[ARFC] Aave Buyback Mechanism Upgrade – Utilize Ekubo TWAMM on Ethereum Mainnet

Governance
1

Author: Moody Salem – Founder of Ekubo, Inc.

Date: 2025-04-05

Summary

This proposal recommends replacing Aave’s current AAVE buyback process (managed by the Aave Finance Committee (AFC) and service provider Tokenlogic on Ethereum mainnet) with an automated Time-Weighted Average Market Maker (TWAMM) solution via the Ekubo Protocol. Ekubo’s TWAMM extension enables continuous, passive execution of large swaps over time, allowing the DAO to convert treasury assets (e.g. USDC from the collector contract) into AAVE with minimal price impact and without active management. By executing buybacks directly on-chain through governance (via an AIP) using Ekubo TWAMM, Aave can streamline the buyback program – improving execution efficiency, simplifying operations, and removing the need for a dedicated committee or external market maker. We propose an initial pilot using a modest portion of protocol revenue (approximately $100k–$500k) to perform time-weighted AAVE buybacks via Ekubo TWAMM. This trial will allow the community to assess performance and safety before scaling up the approach as a core component of Aavenomics.

Motivation

Current Buyback Mechanism & Its Limitations: Under the recently approved “Buy and Distribute” program, Aave DAO’s excess revenue is allocated to buy AAVE on the secondary market to reward stakers and bolster tokenomics. The implementation of this program on Ethereum mainnet currently relies on a multisig Aave Finance Committee (AFC) (with members from Chaos Labs, Tokenlogic, Llama, ACI, etc.) to execute buy orders, guided by Tokenlogic’s treasury management proposals​. Each month, Tokenlogic determines which treasury assets (stablecoins, etc.) to liquidate for buybacks, and the AFC either executes on DEXes or works with market makers to purchase AAVE, typically in fixed tranches (e.g. ~$1M per week for 6 months as per Aavenomics)​.

While this system establishes a formal process, it has several drawbacks:

  • Operational Complexity: Manually splitting large buy orders over time requires coordination and timing. Even with a committee, executing trades in chunks (and possibly timing the market or working with market makers) adds overhead and risk of human error. As Paradigm’s research notes, breaking an order into pieces introduces operational complexity and increased work

The current approach demands active management (regular proposals, multisig approvals, possibly negotiating with market makers) which is resource-intensive.

  • Execution Inefficiencies: Executing $1M weekly buy orders in discrete batches can incur slippage and market impact, especially if done via market orders on DEXs or OTC deals. Large, infrequent buys may move the price or be front-run by arbitrageurs. Even with professional market makers, the protocol might not always get the optimal average price, and there could be fees or spreads involved in off-chain trades.
  • Oversight and Custody: Relying on a committee and third parties introduces custodial and oversight considerations. Funds must be approved to the committee (via on-chain allowances) and then spent by signers. This adds latency and trust assumptions, which could be streamlined with a more automated solution. Reducing the number of touchpoints (and intermediaries like market makers) would decrease operational risk and simplify governance oversight.

Why TWAMM (Ekubo) is the Solution: The TWAMM (Time-Weighted Average Market Maker) concept, first proposed by Paradigm in 2021, is designed exactly to help traders (or protocols) execute large orders smoothly over time on-chain​. In essence, a TWAMM breaks a large long-term order into infinitely many infinitesimal pieces and executes them continuously against a constant-product AMM, achieving a uniform execution over a chosen interval​. This yields an outcome similar to an ideal algorithmic TWAP (time-weighted average price) execution that a TradFi broker might achieve by slicing an order into small bits over hours​. Ekubo Protocol has implemented this TWAMM model as an extension to its AMM, allowing users to create “DCA (Dollar-Cost Average) orders” on-chain. Key benefits of leveraging Ekubo’s TWAMM for Aave’s buybacks include:

  • Superior Execution & Reduced Price Impact: Because the TWAMM executes continuously (e.g. every block or every few blocks) over a long period, the AAVE purchases are spread out, avoiding large instantaneous trades. This significantly mitigates price impact compared to lump-sum buys. DCA orders effectively give the user the time-weighted average price over the order duration, with efficient pricing and minimal fees for large swaps​. Any short-term price deviations are arbitraged by the market, ensuring the order fills at a fair market rate over time​. In short, the DAO can expect to get a better average entry price for AAVE with lower slippage using TWAMM than via weekly manual purchases.

