[RFC] Alchemix veCRV whitelist & Recompensation Request

Proposals
#1

Summary

Alchemix is requesting feedback on this draft proposal, in the context of the alETH/ETH pool exploit. It is anticipated affected LPers of the July Vyper exploit will be receiving recompensation in the form of 1-year vested CRV tokens. Alchemix is requesting to instead receive these tokens up front, as well as a veCRV whitelist for its treasury, provided all recompensation CRV is max-locked in a mix of veCRV and sdCRV. Alchemix is strongly aligned and dependent on Curve as the home of liquidity for alETH and alUSD and their respective AMOs. Alchemix is further aligned through its significant holdings of CVX and sdCRV.

Abstract

Alchemix is requesting a veCRV whitelist for its multisig treasury, 0x9e2b6378ee8ad2a4a95fe481d63caba8fb0ebbf9, and is requesting to receive the expected CRV recompensation for the Vyper exploit up front, with the stipulation it is max-locked in a mix of veCRV and sdCRV.

Motivation:

It is the understanding of the Alchemix team that Curve intends to refund LPers through a 1 year vest of CRV because the necessary amount of CRV to refund LPers of all affected pools cannot be released in liquid form. Alchemix is proposing to receive the entire refund up front, with the stipulation that it is locked in a mixture of sdCRV and max-locked veCRV.

The benefit to Curve DAO is that CRV allocated to Alchemix would be unable to enter circulation for at least 4 years, if ever. Additionally, creating a minimum 4-year locked position instead of vested CRV incentivizes Alchemix to continue to incentivize alAsset liquidity on Curve - an alignment Alchemix is happy to continue.

The benefit to Alchemix is the flexibility of the veCRV whitelist in that Alchemix does not have to rely fully on liquid lockers. Additionally, this allows Alchemix to reback alETH swiftly by selling portions of the current CVX and sdCRV positions in the treasury, without zeroing out its voting power for a significant period of time (which would reduce alignment).

It is understood by Alchemix that Curve wishes to make exploit-affected LPers whole, and due to the size of the necessary funds, the funds cannot be distributed as liquid/unlocked CRV. Alchemix is requesting flexibility in how the CRV is used (the same flexibility we would have if it were possible to refund LPers up-front), so long as it meets the primary goal of Curve DAO that the funds are limited in their ability to circulate. This proposal requests that flexibility in return for greatly decreasing the chance that these tokens ever enter the circulating supply.

The precedent here is JPEG’d whitelist request. veCRV whitelist request for JPEG'd DAO. The only difference is that Alchemix would lock in sdCRV and veCRV, instead of only veCRV, but the net result - the entire refund being locked for 4+ years - is the same.

Specification (Note: Proposal is currently a request for comment).

Whitelist veCRV lockup for the Alchemix treasury 0x9e2b6378ee8ad2a4a95fe481d63caba8fb0ebbf9
Grants council to send the Alchemix AMO portion of CRV recompensation to a Curve DAO-controlled wallet. There will then be an on chain vote to disperse this CRV to the Alchemix multisig treasury wallet. The received CRV would all be max-locked in veCRV and sdCRV by Alchemix.

FAQ - Is Alchemix launching a wrapped veCRV derivative?:

No. Alchemix has considered alCRV or alCVX in the past, but these are infeasible with the current codebase/economic model and would also require whitelisting of a separate contract specifically for alCRV if it ever were pursued.

For:

The benefit to Curve is that the portion of CRV used to compensate Alchemix will be out of circulation for at least 4 years, instead of vested over the course of 1 year. For Alchemix, dependency on liquid lockers is reduced and alETH can be rebacked in much less than 1 year without decimating Alchemix’s Curve voting power, which maintains Alchemix’s alignment with Curve.

Against:

Alchemix is requesting special treatment as an affected LPer of the July Vyper exploit.

  • (temp check) FOR
  • (temp check) AGAINST

0 voters

4 Likes
#2

I think that this proposal makes sense.
For me CRV locked at perpetuity > CRV Locked for 4 years > CRV vesting over 1 year, so this proposal seems to be an improvement compared to the initial idea to vest over 1 year.

A few question marks a bit indirectly related with this proposal:
1/ Is there a scenario where alETH holders could accept a partial lock in the form of a debt token that repays itself with aleth backing yield or something like that? (users who don’t want that can redeem at backing rate, benefiting users who accept the debt token, etc)
2/ I understand that most of the gap to be closed comes from Coinbase, and that Alchemix is not able to press charges against them. I am sure he already did but would Michael have a connection with Coinbase that could help get this money back? How could we do to maybe help Alchemix in this legal process?
3/ I think it needs to be mentioned in the proposal that should Coinbase hand back the stollen funds after CRV have been given to Alchemix, these ETH need to go to the Curve DAO and be used to buyback CRV.

3 Likes
#3

Thanks for the comments! Before answering, I do want to clarify that this request is specifically for what the Alchemix AMO lost, which was about 1/3rd of the pool. This is not a request on behalf of all alETH/ETH LPers. As it currently stands, its expected they will recieve their pro-rata refund of what was recovered as well as 1-year vested CRV tokens for the rest.

