ARC-50: Astroport <> Redpoint MEV round 2

Note: If you like, you may read my proposal here (better formatting).

Astroport <> Redpoint Proposal

Background

This proposal is an updated version of the first Redpoint proposal posted here. Additional topics covered include a discussion about on-chain and off-chain solutions, a cap on Redpoint profits, and gas fee payments.

:memo: Proposal

Redpoint introduced the technique of atomically packaging user and arbitrage swaps into a single TX to the Cosmos community around August 2022.

Our proposal uses this technique in distinct stages, starting with an off-chain simulation and ending with an on-chain solution.

  • The first stage is an off chain simulator that simulates arbitrage and submits a single TX to the chain containing the user’s swap plus arbitrage swap.
  • The final stage is fully on chain, with Redpoint paying user gas fees for arbitrage swaps.

:bulb:Reliability

:thinking: Redpoint’s v1 simulator and Skip’s auction are less reliable than Redpoint’s on-chain solution. In Redpoint v1, and Skip’s auction, some arbitrage swaps will fail or be less profitable than simulated.

An off-chain simulator cannot always know the order of the TXs in the block it is simulating (geolocation, validator block reordering, etc.)

  • This becomes more prevalent the more active the market gets.
  • More TXs in the mempool means less guarantee of TX ordering (network effect).
  • TXs can fail on-chain that pass simulation (example: slippage tolerance is violated).

Therefore, off-chain simulations (including auctions) will be unreliable compared to on-chain.

Off-chain Simulator (Stage 1)

Redpoint’s off-chain simulator submits swaps to a smart contract that guarantees there are never TX failures, even if the arbitrage swap fails. Astroport earns a fixed percentage of arbitrage revenue.

Redpoint simulations take less than 1 millisecond (plus HTTP request time). There is very little “lag time” (where unknown TXs are submitted to the chain. See Reliability section).

On-chain arbitrage capture by Redpoint

:information_source: Redpoint’s on-chain arbitrage smart contract knows pool price and liquidity at the time of the trade. It will never be inaccurate as a result of incorrect data or have the reliability issues above.

  • Redpoint will start by offering 3 and 4-pool cyclic arbitrage on Astroport, which will capture the majority of arbitrage, and is the most common technique used by on-chain trading bots.
  • Redpoint’s solution will gradually integrate with other on-chain DEXes that offer sufficient liquidity for arbitrage purposes (e.g. on Terra, top Terraswap pools).

Redpoint will continuously evaluate cross-chain strategies as they become viable (e.g. interchain queries, interchain scheduler, swap and forward) and collaborate with Astroport on the preferred way forward.

Scenario: Off-chain vs. on-chain solutions

Redpoint’s off-chain solution can be visualized as follows.

Comparatively, Skip’s auction solution takes longer to run, and thus more TXs may squeeze between Searcher estimates:

In fairness to the Skip team, TXs can also squeeze between Redpoint’s offchain solution. The comparison being drawn is the time-to-run and the consequential impact on accuracy.

Contrast this with Redpoint’s on-chain solution.

  • Note that it’s possible for the TXs from TFM and Unknown (image above) to attempt their own arbitrage capture.
  • What’s important is the on-chain solution will capture a greater percentage of total arbitrage, because it knows all pool prices at the time it runs, including any impacts of TXs that add/remove liquidity, swap, etc.
  • By contrast, it’s indeterminate whether the off-chain solutions above will know about the sandwiched TXs, or be able to predict where they will end up in the block.

Redpoint Fees

:point_right: Note that the fee schedule has been updated after feedback from the community, and to provide incentive to complete an on-chain solution which will maximize revenue for the Astroport community.

  • Redpoint will charge 5% of arbitrage revenue capped at $200,000/year per Astroport deployment while off-chain simulator is active.
  • Redpoint will charge 5% (year one and two) for on-chain solution, allowing community discussion after year two.
  • Redpoint will pay user’s gas fees for the arbitrage swaps in all stages of our proposal.

Proposal Highlights

  • Earnings cap for v1 (simulator) which provides incentive for long term decentralization (no dependence on Redpoint or another company’s off-chain infrastructure).
  • Clear fee distribution with no guesswork about searcher incentives.
  • On-chain solution has significantly greater upside for Astroport (see Scenario), and removes dependence on off-chain infrastructure.
  • Fully on chain solution more accurate and covers more clients than just Astroport frontend.

:loudspeaker: We believe Redpoint’s approach matches the forward-thinking and innovative spirit of Astroport!

Thanks everyone for reading and feedback!

BTW: I welcome comments from the Skip team if they want to comment, it is only fair considering their solution was mentioned.

And thanks again all for the feedback & comments on v1 of the proposal, and for checking this one out!

Update: edited title to say ARC-50.