Paper: arXiv 2407.11716

Authors: Walter Hernandez Cruz, Jiahua Xu, Paolo Tasca, Carlo Campajola

Abstract

Fiat-pegged stablecoins are by nature exposed to spillover effects during market turmoil in Traditional Finance (TradFi). We observe a difference in TradFi market shocks impact between various stablecoins, in particular, USD Coin (USDC) and Tether USDT (USDT), the former with a higher reporting frequency and transparency than the latter. We investigate this, using top USDC and USDT liquidity pools in Uniswap, by adapting the Marginal Cost of Immediacy (MCI) measure to Uniswap’s Automated Market Maker, and then conducting Difference-in-Differences analysis on MCI and Total Value Locked (TVL) in USD, as well as measuring liquidity concentration across different providers. Results show that the Silicon Valley Bank (SVB) event reduced USDC’s TVL dominance over USDT, increased USDT’s liquidity cost relative to USDC, and liquidity provision remained concentrated with pool-specific trends. These findings reveal a flight-to-safety behavior and counterintuitive effects of stablecoin transparency: USDC’s frequent and detailed disclosures led to swift market reactions, while USDT’s opacity and less frequent reporting provided a safety net against immediate impacts.

Complexity vs Empirical Score

  • Math Complexity: 6.5/10
  • Empirical Rigor: 7.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper employs advanced mathematical measures (Marginal Cost of Immediacy adapted to AMMs) and econometric analysis (Difference-in-Differences) to investigate liquidity dynamics. It is data-heavy, utilizing real-world on-chain data from Uniswap pools around a specific historical event (SVB collapse) to test its hypotheses empirically.

Research Flowchart

  flowchart TD
  A["Research Goal:<br>Effect of TradFi Shocks<br>on Stablecoin Liquidity"] --> B["Methodology<br>Marginal Cost of Immediacy MCI<br>on Uniswap AMM"]
  B --> C["Data Inputs:<br>USDC/USDT Pools<br>SVB Collapse Period"]
  C --> D["Computational Processes:<br>Difference-in-Differences Analysis<br>on MCI & TVL"]
  D --> E["Key Findings:<br>Flight-to-Safety Behavior"]
  D --> F["Key Findings:<br>Transparency Paradox:<br>Higher Transparency =<br>Swift Market Reaction"]

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 adapts the Marginal Cost of Immediacy (MCI) measure to Uniswap’s AMM and conducts a Difference-in-Differences analysis on MCI and TVL for USDC and USDT.

Source sentence
by adapting the Marginal Cost of Immediacy (MCI) measure to Uniswap’s Automated Market Maker, and then conducting Difference-in-Differences analysis on MCI and Total Value Locked (TVL) in USD

Assumes 3

  • USDC is treated as the asset of interest and USDT as a control because they are similar except for transparency. market
    Source sentence
    we consider USDC as our asset of interest and keep USDT as a control. This choice is dictated by the fact that the two assets are similar in most aspects (i.e. type of backing, adoption, liquidity) except for the exposure that USDC disclosed to the SVB bankruptcy
  • Liquidity should react to shocks, following traditional financial theory. other
    Source sentence
    The underlying assumption that liquidity should react to shocks is well-rooted in traditional financial theory.
  • Miner Extractable Value (MEV) bots have negligible impact on the study. other
    Source sentence
    we consider negligible for our study

Data

  • Uniswap V3 subgraph liquidity pool data · Uniswap crypto-onchain public
    Top USDC and USDT liquidity pools on Uniswap
    Source sentence
    We analyze ten key Uniswap liquidity pools (Table 1) representing our control (USDT) and treatment (USDC) groups, selected for their consistently high TVL and volume

Costs

Unclear - sources disagree The platform classifier and the paper-text extraction give different answers, so no value is asserted.

Platform classifier: Not applicable

Paper text extraction: Modeled - The analysis incorporates the explicit trading fees of each pool (e.g., 0.01, 0.05, 0.3) when measuring marginal cost of immediacy.

Source sentence (the authors' words)
Table 1: Liquidity Pools for DiD analysis with trading fees

Finds 5

  • USDC’s TVL decreased about 19.40% relative to USDT during the SVB event. statistical
    Source sentence
    USDC’s TVL decreased ∼19.40% relative to USDT
  • USDT’s liquidity cost increased roughly 241% relative to USDC. statistical
    Source sentence
    USDT’s liquidity cost increased ∼241%
  • Liquidity provision remained highly concentrated, with Gini coefficients above 0.9 for most pairs. statistical
    Source sentence
    The Gini Coefficient remained relatively high (above 0.9) for most USDC and USDT trading pairs throughout March 2023
  • Higher-fee pools tended to have considerably more providers than lower-fee ones. statistical
    Source sentence
    higher-fee pools tended to have considerably more providers (e.g., USDCWETH3000, WETHUSDT3000) than lower-fee ones (e.g., WBTCUSDC500, DAIUSDT500)
  • The results indicate a flight‑to‑safety behavior, with USDC’s transparency prompting swift market reactions and USDT’s opacity providing a temporary buffer. other
    Source sentence
    Our results suggest that USDC’s transparency and detailed disclosures led to swift reactions, while USDT’s less frequent and opaque reporting provided a temporary buffer against immediate impacts

Checked by others

Supported by

  • DeFi: data-driven characterisation of Uniswap v3 ecosystem & an ideal crypto law for liquidity pools (2022)
    On the question: How do fee levels affect the number of liquidity providers in pools?
    Compare the findings

    This paper: Higher-fee pools tended to have considerably more providers than lower-fee ones.

    That paper: Liquidity providers prefer pools with strong liquidity taker activity, low price volatility, and higher fee tiers, while liquidity takers favor low‑fee, high‑volatility pools with ample liquidity.

    Both findings show higher‑fee pools attract more providers: A notes providers prefer higher‑fee pools; B reports higher‑fee pools have considerably more providers.

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.