  • Passive, Automated Execution (Simplicity): Once the TWAMM order is set up via a smart contract, it requires no active management or intervention. The contract will keep executing the trade fragment by fragment, block by block, according to the predefined rate, until the order completes. This is a “set-and-forget” strategy – no need for continuous monitoring by a committee or the treasury team. This addresses the operational complexity issue: the smart contract effectively performs the algorithmic trading strategy that humans or market makers would otherwise have to handle. The DAO simply defines parameters (amount, duration) and initiates the order; the rest is handled on-chain. This not only reduces workload but also eliminates the risk of manual error in timing or sizing trades​.

  • No Intermediaries or Custodians Needed: By using Ekubo TWAMM, the DAO can execute the buyback entirely on-chain through an AIP, without relying on a multi-sig committee to manually trade or any external market maker services. The collector contract’s funds (e.g. stablecoins earmarked for buybacks) would be directly swapped for AAVE via the TWAMM smart contract. Execution is enforced by code, so we remove two layers of intermediaries: the committee (for execution) and any third-party liquidity providers. This enhances security (fewer hops where something could go wrong or funds could be mismanaged) and aligns with DeFi’s trust-minimized ethos – governance controls the funds and the mechanism directly.

  • Efficient and Cost-Effective: Ekubo’s TWAMM implementation is built into a highly gas-optimized AMM (Ekubo V2, recently live on Ethereum)​. Unlike executing many separate Uniswap trades (which would cost gas each time and potentially incur higher fee rates), the TWAMM order is integrated into a single contract flow, amortizing gas costs and trading fees. Paradigm’s research notes that TWAMM’s design amortizes gas across virtual sub-trades and even reduces sandwich attack susceptibility by executing smoothly between blocks. Additionally, Ekubo nets out opposite order flow in its TWAMM pools (i.e. matching sellers and buyers over time) so that only the difference is actually swapped each block, which further improves price and fee efficiency​. In practical terms, Aave’s buyback via TWAMM should incur lower slippage and possibly lower total fees than either doing a giant single swap or many manual smaller swaps. The Ekubo protocol’s recent deployment on Ethereum has been audited 2 firms (Plainshift and ABDK Consulting). Early usage indicates over 20% gas cost savings compared to traditional AMMs due to its architectural optimizations​.

  • Simplicity of Governance Process: The TWAMM approach can simplify governance in the long run. Rather than monthly or quarterly treasury management proposals to authorize buyback budgets and manual execution (as is currently the case​ [src], the DAO could schedule buybacks via on-chain proposals that directly interact with the TWAMM contract. For example, an AIP could transfer a specified amount of USDC from the Collector contract to the Ekubo TWAMM contract and initiate an order to buy AAVE over the next 30 days. This single proposal replaces a month’s worth of committee activity. Fewer proposals and less micromanagement of execution details will free up governance and contributors to focus on higher-level strategy. The DAO will still retain full control – it can choose parameters of each TWAMM order and can halt or modify future orders if needed – but second-to-second execution is automated.

In summary, adopting Ekubo’s TWAMM for AAVE buybacks offers a more decentralized, efficient, and low-touch mechanism to achieve the same goals as the current buyback program. It leverages innovative DeFi tech to benefit Aave’s tokenomics, and showcases Aave DAO’s commitment to cutting-edge, trust-minimized solutions.

Specification

1. Pilot Implementation: As a first step, it is proposed to initiate a pilot buyback using Ekubo TWAMM on Ethereum mainnet. The pilot will use a small allocation of the Aave treasury’s excess stablecoin revenue (for example, $100k – $500k equivalent in USDC or USDT from the Ethereum collector contract) to purchase AAVE over a defined time period. One or two strategic assets from the collector will be selected as the source of funds – preferably high-liquidity stablecoins such as USDC (and/or USDT), which constitute a significant portion of the DAO’s revenue holdings. Concentrating on one stablecoin at first (e.g. USDC) simplifies execution and reduces variables in the trial.

