Reduce Elected Validator Count to 45 - DRAFT

Governance
1

This thread is a direct iteration on the Strengthen Tokenomics: Cut Validators from 110 to 55 thread and discussion that took place in the last Governance Calls.

As shown in the Celo Treasury and Governance Dashboard, the Celo Treasury is running low, with approximately 76M CELO and 2.5M in stable assets remaining. At the current spending rate, this provides roughly 12–14 months of runway. This is a situation we must address urgently by reducing spending across the board and being more intentional about how we allocate resources in 2026 and beyond.

I analyzed the current validators and ranked them based on publicly available information and my personal assessment of their contributions to Celo, both in terms of validator performance and additional contributions, such as governance participation, solution development, and ambassadorship.

Because there is currently no clear path or timeline toward decentralized validation, and given the rapidly shrinking Celo Ecosystem Treasury, I believe Celo must reduce the rewards spent on validators.

Based on my analysis, I believe the right step is to reduce the elected validator set from 110 to 45, with a maximum of 3 validators per group.

At present, there are 136 validators, with 52 groups elected, costing the Celo network approximately $2,750,000 per year.

According to my simulation, after the reduction, there would be around 75 validators remaining, with roughly 33 groups — the most aligned and active ones — continuing as active validators. This would reduce annual network costs to about $1,125,000.

If this proposal passes, validator incentive costs would decrease by 59%, while the number of validator groups would decline by about 36%, with the least value-aligned and least active groups likely exiting the network.

Below is a screenshot of the anticipated impact of certain amounts of elected validators. My full sheet can be found here.

image
image1621×196 40.3 KB

I recognize the long-standing relationships and significant impact validators have had on Celo, but I truly believe this is the right step for the network’s long-term growth and health.

Hopefully, the remaining validators will continue to collaborate closely to provide essential infrastructure and support for the Celo network — and play a key role in realizing our shared vision, aligned with Celo’s Vision 2030.

Finally, I believe the Score Management Committee can play a crucial role in ensuring that the remaining validator network coordinates effectively, and in providing valuable input for foundation delegation programs and the overall growth and development of the Celo ecosystem.

I look forward to receiving final input, as I aim to turn this into a formal proposal to submit for a vote soon.

3 Likes
2

Thanks @LuukDAO for taking the time to put this forward. At a high level I believe this would be a step in the right direction and I am supportive of us exploring this.

I do like the idea of more closely looking at criteria to see what groups are having what impact on the ecosystem beyond the direct services provided – it is incredible that Celo has been able to support so many great public goods efforts (per Luuk’s post, agree eg Grassroots Economics is an awesome example of this) through validator rewards, I think it’s something to be proud of as a community.

3 Likes
3

The issue isn’t validator set size — it’s CELO price.

When CELO trades higher, fewer tokens need to flow into Mento to maintain the reserve, the treasury’s runway extends, and public-goods funding expands. Every metric people are worried about improves when CELO price strengthens. This proposal treats symptoms instead of addressing the real cause — CELO’s lower-than-expected market price and the resulting dilution effect.

Validators today aren’t the source of sell pressure; they typically convert cUSD → USDC, not CELO. The primary sell pressure still comes from investor and early-employee unlocks, not validator rewards. Cutting 60 percent of the validator set won’t materially improve price or emissions — it only weakens decentralization at a moment when the network still relies on a centralized sequencer.

Validators now provide distributed infrastructure and RPC capacity, which are vital for redundancy and credibility while Celo transitions toward decentralized sequencing. Reducing that diversity sends the wrong signal.

If sustainability is the goal, better levers exist: adjust staking yields, optimize treasury spend elsewhere, and focus on CELO demand drivers. Cost-cutting at the expense of decentralization is the wrong trade-off for long-term health.

Finally, the “CeloPG Validator Analysis” is a subjective assessment — not a data-driven justification for systemic change. Decisions that impact network structure and decentralization should be based on transparent, quantitative criteria (performance metrics, uptime, participation rates), not on personal impressions of contribution.

