[SWIP-3] Deploy sETH2/ETH and sETH2/rETH2 liquidity pools on Uniswap V3

Fellow DAO members,

Last week we published a proposal to deploy the liquidity pools for sETH2/ETH and rETH2/ETH on Uniswap V3. It has engaged the most people to date and sparked a lively discussion about the peg, arbitrage, trading fees, application of sETH2 in DeFi and other topics. We commend everyone who joined the conversation and present an updated version of the proposal that includes feedback to your comments.

In particular, we officially propose deploying the sETH2/ETH and sETH2/rETH2 pools on Uniswap V3 together with a proprietary farming contract that allocates the more rewards the tighter one’s liquidity is centered around the market price. We propose that the trading fees for the sETH2/ETH and sETH2/rETH2 pools be set to 0.3% and 0.05% respectively. The proposal about the size of farming incentives will be published separately.

Motivation

Let’s keep it simple: the liquidity pools for sETH2 and rETH2 would enable early exits from staking, compounding and integrations of sETH2 into the wider DeFi ecosystem, and deploying them on Uniswap V3 would allow StakeWise to achieve these objectives with ~5x less liquidity than was previously deemed necessary.

The majority of the commenters seemed to support the idea of setting a 0.3% trading fee in the sETH2/ETH pool as means to boost trading revenue for liquidity providers and hence attract more liquidity into the pool. The development team believes that this idea has good merit until the natural arbitrage opportunities emerge in Phase 2 (e.g. buying sETH2 at a discount and redeeming at par directly from the Pool). However, we expect that the DAO will likely be forced to migrate liquidity to a pool with a 0.05% trading fee after Phase 2 to maintain the liquidity pool’s viability as a bridge between sETH2 and ETH. For now, the DAO is likely better off setting a higher trading fee to ensure the LPs are well compensated for providing capital.

Now for the rETH2 liquidity. Upon reading @ottodv’s suggestion about setting up an sETH2/rETH2 pool instead of an rETH2/ETH pool, we consulted a prominent market maker about the idea and reached a conclusion that having this pool is indeed a far better choice. There are several advantages to doing this:

  1. An sETH2/rETH2 pool offers direct swaps between rewards and principal, which is a very attractive proposition for stakers looking to compound their staking rewards.
  2. Users withdrawing their rewards by selling rETH2 for ETH would contribute to the trading revenue in the main sETH2/ETH pool, as the rETH2 -> ETH trades would always go through it.
  3. Most anyone could provide liquidity in this pool, because the capital requirements are very low and all that is needed is sETH2 and rETH2 - the assets that are already in everyone’s possession. (@Skunk747 we hear your concerns about the high capital requirements to earn extra $SWISE from LPing in the sETH2/ETH pool, and this pool should address these concerns)

In order to maintain the feasibility of compounding, we suggest to set the trading fee in the sETH2/rETH2 pool to the lowest possible level of 0.05%. We look forward to your thoughts on the change of strategy for the rETH2 pool and the suggestion about the trading fee!

Finally, the “withdrawals” form suggests that the amount of ETH looking to exit the platform is within our previous expectations. Hence, we will reactivate the previous proposal to seed the sETH2/ETH liquidity pool with 0.6% of $SWISE per month in the separate forum post. Similarly, the proposal to seed the sETH2/rETH2 pool (formerly rETH2/ETH) with 0.2% $SWISE per month will be reactivated shortly.

Specifications

  1. Deploy sETH2/ETH liquidity pool with a 0.3% trading fee on Uniswap V3 and encourage provision of liquidity in a tight range by assigning more farming incentives to the LPs in tighter ranges
  2. Deploy sETH2/rETH2 liquidity pool with a 0.05% trading fee on Uniswap V3 and encourage provision of liquidity in a tight range by assigning more farming to the LPs in tighter ranges

Risks

One important consideration about the sETH2/rETH2 pool is that the farming incentives attached to this pool (and consequently rETH2 itself, as per this proposal) are likely to cause an appreciation in rETH2 vs sETH2 and ETH, owing to rETH2 being a rarer asset. Hence, LPs must be wary of setting too tight a range around the market price and should remember to keep the trade-off between the size of farming incentives and the risk of impermanent loss (i.e. liquidity becoming “inactive”) in mind. More on this is coming up in our guide for the Uniswap V3 LPs.