2. TWAMM Order Parameters: The DAO (via governance) will execute an AIP that interacts with the Ekubo Protocol’s contracts to create a TWAMM (DCA) order. The order will specify:

  • Sell Asset: USDC (from the collector contract). The AIP will transfer the pilot amount of USDC into the Ekubo TWAMM contract/pool.
  • Buy Asset: AAVE token. We will leverage an Ekubo AAVE/USDC TWAMM-enabled pool. If such a pool does not exist yet, part of this proposal is to coordinate with Ekubo or liquidity providers to ensure an AAVE-USDC pool on Ekubo is initialized with sufficient liquidity. (Ekubo V2’s extensibility allows new pools to be created permissionlessly, so the DAO can seed a pool if needed.)
  • Sell amount: e.g. $250,000 worth of USDC
  • Start time and end time: a reasonable time span to spread the trade out – for instance, 2 weeks (14 days) or 1 month (30 days). This can be adjusted based on community preference, but should be long enough to clearly average out the price and demonstrate TWAMM’s continuous execution. For the pilot, a shorter duration like 1-2 weeks could be chosen to observe results sooner, given the smaller amount, but even that will involve hundreds of block-by-block executions.
  • Execution Rate: The above parameters implicitly set a continuous execution rate (e.g., ~$250k over 2 weeks is roughly $17.8k per day, which the TWAMM will execute pro-rata each block). No further action is required once the order is submitted.

The AIP will include the necessary calls to approve and fund the TWAMM order. Specifically, the Collector contract (which holds the USDC) will need to approve the Ekubo Orders contract to spend the specified USDC amount, and then the order creation function will be invoked. This can all be encoded in the on-chain proposal transaction payload. After execution of the proposal, the TWAMM order will start executing at the specified start time.

3. Monitoring and Completion: Once live, the TWAMM order will execute continuously for the duration of the order. Anyone can see the order’s progress via Ekubo’s interface or on-chain data – seeing how much USDC has been swapped for AAVE over time and the average execution price. After the order duration, all of the USDC will be fully converted to AAVE and the final amount of AAVE acquired will be withdrawn back to the DAO. Mechanically, this withdrawal requires an additional call to be made to “collect the proceeds” of the buyback, which can be made at any time during order execution. These calls will be added to a subsequent proposal. In the future, a separate contract can make permissionless the creation of buyback orders and withdrawal of proceeds to the DAO, as is implemented for Ekubo Protocol on Starknet [src].

4. Evaluation Metrics: During and after the pilot, we will evaluate:

  • Execution Price vs Market TWAP: Calculate the effective average price paid per AAVE through the TWAMM vs. what the prevailing market prices were. We expect them to be very close (validating that we achieved a true TWAP execution). Any deviation or unexpected cost will be analyzed.

  • Fees: Summarize the total fees paid to the pool.

  • Operational Simplicity: Qualitatively, the pilot should demonstrate the ease of execution from a governance perspective (one proposal and done). We will document the process to ensure it’s reproducible and clear for larger size.
    5. Scaling Up and Integration: If the pilot is successful, a subsequent proposal will be formulated to scale up the TWAMM buyback mechanism. This could involve:

  • Allocating a larger portion of the weekly/monthly revenue to ongoing TWAMM orders. For example, instead of $1M weekly via committee, the DAO could authorize, say, a continuous rolling TWAMM order of $4M per month (roughly equivalent to $1M/week, but managed as a perpetual cycle).

  • Including many assets: e.g. use USDT, ETH from the treasury

  • Automation: a separate contract can be created and configured to trigger a fixed amount of buybacks each week

  • Integrating this into Aave’s broader treasury management: The DAO might decide to make TWAMM buybacks a standing strategy. Tokenlogic (or another entity) could still be involved in advising on how much to allocate based on treasury health, but execution would no longer require their active role. In fact, Tokenlogic’s mandate could shift to just recommending parameters for TWAMM (frequency, amount) each quarter, rather than handling the trades.

Crucially, any scaled program would come back to governance for approval. By demonstrating the concept with a pilot, we minimize risk before committing larger funds.

Next Steps

  1. Invite community & service providers to provide feedback on this proposal with the goal of reaching consensus.
  2. If consensus is reached on this ARFC, escalate this proposal to the Snapshot stage.
  3. If the Snapshot outcome is YAE, implement the proposal via AIP.
  4. Review the outcome of the buyback order

Copyright

Copyright and related rights waived via CC0.

5 Likes
2

Hi Moody,

Thank you for your post. I wanted to bring attention to the forum that a TWAP solution has already been implemented as part of the Finance Steward initiative. It’s in the final stages of the review process and should be operational by next month’s buybacks. It uses COW Swap under the hood as the existing swaps currently do.

