Rethinking Liquidity Mining Gauges

Tokenomics
#1

I’ve been chewing for a few weeks on the idea that the liquidity mining gauges are currently the wrong abstraction for most users.

The gauges are meant to crowdsource the answer to “Where does Marinade need more liquidity?”

Instead, the question people hear is “Who should get a chunk of our treasury this week?”

When you see it that way, it is perfectly reasonable for people to select their own pool, even if it does nothing to increase mSOL adoption or liquidity overall.

A potential solution would be a UI layer on top, a “Liquidity Mining Wizard” which asks users to think in terms of areas first, then provides with a preset distribution that they can use or change.

We currently ask Joe User to vote for a pool, so obviously they’d vote for mSOL in Solend if that’s where they have their stuff - they profit, and it is familiar. They don’t think of what value the other pools might add.

The LM Wizard would first show them higher-level options, and ask them which areas they think Marinade needs liquidity on. They may not only choose Lending Collateral, but also select Trading for MNDE, at which point we can present an allocation preset that also includes MNDE/USDC pools, which they can then tweak.

Even from a selfish player standpoint, this may prompt them to think “crap, even if I farm all that MNDE I can’t dump it without slippage, better get MNDE/USDC some liquidity”.

Meanwhile, someone with strong opinions (eg. a protocol holding MNDE) is free to ape into their own pools if they want to, disregarding the wizard.

In the ideal situation, we’d get close to answering the question the protocol cares about, while the worst-case scenario is the same behavior as now.

Thoughts?

3 Likes
Are Gauges increasing mSOL adoption?
#2

This proposal makes sense to me.

In driving mSOL liquidity, there are Atrix, Orca and Raydium AMMs, each having no significantly more TVL than each other. This gives a very diverse spread of MNDE incentives to each AMM, which looks acceptable to the average MNDE holder.

However on the lending side, Solend dominates lending TVL and userbase, which leads to most of the MNDE centralized onto 1 protocol. This has led to understandably concerns that Solend is getting lots of rewards (which i understand where these people are coming from). As for my own concerns, we are also worried that too much incentives will flow towards non-AMM venues and weaken overall liquidity for mSOL.

@PlayerOfBits’s idea to lets users to choose the higher level options first makes sense in Solend (and also Marinade’s) goal to keep liquidity and collateralization balanced, while also serving as a better UX for users to make decisions.

Disclaimer: I am a contributor to Solend.

3 Likes
#3

I think what’s playing out now is exactly as designed in the game theory of liquidity gauges, just like Curve wars. From the perspective of the protocols engaging in buying and holding an MNDE treasury, (which are the ones moving the needle in terms of MNDE distribution), they are in general doing this in order to add value to their own users and to grow their own protocol, even while I’m sure like us at Hubble being big supporters of mSOL and Marinade in Solana DeFi. eg. (random example) DEX A I doubt would vote for liquidity on DEX B for the overall good of Marinade. They’ve invested in MNDE, and want to use it on their own DEX, whatever the overall health of DEX B liquidity.

The ‘Joe users’ you mention, for sure exist. But firstly I don’t imagine their voting has much influence (one could analyze this), secondly as you mention, a good proportion would also vote for where their personal interest is depending on their assets held.

So in essence by ‘decentralizing governance’, the impact is that governance is controlled by those with the biggest wallet. ‘Centralized’ governance controlled by Marinade, would probably be the easiest way to achieve what you suggest, like big picture planning, and directing MNDE incentives more to areas and protocols where the liquidity is needed from the ‘overall good of Marinade’ perspective.

1 Like
#4

I support the goal of this proposal, though I’d prefer a simplified version in which the DAO agrees on a slimmed down version of the gauge selection

@Soju agree, liquidity emissions have positive externalities that B/L or DOV don’t have.
(Within liquidity, there is also an implicit hierarchy of LPs that provide the biggest positive externalities: mSOL/USDC, -SOL and — often neglected — -MNDE.)

But happy to move forward with the LM wizard if that’s consensus

1 Like
#5

Sounds good.

Regarding improving liquidity for MNDE. If there were a CLOB where you could trade MNDE/SOL - MNDE/USDC or MNDE/mSOL I’m sure this would help. Is there anyone working on getting MNDE listed on an exchange ?

LPing on AMM is pretty much guaranteed to lose you money.

Also most people have the impression that if a token is not on an exchange it’s most likely a scam coin . having MNDE tradeable on an order book, preferably on a popular exchange, will make it more approachable to the general public and speculative trading will provide the liquidity needed. perhaps.

3 Likes
#6

Not to my knowledge. MNDE isn’t even on Tradingview

I have written about the inherent conflict of marinade, ie the tradeoff between promoting mSOL at the expense of MNDE. IMO the pendulum needs to swing in the other, more sustainable direction

2 Likes
#7

Agree with the spirit of this discussion and think that @Soju said it best here:

I also think it would be valuable to keep a record of responses to this LM wizard. Only thing I would caution against is making the UX more complicated / slower for everyone by default. Probably worth including the option to skip for users who already know where they want to direct MNDE liquidity.

1 Like
#8

I think this idea is well intentioned and might help but I think building delegation and supporting that well would be a better long term solution

2 Likes
#9

I generally agree, and delegation is on the roadmap. This idea is likely something that came to mind too late. :slight_smile:

I do think it behooves us to keep this sort of thing in mind, though, since we may need to design other systems where people may not hear the same question we are asking.

1 Like
#10

Adding my two cents here, from ideas I’ve read elsewhere:

  1. Having a gauge that says “MNDE back to Marinade” is a good way to give MNDE holders a say in the amount of MNDE distributed.

  2. Maybe adding some form of quadratic-vote-interpretation to the gauges->percentage translation, in a way that, for example, will require a ton of votes to move a gauge from 40% to 45%, but a small amount of votes to move a gauge from 1% to 5%

3 Likes