Update Liquidity Incentivization (for SETH2/ETH and SETH2/RETH2) with $SWISE to incentivize a Desired Range around 1:1 rather than (just) the Active Range

thatexactmike · June 27, 2021, 4:41am

Background

On 2021–06-23 Stakewise DAO began a farming campaign for liquidity on Uniswap v3 pairs for sETH2/ETH and sETH2/rETH2. Soon after the campaign began the price of sETH2 in ETH in the sETH2/ETH pool became between .95 and .96 ETH.

$SWISE farming incentives were targeted at the Active price, so liquidity providers (LPs) in a tight range around 1:1 sETH2:ETH were not active and thus were not incentivized with $SWISE. However, LPs whose liquidity included a price range of about .95 ETH are incentivized with $SWISE. These LPs are at risk of losing their sETH2 at a ~5% discount from their initial investment in terms of ETH (unless they obtained the sETH2 by swapping at a lower price).

This is creating a situation where the goal of the DAO to maintain a tight price range for sETH2/ETH around 1:1 is not being reinforced by the incentivization and LPs who desire to benefit from farming rewards by providing liquidity in that range without risking their sETH2 being sold at a undesirable discount cannot do so.

There is a case to be made for stakers who want an early exit to take a discount on their sETH2 to be able to exit early but I understand most of the DAO participants desire to maintain a price for sETH2 closer to 1. A small discount can help ensure willing buyers of sETH2 for stakers who do find themselves needing to exit the position early. On the other hand, a premium for sETH2 is also probable in certain market conditions where the staking queue becomes a significant factor for potential stakers (or for rETH2 holders wanting to compound their positions in smaller increments). Together, these considerations push us to desire a tighter price range around 1:1 than we are currently experiencing.

If my understanding of how Uniswap v3 pricing works is correct (all the available liquidity in a certain range needs to be converted before the price will move out of that range), at the current time given the liquidity amount and ranges in the pool (see the image above), for the price of sETH2 in the sETH2/ETH pool to reach a closer approximation of 1:1 it would be necessary for about 500 ETH to be swapped for sETH2 (plus an additional amount to offset any additional sETH2 to ETH swaps that occur) before we are back in that range. This situation could be somewhat mitigated now and in the future by making sure that incentives for LPs are aligned with the goal of keeping a tight 1:1 range. If sETH2 liquidity providers adjust their liquidity range to the Desired Range they will reduce the 500 ETH required to be swapped to a smaller amount of ETH required to be swapped for sETH2 to reach it’s approximately 1:1 range with ETH.

It might also be desirable to have some incentivization for the current active range when that diverges from 1:1 due to extraordinary market forces, so that LPs willing to engage in those transactions can benefit and the market can be more efficient for those who want to transact outside of the Desired Range.

Idea

Let’s think of incentivizing liquidity in terms of a Desired Range and the Active Range.

My idea is that the DAO should update the $SWISE incentives for sETH2/ETH LPs (and possibly also for sETH2/rETH2 LPs) to reward liquidity provided in the Desired Range.
The Desired Range would be the range near 1:1, perhaps the range of .99402-1.006 (about 1%) or the range of .98807-1.0121 (about 2%).

This would cause some LPs to tighten their ranges to the Desired Range and this would enable the price of sETH2 to return to the desired range in a shorter amount of time (both now and in the future when market forces may temporarily push it out of the desired range). Below is a screen snip of what it would potentially look like to add liquidity in the Desired Range:

Additionally, if technically feasible (and what can’t tsudmi do? maybe we should call him houdini tsudmi?), perhaps we should also continue to incentivize the Active Range when it falls outside the Desired Range so that when market forces do push the price out of our Desired Range willing LPs can be rewarded for responding to those market forces.

