[PROPOSED] Serum Fee Reduction Proposal

Fee sharing happens now where the GUI hoster gets a % of fees, ecoserum gets a % of fees and the rest goes to serum

I think the following fee structure for non-stable markets may be better considering the recent concerns of SRM and MSRM sharing. There will still be a benefit for holding SRM and MSRM while trading, but the fees tiers will be much closer together.

I also propose that stable market fees be reduced further too, to 1 bp taker + 0 maker, instead of the 2.5 I proposed initially.

I think this will happen, its just timing and details… keep your thoughts coming

Actually, I think the following fee structure for non-stable markets would be better since it’s more competitive. I’ll start an on-chain proposal for this thread in a few days and link to it when started.

Why remove maker incentive? I’m genuinely curious on the thoughts here.

1 Like

A maker rebate requires an extra fee for the taker if Serum wants to keep some fee revenue for itself. For example if Serum wants a 3 bps fee on every trade for itself, and the maker rebate is 2 bps instead of 0 bps, the taker fee needs to be 5 bps (5 bps - 2 bps = 3 bps).

I think Serum should minimize fees to get as much activity as possible. Serum isn’t just an exchange, it’s a platform for other projects to compose with, so any fees have an extra negative impact. But I think Serum’s tokenomics require some sort of a non-trivial fee for buy back and burns since the SRM unlock schedule is very inflationary otherwise. All this considered, there doesn’t seem to be bps available for a maker incentive.

In your opinion the only incentive for MMs should be to capture the spread?

In Serum’s case, yes, I think that’s the best option for the long-term.

1 Like

I think it is interesting and I think you are proposing in the right direction. However, I also think that at the same time this needs to be integrated with a token burn / utility / tokenomics mechanism of sorts in order to keep the SRM token viably held or utilized. If and when the fee is reduced to what you have proposed (which is a 80-90% reduction), then we need to ensure the lack of “buy and burn” would be replaced by something else. What do you think?

I think Serum is in an interesting situation because of its unique tokenomics. Either it has to sacrifice product (with fees) for token price (with buy back and burns), or sacrifice token price for product. Uniswap V3 (basically Serum on Ethereum) doesn’t have to make this tradeoff because of UNI’s different tokenomics, and that results in much more volume and a better UX since there is no protocol fee or holding mechanics.

Like you mentioned in the tokenomics thread, many people are focusing too much on the price of SRM. I think that we should optimize for product experience first and forgo some buy backs and burns, even if that results in a short-term dip in price, because it will be best for the long-term. But those tradeoffs are best discussed in the tokenomics thread IMO, I want to focus on product experience.

I know this this was deleted, but wanted to respond if anyone checks the edits.

Atrix is 100% based on Serum. Without Serum, Atrix would not be possible. Atrix doesn’t have any internal AMM to fragment liquidity or generate extra fees for Atrix. All 57 pools (most permissionlessly created by the community) and 33 farms provide and incentivize liquidity directly on the orderbook. I believe the SRM grant has been successful in incentivizing liquidity to migrate to Serum, which ultimately improves the product experience for all Serum users through better liquidity and a simple UX.

I don’t want to come off as unsympathetic to the community. I saw your replies in the tokenomics thread and I truly understand the Serum community’s concerns about tokenomics and how lower fees reducing buy back and burns would affect SRM’s price. But any real long-term solution for Serum is going to be painful, for someone in some way. It’s just the tokenomics constraints we have to work with, unfortunately. Maybe the solution is token burns, token redistribution, or even doing nothing, but I think low fees have to be a part of the long-term goal of Serum. There just doesn’t seem to be another way in this ultra competitive landscape.

Thanks @kaiba. Keeping the high fees like I suggested before would be best for SRM holders, but you’re right that Serum is stuck in a difficult position. Pausing this proposal until we figure out tokenomics would be a good idea, since these new fees will really lower the current burn rate.

Growth of Serum should not be slowed down b/c of some concerns with tokenomics.

Folks from Atrix and Mango obviously agree a fee reduction is a great move. Would also like to see thoughts from @JumpCrypto, @AlphaRay, and any other major players in the Serum ecosystem.

Definitely for a fee reduction but would also like to see a maker rebate on these fees as well. In general, having a maker rebate is also the approach most centralized exchanges take for good reason. Rebates would help market makers quote tighter and help with the general look of Serum’s orderbook. Our proposal would be to set the range of taker fees between 7.5bps and 4bps and a range of maker fees between 0 and -2.5bps. This would keep the fees competitive with most major centralized exchanges (FTX, Okex, Bybit all have fees in a similar range) without sacrificing as much profit for Serum.

Having a maker rebate would also open Serum up for composing stable swap AMMs such as Mercurial and broaden its playing field. Without a maker rebate, a stableswap AMM would need to put a spread that is as wide as their LP fee but orders that far into the book on a stableswap would be unlikely to transact and make them uncompetitive.

For example, Mercurial and Saber charge 4bps on their stableswaps. If they want to make 4bps of profit on the orderbook without any maker rebates, they would need to quote 4bps away from the mid price. On unvolatile markets such as USDT/USDC large market makers would easily price them out.

While this could be beneficial to the end user, it isn’t necessarily the optimal play for Serum since project composability is an important factor for its success. I understand there is another post up for stableswaps but it seems most of the conversation is here and a blanket solution seems to be the fastest way to continue moving forward.


An on-chain proposal has been created with the following items that were discussed in this thread:

  • Updated 40% fee rebate for GUI hosting
  • 4 bps → 3 bps taker, 0 bps maker fee tiers for non-stable markets
  • 1 bp taker fee, 0 bps maker fee tiers for stable markets

Vote here: Governance | Solana

Have you populated the instructions section so the changes you propose get executed onchain? Is that even possible for this subject matter rn?


An updated on-chain proposal was created covering just the fees, not the GUI rebate, because that part may not have been discussed enough yet. Covers the following:

  • 4 bps → 3 bps taker, 0 bps maker fee tiers for non-stable markets
  • 1 bp taker fee, 0 bps maker fee tiers for stable markets

@MJP instructions section is empty since these changes may be very involved to implement for the DEX. The idea is that Serum contributors will implement if this proposal is passed like mentioned before.

Vote: Governance | Solana

1 Like

Handling maker orders at scale requires running your own crank and derisking your maker order is magnitudes more complex than takers on the engineering side. Each time you want to change your maker order you have to pay a solana fee as well.

Basically you are probably going to pay .003 cent in fees just to get an order filled if you are lucky – but likely you will be paying closer to 5-10 cents before an order gets filled – which means that makers are already paying significantly more fees than takers in the average case right now.

For more context, both Raydium and Orca are almost always better price execution than serum because of how difficult being a maker is on serum and this proposal only seems to make me believe that the gap will be bigger if it passes.

You have taught me something, thankyou.

Do you have any suggestions on what would be a better plan of attark? To create a more competitive fee profile, to create a platform for high volumes and growth of the ecosystem.

i meant .3 cents in fees woops.

i think alpha ray suggestion is fine. 10bp on taker is reasonable across fee levels and maybe the MSRM /serum is only for fee reductions between 0 to -6 bps.

this leaves 2bps for hosts and 2 bps for burn/community.

But looks like this movement where makers get rekt will pass regardless so :stuck_out_tongue: