[ARFC] Deploy a Lido Aave v3 Instance

[ACI]

2024-07-17: following discussion with Risk service providers and community, Risk parameters have been updated ahead of AIP vote to reflect recommendations

[ARFC] Deploy a Lido Aave v3 Instance

Author: ACI (Aave Chan Initiative)

Date: 2024-06-24

---

Summary

This proposal suggests joining the Lido Alliance as the Aave DAO and deploying an Aave v3 instance focused on the Lido ecosystem.

Motivation

Aave and Lido have historically seen symbiotic growth, with stETH being one of the premier collaterals on Aave and leveraged staking being one of the most profitable use cases for both Aave DAO and Lido users.

Lido is in the process of launching the Lido Alliance, which will reward novel use cases of staked ETH and support further growth of the Lido ecosystem. This ARFC proposal suggests that Aave supports Lido Alliance efforts by deploying a new Aave V3 ETH market for Lido. This Aave v3 instance will be designed and tuned to support stETH leverage loopers. The deployment will only include wstETH and wETH assets with E-Mode enabled.

Lido has committed incentive programs and ecosystem support for this instance in order to bootstrap liquidity and promote additional programs within the Lido Alliance.

This market will be bootstrapped with wETH from the Ahab program to attract wETH liquidity.

wETH suppliers will also be proposed to be eligible to a dedicated Merit Boost

Incentives from Lido will be detailed ahead of the AIP deployment.

Specification

The Lido Alliance’s Aave v3 instance implements the following:

• The borrow cap of wETH will be set to 90% of supplied wETH, with updates tightly controlled by the risk steward. This will ensure that stETH/wETH loops are consistently profitable and can’t go into negative territory.
• E-Mode LTV & LT are set 50 bps above all other Aave implementations, making it the most efficient loop venue in the industry.
• For the first 6 months of the Lido Alliance Aave v3, wETH slope1 will be set at 2.50%, ((currently 2.7% on mainnet and on L2s) and wETH RF set at 10% (currently 15% on all markets). This will make it the most profitable venue to loop stETH & wETH.
• wstETH RF will be 5%.

The deployment of the Aave V3 Lido Ecosystem Aave will be done by Catapulta on behalf of the Aave DAO.

Parameter	wstETH value	wETH value
Isolation Mode	No	No
Borrowable	Yes	Yes
Collateral Enabled	Yes	No
Supply Cap	650K	900K
Borrow Cap	12k	810K
Debt Ceiling	NA	NA
LTV	80.00%	0%
LT	81.00%	0%
Liquidation Bonus	6%	5%
Liquidation Protocol Fee	10.00%	10.00%
Variable Base	0.0%	0.0%
Variable Slope1	3.5%	2.50%
Variable Slope2	85%	85%
Uoptimal	45%	93%
Reserve Factor	5.00%	10%
Stable Borrowing	Disabled	Disabled
Flahloanable	Yes	Yes
Siloed Borrowing	No	No
Borrowed in Isolation	No	No
E-Mode	Yes	Yes
E-Mode LTV	93.5%	93.5%
E-Mode LT	95.5%	95.5%
E-mode Liquidation Bonus	1%	1 %

Useful Links:

Lido Alliance proposal: Lido Alliance: An Ethereum-Aligned Ecosystem - Proposals - Lido Governance

[TEMP CHECK] Deploy a Lido Aave v3 Instance

Disclaimer:

The Aave Chan Initiative is not directly affiliated with Lido and did not receive compensation for creating this proposal. The ACI and its employees hold stETH.

Next Steps

1. If consensus is reached on this ARFC, escalate this proposal to the ARFC Snapshot stage.
2. If the ARFC snapshot outcome is YAE, publish an AIP vote for final deployment and enforcement of the proposal.
3. Work with service providers and Lido Alliance to release and promote the new Aave v3 instance.

Copyright:

Copyright and related rights waived under CC0.

[EzR3aL]

Could you please provide information regaring the questions form @WintermuteGovernance seen in the TEMP CHECK here [TEMP CHECK] Deploy a Lido Aave v3 Instance - #5 by WintermuteGovernance

[ACI]

Answers to @EzR3aL and @WintermuteGovernance questions here:

The benefit of new instance is it allow wstETH users best experience for looping and keep Aave DAO as main venue for largest Lido users. It also allow DAO to be more aggressive on main Aave v3 instance where borrowers are willing to pay higher rates for WETH borrow without negative impact on our long term users.

We expect that new instance to bring net new inflows of WETH because of Lido and potential Merit incentives. So this should grow overall Aave deposits of WETH and wstETH. This also keep isolated risk to onboard Lido Alliance assets like the new LRTs from Mellow.