IF we incentivize both the Desired Range and the Active Range, it would need to be determined what happens when those overlap and when they diverge, and the available $SWISE incentive would need to be prorated between the LPs providing liquidity in these ranges. However, I believe the most important part of this idea is to incentivize the Desired Range over the Active Range so that the Desired Range is best supported in the market.

Disclosure

I provide liquidity to the pool and currently do so in a range close to 1:1. Some of my provided liquidity was swapped from the pool at a discount rate so I stand to benefit if the price in the pool rises closer to 1 ETH per sETH2.

ottodv · June 27, 2021, 9:24am

This may be the number 1 thing to resolve right now, as I don’t think it was ever anyone’s intention to incentivize a price at around 0.95.

The way I see it, we should indeed incentivize the Desired Range (at 1% or maybe 2% spread as you pointed out).

As for the Active Range, I think we should not incentivize liquidity between the current price and the target price (desired price) of 1:1, but a 1% range just below the current price.

Example with the current price at 0.95, any liquidity between 0.94 and 0.95 will be rewarded. The incentivization thus rewards people for building a wall that stops a price decline. If the price moves up to say 0.96, then the incentivized range also moves up to 0.95-0.96.

Bottom line is that liquidity that blocks a return to the Desired Price (1:1) does not get rewarded.

Edit: The Below Active Range incentive should only work in 1 direction: up for sETH2, if 1 sETH2 is worth more than an ETH that’s not a problem, we can still maintain the Below Active Range incentive even when the price is at parity or the sETH2 is higher than ETH.

dreth · June 27, 2021, 11:31am

Great writeup Mike, at first I was not entirely in agreement, but I hadnt considered the risk of principal loss for people which staked using a 1:1 peg back when it was not possible to stake through the pool.

I think the idea is great. I think in combination to my rewards flattening proposal, this could be a great combination that benefits not just smaller wallets but also bigger ones, which do have a lot to lose if they were to lose ~5% of their capital on a whim because of a massive whale trade or several of those.

I fully support your idea, I think that 1:1 range should theoretically count as “active liquidity” even if the $SWISE farming APR is maybe slightly lower than the liquidity within the active trading range.

Haahah the dmitri power bro

jayschoenherr · June 27, 2021, 6:31pm

YES. I pulled my liquidity out from the active range because of this. I’m not interested in selling my sETH2 for a 5% discount! If liquidity is incentivized tightly around 1:1, I’ll put it back in. Thanks.

thatexactmike · June 29, 2021, 5:38pm

The Active Range has moved much closer to 1:1 than when I wrote this and market forces (Liquidity more and more being provided in a tight range around 1:1) will likely help to keep it in that range. However, I still think it would be in the best interest of the Stakewise DAO and protocol to have the primary incentivization of Farming rewards ($SWISE) be allocated to a narrow Desired Range around 1:1 rather than the primary incentivization following the Active Range. I’m not against SOME incentivization for the Active Range, but I think we should incentivize the 1:1 Desired Range MORE. In our current situation, more liquidity would have moved/be moving to the Desired Range if this were the case. We are incentivizing those who are already benefiting from the price of sETH2 slowly but surely going up to the 1:1 Desired Range when that incentivization probably is not necessary since many of these LPs are swapping ETH for sETH2 at a discount on sETH2, then providing LP at a higher tick range. So, they are already making around %.05 percent (or more) per tick the price moves higher. There’s nothing wrong with this activity - ultimately it brings us closer to 1:1. But there’s also not as strong a reason to incentivize the current Active Range with $SWISE as there is the incentivize the 1:1 Desired Range.

ottodv · June 30, 2021, 10:18am

I have changed my mind on a few aspects. I think it’s not a good idea to incentivize the Desired Range if it’s outside the Active Range, because it incentivizes creating a large sETH2 buffer between the current price and the desired price (parity). Given that there is even less chance of the sETH2 being traded, it’s basically a free hand out with no positive effect.

Another aspect to consider is that ETH has no returns from staking while sETH2 does, so it’s more rewarding to provide sETH2 as liquidity than it is ETH.