5 Likes
4

@rene_celo is there anywhere we can get project management updates or information on the decentralized sequencer concept? Clearly it’s not a huge priority, which is understandable, but the difference between “parked for now” vs “ready in 6 months” vs “ready in 18 months” really does have an impact on all of these reward concerns.

i.e. If “ready in 6 months” then I would favour just doing nothing. The governance human-resource costs and discussion/execution is barely worth it.

If “parked for now, no further information” then we probabably should look at the RPC rewards program holistically. Is it worth continuing what was a stop-gap / temporary measure to keep technically involved actors in and around the protocol if there’s nothing planned? Would it be better to close the program entirely and put efforts into running a version of The Great Celo (Stake) Sequence Off when things are at the testing phase? (Aztec has been running a decentralized sequencer set with > 10k sequencers successfully for some time, so there is a precedent for amazing levels of decentralization)

If “ready in 18 months” then maybe some sort of halfway measure, reduced rewards, rewarding value-add or subjective measures like @LuukDAO has mentioned could make sense.

4 Likes
5

Simulation results ran directly on the smart contract electNValidatorAccounts(22,45): D'Hondt 45 validators election 03-11-25 - Google Sheets.

Those highlighted in red would no longer be elected.

2 Likes
6

I agree with much of what @Patrick says, especially comments in the recent governance call about independent groups often being highly aligned with technical or value-add contributions - I share this view and have concerns that are hard to quantify about the “personality” of the network changing if infrastructure is structured such that smaller groups are no longer viable.

Of course this doesn’t mean the protocol should be a welfare project for smaller groups that can’t self-delegate. It was said many times, especially after the Foundation delegation program ended that the longer-term goal was for groups to market themselves effectively to acquire external stake.

To steel-man the side of the voices that are looking at this as a pure budgeting line-item, it is true that currently the RPC providers are not commercially providing direct value as endpoints. How many browser wallets would be willing to move their default RPC to any particular group’s endpoint, or even a top-level load-balanced version without strict QoS and SLAs attached to it? (Especially when they might already have well-managed nodes by their own commercial suppliers Alchemy/Quicknode etc).

It is great however, that there are dozens of companies or individuals that are comfortable operating and monitoring the Celo OP stack however, that does have some non-zero value.

So I am sensitive to the idea that the RPC rewards was a stopgap idea to keep engagement in a technical group. If the program itself doesn’t make financial sense anymore all things considered, I’m OK with understanding that as a driving force for this discussion. The somewhat implied or undercurrent narrative that validator rewards (then) and RPC rewards (now) are just a spend item to be hyper optimized is over-simplified and somewhat scape-goaty (as @yomfana ‘s comments).

There are many gigantic CELO spends from the treasury over the last 24 months, that are essentially “too big to fail”, with even cLabs and the Foundation taking operational spends that never occurred before. CICLOPs is completely lacking in transparency and can be approved and spent with literally no one here knowing where it has gone and why (of course, I understand why this program exists). These are all trusted and foundational actors in the Celo ecosystem, and obviously there’s no malintent here, but it’s clear what @Patrick was saying, token price is likely the upstream force in the situation we are in, rather than any individual program suddenly becoming a bad spend. That said, I do expect Season 2 will be a fair bit more tighter especially in the community / building space given this entire conversation.

Regarding implementation here - if there’s support for some restructuring like @LuukDAO ‘s suggestion, why don’t we just tweak a single number: the daily reward by the percentage we want to save? It’s easier to implement, and saves the endless back and forth with spreadsheets on group restructuring and so on. The current daily reward I think is near $90 per validator per day. If there’s consensus to reduce the spend of the program, we should just adjust that down.

If we go a decentralized sequencer future, I would think a reduction in validator count could be a mistake, we should have as many sequencers as our latency can handle. As far as I understand there’s no technological reason not to.

5 Likes
7

Agree. Btw, what’s a Score Management Committee? And, still trying to understand the real purpose of that “CeloPG Validator Analysis”. If its a kind of personal biased subjective assessment, then it feels quite disappointing and lacks clear understanding/context. Also, decentralized systems and its incentive mechanics to keep an healthy and secure network, should not be based on human interpretation or “opinions”.

