[GIP-31/1] GEAR Strategy Cider’ed Liquidity

Proposals
63

Still catching up on everything, so this is a comment purely on some systematic stuff related to sell pressure. Does not address all comments, as there are plenty of other great considerations between both proposals (simplicity, regulatory uncertainty, timing etc)

I largely agree with @grin on the questions to solve. The buffer will need more eventually, but a fraction of the 1.5% of tokens intended for the initial auction would solve that. And the 500k from the strategic round 2 should be used for POL for a baseline of liquidity.(begin to seed the ETH side). I would also like to proceed with balancer as I’m not sure that curve v2 pool parameters are well understood (also hard to set custom liquidity parameters for a token we don’t really have price info on, right?), and we have some initial balancer incentives lined up, plus a more effective bribe multiplier on balancer (which I hazard to guess will close the gap over time, but not quickly).

I will note, in most outcomes I don’t think it matters which proposal we use. If no one wants to buy GEAR, then both proposals end up in a similar place (GIP32 doesn’t raise much liquidity and then has to liquidity mine, thru straight up LMing or thru cider’s proposal, or another approach).

  1. Ie, if 1b tokens are sold by sellers and no one wants to buy em, doesn’t really matter what token launch proposal is chosen. The token price will drop, liquidity mining will be less effective.
  2. If 250m tokens are sold and ppl really really want all 1b, then it doesn’t really matter if we do gnosis auction or not.
  3. If 250m tokens are sold and people really only want 150m, then bad for token. In this scenario, removing 150m of sell pressure (by using Cider’s method) is better.

Note, what is being discussed here is token price, but purely from the perspective of how to minimize token sell pressure created by the DAO while achieving the goals/reasonings for unlocking the token and liquidity mining and allowing the market to buy/sell as they deem fit. Ie, goal is not “maximize token price” (which is why ideas like “lock everyone forever, burn GEAR” etc are not considered). I do want to note that gnosis auction does to some degree guarantee that demand is leftover due to the nature of how it prices the asset, but Cider’s method definitely helps a lot in a specific window of demand/supply - and that window is more likely in a bear market.

Another item I don’t think has been mentioned is that bribing. If we do liquidity mining, does take at least bi-weekly multisig transactions. Do we have a multisigner willing to take responsibility in queing those? Or a dev willing to build a tool to user a keeper or something to send the necessary GEAR each epoch?

I also want to toss out the idea of a public gnosis auction (take with grain of salt for now just documenting), ie anyone can supply the GEAR that currently owns GEAR that isn’t vesting. Would require a custom contract that at a minimum has to allow anyone to supply. Would be cool if individuals could set a minimum price. Combine that with the uniswap LPing (buyers recieve an LP token rather than GEAR, reserve a spread to pay LPers). Primary benefit here is that ppl are more likely to supply because they aren’t worried about underselling. Granted, if they are correct and they are underselling then most concerns are solved anyway. Ie, this may be a problem that solves itself, but it would make it easier to attract GEAR suppliers if there is worry that ppl that wish to sell think $0.015 is undervalued.

64

gnosis auction prices via the intersection of supply and demand curve, meaning there will be by definition a large chunk of bids that do not get filled, below the final settlement price.

1.5% supply would be added via gnosis auction, not 15%

In general, cider’s method still is less dilutive than the gnosis auction in terms of how much demand is absorbed.

65

This is a good study, but I don’t agree with using the same total liquidity for the cider’d raise. I believe you need to account for mercenary liquidity being far less sticky than gnosis raised liquidity. I would like to see what the result of the cider’d study is, assuming that 50% of LPers pull out pretty immediately (ie, take half of POOL INITIAL GEAR and POOL INITIAL $) and see what price that comes out to. That matches what is estimated in the original proposal.

If a half of members leave this setup and only 50% of the original LP remains, that is 10% APR for a year (or 20% annualized if the duration is 6 months). Solid enough? Depends on the secondary market price of $gears of course, can be more or less.

And as a note I believe 20-25% is a fair estimate of the APR that LPers would settle on, looking at various other defi gov token pools.

67

Grazie mile for the great responses. I 've working with help of fellow hodlers on addressing the latest concepts (price discovery inside Cider’ed Liquidity, and dumpLPunwrap protection). Will share tomorrow. Audit is managable as well - I 'm getting the final numbers.

And as a note I believe 20-25% is a fair estimate of the APR that LPers would settle on

Made solution for this part as well.

I don’t know how koala 'll make the vote. Please either include Cider’ed Liquidity as an option there, or make it YES-NO vote to start with, and concrete plan to be voted on this week. Maybe?

68

Yeah not fully decided but whatever it is the official vote will definitely not ignore this proposal.

I’m leaning towards something more like this right now, curious on your thoughts: [GIP-32] GEAR Liquidity Raise and Enabling of Token Transferability - #18 by ov3rkoalafied

(Read that post and the two comments)

Also how does LM work in this scenario? How do we get a curve or a balancer gauge in time for liquidity to go live? Ie if we don’t have rewards everyone will pull out of LP until we do.

