[ARFC] Add YFI to Ethereum v3

Title: [ARFC] Add YFI to Ethereum v3
Author: @Llamaxyz - DeFi_Consulting (@matthewgraham), ACI (@MarcZeller) and 3SE
Dated: 2023-02-18


This proposal presents Aave the opportunity to onboard YFI to the Ethereum v3 Liquidity Pool.


Yearn is a decentralised suite of products helping individuals, DAOs, and other protocols earn yield on their digital assets. YFI was originally listed on Aave v1 and included on Aave v2 when it was launched in late 2020.

Any YFI nominated revenue could be converted to veYFI, if Aave chooses, and be used to participate in Yearn Finance’s tokenomics. The staked aToken being developed by @AaveLabs enables Yearn’s vaults to be integrated as collateral and the veYFI holding could be used to boost rewards if Aave elects to do so.

This ARFC is a joint effort with ACI and 3SE to allow a selection of v2 listed assets to be onboarded on V3 via the governance process. Subject to a favourable Snapshot vote, @Llamaxyz will submit an AIP for voting.


Yearn Finance is the market leader for yield aggregation with $471m in TVL across four deployments and is ranked 22nd on the Defi Llama TVL dashboard. There is a little over $6m of deposits in the frozen YFI reserve on Aave v2 Ethereum. YFI is not currently listed on either the USDC or ETH Compound v3 liquidity market.

With the deployment of the staked aToken, there is the potential to integrate Yearn’s vaults as collateral. This is not likely to be realised in the short term. However, it is possible and something for the Aave community to consider at a later date.

The risk parameters provided within are the same as those provided by Gauntlet within the initial forum post discussing which assets to onboard to Ethereum v3.

The interest rate provided in the proposal mirrors the v2 deployment with exception of the parameters specific to stable interest rates. The stable rates proposed are the somewhat generic listing rates as stable borrowing is disabled. However, if stable borrowing was enabled, the suggested parameters are suitable and don’t create any undesirable dynamics between stable and variable rates.

We note that this proposal is to be reviewed by Gauntlet, with feedback incorporated, prior to progressing to Snapshot.


Ticker: YFI

Contract Address: 0x0bc529c00c6401aef6d220be8c6ea1667f6ad93e

Risk Parameter Value
Isolation Mode YES
Enable Borrow YES
Enable Collateral YES
Loan To Value 50.00%
Liquidation Threshold 65.00%
Liquidation Bonus 10.77%
Reserve Factor 20.00%
Liquidation Protocol Fee 10.00%
Borrow Cap 10k
Supply Cap 200
Debt Ceiling 16,187k
Variable Base 0.00%
Variable Slope1 7.00%
Variable Slope2 300.00%
Uoptimal 45.00%
Stable Borrowing Disabled
Stable Slope1 7.00%
Stable Slope2 300.00%
Base Stable Rate Offset 3.00%
Stable Rate Excess Offset 5.00%
Optimal Stable To Total Debt Ratio 20.00%

Note: Stable borrowing is disabled, however parameters are provided in case stable borrowing is enabled at a later date without amending any parameters.


Copyright and related rights waived via CC0.


Hello, and thank you, @Llamaxyz, for submitting these proposals to the forum.

Since the proposals are presented in separate threads to allow governance granularity, the ACI’s response will be variations of this current post on each thread.

YFI wasn’t a particularly popular collateral in V2, there’s little to no use case to borrow YFI, so expected revenue from usage of YFI as collateral and collector collection of YFI tokens are not expected to be significant (Aave V2 collector contract currently holds ~0.5 YFI & aYFI).

The ACI is not supportive of this proposal, even in isolation mode and with strict caps.


Chaos Labs is looking into the initial parameters listed on this post and will collaborate with Gauntlet to provide updated parameter recommendations per our joint New Asset Listings Framework.


Hey Folks thank you for putting this one up. Yearn has a simple Aave V3 Lender strategy that has already been through security review and approved. It will be able to fill the 200 YFI supply cap in this proposal with yvYFI: 0xdb25cA703181E7484a155DD612b06f57E12Be5F0

Yearn is planning to use Aave V3 with more complex strategies in the future.

Let me know if you have any questions please!

1 Like

A bit confused by this proposal -

Notice Llama cites the ACI as an author yet Marc seems to express opposing opinions to adding YFI.

Was this done in error or purposeful?

In general, ACI and Llama support each asset being presented individually to the community for inclusion on Aave v3. This allows the community to discuss each specific proposal in isolation. We are proposing most assets for migration to v3 and both teams will share views/opinions/vote as individuals on each listing. In a nutshell, we are aligned on process and vote as individuals.

The Aave-Chan Initiative is dedicated to supporting the governance process and is happy to provide feedback and write ARC, ARFC & AIPs when needed.

However as @MatthewGraham pointed out, the ACI will vote independently and according to our delegate platform.

