PIP - 36 : Multichain Staking and Gas Refund Consolidation

Governance Proposals
1

PIP - X : Multichain Staking and Gas Refund Consolidation

Abstract

Following the successful deployment of liquidity onto Optimism, the following proposal establishes a framework for Social Escrowed Staking of PSP outside of Ethereum Mainnet. Additionally, it suggests a way to make the Gas Refund Programme compatible with this new staking framework by changing it to distribute sePSP1 in Optimism as in Ethereum.

Goals & review

PSP staking has always been exclusive to Ethereum Mainnet. Initially, this limitation came from the nature of ParaSwapPools not being cross-chain, but afterward the main limitation was concerning liquidity, as part of the PSP 2.0 design was to make it as cross-chain compatible as possible.

Because of these limitations, many potential users have been unable to participate in the Social Escrow system, as their stake and potential rewards for their activities were not proportionally beneficial to the gas fees incurred. This proposal aims to address both of these identified problems by introducing a new multichain staking framework, and additionally, it proposes Optimism as the second chain where Social Escrow staking should be available.

Additionally, in preparation for this multichain staking ecosystem, this proposal also sets out to amend the current Gas Refund Programme to consolidate it along with sePSP staking. Currently, Ethereum mainnet receives sePSP1 while the rest of the chains where the GRP is available receive PSP as refund. Additionally, some chains currently do not have sufficient liquidity to be able to easily trade and bridge PSP. Both of these problems can easily be solved if multichain staking becomes available, as now all rewards could be claimed from either one of the new staking chains, giving users also a low-fee chain to claim them in.

Thanks to the recent developments of liquidity in Optimism, there is sufficient liquidity to allow staking to expand to a lower-fee Layer 2 rollup.

Means

Multichain Staking Framework

During the research stage for designing this framework, different solutions were identified for how multichain staking could be handled. For example, each chain and fees could be isolated for a simpler experience, or a single claim chain could be voted on by the DAO. To decide the best framework to build on, the following criteria were taken into consideration to ensure the best staking user experience:

  • Stakers have to be able to choose which chains to interact with.
  • For users who are still only staking on a single chain, there cannot be any differences compared to the prior experience.
  • The system should allow participation on multiple chains.

Following the aforementioned criteria, the best solution proposed was to move forth with a Proportional Staking Distribution system. In it, users can stake as per usual in different chains, with the difference happening at the point of distribution.

To decide how to split the rewards across different chains, the ParaBoost Score is calculated based on the added stake scores from all chains. Afterwards, rewards are split proportionally on each chain relative to the stake score of the user on each chain. For example, if 33% of someoneā€™s stake score is on Optimism and 67% on Ethereum mainnet, the rewards will be bridged and distributed following that ratio.

1137Ɨ1491 303 KB

Selection of Optimism

With over $1.1M USD in TVL as of the time of writing, Optimism is currently the best candidate for the next chain to implement staking on simply because of the liquidity available to allow for user onboarding onto Social Escrow. This is also a possibility that has previously been voted in the Optimism Deployment proposal, also marking previous approval for this move.

Due to the 80/20 flexibility of the pool, as well as development compatibility with what has been done in Ethereum mainnet, we propose a BeethovenX pool to be the best form of deployment for sePSP2 on optimism.

Pooling Boost clarification

As multiple sePSP2 pools are going to be available, the current pooling boost is going to become either incompatible or Ethereum mainnet exclusive. We would like stakers in Optimism to also be able to receive a Pooling Boost from participation, and thus an amendment to the current boost is required .

To remove this limitation, we propose moving from the current minimum being Balancer pool BPT denominated, and instead for it to be calculated based on the total amount of PSP pooled in all sePSP2 pools. As of the time of writing , 10 000 sePSP2 ( the pooling boost minimum), equates to approximately 60 000 PSP, and as such we propose this to be the new minimum value to be eligible for a pooling boost.

