BIP #10: Participate in KeeperDAO's coordination games

EDIT: the post was edited on the 9/2/2022 to formalize it as a BIP in the title, and to supply additional technical details.

Short summary

Share MEV in the rebalance process with KeeperDAO, on the expense of some part of the DAO fees.
In return:

  1. get more efficient and secure rebalance process.
  2. help in seeding new backstop markets (KeeperDAO will deposit $5m in B.Protocol’s BAMM’s

Overview

B.AMM and MEV

B.Protocol’s BAMM democratizes liquidations and eliminates most of the liquidation process’ MEV.

However some MEV is still applicable in the following components:

  1. The B.AMM incentivises keepers to call the liquidateBorrow function by giving them a certain % of the liquidated collateral.

  2. The rebalance process offers the seized collateral for sale, and often offers it at arbitrage prices.

In both components actors can benefit from GPA (gas price auctions) which give rise to MEV.

KeeperDAO’s coordination games

KeeperDAO’s Coordination Game is a protocol that defines a set of incentives that form an equilibrium in which keepers’ utility is maximized when they coordinate, rather than compete for order execution.

This gives rise to a win-win scenario in which Keepers share their profits with transaction originators, rather than with block producers.

B.AMM coordination games

In this proposal we focus on our LUSD backstop, and hope to extend it to more integrations in the future. Moreover, in the LUSD B.AMM only the MEV in the rebalance process can be mitigated, as the call for liquidateBorrow is done outside the B.AMM.

In order to participate in the game the DAO will set the rebalance fee to 1%, which effectively increases the ETH sell price by 1%, however the effective fee will remain zero for whitelisted parties, namely to KeeperDAO’s keepers.

Hence, as long as the B.AMM imbalance is medium/small the offered sale price will be beneficial only to the keepers who participate in the game, while for the rest of the world it will be less attractive than the current market price. While for bigger imbalances, the offered discount will outweigh the fee, and hence non-whitelisted parties could also benefit from buying…

In return, KeeperDAO will commit to seed B.Protocol’s backstops integrations, up to an amount of $5M, starting with Liquity integration. Here you can read their governance proposal on this matter.

Technical Details

It is proposed to set the fee value in the LUSD B.AMM to 1%, and to set a new admin to the fee pool.

The new admin is a smart contract with a special swap that allows whitelisted keepers to buy ETH from the B.AMM and get the fee rebate, in a single transaction.

EDIT:
Full implementation of the new admin contract will be provided before any vote will take place.
The coordination game will take place on the recently deployed BAMM, namely, the BAMM with address 0x00FF66AB8699AAfa050EE5EF5041D1503aa0849a, and it is proposed to:

  1. Set fee to %1 by calling setParams with _A=20 (no change) and _fee ==100 (1%).
  2. Set the fee pool contract (which is deployed in address 0x7095F0B91A1010c11820B4E263927835A4CF52c9) to 0xB67200fEfCBf36B0CA3D70d59ebf90D0B9B38f82. This is a smart contract that implements the integration with keeperdao. Its source code can be found here.
5 Likes

Questions:

  • I’m reading this as the B.AMM fee parameter being maxed to facilitate the coordination game utilization, but thus also using all the potential “BPRO hodler” potentials of this fee parameter. Is that correct? I.e., DAO no longer has any control over this 0-1% fee or the funds it generates for all cooperation game B.AMMs?

  • How do I read the duration of this? Is this for all B.AMMs using coordination game while the $5M are “ready to be used”? What if the $5M are revoked or coop ended? Is the rebate (1% fee) then returned to DAO admin? Will coordination game then still be used without the rebate/fee?

More of a KeeperDAO question:

  • From skimming the KeeperDAO part of the proposal, they profit from having the turnover of our B.AMM, in which they get ~20% of the MEV extracted from the operation, while B.Protocol gets ~80%. Would this be correct? Where do these figures come from (e.g. they mentioned the ~80% in their proposal)?
5 Likes

In general the plan is for the BPRO DAO to remain in control. The participation in the game is a DAO decision (for both KeeperDAO and B.Protocol) and can be revoked at any time by the DAOs.
At least this is what the current proposal is about (to my understanding - and this is how it is will be technically implemented).

