Gauntlet LRT Core Vault Market Update 4/24/2024 (ezETH price volatility)

Vaults Gauntlet
1

Simple Summary

On April 24, ezETH experienced a liquidity crunch that resulted in a market price drop starting at 2024-24-04 02:25 UTC. The ezETH/WETH/0.86 LLTV market accrued 10.96 WETH in insolvencies, of which, 7.12 WETH was socialized to the Gauntlet LRT Core Vault.

As shown in the chart below, users who deposited in our Vault prior to 2024-04-22 11:27 UTC are net positive in their supply position due to accrued interest. Users supplying into LRT Core on 2024-04-22 and withdrawing after 2024-24-04 realized a slight loss.

The LRT Core vault whitelists a range of liquid restaking token (LRT) collateral markets and optimizes risk-adjusted yield. The purpose of this vault was to capture riskier yield opportunities for WETH suppliers with a higher risk appetite. Users with lower risk appetite have a variety of alternative vaults to choose from, such as our WETH Prime vault.

LRT Core Shares to Assets Conversion History
LRT Core Shares to Assets Conversion History625Ă—530 53.1 KB

ezETH Price Volatility Timeline

On April 24, 2024:

  • 02:25 UTC: The ezETH/WETH oracle first dipped below 1.0
  • 02:29:47 UTC: The first liquidation occurred at block 19722269
  • 02:52:35 UTC: 145 of the 146 total ezETH liquidations occurred
  • 13:00:47 UTC: Final liquidation

ezETH Liquidations

LRT Core contributed 65% of the ~7.6k WETH supplied in this ezETH/WETH/0.86 market at the time of these liquidations; the insolvencies socialized to our depositors come out to be 7.12 WETH, representing .0011 (11 bps) of WETH supply to LRT Core Vault.

ezETH/WETH Liquidity

The WETH Liquidity in the ezETH/WETH Balancer pool decreased from 4K WETH down to as low as 500 WETH before rebounding to 1300 WETH between 2024-24-4 00:00 UTC and 2024-24-04 08:00 UTC. A combination of factors led to liquidity decreasing:

  • ezETH holders swapping out of their position
  • Liquidators liquidating ezETH positions swapping out of the ezETH they claimed

The underlying price oracle used in this market relies on a few liquidity pools:

  • ezETH/WETH Balancer pool
  • ezETH/weETH/rswETH Balancer pool
  • ezETH/WETH Curve Pool
  • ezETH/WETH Uniswap V3 pool

Decreasing liquidity in all of the pools put downward pressure on the underlying price oracle used in the ezETH/WETH/0.86 LLTV Morpho Blue market, triggering further liquidations, which in turn led to even lower liquidity in these pools as liquidators swapped out of ezETH to realize their profits.

For more data and charts on ezETH price trajectory, liquidity, and liquidations, please see the Appendix.

Background and Overview of Gauntlet Vaults

Gauntlet operates multiple vaults with different risk profiles, and correspondingly, different yields. To name a few of them and their strategies:

  • Gauntlet WETH Prime: lend WETH against blue chip collateral (wstETH, sDAI), averaging ~3% APY
  • Gauntlet USDC Prime: lend USDC against bluechip collateral (wstETH, WBTC), averaging ~8% APY
  • Gauntlet LRT Core: lend WETH against LRT collateral (ezETH), averaging ~20% APY

Morpho Blue’s well-designed architecture isolates risk within each individual lending market, preventing contagion across markets; risky positions in one market do not affect others external to that market. This enables Gauntlet to implement differentiated yield strategies across vaults allocated to distinct markets.

