DeFi Lending During The Merge

Lioba Heimbach, Eric Schertenleib, Roger Wattenhofer

TL;DR

This paper analyzes how the two largest DeFi lending protocols, AAVE and Compound, navigated the systemic risks posed by the Ethereum Merge. It reveals that while proactive DAO interventions successfully averted a mass liquidation crisis, they failed to adequately compensate lenders, allowing arbitrageurs to extract over $13 million in value through "hard-fork-arbitrage."

What Problem Does This Paper Address?

In September 2022, Ethereum transitioned from Proof-of-Work (PoW) to Proof-of-Stake (PoS), an event known as "The Merge." Because a segment of the community planned to maintain a PoW fork (ETHW), anyone holding ETH at the time of the fork would receive an equivalent amount of ETHW tokens.

This created a massive incentive for "hard-fork-arbitrage": users could borrow ETH just before the merge to double their token holdings on the new chain. For lending protocols like AAVE and Compound, this threatened to cause a "liquidity crunch" where utilization hits 100%, interest rates skyrocket, and lenders are unable to withdraw funds. The authors investigate whether the protocols' emergency interventions—such as pausing or capping borrowing—were effective and who ultimately won or lost in this high-stakes event.

Methods

The researchers performed an empirical post-mortem using a combination of on-chain data and financial modeling:

  • Infrastructure: They ran an Erigon archive node for Ethereum data and a Geth node specifically for the EthereumPoW (ETHW) chain to track activity across both forks.
  • Debt Tracking: They identified major borrowers and used wallet labels (from Etherscan and OKLINK) to trace borrowed funds as they were moved to centralized exchanges (CEXs) to be sold.
  • Financial Modeling: They calculated a "break-even borrowing rate"—the maximum interest a rational actor should be willing to pay to acquire the "free" ETHW tokens—and compared it to the actual interest rates charged by the protocols.
  • Risk Analysis: They monitored "health factors," liquidation events, and the accumulation of "bad debt" (uncollateralized loans) during the volatility.

Datasets

  • Ethereum & EthereumPoW Blockchains: Primary source for transaction logs, event logs, and state queries.
  • Price Data: Hourly price data for ETH and ETHW from Yahoo Finance and Chainlink price oracles.
  • Wallet Labels: Metadata from Etherscan and OKLINK to identify exchange-owned addresses and high-profile entities like Alameda Research.

Key Insights and Ideas

  • The "Lender's Dilemma": While protocols survived, lenders were the primary victims. They faced 100% utilization (meaning they couldn't withdraw their ETH) but weren't paid interest rates high enough to match the value of the ETHW tokens they were effectively "giving" to borrowers.
  • Ineffective Interest Models: Standard interest rate curves are designed for "normal" market conditions. During the Merge, even a 100% APR was "cheap" compared to the 3% instant windfall from ETHW tokens, making the interest rates insufficient to deter borrowing.
  • Systemic Risk from stETH: AAVE allowed staked ETH (stETH) as collateral. Since stETH holders did not receive ETHW, many used stETH to borrow ETH. If ETH borrowing rates had spiked high enough, it could have triggered a "liquidation spiral" for 20% of the entire stETH market, potentially destabilizing Ethereum’s consensus layer.
  • Centralized vs. Decentralized Response: While DeFi protocols struggled with interest rate models, centralized exchanges like Binance simply withheld ETHW from borrowers, a more "equitable" but arguably less "permissionless" solution.

Key Results

  • Massive Arbitrage Extraction: Arbitrageurs extracted over $13 million USD in ETHW tokens. A single entity, Alameda Research, extracted over 300,535 ETHW.
  • Interest Rate Disparity: The "break-even" borrow rate was often orders of magnitude higher than what protocols actually charged. For example, AAVE's max rate was capped at 103% APR, while the arbitrage opportunity justified rates much higher.
  • Utilization Spikes: Borrowing hit 100% of available funds on AAVE, completely drying up liquidity for lenders.
  • Intervention Success: Despite the volatility, no spike in mass liquidations or irretrievable bad debt materialized, suggesting that the DAO interventions (capping/pausing borrowing) successfully protected the protocols' solvency.

Limitations and Future Work

  • Lower Bound on Extraction: The $13M figure is a conservative estimate; the authors only tracked direct transfers to known exchange addresses and likely missed indirect or "hop" transactions.
  • Foreseeability: The Merge was a highly anticipated, "one-off" event. The authors note that protocols might not have the luxury of weeks of DAO discussion to respond to sudden, unforeseen market shocks.
  • Future Research: The paper suggests a need for more dynamic interest rate models that can automatically adjust to event-driven arbitrage opportunities without requiring manual DAO intervention.

Summary updated

Abstract

Lending protocols in decentralized finance enable the permissionless exchange of capital from lenders to borrowers without relying on a trusted third party for clearing or market-making. Interest rates are set by the supply and demand of capital according to a pre-defined function. In the lead-up to The Merge: Ethereum blockchain's transition from proof-of-work (PoW) to proof-of-stake (PoS), a fraction of the Ethereum ecosystem announced plans of continuing with a PoW-chain. Owners of ETH - whether their ETH was borrowed or not - would hold the native tokens on each chain. This development alarmed lending protocols. They feared spiking ETH borrowing rates would lead to mass liquidations which could undermine their viability. Thus, the decentralized autonomous organization running the protocols saw no alternative to intervention - restricting users' ability to borrow. We investigate the effects of the merge and the aforementioned intervention on the two biggest lending protocols on Ethereum: AAVE and Compound. Our analysis finds that borrowing rates were extremely volatile, jumping by two orders of magnitude, and borrowing at times reached 100% of the available funds. Despite this, no spike in mass liquidations or irretrievable loans materialized. Further, we are the first to quantify and analyze hard-fork-arbitrage, profiting from holding debt in the native blockchain token during a hard fork. We find that arbitrageurs transferred tokens to centralized exchanges which at the time were worth more than 13 Mio US$, money that was effectively extracted from the platforms' lenders.
BibTeX
@inproceedings{heimbach2023defi,
  author = {Lioba Heimbach and Eric Schertenleib and Roger Wattenhofer},
  title = {{DeFi Lending During The Merge}},
  booktitle = {AFT},
  year = {2023},
  url = {https://openreview.net/forum?id=M1fld5FkkI},
}

Loading related content...

Figure