Paper: arXiv 2408.07653

Authors: Wei-Ru Chen, A. Christian Silva, Shen-Ning Tung

Abstract

This paper presents a comprehensive statistical analysis of the Web3 ecosystem, comparing various Web3 tokens with traditional financial assets across multiple time scales. We examine probability distributions, tail behaviors, and other key stylized facts of the returns for a diverse range of tokens, including decentralized exchanges, liquidity pools, and centralized exchanges. Despite functional differences, most tokens exhibit well-established empirical facts, including unconditional probability density of returns with heavy tails gradually becoming Gaussian and volatility clustering. Furthermore, we compare assets traded on centralized (CEX) and decentralized (DEX) exchanges, finding that DEXs exhibit similar stylized facts despite different trading mechanisms and often divergent long-term performance. We propose that this similarity is attributable to arbitrageurs striving to maintain similar centralized and decentralized prices. Our study contributes to a better understanding of the dynamics of Web3 tokens and the relationship between CEX and DEX markets, with important implications for risk management, pricing models, and portfolio construction in the rapidly evolving DeFi landscape. These results add to the growing body of literature on cryptocurrency markets and provide insights that can guide the development of more accurate models for DeFi markets.

Complexity vs Empirical Score

  • Math Complexity: 4.0/10
  • Empirical Rigor: 8.0/10
  • Quadrant: Street Traders — empirical and applied, lower math complexity

Why this score: The paper focuses on empirical statistical analysis of token returns using standard econophysics methods (heavy tails, clustering), without advanced mathematical derivations. It is heavily data-driven with specific data sources (Binance, Coinbase, Uniswap), code links, and multiple asset comparisons, making it highly backtest-ready.

Research Flowchart

  flowchart TD
  A["Research Goal<br>Analyze stylized facts<br>in Web3 ecosystem tokens"] --> B["Data Collection<br>Historical price data<br>CEX vs DEX tokens"]
  B --> C["Statistical Analysis<br>Probability distributions<br>Tail behavior & Volatility"]
  C --> D{"Comparison Analysis<br>CEX vs DEX Markets"}
  D --> E["Key Finding 1<br>Heavy tails → Gaussian convergence<br>across time scales"]
  D --> F["Key Finding 2<br>Volatility clustering present<br>in both CEX and DEX"]
  D --> G["Key Finding 3<br>Arbitrage maintains<br>price alignment across markets"]
  E --> H["Outcomes<br>Risk management models<br>Portfolio construction<br>DeFi pricing frameworks"]
  F --> H
  G --> H

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 presents a comprehensive statistical analysis of Web3 tokens and proposes that the similarity of stylized facts between CEX and DEX markets is attributable to arbitrageurs maintaining similar prices.

Source sentence
This paper presents a comprehensive statistical analysis of the Web3 ecosystem, comparing various Web3 tokens with traditional financial assets across multiple time scales.

Assumes 2

  • Log-returns are used as the measure of returns for analysis. statistical
    Source sentence
    Following standard financial literature practice, we analyze stylized facts using log-returns [log (price(T)) −log (price(T −t))] over a given time interval t.In this work, t is, for the most part, one hour, but we will also look at t on the order of days and months when illustrating the central limit theorem and shorter (minutes) when looking at DEX price formation. We choose log-returns because it allows us to
  • Trading fees are fixed at 5 basis points and 30 basis points for the examined liquidity pools. other
    Source sentence
    Let Z represent the pool price, S the CEX price, and γ the fee constant (where γ = 1−0.3% for 30bp pools and 1 −0.05% for 5bp pools). The optimal target pool price after arbitrage is theoretically: Zopt =      γ−1S if Z > γ−1S Z if Z ∈[γS, γ−1S] γS if Z < γS (5) The interval [γS, γ−1S] is termed the "no-arbitrage region," where arbitrageurs cannot profit due to trading fees.

Data

  • Binance and Coinbase token price data · Binance, Coinbase market-prices public
    Web3 tokens (selected tokens with >2‑year history)
    Source sentence
    Our data sources include Centralized Exchanges (CEX), specifically token prices from Binance and Coinbase 2, two of the largest and most liquid cryptocurrency exchanges.
  • Uniswap ETH price data · Uniswap crypto-onchain public
    ETH paired with USDC, USDT, DAI in 5bp and 30bp pools on L1 and L2
    Source sentence
    We also gathered data from Decentralized Exchanges (DEX), focusing on Ethereum (ETH) prices from Uniswap, a leading DEX. Here, we considered 5 basis point (bp) and 30 bp liquidity pools (L1 and L2), examining ETH pairs with major stablecoins: USDC, USDT, and DAI.
  • SPY ETF price data market-prices public
    S&P 500 proxy
    Source sentence
    For comparison with traditional financial markets, we included SPY ETF prices as a proxy for the S&P 500 index and the EUR to USD exchange rate. 3.2.
  • EUR/USD exchange rate data market-prices public
    Euro to US dollar
    Source sentence
    For comparison with traditional financial markets, we included SPY ETF prices as a proxy for the S&P 500 index and the EUR to USD exchange rate. 3.2.

Recognised sources: Coinbase Binance Yahoo Finance

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 (Fees of 5 basis points and 30 basis points are incorporated into the arbitrage model and pool price analysis.); the classifier found no reported trading result, so costs are not applicable.