2 Likes
3

There are some major differences between these approaches:

  • Per the docs, AaveSwapper is a permissioned smart contract, where the guardian is an address to be chosen by the DAO to more easily cancel swaps without relying on governance. In other words, it requires guardrails due to the complexity of the Cow auction and the steward contract.
  • The contract relies on Chainlink oracle for protection against slippage (i.e. bad order execution)
  • Cow TWAP orders execute in batches with fixed intervals, whereas Ekubo TWAP orders execute continuously up to once per block. Therefore, Ekubo TWAMM is more likely to provide the time weighted average price.
  • Opposite Ekubo TWAMM orders can be placed by anyone, whereas Cow solvers would have first look at the buybacks flow if they all go through Cowswap. The Cow auction is also permissioned.
  • We have been using TWAP orders to do automated buybacks of EKUBO tokens for the Ekubo DAO since August 2024 (see the RevenueBuybacks contract and our dashboard)
  • Our TWAP extension and Order creation contracts are audited by ABDK audits, and the DAO can directly create the orders by calling these contracts

IMO the DAO should trial the implementation of buybacks via Ekubo TWAMM regardless of the current status of the Cowswap integration.

4

A few points to clarify:

  • AaveSwapper is permissioned means that nobody can just decide to perform swaps on behalf of the DAO. This is the same flow that is being suggested by you where the DAO would via governance call the Ekubo protocol. The guardrails are not due to the complexity of COW Swap, but rather because you wouldn’t want anyone to just randomly trigger swaps on behalf of the DAO.
  • The TWAP orders do not rely on Chainlink oracles as can be read in the documentation. The slippage protection is for market orders, and has been in use by the DAO since 2022.

I understand why you would want the DAO to trial Ekubo and I cannot make that call for the DAO, I am just pointing out that a solution is already in place to automate the buybacks, the development work has already been done and a sample trial transaction performed.

As the original post does not specify what party would be responsible for reviewing Ekubo as a protocol and implementing the AIP for the trial phase (and then a subsequent FinanceSteward implementation), I thought it worth mentioning that all that lift has already taken place.

5

For the trial it would be initiated by the DAO. If the trial goes well, it can be automated in the same way it is done with Ekubo’s Revenue Buybacks, in which anyone can trigger new orders using protocol revenue.

Can you clarify:

  • For what reason do you need the guardian role that can cancel a swap?
  • Who is intended to be able to place market orders?

Ekubo, Inc. will implement the AIP for the trial, and if it goes well, create any additional smart contracts for continued buybacks and fund the audits. Note we already plan to create and have audited a revenue buybacks contract for the Ethereum deployment of Ekubo Protocol.

6

As I stated previously, it is not my call to make, but rather to provide solutions to the DAO as needed and as efficiently as possible and I am pointing out that a solution currently exists. I have no preference on what underlying protocol is used myself.

Again, I understand why you would want to bring the attention of Aave into your protocol, and to handle swaps through your pool. However, I don’t think the reasons provided above bring much value or are honest improvements on the existing solutions.

You have multiple times brought attention to how having the swap functions be permissioned is a hindrance to the DAO. Why would the DAO want anyone to create or cancel swaps at any time? The finance committee exists to perform these tasks on behalf of the DAO in order to minimize governance, but otherwise, governance maintains full control. Governance sets budgets, etc.

I have not mentioned COW Swap until now as a comparison because to me the important thing here is how this all works for the DAO, not what solution is used under the hood but since you have mentioned it a few times, I wanted to ask:

What benefit exists in getting AAVE from an Ekubo USDC/AAVE pool compared to COW Swap getting liquidity from different places, both on-chain and off-chain? Are we sure the DAO will get the best price using just your pool?

You mentioned your contracts are audited, so are COW Swaps, and also the steward contracts.

COW TWAP orders can also be configured to work across multiple hours/minutes if needed, across any number of days. They can be monitored as well.

It could very well that the DAO ends up utilizing Ekubo TWAP orders, I just need to see what value is added compared to the costs that arise from this proposal such as:

  • Liquidity concentration risk (Ekubo AAVE/USDC pool) vs. on-chain/off-chain liquidity from COW Swap
  • Smart contract risk analysis to be performed by a DAO team on Ekubo vs. 3 year history with COW Swap
  • Development time of FinanceSteward to incorporate Ekubo
1 Like
7

A la why should we build permissionless products in DeFi, when we can just trust {individuals/committees/institutions}? The best answer is efficiency. It’s much easier to get a loan from Aave than the bank.