Parameterization of the LRT Core Vault

LRT Core allocates user-supplied WETH into three ezETH collateral markets:

  • ezETH/WETH/0.86 LLTV
  • ezETH/WETH/0.77 LLTV
  • ezETH/WETH/0.625 LLTV

Prior to the ezETH/ETH price drop on 2024-24-04, LRT Core had a total of 5k WETH in the 0.86 LLTV market, 2k WETH in the 0.77 market, and 400 WETH in the 0.625 market. The collateral at risk under a 5% price drop was stable around ~1.5k WETH, 10% price drop at ~2k WETH on the 0.86 LLTV market. In the 0.77 and 0.625 LLTV markets, there were minimal liquidations at risk up to a 10% price drop. Considering the liquidity in the relevant ezETH liquidity pools in the past week and liquidations on other protocols, we concluded that the level of insolvency risk was acceptable when considered with the consistent yield accrued through borrower interest in these markets.

Currently, we observe no outsized risk of cascading liquidations. Given the recent volatility of ezETH’s market price and the liquidity landscape of ezETH, Gauntlet has started to slowly reallocate WETH supply out of the ezETH/WETH/0.86 LLTV market into lower LLTV tranche markets to maximize risk-adjusted returns. We will continue to update the community on our strategy for the LRT Core Vault going forward.

FAQs

  • Why does the vault allocate to three ezETH markets with different LLTVs?
    • When determining appropriate liquidation loan-to-values (LLTVs) and allocation amounts across various markets, we consider a combination of factors, including market liquidity, existing risky borrower positions, and external market risk considerations, such as highly leveraged positions on other protocols. Our analysis revealed that the ezETH market had sufficient liquidity to support more risk-averse borrowers who would maintain lower leverage compared to those in the 0.86 LLTV market. Consequently, borrowers in the 77% LLTV ezETH/WETH market primarily maintained positions with sufficiently high health factors, resulting in only 8 liquidations, in contrast to the 138 total liquidations observed in the 0.86 ezETH/WETH market. Furthermore, only the borrowers in the 86% LLTV market accumulated any insolvent debt.
  • Will LRT Core move to markets with different oracle setups (ex: an exchange rate oracle) going forward?
    • We are looking into this. It is likely that an exchange rate oracle would have led to fewer borrowers getting liquidated. Note that our WETH Prime vault allocates WETH into wstETH/WETH markets that use exchange rate oracles.

Next Steps

Given the recent volatility of ezETH/ETH’s market price, Gauntlet has reallocated a portion of LRT Core’s WETH supply into the 0.77 and 0.625 markets to maximize risk-adjusted interest yields. We will continue to update the community on our strategy for the LRT Core Vault here going forward.

Appendix

ezETH/ETH Price Trajectory

The chart below compares the RedStone and Chainlink ezETH/ETH oracle price updates around the time of the ezETH price volatility. While the feeds were in sync for the most part, the RedStone was quicker to respond to the price action and sent more price updates in the below timeframe than the Chainlink oracle.

image 2
image 2618Ă—532 63.6 KB

image 3
image 3616Ă—532 62.6 KB

Similar to the stETH dislocation event, there was a significant reduction in liquidity leading up to the large ezETH dislocation. Although WETH liquidity has partially recovered in the Balancer pool, it remains 65% below the levels seen before the dislocation event.

image 4

image 5
image 5739Ă—610 53.1 KB

Liquidations that resulted in Insolvencies

Here are the liquidations that resulted in insolvencies:

9 Likes
2

Hello,

I am curious what Gauntlet’s stance is on using single collateralised vaults using extremely unproven, untested, locked, and young tokens such as ezETH? Although the yield may be “higher” it is quite evident that you can have a similar yield while adding other assets as collateral to reduce this overall risk? Why wasn’t this taken into consideration?

Finally what other risk management practises will Gauntlet start to use moving forward as a leading risk manager in the space?
I.E.)

  1. Issuing vaults with more than 1 collateralized asset
  2. Issuing vaults where the collateralized asset has a lower risk, etc. no luck-ups
  3. Increasing the timelock of Gauntlet vaults for depositors to exit
  4. Adding trustless vault guardians for depositors
  5. Adding information about smart contract risks with tokens such as ezETH

Thank you!

3

Hi @PRado , thanks for your questions.

