[ARFC] Ethereum v2 Collector Contract Consolidation

title: [ARFC] Ethereum v2 Collector Contract Consolidation
author: @llamaxyz - @MatthewGraham & @Dydymoon
created: 2022-12-02


@llamaxyz proposes consolidating the Aave v2 Collector Contract holdings by swapping a portion of the long tail assets to USDC and redeeming assets from the Aave AMM deployment.


The Aave v2 Collector Contract has accumulated a holding in several assets over time. Assets from the v1, v2, RWA, ARC and AMM depoloyments all accumulate in the Ethereum v2 Collector Contract. This proposal is the first of several intended to consolidate the various Collector Contracts and Treasury holdings.

This proposal seeks to consolidate 17 assets in the v2 Ethereum Collector Contract to USDC and redeem ammTokens for the underlying assets.


The DAO is currently accumulating over 60 assets in the v2 Collector Contract from various Aave deployments. Many of these holdings are of small economic value, are undercollateralized assets, or exhibit volatile pricing behaviour. Upon consolidating the DAOs assets into USDC, which most of the DAO’s costs are denominated in, Llama intends to present the community with suggested recommendations for deploying a portion of the larger holdings.

The AMM Liquidity Pool, v2, attracts very little capital and usage relative to other deployments. A number of the assets are frozen and, with v3 being deployed on Ethereum soon, it makes sense for the DAO to relocate this capital away from from the v2 AMM deployment. This proposal enables the following ammTokens, aAmmDAI, aAmmUSDC, aAmmUSDT, aAmmWBTC, and aAmmWETH to be redeemed periodically by calling a contract. Any actor will be able to interact with the contract via Etherscan. Llama, or anyone else, can redeem these ammTokens periodically with the underlying asset to be held within the v2 Collector Contract.


The following assets held within the Ethereum v2 Collector Contract are to be converted to USDC via swap contract.

Users will be able to deposit USDC and receive the tokens mentioned in the table above, plus a small premium. The premium is expected to reimburse the cost incurred from swapping the applicable token to USDC along with associated gas costs. The methodology enables actors to support Aave in swapping one asset for another by creating an arbitrage opportunity in a decentralized manner.

The following assets are to be redeemed from the AMM deployment.


The below outlines at a high level the flow of events this proposal will achieve:

  1. Redeem aTokens and ammTokens for underlying asset
  2. Deploy a Swap Contract whereby actors can deposit USDC and receive a selected token stated in the above table at the Chainlink Oracle price + premium
  3. v2 Collector Contract receives USDC


Copyright and related rights waived via CC0.


The Aave-Chan initiative is supportive of this proposal.

a small remark, to our current understanding, the value of renFIL is heading toward zero in the short term, while this represent a very small amount of the total, we’re supportive of getting this AIP getting enforced before this deadline.

The premiums are probably more generous than needed but will surely help to resolve this operation quickly by attracting MM willing to do a quick arb.

Thanks you llama for this proposal.


Supportive of this proposal and the consolidation of v2 collector asset and redemption of AMM assets.

Thank you to @MatthewGraham and the Llama team for incorporating community feedback and for publishing a clear ARFC for the forum to see.

We thank you for the initiative.


We are in full support of this proposal, thanks @Llamaxyz for putting this forward!


I guess the MM are indeed attracted! Everyone wins.

Hi @MarcZeller,

We increased the premiums on this proposal relative to the BAL acquisition due to the desire to close this out quickly and also to accomodate a broader range of market conditions.

Given the holding sizes, it would be easier for actors to perform the swaps manually rather than create a bot. The smaller trades are quiet gas sensitive. Although block space is cheap at the moment, we did want some buffer to ensure things were swapped out fairly quickly.

We did not consider the time it would take someone to sit there and do all the transactions. We made the assumption that if no one actioned the swaps, then we would likely end up doing this ourselves to ensure it gets done. Ideally, we hope there is enough incentive in the premium to ensure actors invest the time to complete the swaps. Fine tuning the premium for cost minimisation would likely lead to a slow close out rate for a small cost savings.