Aave DAO will be the sole recipient of revenue generated from the Lido instance.

Our expectation is that wtstETH and wETH migrated is offset by growth in deposits to that instance over the long term as it will be the most attractive place to loop stETH on chain. It will also benefit from Lido marketing and incentives. Long term additional Lido Alliance assets like LRT can be added to add further revenue streams for Aave DAO.

Risk stewards have agreed to adding this instance to their scope of work, and can confirm here.

Coverage by the Safety Module for this instance would need a separate governance vote.

To recap benefits for Aave DAO are many. This will solidify our position as venue of choice for wstETH loopers; improve economics of wETH borrow for our users and DAO as we can go with higher rates later; and drive inflows from Lido Alliance incentives and new assets in risk isolated manner. Long term this grows Aave TVL, market share and revenues.

We have asked Lido and the other service providers to weigh in with further details on their respective scopes so that the community has full clarity on this proposal.

[ACI]

Copying @McNut’s comment from the TEMP CHECK thread here for visibility:

As a Lido DAO contributor, we are very grateful for the continued collaboration from the Aave community. Much of the success of stETH can be attributed to the quality of the marketplace that Aave has built. In fact, Aave has single handedly contributed more to the utility and adoption of stETH than any other protocol.

This proposal to form a Lido Aave v3 instance has the potential to bootstrap the next phase of stETH growth. The staking (and re-staking) industry is evolving faster than ever and many contributors to Lido DAO believe this proposal can help provide the market with some maturity and provide users with time-tested stability. The Lido Aave v3 instance is not only an opportunity to isolate risks, but a mechanism to provide dedicated incentives for stETH users and the growth of Lido Alliance around it.

To address a few of the questions that have been raised:

• Aave DAO will be the sole recipient of revenue generated from the Lido instance. The instance is to support stETH and the Lido Alliance not to generate revenue directly from it.
• While a lower RF may optically have an impact on Aave DAO revenue, through growth we believe it will accelerate value accrual to the DAO. It is our opinion that this instance can be the foundation for a new phase of growth in collaboration with Lido Alliance members, more than offsetting the lower RF.

In effect, there is an opportunity to uniquely support as well as mitigate the risk from the growth of stETH. The proposal marks the beginning of more explicit support from Lido DAO as it promises to cultivate the stETH ecosystem. With that in mind, the reWards committee has pre-committed 225 stETH as a first step to incentivize the adoption of this instance.

In parallel, contributors at Lido DAO are working to coordinate additional forms of support to ensure the success of the new instance.

[pcx]

Is this on multi chain?

[ACI]

This proposal is for Ethereum mainnet only at the moment.

[ChaosLabs]

@WintermuteGovernance - to confirm, we would support Lido Alliance’s Aave v3 instance on mainnet as part of our engagement, similar to any other new Aave deployment.

Summary

We support @ACI’s proposed deployment and its looping-focused initial parameter recommendations. We expect the Lido Alliance instance to cannibalize some of Aave V3’s existing wETH and wstETH liquidity on mainnet. However, depending on incentives, it could generate new net inflows of both wETH and wstETH protocol-wide.

Additional Considerations

• Seven of the ten largest wETH depositors on Aave v3 mainnet are borrowing assets other than wstETH against it, while the remainder are deposit-only. It may be impossible to attract the former type of user to a wETH-wstETH isolated instance, given that wstETH will be the only other borrowable asset; this could limit the overall growth of this instance. Attracting non-borrowing WETH depositers will require strong deposit incentives to compensate for the lower Slope1 on the new instance.

• The migration of wETH and wstETH to the new instance will take time, and this instance will be the first of its kind. For these reasons, we recommend starting with supply and borrow caps roughly half as large as those currently proposed. Overall, this instance does not present new risks to the protocol.
• We also suggest reducing wstETH’s UOptimal to 45%, given its limited borrow demand.

Specification

Parameter	wstETH value	wETH value
Isolation Mode	No	No
Borrowable	Yes	Yes
Collateral Enabled	Yes	Yes
Supply Cap	650K	900K
Borrow Cap	12k	810K
Debt Ceiling	NA	NA
LTV	80.00%	82.00%
LT	81.00%	83.00%
Liquidation Bonus	6%	5%
Liquidation Protocol Fee	10.00%	10.00%
Variable Base	0.0%	0.0%
Variable Slope1	3.5%	2.50%
Variable Slope2	85%	85%
Uoptimal	45%	90%
Reserve Factor	5.00%	10%
Stable Borrowing	Disabled	Disabled
Flahloanable	Yes	Yes
Siloed Borrowing	No	No
Borrowed in Isolation	No	No
E-Mode	Yes	Yes
E-Mode LTV	93.5%	93.5%
E-Mode LT	95.5%	95.5%
E-mode Liquidation Bonus	1%	1 %

[ACI]

The current proposal has been escalated to ARFC Snapshot.

