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

Nice work sir, thanks for fighting through

1 Like

It seems like you are consideing arbitraging between pools , so what if mev bot ‘buy first,sell later’,I don’t think it’s realistic and reasonable to limit trading behaviors,so there is mev profit captured by bot if a big buy order gets sandwiched on Curve.In this scenario, it doesn’t matter to make gear still be non-transferable despite you prevent arbitrage between pools.
Another issue is about the fee collected in fairtrading, your initial thought is put these profits into LM contract, but how much will the fee be within 14 days to support the 6-month LM? I personally prefer it be tranferable on trading stage.

@KEVINZ ::
The core idea is this: if we just set up a normal pool and made $gear transferable, sellers would start a bidding war on ETH protocol level to sell first and get a better price. The profits from those bidding wars would go to Flashbot operators.

The fair trading phase is designed to spread out these sell transactions over time and route these bidding wars profits to LPs in the process. E.g., you could pay 30% to push a flashbot bundle in a “normal pool”, or you can pay the same 30% here to sell earlier than others - the difference is where profits go.

Same goes for, ex. 15% fee - if you pay 15% to flashbots, you’ll get in close to a middle of the pile, which corresponds here to selling in the middle of the fair trading period.

The actual discount values, are of course negotiable, feel free to tell if they seem too low / too high compared to a typical flashbots premium

1 Like

@ 0xcider Thanks, I get what your point is,personally support the overall framework but some parameters should be adjusted, mainly on decreasing fee and fair trading duration(like 14% and 7 days )

1 Days for accepting $gear & $eth. 2+2 days sounds enough.
2-4 Min amounts for the cider contract: $gear and $eth. IIRC 2.5M USD of ETH allowed for a 25% dump of circulating supply? Sounds good enough, but dunno if that accounted for the tokens that will start to unvest at the 15th of December + unlocked tokens from LP farmers (if they’re unlocked). If LP farmers are unlocked, it’ll likely be easy to fill the $GEAR side.
5. Both 7 days and 14 days of fair trade duration sounds good. Those worried about tokens unvesting would probably prefer 7 days, so I’d go with that, even if it’s in the middle of Christmas season. Both durations should do enough of a good job to prevent MEV to be extracted
6. No opinion on selling fee % amount. Anything above 15% should likely do the trick.

Auto-returning funds in case of min-amount not achieved sounds good.
Fee on buying sounds silly. Can’t see the point in doing it.
I’m not a right-curvooor so I don’t have an opinion on curve pool parameters.

Great job on this. Fun times and I think this definitely improved the status quo for launching liquidity in a few things.


Problems persisting imo:

  1. I worry about the amount of sticky liquidity after FairTrade is over, even with seller fees being distributed to LPs. Harder to see the incentive to LP later on, but I don’t have a solution for this.
  2. It’s a bummer we’ll lose the Balancer incentives that were discussed previously.
1 Like

Hi 0xCider. Great plan if this can be all put into code. The multistage liquidity building process and letting gear dumpers exit early is very well thought out.

The only change that I can propose is to swap ETH for FraxBP as the pairing asset. I suggested this in the previous thread and will explain a bit more here.

GEAR will need 2 seperate LPs at some point for arbitrage purposes. In your current plan the idea is to use ETH in both the balancer and curve pools. Frax BP as the pairing asset for curve addresses the point you make in your plan that bribes are expensive and probably should not be used. I agree.

However if you use FraxBP, Frax as a protocol will do a few things:

  1. Frax will distribute the CRV+CVX revenue it earns back to the GEAR metapools proportional to the amount of total FRAXBP. This is “rev share” on the project level between DAOs.

The amount of incentives is proportional to the size of each metapool compared to the total fraxbp TVL. For example, if fraxbp has $1B TVL and a metapool has $20m TVL, then 1% of the fraxbp Votium bribes go to that metapool since $10m/$1B=1%

This leads into point 2

  1. Frax will use it’s entire CRV and CVX voting power to vote for gauges on both platforms. Frax will ensure gear gets a gauge and that it always recieves rewards.

  2. CVXGearFraxBP Lp will also get a gauge with Frax. This creates long term locked liquidity on a 1 year time frame. Mechanics can potentially be built into this release system to lock liquidity for 1 year initially to take part and ensure no one pulls LP after the full release of all tokens.

  3. Gear should commit to having it’s ETH pair LP on balancer. 500k of ETH can be set aside by the DAO to pair with GEAR at a later date.

Hope this gives a few ideas on why FraxBP pairing would be best for GEAR curve.

