[ARC] Add support for wstETH on Arbitrum Aave V3

Proposal: Add support for wstETH on Arbitrum v3

References:

Project: https://lido.fi/
Whitepaper: lido-dao/README.md at master 路 lidofinance/lido-dao 路 GitHub
Github: lido-l2/contracts/arbitrum at main 路 lidofinance/lido-l2 路 GitHub
Documentation: https://docs.lido.fi/
Audit: audits/Oxorio Lido L2 Smart Contracts Security Audit Report 07-2022.pdf at main 路 lidofinance/audits 路 GitHub
Dune: 鈥嬧wstETH on Arbitrum
wstETH: Wrapped stETH: wstETH Token | Address 0x5979d7b546e38e414f7e9822514be443a4800529 | Arbiscan
Chainlink: EACAggregatorProxy | Address 0xB1552C5e96B312d0Bf8b554186F846C40614a540 | Arbiscan
Governance: https://research.lido.fi/
Twitter: https://twitter.com/LidoFinance
Discord: Lido

Summary:

This ARFC presents the community with the opportunity to add wstETH to the Arbitrum v3 Liquidity Pool.

Motivation

The stETH Reserve on Aave v2 is the largest Reserve across all Aave deployments with $1.52B in deposits, exceeding USDC ($0.88B) and ETH ($1.37B). This is partially due to several communities having built products that deposit stETH and borrow ETH as part of a rewards-maximizing strategy. The demand for stETH has increased as a result of listing stETH on Aave v2, which has increased the Lido DAO takes a small percentage of rewards received by staking ETH.

Listing of stETH by Aave has enabled Aave to offer the best return on ETH, in ETH terms, across major lending markets, continually outperforming Compound.

By listing wstETH on Arbitrum, the Aave community is helping create an environment capable of replicating a similar outcome to that of the Ethereum Mainnet Aave v2 Liquidity Pool.

The utilization of wETH on Arbitrum is 32.23%, compared to 49.56% on the Ethereum v2 wETH Reserve. Listing wstETH on Arbitrum v3 will increase TVL for Aave and generate new demand for wETH borrowing, creating demand for wETH deposits and driving growth for Aave DAO.

Specification

1. What is the link between the author of the AIP and the Asset?

Jacob is a full time contributor to the Lido DAO.

2. Provide a brief high-level overview of the project and the token?

鈥嬧婰ido DAO鈥檚 Ethereum liquid staking protocol enhances the security of the Ethereum network by enabling more users to stake ETH. When they stake ETH through the Lido Protocol software they receive stETH. stETH tokens are fungible and liquid. The total balance of stETH in existence is based on the total amount of ETH staked via the Lido protocols plus total staking rewards minus any slashing applied on validators. stETH rebases daily.

Due to the rebasing nature of stETH, a user鈥檚 stETH balance changes daily as staking rewards are received from the network. As some DeFi protocols require a constant balance mechanism for tokens, Lido protocols offer wstETH, which is a wrapped version of stETH. wstETH keeps a user鈥檚 token balance fixed and uses an underlying share system to reflect users鈥 staking rewards.

Example:

  • Wrap 100 stETH to 99.87 wstETH
  • Continue to receive rewards on your wstETH
  • Unwrap wstETH, receive 101 stETH

When redemptions are enabled on the Ethereum network, by the Ethereum Foundation, wstETH can be redeemed for unstaked ETH and accumulated rewards.

wstETH on Arbitrum: 0x1dca41859cd23b526cbe74da8f48ac96e14b1a29

3. Explain positioning of the token in the AAVE ecosystem. Why would it be a good borrow or collateral asset?

The below shows the effect that listing of stETH has had on the Aave v2 Ethereum Mainnet deployment. stETH drives most of the wETH borrowing demand and the resulting fee-revenue. The wETH reserve generated approximately $125k in revenue during November, $217k during December and has already generated $280143k by mid January 2023.

4. Provide a brief history of the project and the different components: DAO (is it live?), products (are they live?). How did it overcome some of the challenges it faced?

