This, however, strikes me as incentivizing the exact opposite of what this is trying to achieve. Asymptotically reducing staking incentives after a given staking ratio is likely to hurt solo stakers before it hurst LST stakers.
Set aside infra, capital requirements, and knowledge needed to run a solo validator. I think this boils down to (lack of) liquidity. The sacrifice solo stakers make when validating is lockng up their ETH in contract. LSTs don’t make this sacrifice, they still had their LST that they can use on DeFi, borrow against, lend, or immediately exit at any point in time. In a future where the value of holding your ETH excels the rewards, I don’t necessarily see it hurting LSTs, if you assume the 1:1 parity of :ETH will hold.
If anything, this would deincentivize existing solo stakers from continuing to stake. And the downsize in staking ratio will likely come from solo stakers. The incentive for multiple LST stakers to collectively withdraw their ETH does not seem obvious to me. They would not hurt, they are liquid.
Maybe this is from a place of naivete, but if the underlying worry is a single tbtf (too big to fail) LST, then a native staking approach might be a higher priority to explore right? The issuance ultimately is a value that reflects the networks’ willingness to pay for its own security, which coupled with the burn has actually resulted in a negative issuance since the merge overall. Is there not a risk of overtuning this, and underpaying?
Also, in a second-order effect, restaking should be considered when thinking of staking economics. This is the stake that secures ethereum that also secures an AVS, in essence, you could say this stake is “less secure” for the base layer. One, if we target a stake percentage, and most of it gets restaked isn’t that dangerous to the base layer? Two, if the issuance is reduced, isn’t it more likely that the targeted stake all get restaked so people retain the same staking return? In a world where we overload the security of the network with additional terms, is it really valid to say we are oversecuring the network? Today almost ~10% of all staked eth is already restaked in eigenlayer.
What is the current time required to withdraw staked ETH (given current withdraw queues and whatnot)? Knowing this would help analyze the time-value-of-money cost to low-liquidity stakers.
Restaked ETH does not secure Ethereum, generally speaking.
For any given amount of stake to be actually securing Ethereum, the actor that is taking on the capital risk of staking failure/chain split MUST be the same actor that is making decisions about what algorithm (client) to run. As soon as you either outsource algorithm choice to someone else or you sell the risk to someone else the game theory for staking falls apart and your “stake” needs to be treated as attacker controlled stake for mechanism design calculations.
The above game theory issue is the reason why solo staking is so important. There is no such thing as “too big to fail”. We can wipe out 99% of stake and everything will be fine as long as the wiped out stake is of people who chose the wrong algorithm (e.g., chose a super-majority client, or a censoring client). There will be a long period (weeks) of no finality during recovery, but the system is designed to self heal from nearly all staked ETH defecting.
I disagree here, it is much more nuanced and 10% ETH being re-staked does not mean Ethereum is 10% less secure. You may have principal-agent relationships (in fact you will have such relationships) between operators and delegators, and some of the capital at stake may also be burdened with other conditions, all of this is important to consider the security of Ethereum, not only the worst-case pessimistic view.
Something that could happen is a large EigenLayer slashing event, in which case the first-order consequence is a reduction of the “dollar-amount” security of Ethereum (as long as EigenLayer slashings are properly surfaced to the protocol, which is the point of EIP-7002 for instance). Since lots of validators are exited, the consensus-offered yield will decrease, inciting more validators to join the staking set and recovering the equilibrium staking ratio.
A bit of weirdness may happen during the transition, but is in my opinion mitigated by the following. Let’s say an adversary wants to benefit from the temporary lower amount of stake after a large EigenLayer slashing event. There can be two cases. If the adversary is the one to trigger the large slashing event, then its own stake will be penalised and exited from Ethereum, largely preventing them from pursuing an attack on Ethereum (e.g., a safety fault). If the adversary is not the one, then it cannot really predict in advance that a large slashing event will happen, it needs to command a large enough amount of (non-EigenLayer-encumbered) stake to launch the attack as soon as the slashing event happens. At the end of the day, the bounds provided to us by the theory of consensus are binding: If we have an adversary with 1/3+ of the total active stake, a safety fault of FFG is possible in theory.
One can certainly try to model the system with cow shaped cows rather than spherical cows, but things get really hard and really complicated really fast. In general, I think when designing systems we should focus on designing a system that is resilient to the worst case scenario with rational actors rather than hoping that we get all of the numbers right for the fuzzy and hard to calculate things.
If I am accepting delegated stake and operating a node, and in return I earn say 1 ETH over 10 years for doing so, if an attacker offers me 1 ETH today in exchange for running their software instead of canonical software, the rational decision (with spherical cows) is to take it. Of course, there are fuzzy things like threats of violence (legal risks) and reputational damage that I perhaps should think about, but we have no way of measuring those things accurately. How much do the operators of some random delegate value their reputation? What if their reputation is already in the gutter for unrelated reasons? What if they operate anonymously? What if they live in a country where bribing law enforcement is cheap and easy?
