CIP: <to be assigned when moved to phase 2>
title: Solver rewards budget renewal and update of CoW DAO bonding pool operations
author: Haris Angelidakis, Andrea Canidio, Felix Henneke
status: Draft
created: 2024-07-10
Simple Summary
The previous solver rewards budget (for the solver competition on mainnet) was defined and approved in CIP-36, about 5 months ago, committing 8M COW tokens for the solver competition, projected to last until the payout of July 30, 2024. The estimate proved to be slightly optimistic, with the allocated tokens projected to be depleted after the payout of July 16, 2024, instead of the original estimate of July 30, 2024.
This CIP proposes adjusting the rewards for the solver competition on mainnet, namely removing the consistency rewards. It also proposes that the program be fully funded (whenever possible) by allocating (part of) the COW tokens the DAO has repurchased with revenue earned from its products.
Moreover, changes specific to the CoW DAO bonding pool operations are proposed in order to further encourage decentralization, as well as mitigate the risks associated with the CoW DAO bonding pool. Specifically, a service fee is proposed to be paid by solvers that have been in the CoW DAO pool for a sufficiently long time. The revenue from this service fee will stay in the Solver Payout Safe and will be redirected to further align the solvers’ evolution with CoW DAO’s goals and vision.
Motivation
At the time of writing, there are 24 solvers competing to provide the best possible execution for CoW Protocol users on mainnet. This dense network of solvers has proved to be very reliable and competent, providing very robust quotes, doing private market making, striving to improve their view of the available onchain liquidity, and taking advantage of the structural benefits of batching. All these are largely due to the solver rewards that solvers receive in COW.
Part of these rewards are currently disbursed as consistency rewards to solvers who do not win batches but submit valid solutions. Consistency rewards were initially introduced to incentivize participation in the solver competition, which does not seem necessary anymore. Also, given the way they are calculated, they commit the protocol to spending at least a given amount of COW per week. Finally, for the past several months, consistency rewards were minimal or zero, which did not seem to affect the competition. It is therefore proposed that these rewards be eliminated.
Moreover, with CIP-34, the protocol has proved that it can generate significant revenue, while still providing very competitive prices to users. Because of that, and as a first very important step towards a more sustainable and long-term plan for the rewards mechanism, this CIP proposes that the protocol revenue, either accrued in COW or post-buyback of COW (as specified in CIP-36), will be used to provide the necessary COW for solver rewards, among other important initiatives as determined by the DAO, and in the case where the amount of COW accumulated by the Treasury that is generated from protocol revenue is depleted, then a new proposal should be put together to ask the DAO for additional COW tokens. This CIP reserves the possibility for solver rewards to also be funded from other sources in the future, to ensure a robust and adaptable system.
In addition to the changes in the solver rewards program, it is also proposed that certain changes regarding the operations of the CoW DAO bonding pool are enacted. The CoW DAO bonding pool, with the processes around it, has allowed multiple solver teams to continue participating in the competition, thus leading to today’s state of the art solver competition that the protocol is running.
However, the minimal requirements associated with joining the CoW DAO bonding pool and the very reasonable limitations solvers face when being part of the pool has also led to the unfortunate situation that a significant portion of the solvers are now part of the CoW DAO pool, and do not have any meaningful incentives to actually leave the pool and create their own bonding pool. This has important downsides, such as:
- All these solvers heavily rely on the default open-source so-called “driver” implementation operated by CoW DAO, that needs to be continuously maintained and updated by the core team, thus creating a centralized point of failure for multiple solvers (besides the constant pressure on the core team to continue the development of the default driver).
- Having multiple solvers in the same pool means that misbehavior of a solver could potentially (and in cases of significant losses of funds) lead to the closing of the pool itself, thus also immediately removing several solvers from the competition.
- Fully relying on the core team to provide robust driver implementations of the driver in any blockchain the DAO decides to expand to, thus making such expansion much slower than it could have been possible.
For all these reasons this CIP proposes that the CoW DAO bonding pool sets additional rules that would still allow it to serve its original purpose, i.e., to facilitate the easy onboarding of solvers in the competition, while also ensuring that the pool itself does not create a decentralization barrier.
Specifically, the proposal is that all solvers that have been in the CoW DAO bonding pool for more than 6 months start paying a service fee, in the form of withholding 15% of the weekly COW rewards solvers are receiving. The COW collected from this service fee will remain in the Solver Payouts Safe and can be used for various reasons, such as bounty-type programs within the Solver Competition to encourage adoption of new features (e.g., support for partially fillable limit orders, integration of certain liquidity sources, CoW AMM support etc) and to further align the solvers’ evolution with CoW DAO’s goals and vision. We highlight that the service fee will also apply to solvers that create a reduced bonding pool, as specified in CIP-44, and we propose that the service fee is waived for the first 3 months of a solver operating within a reduced bonding pool; the rationale being that the reduced bonding pool should be seen as an intermediate step that will eventually lead to a solver either creating or joining a full-sized bonding pool.
Specification
Transfer 4M COW from the CoW DAO Treasury Safe (0x616dE58c011F8736fa20c7Ae5352F7f6FB9F0669) to the Solver Rewards Safe (0xA03be496e67Ec29bC62F01a428683D7F9c204930) on Wednesday July 24, 2024, and moving forward, provide a mandate to the Treasury Core Team to fund the Solver Rewards Safe with COW from Treasury allocation, whenever needed. In case it happens that there is not enough COW to cover the needs of the solver rewards program, a new CIP should be put in place to ask the DAO for additional COW.
Regarding the service fee for solvers within the CoW DAO bonding pool, the starting date will be Tuesday, August 20, 2024, 00:00 (UTC), meaning that the first service fee will be paid on Tuesday, August 27, 2024. Specifically, the evaluation of whether a solver has been in the CoW DAO bonding pool for more than 6 months will happen on the first day of each accounting period, and this will determine whether a service fee should be paid by each solver for each accounting period.