Hello all,
I am replying on a personal capacity because many technical aspects discussed here have been misrepresented or are outright incorrect so I believe it’s important to make sure everyone is on the same page, in order to have a healthy discussion.
First i want to address ACI allegations that the new swap adapters are using balancer/morpho flashloans, therefore “bringing volume to the competition”. This is factually incorrect and a pretty surprising statement since it’s quite easy to verify the reality onchain. The CowSwap adapters have been deployed with an enforced Aave flashloan factory, therefore from both a flashloan volume and fee perspective, the behavior of the adapters is unchanged. Again pretty easy to verify onchain, as the flashloan factory is available at this address https://etherscan.io/address/0xdeCC46a4b09162F5369c5C80383AAa9159bCf192#code (same address on all the networks). Other users in the thread have already verified how the flashloan fee is correctly collected upon execution. It would be great if for once ACI could validate their claims onchain instead of creating pointless sensationalism.
Second, multiple posts here refer to a “privatized 10M/year revenue flow”. The details of the revenue coming from the paraswap (now Velora) are available in this (great) dashboard by TokenLogic https://aave.tokenlogic.xyz/revenue.
This is the revenue for 2024:
Total: 613,258 USD
Revenue for 2025:
Total: 1,160,615 USD
(Cowswap adapters went live on nov 17th so the last quarter is slightly affected)
So the revenue is an order of magnitude smaller than what has been advertised here.
Also important in my opinion is to discuss how this revenue was generated in the first place: There was never an explicit fee stream to the DAO embedded into the adapters (meaning, there was no onchain, direct fee accrual mechanism for the DAO) neither a direct agreement for fee sharing that was agreed upon with the DAO. The revenue was coming purely from the Paraswap (now Velora) referral program, which shares the surplus of positive slippage trades with referrers (here the details https://docs.velora.xyz/integrating-velora/integrating-velora-overview/fee-sharing#partners-fees-overview). There was never an ad hoc agreement with the DAO, and the result of the integration and consequently the revenue could change tomorrow or anytime Velora decides to modify the terms of the referral program.
I also want to answer this:
SVR liquidations have been a large implementation success and have yielded significant revenue for the DAO. The Aave V3 liquidations are extremely efficient, and the protocol sustained large volatility events (most recently the 10/10 event) without noticeable bad debt accrual. The promoted V4 liquidation engine will be hurtful to this revenue to the tune of 10+ millions of dollars per year. Aave Labs used Aave DAO’s brand assets and social accounts to promote their new V4 liquidation engine and issued communications suggesting that the most Lindy and safest lending protocol liquidation engine was inferior.
Is it safe to expect another large Aave DAO revenue stream will be lost in favor of V4?
I dont even understand whats the point of this question in this thread as clearly has nothing to do with the topic, and i would like to highlight that ACI would already know the answer if they had cared enough to attend the calls that we had for SPs for longer than 10 minutes, or at least take a glance at the docs or the code. Anyway I will once again help, hopefully this trend of having to perpetually educate ACI members will stop at some point. First of all calling “extremely efficient” the V3 liquidations is a pretty gross misrepresentation - you can literally find onchain thousands of liquidations, especially smaller ones, where liquidators could have earned much less and still guaranteed protocol solvency. Overliquidating users is never a good thing, and there are protocols out there outcompeting Aave on this specific aspect already. The goal of the new V4 liquidation engine is to give risk manager options. It can be configured more aggressively (so to be more competitive towards users, but increasing the risk for the protocol) or more conservatively (so to have even more safety) than V3, while still being more gas efficient. Ultimately whether or not revenue will be impacted is only dependent on the configuration risk managers will propose, but this will leave the door open for them to tune it whenever it’s needed to match protocol and DAO needs. This is not something Aave Labs will decide in any shape or form.
I hope this was helpful, the goal is just to present truthful information so that we can have a healthy discussion without fake sensationalism.
EDIT because i just happened to reread this sentence and i find it pretty funny
Aave Labs used Aave DAO’s brand assets and social accounts to promote their new V4 liquidation engine and issued communications suggesting that the most Lindy and safest lending protocol liquidation engine was inferior.
Glissing on the way the sentence is framed that i find slightly amusing, from a purely technical perspective, the Aave V3 liquidation engine is, in fact, inferior, that’s pretty much an objective statement. That doesnt mean that is bad, by any means, but i feel silly even saying this, turns out that new protocol iterations are by design meant to provide better technology than what previously available, and that’s exactly what V4 is for. The DAO in the end paid for this exact reason and it would be disappointing if that wasnt the case. If you need more details to better understand V4 you can reach out as other SPs have been doing, would be nice to have everyone up to speed.