[RFC]: Rethinking the EUL distribution

Archive
, ,
1

Summary

The purpose of the EUL distribution is to decentralise governance of the Euler ecosystem among its users. The current system designed to achieve this is working sub-optimally. This proposal suggests a range of modifications to the current system to help widen the distribution. These include:

  • creation of staking contracts designed to distribute governance votes to lenders (in addition to borrowers, who are the target of the current system)
  • modifying the existing EUL distribution to increase the number of eligible markets from 10 to 40, whilst scrapping the quadratic voting on the gauges, and splitting a fixed amount EUL between staking (15,000 per epoch) and gauges (40,000 per epoch)

Staking

One of the main community complaints with the current EUL distribution mechanisms is that they do not distribute governance rights to lenders (only borrowers). Since the number of lenders on Euler outweighs the number of borrowers by roughly 5:1, this seems like a lost opportunity to help distribute governance more widely.

To remedy this, the proposal here is to create new staking vaults for eTokens. These would be Synthetix-style staking contracts where users deposit, for example, eUSDC, and receive a constant stream of EUL over time. To begin with, there should be a small number of such vaults, hosted over a short period of time, as a trial. The following three vaults are proposed initially: eUSDC, eUSDT, and eWETH. Each would receive an initial allocation of 5,000 EUL, to be distributed over a period of 6 epochs (roughly 3 months). Total cost: 5,000 EUL * 3 * 6 = 90,000 EUL. If this proved successful, it could then be rolled out with more tokens or to a wider set of markets.

Eligible Markets

Another concern with the current EUL distribution scheme is that it only includes 10 community-selected markets (voted on through the Gauges). This number worked well when Euler launched, nearly a year ago, and had only a handful of active markets. Today, however, there are over 100 markets on Euler and outside the stablecoins and majors (ETH and BTC), only a few longer-tail markets are eligible for a distribution in any given epoch.

Among those additional longer-tail markets, a large proportion of the EUL distribution is currently making its way to only a small number of users. For example, around 15,000 EUL per epoch is currently making its way to just 19 users of SDL, 15 addresses of STG, and a meagre 6 addresses of HMT. This is a result of the limited number of markets to which EUL is distributed, but is also compounded by the quadratic voting mechanism used in the gauges, which was initially designed to ensure long-tail markets did not get washed out in the voting by markets like USDC and ETH. In practice, however, it is clear that the limited number of users of long-tail markets does not justify the quadratic system.

The proposal here, therefore, is to remove the quadratic mechanism and increase the number of Gauges from 10 to 40 to allow users to vote on the WETH lending market in addition to all markets with a Chainlink oracle. Why Chainlink-only markets? Because all Chainlink markets on Euler have the highest oracle rating possible, whereas markets using Uniswap oracles have quite variable oracle ratings. In general, EUL should probably not be distributed to any market without a consistently strong oracle rating. In this way, Chainlink support would also help limit ā€˜fakeā€™ tokens being listed on Euler and used to game the distribution.

Whilst the number of eligible markets is increased, the amount of EUL should also be capped. In recent times, around 55,000 EUL has been distributed each epoch. This number is set to gradually increase over time, a design which was intended to match growing user adoption of DeFi with increasing governance votes. Given the depressed state of the market, however, it is clear that increasing the amount of EUL distributed beyond its current level will only further centralise governance (since user growth on Euler and most dApps is rising slower than it was before).

Since an additional 15,000 EUL per epoch will be moved to the staking contracts the proposal here is for a freeze on the EUL amount distributed each epoch to 55,000 EUL, where 15,000 would go to the staking contracts, and 40,000 would be used in the updated gauge system.

Update:

A concern voiced by a number of people across various community forums, including @Jib0xD and @itbresearch here, is that distributing to too many markets could dilute the distribution too much, especially among popular existing markets. A modified solution is therefore proposed that would follow the 80-20 rule, and distribute 80% of the new distribution equally among the current top 4 gauge markets (USDC, WETH, USDT, and wstETH), holding their current distribution amounts fixed each epoch, with the remaining 20% distributed via gauges across 40 additional markets (as above).

This modified proposal strikes a balance between favouring distributions to the most popular markets with the most users, whilst still enabling smaller market communities to engage. The modified proposal would leave the current top 4 gauges receiving roughly the same EUL distribution that they receive today, whilst opening up broader community participation from holders of other assets.

In terms of specifics, the top 4 assets would remain relatively unchanged: USDC would change from 8,286.29 ā†’ 8,000, WETH would change from 7,991.66 ā†’ 8,000, USDT would change from 7,028.52 ā†’ 8,000, and wstETH would change from 7,108.70 ā†’ 8,000. All assets with a Chainlink oracle would then receive a share of 8,000 EUL as determined by the gauges. This modified proposal will go to vote via Snapshot and will take effect from epoch 17 if passed.