Vote will start tomorrow, we encourage everyone to participate.

[TokenLogic]

@ChaosLabs

Attracting non-borrowing WETH depositers will require strong deposit incentives to compensate for the lower Slope1 on the new instance.

We share this view with @ChaosLabs that attracting wETH deposits will be the main focus area for making the Lido Instance a success. If we are to use Liquidity Mining (LM) as a means of attracting wETH deposits, lessons from past LM campaigns indicate that users will farm deposit rewards by simply depositing and borrowing the same asset.

Users can Deposit wETH, borrow wETH and loop knowing the LM on the deposit side will offset all borrowing costs at low utilisation and then at higher utilisation, there is still a positive yield due to the relationship between the Reserve Factor (RF) (low), Uoptimal (high) and Borrow Cap near the Uoptimal. This type of user, whilst generating revenue for the DAO, is a short term mercenary user sustained only by the LM rewards. If we want to attract sustainable ETH deposits, the wETH looping strategy is not something that should be viable.

We suggest disabling wETH as collateral to prevent the recursive wETH looping strategy from being possible and enabling the LM rewards to attract new wETH deposits.

The wETH deposit rate should be sufficient to attract long term passive deposits and to do this successfully, it needs to be a viable alternative to passively holding wstETH. The yield from wETH reserve borrowers is not enough to generate the yield due to the Borrow Cap being set at the Uoptimal and the Slope1 configured less than the wstETH Index Rate. Therefore LM rewards is the ideal tool for helping bootstrap ETH deposits, but only if users can not loop wETH to farm the rewards.

Moving the Borrow Cap such that it is higher than the wstETH index rate means the wstETH yield maximising strategy is not profitable without incentives. Pursuing this as a growth strategy would lead to a reduction in AUM as soon as the LM was to stop.

Longer term, we can pivot from using revenue generated from the main Aave v3 instance on Ethereum to using revenue generated from the Lido instance to create a self sustaining liquidity market. Revenue generated from wstETH and Symbiotic LRTs reserves can be redirected to wETH depositors. wETH deposits can then earn a higher yield than the wstETH index rate and depositing wETH into the Lido instance of Aave v3 becomes a viable alternative to passively holding LSTs. Provided the users position does not exceed (1-Uoptimal) x Reserve Size, there will always be instant withdrawals into ETH which is something LSTs do not offer. The slashing risk of LSTs is then replaced by the possibility of Bad Debt which is backed by the DAO’s Treasury and Safety Module, subject to governance.

Users do not take spot price LST or LRT risk due to Aave’s CAPO oracles which means the ultimate risk profile of using the Lido instance is derived from the Aave Protocol and Lido Protocol and unwinding looping strategies can be executed gradually via Lido’s withdrawal contract that is topped up daily. This is a point of difference to other liquidity markets and targets the most risk adverse user segment of the market.

By doing this, the Lido instance of Aave v3 can become self sustainable and net positive cash flow for the Aave DAO whilst not introducing any non ETH correlated asset exposure.

Alternatively, once LM has come to completion wETH can be enabled as collateral and a GHO facilitator added. Why GHO specifically, 10x the revenue for every unit of debt relative to other stables. This means a lot less GHO debt for the same amount of revenue compared to onboarding other stablecoins like USDC. The lower the stablecoin debt, the lower the overall risk profile of the Lido instance of Aave v3 and the better positioning within the market as the lowest risk ETH yield.

Our preference is to focus on optimising the Borrow Rate and distribution of the RF revenue streams to create a sustainable ETH sink that is a viable alternative passive LST and LRT holdings and to avoid introducing any non ETH-correlated assets for the time being. We also recommend disabling wETH as collateral whilst there is any LM campaign.

[EzR3aL]

Agree here…
If wETH is enabled as collateral the exact situation as described will happen. Which will bring in high deposits but only as long as LM will be active. This will result in anger by smaller user not being able to deposit their token as the pool might be full but also in large outflows after the campaign ends.

So yes, bootstrap without being a collateral and when the market is net positive and self sustainable the asset can be activated again

[Terminal]

@TokenLogic

We suggest disabling wETH as collateral to prevent the recursive wETH looping strategy from being possible and enabling the LM rewards to attract new wETH deposits.

The concerns of abuse of the rewards are valid, however, this approach will also punish users who legitimately want to borrow wstETH with wETH as collateral. Even deeper into this, wstETH is also open to the same abuse.

