[RFC]: Promote cbETH to collateral tier

Author(s): Seraphim Czecker
Submission Date: 31.10.2022
RFC: Promote cbETH to collateral tier

Simple Summary

This is an RFC to temperature check the idea of Coinbase ETH (cbETH) as collateral asset. A potential eIP would also include increasing borrow factors and decreasing reserve factors.

The proposed parameters are: collateral factor (CF) of 0.50, borrow factor (BF) of 0.80 and reserve factor (RF) of 0.05.

Edit: Also to change pricing from Univ3 to Chainlink.

Motivation

While the liquidity of cbETH is much lower than on stETH, cbETH could present a huge commercial opportunity for Euler as it did for wstETH (130mil TVL on Euler). Lots of funds expressed interest in trading cbETH/stETH and cbETH/WETH which is so far not possible in DeFi.

Lending cbETH on Euler would also be potentially extremely popular as Euler allows 2-way trading. This way, lenders will receive additional APY as they do on wstETH.

While a non-zero collateral factor is increasing systemic risk, this is mitigated by the presence of a Chainlink Oracle: cbETH / USD | Chainlink

With a CF of 0.5, it would take at least a 100% increase in the oracle price to meaningfully drain the protocol. Given the Chainlink oracle includes Coinbase-derived pricing, it’s very unlikely funds won’t sell if cbETH/ETH trades above 2.

The bigger risk is bad debt due to lack of liquidity onchain to sell the collateral. Should the market grow much bigger than onchain liquidity, it could be problematic. However, given Euler’s architecture, it is possible for a fund to liquidate users without selling the collateral. This presents an easy trade to buy cbETH from users at discount.

Specification

  1. What is the link between the RFC author and the asset?

None. The proposer is the head of risk at Euler Labs and has no link to cbETH.

  1. Provide a brief description of the asset

Coinbase is supporting liquid staking for its ETH stakers with Coinbase Wrapped Staked ETH (cbETH), where the staked asset is Ether (ETH), and the staking provider and token issuer is Coinbase. With cbETH, Coinbase aims to contribute to the broader crypto ecosystem by supporting high-utility wrapped tokens and open sourcing smart contracts.

cbETH follows the cToken model, which allows it to be ERC-20 compliant and easier to integrate with DeFi more broadly. cbETH can be unwrapped for staked ETH plus accrued rewards net of Coinbase staking fees and any network-imposed penalties. Rewards and penalties affecting Coinbase staked ETH change the conversion rate between cbETH and underlying staked ETH.

  1. How is the asset primarily used?

cbETH allows users who have staked with Coinbase to sell or transfer the ownership of their staked ETH. cbETH is currently traded on Uniswap, Curve, and the Coinbase trading platform.

  1. What are the volumes and market capitalisation?

References

Project - https://www.coinbase.com/price/coinbase-wrapped-staked-eth 3

Whitepaper - https://www.coinbase.com/cbeth/whitepaper 2

Documentation - cbETH Intro | Using cbETH | Sourcing cbETH on Coinbase 1

Github / source code - Source code | Github 3

Ethereum contracts - Coinbase Wrapped Staked ETH | Etherscan 1

Chainlink oracle - cbETH / USD Chainlink Price Feed 4

Audit - Coinbase Liquid Staking Token Audit - OpenZeppelin blog 2

Twitter - https://twitter.com/coinbase 1

Blog - https://blog.coinbase.com/

Support - https://help.coinbase.com/

6 Likes

A great addition to Euler - I have full faith in Coinbase (given their proven track record with USDC working alongside circle so well). I saw this was the 12th largest DeFi asset by TVL about a week ago and was about to submit a proposal myself - thank you Seraphim. As such, I fully support this RFC and look forward discussing it further.

2 Likes

In full support of this, thanks for putting it together.

cbETH is certainly behind in the liquid staking derivitives race but with a force like Coinbase behind it, I think utilisation within the DeFi space will pick up sooner rather than later.

1 Like

Being the first to provide cbETH would open up over 1.2B of additional collateral to Euler, all of which is high quality. There are certainly a lot of traders (including myself) who use Euler for arbitraging LSDs, and this only makes Euler a more attractive place to deposit.
Thanks for submitting this proposal.

3 Likes

I think cbETH would make acceptable collateral.
My only concern is currently ~80% of their volume is on their own exchange and there is very little actual onchain liquidity relative to its total market cap.

Would this pose any problems liquidating positions?