The below provides a brief overview and some key dates of interest providing insight into the history to stETH and wstETH:

  • stETH deployed on 17th December 2020
  • stETH listed on Aave v2 on the 27th February 2022
  • wstETH deployed 5th August 2022
  • wstETH to be listed on Ethereum v3 deployment discussion 11th October 2022

Lido Protocol stETH is considered the most dominant staked ETH utility token across the industry. stETH tokens make up 29.5% of all staked ETH deposits.

5. How is wstETH currently used?

Aave v2 on Ethereum is the largest holder of stETH.

Deploying wstETH on Arbitrum, along with on other networks like Optimism, will make the token available to Aave users on Layer 2 networks in the same way stETH is available on Ethereum Mainnet. It will also enable DeFi protocols and products on Arbitrum to build interesting and exciting new use cases by utilizing integration with Aave to access the wstETH market.

InstaDapp, Index Coop, Galleon DAO, and others have all built products on top of Aave utilizing the recursive stETH/ETH strategy. By adding wstETH to the Arbitrum Liquidity Pool, Aave moves closer towards enabling developers to deploy similar products on Arbitrum.

On Arbitrum, wstETH is mostly used for providing liquidity on various incentivised decentralized exchanges.

The chart below shows a clear upwards trend for wstETH migrating to Arbitrum. The introduction of the Curve Liquidity Pool in November had a material impact on the distribution of wstETH on Arbitrum.

6. Emission schedule

There is no emission schedule.

7. Token (& Protocol) permissions (minting) and upgradability. Is there a multisig? What can it do? Who are the signers?

For details on stETH, which is already listed on Aave v2 Ethereum, please refer to the forum post shown below:

The custom bridging contracts for the Arbitrum native bridge have been developed by the Lido DAO team, in collaboration with the OffChain Labs team. The code was submitted for audit and the audit report can be found below:

The Arbitrum bridge was deployed by Lido DAO on L1 and L2 (Arbitrum One). The upgrade of Lido鈥檚 L1 requires a Lido DAO vote. Just like the L1 part of the setup, the L2 deployment of the Arbitrum bridge contracts is also designed to be upgradeable. The L2 upgrades are implemented via governance approval on Ethereum and leverages the Aave Governance Cross-Chain bridges contracts. As a result of the collaboration between Lido DAO and Offchain Labs, bridging wstETH tokens works the same as regular Arbitrum bridging and the user experience of the native bridge is the same. The parts of Arbitrum bridging which are controlled by Arbitrum/Offchain Labs can鈥檛 be changed by Lido DAO.

For security, Emergency Brakes multi-sigs have been set up on Arbitrum. The multi-sig can pause deposits/withdrawals. However, only the Aragon Agent is capable of resuming bridge operations. The Arbitrum multi-sig address is shown below:

The multi-sig composition is 3 from 5 of the following signers:

  • @psirex with address 0x2a61d3ba5030Ef471C74f612962c7367ECa3a62d
  • @vsh with address 0x2a96805188e583dd760785A0dE93128504DDd5c7
  • @kadmil with address 0x6f5c9B92DC47C89155930E708fBc305b55A5519A
  • @ujenjt with address 0xdd19274b614b5ecAcf493Bc43C380ef6B8dfB56c
  • @folkyatina with address 0xCFfE0F3B089e46D8212408Ba061c425776E64322

8. Market data (Market Cap, 24h Volume, Volatility, Exchanges, Maturity)

  • Market capitalisation: $7,691,227,377
  • 24H Volume ~$50M (stETH + wstETH)

Decentralized exchange liquidity pools

Balancer

Curve

Kyber Network

9. Social channels data (Size of communities, activity on Github)

10. Contracts date of deployments, number of transactions, number of holders for tokens

The below applies to just wstETH on Arbitrum:

Number of holders for token: 14,679 (liquidity pools are recorded as 1 address)

3 Likes

Hello and thanks for publishing this ARC following the standard.

having the ability for users that do not own 32 ETH and can鈥檛 afford L1 tx fees to be exposed to staking revenue is crucial for the democratization of Ethereum.

L2 are a fitting solution for this and onboarding wstETH on Aave will help the adoption of this option.

The ACI is generally supportive of this proposal and wait for comments from risk teams and suggested risk parameters before stating a definitive opinion.

