Executive summary
In this proposal, we suggest allowing certain DAO-approved Vaults to command 100% minting threshold (otherwise known as LTV) for osETH, enabling automatic issuance of osETH against deposits and maximizing capital efficiency. To mitigate the risk of minting osETH without providing excess collateral, we suggest setting a high bar for Vaults that wish to do so. We propose that the DAO require such Vaults to pass technical & commercial due diligence, have a TVL of at least 10,000 ETH / 5,000 GNO, not exceed a threshold fee level, and post a significant (5M SWISE on ETH / 1M SWISE on GNO) bond in order to receive this capability. These requirements will ensure that only the most advanced and committed Vaults will receive the top perks of using StakeWise, derisking the rollout of the feature that is crucial for consumer liquid staking products, large-scale integrations, and highly efficient DeFi strategies.
Motivation
With the recent introduction of granular LTV threshold setting for Vaults, the StakeWise DAO has achieved the capability to increase or decrease the osETH minting threshold based on any given Vault’s risk parameters.
This capability is exciting because it opens up StakeWise Vaults usage for:
- more liquid staking integrations (e.g. an exchange running an internal staking service with osETH as a token for DeFi, like Coinbase and cbETH),
- consumer-focused liquid staking products (have some cool stealth ideas here hehe), and
- highly efficient DeFi strategies (e.g. leveraged staking with 14x loops, creation of osETH-based LRTs, and more).
In short, the Vaults tech is capable of handling any staking use case, but with 100% minting threshold, it will handle any liquid staking use case, too.
However, enabling this capability comes with a risk, for two reasons:
- removal of overcollateralization introduces the risk that slashing in a 100% LTV Vault will affect osETH more broadly, and
- osETH APY calculation mechanics are changed whereby a 100% LTV Vault with the lowest APY will always dictate the osETH APY in order to maintain at least a 1:1 ratio between stake and osETH across different Vaults.
While point 1 is rather straightforward, point 2 is less digestible, so let us illustrate:
Imagine a scenario where there is at least one Vault with 100% minting threshold for osETH.
Using the current osETH APY calculation mechanics that determine the APY based on the weighted average of APYs in various Vaults (using their amounts of osETH minted as weights), a 100% LTV Vault may become undercollateralized whenever its APY is less than the osETH APY.
To see this, assume the osETH APY has been 3.0% and the Vault APY has been 2.9%. Over a year’s time, a 100 ETH deposit in the Vault will earn 2.9 ETH in rewards, while the value of minted osETH will grow by 3 ETH, leading to a 0.1 ETH shortfall.
Unlike with 90% LTV Vaults, there is no buffer to cover for this discrepancy in the case of 100% LTV Vaults. If the osETH APY outpaces the Vault APY, undercollateralization will ensue.
To combat this, the new osETH APY calculation mechanics will always set the osETH APY to be the lowest of the APYs among 100% LTV Vaults.
These risks mean that enabling 100% LTV minting for any given Vault must require heavy due diligence and certain guarantees from Vault owners in order to shield the StakeWise DAO and osETH users from a potential threat.
In the absence of on-chain methods for controlling these risks, we believe the StakeWise DAO should adopt certain off-chain standards that any Vault willing to receive 100% minting capability should meet.
Conceptually, the goal is to have 100% LTV Vaults operated by teams with a) an impeccable track record, b) appropriate risk management processes in place, c) long-term plans for their Vault, and iv) skin in the game.
Here are the hard requirements that we believe must be met for any team looking at opening a 100% LTV Vault:
Must have ≥10k ETH / ≥5k GNO staked for at least 1 month.
This threshold ensures that the effect of unstaking flows, using own MEV escrow, and receiving new deposits on the Vault APY is minimized.
Fee ≤5% (≤15% on Gnosis).
This threshold seeks to prevent osETH APY dilution from Vaults with high fee percentages.
Vault must consistently rank above the median performance for all stakers in the network.
This ensures an above-average performance from Vaults that have the biggest impact on the StakeWise DAO product.
Must be updated to Vaults version 3 (v3).
This ensures that the Vault leverages the latest exit queue mechanics and has access to the newest StakeWise features. Read more about the v3 upgrade here.
Vault Admin must lock 5M SWISE for ETH or 1M SWISE for GNO in a DAO-controlled address. SWISE is returned on exit and used to cover losses in case of slashing.
This ensures that the Vault operator has skin in the game and is incentivized to perform to the highest standard.
Vault must pass a DAO vote for inclusion in the osToken vault list.
This ensures that the DAO is ultimately in control of which Vaults get access to the feature.
All of these criteria must also hold after the Vault has received the 100% minting capability, and throughout the Vault lifetime.
Specification
The introduction of this feature does not require calling on smart contracts - instead, we seek DAO approval for these requirements as a policy for all Vaults seeking 100% LTV capabilities.
Considerations
We need your feedback on the proposed parameters, especially around the max fee levels and the size of a SWISE bond.
Our choice of the maximum fee is driven by the market average level; go higher and you risk negatively affecting the osETH / osGNO APY during times when the pressure on LST yields is as strong as ever. However, with this requirement we risk pushing away those potential integrations that seek to capitalize on an exclusive access to some user groups, i.e. are capable of extracting higher fees because of their captive user base (e.g. exchanges). Our suggested approach is to start low (i.e. 5% / 15% fee max for ETH and GNO, respectively), and relax the requirement if it makes sense for the StakeWise DAO commercially.
Our choice of the SWISE bond size is driven by considerations around the current price of SWISE and the economics of running a large (>10,000 ETH / >5,000 GNO) Vault at a 5% / 15% fee. If the required investment in SWISE is too high, we might not see any Vaults with 100% LTV at all, hurting our growth prospects. At the same time, the bond should be significant enough to force the operator to have substantial skin in the game. We suggest to start off at the proposed levels (5M SWISE for ETH, 1M SWISE for GNO) to test out the demand and adjust this amount based on the results and operator feedback.
Interestingly, SWISE bonding mechanisms for 100% LTV unlock a path where the token can be lent out to operators / staked to various Vaults for a prolonged period of time in exchange for an interest rate or a revenue cut, adding more utility to SWISE and encouraging long-term ownership.
Discussion
As always, we welcome questions, comments, and feedback about the proposal, so do not hesitate to post and engage!
Vote
We have created a Snapshot vote for this proposal that will last for 10 days. Vote with your SWISE here: https://forum.stakewise.io/t/swip-24-establish-a-policy-for-enabling-100-oseth-minting-in-select-vaults/1724
PS. edited TVL threshold for Gnosis to 5,000 GNO from 2,000 GNO (typo spotted).