Deployment of Optimism Upgrade Module

Finally,to ensure that there is some liquidity for people to be able to upgrade their stake on Optimism, this proposal suggests the introduction of an sePSP1 upgrade module, similar to that in Mainnet. Due to staking being new on this chain (no migrations present) and lower gas costs, we expect the demand in this chain to be lower, so believe a good initial amount of liquidity for the safety module should be around 500 000 PSP.

Gas Refund Consolidation

To ensure the scalability of the system and avoid further confusion on the user experience, this proposal also wishes to amend the way the GRP is currently working.

Rather than running the Gas Refund Programme only on chains PSP is (and claims being done on each chain individually), the new amended version will send all available refunds proportionally to sePSP1 enabled chains, in a way similar to that of how staking rewards are being proposed to be split.

If implemented this way, the new version of the GRP also removes the existing limitation of Gas Refunds being doable only on chains PSP is deployed in. Because of this, in addition to the gas from existing chains being claimable from Ethereum and Optimism, we propose for all technically viable chains to be included in the GRP following a vote for it.

Implementation Overview

The following steps will have to be implemented for this proposal to be needed for this to come to fruition:

  • Creation of an 80/20 PSP/ETH pool on BeethovenX in Optimism
  • Amendment of Gas Refund Programme to only refund on staking chains in the form of sePSP1 the total eligible refunds
  • Deployment of sePSP1 and sePSP2 tokens on Optimism
  • Amendment of the Pooling Boost minimum from 10 000 BPT of sePSP2 to 60 000 PSP staked across all chains
  • Creation of an sePSP1 upgrade module with 500k PSP

As the proposal requires two major implementations, we suggest the following steps for deployment of these initiatives:

  1. Deployment of staking pools and staking on Optimism
    1.5 Amendment of Optimism gas refund rewards from PSP to sePSP
  2. Re-direction of existing gas refunds for all eligible chains to Optimism and Ethereum, and alignment of GRP distribution relative to sePSP reward distribution
  3. Addition of viable, but currently not present chains onto the GRP system with future governance votes

Voting Options

  • For

  • Against

  • Abstain

7 Likes
2

Great proposal, quality work as always.

My only concern is that opening up staking to other chains will, I think, reduce liquidity on Ethereum.
Right now liquidity on Ethereum (Balancer) is a $4.85M, do we consider it enough?
How much of that liquidity is going to be transfered from Ethereum to Optimism?

One of the main achievement of the Social Escrow model is that ParaSwap has managed to build a strong liquidity, which is one of the biggest challenges for a DeFi protocol.

Are we OK to give up on that achievement in favor to promote another chain?
Can we plan on adding more PoL to compensate that probable lost of liquidity on Ethereum?

3 Likes
3

The timing of this proposal couldnā€™t have been better; It came when I personally was preparing to propose it.

More importantly, this proposal will allow retail investors to experience Social Escrow which I beleive is in important for mass adoption.

I support every aspect of this proposal except the proposition to suspend GRP on non-stakibg chains. While it makes sense to refund gas in SePSP1 (as it currently stands), I do not see the reason why deployment of staking on Optimism should impede payment of refund in other chains. Most of the refund budget goes to Ethereum mainnet so itā€™s not as if we are spending excessively on any if the other chains. While the option to stake on layer 2 is expected to attract retail investors, suspension of GRP on other chains could make us loose some users (especially those who trade heavily on other chains).

To keep things simple, I propose the base criteria to be eligible for GRP should be SePSP2 staking in either Ethereum mainnet or Optimism with a minimum Parascore of 10k. This should qualify any user to get a min refund of 25% across every supported chain to be paid in SePSP1.

Edits:
What if we eliminate claiming for the GRP all together. Is it possible to auto-claim and autocompound the GRP?

1 Like
4

You have raised a valid concern and I share the same sentiment. But, I assume non of the liquidity we currently have on Ethereum will be transfered to Optimisms. Instead, the DAO shall find a means to bootsrap new liquidity on Optimism. Thereā€™s a mention of PSP/Eth pool on Beethoven but there was no mention of source of funding.

5

A proposal that makes sense for our small shareholders.
Do we have any idea how many addresses would be interested in this proposal?

