Liquidity Mining Program

[adrian]

Serum should launch a liquidity mining program to grow the Serum ecosystem.

Protocols that provide TVL to Serum should be eligible to apply for SRM grants and provide SRM emissions to their users.

The purpose of the grant would be to offer SRM incentives alongside participant protocols’ existing or planned liquidity mining incentives. Participant protocols include automated market makers (AMMs) built on Solana that share liquidity with Serum.

The SRM incentives being offered would only be applied to eligible markets that interact with Serum’s order book by placing open orders.

The program would include the following eligible markets :

• BTC/USDC
• ETH/USDC
• SOL/USDC
• SRM/USDC
• X/Y (an optional pair of the projects choosing - as long as it exists on Serum e.g. USDT/USDC)
• X/USDC (optional - reserved for the participating project’s token)

The program could be offered in rounds that last for 3 months. Projects can apply to be included in the next round and lock in their SRM grant for review by the DAO before a pre-determined date.

The amount of SRM offered per grant should be enough to provide approximately 10% APY, for a total value locked of USD $300m. Projects that provide a higher percentage of their liquidity to Serum could be eligible for larger grants that allow them to boost the 10% APY up to 20% APY.

The parameters should be subject to change in later rounds.

For example, if SRM is valued at $10, a project can apply for a grant of between 750K SRM and 1.5M SRM, depending on the percentage of its liquidity it sends to Serum. This should be enough to satisfy between approximately 10% and 20% APY for 3 months with 300M Total Value Locked.

[kaiba]

Hi everyone! I’m kaiba, founder of Atrix (atrix.finance), a Serum-based AMM project. My day-to-day role is being an Atrix protocol core developer and managing the rest of the team. Very excited to be part of this first official community discussion.

For context, Atrix allows projects to permisionlessly create Serum pools and farms, so new and existing projects can easily bootstrap liquidity on Serum.

We think this proposal adds a ton of value to the Serum and Solana ecosystem as a whole. Looking at Solana on-chain data, Serum is lagging behind some of the other top Solana projects in terms of TVL (as of Oct 7th, 2021), as shown in the screenshot below from DefiLlama - DeFi Dashboard.

Protocols such as Saber and Orca, which do not provide any liquidity to Serum, combined currently have multiples more value locked than Serum. Of course, these protocols provide value to the ecosystem as they are, but we think it would be better if Serum is a central source of liquidity on Solana, rather than multiple fragmented protocols. This would lead to better prices and execution for traders and Solana programs that swap tokens internally.

Therefore, incentivizing liquidity for projects that contribute to Serum’s liquidity (Atrix, Raydium, etc…) would benefit not just Serum, but more broadly most users of Solana in the form of better pricing.

Furthermore, we propose that it would also make sense to calculate the amount of SRM granted as a function of not just protocol TVL, but specifically protocol TVL which is deposited into Serum.

---

To illustrate:

Protocol A: $1B TVL, with 15% deposited into Serum ($150M).
Protocol B: $500M TVL, with 90% deposited into Serum ($450M).

Grants according to the current proposal:
Protocol A: $30M (~10% APY on 300M TVL cap)
Protocol B: $57M (~20% APY on 300M TVL cap)
Total granted: $87M

SRM USD value granted per USD of Serum liquidity:
Protocol A: $30M / $150M = $0.200
Protocol B: $57M / $450M = $0.127

In this example, Protocol B is providing 3x ($450M / $150M) more liquidity to Serum than Protocol A, but is receiving 37% less grant value per USD of Serum liquidity than Protocol A. In other words, Serum is spending 57% more on each unit of Serum liquidity provided by Protocol A compared to Protocol B. This is a suboptimal use of SRM when viewed through the lens of liquidity of Serum per SRM granted.

Furthermore, incentivizing liquidity more on Protocol B would lead to even more TVL on Serum, as Protocol B’s APYs would be increased (leading to more deposits) and as Protocol B deposits 90% of all TVL into Serum vs the 15% that Protocol A does.