4 Likes

Hi it鈥檚 Peter from Offchain Labs, the company behind Arbitrum. We鈥檙e huge supporters of this proposal and have had a great experience working with the Lido team.

As a long time stETH holder, adding support support for wstETH would unlock more flexibility and further strengthen Ethereum鈥檚 Defi initiatives on Layer 2 Ethereum.

Let鈥檚 work together to make this a success! If you ever have any questions, I can be reached via email (phaymond@offchainlabs.com).

2 Likes

Hello,

Random AAVE user/nobody. I support this as well and wouldn鈥檛 mind seeing it launched on Optimism as well.

1 Like

Hi all, Gauntlet and Chaos Labs are working to provide recommendations for the initial parameters and will share our analyses with the community in the coming days.

4 Likes

Hello!

We are supportive of extending Aave鈥檚 LSD reach across cheaper fee venues and much like our support for wstETH on Optimism, we believe wstETH on Arbitrum is a great addition.

We look forward to supporting this vote under the risk recommendations made by Gauntlet and Chaos Labs.

2 Likes

Joint Recommendation:

Given the similarities between Chaos鈥 and Gauntlet鈥檚 recommendations (analyses below), we recommend moving forward with Gauntlet鈥檚 recommendations as they are slightly more conservative.
Regarding supply and borrow caps, we recommend starting with the lower recommended caps provided by Gauntlet and considering increasing them in the future depending on supply and borrow demand.

Parameter Recommendations

Symbol Isolation Mode Borrowable Collateral Enabled LTV LT LB RF LPF Debt Ceiling Supply Cap Borrow Cap
wstETH NO YES YES 70% 79% 7.2% 15% 0.10 N/A 1200 190

eMode - LT 93%, LTV 90%, LB 2% (conditioned on stETH/ETH oracle and other necessary oracles on Arbitrum)

  • The Snapshot should include options a Yae/Nae option for emode for the community to vote on.

IR Curves

Parameter Recommendation
Base 0.0025
Slope 1 0.045
Uoptimal 0.45
Slope 2 0.8
Reserve Factor 0.15

Gauntlet Analysis

Symbol Isolation Mode Borrowable Collateral Enabled LTV LT LB RF LPF Debt Ceiling Supply Cap Borrow Cap
Gauntlet wstETH NO YES YES 70% 79% 7.2% 15% 0.10 N/A 1200 190

Gauntlet has previously provided wstETH param recommendations for Aave v3 Ethereum and Aave v3 Optimism, which serve as the foundations for the following wstETH initial parameter recommendations on Aave v3 Arbitrum.

As noted in our wstETH param recommendations for Aave v3 Optimism, parameterizing new markets provides a specific challenge due to the lack of user data to train simulation models. That said, we anticipate wstETH to act similarly on Aave v3 Arbitrum as on v3 Optimism, and thus made adjustments to the LTV, LT, Supply Cap, and Borrow Cap from our v3 Optimism recommendations that we put in place here.

We recommend the following initial parameters for wstETH on Aave v3 on Arbitrum:

Isolation Mode - No

Borrowable - Yes

Collateral Enabled - Yes

eMode - Yes, LT 93%, LTV 90%, LB 2% (conditioned on stETH/ETH oracle and other necessary oracles on Arbitrum)

wstETH 鈥榮 market risk in eMode on Aave v3 Arbitrum is relatively minimal given the conservative proposed borrow and supply caps and its correlation with the LSD assets, thus we recommend eMode settings the same as those we recommended for wstETH on Aave v3 Optimism. Note the slippage incurred on Arbitrum when swapping wstETH for WETH:

Supply Cap (in tokens) - 1200 / Borrow Cap (in tokens) - 190

Gauntlet recommends reducing the Supply Caps and Borrow Caps from our v3 Ethereum and Optimism recommendations to 1200 and 190, respectively. While wstETH on both Arbitrum and Optimism are less liquid than wstETH on Mainnet Ethereum, wstETH is less liquid on Arbitrum than on Optimism, and a supply and borrow cap closer to that recommended for Optimism (6000 and 940, respectively) increases slippage to nearly 70% in the case of a whale liquidation. Further, a sell of 330,000 wstETH to WETH on Ethereum yields a ~1.0% slippage:

compared to 900 wstETH to WETH on Arbitrum:

which is a difference of over ~99%. These diluted caps will mitigate unknown tail risks while giving the market ample time to naturally evolve and grow. In the case that the wstETH market grows quickly due to its greater use cases as compared to stETH, we can meet rapidly increasing demand with proactive updates as needed.

