0% of indexed voting power
Router Taker Fee Proposal: Lower the Deepstate Router Taker Fee from 10 bp to 2 bp
PROPOSED BY
The goal of this proposal is to make the Deepstate book a competitive route for takers. At 10 bp the book wins almost nothing, while the best route costs 1-4 bp a side and 2 bp is the peer rate. This proposal will lower the router taker fee from 10 bp to 2 bp. The fee recipient, the maker fee (0) and fill stay unchanged, so 0x's Settler keeps working.
Takers will pay 2 bp instead of 10 bp. At 2 bp the book can win large trades it loses today, but 2 bp alone does not win everyday trades, and a cut makes stale quotes cheaper to hit.
This proposal makes one change, setFeeConfig(0xf5011f7f3b09d2a6d1cf74a7f50400b37a6ea82c, 2) on the router; because the Rewarder Factory V3 owns the router, the proposal first returns the router to the Governor (returnDeepstateOwnership()) and hands it back afterwards (transferOwnership). KyberSwap's best route cost roughly 1-4 bp a side at $10k-$100k (three overnight reads, 10-22); Uniswap v3 alone costs 5-8 bp at $1k-$19k (10-09). At 10 bp, over 49 hourly reads from 10-17 14:48 to 10-19 14:22, the book beats the best route only from a median $224k-$301k (sells) or $270k-$327k (buys), at a maker margin of 0.5-8 bp over a Lighter hedge; in the 30 days to 10-21 one trade on other venues ($249k) cleared that, out of $28.6M/day. 2 bp is the peer rate: Kuru's new books charge 2 bp (from 10-15) and Aster's stock perps 1.25 bp (from 09-07).
The book only wins a trade when the trade is big enough that every other route costs more than the book. We call that size the crossover. It depends on two things: the taker fee, and the maker's margin (how much a maker charges above the price it can hedge at). Over the 30 days to 10-21:
- At today's 10 bp, the book wins only sells above $224k and buys above $270k, and only with a 0.5 bp margin: 1 trade. At a 2 bp margin or more, it reached none.
- At 5 bp, with a 2 bp margin, the book wins sells above $175k and buys above $212k: 1 trade ($8k/day). With a 4 bp margin it also reaches 1 trade. Only a very thin 0.5 bp margin brings it down to $98k / $169k: 5 trades ($26k/day).
- At 2 bp, with a 2 bp margin, the book wins sells above $68k and buys above $110k: 16 trades ($54.5k/day). With a 4 bp margin, $124k / $186k: 5 trades ($26k/day). With an 8 bp margin, $218k / $267k: 1 trade ($8k/day).
So 2 bp with a 4 bp margin reaches what 5 bp reaches only with a 0.5 bp margin.
Even the best case is small: 0.19% of the NVDA volume on other venues (6% of its volume in trades over $25k). A 2 bp fee wins some large trades, not everyday ones.
If this proposal passes, I will draft a follow-up proposal to rebate part of the taker fee to makers who stake 2DEEP: up to 50% of the fees paid on trades filled against their orders. A staked maker could then quote up to 1 bp tighter, so takers' all-in cost (price plus fee) could beat what the 2 bp fee alone allows. The rebate needs new contracts, so it would be a separate proposal with its own review.
ACTIONS (all three, in this order; ETH value 0 for each)
- target 0xb72e42cc8e835b2a79c7e50cdb8986668e3cd59a (Rewarder Factory V3)
calldata 0x248f4949 (returnDeepstateOwnership())
- target 0x6cf19308C22FC82ea620Fa0B3E94948d20f27B96 (Deepstate router)
calldata 0x350e9756000000000000000000000000f5011f7f3b09d2a6d1cf74a7f50400b37a6ea82c0000000000000000000000000000000000000000000000000000000000000002 (setFeeConfig(Governor, 2))
- target 0x6cf19308C22FC82ea620Fa0B3E94948d20f27B96 (Deepstate router)
calldata 0xf2fde38b000000000000000000000000b72e42cc8e835b2a79c7e50cdb8986668e3cd59a (transferOwnership(Rewarder Factory V3))
CAST 0 2DEEP OF VOTING POWER
VOTING ENDED · EXECUTED