TLDR, my reasoning below (largely in agreement with GRIN)

  1. 2+2 good
  2. Modify variables to achieve a range of $0.0125-$0.0175
  3. DAO should supply $500k if ETH goal isn’t met.
  4. 7 day duration.
  5. 30%->0% good, too low will defeat the purpose of it.
  6. Failure → return funds is good
  7. I’m unsure on the buy fee → does removing it make it so ppl will once again flashbots to buy/sell? Ie, it seems like this should apply to both buy AND sell to have the desired effect.
  8. Curve params → mimic uni v2 as closely as possible.

Agreed on 2+2. Also agree that 7 days is fine, later is also new years anyway so yeah just do 7 days. Some people might be busy with family, other people might have days off, so kinda washes out. Flashbots MEV is like an instant thing, and given the main reason for this period is to replace that 7 days feels like plenty.

For people that wish to LP but don’t particularly want to sell, they can deposit GEAR and ETH - but that takes some capital. I’d like to see the maximum raised to 0.02 so people at least know if they LP there’s a chance they still sell at the most recent valuation. If we don’t hit the max that’s fine, we don’t need $6m of liquidity total so it’s OK to risk overshooting a tad IMO. And the use case of depositing GEAR AND WETH is still intact, which I think will provide a decent amount of ETH supply (plus the DAO $500k).

I like the justification for the 30% fee based on flashbots experience, so I’m fine keeping that. I’d rather have the first day be low activity then set the fee too low and end up not actually solving the flashbots issue (ie, too low of fee = potentially doesn’t actually solve intended problem).

Curve params: the doc reccomends a minimum A of 4000 but the sim shows that as very flat. Not sure II understand their reccomendation. I would like to immitate something like uni v2 as closely as possible, simply for LPer simplicity. This system is already new enough, no need to add brand new LPing curves into the mix with it.

Lastly I want to propose that if this is executed cider’d does get paid for this at whatever hourly rate our other devs are at (ie, basically considered a contributor for this effort).

I think the DAO could commit a small amount of GEAR for the first ~month. The intent being that during that month, we would pursue a crv/cvx gauge - if it takes a bit longer, that’s fine. Then once the gauge is live, we can do a proposal to decide if we should keep LMing, how much, and via what method (straight LM or bribes). Could also consider FBP at this time.

1 month of extra LM from the DAO, at 10% APR (a bit low cuz there will also be APR from the cider’d fees) on let’s say $3m of liquidity would be 0.0167% of supply, or 1.6m gear tokens. Extremely tiny, esp when compared to gnosis auction being ready to release 150m tokens on the market.

I don’t think set the fee still be 30% is appropriate for dealing with mev when the duration decreases to 7 days,seller needs to wait for 5 days before the discount drops below 10% , in this scenario,it is likely to see a significant increase in trading activity in the last two days of that duration,which may lead to bidding war again,2% daily decrease seems more smooth,just like initial idea, 30% with 14 days, lower the fee as decrease the duration.

Made a thread:

1 Like

Good morning, boys&girls.

I have shared code with developers and auditors, will also open it to you all in two days. I 'mcraping rusty parts and waiting for the auditor deadlines. Will keep you posted, moving fast ~~~

TLDR, my reasoning below (largely in agreement with GRIN)

  • 2+2 days seems to suffice every one, ok.
  • Here I see a lot of discussions. The numbers from ivangbi post are fine. Any one against-it? 0.015-0.30 looks to be supported by a few members, I don 't mind it very much. Ok?
  • Sell fee 25% as the fee start is enough, ok.
  • No more than 1 week for FairTrading, ok.
  • FairTrading fees go to LPs for 6 months duration.
  • Buy fee: no body seems to want care. Keep it at 0% = simpler logic?

I am observing consensus, glad. @ov3rkoalafied can report to me final numbers any time.

ideas

1) DAO capital

Suggest to purpose $ X K treasury regardless: either top up 0xcider (if < min) or in Balancer:

  1. I suggest for the DAO to supply up to 400 ETH of its own assets (20% of 0xcider min) in $eth in case the min range is not hit. In case the min is hit, the DAO does not need to do anything. Practically speaking, financial multisig should pre-sign the transaction to be ready during the last blocks, and withdraw the difference later. That is ok?

  2. In case 0xcider doesn 't need any capital (strictly = 0), DAO can make a GEAR/FRAXBP pool on Balancer (whatever asset the Balancer-Aura-Frax were offering the highest rewards for). This will strongen the relationship between DAO, and make liquidity deeper. More detailz:

2) Extra LM for LPs

ov3rkoalafied: I think the DAO could commit a small amount of GEAR for the first ~month.