1 Like
8

The Bad (Centralization Creep)

  • Subjectivity is poison: Ranking on “ambassadorship” or “solution development”? That’s a recipe for insider bias—who decides what’s “aligned”? It risks alienating diverse groups (e.g., smaller operators from non-Western regions that Celo claims to champion) and concentrating power in a handful of cliques. True decentralization thrives on permissionless entry and objective signals like stake-weighted voting or automated performance audits, not a spreadsheet from one governance whale.

  • Slippery slope: Start with 110 → 45, and what’s next? A “Score Management Committee” sounds noble, but it could evolve into a validator cartel, undermining Celo’s proof-of-stake ideals. We’ve seen this in other chains—Cosmos or Polkadot hubs get fragmented when cuts feel unfair, leading to forks or exodus.

  • Misses the root: Why not tackle treasury drain elsewhere? The forum’s buzzing with ideas like accelerating stCELO staking for more organic security or optimizing Mento reserves. And hey, if the network’s truly decentralized, let token holders vote on reward pools via on-chain governance, not forum drafts.

5 Likes
9

This is reactionary cost-cutting disguised as optimization. It is built on subjective analysis, misleading framing, and a fundamental misunderstanding of both validator economics and the broader ecosystem’s fiscal challenges. A lot of my points from here are applicable to the proposal as well, so I won’t reiterate them over and over again.

Let’s establish some context. When the Seasonal Intents were released and ecosystem grants were distributed in the first week of August, CELO was trading at approximately $0.30 (roughly the same price it held two weeks ago). The Community Fund runway was entirely predictable at that point. Everyone involved in governance knew exactly what effect those distributions would have on treasury sustainability.

Back then, there were no emergency measures. No proposals to address “excessive spending.” No concern about “runway risk.” The treasury’s trajectory was already baked into the decision-making process.

Now, conveniently, as CELO’s price trends downward, validators have suddenly become the scapegoat for the ecosystem’s spending concerns. This is not principled fiscal policy; it’s opportunistic blame-shifting. If the argument truly hinges on opportunity cost and runway, then the rational step is obvious:

wait for Season 1 performance reports. Let the community assess all ecosystem programs based on actual deliverables and measurable performance, then adjust funding proportionally.
Why are validators being singled out before there’s any transparency on other ecosystem programs have performed?

Your “analysis” is fundamentally subjective and demonstrably inaccurate. The simulation results you reference show that your “likely” validator count projections are off from what would actually occur under the proposed election mechanics (also posted above). This is not rigorous modeling, it is working backward from a desired cost savings number and figuring out in reverse how many validators need to be eliminated to achieve it. The target of 45 validators is entirely arbitrary. There is no technical justification for why 45 is optimal. There is no game-theory modeling of how this concentration affects governance or collusion risk. It is simply: “I think we should save money, and this number saves money.” This is not how you design critical infrastructure policy.

This is wishful thinking at best, and dangerously naive at worst. Validator elections are decentralized and permissionless. Human scoring systems, however well-intentioned, will have zero binding effect on election outcomes. The mechanics of the election are determined by vote weight, not subjective assessments of “alignment” or “activity”. Let’s be brutally honest about what this proposal actually achieves:

  • Small validators get eliminated. Groups like @celocolombiano, who have worked for years to secure a single elected slot, get wiped out overnight. Their grassroots progress becomes meaningless.
  • Well-capitalized groups consolidate further. Entities like (insert any new group in the last year with > 4 validators), which already have setup capital and five elected validators in months, become entrenched.
  • Barriers to entry skyrocket. The validator set becomes a gated club rather than a permissionless network.

Celo isn’t a “charity chain,” true, but it was built on openness and inclusion. Eliminating smaller, independent operators without hard data or transparent justification isn’t optimization; it’s institutional capture disguised as efficiency.

And then there’s this gem:

That line would be darkly comedic if it weren’t so consequential. The implicit message is clear:
“The beatings Validator cuts will continue until morale improves”.

2 Likes
10

I have a neutral opinion on this, but can this really solve the fundamental problem of tokenomics? Celo highlights the stablecoin transaction count, but the price of the CELO token has been plummeting for a long time. We need to identify and address the root issue of Celo’s tokenomics — the number of validators seems like just a temporary fix.

