Mis-use of Marinade gauges by closed AMM

Governance
, ,
#41

Our mSOL-USDC liquidity is permanent, unless Marinade violates the terms of the grant, Lifinity decides to dissolve, or something like that. If I remember correctly this was discussed somewhere in the thread of the proposal.

Whenever market conditions get tough, market makers / LPs are incentivised to withdraw their funds.

As an aside, I mentioned this earlier but one of the benefits of Lifinity is that we don’t decrease the liquidity we provide even as market conditions get tough. It’s stable regardless of the price of MNDE.

2 Likes
#42

But also less MNDE to Lifinity doesn’t mean less liquidity, which is beneficial to Marinade. It goes both ways.

1 Like
#43

If every protocol would do this, you get the MNDE ponzi that @Hao is describing, without driving any actual value to mSOL/mDAO.

This is entirely dependent on the exact model being used. For example, it’s not true of Lifinity’s model, because the MNDE acquired ultimately belongs to token holders. Protocols permanently keeping all rewards for itself is just one possible model, and I certainly don’t recommend it. As I mentioned in a post above, protocols could simply delay the distribution of MNDE rewards to LPs, using it to lock and vote in the meantime. Ultimately, it’s a matter of willingness to withhold instant gratification (lock and vote now, distribute later).

1 Like
#44

so the conclusion is that it is in the best interest of mDAO to send less MNDE to Lifinity and put it to better use elsewhere? :stuck_out_tongue_winking_eye:

1 Like
#45

I think this thread is looking more and more about people looking after their own interests and loosing sight of the bigger picture. mSOL adoption and increased use. Lifinity partnership came about as there was a need for ‘Deeper’ liquidity for mSOL and the fact Lifinity concentrates liquidity by up to a factor of a 100X is a net positive in this regard. Its disingenuous to suggest Lifinity has not lived up to its side of the bargain here, and apart from offering a valued service, it executed professionally every step of the agreement that was voted upon.

Ive been researching Lifinity since late December 2021. Ive used SRM and RAY since day 1. Most of you know me as ive been on Marinade since the beginning.

Traditional LM based Dex structure is already outdated. Its simply unsustainable and by its nature cannot last long term. Witness race to the bottom on fees to just try grow TVL. And no protocol appears to be looking at a traditional profit making business model.

POL / Concentrated Proactive AMM is the present and the future. Pyth and Jump are 100% committed to Oracle driven stock exchanges and have very deep pockets. Lifintiy is the closest thing right now to that and most resembles a tradfi MM.

This is why its great Marinade community voted to add lifinity as a partner and seek diversification over traditional lp pools.

I invested in lifinity initially in the flares and then at IDO. I follow them daily. I have Witnessed the team deliver and the culture & execution is exemplary.

To me this its important that we acknowledge the benefits the Lifinity partnership has delivered. Its a better solution than what we currently have in place in every way imho, and its moving with the times. Its much cheaper and more efficient. Its brought MNDE / mSOL awareness and adoption in lifinity community, so much so we voted to lock up our MNDE from the guages. And this is a bad thing!!! Should we dump it as soon as we receive - would that be better :thinking: Also a hell of a lot of the community bought MNDE and Octopus NFTs…this is helping to drive Msol / MNDE adoption organically which is the strongest and stickiest imho.

Also, anyone try trade mSOL on FTX? I have and the spread is terrible (was 30 points)…actually discourages SOL folks from holding mSOL. Since Lifinity ive been back holding more mSOL and therefore confidently trading knowing im working with a tight spread. Dont underestimate how important this is to mSOL holders.

https://twitter.com/PigFahy/status/1553449137013760004?t=mX_lMYRcRfYXRtxSPcArPw&s=19

In conclusion we dont want this to turn into a witch hunt for a partner that has done everything asked of it and more. Lets be careful as we wanted a decentralized DAO. If something needs amending then we are free to table a vote. Buy your MNDE to play.

4 Likes
#46

totally agree on Lifinity’s contributions, I don’t think anyone here is trying to undermine the benefits the partnership has brought to the ecosystem.

