JIP 1: Expanding Validator Set; Protocol Development

[andrewt]

1. JIP 1: Expanding Validator Set; Protocol Development

2. Abstract

The Jito Foundation is proposing expanding the Jito Stake Pool validator set to 200 as well as capping the maximum permissible validator commission for Stake Pool participants at 5%.

3. Motivation

In October 2023, the Jito Network began leveraging a transparent, algorithmic process for selecting validators to participate in the Jito Stake Pool validator set. This change enabled validators to measure their status against other validators in order to know how to improve their chances of receiving Stake Pool delegation, and was the first step towards the forthcoming launch of StakeNet, which will enable Jito governance to monitor and adjust the Stake Pool inclusion parameters in a more fully decentralized manner.

Since then, activity on Solana has skyrocketed and JitoSOL TVL has likewise more than doubled to a total of 8.3m TVL as of today. Additionally, since Jan 1st, the Jito Network’s TVL has grown 23% in SOL terms despite the absence of a points incentive program.

The number of high-quality, high-performing validators has likewise increased. When the Jito Network expanded the validator set in October, it was unclear if there were 100 validators who would be capable of meeting the requirements; now, there are far more high-performant validators for the algorithm to choose from. See this spreadsheet that tracks the eligible JitoSOL validators and the increase in quality of candidates since epoch 480.

Additionally, certain large stakers are considering expanding their amount allocated to stake pools, but only if their stake can be sent to a larger and more decentralized validator set. In order to accommodate the growing allocation, the stake pool must further decentralize.

Given the expansion of the Jito Network’s TVL, the growth in the number of quality validators, and interest from large staking parties, now is an ideal time to expand the validator set and lower validator commissions.

4. Key Terms

Stake Pool Validator Set: The group of validators responsible for receiving delegation from the Jito Network and maintaining the quality and returns of the underlying liquid staking token.

Liquid staking token: A liquid staking token is a type of cryptocurrency token that represents a stake in a blockchain network but, unlike traditional staking methods, allows the staker to retain liquidity.

Validator Commission: The current validator commission cap is 10%, and the average fee for Stake Pool participants is currently 1.54%.

Churn: In the context of the Jito validator set, churn refers to the process of replacing poorly performing validators with new or better-performing ones over time. It is a dynamic process that ensures the network remains secure, efficient, and decentralized by continually optimizing the validator set.

5. Specification

The Jito Foundation is proposing two changes to the algorithm for determining Stake Pool delegation. These changes would be executed by the Jito Labs team immediately following passing the governance process.

First, the Jito Foundation is proposing to expand the number of eligible validators to receive delegation to 200. Additionally, we are proposing imposing a cap on validator commission at 5%, a decrease from the current 10% that better reflects current competitive dynamics.

The Foundation believes that these two metrics – the number of eligible validators and the validator commission cap – are innately linked and should be addressed in the same JIP. Because more validators will be entering the Stake Pool, the average commission may rise as well, as tail-end validators tend to have higher commission rates. This new cap allows tail-end validators to maintain higher commission rates while still being included in the pool, without raising them so high as to potentially impact the performance of JitoSOL.

Additionally, the commission cap also acts as a guardrail to ensure performance doesn’t suffer if there are changes to the potential validator set during the process of churn. Without it, if some validators became ineligible for various reasons, the pool would be forced to delegate to the top 200 validators even if they had commissions at 8% or much higher. This could have a significant impact on JitoSOL yield and potentially TVL.

This proposal does not impact any other delegation parameters, including superminority exclusion, the MEV commission cap of 10%, running the Jito-Solana client, vote credit averages, performance weighting, or the rebalancing formula. Users can read more about current Jito delegation parameters here.

6. Benefits/Risks

When considering all members of the Jito community, in the short term this proposal most benefits both retail and institutional JitoSOL stakers because it has the potential to increase both decentralization and validator diversification.

Meanwhile, this proposal may decrease the amount of stake validators receive in the short term. However, it significantly increases the decentralization of the validator set and sets the backdrop for further growth of the protocol – an effect that will ultimately lift all boats and benefit all validators on an intermediate-to-long time horizon.

We anticipate that the proposal will lower the average stake validators are currently receiving. At current, >20% protocol growth rates, top-performing validators will reclaim their pre-expansion allocation in SOL terms in under a year.

We also believe this proposal benefits smaller validators, in that they will have a stronger chance of entering the Stake Pool.

Ultimately, over time we believe that this proposal benefits all members of the Jito community.

7. Outcomes

As of today, there are effectively 126 validators receiving significant allocation from the Jito Stake Pool. The October 2023 proposal targeted a validator set of 100, but due to churn – less performant validators being cycled out and new entrants coming in, a phased process that takes place over two or more epochs – the amount of validators receiving delegation often exceeds that target.