4 Likes
11

Cutting validator spending is one solution as part of a broader package of necessary tokenomics reforms. Price is a function of supply and demand and we’re reducing emissions supply. We have to start somewhere. If you want to start putting up additional proposals to strengthen tokenomics, I am supportive, please initiate! FWIW, I’m also working hard on the demand side of the equation to generate more stablecoin transactions, but we need to tackle supply side in parallel.

Why start with validators? IMO, there are two ugly truths that we’ve been too timid to acknowledge:

1)Network is paying “validators” to run RPC nodes and the network doesn’t need those RPC nodes.

2)Market is not asking for decentralized sequencing as evidenced by MegaETH fundraising and espressosys.com dropping decentralized sequencing support plans. Maybe shared senquencing for interoperability (faster finality guarantees + comms with other chains), BUT not decentralized sequencers.

To be frank, structurally speaking, Celo could remove ALL validators. Finality comes from blocks derived from the L1, having decentralized sequencer could help censoring but you can get an uncensorable route posting to L1 directly. As an ex-validator (Tango Group), I recognize the value validators bring to the community and ecosystem, so I’m not suggesting we do this, but I think leaving validators numbers untouched in a post L2 world is also unreasonable.

2 Likes
12

Of course, RPCs are a commodity and easy to find. But the program wasn’t really started because it’s providing hard business value, it was partially to keep an engaged technical group understanding how the OP stack works, and (when there was a real possibly we would be involved in some sort of decentralized sequencing future state) to keep us on standby and involved in whatever comes next. At least that was my take from listening to @marek speak earlier this year before the L2.

Maybe things have changed and I’m OK with the revisiting the appropriateness of the program today. Feels somewhat small-fry compared to other programs, nor does it free up cash directly that can be part of a budget from the Celo Community Fund (although affects emissions of course) but I totally get it as part of a number of levers we can start pulling tokenomics-wise.

Anyway with @rene_celo approval for this, it’s clearly going ahead in some format. What do you think about my idea of just reducing the daily remuneration by X%? If the goal is save money, that seems the straightest line?

4 Likes
13

We’re all aware that changing validator rewards is not a silver bullet solution, but a step in the right direction to get our treasury spending under control and increase scrutiny across the board to prioritize activities that have a high likelihood of contributing to CELO demand increase.

With that said, I truly believe reducing validator count > reducing daily rewards per validator for two reasons:

  1. Given that validators have a cost to operate and maintain uptime, the lower we make the rewards, the larger the relative share that is spent on maintenance / paid to third parties. From that angle, it seems better to reduce node count and ensure each node is still profitable with a good margin for the operator.

  2. It seems like overkill to have a single group run as many as 5 nodes; we see value in limiting that to 2 or 3. In addition, the reduction in nodes will likely make the validator slots more competitive.

I hope for initiatives to be started where active validators that do more than just validate to be able to receive additional Celo votes pointed to them, and would support such initiatives.

A question to the stakeholders here, do you feel a reduction to 45 elected validators with max 3 nodes per group is the right balance between cost reduction and potential loss of stakeholders, or would you consider going lower, potentially even to the point where we would operate a period without elected community validators?

14

hi Aaron - probably a better question for the cLabs crew to chime in on, I will flag to them also, I know they are actively looking at this now.

3 Likes
15

I agree. If the same level of rigor that’s being applied to validators were also applied to ecosystem spending, then it would be fair. Right now, a lot of programs are getting a free pass for writing glossy reports that say very little. I haven’t seen a single detailed expenditure report from any major ecosystem recipient. It’s always the same pattern: five lines of budget requests, and six months later, five lines of “financial reporting” stating that all funds were used up.

If you actually optimized ecosystem spend e.g. salaries based on Purchasing Power Parity and industry standards, and cut redundant or low-value roles, the savings would be substantial. It’s absurd that some people with minimal responsibilities and basic skills are earning the same as highly experienced engineers in the same locality. No one calls that out :person_shrugging:.

From a security perspective, yes. In theory, we could shut down all community RPC nodes today and the chain would still run. But from a cost perspective, the network would still end up paying someone else to do the same job. For example, CICLOPS takes around $3M per season from the community treasury to “partially fund RPC providers,” yet there’s no public breakdown of who those providers are or how that money is allocated.

