You are reading an excerpt from our free but shortened abridged report! While still packed with incredible research and data, for just $40/month you can upgrade to our FULL library of 60+ reports (including this one) and complete industry-leading analysis on the top crypto assets.
Becoming a Premium member means enjoying all the perks of a Basic membership PLUS:
- Full-length CORE Reports: More technical, in-depth research, actionable insights, and potential market alpha for serious crypto users
- Early access to future CORE ratings: Being early is sometimes just as important as being right!
- Premium Member CORE+ Reports: Coverage on the top issues pertaining to crypto users like bridge security, layer two solutions, DeFi plays, and more
- CORE report Audio playback: Don’t want to read? No problem! Listen on the go.
CRV: Background and Borrowing
Michael Egorov’s $100m CRV backed loans have brought to light some of the risks in DeFi lending. This post will break down the risks, track the timeline, suggest some fixes, and highlight the importance of risk management in DeFi lending. Not financial advice but DeFi lenders and protocol governors should be more aware and proactive about long tail risk.
The CRV token was launched in August 2020 and became a hot topic in DeFi lending. By October 2020 the Curve team proposed to add CRV as collateral to Aave v2 which was approved by Aave’s governance without proper risk assessment. By December 2020 Michael Egorov started borrowing on Aave, accumulating $1.9m debt against a new collateral asset.
By February 2021 Egorov’s borrowing was $2.4m and his activities started to get noticed. Avraham Eisenberg saw an opportunity to short CRV and force liquidation, which failed but exposed the vulnerabilities in the Aave CRV market. Aave then made CRV non-borrowable; other markets followed but didn’t address long-tail collateral risk.
Egorov kept borrowing, and by June, DeFi alarm bells were ringing. Gauntlet recommended emergency measures to reduce Aave v2’s $65m USD debt exposure, but the proposal failed due to community opposition. The debate highlighted the divide in DeFi governance on risk management.
Although LTV was adjusted slightly Egorov borrowed more on Fraxlend, Abracadabra and Inverse Finance and accumulated over $105m debt. This is 5x CRV’s daily trading volume and a systemic risk to CRV’s market. Egorov’s monthly borrowing costs are $3.4m which increases global leverage and liquidity risk.
crvUSD
Much of Egorov's borrowing activities utilized the crvUSD stablecoin. Here is a reminder as to how that product works. Curve's native stablecoin is crvUSD, and it leverages Curve’s unique AMM model. On May 17, 2023, crvUSD launched its user interface for crvUSD, signaling the debut of its much-anticipated stablecoin. So far, crvUSD accepts collateral including : sfrxETH , wstETH, wBTC , wETH.
At the heart of crvUSD is its groundbreaking liquidation mechanism known as 'LLAMMA.' This system offers a borrower-friendly approach by converting a borrower's collateral into a Liquidity Provider (LP) position, thereby enabling a continuous adjustment of collateral using a special-purpose Automated Market Maker (AMM). This method, termed 'soft liquidations,' presents a less volatile alternative to the forceful liquidation procedures commonly seen on other borrowing platforms.
LLAMA
crvUSD is the outcome of repeated research into minimizing the impact of liquidations. The name of its stablecoin model is LLAMMA (Lending-Liquidating AMM). Its distinguishing characteristic is an automated liquidation that occurs in stages during the minting/burning process. The approach not only allows stablecoin users to disperse depeg or liquidation risks by diversifying positions into multiple bands, but it can also mitigate the consequences of a quick decrease, which has proved crippling for DeFi platforms, by liquidating assets in phases.
The Lending-Liquidating Automated Market Maker Algorithm (LLAMMA) facilitates a smoother liquidation process for borrowers as their collateral value dips compared to their debt position (e.g., loan health). This is a far cry from the abrupt liquidation common to other lending platforms like Maker or Aave.
In essence, LLAMMA cleverly transforms your collateral into a liquidity provider (LP) position – a pairing of a volatile asset and a stablecoin. Unlike the instant liquidations that lead to a total loss for the borrower, the LLAMMA mechanism minimizes potential losses and limits the protocol's risk of accruing bad debt. As the value of the collateral declines, crvUSD will leverage the liquidity of Curve pools to automatically rebalance the composition of the collateral. In this way, the platform will convert ETH proportionally to other stablecoins like USDC, creating a larger buffer and keeping a "safe" distance from the liquidation price. To prevent liquidation, debt is used to regulate the distance between the current price and the liquidation price.
To illustrate, if a borrower places ETH as collateral and borrows crvUSD, the collateral will be entirely in ETH. However, should ETH's price fall, the protocol will slowly convert some of the collateral to crvUSD. If the ETH price rebounds, the borrower can use this cash to reinstate or "de-liquidate" their LP position by buying back ETH using crvUSD.
Leverage and Global Risk Management
DeFi lending protocols lack risk management culture and don’t have the infrastructure to adapt to global leverage and liquidity. Risk management requires constraining LTV ratios or per asset borrow limits to safe levels. CRV’s leverage is unsustainable and more CRV debt is dangerous. Protocols need to prevent more exposure which means partial repayments, more collateralization and rebalancing across protocols.
However, a panicked liquidation approach would be disastrous and lead to huge value loss. Instead protocols should do gradual parameter adjustments to force partial repayments including:
-
Set LTV to 0
-
Gradually increase liquidation thresholds
-
Increase interest rates
These will minimize value leakage and maximize returns to lenders. Individual lenders should withdraw from risky markets and participate in governance forums to advocate for risk management.
The broader context of volatile DeFi assets
DeFi governance tokens like CRV are inherently volatile and concentrated among founding teams, they don’t have the broad and diverse holder base of assets like ETH or BTC. This is called “absolute liquidity depth” and means the market cap is not equal to realizable value. Even top tokens like UNI don’t have guaranteed large buy demand.
Lenders need to monitor exposure across volatile assets and adjust LTVs according to global leverage. Protocols like Aave and Compound, which can’t adjust interest rates per asset, should have very conservative LTV and liquidation thresholds for risky assets.
Long-tail assets in DeFi lending
Securing lending against long-tail assets with concentrated ownership is hard. But with a conservative sizing and a callable loan model, you can mitigate the risks. This model allows loans to be unwound without liquidation and protects the back credit. For example, if a new stablecoin increases leverage against an asset, immediate callable loans can protect protocol credit.
Egorov’s situation is bad, but market-making and price discovery across assets is important. Limiting CREDIT issuance to a subset of daily volume will keep protocol risk low as long as global leverage is under control. The Credit Guild’s flexible lending terms can do this without much political debate.
Borrowable assets
A recent Twitter thread discussed the borrowability of collateral assets. Michael Bentley of Euler said non-borrowable markets are dangerous while Nour Haridy of Inverse said large asset holders want their tokens to not be borrowed so they don’t depress the market price. A developed borrowing market is good for collateral evaluation, but the core issue is that collateral is mispriced based on market unit price rather than realizable sale price.
In the case of the Credit Guild collateral will not be rehypothecated but other system versions can support lending of many assets including ETH after Credit v1 launch.
Conclusion
The scenario of a protocol founder borrowing $100M+ against his own token is a stark reminder of the risk to lenders. Despite the community backlash against Gauntlet’s risk management the LTV to 0 is a prudent and not persecutory measure. This report is not to condemn Egorov who borrowed from willing lenders but to highlight the risk to lenders many of whom underestimate it.