My argument here is that it is much more efficient to execute buybacks via TWAMM, both because there is no need for a committee to execute the orders (whom I assume receive compensation for this task among their other responsibilities), and because the design of TWAMM enables users to get the time-weighted average price with very low overhead and execution cost. If you want to learn more about TWAMM, you can read about it here: TWAMM - Paradigm

The trial will demonstrate that the DAO receives the time weighted average price over the period, and it will be trivial to measure the performance objectively as well as compare with cost-adjusted price that the buybacks via Cow receive.

The goal, of course, should be getting as close to the TWAP as possible after accounting for all costs to the DAO. For example, if you give the responsibility solely to an individual, they could time their trades and receive a better-than-expected price, but they could also mis-time their trades and receive a worse-than-expected price.

If tens or hundreds of millions are meant to be directed to buybacks, a difference of 1% more than pays for the cost and risk to the DAO of doing a trial with Ekubo’s TWAMM.

Since you’ve mentioned it twice now, yes it is obviously in my interest, but it doesn’t need repeating–I already shared I’m the founder of Ekubo Inc. I think it would make for better discussion not to belabor that point and focus on the benefits and drawbacks of the solution I’m proposing.

1 Like
8

I think there might be a misunderstanding in what we both envision the solution looks like and the current capabilities are.

The current existing solution can:

Set a swap for X amount, for Y parts, and can get more granular than that with time of day and how long to spread the swap for.

For example, I want to swap 1 Million USDC for AAVE, over a month. I want to do this every Monday, and I want the swaps to occur over 5 hours. This can currently be done.
Or I want to swap that 1 Million every 3 days, so 10 swaps of 100,000 and I want the swaps to only take place from 12PM-2PM, etc.

The only thing that needs to be triggered by a committee would be to set the initial parameters. You can even set the whole year if you really wanted to, that is why I don’t think the new solution brings extra value. Using Ekubo, the same thing needs to happen, the DAO/committee needs to approve the funds to be spent by Ekubo, set the initial parameters, etc.

One other thing I want to mention is that unlike Ekubo, the existing solution provides the acquired funds to the DAO after each portion of the TWAP swap, there’s no need to wait for it to finalize at the end of the set period.

I respect all your contributions to the ecosystem and have referenced a lot of your GitHub contributions over the years. I hope this last post clarifies why I don’t see why yet another solution that performs the same functionality and would require extra DAO resources would be beneficial at this time unless it significantly yields more revenue for the DAO.

Lastly, I want to reiterate that I by no means speak for the DAO and would not even dare pretend to be the one person who will make this call. I only brought to attention an existing solution that you might not have been aware of during your initial post, and then clarified points of misunderstanding and added all the relevant information for the community members to be fully informed.

1 Like
9

Can you set it up to do a $5 swap every block for an entire month? Because that’s how TWAMM works if you want to do a $1m buyback, and that’s why it gets you closer to the time weighted average price.

Not sure where you got this idea. You can withdraw proceeds in the same block that the order executes.

There is minimal risk in a trial since we will do all the work. The real risk is in ignoring it because you think the current solution is perfectly sufficient. (For clarification, neither solution produces revenue, so the correct measurement is once again how close to the TWAP the DAO gets)

3 Likes
10

I think there is still a misunderstanding.
What would be best for the DAO is to see each solution one by one in a table with an example of 1m of buybacks per week.
And then showing parameter like gas costs, efficiency, average price, manual tasks, etc.
That way the DAO can decide what solution would be the best. Everything else is simply not simple explained.

Try to make it simple, not harder to understand. That’s how you get the people to understand and vote accordingly.

Edit:
What the DAO obviously also has to consider is the time/money/knowledge SPs have already put into the existing solution.

And @sendmoodz checking the volume on Ekubo im unsure if the DAO can perform its tasks there.
Even no route for USDC-AAVE

{5CEBD946-23CE-4D5B-AD09-B3975F23E80E}
{5CEBD946-23CE-4D5B-AD09-B3975F23E80E}562×595 15.2 KB

And overall volume for stablecoins is pretty low.

2 Likes
11

When you’re talking about numbers like $1m per week, gas costs are not what makes the process more efficient. The distance between execution price and time-weighted average price is what needs to be measured by the trial and compared amongst the solutions. Per the OP, there are no manual tasks required of any individual or committee once automated by the DAO since the process of creating the orders and withdrawing proceeds can be made permissionless with Ekubo TWAMM.

Also, I must reiterate that the risk to the DAO in using this is only the amount of USDC it commits to buybacks.