Vote

Should the StakeWise DAO deploy sETH2/ETH and sETH2/rETH2 pools on Uniswap V3?
  • Yes - with the trading fees set to 0.3% and 0.05% respectively
  • Yes - but with different trading fees
  • No - I do not support the sETH2/rETH2 pool idea
  • No - I do not support the deployment on Uniswap V3

0 voters

If you vote for an option that involves a “No” or a change in the trading fees, please leave a comment explaining your decision - it will help the DAO members to understand each others’ perspectives and will lead to the solution that benefits the most stakers.

4 Likes

Clarification please. 0.3% for sETH2/ETH pool and 0.05% for sETH2/rETH2 ? Thats how i read it atleast.

Thanks <3

EDIT: Nvm. I’m to stupid to read. It’s explained further down.

3 Likes

Vote is basically unanimous on that first option, including myself.

Great work to the team on constructing the proposal.

I glossed over otto’s reply to that thread but now that I see it, I find that an rETH2/sETH2 pool is an excellent idea. I want to think that the assumptions you make here are going to play out as expected @kiriyha!

I personally think they will, but after the pool is created, there’s some concerns on my part about people swapping rETH2 for sETH2, in particular large wallets.

The top 10-20 wallets holding sETH2 will accumulate rETH2 very fast. With the growth of the protocol (new stakers), and given large wallets wanting to keep growing their stake through compounding, or a lot of smaller wallets wanting to grow their stake, It concerns me that the rETH2/sETH2 pool could end up eventually being depleted, as it is the only place currently available to compound the stake.

Therefore while I agree we should create that pool ASAP, and in order to avoid the depletion of it, it would be great if there were other ways to compound rETH2 earnings, whether that’s now or in the future.

I bring up the possibility of it being eventually depleted because there’s no incentive to hold rETH2 at the moment (unless you pool sETH2/rETH2 and gain SWISE incentives), so I would find it very rare that people would come and swap their sETH2 for rETH2, so most trades will be very one sided, rETH2 --> sETH2, and will come at the cost of liquidity providers “losing” part of their staked ETH, sort of trading the earnings from staking for the earnings of the pool, which I’d guess will be fantastic during the first week, but not so much after a while.

I love this but I always think of the TORN airdrop distribution when I see LP incentives in the shape of governance tokens like SWISE, and I wish it was more common to consider the size of the contribution to the pool when assigning rewards to it. Not like an on-chain communism of sorts where everyone gets the same, but where rewards for monstruously large wallets could be tuned to fit a somewhat tuned log-like curve.

Obviously, the larger your contribution to the pool, the more you earn and the more you SHOULD earn, and sure, whales also like profits, and we all appreciate large liquidity. But given a long enough farming period and fast enough issuance, 2 things worry me:

  1. High inflation leading to price depreciation of SWISE.
  2. Large stakers accumulating SWISE a whole lot faster than smaller wallets (and either holding a larger stake in the protocol or dumping it on the market, nothing wrong with taking profits and de-risking, but this will inevitably lead to price depreciation). This will especially happen with the incentives of the rETH2 pool, as a huge stake --> faster acquisition of rETH2 --> a somewhat one sided pool (a single wallet holds 18.5% of the rETH2 supply, for example)

I think the selection of fees is great.

I have one last, unrelated question and that is, will there be a requirement to deposit LP token or will the rewards be distributed to wallets holding the LP token (i think it’s an NFT now for Uni V3?) corresponding to the specific pool position?


I still fully support this proposal and will vote yes on it, as this is very needed to grow the community and attract more stakers, but I want to see it play out better than how I imagine it.

If there’s anything I’m wrong about, as obviously this is all very speculative, I’d love to be corrected/cleared up.

3 Likes

Hey there, thanks so much for finding the time to outline your concerns and share ideas about how the distribution could be done!

I bring up the possibility of it being eventually depleted because there’s no incentive to hold rETH2 at the moment (unless you pool sETH2/rETH2 and gain SWISE incentives), so I would find it very rare that people would come and swap their sETH2 for rETH2, so most trades will be very one sided, rETH2 --> sETH2, and will come at the cost of liquidity providers “losing” part of their staked ETH, sort of trading the earnings from staking for the earnings of the pool, which I’d guess will be fantastic during the first week, but not so much after a while.

I think absent the incentive we have in the pipeline, namely that of assigning $SWISE rewards to the holders/stakers of rETH2 (likely tied to the insurance mechanism), this would be the case. However, if the incentive passes, it would turn rETH2 into a productive asset and make holding it worthwhile - not only it represents ETH, but it would also earn $SWISE to its holders. And since the supply of rETH2 remains low, we as a DAO 1. don’t need to allocate significant incentives to make it worth holding, 2. the APY from $SWISE incentives for holding it can still be high.