Indeed, it is possible that a small number of addresses will be represented. In that case, ETH liquidity wonā€™t be significantly impacted, leaving room for new ETH entrants without too much slippage. As a bonus, weā€™ll be opening up to the optimism market. All in all, a win-win situation.

Conversely, if small addresses account for a large proportion of liquidity, weā€™ll simply transfer liquidity from eth to Opti.

As Enerow points out, liquidity on ETH is high and the current TVL is more than comfortable for the time being.

If we see too much leakage of liquidity. Perhaps we could think about equivalent compensation using DAO ETH.
Let me explain: we accept a 10% loss of liquidity on ETH and below that, the liquidity will be compensated by the DAO on the 80/20 balancer pool. This is just an example

1 Like
6

I had the same concerns when thinking about the proposal, but ended up noticing that thereā€™s too many factors that could tip overall liquidity one way or another, hereā€™s a few I thought about. For example, we cannot really estimate how many people are not staking sePSP2 right now because of gas concerns that could add new net liquidity onto Optimism ( not just because of the gas intensiveness of sePSP2, but also the routine claims).

Because of this, I believe that the best way to go about it is to begin staking with just a single othrer chain, and then see if liquidity becomes truly imbalanced long term. If it does, some ideas I thought of were:

  • Participation on the veBAL gauge distributions , with rewads being given proportionally to fees generated on each pool (that way stakers in more profitable chains receive more rewards).
  • Boosts for staking in under-pooled chains (more volume being made relative to its liquidity): Not a fan of this one, as it will add complications and could break Rule 2 of the multichain
  • Finally,as you mentioned, additional PoL injections could be explored. We could deploy either more sePSP2 liquidity or a Uniswap V3 targeted range.

As explained in the original proposal, the reason is not technical, but rather to resolve the current mess that has happened due to the current gas refund predating our new liquidity and staking plans. If left untouched, the current system will have chains where staking is not available but refunds are, chains where you receive sePSP1 as opposed to PSP, chains where there is liquidity vs chains that there arenā€™t, etc.

Consolidation is important to be able to keep the current staking system simplified and accessible. With this new system, itā€™s as simple as ā€˜stake,and get already-staked gas refundsā€™!

sePSP1 refunds already count for your staking score and thus auto-compound it, even without claiming first!

Agreed on this, worst case scenario the DAO can mitigate TVL segmentation with solutions like the ones stated above. But imo we can start exploring this once multichain staking goes live!

2 Likes
7

I have mixed feelings.

The idea of including it in your proposal would help prevent abuse of DAO bailouts. It would lay a clear foundation for when the DAO should implement the rebalancing system.

On the other hand, it goes without saying that the aim is not to allocate all DAO eth, so we could also attach a maximum.

Otherwise, I remain convinced of the need for your proposal. Congratulations again.

1 Like
8

Iā€™m not sure to understand why we canā€™t keep 10k sePSP2 as the requirement for Paraboost, as it provides more value to the DAO/to stabilize the price to provide LPs rather than raw PSP no? Why canā€™t we set the same kind of pools for Beethoven and Balancer, both should stayed synced? Iā€™m afraid it will create confusion as price fluctuates to establish a relatively fixed amount of PSP.

Thus, gas refunds on non-staking chains will be suspended until the DAO will approve staking on those chains as well.

Iā€™m not sure to get that part, enabling sePSP1 instead of PSP reward on staking chains is good. But that is not required to disable other chains gas refunds. Previous drop in gas refunds already decreased volumes and fees for the DAO, so Iā€™m not sure that makes sense at all, especially in a proposal that mostly targets multi chain staking. Because except mainnet that would mean stop all the other gas refunds programs: Polygon, Ftm, Bsc. So definitely opposed to this specific brutal modification, as staking on these chains isnā€™t a necessity at all, it will just end these programs with very little notice or good reasons. To add to this argument, the gas refund isnā€™t even enabled on Optimism, which is the main target here, so it shouldnā€™t be part of this proposal.