9 Likes
2

Sounds a good directionā€¦

3

I think this addresses a concern of mine well that tokens that bring little value to the protocol are receiving outsized returns in terms of EUL emission. This solution appears to not only start to address this issue but does so in a careful way that maintains protcol security by prioritising robust assets.

I am hopeful that this proposal passes the temperature check and given itā€™s importance to the protocol I look forward to seeing it discussed further by the DAO.

4

I think these ideas are fascinating and very well intentioned.

Some comments:

Rewarding lenders: good in theory, some thoughts: By distributing to lenders could we not simply be doubling up on siloed distribution and centralisation? The number of lenders may be diverse but the concentration of capital may well still be among addresses of lenders who are also borrowers.

Increasing the number of eligible markets: I think the reasoning here to do this is excellent. What Iā€™m not sure on is the chainlink oracle requirement. Are you proposing that only markets with chainlink oracles receive EUL distribution? I feel this goes against the Euler ethos and value prop of being able to service the long tail while managing systemic risk. Moreover, there are some markets currently receiving rewards on a Uniswap Oracle and arenā€™t ā€œfake tokensā€. e.g. SDL. What happens in this scenario? It may be worth considering a multiple box ticking approach for whitelisting rather than just chainlink. e.g. overall oracle rating, verified token on xyz, existed for X number of days/months, liquidity etc. and if a token meets X number of Y requirements then it is valid for a gauge distribution. One could argue that this is essentially what chainlink does but I feel outsourcing decision making to them rather than the DAO is problematic.

Scraping quadratic voting: letā€™s give it a go.

Fixing EUL distribution: Up for it. Would the Gaussian esque curve be re-introduced at a later stage?

Once again. Thanks for putting this up @euler_mab

Looking forward to discussing. :slightly_smiling_face:

1 Like
5

Thanks for the proposal! I am quite supportive. Just maybe would like to ask for some clarifications:

  1. When we are talking about new staking vaults for eTokens, is it going to be implemented on the Euler Dapp or another platform? There were some ideas in Discord about Curve involvement into the process.

  2. Since stablecoins and majors (ETH and BTC) are currently taking most places provided by gauges for the EUL distribution, would not it be easier to include, for example, all the collateral assets in the distribution by default and 10 to be decided via staking?

2 Likes
6

Generally in favor of the changes proposed. Adding staking for tokens on the lending side should help more broadly decentralize EUL, while also making eTokens more attractive for potential pools in other protocols.

One question regarding the eligible markets: Are you proposing that every epoch 40 assets would all receive EUL rewards? Or that 40 are eligible to be selected and 10 of those with highest votes get the rewards? Iā€™m afraid distributing rewards across 40 assets at the same time could dilute the rewards too much (especially considering distribution would drop from 62k EUL this epoch to 40k EUL) and thus decrease protocol revenues/reserves.

Supporting only assets with Chainlink oracles makes sense to avoid allocating rewards to illiquid assets with a small user-base.

1 Like
7

This is great, changes needed to be made.

There is no perfect way to distribute the tokens so itā€™s important to remain nimble and reassess the changes after a few weeks.

1 Like
8

Yes, I was proposing the first option, where every epoch 40 assets would all receive EUL rewards. I agree that this would potentially dilute distribution. To counter this, I proposed removing the quadratic voting aspect, so that the most popular assets, with the most holders, like USDC/stETH/wBTC etc, should still receive more votes. Itā€™s still a fair concern though.

1 Like
9

  1. The proposal is to host it on the Euler dapp itself. I think there should be a new page on the dapp: ā€˜Stakingā€™, where new pools can be added for eTokens (starting with the 3 above).

  2. I think that is quite a nice idea. It would also help to alleviate some of the concerns of @itbresearch that the distribution could get washed out with too many ā€˜long-tailā€™ assets.

3 Likes
10

I was proposing to use Chainlink as a sort of gatekeeper, yes. I agree it goes against the ethos of Euler somewhat, but I think we need to address the elephant in the room with these more exotic tokens. One additional argument for relying on Chainlink is that, right now, they provide the most secure oracles on Euler. After the Merge, Uniswap oracles are a little cheaper to attack than they once were. So does it make sense to enable markets whose oracle isnā€™t first class? Iā€™m not sure it does. Obviously EulerSwap and its more secure oracle may change that in the future.

2 Likes
11

I dont understand the logic for going with more long tail markets for eul distribution. If the concern is around attracting liquidity then the cost basis for incentivising low caps with eul doesnā€™t seem to match the payoff when compared to high caps (i have not looked at the data for this, so could be wrong, its just a hunch based on how I was using euler a few months ago where i farmed low caps to liquidation in order to access the very high EUL rewards on those lowcaps) It makes far more sense to go with fewer if the goal is to attract liquidity imo. I think there are a number of lowcaps which are being promoted with very high APYs that realistically do not bring as much tvl value than if those funds were allocated to blue chips.