Ie in either scenario, cider, I think you need to build like a 1 month lm contract to use until we secure a gauge on balancer (if you are willing)

69

Zoo keeper, at your service!

  1. 1/2 of GEAR is sold via gnosis auction, relevant parameters from this proposal (need to confirm it is in fact half - depends if bids over clearing price get extra tokens or a rebate)

The suggestion you made is practically what cider contract should be able to do, without create complexity. Some send $gears, some send $eths. Complexity much reduced compared to above. We are on the same page then, technical implementation is of no concern for me here.

Also how does LM work in this scenario?

I 'm working on an article to explain it, here is what was shared as concept so far: Users can pull up CurveMEVSnatcher interface and sell GEAR in Curve Pool. At the same time, at the time of this operation, the user is charged fee = 50% * (transferabilityStart - currentBlock) / (transferabilityStart - auctionEnd) in gear. This is to be used for 2 weeks, generating fees and rewards for LPs from day 0 and for the months to come. How long? Can be decided as well.

How do we get a curve or a balancer gauge in time for liquidity to go live?

I need help coordinating this, but it should not be hard. The LM will be threefold: these rewards as per above, the GEAR extra if the DAO votes for those, the fees from these crazy swaps at the start, and then DAO can also engage in brining or not, that’s to be decided. I 'm not professional with balancer, so I have to use Curve. We “lose” on Balancer extra votes, but Curve-Convex-Frax teams can help in the future. This sounds somewhat vague, but I am sure this is doable to fix.

Ie in either scenario, cider, I think you need to build like a 1 month lm contract

I agree. Easy and cheap to do. Doesn’t hurt!

70

I like the idea of removing the Gnosis Auction since it will probably result in -EV performance because;

  • DAO is doing a public sale
  • Forced to set a price for $GEAR (horrendous idea imo)
  • Removes all demand for token
  • creates a liquidity cushion which can actually result in auction participants losing money right off the bat (creating bad publicity)

However I think using a Curve pool to jump start liquidity is also a bad idea for the following reasons:

  • Curve emissions are hard to compete with
  • Bribes for CRV emissions will put GEAR tokens in the hands of veCRV holders and CVX holders rather than GEAR LPs
  • Creating initial liquidity with curve pools can have really bad UX for initial depositors who are tempted to deposit one sided while liquidity is low, causing them to incur unnecessary slippage
  • Curve team are machines, but they are really unresponsive at times and hard to collaborate with for things like spinning up a new gauge and making UI changes if we opt for direct GEAR incentives for the curve pool

My suggestion is to simply stick with the Balancer pool, but remove the Gnosis Auction idea - which is horrendously mid IQ anyway - and proceed with an initial GEAR liquidity mining strategy with plans to propose a veBAL gauge. The Balancer team is pretty quick with these integrations and provide more support where as with Curve, everything gauge related basically requires a lot of on-chain governance and help from veCRV holders.

So in general I like @0xcider’s idea but just think that a Curve pool isn’t appropriate for spinning up initial liquidity from the community.

Across protocol is launching $ACX tomorrow and they are going with the same strategy I mentioned above:

71

[GIP-32] GEAR Liquidity Raise and Enabling of Token Transferability - #21 by ov3rkoalafied reasons why gnosis auction was not mid IQ when initially proposed but probably is now. Point 2 (forced to set price) doesn’t really apply though, if anything cider’d method needs more price “setting” due to having to set a range and a max price (gnosis auction determines the max price itself).

Your first two points about curve also apply to aura/balancer, right? Since we would use bribes for both. Ie just sub CRV or CVX with BAL or CRV and it’s the same issue?

For #4, but we will have our own staking contract it seems until a gauge goes live. Curve has also reached out from what I understand, so they have interest here. Ultimately I think if devs were familiar with balancer (cider, plus a core gb dev I believe agreed to help cider out if DAO chooses this option) then I’d push for it more. But I don’t know if the extra efficiency and partnership is worth blowback from further delay and more auditing necessary.

Also initial liquidity is created by the contract, thus solving the issue of initial LP slippages. There will be plenty of liquidity before people can interact on an individual level, minus how the contract handles single-sided withdraws and adds (which tbh should just be the same as withdrawing liquidity balanced then dumping half?)

Also the ACX method means ACX will have some super volatile price as people start to LP. It’s a very simple approach, but no buyers are paired with sellers and that means instant volatility and race to buy or sell. I don’t think it’s a great experience and it also doesn’t get us the extra GEAR to LM that cider’s approach does.

72

Cider, can you please update your proposal to reflect what you think is the best option for the undecided items below? I added my thoughts. Then I believe it would be ready for a vote.

