CIP-37: Increasing liquidity for COW token via a programmatic order and an FM-AMM

Governance Closed Proposals
, ,
42

It’s amazing to see so many great ideas here. Thanks in particular to @Yakitori and @cryptocondom for pushing the discussion in the first place. As both a COW holder and core team member – after reading the discussion here – I don’t think that I can support the original text of the proposal.

In examining the path forward, I think it might make sense to break out the key issues and discuss each of them separately. As I see it, we have:

  1. Should we test the FM-AMM? (It seems we agree that we should, but disagree on how to fund it.)
  2. Should we increase liquidity for COW? (Again, it seems we agree that we should, but disagree on how to do it.)
  3. Should we raise another round of funding for CoW DAO? (This point we’ve discussed less, though I imagine our collective answer would be something like “yes, if we can sell COW for a fair price”.)

Point 1 is probably the most urgent to discuss. As a next step we should source proposals for (a) how much we need to fund the FM-AMM, and (b) where the funds should come from. And agree on which is best. Many of you have made suggestions already, including @Yakitori and @koeppelmann, but we need to align to what, specifically, is most ideal.

Point 2 might need another full discussion of its own. It might be helpful for the core team (or anyone else here) to pull all of the ideas discussed here, evaluate them, and then present a plan to the DAO in another thread. This shouldn’t be too drawn-out, but we should thoughtfully consider each of the ideas proposed here.

Point 3 would benefit from analyzing the fair value of $COW. We have a number of investors in this thread that have presumably thought about that question a lot. It might be helpful to share some POVs so that we can ballpark a range to target. If comps say we should be $0.60, for example, then I agree it makes little sense to sell at $0.30 (either openly or privately). Bearing in mind that we can move faster with a higher burn, and we need to be a little conservative given the volatile nature of our industry.

We’re in it to win it! So we shouldn’t be overly-concerned by short-term thinking. But we should absolutely take this thoughtful governance discussion seriously and update the proposal(s) accordingly.

1 Like
43

I have nothing against scaling back the proposal and having another discussion down the road. But 1M COW is way too low: counting the time it takes to write a CIP draft, post it on the forum, discussing it, and then voting, it just give us 1 month before we have to discuss again how to fund FM-AMM. This is a significant overhead when instead we should be focused on building FMAMM (fyi, Cowswap development company employs only 25pp). Also FMAMM may not be fully finished by then.

I think @Anna’s number are in the right ballpark: 3.5 M would give us about 4 months of work. By then FMAMM should be finished, fully tested, and we can have an informed discussion on what to do next.

44

Strongly agree with CC and his well-argued post. COW is one of my bigger bags and it seems there are alternative options for increasing liquidity that won’t artificially constrain the token price.

2 Likes
45

Dear community,

It is great seeing this participation and interesting comments on this proposal.

As a participating member in the ecosystem would like to share some thoughts:

GnosisDAO is a key strategic partner, hence, in my view it is not in CoW DAO strategic roadmap a good option to sell GNO (as per @anna’s comments)

The AUM under Karpatkey report are actually 5.7M if we exclude GNO and CoW.

Lower if we consider that the Grants Program is requesting 0.7M USD and that the CIP for Protocol development is ending.

There is a clear consensus to quickly test the innovative approach of FM-AMM as the DAO may diversify towards a 3rd product and enter the DeX market as a clear competitor to current solutions.

There is a sentiment to not sell CoW in the market as it is currently under valued given it’s fundamentals / comparables.

As such, there’re not that many ETH reserves available for testing of the FM-AMM. Before touching those we should consider:

  • Leveraging existing ETH decreases runway and yield returns
  • MEV Blocker rebate related revenue in ETH will reduce as the protocol moves to driver-solver colocation

If the community is still in favour of leveraging existing ETH, following could be proposed:

  • That the Treasury team frees ETH holdings from the wsETH / GNO pool - counteracted by raising validators with idle GNO that can increase yield.
    ---- Frees 366k USD in ETH
  • Liquidate the rETH positsion
    ---- Frees 222k USD in ETH

This would allow the DAO to launch the current product in 50/50 setup of around 1.2M total TVL. Whether this liquidity would be sufficient to test the potential behind FM-AMM is best judged by @AndreaC ).

The advantage of taking this route would be deeper funding of the AMM from it’s inception. A downside would be a more manual process that reduces the positive impact of showcasing a replicable, automated process for other DAOs.

@Yakitori , the Karpatkey report is a truthful image of the DAO’s funds, as there as been relevant investment in the ecossystem through Development Services CIPs, Grants DAO Funding, Solver Bond funding and other Protocol operational expenses. Nonetheless, it excludes a very significative CoW holdings that are managed directly by CoW DAO.

1 Like
46

Hi @notsoformal, thanks for the insightful analysis!

The advantage of taking this route [leveraging existing ETH] would be deeper funding of the AMM from it’s inception

IMO, if we were to go this route and seed the FM-AMM with existing ETH, we would anyway increase its TVL slowly: starting simple, improving the product gradually as its TVL increases, pausing it if we feel that there are issues to be ironed out, and so on. I think this iterative and incremental approach will get us faster to the goal, compared to working on the smart contract for several months, trying to get it perfect, and then putting more than USD 1.2 million into it.