  1. Perhaps, and we did brainstorm some similar options, but it is quite a bit more technically complex whereas this option is pretty simple and doesn’t require specialty contracts / ongoing redemptions.

  2. Only about 13% of what is left to be closed comes from Coinbase. We are not giving up on that, but it doesn’t make sense to hold up the rest of the process for something with such an uncertain timeline / result.

  3. Yes, I agree that if CRV makes LPers whole now, and then Coinbase later returns funds, those funds would likely belong to Curve DAO. The only scenario this wouldn’t be true is if LPers could return the vested CRV and instead recieve the Coinbase ETH. It gets messy because we don’t know at what CRV/ETH price the recompenstation will be paid, and the price will not be constant. So I agree with the sentiment, but the details are extremely dependent on the state of the market at the time the funds are returned. For example - if everyone got vested CRV/ETH, but CRV/ETH bled out over a full year, then no one would actually be made whole. So they may feel it is unfair if Curve got all of the ETH from Coinbase.

3 Likes
#4

Dont want to sound bias or anything but I think anything up front should only be vecrv. If you want to lock as sdcrv, and nothing wrong with that, then accept it as vested crv and lock as you go. Up front as a liquid token which means can be sold off crv, and you state that selling sdcrv is quite the possibility which means you would be selling the should-have-been-vested/locked crv. Due to the size and other payments, the DAO should see this happen over time, as stated in this proposal itself. “oh but c2tp the received crv is LOCKED, whats being sold is DIFFERENT crv.” well…same result though. its crv being sold from something that didnt vest.

again, obviously coming from me this may sound bias but, I really think anything up front here should just be vecrv. youll still get the sdcrv as you vest. (same would apply to anyone wanting to get cvxcrv or ycrv up front)

All for the whitelist etc.

2 Likes
#5

Hmm but in the eyes of CurveDAO, CRV is not being sold. It is being locked in veCRV. Additionally, what Alchemix would be selling is sdCRV we already own. None of the sdCRV obtained from Curve DAO is being sold.

Ie, in either scenario we need to sell sdCRV or CRV. An unfortunate scenario of Curve DAO refunded vested CRV is that if the price of CRV/ETH drops, LPers won’t even be made whole. With that in mind, we are forced into a situation where we either wait a year and hope CRV/ETH price stays the same or goes up, or we “lock in” the price by selling our sdCRV which is slowly replenished. So essentially, either way we are likely going to end up selling the sdCRV we own.

In the linear vest scenario, we would sell a portion of the sdCRV we own equivalent to the refund and potentially even more CRV or sdCRV as needed to reback alETH, depending on CRV/ETH volatility.

In the scenario outlined in this draft proposal, we would sell sdCRV equivalent to what we receive.

So our proposal compared to the linear vest means:

  1. alETH is rebacked more quickly
  2. The refund is locked for likely >> 4 years rather than sold
  3. Either an equivalent or lesser amount of combined sdCRV+CRV is sold.
  4. Nothing obtained from CurveDAO is being sold, only sdCRV assets Alchemix already owns

This same option could be used for cvxcrv, ycrv, etc. It means a temporary depeg pressure on that asset for more long-term CRV under its control. If there were concerns over the bias, we could potentially route through multiple protocols, but then we would have to sell what we actually lock rather than selling only our own sdCRV.

@C2tP

2 Likes
#6

by transitive property, its the same thing. “we need to sell some sdcrv, so please give us some sdcrv”

Others will have to sell their vested crv over time if they need to, but special rules for you to have it all up front?

2 Likes
#7

Also just to be clear here, I’m up for contemplating this further. I’m just writing my first impression of what I thought when reading this. Feel like vecrv is the proper way to do things if the request is to have it all up front.

2 Likes
#8

I think it’s two separate things:

  1. We need to sell our sdCRV regardless. It doesn’t make sense to wait 1 year for CRV to vest. Waiting 1 year would mean a likely unusable alETH product for much longer chunk of time, as well as the potential to need to use the treasury to make alETH whole if CRV/ETH average price over that year is less than whatever the refund price is. The unfortunate historical truth is that with any defi coin with emissions, that is a non-negligible risk.

  2. Separately, we are asking to lock the CRV we get up front, same as JPEG’d. The deal here is that CurveDAO gets to see this CRV locked for 4 years + instead of potentially circulating in 1 year. For Alchemix, we get the voting power up front. It is the same thing JPEG’d is asking for, with the only change being locking a portion in sdCRV instead of all in veCRV.

So I think with that change in mind, anything related to what we do with our treasury should be ignored.

So the question is - JPEG’d want CRV up front and wants to lock it in veCRV. We want it up front and want to lock in sdCRV. Is this OK? I think the main criticism I am hearing is that because the sdCRV can be sold, this could still have an effect on the price of CRV (ppl may buy sdCRV instead of Curve). However my counterpoint is still that none of this sdCRV is being sold. We will be selling some sdCRV, but as noted above this will be done regardless of if this proposal is accepted.