Anyway, great proposal, definitely agreeing that staking on another chain is needed to improve a broader PSP adoption. However some parts should be discussed separately, the gas refund program being altered for invalid reasons IMO.

Cheers

3 Likes
9

Sure itā€™s important to be clear, but staking on another chain was important and thatā€™s why this proposal on Optimism is released.
Gas refund is currently enabled on mainnet, Polygon, BSC, Ftm. This will just remove gas refund from Polygon, Bsc, Ftm without real reasons than consolidation?
I donā€™t know I feel like weā€™re losing ourselves, users wanted more features because small holders couldnā€™t stake. Thatā€™s the solution to this, why over-complexify this and completely remove Gas refund programs on 3 chains? I donā€™t even find it obvious that a gas refund is gonna be introduced in this proposal for Optimism, so I donā€™t find itā€™s a consolidation.

I think we should keep adding new features rather than invalidating current programs that work well.
Refund program has been enabled on specific chains from the beginning, as staking. And thatā€™s fine for everyone I think.

2 Likes
10

I agree with you.

For Epoch 5 for exemple
$PSP refunded on BNB/Polygon/Fantom represent only about 2% of the total.
But it represent ~43% of the addresses and ~60% of the transactions.

Very few $PSP saved but it will affect half of the participants.

2 Likes
11

If anyone from the DAO proposes a number that they are willing to discuss for this budget, I am open to including it in the final proposal! I just personally believe itā€™s not worth delaying this highly requested framework over a theoretical scenario where a lot of liquidity migrates to an L2, but somehow the token volume does not follow. If it does happen, then we will have specific data to assess, but right now itā€™s difficult to budget with so many theoreticals.

In the current state, the sePSP2 used to calculate the minimum 10k is relative to the BPT value of this pool. Due to the nature of Balancer Pool Tokens, this price is dependent on the pool size , and 1 BPT on Optimism will not be the same as 1 BPT on Ethereum. This means that the idea of measuring a total of 10 000 BPTs across chains is not possible, as the two tokens are not equivalent.

Even if we wanted to ignore that both BPTs are not the same, this would add an additional layer of confusion that even users familiar with DeFi struggle to understand. Many users already were confused in the current state, as they believed 10k sePSP2 referred to 10 000 PSP in sePSP2, by also having to now manually calculate the value of 1 BPT across every chain staking is available in (and this value changing on the daily as liquidity fluctuates), it would add yet another layer of complexity.

In contrast to that, by moving to a total PSP value, this proposal hopes to make it understandable how much PSP they need to have staked across any number of chains to make them eligible. Even if the pooling boost is lost as someone drops below 60 000 PSP, the problem can be intuitively resolved by adding up the total number of pooled tokens across all chains.

As explained in the proposal and some messages above, the current situation where the old Gas Refund Programme had to be made compatible with the new PSP 2.0 system lead to many significant issues and complications. With only a single staking chain most of these issues could be brushed aside, as the explanation was simply ā€˜stake in mainnet, receive sePSP1 in mainnet, PSP in other networksā€™, but with the introduction of a scalable multichain staking network, the rules become much more complicated and borderline inaccessible to understand for new stakers. Hereā€™s a few quirks that will be coming up:

  • Some chains will give you an auto-staked form of PSP with a 30 day cooldown, while others will give you a liquid version. If you are unfamiliar with the benefits of sePSP1 , you will wonder about why if you are for example staking in Ethereum you got sePSP1 in optimism but just PSP in another chain.
  • Some chains allow for staking, while others do not. As more chains become available, this becomes even spottier.
  • Distribution and automation issues : Instead of sending all assets only to staking enabled chains, some will have to be sent to staking chains, while sePSP1 will have to be sent to staking chains and PSP to non-staking chains. This is not seen from the end-user perspective, but adds even more complications for very little gas.
  • These chains do not have significant liquidity, meaning that even if the refunds exist now in theory, in practice you cannot swap more than 10 USD in chains like BSC and Fantom. If a chain receives sePSP2 this issue gets mitigated, which means that both issues will be solved by enabling staking on those chains. A targeted chain-by-chain proposal to re-enable staking will also allow for better liquidity strategies tailored to each chainā€™s situation.

