[TEMP CHECK] Minimum Fee for Public Operator Market and Public Operator Financial Support

[fod]

[TEMP CHECK] Minimum Fee for Public Operator Market and Public Operator Financial Support

This is a proposal for two actions that address problems in the public operator market. For details around that situation, see the following forum posts:

[TEMP CHECK] Fixes to Recent Issues with Public Operator Economics

> The following Temp Check was brought forward by some of the core contributors, and the SSV Foundation was asked to post it on the forum and facilitate the discussion and voting around it. [TEMP CHECK] Fixes to Recent Issues with Public Operator Economics Recently, SSV’s operator economics have not been a high priority, and the current state is one where likely no public operators are running profitably. This has been growing into a serious problem where we risk a large number of operators shu…

[TEMP CHECK] Revisiting the Operator Marketplace and Fee Models

[TEMP CHECK] Revisiting the Operator Marketplace and Fee Models Introduction The SSV Network was created with the concept of an open marketplace where validators and operators could come together to do business. Operators would provide hardware and services to operate validators, and the validators would select a cluster of operators that met their needs for diversity, performance, and price. To help ensure that the marketplace always has a selection of top-quality operators, the Verified Opera…

Both of these actions have been discussed thoroughly, and it seems we might have sufficient support to move forward. A huge thanks to everyone that was part of that dialogue and work. The intention is to append these tasks to the others that recently passed a Temp Check vote.

Action 1:

Create minimum operator fee and set to 0.3% of ETH rewards.

A prerequisite of this is enabling the payment of fees in ETH and the definition of operator fees as % of ETH staking rewards, which have both already gotten preliminary approval through a Temp Check vote but have yet to be passed as a binding proposal.

The specific value of 0.3% has been chosen to target a value slightly above the base expenses of running a low-cost node, using one’s own hardware and network. This is meant to preserve a public operator “free market”, for the sake of maintaining high performance standards, high decentralization, and low costs due to competition. But it keeps a lower bound to prevent a total price collapse due to oversupply and protects from potential bad actors with motivations to accumulate large numbers of validators.

Action 2:

Disperse two payments to verified operators to help cover costs.

This is necessary because all public operator nodes have been operating at significant losses for a long time, due to issues with the public operator price mechanisms and high volatility with the SSV token. The fixes that would allow the potential of profitability are expected to take at least 6 more months. Public operators are being severely strained, and many are on the brink of shutting down. Operators shutting down has costly impacts to the customers using them, including forcing them to withdraw and recreate their validators in some cases. The network has very few tools to mitigate these events, and the expected impacts would be high.

Although our operator market is currently oversupplied, operators with long histories of good performance provide additional value because they raise the trust offering of the network as a whole. It’s more valuable to keep the good operators with long histories that we have, rather than start with new ones without history when the public operator demand expands, as it is expected to in the future.

To limit costs, the line is drawn with operators that have earned Verified Operator status, which carries the requirements of consistent high performance and passing KYC checks. These are the reputable core of the network’s public operator set.

First payment:

The first payment, paid as soon as possible, is for $2000 per year of operation per verified operator node. This would be on a prorated basis, using a period from the later of January 1, 2025 and the date when Verified Operator status was given to the time of this proposal passing.

Payments will be made in SSV tokens, with the price set using a 7-day moving average and the tokens minted. At the time of writing, there seem to be 92 public operator nodes with VO status. Therefore, a rough upper bound for this first payment would be $184,000 in SSV, but would likely be around $150,000 due to many operators gaining this status throughout 2025. The payment amounts and eligibility will be determined by the DAO representative assigned to oversee these tasks.

Second payment:

The second payment will be paid at the same rate, prorated from the time of the first payment to when the following issues are completed and accepted by the DAO representative assigned to oversee them:

-Operator fees are calculated based on ETH staked not validator counts.

-Operator fees are paid in ETH and defined as % of ETH staking rewards.

-The fee change mechanism for operators has been improved.

This payment will also be made in SSV tokens, with a price set using a 7-day moving average and the tokens minted. Assuming these changes take 6 months past the passing of this proposal, we can estimate a rough upper bound as $92,000 in SSV.

Therefore, the expected total cost of this assistance to operators will be approximately $240,000.

-–

Please provide your thoughts, and if there is strong support, we will move forward with a snapshot Temp Check vote. Thank you very much!

[AXBLOX]

Thank you @fod to have come up with this approach to mitigate the issues with the SSV public operators.

AXBLOX has one public operator on the network so of course we would welcome this compensation proposal.

What assumptions did you use in term of total ETH staked with the operator to arrive at a minimum operator fee at 0.3%?

According to @Metanull VO Performance Bot, public verified operators have an average of 89 active validators each.

So for the sake of this example let’s say an operator runs 89 validators with each validator an effective stakes of 32 ETH.