At the same time keeping the bigger picture of the goal to increase mSOL adoption and use in mind, we can’t ignore the reality that the current gauge is going to hurt both mSOL adoption and use in the long run.
Every concern that has been brought up has been solely focused on the exponential feedback loop that the current gauge enables, I don’t see anyone saying Lifinity should not be able to get more MNDE in addition to the grant or even that they did not deserve the 2M MNDE grant. It is just the exponential feedback loop that presents a practical issue because:

a. It encourages other protocols to turn MNDE gauges into a ponzi as outlined in my and other examples; the alternative is that Lifinity will eventually own virtually all MNDE drying up all other gauges and we end up being stuck with 1M worth of mSOL liquidity in perpetuity.
b. Lifinity doesn’t need more MNDE to maintain TVL (as @Durden admitted), so there is real opportunity cost in allocating more and more MNDE to Lifinity every cycle.

I think the goal of this discussion should be how the gauge can be reworked into something that is still attractive for Lifinity, while also making sure it leads to the desired increase in mSOL adoption and use in the overall ecosystem.
The unfortunate reality right now is that the current gauge is not helping to regulate mSOL liquidity but purely a vehicle to drive more value to LFNTY holders at the expense of mDAO incentives.

4 Likes
#47

The way I see it, there are 3 paths forward:

  1. Lifinity begins distributing MNDE rewards so that their share of rewards no longer increases. MNDE dumping accelerates a bit.

  2. Other protocols either begin buying MNDE or temporarily retain users’ MNDE rewards to lock and vote, increasing their share of rewards in lockstep. MNDE dumping decelerates or is reversed into accumulation.

  3. The way the gauges work is changed.

I am of course biased, but in my mind 2 would be the best outcome since it increases the value of MNDE and creates an invaluable narrative around it – the MNDE Wars. 3 could also be good, but we would need a truly innovative idea, one that could be expected to outperform Curve’s proven model. 1 is the least interesting and directly goes against the MNDE Wars narrative since it discourages MNDE accumulation.

6 Likes
#48

The discussion here is if MNDE incentives should go to a closed private pool where:

  • the amount of liquidity does not depend on the incentives
  • the MNDE does not get distributed

The MNDE incentives distribution core objectives are:

  • Increase mSOL liquidity in the market, more MNDE distributed => more mSOL liquidity (Lifinity gauge is failing here, the depth does not depends on incentives)
  • Distribute MNDE governance power in the community (Lifinity gauge is failing here too)

The entire system of incentives is based on the idea of decentralizing voting power, and you’re on a self-reinforced feedback loop generating concentration that grows exponentially. We need to fix this problem.

Gauges were created to distribute incentives to open, scalable, community fed liquidity pools and other community fed mSOL uses.

Before that, we were not distributing increasing MNDE amounts to private pools that can’t grow organically, that makes no sense for Marinade.

6 Likes
#49

The current situation has nothing to do with “Curve’s proven model”, you really should stop bringing that up as an example.
CRV means one DEX is giving out rewards and multiple tokens compete for building up liquidity for their token on that DEX. They are competing for LPs that “I know where I want to provide, WHAT token do I provide liquidity for?”
MNDE is one protocol giving out rewards for liquidity on one token and multiple DEXs compete for providing liquidity for that token. They are competing for LPs that “I know what I want to provide liquidity for, WHERE do I do it?”

In addition, part of the reason CRV wars work is because there are simply not enough alternatives for a protocol to build up good liquidity in other places, so they are more required to participate in CRV wars.
Meanwhile for MNDE, the DEXs and other potential mSOL integrators, don’t need to nearly rely on one pair / token as much. While it probably would suck, Solend, Orca, etc. are gonna survive if they had to remove / lower their TVL in mSOL and mSOL/USDC, esp. with alternatives for liquid staking popping up.
Just saying f it and bailing on mSOL is a very viable option if they deem the MNDE wars to be pointless because of players with “unfair” advantages

For the option you are lining out:

  1. isn’t doing anything unless Lifinity also opens up to external LPs to provide liquidity and farm MNDE, so not really an option unless Lifinity fundamentally changes

  2. is not going to help since Lifinity users can just sit through whatever lockup other protocols pick and outcompete this way, since -again- LFNTY holders do not have their money directly at stake in the pool

To give more color to 3, some viable options in my opinion are:

  1. the Lifinity gauge gets removed, Lifinity can receive more MNDE through additional grants if needed

  2. the Lifinity gauge gets a fixed limit on votes that can be committed to it to avoid exponential growth

  3. the amount of MNDE rewards awarded to Lifinity needs to directly correlate with how much TVL is provided on Lifinity (might be hard to enforce)

1 Like
#50

I think we should not make a specific rule aimed at Lifinity. They are showing us that we have a problem, that we didn’t understand correctly the implications when we created the Lifinity gauge, and that the problem must be corrected.