CB ETH has ~810.1k ETH in TLV

Uniswap:
cbETH | ETH
1.79k | 2.57k | 0.05%
0.12k | 3.27k | 0.3%

Curve:
cbETH | ETH
1.36k | 1.41k

While this may be reflective that there is not yet much cbETH withdrawn from CB onto the chain and that as more migrates on chain these pools may grow to reflect the increased activity.

Additionally just want to note there are two risks that I didn’t really see covered since the above mostly focuses on the token itself which represents the staking opportunity:

  1. Slashing risk: Don’t believe validator infra is all that public. Wasn’t able to find any info on how distributed clients or hosting is.
  2. Unstaking risk: Brian Armstrong had tweeted if they were legally mandated to not build ontop of blocks which include sanctioned address activity they would shut down staking. If this were to occur prior to withdrawals being enabled, this could have a big impact on its price. I think its pretty unlikely to occur, just noting it exists.

Edit: the search feature in uniswap analytics was being a real pain so added links

2 Likes

Hi, 0scar from Exponential. Interesting proposal because the main motivation seems commercial. Long story short, I don’t think cbETH is ready for collateral tier yet.

stETH has a $7b+ market cap and there is about $179M deposited in Euler (~2.5%)

cbETH has a $1.3b market cap and let’s assume a similar share is deposited in Euler, so $32M. With a CF of 0.5, that means the potential liabilities ~$16M exceed available on-chain liquidity (~$2m on Curve and ~$3.5m on Uniswap’s 5bp pool). For reference, stETH has almost $1b in liquidity from the curve pool alone, well above the deposited amount in Euler.

Perhaps a stability pool could mitigate this risk of thin liquidity? Another option is to cap the size of the pool to ~$10M so that liabilities are only as great as available on-chain liquidity.

2 Likes

Hey! Thanks for the proposal! While Im in general supportive to the very idea, I just would like to clarify some things regarding parameters of cbETH on Euler.
So, with

* Market capitalization: $1.23B 
* Average daily volume on the 30 last days: $6,346,540

the collateral factor would be 0,5.

At the same time if we take another asset like UNI, that, according to the Coinmarketcap, has

Marketcap: $5,5 B
Average daily volume: $185 m

the collateral factor is 0,3. Moreover there is a general understanding that UNI CF should be decreased to 0.

So in this regard I just wonder, why a smaller and less liquid asset like cbETH would have riskier parameters than other stronger assets?

1 Like

This is a good point, but I think what’s important to mention here is commercial opportunity. So far, there hasn’t been much upside at all from listing LINK and UNI, which is why Shipoor suggested to decrease it. Imo non-trivial risk without any commercial opportunity is worse than manageable risk with big commercial upside.

cbETH is less liquid but presents a huge commercial opportunity and would bring big traders to Euler.

2 Likes

This would be great for cbETH and Euler. Even if on-chain liquidity is limited for now, cbETH is not a shitcoin. It’s ETH backed and listed on Coinbase. If liquidations happens, arbitrage bots would kick in right away. Adding cbETH and being able to borrow against it would help cbETH to propagate through other DeFi protocols and should improve its own on-chain liquidity.

Full support.

Hi all! I work at Coinbase as a protocol specialist. I’m really excited to see this discussion and will try to get answers to some of your questions.

To address slashing, there are a number of measures in place to mitigate the risk of slashing. Coinbase leverages Coinbase Cloud and several external enterprise-grade infrastructure operators for ETH staking. Leveraging multiple providers reduces the risk of correlated downtime or slashing events that could be caused by any one operator. Beyond node operator diversity, these operators run multiple Ethereum clients, using multiple hosting environments, in multiple geographic regions, all in the same effort to reduce the risk of a correlated slashing event.

On unstaking risk, I’ll try to add color. Taking a step back, cbETH represents ownership of ETH staked using Coinbase’s staking services. In the unlikely event that Coinbase discontinued its staking services for any reason, cbETH holders would have the same opportunity to claim their staked ETH as Coinbase’s other staking customers. Further, unwrapping functionality is available to all geographically eligible Coinbase customers who have the option to unwrap their cbETH back into staked ETH. Mechanically, this means holders of cbETH can already “offboard” into a staked ETH position even if withdrawals of staked ETH aren’t enabled by the Shanghai upgrade yet.

I hope this clears things up. Thanks!

1 Like

Is any info on this diversification public?