Thus I think the incentive should primarily be targeted at ETH in the Active Range. This is the ETH that creates price support.

My suggestion would be to only incentivize ETH up to 1% below the current price.

What happens then if you provide liquidity in the Active Range:

When ETH is converted to sETH2, you gain the staking reward of sETH2 but loose the liquidity incentive of ETH.
When sETH2 is converted to ETH, you loose the staking reward, but gain the liquidity incentive.

My feeling is that this creates a better balance of rewards, whilst also using the incentives to better support the price of sETH2.

thatexactmike · June 30, 2021, 4:08pm

This is helpful feedback. Forced me to think through some considerations but actually ended up confirming (to me, at least) that incentivizing Desired Range is best overall (most incentivization should go to Desired Range).

I don’t think it would create a buffer between Active Range and Desired Range. Look at our current situation, as an example.

Why is there so much liquidity at .9881 ETH per sETH2? I think it is PRIMARILY because incentives are in place for the Active Range, which has accumulated at this point because of the incentivization. Most of the people with sETH2 liquidity at this range would probably rather have it at a higher price (to the right), unless they are really interested in exiting their sETH2 position (some likely are but I suspect not all of them). At this point moving out of the Active Range to the Desired Range is going to take more time and more swaps of ETH for sETH2 than if we were incentivizing Desired Range.

I don’t think this point holds up under examination (except for the price support comment). sETH2 gets the underlying staking fees, but to take on the additional risk and cost of providing liquidity I think requires additional rewards beyond just trading fees (which are going to be pretty minimal at many points in the market cycle). If incentives were targeted at the narrow Desired Range, then both ETH and sETH2 providers are incentivized better than they are now (anyone with ETH in the pool outside of Active Range is getting exactly nothing for having their ETH in the pool at this time)

I think this idea has merit as a price support mechanism, but incentivizing the Desired Range actually accomplishes this because people would be incentivized to LP ETH in the tight range just under the Desired price (1:1).

ottodv · June 30, 2021, 5:29pm

It did, I have been following the situation quite closely and provided liquidity myself in the Desired Range, it was all sETH2 (because the current price was lower), hence there was just more sETH2 between the current price and the desired price.

Well, no because there simply will be no ETH in that range at all if the current price is below the Desired Range, as it currently is.

As above, there will be no ETH in the Desired Range at all when the current price is outside (below) the Desired Range.

Great! That means people will not be over incentivized to put their sETH2 on market, as that pushes the price down.

After all the ETH/sETH2 pool exists for the purpose of making the sale of sETH2 possible, at the least possible loss. The purpose is not to provide a nice discount for stakers and steer them away from depositing ETH in the staking contract.

thatexactmike · June 30, 2021, 6:25pm

I’ve been following closely as well and have been in and out of a number of positions. What you are saying here is simply not correct. If you added sETH2 in the Desired Range (close to 1:1) you would not be creating a barrier between the current price and the Desired price. It is adding sETH2 below the Desired Range that does that, and that is exactly what we are currently incentivizing (by incentivizing the Current Range more than the Desired Range). Also, it’s OK to have the Desired Range be all sETH2. That would mean as people swap ETH in for sETH2 the sETH2 gradually becomes ETH in the lower (left side) of the desired range.

And I am saying that is perfectly OK until the price rises to the Desired Price (1:1). At that point there would be an approximately equal amount of ETH and sETH2 in the Desired Range, all incentivized if we are incentivizing the Desired Range.

