTC pool
    Source sentence
    For the decentralized market data from Uniswap v2, we utilized Etherscan, a leading blockchain explorer for Ethereum. We retrieved all relevant transactions through the Etherscan API, covering the interactions that occurred in the blocks starting from May 19, 2020, when the Uniswap v2 pool was created.
  • CME Micro Bitcoin Futures data · CME market-prices commercial
    tick‑by‑tick data for selected dates in · BTC futures
    Source sentence
    On the futures side, we used data from the CME Micro Bitcoin Futures, which represent one-tenth of the size of a standard Bitcoin futures contract. The futures data are also provided as tick-by-tick (t2t) data, ensuring consistency across both markets.

Recognised sources: Binance Etherscan

Costs

Not applicable The paper reports no trading or execution result; the cost the extractor found is a model parameter.

The paper-text extraction found a cost parameter inside a model (Average gas price 98.68 Gwei, gas limit 50,000 → fee 0.004934 ETH (~$30.58) per transaction, applied twice per arbitrage attempt.); the classifier found no reported trading result, so costs are not applicable.

Finds 4

  • Centralized markets typically lead ETH price discovery. statistical
    Source sentence
    Results show that centralized markets typically lead in ETH price discovery.
  • Futures markets generally lead BTC price discovery, but high‑volatility periods produce mixed outcomes. statistical
    Source sentence
    In futures markets, while they tend to lead overall, high-volatility periods produce mixed outcomes.
  • Hayashi‑Yoshida lead‑lag ratio of 1.86 indicates the centralized exchange leads the decentralized market. statistical
    Source sentence
    The LLR of 1.86 indicates again in this case that the centralized exchange is leading the decentralized market.
  • Arbitrage opportunities during high‑fee periods are unprofitable; mean loss after gas fees is about $45.30 per attempt. negative
    Source sentence
    the mean loss from attempting this arbitrage strategy (after accounting for gas fees) is approximately $45.30. Conclusion on Arbitrage Opportunities These results suggest that during periods of high network activity, such as March 5th, 2024, arbitrage opportunities are scarce due to elevated gas fees.

Checked by others

Supported by

  • Stylized facts in Web3 (2024)
    On the question: Which market type leads ETH price discovery?
    Compare the findings

    This paper: Centralized markets typically lead ETH price discovery.

    That paper: CEX price leads DEX pool price in price discovery, with lagged influence up to twenty minutes and not vice versa.

    Both findings agree that centralized exchanges lead ETH price discovery; A shows CEX leads DEX with up to 20‑minute lag, B confirms centralised markets lead.

Code

No repository by the authors found in the paper.

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.

Cited References (36)

  1. Dynamic Conditional Correlation: A Simple Class of Multivariate Generalized Autoregressive Conditional Heteroskedasticity Models (2002)
  2. An Overview of Market Manipulation (2019)
  3. Spoofing and Layering (2020)
  4. Modelling and measuring price discovery in commodity markets (2010)
  5. One Security, Many Markets: Determining the Contributions to Price Discovery (1995)
  6. High Frequency Lead/Lag Relationships—Empirical Facts (2014)
  7. Likelihood-Based Inference in Cointegrated Vector Autoregressive Models (1996)
  8. Lecture Notes (2024)
and 28 more
  1. Analysis of Financial Time Series (2010)
  2. Price discovery and common factor models (2002) — link
  3. Estimation of Common Long-Memory Components in Cointegrated Systems (1995)
  4. Statsmodels: Econometric and statistical modeling with Python (2010)
  5. OxMetrics: An Interface to Empirical Modelling (2018)
  6. Etherscan: Ethereum Blockchain Explorer (2024)
  7. Refinitiv: Financial Market Data and Infrastructure (2024)
  8. The role of the constant and linear terms in cointegration analysis of nonstationary variables (1994) — link
  9. On covariance estimation of non-synchronously observed diffusion processes (2005) — link
  10. The Econometrics of Financial Markets (1997)
  11. Estimation of Common Long-Memory Components in Cointegrated Systems (1995)
  12. Investigating Causal Relations by Econometric Models and Cross-spectral Methods (1969)
  13. Binance: Cryptocurrency Exchange (2024)
  14. Algorithmic and High-Frequency Trading (2015) — link
  15. Estimation and Hypothesis Testing of Cointegration Vectors in Gaussian Vector Autoregressive Models (1991)
  16. Decentralized Finance: On Blockchain- and Smart Contract-Based Financial Markets (2021) — link
  17. Smart Contracts and Decentralized Finance (2022) — link
  18. Behavior of Liquidity Providers in Decentralized Exchanges (2021) — link
  19. Fragmentation and optimal liquidity supply on decentralized exchanges — → 2307.13772
  20. On The Quality Of Cryptocurrency Markets: Centralized Versus Decentralized Exchanges (2021) — link
  21. Trust in DeFi: An Empirical Study of the Decentralized Exchange (2021) — link
  22. Price Discovery on Decentralized Exchanges (2024) — link
  23. Coexisting Exchange Platforms: Limit Order Books and Automated Market Makers (2021) — link
  24. Flash Boys 2.0: Frontrunning, Transaction Reordering, and Consensus Instability in Decentralized Exchanges (2019) — link
  25. Uniswap v3 Core (2021) — link
  26. Ethereum poised for high volatility as Grayscale CEO departs several days before SEC’s ETF decision (2024) — link
  27. What role do futures markets play in Bitcoin pricing? Causality, cointegration and price discovery from a time-varying perspective? (2020) — link
  28. Investigating the Efficiency of Bitcoin Futures in Price Discovery (2022) — link

1 of 36 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.