Hence, we believe to have this covered :slight_smile: In fact, the savviest of stakers will probably accumulate rETH2 in anticipation of this proposal passing, thus earning some handsome yields from $SWISE!

I love this but I always think of the TORN airdrop distribution when I see LP incentives in the shape of governance tokens like SWISE, and I wish it was more common to consider the size of the contribution to the pool when assigning rewards to it. Not like an on-chain communism of sorts where everyone gets the same, but where rewards for monstruously large wallets could be tuned to fit a somewhat tuned log-like curve.

I think adding some sort of a dampening factor to the rewards earned by the largest of wallets must be a decision taken by the DAO. I personally find it worth exploring!

Thank you for your support :heart:

3 Likes

Awesome, that does make me feel a lot more at ease with this. I’d almost forgotten about the slashing insurance plans!

I think I could produce a small writeup with the idea here on the forum, it’d require some math fun, so I might take a bit to write it, fortunately got some free time these days so I might take advantage of it.

and thank you Kirill!!

3 Likes

i voted no, i do not like the risk transfer elements from reth to $swise

4 Likes

i voted no, i do not like the risk transfer elements from reth to $swise

I was wondering where the “No” vote came from - thanks for commenting about this! Would you mind elaborating on your concerns a little bit?

Also, welcome to the StakeWise forum!

2 Likes

it is less of a concern about setting up a liquidity pool for reth but the implications of subsidizing reth with $swise incentives. id prefer any reth pool to not have $swise incentives tied to it. main concerns are below, happy to discuss further or offline as needed

  1. misalignment: providing $swise incentives for holding more reth reduces skin in the game and creates asymmetric governance. this gives more voting power to holders of reth over time instead of seth, which i believe creates an inherent conflict of interest

  2. liquidity: naturally the reth:seth pool will have limited liquidity given the ratio of rewards but i think adding tiered rewards for providing tighter liquidity of reth:seth will actually further reduce trading and therefore not provide the yield kicker this is intended to bring

  3. risk transfer: instead of reth trading like a zero coupon bond, this will artificially impact price discovery of reth. i think this could also result in selling pressure on $swise (full disclosure: no $swise position). depressing the value of governance to prop up reth value does not seem like a worthy trade-off

3 Likes

Regarding your concerns tthis migth be a good read.

Could sou elaborate on that part? I’m not sure I’m completely following. Thanks

2 Likes

@slizz_lord The points you make are reasonable, stay tuned for when @kiriyha writes that the proposal to add the incentives, that way you can explore it more in detail.

Keep in mind that these incentives might be temporary depending on the discussion and the DAO’s decision. I reckon that the sETH2/ETH pool holds priority in terms of incentives, so you make a valid point.

I agree with this, we would have to see if simply holding rETH2 is proposed to be incentivized. I personally think the perfect approach should be a way to somehow stake rETH2 without requiring to go through the liquidity pool, but I’m not sure how that could be done.

Also thanks for the shoutout to the thread @George \o/

2 Likes

@George thanks for the link. havent read in detail but not opposed to this. however, i dont think this accounts for the reth/swise misaligment concept i raised.

@George the risk transfer is conceptually a view that there is no free lunches. the return for reth is naturally implicit vs explicit. since it doesnt produce rewards on its own, the rewards must be discounted from the initial purchase price. i.e., someone buying 1 reth would pay (illustratively) 0.7 eth and their return is receiving 1 eth when eth2.0 becomes live. therefore, the buyer of reth takes into account the time duration of eth2.0 and their own discount rate. the seller would naturally take the opposite position; the view of staking 0.7eth generates a better return than holding reth to maturity. the ultimate pricing could be more complex than this if an investor factors in eth price volatility. adding $swise rewards for reth would (i) reduce this pricing outcome as $swise rewards would be factored in and (ii) a reth/seth holder may not want $swise and thereby sell their incentives on the open market and depress $swise value. therefore you add risk to $swise by de-risking reth.

i realize that is a lot so let me know if any is unclear.

@dreth @kiriyha is there any possibility of turning reth/seth pool into a traditional bid/ask market vs LP? i think the iceberg model might be better for price discovery for reth than a LP.