---

We propose a more flexible calculation of SRM granted in proportion to Serum deposits, NOT protocol TVL which includes non-Serum liquidity (since that does not benefit Serum in any way):

Protocol A: $1B TVL, with 15% deposited into Serum ($150M).
Protocol B: $500M TVL, with 90% deposited into Serum ($450M).
Total Serum deposits: $600M

Grants given:
Total grant USD value available (for example): $90M
Protocol A: $22.5M ($150M / $600M * $90M)
Protocol B: $67.5M ($450M / $600M * $90M)

SRM USD value granted per USD of Serum liquidity:
Protocol A: $22.5M / $150M = $0.15
Protocol B: $67.5M / $450M = $0.15

Here, Serum isn’t suffering any losses due to inefficient grant distribution, and projects are treated fairly with respect to the amount of liquidity contributed to Serum.

---

Would love to hear any thoughts in response to this proposal!

[bartek]

TVL is not the best metric for Serum - what really matters is transaction volume. My understanding is all open orders and unsettled trades are counted as Serum TVL. No one wants to hold a long standing open order on Serum because this money can be used in borrow-lending protocols.

I think we can’t compare Serum to AMMs because AMMs are inflating their tokens by issuing so called “rewards”. Serum is so far deflationary because of token burns proportional to transaction volume.

In my opinion Serum business model is more sustainable than AMMs business model which is not sustainable long-term because of inflation which attracts users/volume.

AMMs were created because of technical limitations of Ethereum. I can’t understand why so many projects are trying to copy AMM logic to blockchain which is not constrained by Ethereum limitations.

What we should focus on is to increase transaction volume - we need something like 3commas for Serum DEX (with super simple interface) to allow retail investors to engage in market making using simple strategies like DCA.

[AusteritySucks (Dr. AusteritySucks)]

I don’t understand why liquidity is seen as a concern that should be addressed with any urgency. Serum liquidity is fine, especially on the pairs that you mention there.

Raydium projects liquidity on Serum as well for lesser used pairs where people passively can provide via AMM.

There are many problems on Serum to solve and I don’t think liquidity is one of them.

[theox.sol]

Dev from Almond, wanted to throw in my thoughts on this program.

TVL matters in many ways for a project, which is why even the founders of Solana heavily publicize it as a key metric to measure growth. Ultimately, like it or not, this is the number the average user coming from Uniswap or PancakeSwap looks at to determine if a project is successful.

Big number good, small number bad.

Like others mentioned, volume definitely matters, but I think this program helps with that too.

Anytime the Serum market price moves, AMMs would have provided liquidity for that price move, generating volume for Serum. The amount of volume generated for Serum is directly proportional to the TVL these AMMs provide to Serum and the price move they capture and facilitate. Of course, the AMMs would then again adjust to the market price, being in position to generate even more volume for the next price move.

This program would also be a great way to incentivize more protocols to shift towards using Serum, which is what Serum really wants. The stronger network effect Serum has through protocol integrations, the more staying power it has.

[JumpCrypto]

Jump Crypto is a longtime and long-term partner of Serum. We are excited about Serum’s efforts to aggressively grow the Serum ecosystem.

We agree that many everyday users look at TVL to judge what platform to use, and that it is worthwhile to emit SRM to boost TVL now, while we are still in the early months of establishing Solana’s DeFi ecosystem.

Most of the other protocols with high TVL in the screenshot posted by @kaiba are emitting native tokens at 20+% per year, with the exception of Saber and the liquid staking solutions. And Saber’s TVL is helped by (1) being a stableswap and (2) Sunny emissions over 20%; while the liquid staking solutions are clearly in a completely different vertical. Serum is really the only swap protocol in the screenshot without significant emissions.

So it reasonable to imagine that Serum TVL could be raised substantially with partnerships where Serum offers ~10% APY in rewards and the partner protocol also offers 10+% in their native token. This is a good plan. We also agree with the timeline proposed (try for 3 months; easy to extend if it is working) although it would be nice to clarify how many such partnerships Serum seeks to do. 3? 5? Enough to sum to a particular additional TVL? For definitiveness’ sake, we think it would be great to commit to pursuing partnerships under these terms to boost Serum TVL by 1B USD, which (for 3 months) corresponds to 3-6M SRM emissions (assuming SRM at $8).