Based on churn, the proposed expanded delegation strategy will not always align with the targeted 200 validators, but with steady validator performance will drift towards this targeted number of validators over time, ultimately increasing the number of validators in the Stake Pool set to above 200 validators.

Meanwhile, the maximum current allowable validator commission rate is currently 10%. However, among the top 200 no validator exceeds 5%, and among the top 100 rarely does it exceed 4%. To match this new competitive dynamic, the proposed 5% cap will only have effect on the tail-end of validators in the Stake Pool set.

8. Cost Summary

This proposal will incur no cost to the DAO.

[0xNallok]

1. Is there any economic analysis on the impact to performance of the jitoSOL pool by enforcing this? My guess is the top 100 (as provided by the average) are well below the 10% mark, this would just reduce the overall set which could qualify?

2. Any thoughts on keeping the commission standard in line with SFDP re: commission? Such that operators wouldn’t need to juggle the two? I see some pros and cons with that, but might be worth exploring with SF so that coming into Solana as a validator isn’t the same minefield as it is today. I’m 50/50 on the value this brings, but all in favor of ensuring there’s healthy competition not just pockets of isolated information.

3. Any thoughts on reducing the MEV commission to bring it in line with the validator commission?

4. I’d like to propose in this a regular review of these standards and processes, such that it is expected at some cadence or at certain thresholds for SOL staked (or unstaked). This is likely something which has both but should be established as a process undertaken by the Jito Foundation and governance.

Overall I agree with this JIP and believe the capacity to dynamically adjust to current market conditions is paramount to the success of Jito Foundation and jitoSOL.

[laine]

Thanks for the proposal

I see a detailed explanation and reasoning for the expansion to 200 validators, though I feel the term decentralization is used a little excessively where it would be better expressed through “breadth” or “size”. E.g.

@andrewt

it has the potential to increase both decentralization and validator diversification.

The proposal doesn’t directly touch on factors most commonly associated with decentralization or diversification, such as geolocation or stake concentration, but rather purely on the empirical size of the set. It does not change the measure of diversification of that set beyond the existing parameters.

That being said, an expansion to 200 validators seems appropriate for the pool’s current TVL and is a welcome move, for its ability to support more validators and further broaden the pool’s impact on the network.

However, I don’t believe there is a explanation or justification for the change in max commission from 10% to 5%. This seems like a somewhat unrelated change that is conflated with the separate expansion of the set.

It is important for a pool to be competitive, but to also wield its power responsibly, where it is able to pressure validators into certain behaviours. Therefore such a change in max commission should be fully understood and examined as to the broader implications for the network, Jito being one of many stakers to most of these validators.

Rewards are volatile and frequently changing. While they are currently generally larger across the board (block, staking & MEV), this may not hold true. And for smaller validators in particular, especially those with <100k stake, this sort of reduction in commission may materially affect their ability to operate a profitable business.

All this is to say that a change in max commission is not something I outright object, but I don’t believe this proposal suitably explains the reasoning for it, and the value, without context, appear arbitrary. In general, a step-wise comission cap might be more welcome, where validators with lower commission might receive an additional bonus, or where validators above a certain total stake threshold must maintain a lower commission, to further assist smaller validators in their initial journey to sustainability.

I feel given the above the commission change should either be moved to a separate JIP (1st preferance) or this JIP should be expanded to appropriately explain the reasoning, motivation and evidence for it.

Last nit, this definition is incorrect, commission is the share of staking rewards the validator receives, there is no connection to block rewards whatsoever.

@andrewt

Validator Commission: Refers to the percentage fee of block rewards that validators take from their delegated SOL stake.

I agree with 0xNallok that a regular review should be mandated, for example either with every 25% increase or reduction in Jito TVL (from the value when this JIP passes) or on a semi-annual basis perhaps.

[7Layer]

I think the 5% commission cap is good and still pretty generous. Most validators would have a hard time attracting stake with more than that and I think it makes sense to have a commission cap that keeps stakepool APY competitive.

Having 200 validators also seems fine since vote credit performance is basically random these days outside of downtime and a cutoff at 100 versus 200 probably won’t yield too much of a difference. Looking forward to more criteria related to geographical decentralization in the future though, perhaps easier when there are more block engines.

[Chainflow]

I see the initiative as a strategic step towards enhancing network resilience and promoting equitable participation. Expanding the validator set to 200 and capping the maximum commission at 5% could foster greater decentralization, ensuring a wider distribution of governance and operational responsibilities across a more diverse set of validators. This approach could also make the network more attractive to a broader range of stakeholders by lowering entry barriers for validators and offering a more competitive and fair compensation model.

However, it’s crucial to carefully consider the impact on existing validators, particularly in terms of stake distribution and potential short-term decreases in earnings. Balancing these changes with the network’s long-term growth and the overall health of the ecosystem will be essential. Ensuring transparent communication and providing adequate support and resources for validators to adjust to these changes will be vital for maintaining network stability and validator morale.