Finds 5

  • DEX tokens exhibit similar stylized facts (fat tails, volatility clustering, aggregated normality) to CEX tokens despite different trading mechanisms. statistical
    Source sentence
    Furthermore, we compare assets traded on centralized (CEX) and decentralized (DEX) exchanges, finding that DEXs exhibit similar stylized facts despite different trading mechanisms and often divergent long-term performance.
  • The similarity between CEX and DEX markets is explained by arbitrageurs striving to keep prices aligned. theoretical
    Source sentence
    We propose that this similarity is attributable to arbitrageurs striving to maintain similar centralized and decentralized prices.
  • ETH shows a leverage effect, though with lower significance than traditional assets. statistical
    Source sentence
    Based on these literature findings and our results in Figure 8, we can confidently state that ETH shows the leverage effect, albeit with lower significance than other stylized facts in this study.
  • Time‑reversal asymmetry (TRA) is present for ETH and comparable to SPY and EUR, regardless of exchange type. statistical
    Source sentence
    ETH exhibits TRA comparable to SPY and EUR, regardless of the exchange type (DEX or CEX).
  • CEX price leads DEX pool price in price discovery, with lagged influence up to twenty minutes and not vice versa. statistical
    Source sentence
    Figure 18 suggests that CEX leads DEX in price discovery. The price discovery appears to happen on CEX, and only when it is profitable does the price on a DEX change to match the new CEX price [1, 10, 11].

Checked by others

Supported by

  • Price Discovery in Cryptocurrency Markets (2025)
    On the question: Which market type leads ETH price discovery?
    Compare the findings

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

    That paper: Centralized markets typically lead ETH price discovery.

    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.

  • Automated Market Making and Decentralized Finance (2024)
    On the question: Do arbitrageurs align prices between CEX and DEX, creating arbitrage opportunities?
    Compare the findings

    This paper: The similarity between CEX and DEX markets is explained by arbitrageurs striving to keep prices aligned.

    That paper: Significant arbitrage opportunities exist between Binance and AMM rates.

    A finds significant arbitrage between Binance and AMM rates; B explains the CEX‑DEX similarity as arbitrageurs keeping prices aligned, supporting A's observation.

  • Decentralised Finance and Automated Market Making: Execution and Speculation (2023)
    On the question: How do arbitrageurs align prices between centralized and decentralized exchanges?
    Compare the findings

    This paper: The similarity between CEX and DEX markets is explained by arbitrageurs striving to keep prices aligned.

    That paper: Significant arbitrage opportunities exist between Binance and AMM rates.

    Both findings note arbitrage aligning Binance (CEX) and AMM (DEX) rates—A shows significant opportunities, B explains the resulting price similarity.

Qualified by

  • International Financial Markets Through 150 Years: Evaluating Stylized Facts (2025)
    On the question: Is a leverage effect present in the asset returns?
    Compare the findings

    This paper: ETH shows a leverage effect, though with lower significance than traditional assets.

    That paper: Mixed evidence is observed for the leverage effect, which is not uniformly immediate across all assets.

    A finds ETH shows a leverage effect with lower significance; B finds mixed evidence and non‑uniform immediacy across assets.

  • A Comparison of Cryptocurrency Volatility-benchmarking New and Mature Asset Classes (2024)
    On the question: Do cryptocurrencies exhibit a leverage effect similar to traditional assets?
    Compare the findings

    This paper: ETH shows a leverage effect, though with lower significance than traditional assets.

    That paper: The daily leverage effect is inverted for crypto (negative sign) compared to equity (positive sign).

    A reports a negative daily leverage effect (sign = –) versus equity’s positive sign, while B finds a positive leverage effect for ETH but with lower significance.

  • Forecasting Bitcoin Volatility: A Comparative Analysis of Volatility Approaches (2024)
    On the question: Do crypto assets exhibit a leverage effect in volatility models?
    Compare the findings

    This paper: ETH shows a leverage effect, though with lower significance than traditional assets.

    That paper: Leverage effects were inconsistent across models; only the eGARCH indicated a significant leverage effect.

    Both find a leverage effect; A sees it only with eGARCH, while B finds it for ETH but with lower significance.

Code

https://github.com/silvaac/token_style — the authors' own repository

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

Cited References (20)

  1. Econophysics review: I. Empirical facts (2011) — link
  2. Empirical properties of asset returns: stylized facts and statistical issues (2001) — link
  3. DeFi and the Future of Finance (2021)
  4. Introduction to Econophysics: Correlations and Complexity in Finance (1999)
  5. Theory of Financial Risk and Derivative Pricing: From Statistical Physics to Risk Management (2003)
  6. Ethereum Whitepaper (2013)
  7. Blockchain and Distributed Ledgers: Mathematics, Technology, and Economics (2022)
  8. Bitcoin: A Peer-to-Peer Electronic Cash System (2008)
and 12 more
  1. Time reversal invariance in finance (2009) — link
  2. Seasonality in the cross-section of stock returns (2008) — link
  3. Stylized Facts of High-Frequency Bitcoin Time Series — → 2402.11930
  4. Stylised facts for high frequency cryptocurrency data (2019) — link
  5. Growth rate of a liquidity provider’s wealth in $xy=c$ automated market makers (2020)
  6. Automated Market Making and Arbitrage Profits in the Presence of Fees — → 2305.14604
  7. Improved Price Oracles: Constant Function Market Makers (2020) — link
  8. The Zumbach effect under rough Heston (2020) — link
  9. An arbitrage driven price dynamics of Automated Market Makers in the presence of fees — → 2401.01526
  10. Exponential distribution of financial returns at mesoscopic time lags: a new stylized fact (2004) — link
  11. A random matrix theory approach to financial cross-correlations (2000) — link
  12. Quantifying the behavior of stock correlations under market stress (2012) — link

3 of 20 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.