2 Likes
12

These are valid points. Raises the question of what is the main aim quite well actually. Is it to simply attract liquidity? My understanding of the aims of the proposal, was more of a wider distribution than just liquidity. But what you propose may achieve both. :slight_smile:

13

These are good points.

One thing to note is that EulerLabs are currently developing a new solution that will enable a much larger set of assets to be used as collateral on Euler (without altering the protocolā€™s risk profile). This could bring a lot of extra utility to markets that are currently relatively inactive.

The other point is that the EUL distribution is supposed to help decentralise supply. Itā€™s not obvious whether you do this by targeting a small number larger markets (with more users) or a broader range of smaller markets. People have varying opinions on this. And so one argument for broadening the Gauge system is that this decision is left entirely up to EUL token holders themselves. Rather than go through governance (as we are now), the Gauges facilitate this decision.

A major argument against the Gauges is that when people stake their tokens, they lose the ability to vote in ordinary governance. This is something that I think should be improved. Otherwise people are torn between the two. A solution here might be to create gEUL when EUL holders deposit into the Gauges. And then let gEUL holders have a say in governance.

2 Likes
14

This last solution (gEUL) is a big positive step and I would be very much for it.

15

Please see the update at the end of the proposal.

16

Completely favour of the 80/20 amendment. Keeping the rewards more or less the same for WETH WSTETH USDT will make sure WSTETH lending APYs are competitive. WETH rewards make WSTETH/WETH trade very attractive too, and is a crucial market for Euler.

17

Re: the updated 80/20 suggestion: I think focusing on the core assets is a good way to do things for now.

As mentioned, this should free up some EUL that can then take place in governance votes, which is very positive.

Ultimately, this isnā€™t a perfect solution but a net positive step until EulerSwap is launched and a new system can be put in place.

18

I am very against the proposal for a number of reasons as discussed in discord, as well as a few more not mentioned there.

  1. What comes first? Chicken or Egg? Borrowers or lenders?

The impression that the market has shown has been that in order to onboard capital, you need to provide lenders good rates.

That said, what leads to good rates? Borrowers.

And how do you get borrowers? You incentivize them. Exactly what Euler does now.

This creates a flywheel effect where lenders earn pure yields (not worried about Euler token principal risks) that can be advertised as higher than other competing venues.

These higher yields then attract deposits, which lower rates and attract more borrowers, and all parties are happy. Lenders get a high pure yield, borrowers can profit from some minor arbitrage and your power users are rewarded.

Moving away from the current incentives without that interaction being discussed seems very ill advised. From a pure capital onboarding perspective, I just cant imagine how lowering the lending rates would onboard more lenders, regardless of whether they might earn slightly higher amounts when accounting for incentives.

As the team has noticed, the market is inefficient for borrowers, there is nothing that indicates that all of a sudden it would become efficient when you introduce lenders, meaning the lenders likely just end up with a worse rate and worse UX.

  1. Moving away from borrower incentives hurts integrators.

As an Euler integrator, we would MUCH prefer to never touch Euler tokens for our users. With this change we either end up taking that extra yield from our users and keeping the tokens, or alternatively bloat our gas costs.

The current incentive structure gives integrators a pure yield that is higher than it otherwise would be.

Giving lenders incentives breaks this interaction and no longer leaves Euler as differentiated from an incentives point of view.

  1. Why would you require users to stake rather than include in current distributions?

This one is just more confusing from a mechanical level. Youre telling me lenders have an extra step just to get access to the same yield (or close to it) that they otherwise would have?

Youre also telling integrators (Swivel) that we need to figure out a way to stake our deposits to even get a benefit, likely requiring additional minor audits for each integrator.

I just dont see how it makes much sense to push in this direction without clear and concise reasons / data that would indicate that the first point is incorrect, and that for some reason inefficiencies will be resolved by adding lending incentives.

20

JTraversa, if you read the proposal carefully you can see that borrowing incentives arenā€™t going away. In fact, borrowers will get slightly more EUL for borrowing USDC USDT WETH WSTETH than before.

It adds lending incentives for USDT USDC and WETH primarily to decrease interest rate volatility. Especially for WETH, itā€™s good that lenders arenā€™t pulling out so the WSTETH/WETH trade remains more competitive vs Aave.

21

I think thereā€™s been some misunderstanding here. The proposal is to add additional staking options for lenders, not to remove a distribution to borrowers. In fact, borrowers will still be receiving around 75% of all the distributed EUL per epoch. And in absolute terms, getting pretty much the same amount they were receiving until only very recently in August.