As we work to boost TVL, we might as well focus on making it as productive as possible:

• We should ensure that a reasonable portion of the partner protocol’s liquidity is close to the prevailing market price so that it can interact with takers and promote trading volume. For example, imagine a new unhelpful protocol which allows people to deposit USDC and then bids SOL/USDC on Serum at 0.01. That liquidity will never trade and will get a ‘free’ 10% SRM emission. That is an obviously extreme example but, in general, different protocols will be more or less aggressive about posting liquidity near the inside best bid/offer, and we should have a way to reward the ones that are more aggressive.

• There should also be an incentive for protocols to route taker volume to Serum. That won’t boost TVL but it will promote more organic liquidity as part of the virtuous cycle between takers and makers.
• Both of the above issues could be addressed by including a simultaneous incentive for volume. We would suggest emissions of 50% of the above TVL incentives (1.5M-3M SRM) to accomplish this.

Lastly, although emissions will boost Serum TVL, it is important to consider that the partner protocol gets much of the credit from a brand-building perspective. It would be unfortunate for Serum to emit a bunch of its tokens, boosting NewSwap’s APY and driving new users to NewSwap, only to have NewSwap shut off its Serum integration 6 months from now after growing into a big brand. We will leave it to the foundation to judge which partners are long-term in nature, but it is especially important to ensure that Serum is accruing taker volume (which promotes organic liquidity, which cements Serum’s long-term status as the premiere DEX) as part of the partnership.

[Nikita (Nikita B)]

Nice to see Serum adopting a more aggressive stance. Protocols will tend to extract value from SRM grants so loopholes patches are needed.

[JiangNotYang]

Hi everyone. Serum supporter / investor here. Many great ideas being posted and glad to see that Serum is taking an aggressive approach to attract volume. Also it is great to have this forum to bring many incentivized individuals to discuss how to increase the value that Serum provides the whole ecosystem.

While I have no particular conviction with liquidity mining program, I think it would not hurt to try it with a fixed time frame as suggested. We should assess the effectiveness of the program, perhaps by observing whether the program permanently or temporarily boost volume x months after the liquidity mining program ends.

I’m sure most of us here agree that using tokens to effectively “buy” volume must accrue some sort of long term value to Serum or it is not worth doing. I think value accrual happens when users recognize the value of a CLOB, which should be quite obvious to everyone here.

I agree with bartek - lowering the barrier to entry to use Serum DEX by everyday folks (e.g. probably most buy side folks and retail traders) is of paramount importance. Perhaps another way to effectively remove friction in using Serum is building out APIs for dumb people like me , such as:

• Placing order
• Observing order status
• Transaction history by market or account or price currency, etc.
• Listing all markets in plain English

I understand there are a number of nuances on the tech side due to the fact that the CLOB is entirely onchain - quite a technical accomplishment to say the least. But to the ultimate end user, ease to use is important. I still feel a bit jittery here and there between order execution and settlement, just as an example. It is just not buttery enough yet.

To paraphrase most of the rich folks: Serum DEX needs to be 10x better in user experience than existing options (AMM, CEX) in order to incentivize switch.

[MJP]

Hi all,

My thoughts are as follow, though I am not big brain like people in this thread:

The Serum Foundation should pursue an aggressive program to both increase TVL, the headline number the market looks for, and the frequency of fee generating transactions that sit on the CLOB/CLAAOB/Permissioned CLOB.

This mission should through the use of soft power, actively seek to maximise the success of projects that send 100% of their flow to the CLOB, and are 100% long term loyal to the foundation.

Projects that only send part of their flow, or are proactively pursuing strategies to undermine the Serum Foundation and its SRM tokenomics, naturally should not receive the same level of Foundation support.

