A proposal to implement risk and market parameter changes in the Liqwid v1 ADA market. Given the upcoming launch of Liqwid v2 features which includes risk controls implemented in Aave v3 (borrow caps, supply caps, efficiency mode, isolation mode) the protocol is able to support more Cardano native tokens starting with Optim Bond Tokens. As a result we suggest the ADA market undergo the following updates:
Implement a borrow cap at 80% of total ADA market supplies
ADA Risk Parameter Updates: Increase the maxLTV from 75% to 80.9% and the LiquidationThreshold from 80% to 81%.
Increase the base rate from 2 to 3%.
Increase the utilMultiplier from 25% to 40%.
Reasoning
In order to securely onboard new collateral types to the ADA market a borrow cap is used to limit the amount of ADA borrowed from the liquidity pool to a certain amount to minimize liquidity pool insolvency risk.
Increasing the maxLTV from 75% to 81% allows ADA suppliers to open more capital efficient loans than previously possible (75% maxLTV is ~133% minimum collateral ratio and achieves 4x leverage, 80% maxLTV is 125% minimum collateral ratio and achieves 5x leverage). Increasing the leverage ADA suppliers can achieve borrowing stablecoins in the protocol from 4x to over 5x has been a request of multiple community members since v1 launch.
Important to note other than Liqwid v1 there are no other lending/CDP protocol on Cardano that use a liquidation buffer and there’s not strong user demand for this feature evidenced by protocols that have been live on mainnet for 1year+ still having not implemented it. Instead in these protocols users accept they must manage their debt positions or risk liquidations as soon as they reach their minimum collateral ratio/maxLTV. This is the standard set nearly unanimously by the lending/CDP protocols on Cardano and Liqwid should follow in this direction with a large reduction. Therefore we propose liquidation buffers be significantly reduced but not fully removed.
The only collateral asset that will continue to have a wide liquidation buffer is DJED. All other assets will have a 0.1% liquidation buffer.
A notification will be included in the app warning users of this in the borrow modal if this vote passes.
3. Increasing the base rate to 3% to properly configure the yield curve to reflect the ~3% risk free ADA staking reward rate. This will ensure Liqwid ADA suppliers are always earning a Net Supplier APY (real yield from borrower interest fees and staking yield alone) above 3% before including LQ emissions, even during periods of low market utilization as the base borrow rate at 0% utilization will be 3%. At current ADA market utilization of 3.17% this translates to a real yield of 3.009% as opposed to the current 2.78% Net Supplier APY.
4. Increase the utilMultiplier from 25% to 40% will accelerate the growth in borrowing costs as the ADA market sees more adoption and utilization. ADA suppliers will earn increased yields for their deposits as the demand for ADA loans in the protocol grows.
Implementation
The multisig transaction to update these ADA market and risk parameters has been tested on Preview.
Do you support these updated ADA Market and Risk parameters to securely onboard additional Cardano native token collateral types to the ADA market?
I included some additional context in the Reasoning section for each proposed update. The liquidation buffer is being reduced but not fully removed. The reasoning is Liqwid v1 remains the only lending/CDP protocol on Cardano that uses any liquidation buffer and there’s not strong user demand for this feature evidenced by protocols that have been live on mainnet for 1year+ still having not implemented it and with no real push from Cardano DeFi users to implement buffers. Instead in these protocols users accept they must manage their debt positions or risk liquidations as soon as they reach their minimum collateral ratio/maxLTV. This is the standard set nearly unanimously by the lending/CDP protocols on Cardano and Liqwid should follow in this direction with a large reduction.
The only collateral asset that will continue to have a wide liquidation buffer is DJED. All other assets will have a 0.1% liquidation buffer.
A notification will be included in the app warning users of this in the borrow modal if this vote passes.
Hey DC, have you looked at or considered looking at a 3rd party tokenomic optimization/economic security analysis? Just my 2 cents is that for DeFi (of any sort) it does provide some value specifically like how AAVE uses Chaos Labs for example. Just curious about your thoughts since the TVL since Liqwid has opened up is definitely growing.
We have our own Risk Management framework, but we cannot compare today Liqwid’s TVL to AAVE’s TLV, despite having soon the features-parity.
We have analyzed in details the work AAVE team (incl. Chaos Labs work) did, and once we have implemented the borrow and supply caps, we will be able to control our risk exposures in a more granular way.
If we grow once to ~$1B TVL, then having a third-party making dedicated market and volatility analysis would make sense.
Thanks for the info here! Definitely understand the need to have a partner at a certain TVL milestone but would love the chance to maybe talk about spot helping/working with Liqwid on future risk and economic security analysis. Or at least to have a chat about it! Thanks Florian
Would be great to see a risk focused subDAO team to make future recommendations on market and risk param updates. You are right AAVE and MakerDAO have entire 3rd parties that outsource this work to. It’s important and scales with the protocol’s TVL.
@FlorianVolery and @DC1 just wanted to reach out and follow up on this from a couple of months ago. We briefly exchanged on the topic of doing some economic security analysis and risk management. I know that we didn’t really get that far, but wanted to reach out again. I’m from Token Dynamics (https://tokendynamics.xyz) and we’re just wrapping up doing the whitepaper and economic security analysis for Hoyu (https://hoyu.io/) a new lending protocol that is EVM-based. If you’re interested to see the work we did for Hoyu as an example, I can share with you guys some of the deliverables we have prepared for Hoyu to give a glimpse of what we are about and do. Would love to at least have a call with you guys to open conversations about this.
The recent oracle vulnerability around oracles is one we would love to chat about too. I’ve been around in the Cardano space a while and understand that the infra and tooling to support various protocol features are still in the works and being developed, but we’d love to help build and make Liqwid more robust if you’re open to the conversation! Thanks.