The pool has to have enough liquidity of both resources to serve it’s purposes (and that purpose is both efficient exit AND entry to staking) and be healthy pool that maintains it’s peg near 1:1, thus both sETH2 and ETH liquidity need to be incentivized. Sometimes there will be a discount for ETH and sometimes for sETH2 and market participants will take advantage of that for their purposes. But as a protocol it is in our interest in general to keep the price close to the 1:1 peg and incentivizing the Desired Range is one way to help do that.

thatexactmike · June 30, 2021, 7:25pm

For others who may be following along, what I’m proposing is that $SWISE farming incentives for providing liquidity to the sETH2/ETH pool be aligned to a Desired Range close to 1:1. That range could either a Narrow Desired Range, .994-1.006, or a Broad Desired Range, .9881-1.0121.

As you can see from the image below, the Narrow Desired Range would include 3 ticks and the Broad Desired Range would include 5 ticks. The middle tick of either of these is the 1:1 tick. The Broad Range would actually cover the Current Tick (.9881).

Benefits of this change would include elimination of the situation that currently incentivizes liquidity off-peg (I’m not proposing we eliminate Active Range incentivization entirely, but that we shift the higher percentage of incentivization to Desired Range) and would also potentially have the significant benefit of reducing the number of times LPs need to adjust their positions since the incentivized range would be tight around 1:1 along with liquidity so that risk of losses of sETH2 are reduced and risk of missing out on incentives is also reduced or eliminated.

ottodv · June 30, 2021, 7:47pm

I am afraid you’re the one who is “simply not correct”. The Desired Range starts below the peg. If it is a 1.2% range it starts at 0.994, if it’s 2.4% it starts at 0.988. There’s no ETH above those points and has never been.

It serves no purpose whatsoever to incentivize providing sETH2 as liquidity, if the price of sETH2 goes above 1:1 anyone would be happy to sell, because they can simply restake the ETH with the profit they earned at the parity rate using the Deposit mechanism of the staking pool.

Incentivizing sETH2 below parity, as you propose, simply ensures the price will never reach parity.

The purpose of this pool is literally only to make an exit from staking possible before the merge of ETH2 and ETH1. The better the rate at which one can exit, the more attractive the pool becomes for new stakers.

thatexactmike · June 30, 2021, 9:43pm

Thank you for continuing to engage. I apologize for my wording as it was not as objective or diplomatic as it should have been. I do, however, continue to disagree.

There’s more nuance to the situation with sETH2 than you are describing here.

First, if the price goes above 1 it will likely be due to a significant queue for staking to new validators, which is a possible outcome. In that case some may not wish to sell, even above 1 and some may wish to buy above 1 (with their ETH) to avoid the delay. Desired Range spans the range around 1:1, not just the .994 tick. I think (based on previous conversations here in the forum and in the Discord) that most Stakewise DAO participants expect and even desire the price to move around in a tight range above and below 1.

Second, incentivizing the Active Price right now is doing exactly what you are arguing against. It is incentivizing liquidity below peg.

Third, I disagree with your statement of the purpose of the pool. The pool actually serves a number of purposes for different market participants at different times. However, I am in agreement with you that a closer price to 1:1 is desireable for stakers. I disagree in how we achieve that.

Summarizing, I think the current situation actually demonstrates well the flaw in our current system. We have the greatest amount of liquidity at .9881 tick (over double any other tick) precisely because that is what we are currently incentivizing. The current system is actually building the wall of liquidity between current price and desired price that you are concerned about. Trying to visualize this, here’s the current situation:

Another way of looking at it that might clarify further, here is our current liquidity situation (in the range around 1). Only the first bar is incentivized and it is the Active tick. The total Liquidity pictured is about 5816 sETH2.

This is the distribution of that same amount of sETH2 if we incentivized the Narrow Desired Range around 1 (.994-1.006) and people responded by adjusting their positions across these 3 ticks:

This is the distribution of that same amount of sETH2 if we incentivized the Broad Desired Range around 1 (.9881-1.0121) and people responded by adjusting their positions across these 5 ticks:

ottodv · July 1, 2021, 5:51am

Likewise.

With all due respect it would help if you did not conflate the current system with what I propose.