Suggest doing it for 4 months (between 3 <> 6 in generalé) with open_end like the regular Liquidity Mining for passive suppliers. 1 month is too short, you will have to make a new program again soon after it. In the contrary, 4 months is enough for the market to stabilize and figure itself out. You can always vote and stop after ~2 months, easier rules.

1 month of extra LM from the DAO, at 10% APR (a bit low cuz there will also be APR from the cider’d fees) on let’s say $3m of liquidity would be 0.0167% of supply, or 1.6m gear tokens.
NEW: 4 months of extra LM from the DAO, at 10% APR (a bit low cuz there will also be APR from the cider’d fees) on let’s say $3m of liquidity would be 0.0668% of supply, or 6.68m $gears. Ok?

@ov3rkoalafied idea no2 needs a new proposal separately.

2 Likes

Perhaps this is an argument to change to 30%->0% with 3-4 days instead of 7-14 days for the fair trading period. @0xcider any reason for the fair trading period to be 2 weeks, or even 1 week? IMO 1 week is the longest it should be, but 4 days maybe OK? gets the “delay” down a bit more too. Curious what if any reasons are against it. Obviously if you make it too short then no one gets to participate, but I feel like anywhere 4-7 days might be OK.

So the remaining action items to finalize cider’d:

  1. finalize fee range [dao discussion]
  2. finalize fairtrading as 7 days, or potentially less [dao discussion]
  3. finalize 0% buy fee [dao discussion]
  4. write and launch proposal for additional light LMing program

Items 1-3 maybe should go to vote as a “finalizing” proposal should cider’d pass, before it launches (and none of those really affect code or audit right? or would any of those need to be figured out before dec 5th?)

EDIT: If we do a “finalizing” proposal, then I think that would be a good place to also propose payment for cider’d for his dev work.

4 day 's good. 7 day 's good. Less then 4 is too short. Instruct me what to hardcode.

Items 1-3 maybe should go to vote as a “finalizing” proposal

Yes, extra finalizing proposal with multiple choice?


In total, needed details & proposals for:

  1. 0xcider finalizing proposal [include all variables]. I will communicate on the audit prices and the name chosen before tomorrow. It’s likely ready sooner than I thought.
  2. LM proposal, this topic: [PRE-GIP] GEAR smol LM proposal - #3 by 0xcider
  3. DAO funds usage proposals: either top up 0xcider if the min is not reaching, or use $300,000 K in Baancer pool around middle of month January. Details are above.

Timing the market intensifies :innocent:

audit

Who is taking care of it - Chainsecurity :star_struck: Start: Dec5. End: by the end of next week Dec11.

@ov3rkoalafied costs are to be within 16,000 USDC or less. Please include in final cider remarks.

I shall be sharing code with everyone by tomorrow evening, final commit. Anyone can look then.

While audit is ongoing, I will finish a simple interface for the 0xcider stages. You will not require an interface though, it will be either calling 1 function (for $gears holders) on Etherscan - or directly send $eths to the contract. Easiest interaction ever, boys&girls!

timing

Monday Dec 12 is when multisig transactions can be prepared, while everyone is back to work. Should the audit find no major flaws, start of 0xcider can be scheduled as Dec13 Tuesday.

I will help instruct the multisig on what is needed, with the help of van0k if required.

This is manageable should there be no major flaws in code or multisig delays. Worst case scenario it would add a few days, but no more than a week. $gears shall all be liquid before end of Dec!

Reminder:

Early contributors are not able to claim their tokens before full transferability, because that would require vesting contracts to get “transferability” role. They are not getting it in my plan. No early member is able to take part in 0xcider, it 's only accessible to the community.

3 Likes

Almost final code-base. Make a freeze for auditors tomorrow.

2 Likes

Also in favour of 4 days being sufficient rather than 7 :handshake:

1 Like

4 days should be sufficient, more than enough time for people to do what they want.

The point of the fee is to turn MEV into long-form selling/buying. Liquidity can have a race to both buy and sell. If we only have a fee for selling GEAR, does that not create some artificial imbalanced buy pressure? We aren’t trying to be ichi here, don’t want to create a stairs up elevator down effect. I’m still leaning towards keeping the buy fee. The main reason I can see to remove it is if we are worried about demand. Ie, stairs up elevator down could be a problem in a bull market but in this more depressed and potentially more rational market it may not be a problem?