LTV - 70.0% / LT - 79.0%

For LTV and LT, we recommend dividing the LTV and LT of our wstETH v3 Ethereum recommendation by v2 WETH LTV (LT) / v3 Arbitrum WETH LTV (LT). The v3 Arbitrum WETH LT is 82.5%, compared to 86% on v2. Since the wstETH v3 Ethereum LT and LTV recommendations are derived from v2 Ethereum, this would allow us to initialize Arbitrum wstETH with the same relative risk profile for stETH on v2 (as compared to WETH). This results in an LT of 79% and LTV of 70% for wstETH on v3 Arbitrum.

Liquidation Bonus - 7.2%

Gauntlet recommends initializing with the same LB for wstETH on v3 Ethereum, adjusted by the ratio between WETH v2 LB and WETH v3 Arbitrum LB to account for implicit behavioral differences between liquidations on Arbitrum vs Ethereum. This results in 7.2%.

Liquidation Protocol Fee - 10%

Gauntlet has recommended a 10% LPF for the LPF of wstETH on v3 Ethereum and Optimism. For v3 Arbitrum we recommend the same. This, again, will allow for protocol revenue from LPF while minimizing effects on liquidator behavior and user experience. The LPF is a percentage of the LB that ultimately contributes to Aave reserves. LPFs, like LBs, are intimately linked to liquidation behavior, as they serve as the primary incentive for liquidators. As such, more data is needed going forward to conduct analyses on what optimal LPFs should be. A high LPF and unadjusted LB may deter liquidators from conducting liquidations due to fears of unprofitability. A high LPF with an upwards adjusted LB to ensure bonuses for liquidators similar to v2 may cause too much of the collateral to be eaten up by liquidations, which also impacts liquidator behavior.

Reserve Factor - 15%

Gauntlet recommends wstETH reserve factor to be that of the reserve factor for recommended wstETH on Ethereum and Optimism, which is 15%.

IR Curve

While it is difficult to formulate Interest Rate curves for new assets, we recommend conservative parameters that match those recommended for wstETH on Aave v3 Ethereum and Optimism.

Parameter Recommendation
Base 0.0025
Slope 1 0.045
Uoptimal 0.45
Slope 2 0.8
Reserve Factor 0.15
![wstETH 532x500](upload://fuvCVURe6Wxb6mUupZMmMlo2WxP.png)

Chaos Analysis

Symbol Isolation Mode Borrowable Collateral Enabled LTV LT LB RF LPF Debt Ceiling Supply Cap Borrow Cap
Chaos wstETH NO YES YES 68.5% 79.5% 7% 15% 0.10 N/A 2400 1320


We recommend the following initial parameters for wstETH on Aave v3 on Arbitrum:

Isolation Mode - No

Borrowable - Yes

Collateral Enabled - Yes

LTV - 68.5% / LT - 79.5%/LB - 7%

For the above parameters, we recommend the same ones that we recommended and were set for wstETH on Ethereum V3, as they are set conservatively and we anticipate similar asset volatility.

Liquidation Protocol Fee - 10%

We recommend that the LPF be in line with our recommendation for wstETH, cbETH, and rETH on Ethereum V3. This is a default parameter configuration that should be revisited with more liquidator behavior data.

Reserve Factor - 15%

We recommend that the RF be in line with our recommendation for wstETH, cbETH, and rETH on Ethereum V3.

1 Like

Hi Everyone :wave:

@llamaxyz will create a Snapshot to advance wstETH towards being listed on Arbitrum starting Monday 6th February.

  • YAE - wstETH with eMode
  • YAE - wstETH without eMode
  • NAE
  • ABSTAIN

We will provide an option to list wstETH with and without eMode. We note here Aave elected to list wstETH on Optimism without an eMode at launch.