The total stake is

32×89=2,848 ETH

With an ETH APR of 3.1 %, the total yearly staking rewards:

2,848 ×0.031= 88.28 ETH/year

With a minimum operator fee set at 0.3%:

88.28×0.003 ≈ 0.2648 ETH/year

With an ETH price at $3,430:

0.2648×3430 ≈ $908/year ≈ $76/month

If we compare with Lido Simple DVT x SSV, the clusters are composed of 7 operators, 10% of total staking rewards go to the module (2% Lido’s treasury + 8% module), SSV takes 0.75% as the infrastructure fee but waived for the first year so each operator currently earns about 0.08/7=1.14% of total staking rewards.

Maybe the 0.3% floor helps prevent a total price collapse due to oversupply, but at first glance, it seems a little low to be competitive. Still, it’s hard to tell, for instance, our public operator has 8 validators but the total ETH is quite high because of the consolidated validators since Pectra. We have close to 6,000 ETH staked via our operator which would bring in about $1,913 per year of rewards.

But the obvious question is: when we push the change to operator fees based on a percentage of ETH staking rewards instead of validator count, will the validators stay with us? How many consolidated validators presently use SSV public network for the extremely lower-cost staking opportunity?

[fod]

Thanks @AXBLOX !

For the 0.3% value, my focus was on my estimation of a cheap “solo staker” system that is powerful enough for consistently good performance. The purpose is to try to maximize decentralization within the public operator market by allowing these small operators to exercise their cost advantage over larger operators running cloud systems or other professional infrastructures with higher overhead costs. For this, my “breakeven” target was around $500 / year for 50 validator shares (1600 ETH), and my “moderate success” target was $2000 at 200 validator shares (8000 ETH). Also consider that ETH (and as a result this fee) will likely trend upward long-term.

$3500 ETH, 32 ETH per validator, 3% staking rewards, 0.3% operator fee, 50 validator shares (1600 ETH):
$3500 x 32 x 0.03 x 0.003 x 50 = $504

I know that Lido and others are paying operators much more, but we are not competing with other staking services on operator salaries. Instead we are competing for TVL based on staking fees, and we should seek to minimize these fees.

I am arguing that DVT as a staking technology has a massive advantage that is not being exploited yet, and the “market rate” of a good operator (near 100% performance, KYC’d, 1+ year of history) should be far lower than what Lido and others are paying. DVT greatly lowers both the trust and performance requirements of operators, and the technical task and hardware requirements of running an SSV node have gotten much easier over time. This role can easily be filled by the best of the “solo staker” demographic (we have many here with an outstanding performance history). If we can leverage this to keep our “market rate” to <2% for a cluster of 4 operators, then that’s a massive advantage over competitors (including private operators).

This is intertwined with our primary problem: too few stakers choosing public operators. I know cost is just one piece of this (one that is obviously not sufficient by itself), but we can improve the flow of ETH to the public operator market with more work. And I believe it’s our responsibility to solve this problem, not just for operator profitability, but for the decentralization and robustness of SSV and ETH staking as a whole.

when we push the change to operator fees based on a percentage of ETH staking rewards instead of validator count, will the validators stay with us?

I think so, simply because we will still be extremely cheap, and the IMP is still active. But we should be careful and sample the opinions of stakers. And we should always be pushing to improve the value proposition that public operators offer so that we’re the obvious choice.

[alonmuroch]

I’ve been struggling with what I think about this proposal of paying extra to solo operators.

On the one hand, we want to strengthen their operations and make them more prominent—they’re the backbone of Ethereum’s security.
On the other hand, it feels like we’re ultimately subsidizing the wealthiest participants (the stakers).

What about a slightly different approach:

1. Allow operators to adjust their fees more quickly (with DAO-controlled parameters).
2. Offset previous losses via the IMP for the users who stake.

This feels like a fairer approach than placing the entire “cost” on SSV holders.

[fod]

Thanks for the thoughts, Alon. I’m open to alternatives if there are viable ones.

Using the IMP has the potential to be a temporary band-aid for fee adjustments, since it can be handled off-chain.

Allow operators to adjust their fees more quickly (with DAO-controlled parameters).

This seems challenging without moving forward with a “full solution”. Don’t forget that addressing validator consolidations (pricing in terms of ETH staked, not validator count) was another required item for this to be meaningful. This was part of the recent Temp Check, and it seems like it’s not easy/quick.

[edit: See comment below… maybe everything can just be pulled into IMP deductions, including validator consolidations, and we can handle everything how we want there? I also missed this relevant post from Alon until now: Solving Max Effective Cluster Balance with SSV Staking]

Do you mean something like allowing operators to set fees in terms of % ETH rewards but still having it denominated in SSV? Still need to be cautious about SSV volatility (customer hassle, liquidation risk, malicious operators, etc.), but I think this could at least be a step in the right direction. Might not even be worth the effort without validator consolidations fixed though.

Moving everything to ETH fees (as % of ETH rewards, using ETH staked for calculation) seemed like the agreed upon solution. This was also part of the Temp Check that passed. Any hope to simply expedite all of this?