For these reasons, this proposal considers consolidation necessary for a multichain staking framework. If you have an idea on how to resolve these issues while removing the entry-level complications and splitting liquidity across multiple chains Iā€™m interested to hear it!

1 Like
12

A very interesting proposal to enable PSP2.0 to reach as many people as possible.
And as you can see from the initial reactions, the main body of the proposal has so far met with consensus!

On the risk of fragmentation of liquidity
I donā€™t have a crystal ball, my opinion is that I donā€™t see a large wave of existing TVL switching to Optimism for two reasons:

  • Large portfolios favour mainnet (imo)
  • For small wallets, the unstake and bridge fees would take a long time to recover if only the difference in the cost of the claim makes the difference. It is much better to stay on Ethereum and optimise claims with the new 6 epoch period rather than transferring to Optimism.

On the other hand, perhaps new stakers will favour Optimism and the growth of the Ethereum pool will be slowed.
(Yes, I know Iā€™m theorising a lot for someone who doesnā€™t have a crystal ball).

On the tools to rebalance liquidity if necessary.
Even if balancers gauge can be useful in their own way to increase pool returns, Iā€™m not convinced that they can be a tool for arbitraging and rebalancing liquidity. The cost of changing pools and networks seems daunting.
Only the addition of liquidity owned by the protocol seems to me to be a coherent way of rebalancing pools in extreme cases (even if I donā€™t believe the opportunity will present itself).
Should we be planning for this type of contingency right now? I donā€™t think so. But itā€™s mainly linked to whatā€™s described above. And my predictions are certainly no guarantee.

On the GRP
Iā€™m going to have a rather clear-cut opinion. The Social Escrow is too complicated, it shouldnā€™t be made more complicated, it should be simplified to attract more people.
Iā€™m therefore of the opinion that GRP should only be activated on channels that have staking. This way we can have a simple rule: GRP = sePSP1.
I understand that this could be seen as a regression in terms of the service provided, however, as explained above, GRP on the chains mentioned costs very little $PSP but reaches a lot of participants. I just have another interpretation of the same figures: given the very low fees for these chains, is GRP really an argument for using ParaSwap on these chains?

1 Like
13

For Epoch 5 alone :

  • Transactions refunded by the GRP made on Polygon + Fantom + BNB (5912 tx) represent 60% of all chains transactions (9708 tx)
  • Wallets participating in the GRP on Polygon + Fantom + BNB (412 addresses) represent 43% of all chains wallets (957 addresses)

Do we want to exclude off the GRP 60% of the transactions and 43% of the addresses for the sake of simplification?

2 Likes
14

As mentioned a couple posts above, itā€™s not just about simplification, itā€™s about the current state of the GRP not being compatible with multichain staking and accessibility.

Of the numbers you mentioned above , 2860 txā€™s among 128 addresses received a total of around 15k PSP in chains where selling just 10 PSP can result in a 42% (BNB Chain) and 20% (Fantom) gas impact. These means that these hundreds of addresses are receiving a very negative experience and perception of the Gas Refund, having to stake in another chain to get illiquid rewards.

As always, Iā€™m more than open to constructively discussing alternative solutions on how we can make the GRP compatible with multichain staking, but in the current state the system cannot be left untouched in my opinion, as for many users it is simply creating an inaccessible experience that once understood gives rewards that cannot be even used on that chain. In exchange for leaving these ā€˜legacyā€™ systems on, many users are having a subpar experience with the GRP and staking in general.

Letā€™s focus on consolidating GRP with sePSP1, and then work as fast as possible to make the experience enjoyable on those chains in a scalable and Social-Escrow focussed way :smile:

1 Like
15

Youā€™re right and thatā€™s why I think we should start talking about adding PoL on those chains to improve the experience for those 43% of participants instead of excluding them, until we add staking support (sePSP1/sePSP2) on those chains.