The control is still in the hands of the DAO, who can set it to different values in the future.

In the traditional world both parties would seat down and form a precise legal agreement. In the blockchain world the alternative is either a smart contract agreement, or a DAO vote that can always be undone.
As both projects have some history of competing with each other, and as the coordination games still have big off-chain component, and as the deployed B.AMM is non upgradable and was not initially designed to be integrated with KeeperDAO, we preferred to have a quick collaboration to test things out, and formalize it with DAO decisions, rather than with smart contracts.
LUSD is only the first integration, and things might become more formal in the future.

But to answer your question, at this point the KeeperDAO and B.Protocol DAO can technically revoke the decisions at any point. Technically even if the DAOs commit not to revoke it does not matter if it is not forced by a smart contract, which at this point will be hard to do.

Will try to have someone from their team to respond it.

2 Likes

Thanks. Let me try again. The DAO is deciding to use the max of 1% fee to get a better rebalancing deal for those providing liquidity to the backstop by giving KeeperDAO “the advantage” and returning some of the extracted MEV. In doing so the DAO is for the first time actively deciding to use this fee. If they wanted to “undo” it or otherwise utilize the fee it would simply require another DAO vote to set back the original fee pool admin and/or adjust the fee itself. Correct?

2 Likes

yes. The B.Protocol DAO remains with full control.

2 Likes

Hi @TragedyStruck it’s Hazard from KeeperDAO. Thanks for your great questions.

How do I read the duration of this? Is this for all B.AMMs using coordination game while the $5M are “ready to be used”? What if the $5M are revoked or coop ended? Is the rebate (1% fee) then returned to DAO admin? Will coordination game then still be used without the rebate/fee?

From our perspective this is a long-term strategic decision. At KeeperDAO our primary objective is to relay as much transaction flow as possible through our Coordination Layer, which will result in the most possible MEV being internalized to protocols and users at the application layer.

As you know we have our own liquidation backstop technology, Hiding Vaults, but whereas that is tailored for individual borrowers, the current landscape around protocol-level backstops is far more interesting from the perspective of transaction flow and the value we can add.

In that landscape, we strongly believe that B.AMM or B.AMM-like technology is currently the best route forward. As @yaron points out, we have historically been competitors. But I believe that our interests are in fact complementary. We at KeeperDAO want to see B.Protocol become the standard for how lending protocols backstop their risk.

If we wanted to capture liquidation flow ourselves, what would it take? We would need to go door-to-door to each and every lending protocol, figure out a way to provide strong liquidity guarantees to each, make separate arrangements with each, separate integrations with each, and keep innovating at the same time. That’s a whole project in itself, and it already exists – it’s called B.Protocol.

B.AMM offers us an abstraction layer where we can add value to ALL lending protocols that integrate a B.AMM. Therefore, it’s in our interest to see that ALL protocols integrate one! Our goal is to do whatever we can to help B.Protocol eat the market for backstop technology and become the gold standard across all chains and protocols.

The B.AMM is beautiful because it’s a perfect vehicle for synergy between our projects. It lets B.Protocol be the experts at working with protocols, incentivizing liquidity, and catching that initial offloading of risk during liquidation. Then, the problem is converted elegantly into one of efficient on-chain arbitrage and portfolio rebalancing, and that is where we are experts. Working together gives us a truly powerful product/service combination.

That said, we have also been asked about specific terms in our own governance thread, maybe we should draw some up. But as @yaron points out, we planned this to be an informal collaboration to start, and perhaps become more protocolized as time goes on.

From skimming the KeeperDAO part of the proposal, they profit from having the turnover of our B.AMM, in which they get ~20% of the MEV extracted from the operation, while B.Protocol gets ~80%. Would this be correct? Where do these figures come from (e.g. they mentioned the ~80% in their proposal)?

Our Coordination Layer includes governable logic which determines the split of the return. Initially, originators (B.Protocol or the specific B.AMM in this case) receive 80%, and 20% is split up in various ways like burn, going to stakers, or going to the DAO Treasury.

More information is available in our docs, here is a diagram overview:

In the diagram, ‘User/MM’ can also mean ‘Originator’ and will refer to B.Protocol, the B.AMM, etc.

Thanks again for your questions, I’m happy to answer more.

4 Likes