1 Like

This a general comment that maybe somebody else from the community apart from me is wondering about.
There are multiple ARFC/proposals from @Llamaxyz co-authored with multiple other individuals and teams, like in this case DeFi_Consulting, ACI (it seems only partially given @MarcZeller previous comment), and 3SE. This is pretty confusing because Llama is an engaged entity with the DAO so it is important to really identify what comes from Llama.
At least for me, it is difficult to understand why 4 different entities are involved in a single ARFC.

Additionally, as an AAVE holder, for all ARFC/proposals with assets of other communities involved, I would like to see always a full disclosure of any relation of any of the proposers with that community.

1 Like

The ACI is supportive of a culture of transparency in the aave DAO, and will present a governance framework for the ARC/ARFC process integrating mandatory disclosures and a rework of the process shortly

1 Like

Hi @eboado,

The @Llamaxyz account typically lists all the contributors who have participated in generating the post. This provides visibility into the team and also gives context when certain individuals comment on the post for readers who are less familiar with Llama. If the direction is to run a business account, manage it as such, then all service providers should align to this practice. We can remove all individual names as such. I do think this will lead to less public facing discussion.

Myself as an individual, does not contribute to Aave and does not contribute to Llama. I do however contribute through a series of offshore legal entities and ultimately via Llama to Aave, but any rewards do not flow the reverse way, these are separate liability and revenue flows. The company (Defi_Consulting as it is referred to here) is also owned by a legal structure consisting of several entities which makes tracing anything back anything to myself very challenging. This structure was created in response to legal advice received and it offers financial protection for my family and enables me to contribute to Aave/Llama/others indirectly.

Once the trading name, frontend and more business details are ready to be shared publicly, a post will be created. The crux of the issue is why take the legal risk, or potential legal risk, when it can be minimized by using a legal structure. In time, all communications will come from the legal entity. This same logic applies to the relationship with Llama, myself as an individual is not engaged by Llama or any DeFi community for that matter. I can only assume other teams/individuals have similar setups and have not shared such information publicly.

Over time, you will see less of me on this forum and more of the company over time. I also need to move my guardian wallet from myself to one of the various companies as well. Being an unpaid guardian with no upside and plenty of legal liability exposure was not smart, especially when I am one of the more active signers. This will just be another admin task to clean up at some point.

1 Like

Appreciate the work and clarity on this topic. It is interesting that you do not contribute to Llama yet you are seemingly the most active community member in forum discussions on Llama’s behalf. With this, two questions:

a) can you expand on the nature of these “offshore legal entities” and specifically how your work for them ultimately contributes to Llama and therefore to Aave? Is there a separate contract between Llama and these entities for work that effectively subcontract work they were engaged by the DAO to conduct?
(Aside: we have no issue with the structure, but as you are the primary point of contact between the DAO and Llama as a well compensated contributor it is important. Same would be true if any of the key contributors at engaged entities such as BGD, Gauntlet, Chaos, Certora, etc. had a obfuscated relationship with the DAO).

b) are you individually directly or indirectly via any of the above-mentioned “series of offshore legal entities” compensated by anyone other than Llama/Aave specifically for your work on Aave proposals, ARCs, or AIPs?

Above all, we are supportive of a clear disclosures and transparency framework for all relevant engaged entities and major delegates.

Chaos Labs recommends not listing YFI on AAV3 V3 Ethereum at this time. Although our general approach is to promote listing more volatile assets to increase the offering of AAVE protocol, the risk-reward ratio here is not balanced, with no strong signal of significant reward.

Liquidity, Market Cap, and Manipulation Risk

When analyzing market cap and trading volumes of assets for listing, we are looking at the past 180 days, especially in light of the recent market turbulence. The average market cap of YFI over the past 180 days was $234M, and the average daily trading volume was $34M. We find the market cap significantly low, especially when examining the lows to which it fell over the past few months.

1 Like

Gauntlet Analysis

Our original recommendation here excludes YFI from the initial launch of Aave V3 ETH as a result of market metrics and usage.

However, if the community has a strong strategic preference towards listing YFI, below are Gauntlet’s parameter recommendations.

Syntax Description
Liquidation Bonus 8.5%
Borrow Cap 500
Supply Cap 500
Debt Ceiling 6,700k

Given that YFI will be in isolation mode and have a debt ceiling, our analysis shows that a liquidation bonus decrease from the proposed parameters would still properly incentivize liquidators. The liquidation bonus decrease from the original recs will keep the effective liquidation incentive closer to the Aave V2 value. Based on our borrow and supply cap methodology, the borrow cap (assuming the initial post was a typo) and supply cap should both be 500. For the initial parameterization, a more conservative setting of 6.7M, which can be calculated from the supply cap, liquidation threshold, and current price, would be prudent. We can revisit this as the market matures.

1 Like