1 Like
16

While this is a much needed idea that can be written in another proposal (feel free to create a proposal to discuss it!), this still wonā€™t solve the issue of sePSP1/PSP rewards difference, stake-chain compatibility and distribution automation .

I know these do not seem like major problems for users who are already part of the staking system, but after multiple rounds of feedback since the release of PSP 2.0 weā€™ve been consistently receiving confusion and complaints about how to stake to get rewards on those chains. If there is sufficient demand for a chain to receive gas refunds, a proposal can be made to introduce staking there and remove this incompatibility between the two systems while keeping users active.

Finally, it might be misleading to say that this is excluding 43% of participants, many of these addresses interact with multiple chains and a lot of double counting will be going on if we assume every chain has only unique users.

1 Like
17

On the GRP
A step backwards on this subject, I think I underestimated the impact that removing the GRP could have on non-staking chains. On rethinking it, the removal of the benefits of some users could have more disadvantages than the advantages provided by the simplification of the system.
Sleep on it is often useful (:

2 Likes
18

Letā€™s assess the risk / reward

Maximum Risk : Loosing about 60% of the transactions and 43% of the users
Maximum Rewards : Saving 2% PSP + Making the system easier to understand

I think our goal should be to extend the GRP to more and more chains to attract new users not the other-way around.

2 Likes
19

Itā€™s great to read the differing opinions of Chab and Enerow.
They do raise a very important point.

I donā€™t think itā€™s possible to go back on the GRP of these chains, even if itā€™s done in PSP.
You have to put yourself in these usersā€™ heads. They are making volume on these chains and therefore quite naturally receiving refunds. They may be illiquid today, but they have the merit of existing and growing in their wallet.

What about the day when PSP explodes and liquidity naturally opens up on all chains?
How will you justify the fact that these users, who have always been present on these chains, have not been earning PSP GRP for some time now? How will you justify the loss you have caused them by simplifying a system that was working properly?

What we have here is a problem of acquisitivenessā€¦ If you take something away from a group of people without offering something similar or even better in return, then youā€™re bound to come into conflict with that group and the outcome is bound to be negative for the product, i.e. Paraswap.

In my opinion, the best idea is as follows:
We deploy your staking idea on optimism but the refund programme continues to benefit all chains, as it exists today.

1 Like
20

Thanks @0xYtocin for this proposal and also others for the ideas :sweat_smile:

I totally that we should simplify the SE model and GRP. Currently, the only reason PSPs exist on other networks is to reward users for the fees they pay and accept the fact that they dump them.
First, Iā€™m wondering from the numbers provided by enerow, What are the % of wallet that really care about the SE model (impossible to know but we can link some wallet with their mainnet stake)?

Regarding the proposal, Iā€™m not in favor to provide the staking system on all chain because:

  • liquidity could exist mainnet or be frozen, as mentioned above. In my opinion, the price impact for buying PSP is to high due to the 80/20 pool, I donā€™t see how this minimum of 2% could bring interest market maker, and whales.
  • I think non advanced SE user will be lost.
  • regarding the pooling boost:
    • for mimic to align the fees it will be a lot more costly in term of gas and maybe additional setup.
    • for govCo if we bring staking on other chain that means they will have to do additional multisig signature to distribute rewards => bring potential new errors.
    • for the protocol to calculate the rewards of each user it will be really complexe.

I think we should keep it simple as ā€œhelloā€.
What if we are moving back and make PSP only available on mainnet as pooling & GRP are calculated off-chain, we could :

  • rewarding the user directly on mainnet with sePSP1 (as they is no deadline itā€™s perfect)
  • add a new deposit feature using meta-tx / middleware escrow and allow someone else to pay the gas for the deposit in the 80/20 + staking for the user in exchange of a % paraboost :slight_smile: (hello Social Escrow).

PS: stakeDAO doing a similar thing that my last point for their staking contract but in exchange of 0.1% of fees

With this 2 points above, we keep everything consistent and allow user on other chain to access to the party

2 Likes