Offset previous losses via the IMP for the users who stake.

I assume you mean give some portion of the IMP to operators (at the expense of stakers)? If so, this might work, as it’s “fair” for most involved… stakers that have been getting cheap fees would pay to make up for the discount. But there are disadvantages, mainly the inequality between operators. Some operators will get far too much, and some almost nothing. Also unknown response from the stakers for losing income, especially new ones that didn’t get the cheap fees.

My intention was proposing this as a minimal subsidy to reimburse operating losses, not reimbursing lost income potential. I guess it depends on our goal… if it is to preserve our entire verified operator set (my intention), this solution won’t save all of them.

This feels like a fairer approach than placing the entire “cost” on SSV holders.

How about we soften the impact by stretching out the payments over 12 months or something, instead of it being immediate?

I’m personally still leaning towards just minting and paying in some form… simple, clean, fair, no overhead, gives us time to fix everything properly without “band-aids”. Obviously we want to be frugal, but ~$200k is comparatively small (~$17k per month over 12 months).

I understand that it’s not ideal, and I too have been resisting this option until now. But the current situation is one where the DAO deprioritized fixing operator issues and “borrowed” from them for a long time. We waited until it became a large problem, and the “bill” is coming due.

[fod]

If we decided to go this way, seems like everything could be pulled from the IMP as deductions, probably including validator consolidations? So this could maybe be a temporary band-aid for all fee problems.

Similarly it could cover past losses if stretched over time. It also allows equal distribution among all verified operators (if we want that), by spreading the total cost over all stakers and giving the fixed $2k per node. If this includes private stakers, it would decrease the IMP rewards by something like 2% over a year. If only public stakers, then much higher, like 40%. (I think… all rough napkin math) Including private ones seems reasonable, as a small favoring of stakers that chose public operators.

I think this is fine if we want to avoid additional minting and feel the IMP is a better funding source (stakers pay for it instead of holders). Still just a band-aid and comes with some overhead, but probably ok.

[AXBLOX]

@alonmuroch

Allow operators to adjust their fees more quickly (with DAO-controlled parameters).

Yes this is aligned with the urgent fixes proposed from the Temp Check and voted Yes by unanimity.

@alonmuroch

On the other hand, it feels like we’re ultimately subsidizing the wealthiest participants (the stakers).

Are you suggesting that by introducing payments to VO to help cover their costs, we are indirectly subsidizing stakers, since it would help VO keep providing high-quality DVT services at a very low cost for the stakers?

@alonmuroch

Offset previous losses via the IMP for the users who stake.

Do you mean the losses operators are realizing, not the stakers? From my understanding and experience, public operators are currently struggling economically, while stakers benefit greatly from low network and operator fees.

[AXBLOX]

This is probably a topic for a future Temp Check but I want to share my point of view here.

Now that I’m in the new OC, I started looking at the overall situation and challenges of the public operator marketplace. I think the main issue with the SSV public marketplace is a systematic one, how the system is designed and how things work today.

I agree, we should move forward with urgent fixes like making fees depend on ETH effective balance (not validator count) and allowing operators to adjust fees more freely. But currently, we’re seeing that not enough stakers are choosing public SSV operators and over time I realized there are valid reasons for that.

1. In general, most medium and small stakers don’t want to deal with the technical side of staking so they use custodial or turnkey staking services like Kraken or Lido. Those services use SSV for DVT within private setups to maintain control over pricing and operations.

2. There’s also the SLA (Service Level Agreement) angle between pro operators like P2P or A41 and the medium-sized staking services that pool smaller stakers. SLAs set clear expectations for operators, like staying online and performing well. If they don’t, they can be penalized or replaced.

Reason #2 is a big reason justifying why these staking services tend to go with private operators running the SSV infrastructure. At the end of the day, staking providers want guarantees if something goes wrong, they want someone responsible. That’s a lot harder to get that with anonymous public (verified or not) operators who don’t really have skin in the game and could just vanish without any real consequences.

In my opinion, these two reasons explain grossly why more than 85% of operators on SSV are on private setups. This is the systemic reality and the core issue with public SSV operators. It’s good for SSV overall, but if we want the public side to survive, we have to improve some things.

A potential improvement would be to bring a systematic solution to a systematic issue. The Trusted Cluster model proposed by @alonmuroch could be a meaningful step in that direction. By offering cost predictability, off-chain payments and reduced operational friction, it has the potential to make public SSV operators more competitive with private setups.

When combined with accountability features similar to SLA potentially structured between the Trusted Cluster and the Foundation and adopted by already verified operators, Trusted Clusters could help attract more stakers to the public marketplace.

[fod]

These are good thoughts, and although I do generally agree with them, this is the larger overarching issue that will require significantly more effort and thought. Let’s stay focused here on first solving our most immediate problems.

I have been thinking along similar lines though, so I’ll reach out on another thread. I think there’s definitely a lot we can do to close this “gap”, and I’m confident that we can get there.