The decentralized stablecoin sector holds vast market potential, yet most projects have failed midway through their journey. LUNA and UST, once boasting tens of billions in market capitalization, collapsed to zero within a short period. Now, Curve, which commands a TVL of $3.7 billion, has entered the competition.
Recently, Curve released the whitepaper and code for its stablecoin. According to GitHub, the whitepaper was completed in October and is not the final version. Although the whitepaper does not explicitly state the name of Curve’s stablecoin, the code reveals that the full name is “Curve.Fi USD Stablecoin,” abbreviated as “crvUSD.” The inflation of CRV has drawn criticism; however, revenue from the stablecoin’s stability fees and PegKeeper income may help address Curve’s insufficient revenue issue. This whitepaper primarily introduces several innovations of crvUSD: the Lending-Liquidating AMM Algorithm (LLAMMA), PegKeeper, and monetary policy. A Smoother Liquidation Algorithm: LLAMMA In the previous CRV long-short battle, a whale used USDC as collateral to borrow CRV from Aave for shorting. When the value of the CRV debt appreciated to 89% of the USDC collateral, the loan began to be liquidated. Although various lending protocols differ in parameter design, the overall approach is similar: after reserving a certain liquidation penalty ratio (to incentivize liquidator participation) and market fluctuation buffer, they maximize capital utilization as much as possible. When market liquidity is insufficient, bad debt arises, as happened with Aave. Related reading: “Whale Shorts CRV, Aave Pays Millions? Why Do On-Chain Lending Protocols Generate Bad Debt?” Short-term concentrated large-scale liquidations can also have a severe impact on the market. As shown in the figure below, during the market decline in June, a single liquidation/stop-loss operation in MakerDAO directly caused the market price of ETH on Uniswap to drop from $1,300 to below $1,000. To mitigate the impact of the liquidation process, crvUSD adopts this smoother liquidation algorithm, LLAMMA, which is the primary innovation of crvUSD. First, crvUSD is still issued through over-collateralization, but it employs a special-purpose AMM to replace the traditional lending and liquidation process.When the liquidation threshold is reached, liquidation does not occur all at once, but rather as a continuous process of liquidation/de-liquidation.
For example, using ETH as collateral to borrow crvUSD. When the value of ETH is sufficiently high, the collateral remains unchanged, similar to traditional collateralized lending. When the ETH price falls and enters the liquidation range, ETH begins to be gradually sold off as the price declines. After falling below the range, the collateral becomes entirely stablecoins, and further price declines do not cause changes, which is also consistent with other lending protocols. However, within the intermediate liquidation range, if ETH rises, Curve will use stablecoins to repurchase ETH on behalf of the user. If the price fluctuates within this intermediate liquidation range, the process of liquidation and de-liquidation will repeat continuously, with ETH being sold and repurchased repeatedly. This process resembles the impermanent loss hedging after providing liquidity in an AMM, such as providing ETH/USDC liquidity on Uniswap. When ETH rises, ETH is passively sold; in order to prevent dramatic changes in coin-denominated capital during liquidity provision, one must buy ETH from the market. Similarly, when ETH falls, ETH is sold to hedge. Compared to one-time liquidation lending protocols like MakerDAO, if the market rebounds after a liquidation event, users in MakerDAO are left with only a small residual value after liquidation, whereas in Curve, ETH would be repurchased during the upward move. In tests conducted by the Curve team, when the market price fell 10% below the liquidation threshold and then rebounded, the user’s collateral loss was only 1% within a 3-day time window. Although this algorithm can indeed reduce user losses from liquidation during extreme market conditions, and even keep losses to a very low level after price recovery, it is not entirely without drawbacks. It may be more prone to triggering liquidation. For example, if ETH is pledged as collateral to borrow stablecoins under the same conditions on Aave and Curve, when price fluctuations are relatively small, positions on Aave may not be liquidated, while Curve may experience the liquidation/de-liquidation process, resulting in small losses for the user. Automatic Stabilizer and Monetary Policy With sufficient collateral, the price of crvUSD can be pegged to 1 USD. However, prices in the secondary market are volatile, so the price of crvUSD will fluctuate slightly around 1 USD.When the price of crvUSD is above its peg (ps > 1), i.e., higher than 1 USD, due to increased demand, the PegKeeper can mint crvUSD without collateral and deposit it unilaterally into a stablecoin swap pool, driving the price of crvUSD down. Even though the minting process is uncollateralized, it can be implicitly backed by the liquidity provided in the liquidity pool. When the price of crvUSD falls below 1 USD, the PegKeeper can withdraw some crvUSD liquidity, allowing the price to recover to 1 USD. In this process, the PegKeeper effectively sells crvUSD when it is above 1 USD and buys crvUSD when it is below 1 USD, which is profitable and maintains the price stability of crvUSD. A similar pegging mechanism is also adopted in Frax’s AMO, which can avoid the influence of centralized stablecoins in the MakerDAO PSM mechanism. Monetary policy controls the relationship between the PegKeeper’s debt (dst) and the crvUSD supply. For example, when the debt/supply ratio exceeds 5%, parameters can be changed to incentivize borrowers to borrow and sell the stablecoin, forcing the system to burn dst. When the debt/supply ratio is low, borrowers are incentivized to repay loans, causing the system to increase debt. Thoughts and Summary Although the collateral example in the whitepaper is ETH, the more important assets on Curve are those in the stablecoin pools. If LP tokens from pools such as 3pool could be used as collateral, capital efficiency would be improved. The Curve team itself controls a large amount of veCRV as voting power, which can guide liquidity between crvUSD and pools such as 3pool, facilitating the cold start of crvUSD. This is an advantage that other stablecoin projects do not possess. Curve does not intend to issue a new governance token for the stablecoin. Currently, Curve operates as a DEX, where trading fees for major pairs are only 0.05% or 0.01%, resulting in limited revenue. The introduction of stability fees and the PegKeeper may improve Curve’s revenue situation. Most lending protocols use Chainlink oracles, while Curve uses the price oracle from its own DEX. Therefore, only assets already listed on Curve may be used as collateral, which also saves on oracle costs.
The ETH mentioned in the white paper example is already included in the tricrypto pool.
Since Curve provides lending based on transactions, if the borrowing cap and liquidation threshold for each token are controlled according to liquidity, it is theoretically possible to completely eliminate bad debt caused by untimely liquidation. Statement: The article only represents the author’s personal views and opinions, and does not represent the views or positions of this website. All content and views are for reference only and do not constitute investment advice. Investors should make their own decisions and transactions. The author and this website shall not bear any responsibility for direct or indirect losses arising from investors’ transactions. Disclaimer: This article is only intended to provide market information. All content and views are for reference only and do not constitute investment advice, nor do they represent the views or positions of this website. Investors should make their own decisions and transactions. The author and this website shall not bear any responsibility for direct or indirect losses arising from investors’ transactions.