We also need general rules on gauges creation in the future, so MNDE distribution is always aligned with Marinade objectives. That’s basic. The Lifinity gauge should be disabled because it does not aligns with Marinade core objectives for MNDE distribution at all, and the problem was not noticed on the original proposal.

Any gauge with the same problems should be disabled/never created.

The Lifinity gauge is failing all core objectives. It is not that Lifinity is getting exponentially more MNDE because doing a great job at building mSOL liquidity (liquidity is fixed) or decentralizing MNDE governance (MNDE is concentrated)… it is the exact opposite of that.

3 Likes
#51

sorry I missphrased in my last post using Lifinity as scapegoat while this should be applied to “closed AMM” gauges in general.
Very much agree that this needs to be a general rule moving forward so we don’t have the same conversation again whenever the next “closed AMM” shows up

1 Like
#52

Note: I’m cool with what Lifinity does as closed AMM providing mSOL/USDC liquidity. We paid a fixed 2mm MNDE grant for that and for the development of the oracle, and Lifinity is getting close to $430k/year in profit from the user’s swapping in their private pool. That’s good.
But the gauge itself is an exponential problem, it is contrary to Marinade objectives and should be disabled.

2 Likes
#53

Repeating myself as I seem to be doing a lot of but…

  • the amount of liquidity does not depend on the incentives
  • Increase mSOL liquidity in the market, more MNDE distributed => more mSOL liquidity (Lifinity gauge is failing here, the depth does not depends on incentives)

The fact that Lifinity’s liquidity doesn’t change benefits Marinade when MNDE price decreases.

  • the MNDE does not get distributed
  • Distribute MNDE governance power in the community (Lifinity gauge is failing here too)

It does get distributed, just not immediately. It’s part of the grant’s terms. If it were so obviously bad, surely it would have been called out while the proposal was being discussed.

2 Likes
#54

It wasn’t called out properly. That was a failure on our part. I was opposed but because it was too expensive in my view, but I wasn’t familiar enough with Lifinity and I didn’t realize that the pool was also closed and fixed. That was a failure on my part, I don’t know how many others were not familiar with Lifinity

2 Likes
#55

  1. It stops our accumulation of MNDE, which many have voiced as a concern.

  2. You could say the same of other protocols – they can just sit through Lifinity’s lockup and outcompete this way. The situation is symmetric. LFNTY holders do have their money at stake in the pools. Lifinity’s POL belongs to token holders.

Regarding changing how gauges work:

  1. This would violate the terms of the grant and would therefore terminate it.

  2. Very difficult to determine how much, and the “right” amount will likely change with time. Simply having us not lock rewards seems like a simpler solution.

  3. I think this could make sense (it’s similar to what we originally proposed – incentives based on volume generated), but it should be applied to all protocols since all protocols can be measured by TVL. Unfortunately, this would rid MNDE of a lot of its utility (no more liquidity mining gauge voting).

2 Likes
#56

Just checking: -

So if Lifinity give out all the MNDE to community right away and we all sell it immediately it’s fine. Is this correct?

As a person who has held MNDE since 86 cents (I entered mSOL staking in Aug 2021 so picked up several thousand MNDE and have never sold a single one) this does not benefit me.

2 Likes
#57

  1. “Ponzi profiters voice concerns about stopping the ponzi” - mild shock

  2. No it is not the same, you still don’t understand my point. If I buy $100 worth of LFTNY I don’t need to care at all about what mSOL/USDC is doing, I am guaranteed a risk free MNDE reward after whatever period is needed. Meanwhile I’m earning profits from all the other pools that contribute to LFTNY value. Even if the mSOL/USDC pool gets completely REKT, impact on LFNTY is very limited. My $100 investment also did NOTHING for mSOL liquidity.
    If I buy $100 of mSOL and USDC and deposit it onto a DEX I have huge opportunity cost, I am directly exposed to mSOL/USDC, sitting it out compared to someone who simply bought $100 LFNTY is way more expensive for me. I am much less incentivised to provide liquidity in the first place, although my $100 is 100% net mSOL liquidity on top.