If any recommendations to amend the premium were to emerge, then we are happy to tweak the post and incorporate community feedback. If not, we would like to proceed to Snapshot. Our current plan is to post a Snapshot on the 6th December with voting commencing on the 7th December, which accommodates the 5 days discussion time requirement for governance.

1 Like

Would be great if the ppl from 1inch (@roundelephant), paraswap and/or 0x would chime in.

From my limited understanding of how dex aggregators/routers work, I guess it shouldn’t be to hard to define a common interface that could be used in future PSMs to allow simple integration with them. Having up to 500bps premium seems quite excessive to me and even 0bps could be attractive as there’s 0 slippage for all supported amounts (ofc some premium would be fine).

Thanks for the initiative.

How did you come up with those prices? Have you think about using Chainlink’s prices (+ premium)?

PS: It seems the convention in the community is to use ARC instead of ARFC, fyi.

as a side-note while the “official Aave Request for Comment” terminology is ARC, the emergence of the ARC market might create some confusion,

using ARFC is an alternative that makes sense to me, and I have no opposition to its usage.

1 Like

Hi @miguelmtz, @sakulstra, @MarcZeller and @Callen_Wintermute,

Thank you for commenting on the publication. It is great to see the conversation evolving.

We are very open to changing the premiums on this proposal and would welcome suggestions towards lowering them.

Our general logic follows that an actor has to perform a couple approval transactions, deposit USDC, receive the asset and then swap the asset to USDC completing the road trip. Using an Oracle to anchor the price means, it becomes most economical to perform the swap when spot price exceeds the oracle price. The Chainlink oracle price is influenced by CEX liquidity whereas most actors will likely use DEX liquidity. When DEX pricing exceeds the oracle price, this is when most DeFi actors will perform the arbitrage transactions.

Using the numbers in the table, the premium is equivalent to a total of $9,959.42. If people want, we can optimize and save a few thousand. To do this, the focus should be on AMPL, RAI and UST as they make up $6,445.88, $1,910.04 and $928.35 respectively. The next holding is FRAX at $217.36, which offers diminishing returns on our time by optimizing further.

Regarding ARFC, we were asked to trial this and see what the feedback was from the community. Thank you for asking though, as I am sure many people will be wondering the same thing.


even as a famous penny-pitcher, imho, delaying this proposal any further for a “potential” optimization that is lower than a single day of protocol revenue is unwise.

For more significant amounts, optimization is welcome. In the current situation, the ACI is supportive of Llama’s proposal and expressing support for escalation to the next governance steps.

Agreed, doesn’t make much difference in this case.

That said my comment was meant in a more holistic sense.

It’s not the first proposal and won’t be the last proposal targeting to swap treasury assets:

  • there was/(still is) the BAL acquisitions
  • there is this proposal
  • it seems to me that soon there might be a CRV acquisition
  • when v2 is sunset there probably will be some USDT.e → USDt swap on ava needed

With an entity more involved in treasury management these things will get much more frequent.

In the past these x bps discount points seem somewhat arbitrary to me.
The only reason they are needed is that they aren’t integrated into dexes.
The current BAL-USD bonding curve essentially outperforms every dex there is as it has 0.5% positive slippage even for big amounts.

I see that integrating a bunch of “low cap” PSMs with custom integration code needed is not so attractive for a dex. That’s why I think down the road coordinating with such an entity for a more standardized approach would be a good idea.
Might also make sense to use some pool primitives that already exist instead of deploying custom solutions.

Anyhow, you’re right - doesn’t really matter in this case.


That’s a good point about Chainlink’s prices. It would be a matter of identifying the DEX liquidity and see the weight of the DEX price over CEX price… which makes it a bit cumbersome.

The proposed approach looks good to me, with a slightly reduction of premiums if possible.


Hi Everyone,

Thank you for all the feedback and conversation on this publication. We have amended the premiums and included an additional column that shows the dollar value of the premiums. This table has now been included in the original ARFC post.

Premium USD value $9,959.42 reduced to $7,790.94. A 21.8% reduction.

A Snaphot vote has been created:


Thank you in advance to all those who participate in the vote.