Also, most likely, we would still need to deploy additional resources: there are currently approximately 10 M TVL in various COW pools, and adding USD 1.2 M (with a better mechanism) is unlikely to change the experience of COW traders drastically.

For these reasons, I still believe it is better to start by deploying additional resources (perhaps much fewer than initially proposed) and then move existing ETH into FM-AMM at the end of the experiment. As you mentioned, choosing this route also has the advantage of “showcasing a replicable, automated process for other DAOs.”

Finally, thanks for pointing out very clearly which resources are, potentially, already available!

47

Well from those 10M, ~6M are allocated in the Balancer COW/GNO pool, which we know already is quite inefficient, so by reallocating part of those GNO (staking) and COW in more efficient WETH pools (e.g Univ3, FM-AMM) would make for it, and with a smaller TVL amount (e.g 2M we could already be seeing some notable results)

Is quite hard to pass a voting process to sell tokens, so what I would propose is to match any amount of sold COW tokens with a equivalent in liquidity being added to the pool, mitigating any effect, with the difference that liquidity would be added ‘‘today’’, while sells would occur over a period of time.

For example, adding 588K USD in liquidity (based on @notsoformal proposal)+ the equivalent in COW tokens, would ‘‘free’’ the DAO to sell another 588k USD in COW tokens for a period of 12 months, ideally through OTC strategic deals (it would be an amazing opportunity to get some high-quality investors onboard) but also could be on the market.

I think you were aiming for 1.5M, so why not adding 750k USD in liquidity, and creating a structure or more detailed proposal with:

  • Where we are getting those 750k USD from (I think most agree where the first 588k will come from, which is a big step forward)

  • How the equivalent 750k USD in COW tokens would be deployed on the market and for how long? Imo OTC is the best, since you not only get liqudity in, but added value (e.g a Wintermute or Cumberland onboard would be sweet!)

  • Where are we getting the COW tokens to match that 1.5M in liquidity from, although I think most will agree on taking them from the GNO/COW pool!

@cryptocondom curious to hear your thoughts on it

1 Like
48

I do not believe adding 10m liquidity for sell side pressure on COW in the cowAMM is appropriate. Agree with re-allocation of existing pools to pair liquidity and use a much smaller allowance of COW over time to “try out” the tech.
If the dao need funds prior to their raise in 2H, it seems more reasonable to do vested OTCs than sell tokens with one sided liquidity.

49

Oh I didn’t know that the DAO was raising funds in 2h, is this open to everyone or just selected strategic investors? I would be more than happy to contribute.

On the other side with my proposal only 2.2M COW tokens (potentially reduced if the price of COW increases) would enter ‘‘circulation’’ and be sold over a period of 12 months or so, but in exchange we would get 750k USD of fresh liquidity TODAY coming from treasury assets (mainly exisitng ETH), everything else would be taken/generated from re-allocation of assets and liquidation of assets as proposed by @notsoformal in order to reach the 1.5M TVL that @AndreaC pointed out as optimal to test the effect of the FM-AMM

51

Dear all,
Thanks again for all your comments and suggestions! Let me make two points:

  • Foundraising: CoWDAO will need to fundraise at some point. The amounts involved will be much larger than those discussed here and will require long discussions with potential investors. So, whereas the proposal under discussion “looks like” fundraising because it sells COW for ETH, I think it should be kept separate from the fundraising. For example, here, the main objective is to move fast and launch a new product ASAP; in the fundraising, the main objectives will be financial.
  • what funds to use first we all seem to agree that we should use a mix of existing liquidity and new liquidity. With respect to the new liquidity, the idea is to sell between 2.2 M COW (as per @iznogud07 calculations) to 3.5 M calculations (as per @Anna calculations). With respect to existing liquidity, we are looking at around 600k USDC currently held in ETH (per @notsoformal calculations). The main disagreement seems to be on the timing: whether to start with new liquidity and then use existing liquidity or vice versa.

Again, the goal here is to launch a new product ASAP. The faster and most flexible way to build and progressively “try out” CoWAMM (as @cryptocondom said) is to start by using a dedicated allowance and then redeploying existing liquidity.

As for how many COW should be sold during this initial period: honestly, the various proposals are not dramatically different from each other. In my view, the highest amounts give more time to the development company to focus on building the product before reopening the discussion. Because time is also scarce, I would prefer larger numbers. But, honestly, I think selling 3M COW in the initial phase should work.

52

As far as we sell those 3M COW (at current prices) in a 9-12 months period but we add equivalent ~750K-800k in liquidity from ETH re-allocation to the COW/ETH pair (in the FM-AMM pool) between ‘now’ and 3 months you have my vote!.

Note: As mentioned before COW should be taken from the GNO/COW pool (which currently has ~1.5M USD in COW tokens), not adding new COW tokens in circulation, so we can slowly reduce the impact of that pool and put that GNO to generate some real revenue :slight_smile: ).