  3. I don’t see an issue with adjusting/amending a previous proposal, esp. if the proposal was voted on with wrong/incomplete assumptions

  4. not hard at all. We know we get 1M liquidity from Lifinity, add some bonus on top because it is supposedly more concentrated and we can easily come up with a number that gives Lifinity a reasonable APR.

  5. on any other protocol more MNDE rewards = more benefit for mDAO, usually in form of more TVL. That’s what gauges should be for.

EDIT Can’t get the numbering correct but 3 is about point 4, 4 is about point 5, etc.

1 Like
#58

Hey I just caught up on this thread and wanted to chime in. Sorry long lol

TLDR; The original grant was probably the problem, fixing it by shutting off the gauge is probably the only solution at this point.

There seems to be a disconnect on the issues

  • the amount of liquidity does not depend on the incentives

@Durden claims this is a benefit because when MNDE price goes down liquidity stays the same. I would say this is more like a coincidence (and using the same logic removing the gauge liquidity stays the same).

I have always thought about incentives in terms of how a protocol can achieve some metric, in marinades case this would be driving liquidity efficiently and total stake (in solend for example it’s driving borrow demand mostly). if I am a MNDE holder and I see I can pay 1 MNDE per day per msol liquidity on one protocol vs 2 MNDE per day per msol liquidity I would obviously choose the former because I can get twice as much liquidity at the same cost. This is a central tenant of of any emissions program. Gauges are a way to decentralize these decisions, which is good, but sometimes create perverse incentives as we are seeing right now. Agreement notwithstanding, marinade dao has no incentive to give emissions to lifinity AND lifinity has no incentives to provide liquidity for MSOL (if they dont find it a profitable endeavor without incentives) which i think is kind of a core problem (even more than the feedback loop). The feedback loop has it’s own problem but even if MNDE was distributed this issue still exists.

I think we as a DAO should be looking at this is in terms of efficiency of paying for liquidity. I dont have exact number but a good lower bound estimate is that by end of year 1 of gauge LIFINITY will have somewhere in the range of 7~10 million (not counting reinvesting just looking at current rates). In addition i estimate lifinity will make $40~50k worth of rev from staging gauge voting (if played optimally). Not exactly what we thought when we initially voted for that $1mil of locked liquidity eh?

It’s part of the grant’s terms. If it were so obviously bad, surely it would have been called out while the proposal was being discussed.

I will say at least for solend, we were not aware of this effect, and were already on the fence about the original grant proposal (mostly because lifinity liquidity is not useful during liquidation, will try to get data on this, and that it already seemed on the high end of cost) and I think this would have pushed solend to voting no.

Distribute MNDE governance power in the community (Lifinity gauge is failing here too)

I wanted to just address this quickly so i wasn’t ignoring anything. I agree in principal. However every dao member/representative has different motivations and focuses. So while I think this is important it is not a core reason solend participates in marinade governance so will defer to others expertise

Given the above I feel removing the gauge is the only viable option I have seen to move forward, but am open to being convinced otherwise

2 Likes
#59

  1. I was pointing out the benefit of us stopping accumulating because you said there was none :person_shrugging: And I wasn’t going to bring this up but since you keep calling what we’re doing a ponzi and insinuating that we are acting maliciously, let me point out that it’s not even close: “A Ponzi scheme is a form of fraud that lures investors and pays profits to earlier investors with funds from more recent investors” (Wikipedia).

  2. The impact of mSOL getting rekt on LFNTY is proportional to the value of mSOL in Lifinity’s portfolio, just like any other LP. So no it’s not “very limited”.

  3. Lifinity can’t be held responsible for Marinade changing its mind after the fact. This should go without saying , but if the terms are changed, Marinade should not expect Lifinity to fulfill whatever promises it originally made. Can Lifinity also unilaterally change the terms and expect Marinade to just accept it? Of course not.

  4. I doubt it will be easy for Marinade and Lifinity to reach an agreement, and again the “right” amount will likely change over time, so choosing a fixed number to be used forever will likely not work well.

  5. On any other protocol less MNDE rewards = less benefit for mDAO, but this is not true in the case of Lifinity! You keep stating the one side of the equation while ignoring the other.

3 Likes
#60

I feel like It is very obvious that Lifinity is receiving an outsized amount of MNDE for the liquidity they provide. So this point is kinda moot because even if the rewards is reduced by 50% it is still 4-5x more than whirlpool’s MNDE per $1 tvl. So it’s fair to look at it only at @DerMitOhne‘s way.

1 Like