Your graphic commentary is wrong, the part you labelled “No $SWISE Incentive” is in fact partly (the bottom part) incentivized by the current system. It literally blocks the ascent of the sETH2 price. And I do not see any reason to incentivize that.

The main thing that helps us in our goal is the blue liquidity right below the current price of 1.016, which prevents the price from slipping further by absorbing sales. Incidentally the only thing in your graph that you do not label.

The red bar you label as a wall is in fact not the wall you think. It’s the range between 1.012 and 1.018, which the current price of 1.016, you should understand that 2/3rd of that is actually sETH2 and not ETH. So it’s not much of a wall, except in the sense that there’s a 2k sETH2 wall in that range preventing the price from going up.

Yes, but the increase above parity would be tiny, if any. A 2 week queue would only justify a 0.2% premium at most. Yet given that the trading fee is 0.3% it wouldn’t make sense to buy an sETH2 at parity or above. Even with a long activation queue.

So again there’s definitely NO need to incentivize liquidity above parity, as it literally serves no purpose whatsoever.

In fact it would be just an easy way for every sETH2 holder to collect farming rewards for doing literally nothing. We should not be wasting the rewards we have on things that do not help the protocol.

There’s also no point to incentivize sETH2 liquidity below parity as all it does is increase the supply of sETH2 blocking the path back to parity.

There’s currently more that 5k sETH2 blocking the way back to parity, while there’s only 2k ETH supporting the current price.

It’s really not rocket science to understand which one is more important in our endeavor to return to the desired price point (or close to it). Hint it’s not the 5k sETH2 that is currently in the pool.

thatexactmike · July 1, 2021, 3:03pm

Thanks, I can see we are interpreting the data very differently. You seem to be focused on keeping the price from slipping to the left. I don’t perceive that as much of a danger and the way that liquidity in v3 works I think prevents that without an event that causes it externally as everything to the left of the current price is ETH (there’s no sETH2 to purchase there). We’d have to have a lot more sellers of sETH2 enter for that to be a risk. Market motivation right now is likely for holders of ETH to buy the remaining discounted sETH2. There’s not much reason that most of the remaining sETH2 in the pool that is being provided at .9881 couldn’t be provided at the .994 level instead to move the price upward more quickly. I’m more focused on bringing the price to 1:1 and the range immediately around that. Once that happens my proposal actually includes the building of the ETH wall at the range of .994 as that is part of what we would be incentivizing.

You mentioned that I’m not responding to your proposal. May I suggest that you have not made a complete proposal and I would actually welcome you posting your own Idea on another thread here with a more fully described mechanism that you are proposing. I’m not sure we are adding clarity here since we sometimes seem to be defining terms and perceiving priorities differently. Maybe there is an entirely different approach than what I am proposing that the community should consider and I think you should develop that out for conversation.

thatexactmike · July 1, 2021, 3:23pm

This sounds like you want to propose incentivizing liquidity only at exactly the 1:1 level. That is likely a very interesting proposal that might gain traction with the community. But that’s not what you’ve proposed previously in this thread. Please do post an idea here in the forum with your arguments for that. I think that would be a good discussion for us to have. It’s different from what I’ve proposed here, and from our current system (Active liquidity is incentivized as long as the active price is included in the range).

Just reviewing, here’s what you said when you initially stated you had changed your mind:

So that’s a bit different from what you most recently said above.

thatexactmike · July 1, 2021, 3:43pm

I understand that the ‘wall’ is a mix of ETH and sETH2. The sETH2 there is a barrier to moving to 1:1 as it all has to be swapped for ETH. Once all of it is swapped for ETH, at that point, THEN it is a wall of ETH to slow the price from dropping to the left. But until that happens it is a wall that is slowing the rise of the price of sETH2 to 1:1. What I’m proposing is incentivizing people to move their liquidity to the right towards 1:1. That would reduce the supply of sETH2 at the lower prices meaning fewer swaps of ETH for sETH2 required to bring the price up.

kiriyha · July 3, 2021, 3:10pm

Hey @thatexactmike and @ottodv, I’ve been following this discussion and am amazed at the thought and effort you guys put into the ideas for solving the peg. I believe your disagreement stems from the fact that both ideas aim to incentivize providing liquidity to achieve the goal of restoring the peg, yet this is simply impossible. At the same time, I think that both ideas have merit once the peg has been achieved, but (IMHO) do not offer much improvement over the current system (which also sucks when there is no peg btw) unless they’re modified slightly.