Risk Parameters

The following parameters are to be used which are derived from the Chaos and Gauntlet post. @Llamaxyz opted to include the more conservative value for the initial listing with an expectation that the parameters will be actively maintained after the listing goes live.

Parameter Value
Isolation Mode No
Borrowable Yes
Collateral Enabled Yes
Supply Cap (in tokens) 1,200
Borrow Cap (in tokens) 190
Debt Ceiling 0
LTV 70.0%
LT 79.0%
Liquidation Bonus 7.2%
Liquidation Protocol Fee 10%

ETH eMode Category

Category Assets Included Loan to Value Liquidation Threshold Liquidation Bonus
Ethereum Correlated wstETH wETH 90% 93% 2%

Borrow Rates

Category Value
Base 0.25%
Slope1 4.50%
Slope2 80.00%
Uoptimal 45.00
Reserve Factor 15.00

Regarding the more qualitative assessment of wstETH on Arbitrum, wstETH on Arbitrum is very similar to Optimism as shared here.

As the Arbitrum bridge is owned by Lido DAO on Ethereum, no action can be taken on the bridge without explicit Lido DAO approval with exception to pausing the bridge. L1 and L2 upgrades are implemented via governance approval on Ethereum and utilise the Aave Governance Cross-Chain bridges contracts.

There is a 3 of 5 multisig on Arbitrum that performs the Emergency Brake role, which can pause deposits and withdrawals in the bridge contract. Do note resuming operations can only be performed by the Aragon Agent. Details about the signers on the multisig are mentioned in the original proposal.

We would like for @bgdlabs to share with us how the oracles will be constructed for each scenario with and without an eMode. There are some more technical considerations with respect to if it is just listing wstETH as collateral relative to if wstETH and wETH are included in an Ethereum eMode Category. Ideally, some insight into this aspect before the Snapshot vote would be ideal.

2 Likes

Hi Everyone :wave:

A Snapshot has been created.

https://snapshot.org/#/aave.eth/proposal/0x402647b83c436aecbe2fe404870f08767b5509225cbe606913e50801f87f5db8

Thank you in advance to everyone who participates in the vote.

For clarification to the community regarding the pricing of wstETH on Arbitrum.

Currently, it is possible to price wstETH on Arbitrum composing 2 existing Chainlink price feeds: stETH/USD and wstETH/stETH. It is technically feasible to use that composed oracle with the parameters recommended, including the suggested eMode.

But with eModes including Liquidation Thresholds of the suggested levels (93%) and above, the problem described HERE becomes way more meaningful, as small deviations caused by de-synchronization carry higher risk.
In this specific case, given the deviation configurations of the 2 previous Chainlink feeds, the system should work correctly, as the risk added is not really high. But exists, so for the sake of its minimization, we have already contacted Lido and Chainlink to spin up a wstETH/ETH extra feed, in order to compose it with ETH/USD.
This way, ETH (price via ETH/USD) and wstETH (to be priced via ETH/USD + wstETH/ETH) will have almost perfect 鈥渟ync鈥 volatility-wise, with their common ETH/USD layer.




To summarise, eMode activation is legitimate and the oracle dependency should not be a blocker for it. We will help to spin up the right oracle adapter; depending on the availability of the new wstETH/ETH in any case, one or another composition.
Of course, if we would notice any non-expected risk, we would comment here on the forum if any post-Snapshot change in configs is required (always would be on the conservative direction).

5 Likes

Hi Everyone :wave:

We would like to let you know that @Llamaxyz will be preparing this AIP for vote. We are currently working with Lido and @bgdlabs teams to achieve this and intend on having a proposal live in the coming weeks.

Tagging: @MarcZeller for visibility to ensure no duplication of efforts.

2 Likes

Thanks for the Tag,

The ACI will let llama post this AIP.

2 Likes

AIP-163 is live so please vote if you hold AAVE!

Small note on the AIP text: under Risk Parameters, the following values should be:

  • Stable Slope1: 0.00%
  • Stable Slope2: 0.00%
  • Stable Rate Excess Offset: 5.00%
2 Likes