1 Like

Hey, thanks a lot for this! Let me address your concerns in the order of appearance:

misalignment: providing $swise incentives for holding more reth reduces skin in the game and creates asymmetric governance. this gives more voting power to holders of reth over time instead of seth, which i believe creates an inherent conflict of interest

I thought about it and cannot see how $SWISE allocations to rETH2 holders/stakers/LPs reduce skin in the game. Perhaps you are drawing this conclusion because we haven’t publicly considered one important aspect: rETH2 is the first in line to suffer from slashing, because the validators first lose a portion of the rewards, then principal. Hence, holding rETH2 may be considered a more dangerous endeavor for those who perceive sETH2 and rETH2 supply as tranches of the staked ETH pool. From this perspective, allowing rETH2 to earn some yield is desirable to avoid pricing according to the zero-coupon bond principles (I’ll come back to why this is important below).

With regards to concerns about asymmetric governance, I believe the DAO should determine the appropriate compensation levels for staking/LPing with each asset. My personal expectation (and the one that is outlined in the upcoming proposal about the incentives for pools) is that the amount of incentives should offer a decent yield based on the expected amount staked, but not more than necessary, in order to preserve the value of $SWISE and not give away too much voting power relatively to the importance of actions performed for the protocol. The significant difference between the supply of sETH2 and rETH2 must be reflected in the amount of incentives involved in staking/LPing with each, and that’s exactly what we should propose to the DAO. I hope this addresses your worry :slight_smile:

liquidity: naturally the reth:seth pool will have limited liquidity given the ratio of rewards but i think adding tiered rewards for providing tighter liquidity of reth:seth will actually further reduce trading and therefore not provide the yield kicker this is intended to bring

Hmm please pardon my ignorance, but why should tighter liquidity limit the trading volume? Sorry to not be helpful here, but I don’t understand :frowning: Would you mind elaborating further on this for me?

risk transfer: instead of reth trading like a zero coupon bond, this will artificially impact price discovery of reth. i think this could also result in selling pressure on $swise (full disclosure: no $swise position). depressing the value of governance to prop up reth value does not seem like a worthy trade-off

I agree with the last statement and it is the reason why we advise the DAO to not reward staking/LPing with rETH2 too lavishly. As for avoiding the zero-coupon bond dynamic, we believe it is actually desirable given the positive effect of compounding on staking returns. Of course, you have correctly identified that mitigating this dynamic requires subsidizing rETH2 with $SWISE, which should only be done in case:

  1. $SWISE can have a high-enough floor, and
  2. the amount of $SWISE incentives used to subsidize rETH2’s price is minimal.

With the $SWISE staking next in our development pipeline, we believe to have 1 covered. As for 2, it should be up to the DAO to keep the amounts paid high enough to make rETH2 attractive, but small enough not to dent the price of $SWISE. It’s an intricate balance, but one that must be found if we are to offer the compounding possibility to the users.

3 Likes

thanks for the response. i understand the point on slashing but this is only necessary action for turning reth into yield instrument. on a more fundamental level to me, $swise and seth are implicitly linked, but $swise and reth should not be linked. therefore to me, creating the $swise & reth linkage impacts the implicit linkage of $swise & seth. the risk i see here (while i recognize is very unlikely) is reth holders receiving governance could impact the staking fee at the detriment of seth holders.

liquidity: i dont see how incentivizing people to provide liquidity at prices that increasingly disconnect from fundamental value (i.e., Z-bond dynamics) will improve the liquidity of the market. im not super versed in liquidity pool dynamics so i could be missing something here but i dont see any benefit for a third-party to utilize this reth pool.

in summary, i think the concept of turning reth into a yield-bearing instrument requires the added concepts which you highlight in the proposal (and potentially more), but i tend to prefer the simplest solution which would be zbond pricing for reth. otherwise, you need to do significant mental gymnastics to identify, address and appropriately compensate for the incremental risk that is created. i realize im in the minority here and unlikely to ever change my views but think that risk sections of proposals should be more fulsome as most people probably focus on the positives without fully understanding risk

4 Likes

Hey, thank you for the extensive reply and sorry for disappearing from this thread. I think there is a simple compromise to be had between our POVs - owing to the supply of rETH2 vs sETH2 (tied to the network staking yield), the DAO is unlikely to ever give a significant voting power to the holders of rETH2 through the mechanism I previously described.

Meanwhile, the team also considers it important to counter the z-bond dynamics of unincentivized rETH2 because it doesn’t permit compounding, which is one of the advantages our dual token system can offer to stakers at large. To our mind, there is no risk attached to such incentivization as long as it is benefit of the protocol.

Would you mind elaborating more on your concerns as we prepare to flesh out the incentives mechanism attached to rETH2?