What are your thoughts on an instance where rewards gets cancelled if the borrowed asset is the same as collateral?

Requires more developing but should be a more precise solution, even for future rewards seasons.

[TokenLogic]

Hi @Terminal

Users looking to use wETH as collateral and borrow wstETH will be able to do this on the main market where there is plenty of liquidity, although at a slight higher cost to the Lido instance.

Looking at the main market, there doesn’t appear to be a lot of demand for this use case.

Source: Chaos Labs

A more likely scenario, is when Symbiotic LRTs are onboarded this drives borrowing demand for wstETH. This then further makes looping wstETH/wETH more profitable for users. It is this demand for LRT/wstETH that can create the revenue which can then be directed to wETH depositors to offer a competitive yield to staking. Given staking yields are falling at the moment, this might find good traction if the current market conditions are to persist.

[Aqua99]

@ACI

E-Mode Yes Yes E-Mode LTV 93.5% 93.5% E-Mode LT 95.5%

@ACI @ChaosLabs can you explain the rationale of having e-mode enabled if the market is isolated anyway? Why should there be 2 TVLs, LTs and LBs when all the tokens in the market are within the same e-mode?

[MarcZeller]

this market will eventually host lido LRTs from the Mellow ecosystem, having a mode allows for more granularity in risk control on these new assets.

Also, it’s likely this market will host Stablecoins in the future, these will be excluded from emode.

[MarcZeller]

After a discussion with service providers, the proposal was edited ahead of AIP deployment.

We’re aiming for an activation of this new market if governance approves the AIP vote by early next week.

[Terminal]

Thanks for the data, I agree wETH/wstETH borrowers temporary ban is just a low impact secondary issue.

This measure seems fine for now.

I was aiming for a permanent solution feedback of the abuse you exposed. Granuarly cancelling rewards if the same collateral asset is borrowed seems to be the way to go.

[Jommi]

at the same time we should look at doing other Lido-friendly things such as removing LDO from isolated mode on mainnet. It has high liquidity and will only grow from here.

[MarcZeller]

Title: [ARFC-ADDEDUM] Update Lido Aave v3 Instance - Set wETH LTV to Zero
Author: ACI (Aave Chan Initiative)
Date: 2024-07-30

Note: This proposal follows a direct-to-AIP process

Simple Summary

This AIP proposes to update the recently deployed Lido Aave v3 instance by setting the wETH LTV to zero, shifting focus to the instance’s intended main use case of leveraged staking.

Motivation

Following discussions with service providers, the Lido Aave v3 instance was initially launched with wETH as collateral to allow quick growth of the wETH reserve, preparing the grounds for third-party integrations such as DefiSaver to bring wstETH liquidity and usage.

This kickstart strategy was highly successful, attracting $121 million in deposits within 24 hours, surpassing our projections. Now, it’s time to adjust the strategy by cutting the wETH LTV to zero and focusing on the instance’s primary intended use case: leveraged staking.

The motivation for this change includes:

1. Refocusing the instance on its primary use case of leveraged staking with wstETH.
2. Reducing potential risks associated with using wETH as collateral in this specialized instance.
3. Encouraging users to migrate towards wstETH for collateral, aligning with the Lido ecosystem focus.

Specification

This AIP proposes the following changes to the Lido Aave v3 instance for wETH:

• Set wETH LTV (Loan-to-Value) from 82% to 0%

All other parameters for wETH will remain unchanged:

Parameter	Current Value	Proposed Value
LTV	82%	0%
Liquidation Threshold	83%	83%
Liquidation Bonus	5%	5%
Reserve Factor	10%	10%
Variable Base	0.0%	0.0%
Variable Slope1	2.50%	2.50%
Variable Slope2	85%	85%
Uoptimal	93%	93%
Stable Borrowing	Disabled	Disabled
Borrowable	Yes	Yes
Collateral Enabled	Yes	No
Supply Cap	900K	900K
Borrow Cap	810K	810K

[x2daniel]

Hi Marc,

When will this LTV change be implemented ?
I’m asking because I don’t know what the following means : " Note: This proposal follows a direct-to-AIP process"…

Quite honestly, I did borrow WETH with aEthLidoWETH collateral. I didn’t know it was “bad”. Do I need to switch aEthLidoWETH for aEthLidowstETH ? When ?

[MarcZeller]

Hello, as we discussed on X, you did nothing wrong, this instance is centered on wsETH <> wETH leverage staking and passing wETH to LTV zero make sure that’s the main usecase.

Legacy users in a wETH <> wETH position won’t be able to increase their position size but won’t be at any liquidation risk.