I suggest all Foundation power, network, resources, super shadowy devs, influence, is marshalled through a concerted 6 month effort into projects that fit the criteria that promotes long term success for the Foundation.

A liquidity mining scheme can be a part of this, but as we have just seem with the TVL attributed to Atrix/Almond, asset growth can be achieved without the level of Foundation incentive we are talking about here. So what am I saying here?

Sure lets do a LMP, but lets not have it too long, and lets make sure the projects that are the receipient of it are 100% long term loyal and that the LMP achieves something more than just TVL, as we have just seen this is achievable without too many incentives.

Any wasted emission just puts more Serum out into circulation and just highlights the single biggest issue facing the foundation over its life time…the maximum amount of Serum tokens that may be isssued at 10bn.

I know the market says its a meme, but this will not last forever.

[theox.sol]

Will create a separate thread

[AlphaRay]

AlphaRay from Raydium here.

Our thoughts on emissions have always to think of them as more of a budget with a clear goal to grow the project rather than to just send tokens out aimlessly which is why we think that looking at TVL figures for Serum is rather flawed. First of all, as @JumpCrypto noted, this measurement is easily gamed by protocols placing orders far from the mid-price which would never be transacted against. More importantly, placing these impractical orders consumes resources from Serum and Solana which is already having trouble keeping up at times. To put it in numbers: for an AMM which places 5-10 orders on each side of the book, this would result in a 10-20% wastage of resources.

For an AMM, TVL is an important number since the amount of liquidity supplied, trading volume and in turn revenue an AMM can generate is directly correlated to TVL. For Serum, we should look at metrics that Coinmarketcap and Coingecko track on exchanges:

• Spread
• 2% Depth
• Volume
• User count
• Trading pairs offered

Uncoincidentally, these are generally the same metrics centralized exchanges use to measure market makers.

As for the APY, while its important to have a target, the number should really be up to the market to decide by adding or removing liquidity. If the APY is a set number, the system could also be gamed by a farmer adding a large amount of liquidity for a day to mine all the rewards and pulling their tokens as soon as the rewards are all farmed.

[meow (ming)]

Ming from Mercurial Labs here! I have been reading through the various discussions made, and we are in general strongly in favour of this effort. We are in the process of building several Serum focused pools, and having this LM will help tremendously with the success of these pools, and incentivise motivate us to in turn put in the appropriate investment into it.

For example, we will be announcing an institutional vault this week that will be providing liquidity to Serum, while also making the funds available for institutional partners to borrow. This would be the type of innovative systems that will underpin Serum’s position as the backbone of trading on Solana, and this LM will help to bootstrap this pool tremendously.

The main debate as brought up by @AlphaRay and @JumpCrypto is that if all the liquidity in the AMM is going towards the TVL count, it is likely not very efficient as the tail end of the AMM is likely to never going to be taken. However the answer to this question is likely not to be a simple one, for it really pertains to what goals the serum community has at the moment.

Is it to have a higher TVL, in which case the proposal of 100% of liquidity will be correct? Or is it to have “useful liquidity”, in which case a combination of incentivising the first 10-20% of AMM liquidity and volume will be a good combination. Once we have the current strategic goal decided upon, I think the right LM direction will be pretty straightforward.

[adrian]

Thanks everyone for your thoughts and feedback on the topic.

Here is a summary of the shorter term considerations discussed to consider for proposals now:

1. SRM grants should also reflect the usefulness of TVL deposited to Serum, noting the distance between the mid price and orders placed on Serum. In other words, the program should be careful not to reward vanity metrics.

• One way around this could be to just measure trade volume. Grants could be allocated proportionate to the historical trade volume of the TVL that was placed on the book over some period of time.

2. The purpose of the program should also be to grow the ecosystem and incentivize more protocols to share liquidity with Serum.

• Just an idea but maybe there is something built into the economics to make it attractive for new / smaller projects to compose with Serum.

3. Consider how this program this will increase SRM’s circulating supply and make efforts to ensure emissions are not wasted.

• It’s important to be clear about the monthly grant size available. As an example, it could start with 5M SRM per 30 day period. Projects would need to apply for a share of that 5M each 30 day period.