The market will always favor centralized sequencing: it’s easier for MEV extraction, fee capture, censorship, and data monetization. But serious, user-focused chains that actually care about decentralization will not fall for this trap. They’ll continue to maintain a pathway toward decentralization, even if it’s gradual.

Also, for the record, Espresso Foundation posted an update just two weeks ago that the Celo OP Stack integration is progressing well. It was mentioned on this forum about four months back too. Their official site still lists Celo, and their documentation explicitly states that decentralized shared sequencing remains the long-term goal. So yes, it’s happening, maybe not immediately, but definitely eventually. It’s not a question of if, but when.

Agreed. But any change must be done the right way, with a high-quality, data-driven proposal, not one driven by panic or short-term optics about CELO price or treasury runway. Neither of those factors have a direct correlation to validator pay, which comes from reserve emissions.

There are a lot more benefits than just censorship. The fact that Celo advertises itself in both papers and the website as decentralized means that it needs to take every opportunity to make sure it remains that way.

For this and any other similar proposal to have real legitimacy, I’d suggest the following:

  1. Impact report: A clear, data-backed analysis on how validator reduction would affect existing community RPC providers and their users. This means direct outreach to groups, collecting real feedback, and publishing it as-is. No subjective scoring or favoritism toward group A or B. It needs to be professional, methodical, and transparent.
  2. Independent tokenomics analysis: Request the cLabs tokenomics team (if still active) or an independent, credible third party with crypto-economic expertise to publish a short but rigorous analysis on how validator pay actually impacts CELO price, ideally in comparison with other ecosystem expenditures.
  3. Transparency on savings: There must be a clear upfront transparency by the proponents (especially this particular proposal) on what happens to the savings. Will these funds be redirected to the community fund within a year? For what purpose? Without that clarity, this proposal just feels performative and sanctimonious.
2 Likes
16

As a validator since before mainnet, I’m extremely grateful to have had the opportunity to help and support the network over the past five years. I appreciate all the comments here, especially the honest and unbiased takes. It’s refreshing to see genuine discourse. Some of the critical points raised are also valid in the broader scheme of things. I believe the market has shown it does not value a decentralized validator set, nor is purely running validators a viable business (it arguably never has been).

The issue with following the market too closely is that we end up lagging instead of leading with conviction in our beliefs or ideas. We’ve seen this before in decisions made under market pressure: spinning off Mento and removing the core design mechanism for value accrual in the $CELO (RIP $cGLD) token, or spending $100M on the DeFi for the People campaign chasing mercenary capital because it was the meta at the time. If Celo had instead leaned into its original vision of a mobile-first stablecoin network, we might be in the best position for what the market demands now.

So let’s be serious here and look at this through the lens of an unbiased observer. Validators are, more or less, providing a redundant service. The long-term goal has always been to move toward a decentralized sequencer. To be honest, Celo needs a differentiator from other L2s, ideally one that reflects its core values. If such an initiative does emerge in the future, it makes sense that the entities who have spent the past five years validating and now providing RPC infrastructure would participate in providing that service.

With that in mind, we should acknowledge that the current validator selection process and schema are outdated. Validator groups with child validators are no longer representative of a functioning system and would need to be redesigned for any future L2 decentralized sequencer. If we were to move forward with the direction outlined in this draft proposal, I’d suggest that every validator group be limited to a single validator.

That said, I do see a problem with this proposal being initiated by an entity that has used community funds to build initiatives which, intentionally or not, directed votes in a way that now positions them to benefit from this very proposal. However, as others have pointed out, we should avoid evaluating this in isolation. If this proposal is indeed motivated by cost-saving concerns, then we need to look holistically at all entities drawing from the ecosystem, especially the beneficiaries of various sub-programs.

At a high level, I’d argue that none of these entities are revenue-generating businesses. To stop the bleed, the broader solution should be that no program is eligible to receive community funds unless it is a revenue-generating business capable of sustaining itself without ongoing community funding. Of course, there may be rare exceptions such as the Foundation, but setting this standard, even if painful at first, would create a stronger and more resilient network over time.

Lastly, as an independent validator who has used the leverage of that role to support the network through various initiatives like Prezenti and the Celo Governance Guardians (outside of validation duties), it’s disheartening to know that I’m likely to be one of the entities wiped out by this proposal.

3 Likes