53

I think it is a great suggestion for the second part of FM-AMM: we can supplement the liquidity of FM-AMM by liquidating some ETH and also using COW from the COW/GNO pool. It also goes well with @koeppelmann proposal of creating a COW/GNO FM-AMM where CowSwap contributes its GNO and Gnosis contributes its COW.

For the build-up phase: ultimately, FM-AMM will be judged by whether it improves the market participants’ experience relative to other mechanisms. If we seed it using COW from the GNO/COW pool, then we are improving the liquidity of COW on one side (by creating FM-AMM) while making it worse on the other side (by removing liquidity from the GNO/COW pool). It is also possible that FM-AMM development is slower than expected and that, for a period, FM-AMM does not perform as it should. If that happens, the liquidity in the COW market may be significantly reduced for a prolonged period. Overall, although I agree with the benefit you point out, I think that using COW from the COW/GNO pool comes with non-negligible risks.

54

It is great to see the large amount of community engagement and thoughtful suggestions to bring this proposal to fruition.

We see two primary questions to address:

  1. How much do we want to allocate for the FM-AMM COW/ETH pool?
  2. How do we want to source that liquidity?

On 1) We believe starting with a smaller allocation in the early phase to limit risk is the right approach, however we should be ready to scale up quickly. Assuming we didn’t miss this information elsewhere, it would be great if the team could provide context on what a reasonable initial liquidity range might be.

If a pool with 3.5M COW and an equivalent amount of ETH is what we eventually envision as sufficient liquidity when scaling up, why not allocate some of the existing ETH, e.g. equivalent of 1.5M COW, from the treasury funds and in parallel have a TWAP over e.g. 6 months swapping the remaining 2M COW for ETH. This should give enough liquidity for the initial phase and to scale up in the future (in the above example, we’d have 70-80% of the target size after 3-4 months).

We don’t have a strong opinion on the exact split of what amount to source from existing treasury funds vs. TWAP - the daily sell pressure of 2M COW over 6 months from above example equates to less than 1% of daily trading volume currently, so we don’t expect a feasible impact. What might be more impactful here is to showcase this funding mechanism (bootstrapping liquidity via TWAP to deploy a DAO owned FM-AMM pool).

Would love to hear any thoughts on the above!

55

Thank you all for contributing these insightful comments! By now, everyone has had time to read the proposal, comment on it, read the other comments, and form an opinion. We are, therefore, ready to move to the voting phase. To address the concerns raised (both here and via other channels), the proposal on which the DAO will vote is the same overall scheme described in the initial post with three changes:

  • slowly selling 2.5 M COW over a period of 4 months to create liquidity for FM-AMM, (vs 5M in the initial proposal)
  • complement the above with some of the ETH already in the DAOs treasury
  • Remove any reference to the price floor. Note that, for obvious reasons, the TWAP order will have a slippage tolerance (which will be chosen depending on market conditions). So, if the market moves suddenly in a negative direction, the TWAP order will not execute.

The proposal will be on Snapshot in the text in a few days. Please monitor the usual channels for more info. And remember to vote!

2 Likes
56

UPDATE: the proposal is now on snapshot. You can vote here: Snapshot

1 Like
57

Agreed, LGE can be done and benefit holders. Selling that much in these mcap conditions is not wise. Get rid of GNO which literally add no value to this protocol instead of own token.

58

Translation: We are already well-funded from Gnosis, Gnosis is our main project, we don’t care about Cow at all. We will sell for “liquidity”
Why do you need liquidty at all? Last 24h, only 6 transactions made, 2buys-4sells

1 Like
59

These are the end times if you did not notice. You can only longterm hold your head between your hands

60

image
image1037×690 52.4 KB

A little update on how it’s going. Not sure the DAO is getting a great deal here.

61

Hi @Yakitori

Apologies for not providing updates. The CoWAMM COW/WETH pool is here. As you can see, CoWAMM was created with a two-sided contribution (117,134.93 COW and 18.19 ETH, which had equal value at the time). Then, we focused on fixing bugs and improving its functioning. We manually added another 38.9k CoW a few days ago (on 2024/03/28, see here). The module for selling a small amount of COW daily is not yet active. AMM now holds 133,2700.

So, over the period you are looking at, the AMM bought more COW than it sold. This makes sense because COW price dropped relative to ETH and COWAMM runs a rebalancing strategy: when prices move, it always sells the most expensive and purchases the cheapest so to keep the TVL equally shared between the two assets. It is indeed unfortunate that the price of COW dropped relative to ETH, but we don’t think COWAMM has anything to do with it. If anything, we suspect the price drop would have been larger if CoWAMM didn’t absorb some of the sell pressure.

3 Likes
62

Dear all,
we completed an evaluation of CoW AMM so far. TlDR is: CoW AMM is still in beta, and there is ample room to increase the return for its LPs further. Despite this, depending on the method and the pool, the returns to providing liquidity to CoW AMMs are similar to or significantly better (3% extra per month) than the best alternative for passive LPs. Also relevant for the discussion here: we decided to pause the stream of COW tokens to the initial COW/WETH CoW AMM pool. You can find all details here:

1 Like