Some of the less immediate considerations are:

• Volume is more important than TVL for Serum, what else can be done to increase taker volume on the exchange. Agreed and probably warrants a serparate thread.
• The barriers to entry to trade on Serum are high. Refining the experience for the end user is potentially a higher priority. Agreed and warrants a separate thread.

Please challenge anything i’ve mentioned here or let me know if i’ve missed anything!

[MJP]

Just on your point 2, there should always be a differential drawn between a project that wants a grant and their current business model involves some off Serum matching and on Serum matching, and the projects that are nativly 100% serum matching for everything they offer. The priority, with due care and analysis, I would expect to always fall to the latter type of project.

[AlphaRay]

I’d say this isn’t necessarily true. A swap that gives the user the best price and where users know they can go to reliably find the best price leads to higher user growth and is likely to be a net positive for the ecosystem. We can see examples in AMM projects looking to partner with aggregators to find additional exposure and increase their userbase/volume through these means. In the end, a lot of it comes down to the relative size of their userbases, how prominently the aggregators feature the underlying DEXs and most importantly, how much additional volume the aggregator is sending to the DEX.

[MJP]

I know there are arguments both ways, but for me I try and keep things simple. Projects that send 100% of volume to the CLOB should get evaluated differently to those that dont…else how are we going to make the Serum CLOB the best place to do business? These nudges and actions and grants all add up and I think the end destination is the CLOB having the best liquidity possible, which surely is the goal here for the Serum foundation.

[VxS3]

Agree with MJP point, you are still “segregating” volume to non-serum orderbooks.

[nkosi]

I would reiterate the comments above that have highlighted the fact that TVL is a poor metric for an exchange. As the rollout of Uniswap v3 demonstrated very clearly, capital efficiency and returns for market makers are increased by achieving the same trading volume with lower TVL.

As with centralized exchanges, the success of Serum should be measured in terms of transaction volume and fee revenue captured, and ultimately buybacks/burns and staking rewards to SRM holders.

Currently Serum incentivizes liquidity providers to provide liquidity around the bid/ask via maker rebates. A well-designed AMM can place its swap orders on the Serum book using simple market making algorithms in order to capture this premium and to offer swaps with close to zero fee for users. Indeed this was the original plan for Raydium, although they subsequently pivoted into using their own pools of liquidity to execute the majority of trades.

From a game theoretical perspective, if all protocols utilize the Serum order book to execute trades it creates a win for everyone (higher liquidity, lower slippage, lower fees), but in the short term protocols are able to make more profit by charging higher fees to swap using their own liquidity pools.

Rather than granting protocols SRM based on TVL, a simple and effective way to increase trading volume and liquidity on Serum would be to temporarily boost the maker rebates, until it becomes a no-brainer to execute trades this way rather than via fragmented liquidity pools (it may even make it possible to build a zero-fee swap interface that places orders on the Serum book under the hood).

[JumpCrypto]

After seeing a few project-specific proposals, and after fleshing out Volume Incentive Program, we’d like to revise our stance.

We are opposed to the blanket Liquidity Mining Program which seeks to raise Serum TVL by offering grants to composing partners. There may be specific partnerships that are sensible to pursue for the purposes of creating buzz and inflows, but we’d be opposed to most grants where the primary value-add of the composing project is incrementing TVL.

We came to this conclusion because, while we were reviewing on the project-specific proposals, we kept honing in on issues arising from misalignment of interests. Questions of “how tightly does your protocol quote” or “what proportion of your liquidity is sent to Serum and why” are complex questions, but as @nkosi says, there is a simple solution: just reward based on volume.

For other Serum holders, even if you think raising TVL is an important priority, we think you should reconsider the budget of 5M SRM per 30d period. At $8/SRM, this annualizes to $480M per year, which will be over 20% APY if Serum is unable to raise TVL by 2.5B by doing this.

Overall: expending fewer resources on TVL will leave more resources to incentivize volume. We think it’s best for Serum holders to focus on that.