Even if you can accurately estimate (or at least put some bounds on) things like legal/reputational risks and reasonably convert those into ETH denominated numbers, adding complexity to the system can cause a complexity explosion that ends up with subtle bugs in the mechanism design. Take Bitcoin for example, where the economic incentive system is stupidly simple. Yet it wasn’t until sometime after its launch that people realized selfish mining attacks were profitable in a transaction-fee-only future. This is a very subtle attack that was discovered years later despite the system being trivial to analyze. The situation is far worse when your system is not trivial to analyze, which is the case when we start treating restaked ETH as anything other than “attacker ETH”.
Note that these proposals are not actually hoping to “get all of the numbers right”, they are quite agnostic to the ratio between “true operator stake” and “delegated stake” or “re-staked stake”, they simply reason out about the aggregate size of the staking set. Scaling the staking set size by 2x may mean that you get 2x the dollar amount of “true operator stake”, but also may not (e.g., larger staking set => more LST network effects => more incentives to delegate to LSPs). Either ways, there are indeed risks of misalignments between principals and agents, but these risks have more to do with staking UX (by which I mean broadly the set of mechanisms that are in place in- and out-of-protocol between stakers and the protocol) than with the current discussion. To be clear, these mechanisms are important and they help us improve the quality of our staking set, which is a necessary condition to me (see recent post), but not sufficient to achieve sustainable staking conditions.
This is besides the point but selfish mining attacks are also profitable in a block-rewards-driven regime, though yes when you are in a transaction-fee-driven regime you need less hashpower to launch them with +EV. More generally, the proposals made for Electra or this curve with stronger targeting are not more opinionated to me than the current reward curve that we have, they just express different opinions, which appear to me to be broadly more in line with the goals of Ethereum. So it’s not really a problem of adding complexity to the system vs status quo, but deciding what best suits the needs of our network.
does a great job at highlighting the advantages in terms of cost, convenience, and liquidity of SSPs vs solo stakers
does not account for airdrop/points farming from protocols such as EigenLayer. Swell, Puffer, etc. which represents additional yield pressure to drive SSP adoption
correctly suggests that as overall staking ratio grows, solo stakers will be disadvantaged even more, because of SSP’s structural advantages
proposes a solution that tinkers with overall stake ratio rather than trying to address the problematic structural advantages of SSPs
The dominance of SSPs (and in turn, the staking ratio issue) is a result of average ETH holders making rational economic decisions. The threat posed by SSP dominance is not a new topic and yet we have make little progress solving it. And I’d argue the reason we found ourselves in this situation because we have not been willing to address the structural advantages of SSPs over solo stakers
In the end, it’s an incentive problem. And it is best (can only be) fixed with incentives
The real question here is: Are we willing to discriminate against a SSP validator? More specifically:
Is a solo validator the same as a SSP validator? (I’d argue no. at this point an additional SSP validator is a net negative)
If a solo validator and a SSP validator are different, is it wrong to discriminate one against the other? Does it violate principle of credible neutrality?
If it does violate credible neutrality to a certain degree, is it acceptable?
I’d love to see what the community think re: those questions, the pros and cons
But now assuming we are willing to entertain the idea of validator discrimination, here’s a few unsophisticated thoughts on how we might fix the imbalance in structural incentives
First, thanks to community effort, we already have some ways to distinguish solo stakers from a SSP. Of course, the list is imperfect at the moment, but I’m sure we will get there over time. (also, as a zk-illerate, is there a way we can allow solo stakers to attest to their solo staking status while keeping privacy?)
Proposal #1: Yield over Convenience with ePBS + MEV Redirect, a protocol-based approach
Instead of the proposed MEV burn, simply spread the MEVs evenly among solo validators only
If the result of this comes to a 2+% higher yield to solo stakers over SSPs, I’m sure many of those who own more than 32 ETH will be motivated enough to unstake their LST and learn to spin up solo nodes
Proposal #2: Solo Staking Loyalty Points (SSLP) farming, a community-based approach
solo stakers accrue loyalty points
Points should measure how long a solo validator has been active and attestation effectiveness
Points accrued on a monthly basis and non-transferable
Points can be redeem for solo staking achievement NFT/SBT/POAP
Retro-airdrop for SSLP or staking achievement
We can try and persuade projects to pre-commits to an allocation to solo stakers. For example, they can make statement like “We don’t have a token, we might never have a token, but if we ever do, there will have an allocation to solo stakers and the amount will be based on SSLP”
Thanks for the proposal. I appreciate the thoughtful analysis of the cost structures of home staking and I agree with all of them. However, I do not understand why stake ratio targeting solves them. It seems to me that no matter how you slice it, implementing targeting means that staking yield, both nominal and real, will go down compared to today’s levels.