  1. We had always been planning on adding additional LRT collateral assets to the vault. Our primary considerations for this vault were the following:

    • optimize risk-adjusted returns
    • slowly onboard additional collateral assets as there were and continue to be a lot of unknowns in the restaking space
    • avoid overly fragmenting liquidity. WETH suppliers have a choice of multiple vaults to allocate into. It is possible that we could have created a separate vault that would supply into a different LRT collateral market but we wanted to avoid creating yet another WETH vault which would further fragment liquidity.

    At the time of vault creation, there were a number of developments happening in the restaking space such as:

    • the launch of AVSs
    • incentives to provide liquidity
    • start/end of certain LRT airdrop campaigns, which have an outsized impact on liquidity

    Given the various upcoming developments that would each have some potentially large impact on LRT collateral liquidity, we felt it was prudent to start with one collateral asset and slowly add additional collateral asset markets over time.

    It is difficult to say definitively whether we could have generated similar yields at the same volume with a different initial LRT in the vault. Every LRT is still somewhat untested given how new the restaking space is. It is true that other LRTs have withdrawals enabled. This was a risk we felt comfortable with given the amount of borrowing demand and relative profitability of leveraged restaking in ezETH compared to other LRTs. We believed that the potential rewards outweighed the risks, and we were confident in our ability to adequately mitigate the majority of the downside associated with the lack of withdrawals with appropriate supply caps, and reallocations to lower LLTV markets when necessary.

    Another important consideration we had with respect to adding additional LRT collateral was the risk of correlated price drops against ETH. At the time of vault creation, one of the largest LRT liquidity pools were comprised of LRTs being paired against each other, for instance this ezETH/weETH/rswETH Balancer pool. So large liquidations for one of these LRTs would end up routing through this tri-LRT pool and eventually through the ezETH/WETH Balancer pool, which was the deepest source of WETH liquidity for any LRT at the time. The presence of this shared liquidity pool for LRTs meant that if multiple LRTs were to lose par against ETH simultaneously, liquidators for each LST would all compete for the same limited WETH liquidity, exacerbating the impact of the price movement. To mitigate this risk of correlated liquidity issues arising from multiple LRT collateral assets in the vault, we decided to initially limit the vault to ezETH and re-evaluate the inclusion of other LRTs over time, contingent upon evolving market/liquidity conditions.

  2. Yes. We have a number of other vaults such as WETH Prime, and USDC Prime that allocate the supply asset to lower risk collateral assets. As mentioned in the previous point, we also want to be mindful of overly fragmenting liquidity, which is why we did not create a separate LRT vault for a different LRT.

  3. Vault suppliers can withdraw at any moment, subject to available liquidity in the vault. The set of timelocked actions are:

    • initializing or increasing a supply cap on a Morpho Blue market
    • decreasing a supply cap is not time locked
    • setting a new guardian

    For reference, see: Roles & Capabilities | Morpho

    Or directly in the smart contract: metamorpho/src/MetaMorpho.sol at main · morpho-org/metamorpho · GitHub

  4. We are considering this and will update the community as we make a decision. As a reminder to the community, guardians can do the following:

    • revoke a pending market supply cap increase
    • revoke a pending timelock decrease
    • revoke a pending guardian change

  5. Yes, we will look to provide resources and other relevant material in the future.

4

Thank you for the detailed response.

  1. Thank you.

  2. I wonder if there is a way to provide an overview of a risk level in the vaults UI. For a typical depositor all you focus on is APY and it’s hard to determine a health score? Will Gauntlet include that in forum posts in the future?

  3. Seems to me increasing the timelock is a very easy and low-lift method for increasing safety?

  4. Guardians are the community (depositors) thus all 3 things you mention here are huge benefits for depositors to have veto power over. Revoking a supply cap increase in fact could have potentially deflected less damage in the case of the ezETH liquidations. And depositors could have, if they wanted, then opened up a conversation on this and a better decision potentially made. Looking forward to hearing more as I have deposits in a few vaults.

  5. Excellent. Should be standard practise at least in the forum.

Thank you for the response and looking forward to seeing how Gauntlet as a large vault curator improves their vaults and thus the entire Morpho ecosystem.