  1. I think ETH is most popular as the counter asset. I’ve outlined my reasons before as to why I think so.
  2. Extra buffer is good, I believe 10% is fine. I also don’t think there is any advantage on a higher initial penalty.
  3. I’m not sure what the range should be here. I think 150m makes sense as the low point, if ppl believe they can buy lower than the recent raises/LMIng program estimate then they would just wait for live liquidity. I’m debating on 250m or 300m for the high point. I think you have to do 250m, as buyers may not step in knowing that their max price is $0.025 should others deposit ETH after them, and idk that 250m vs 300m max will make a huge difference in if ppl supply GEAR or not. Additionally, this puts the current FDV of $200m right in the middle of the range, meaning an “expected” result could be half of the expected liquidity, which I think would be adequate.
  4. Once LM kicks in and users are unlocked, we may see more liquidity adds (to counteract any liquidity withdrawals). So I think for now we can leave out the DAO funds for liquidity, address that if liquidity seems too low after this all unfolds. This could even be a proposal during the 1 week time where people are able to withdraw.
  5. Need to reflect that initial LP tokens can optionally be staked in an LM contract. I would say that the 10% of rewards should be used for 6 months of LM, so 10/6 would be used for the LM contract (gives us 1 month to get a gauge on convex/crv).

Basically get these specifics in the actual proposal text and it’s good to go, feel free to counter on any of these.

Then I believe ranked choice voting would be best, with options as Gnosis Auction, reasoning here: Discord

73

Afaik Balancer bribes are far less competitive than Curve bribes and trying to prevent volatility is not the DAOs job imo. The volatility will even out as liquidity is added and also some volatility is beneficial to LPs.

Overall I think there is too much effort going into a highly engineered launch tbh, and my alternative proposal would be to just forget all this and make the token transferrable without any liquidity mining. The result would be that ppl will just add some onesided UniV3 liquidity and buyers will be matched with sellers based on natural market dynamics without any overhead. Meanwhile GearBox governance can come up with a liquidity mining solution that won’t have to worry about price settings and have to rush an audit or delay the launch.

2 Likes
74

100%. Lets just vote to make the token transferrable. Everything else will get worked out with time

2 Likes
75

How do you propose LMing v3? I’m only aware of arrakis right now that works, but that would require forcing users to use a specific liquidity range that we determine. Which once again makes it helpful to have less volatility, so we have a better idea on price before opening it up. The reason low liquidity and volatility are not great is because it can make the LMing program ineffective. Hence finding a solution that pairs buyers with sellers in a way that aids in some pre-market price determination and somewhat deep liquidity.

Balancer bribes are a bit better than curve, it’s not a massive difference though. We could do balancer with a delay, or we could do curve with very little delay given cider and the gb devs already are familiar with curve v2. That’s the tradeoff as I understand.

76

We would do liquidity mining where ever the DAO chooses later on, it wouldn’t need to be on Uniswap. Most of the liquidity would simply move to the LM venue and perhaps some will stay, which is actually a good thing because having some UniV3 liquidity is good for discovery since Curve/Balancer pools are not integrated with most charting tools.

77

Aside from balancer being less familiar than Curve, their “internal balances” feature may pose an issue, since it would allow people to trade ETH => GEAR AND back afterwards even before transferability. While this would technically only allow them to push GEAR price in only one direction, they could try to sandwich GEAR dumpers or do some other attack I haven’t figured out yet

This is also somewhat possible with Curve by depositing one-sided, but can’t withdraw or trade back before transferability, so there it’s not an issue

78

I’m 50/50 on Cider’s proposal now… perhaps I was too harsh in the beginning. It has some good points, and I’m open to learning more when the vote goes live

79

At the time of writing, the vote is not posted yet, therefore, I ‘m not aware how a koala will phrase it. I observed different proposals being suggested, summarizing to:

  1. Don’t unlock. This should be a voting option, but doesn’t seem like it will pass.
  2. Just unlock, do nothing. Fwiw, this is lazy. Millie post here. You can argue for or against there on forum. I already explained above why I believe we can do better.
  3. Unlock and do normal LM. Fwiw, this is lazy but ok. Should be a voting option, but wouldn’t be my favorite. Again, we can do better. Millie post here.
  4. Gnosis Auction from the DAO funds. Everyone is against it now. Original proposal.
  5. Gnosis Auction from the community funds. Approximate post here. This concept is similar to my 0xcider model, but technically worse. In case you were a fan of the concept “sell community tokens at GA” - you are factually supporting 0xcider.
  6. 0xcider model. Included price discovery mechanism; early swap-trade FairTrading concept, and reduced complexity for participants.

6 choices for snapshot. Factually, 2-6 are to be summed up into total votes. 1 wins only if 1 > sum of all the other. Currently, not seeing this happen. Alternatively, as for the choice between 2-6, any one with the most votes wins.

That ‘s the best approach isn’t it? @ov3rkoalafied?

80

ser, you can write 10000 lines of code and you cannot stop a token from dumping lol.

It has nothing to do with lazy, if you are worried about the token being dumped then the best thing to do is not have a ton of liquidity readily available for ppl to exit on. You seem obsessed with controlling the way the token changes hands and it’s honestly bizarre… I’ve never seen a DAO so concerned with how the token trades post transferability

2 Likes
81

Give me 21 minute to fix the post and final concept. If you still believe the idea is bad, the voting will decide the plan eventually. I have superbig respect for you, so hope you get to liking it!

82

I should also mention that it might not be a great idea to do a number of proposals if the bulk of the work behind your plan has been done. Might as well proceed with that plan in that case

1 Like
83

It is up to DAO voting eventually, I could be wrong with these ideaz.