There is 0 AAVE in Ekubo currently. However, the current liquidity and volume is not relevant for TWAMM orders anyway.

When an order is created, it selects a specific pool that uses the TWAMM extension, which may not even exist yet. The order is both split into per-block parts and netted against opposite orders, meaning the amount that gets swapped against the pool can be as low as $0. Absent any opposite orders, that means a $1m swap over 1 month only swaps $5 per block. To keep slippage per swap to 0.1%, you only need ~$10k in liquidity in that pool. If there are opposite orders, the amount of volume that is matched does not swap against the pool, further reducing execution costs. If you’d like to understand more, read this: TWAMM - Paradigm

Note aggregators are the ones that keep the pool’s price close to the average price, for which Ekubo’s very recently launched TWAMM is already integrated with Matcha, and integrations are coming for at least Cowswap (via Tycho) and Paraswap in the next few weeks.

This mechanism has already been used to settle millions of dollars of volume on Ekubo Protocol. We even did our initial token distribution via TWAMM orders.

3 Likes
12

Hello,

The AFC is mandated to operate these swaps on Behalf of the DAO.

@TokenLogic built the Aave swap steward infrastructure to swap assets efficiently as part of their service provider Scope.

Unless this proposal removes the AFC mandate to operate the buybacks, any vendor seeking integration can directly contact the AFC co-loads @TokenLogic & @ACI.

If this proposal doesn’t seek to remove AFC mandate, we do not suggest escalating this proposal to the next stage.

5 Likes
13

Thanks for the feedback Marc. Would it be sufficient for the proposal to amend the mandate to specify that additional buybacks can also happen directly via calls to permissionless and immutable smart contracts?

14

This proposal fits in nicely as the AFC is also tasked with ensuring these processes become automated.

Would be great to hear from @TokenLogic and @ACI.

1 Like
15

We are more than happy to build any solution that advances and improves Aave DAO’s tooling.

The most recently proposed upgrade to the Aave Swapper is still to enter production. The buybacks could all be done via AIP each quarter with Admin Role oversight, however they are currently being performed manually to ensure operations continue unaffected.

Once the Aave Swapper has been upgraded, the DAO can access the full capabilities of Cowswap. The question then becomes is Ekubo better that Cowswap, if so great. If not, Paraswap is the most likely inclusion. On Paraswap, we can push a PR to tap into new liquidity sources or factories in a similar way to how we can also push upgrades to the Glue Solver serving Cowswap. The ability to push PRs enables TokenLogic to ensure the best pricing terms are sustained, to the benefit of Aave DAO.

However, we do note, anyone can propose an upgrade to the Aave Swapper tooling. The Aave Swapper is Open Source and has no dependencies on any Service Providers offering/existence within the Aave community.

3 Likes
16

“is Ekubo better that Cowswap” is not fully defined, nor is it in my opinion the correct question: buybacks via TWAMM can be made completely permissionless meaning there is no involvement of third parties such as @TokenLogic. The end goal would be to replace the entire solution and remove the intermediaries, not just Cowswap’s implementation of TWAP orders.

Is there a mandate that says the AFC should seek to receive the maximum amount of Aave when using its revenue for buybacks? For example, should the AFC be attempting to time the market? If not, what is meant by “enables TokenLogic to ensure the best pricing terms are sustained”? IMO a buybacks mechanism should aim to receive the TWAP, as opposed to receiving the “best” price, since that unnecessarily creates a perverse incentive for the DAO to minimize AAVE price.

Note also that thanks to our Tycho implementation, Cowswap solvers would have access to the buyback flow, as well as aggregators such as Paraswap (here is our PR implementing TWAMM pools for Paraswap) and Matcha (already integrated, e.g. ETH->USDC swap). In addition, the top solvers are already working on integrating Ekubo liquidity, including TWAMM pools.

I don’t believe the Aave swapper needs to be upgraded to support buybacks via Ekubo. No additional contracts are needed for the trial. The Ekubo Orders contract can be called directly via AIP to create an order owned by the DAO, and after execution a second call can be made (again via AIP) to collect the proceeds of the buyback. This limits short term development work and smart contract risk.

To minimize risk, I can guarantee on behalf of Ekubo, Inc. a realized execution price within 1% of the average price for up to $250k of buybacks for the trial. Specifically, if the DAO does not receive within 1% of the time-weighted average price over the period, based on Coingecko data, the company Ekubo, Inc. will reimburse the DAO for the difference up to the total amount minus 1% (= $247.5k).