Curve Incident: Lessons & Opportunities
Introduction
The recent Curve exploit hack sent ripples through the DeFi space, with many decentralized exchanges susceptible to downstream consequences, including runaway bad debt. Kinza Finance is on a mission to establish sustainable lending on BNB Chain with security features and measures to ensure protection against even the most unforeseen situations and hacks. It’s only been a week since the Curve exploit incident began, but Kinza Finance has already deployed resources to explore, develop, and implement new features to mitigate potential consequences of an event like this one. This article outlines a few of those development ideas, including the creation of an Interest Rate Controller, as well as an Algorithmic Loan to Value Controller.
The Exploit
On Sunday, August 30, Curve Finance, a leading Decentralized Exchange with $2.5 billion in TVL prior to the exploit, experienced a hack due to a Vyper reentrancy lock malfunction, resulting in a loss of over $50 million. This caused a domino of downstream effects, including a drop in the $CRV token price. Lending protocols, including AAVE V2 (without isolated lending pools), became vulnerable due to large amounts of $CRV used as collateral for >$180 million loans. The Curve Finance founder himself, Michael Egorov, has a loan of around $70 million USDT using CRV as collateral on Aave v2. With a falling $CRV token price, these loans are becoming massively unhealthy and dangerously close to liquidations, and outcomes of bad debt for the protocol. To add fuel to the fire, circulating $CRV on DEX’s is limited, meaning there isn’t enough liquidity available on the open market to liquidate these loan positions and rebalance or return the accounts and protocol to health.
Proposals for DEX actions and responses are popping up across the DeFi space and in frantic governance conversations to minimize damage. However, there has already been incredible damage done, and the incident is still unfolding by the day. Responses to events like these need to occur rapidly, and the best way to achieve this is to build mechanisms, safeguards, and features to react automatically. It’s time to brainstorm.
Evolving Kinza Finance
Kinza Finance has kept tabs on the space to see how other protocols react, and to mine inspiration for protecting the Kinza protocol by innovating and integrating new features and solutions. With many possible options being discussed by the development team, two promising features have been selected for experimentation and testing.
The two features are:
- Time-Adaptive Interest Rate Controller
Interest rates are reactive to utilization. Interest rates for each lending pool evolve over time, depending on how much asset is currently used on the platform. In general, the more an asset is utilized, the higher the Interest Rate becomes, in order to encourage debt repayments and additional supplying. When an asset pool is available, and utilization is low, interest rates also fall to encourage borrowing. This works most of the time, but when an incident like the Curve exploit occurs, utilization rates change rapidly, and current mechanisms are not sufficient to adapt in real-time. An Algorithmic Interest Rate Controller could adapt in rapidly changing conditions, and increase rates sufficiently to force borrowers to repay their loan and prevent bad debt from developing.
In the CRV situation, as the controller forces interest rates to progressively trend upwards, the controller would add an increasing incentive to push lenders like Michael to repay their loan on this platform first. This way, even if the position gets liquidated, it would occur earlier than similar loans on other lending protocols, making use of the limited DEX liquidity first.
How would it work?
The basics of the IR Controller are to automate interest rate adaptation according to the liquidity market. For each asset, there is an optimal Utilization rate, and controllerHalf-Life.
If Utilization is above the target rate, the IR rises as time passes (each Half-Life), encouraging more lending and less borrowing, bringing the Utilization back towards the target range.
This IR adaptive response system attempts to exponentially adjust interest rates to incentivize the needed actions to bring the system back into balance. The team is already experimenting with this feature.
2. Algorithmic LTV Controller
Asset risk attributes and factors are critical for properly managing the LTV’s in each given lending market. In the recent incident, the available circulating liquidity of $CRV in DEX’s, or lack thereof, is playing a big role in the possible negative outcomes to vulnerable lending protocols. For the loans with $CRV as collateral, in case liquidation occurs, liquidation bots can not sell the CRV collateral effectively on the open market because of this lack of liquidity. An algorithmic LTV controller will be integrated into the Kinza Finance asset risk analysis process, so that current DEX liquidity and token price volatility are factored into the risk parameters and LTV equations.
Many lending protocols currently have governance processes to alter and change LTV parameters for a given lending market. However, these on-chain metrics are constantly changing, and updating risk parameters through slow governance processes can lead to delayed decision-making, and negative consequences to the protocol. Building an algorithmic LTV controller that factors in DEX liquidity and price volatility will help this situation. The Kinza Finance team will strive to provide top-level security for safe lending on BNB Chain.
Into the Future
The Kinza team is currently exploring and developing these features, and updates will be given on the progress and possible implementation timelines. These incidents are shocking, with so much capital affected, and many downstream effects unknown. But they also provide an incredible opportunity to rethink possible vulnerabilities, and to develop innovative approaches to ensuring a stable and safe protocol.
Stay up to date with all things Kinza Finance, and join our communities at the following links:
Discord: https://discord.gg/JFXTEp8Nub
Telegram: https://t.me/kinza_finance
Twitter: https://twitter.com/kinzafinance
Website: https://kinza.finance
Documentation: https://docs.kinza.finance
Kinza V1 Mainnet: http://app.kinza.finance