While the proposal presents a positive direction toward decentralization and fairness, thorough analysis and strategic implementation are necessary to mitigate risks and ensure the proposal achieves its intended benefits for the Jito community. I see some comments questioning the reasoning behind the 5% cap and these concerns can be addressed potentially by financial models or previous experiences the team is pulling from to make this choice. Personally I am okay with the 5% cap for now, this can always change. As long as we have metrics(qualitative/quantitative) we can monitor to measure the effects of this cap. The more collective information we have the better decisions we can make as a community.

[mrknc]

Laine has covered it pretty well. Another nitpick I have is, since we are talking about validators, it would be good to have performance score formula specified exactly, as opposed to how it is stated now - “adjusted for commission”

[ValidBlocks]

Overall the proposal looks fine. The commission cap is senseless in my opinion since only 2 valdiators in top200 have >5% commission (based on last snapshot)

Instead of the commission cap I would include max stake cap if the idea is to support small validators. I don’t see why a >1M staked validator would need this kind of “help” since is a self sustaining business by far anyway while other small validators are struggling.

Stake distribution in top200 currently (last snapshot) looks like this:

13 validators over 1M
8 validators between 1M and 500k
58 validators between 500k and 200k
53 validators between 200k and 100k
69 validators under 100k

In the proposal I see a lot of emphasis on decentralization but by just expanding the set is not enough, I would include other metrics like geolocalization (by ASO/City for instance).

I’m glad to see this proposal live, it was much needed.

[richtedros]

Newbie here. Some questions for Consideration:

1.	How will you support smaller validators during this transition?
2.	What measures will be implemented to monitor the effects of these changes on network health and security?
3.	Are there plans to adjust these parameters based on future network growth and feedback?

[Goldberg]

For participants of the Solana foundation delegation program nothing changes through this proposal. They will receive half the delegation from Jito which should be doubled by foundation shortly. They can not afford to loose their spot in Jito so commission stays across the board at 0%.

From the perspective of the outside delegator it seems irrational to pick a validator with more than 0% commission if 20% of the network chose to go that low. Expanding the jito set would therefore fuel the race to the bottom.

Since validators depend more strictly on MEV rewards in that environment, the Jito client may take over too much market share and potentially slow down the network - if Anatoly’s criticism of the way its harmful by introducing the mempool through the back door is valid.

As a user I have to fight with congestion every day recently and more frequently in general. I think we need to be very responsible with incentives and respectful to the health of the network if there’s a doubt on our effect on performance and stability.

[andrewt]

Thanks to @0xNallok and @Laine in particular for their questions about the commission cap.

In our view, this would be a change primarily designed to match – rather than modify – current validator behavior, and we believe the increase in cap and commission rate are tightly related, and make sense to be included together as a JIP.

As @validblocks points out, the commission only begins to creep up into 4% and 5% at the very tail-end of the proposed set. As a result, this new cap allows tail-end validators to maintain higher commission rates while still being included in the pool, without raising them so high as to potentially impact the performance of JitoSOL.

However, the commission cap also acts as a guardrail to ensure performance doesn’t suffer if there are changes to the potential validator set. Without it, if some validators became ineligible for various reasons, during the process of churn the pool would be forced to delegate to the top 200 validators even if they had commissions at 8% or much higher. This could have a significant impact on JitoSOL yield and potentially TVL. We’ve edited the original proposal to include more of this rationale and logic, and hope is assuages delegate concerns.

Additionally, this JIP will help smaller validators, to the extent that there are now more validators who can gain entry into the Stake Pool set, and smaller validators don’t have to set commission to 0% to compete. For reference, a 100k SOL validator at 5% commission generates the equivalent of $150k per year at current SOL prices.

Separately, there has been good feedback from multiple commentators regarding geographic decentralization, periodic reviews of the full stake pool inclusion parameters, and stake distribution. These topics are particularly interesting in terms of how they could potentially be managed by the Stakepool system. We’ll be opening a discussion thread to address some of these proposals for possible ideation into JIPs.

Finally, as all of these policies have an impact on validator earnings, we’re considering commissioning monthly reports benchmarking validator earnings rates. We’ll be opening a discussion thread about who authors this report and if there’s any sort of compensation as well.

[BlockworksResearch]

We agree with the motivation and specifications in JIP 1 and voted in favor.

Given Solana’s growth and the TVL of the JitoSOL pool, it is appropriate to expand the Jito Stake Pool validator set. In terms of changing the commission cap from 10% to 5%, we think the effects at this stage are trivial given the commission of the proposed set.

That said, we echo commenters on the proposal and what seems to be the general consensus that over time these decisions require careful consideration.