Let’s start with why incentivizing the growth of liquidity in order to fix the peg cannot work. The moment the peg is lost, any system that incentivizes putting liquidity into a particular range will enable more price stability (especially so with Uniswap V3), which is detrimental to the goal of returning to parity. In other words, with every incentivized unit of liquidity added to the pool the peg becomes more difficult to restore because the buyers need to sift through more sETH2 to get closer to 1. It doesn’t matter whether we incentivize the ETH part or the Desired Range - it will not restore the peg. Instead, it will either prevent the price from de-pegging further (with ETH in the Active Range or below) and/or make it more difficult to restore the peg (with a wall of sETH2 in the Desired Range whenever Active Range < Desired Range).

Hence, I hope you see that any discussions about incentivizing liquidity in one range or another are futile if they’re aimed at achieving the peg - it is simply not possible to restore it by incentivizing liquidity provision, and hence the argument is a little pointless.

If for a second we assume that the peg has been restored, the priority will shift to creating deep liquidity in a tight range around it. At that point, in my opinion, your ideas could both work, but will require a little bit of tweaking.

In case of @thatexactmike’s proposal, incentivizing liquidity in the Desired Range and setting it to a tight range around 1 means suscepting LPs to the outsized exposure to sETH2 unless the Desired Range is relatively wide. This is because most of the contributed ETH will sit in a tight range, offering very little slippage to the exitors, but also no discount and hence no incentive to go through the liquidity pool to the new stakers. Over time, this absence of two-sided movement may result in the protocol simply incentivizing the staking of sETH2 in the liquidity pool, which of course does not make sense, and will result in low trading fees for the LPs. The solution, of course, is to choose a wider range that would allow for a larger price movement. I believe Mike also recognizes the possibility to choose a wider range, so the idea may not even require tweaking after all - I would certainly explore an incentive like this.

Meanwhile, I believe that @ottodv’s proposal to incentivize people pro-rata to the amount of ETH they contribute to the Active Range is pretty much the same system that already exists, unless we consider it for incentivizing LPs in the range of ±1% below the current price and do it alongside the current system. Again, I believe Otto also mentioned this possibility at some point, so provided we have this tweak, I would be in favour.

With this in mind, I would reiterate that these strategies have little added value unless we return to parity first (or get very close to it). Perhaps, before we can reactivate this discussion, it is worth considering whether the protocol could incentivize the LPs to remove their liquidity and encourage the market participants to buy sETH2 until it is trading close to parity. Only then, I believe, we should re-focus on incentivizing liquidity. But that is just my 2c.

Looking forward to your thoughts on this guys!

ottodv · July 9, 2021, 10:17pm

There’s a big difference: the current system also incentivizes sETH2 in the active range.
My proposal is to stop incentivizing sETH2 completely (except they still get their rETH2 staking rewards as they would anyway even if they were not providing liquidity).

This mechanism discourages price drops and encourages price increases. Hence it slowly pushes the price back to parity.

ottodv · July 9, 2021, 11:36pm

These kind of strawmen are why these internet discussions are impossible (one example among many). The main “focus” I have is to remove the barriers that prevent the price moving towards what you would call the “right”. I want to stop incentivizing barriers to the right, whereas you want to continue to incentivize sETH2 liquidity to the right.

I have seen no convincing argument to incentivize sETH2 liquidity to the “right” of the current price.