FYI - the primary motivation to lock in sdCRV is because we have a significant veSDT position, so locking in sdCRV lets us use the veSDT to its full potential. The primary motiviation of a veCRV whitelist is so that we are not fully dependent on sdCRV, vlCVX, and other Curve derivatives and can use them for the product/features they offer rather than being the only option.

2 Likes
#9

Think here it’s a bit different since Alchemix already has CRV that they can sell anyway, so the vesting has limited impact.
I think vesting is liquid linearly over one year while sdCRV and cvxCRV (minted) are locked forever.
So for me any combination of sdCRV / cvxCRV minted is better than the vesting.

2 Likes
#10

I’m against using funds used to repay hack victims in a way that is preferential to any org. I’d agree to alchemix either:

  • Receive CRV immediately and do a 4-year vanilla veCRV lock
  • Receive CRV as a 1-year linear vest and do whatever needed with it

I’d be pretty surprised if this passes a DAO vote, pretty much in order for this to pass you need to convince Convex that it’s cool you lock CRV with their main competitor. I don’t see that happening, but I’d be very impressed with your diplomacy skills if you manage to pull it off

2 Likes
#11

I want to be clear on our primary motivation here:
We have a hole in alETH backing. Having to wait 1 year for CRV to vest would mean an unusable alETH product for that period of time and no guarantee the CRV is even adequate to cover the hole. Instead, we can plug the hole ourselves while we wait, which is going to require us to decimate our treasury, namely our sdcrv/cvx positions. Either of these options will hurt.

Yes, we can also max lock in CRV. Now we have far less flexibility in the future (have zero access to liquid capital with this voting power in case of any future security events) and miss out on a chance to benefit other protocols building on top of Curve.

Like is the goal here to actually help the projects who got impacted? Or to profit maxi even at the cost of Alchemix, who has been commited to the curve/convex/stakeDAO ecosystem essentially since inception, being severely hindered in the coming year - right as it was finding a fresh PMF with LST vaults.

We think we have a better option here that is better for CurveDAO and better for sdCRV. As noted below, we understand the politics and we are also active CVX holders/voters (388k vlCVX in the treasury), and are willing to add convex to this plan if need be.

I just think it would be a shame if this proposal failed because people want to impart additional restrictions on a plan that already improves the outcome for Curve and Alchemix.

There are two parts here:

1. What CurveDAO wants
I want to reiterate that it seems Curve has decided LPers are entitled to CRV tokens as recompensation. Curve does not wish for these tokens to circulate. As long as we meet those restrictions, why do there need to be additional restrictions? Ie, we have laid out a different plan that still follows those restrictions. Why take autonomy away from us when our approach either meets or exceeds the goals of the restrictions?

Perhaps we need to better define the purpose of these restrictions (ie,why Curve can only give 1-year vested CRV in the first place and the goal of that). I would appreciate insight there.

2. What Convex wants
I’m not naive to the fact that Convex will probably control this vote. I think Alchemix could be amenable to a 50/50 split - 50% locked in sdCRV, 50% in cvxcrv. While this would be somewhat hypocritical of some of the arguments against this proposal, I understand it’s a business decision and Convex may have no interest in benefitting a competitor. Ultimately this is a slightly worse economic outcome for Alchemix just because the efficiency of the veSDT position cannot be maximized, but it is still the same win-win for Curve and Alchemix, with stakeDAO and Convex also benefitting.

tl;dr - fine, business is business and we are certainly open to more groups benefitting from this. But lets not lose the forest through the trees of what the actual goal of recompensation is here, please!

1 Like
#12

why Curve can only give 1-year vested CRV in the first place and the goal of that

There is a Community Fund smart contract controlled by DAO voting, which holds roughly 124m CRV. It disburses CRV through deployment of vesting contracts with a minimum (linearly) vesting period of 1 year.

3 Likes
#13

To further clarify what Chanho said, the Curve Community Fund is the primary way recompensation is to be handled and, in general, to attain community CRV for any purpose requires a 1-year linear vest. The one alternative to receive CRV without a vest involves using funds held by the Curve Grants Council which have already been vested and which are typically reserved for public goods projects benefiting Curve. You can see from the link provided how much CRV the council has to work with.

The grants council feels that using funds for this purpose is outside the scope of Grants, but it is willing to send a requested amount of its vested CRV to a Curve DAO-controlled wallet, making the decision to distribute liquid CRV for this purpose executable by on-chain DAO vote.

4 Likes
#14

Thank you both for the clarification!

#15

I can empathize with ALCX (and of course JPEG’d as well). The fallout from the hack must have been very stressful. While I agree there should not be any preferential treatment, it is my hope an amicable resolution can be reached that satisfies the major stakeholders herein.

2 Likes
#16

Hey all,
We discussed with various parties. Additionally, the market upturn has let us secure a good chunk of additional runway, to the point where we believe we have a path forward to reback with our treasury directly without having to sacrifice our CVX/CRV positions. We will share an updated proposal if necessary, but for now, we can consider this one obsolete.

4 Likes