If CB discontinued staking services switching between liquid staked ETH via CB and non liquid staked ETH via CB doesn’t really address changes in demand for staked ETH. This presents two choices for earning yield which is no longer providing yield.

Imagine the flow would be more non liquid staked ETH converted to liquid to offload in an effort to rebalancing into yield earning staked ETH via other providers still providing staking services. Given the lack of liquidity, imagine most will do so via borrowing ETH against cbETH to avoid price impact (if such an option were available).

Mechanically holders of cbETH cannot offboard to ETH, and would be stuck in one of two assets which are not earning yield actually staking unless CB would only discontinue staking services for liquid staked ETH but not illiquid staked ETH. I find it hard to imagine where one is discontinued but not the other. Could you explain why you think users would “offboard” liquid to non liquid staked ETH in the event staking services are discontinued?

Regarding liquidity incentives

Coinbase isn’t known for incentivizing liquidity, and when they due, a la Maker DAO, it has been set up for pure holding incentives rather than liquidity providing and requires relatively high levels of centralization and minimal parties rewarded.

While USDC was an early stable provider, and benefited from things like 3pool which enabled its liquidity to be incentivized by external parties pairing w/ 3pool, cbETH is late to the game. Most of its more liquid competitors already incentivize liquidity in some way.

How does Coinbase intended to ensure adequate liquidity is provided? What avenues is CB considering if any?

Blockquote
Coinbase isn’t known for incentivizing liquidity, and when they due, a la Maker DAO, it has been set up for pure holding incentives rather than liquidity providing and requires relatively high levels of centralization and minimal parties rewarded.

To be fair, Coinbase had a Curve gauge voted in for the cbETH-ETH pari and the Curve pool is currently earning CRV incentives. Unsure if they acquired CRV directly or through bribes.

Still, really doesn’t address the concern of thin liquidity and how the protocol would be vulnerable if/when a liquidation exceeds the amount of on-chain liquidity

1 Like

cbETH has a 25% take rate thats daylight robbery lmaooo

Really in favour of this proposal. Adding cbETH as a collateral asset would definitely be beneficial to Euler given the quality of the asset and given the historical usage of similar ETH staking assets like wstETH on Euler.

We think that a CF of 0.5 is overly conservative given the quantitative and qualitative profile of cbETH and would advocate for a higher CF of 0.7.

Here are a few thoughts and stats regarding cbETH liquidation risk.

Slippage

Trade size (USD) Slippage - sell cbETH for ETH Slippage - sell stETH for ETH
1,000,000 -0.46% -0.02%
2,000,000 -0.71% -0.03%
3,000,000 -1.03% -0.03%
4,000,000 -1.59% -0.04%
5,000,000 -2.21% -0.04%
6,000,000 -2.99% -0.04%
7,000,000 -3.75% -0.04%
8,000,000 -4.60% -0.04%
9,000,000 -7.44% -0.05%
10,000,000 -9.16% -0.05%

Volatility

Historical max daily drawdown -11.24%
Daily vol 3.76%
Annualized vol 71.81%

Liquidity

Dex Current liquidity Avg daily volume
Uniswap v3 12,865,446 3,154,864
Curve 5,702,275 669,841
Total 18,567,721 3,824,705

We think the price risk of cbETH is relatively low vs ETH given the merge is behind us and staked ETH will be withdrawable in the coming months. The main divergence between stETH and cbETH is market liquidity.

Expected slippage for a $10M cbETH trade is below Euler’s maximum liquidation discount of 20%. Moreover, it seems unlikely that such a large liquidation would occur given that liquidators can conduct partial liquidations. We have also seen that some accounts have been executing Coinbase <> Univ3 arbs. This reinforces the idea that cbETH prices are likely to stay in line between DeFi and CeFi over time.

We think that the adoption of cbETH on Euler will be lower than stETH given that cbETH adoption in DeFI is still relatively low. This level of adoption will limit how much will be deposited initially in the protocol and will limit the amount of cbETH potentially up for liquidation.

We think the risk of adding cbETH with a slighlty higher CF is very acceptable. We’d like to propose amending the proposal to list cbETH with a 0.7 CF.

2 Likes

I like the 0.7 CF and doesn’t seem like there are objections. Will create a Snapshot vote in 24h.

2 Likes

Really appreciate it you chiming in here, Andrew.

Would you be so kind to let us know what happens with cbETH in the unlikely case that Coinbase goes under?

Wasn’t able to find anything about that in the whitepaper unfortunately.