This means the first stakers to become priced out of staking will be the home stakers, for the reasons you’ve already enumerated: they don’t have the economies of scale that large staking pool operators do, they don’t benefit from the value of having their stake liquid, they have higher reward variability, and so on.
I would add to this list: Home stakers will likely not be able to earn as high a yield as SSPs will due to restaking opportunities that require serious hardware.
There is also another dynamic that this post doesn't mention: LST liquidity wars *between each other* in a stake ratio targeting environment.
So far, most new staking providers have been able to gain market share because they can attract new ETH to be staked. In a world of stake ratio targeting, especially one where the target is close to 1/4 which is already what the current ratio is, the LST competition becomes zero-sum. This accelerates the winner-takes-most dynamic the post already mentions, and it cements the position of existing LST operators as it makes entering the LST market increasingly difficult. You might argue that we are already effectively heading there anyway but that it’s just a matter of it happening when we reach 100% of stake vs whatever the targeted stake ratio is, but the argument remains that the world where the LST market is allowed to grow in a non-zero-sum way for a few more years vs the world where the only new ETH coming into LSTs has to come from existing stake are two different worlds. It may well be that giving more breathing room to the LST market brings a sufficient number of new entrants to ensure the a sufficient plurality of SSPs and delay winner-takes-most effects. But of course, all of this is speculation.
At the end of the day, home staking is already not an economically rational choice. LSTs are already more competitive, and the difference seems like it will only grow over time as per the above arguments. This is why I’ve proposed that we should make staking operators legible to the protocol so that we can actually reverse this effect.
Exiting the validator is an extremely manual process which I have not researched and, having a full-time job, it would likely be a weekend project to figure out how to do this safely. It requires re-learning a bunch of things I have likely forgotten around how my validator operates (something which I set up once and only need to worry about occasionally), as well as needing to research how to set up another validator later if I want to stake it again.
I cannot use my home-staked ETH for anything other than ETH staking. Contrast this with LSTs, which can be used in DeFi to get a loan against them, or can be used to participate in vampire attacks in the SSP wars, etc. as noted above by @theSamPadilla. Essentially, LSTs have already attained a significant amount of money-ness. In the future, my home-staked ETH will probably also be forgoing more yield opportunities from restaked AVSs which be run on a residential connection. That’s OK with me; my goal with home staking is to secure Ethereum and ensures it keeps running first and foremost, and I would be home-staking even with negative yields. But I think the pool of home stakers for which this is sufficient motivation is vanishingly small as a fraction of staked ETH.
Assuming your withdraw address is already set, exiting a validator is just running a single command with your validator client: https://launchpad.ethereum.org/en/withdrawals/#how-to-exit I would estimate 5 minutes to read the docs for your validator client and run the command, maybe 10 minutes if you want to be thorough. If you don’t have a withdraw address set, then you’ll need to do that first but it is a similarly easy process IIUC, and new stakers will not have to do this as it is set during staking setup process now.
If your goal is to secure Ethereum, then you should continue to solo stake and not participate in all of the crazy DeFi lego stuff with your stake as that stuff generally weakens Ethereum’s security. Ethereum would be more secure if those people didn’t stake at all. The problem is, we don’t have a way of stopping them from harming the network, so they continue to exist and do what they do and we hope that enough people don’t do those things.
This is an interesting problem, thank you for the detailed writeup and acknowledging solo stakers might suffer more from targeting.
I’m a solo home staker, with several validators and rocketpool minipools (and intention for future DVT participation).
There’s just a few assumptions/comments I’d like to challenge:
I think it’s a misconception that it’s not viable to hold ETH without a yield. Bitcoin is held without a mining yield, and the recent ETF flows have proven that people don’t need a yield to simply hold an asset. In addition, the issuance/inflation is higher for Bitcoin than it is for Ethereum after EIP 1559.
Even so I acknowledge humans will chase small amounts of yield wherever it is, as seen through 2021-2022 and all the problems it caused.
The initial cost of staking hardware really is not expensive compared to the ongoing monetary and non-monetary costs. In addition since it’s a sunk cost it wouldn’t really contribute to the decision about whether to keep operating based on the issuance - that would be more about the ongoing viability of staking and ones belief in Ethereum as a viable long term investment. The ongoing long term costs of staking I have for example are:
Extra internet connectivity, a 2nd internet connection, higher speeds, higher data. Particularly required for multiple validators and/or splitting into DVT clusters or minipools like Rocketpool.
Time cost for updates every 2 weeks, fixing problems with updates (e.g. today the MevBoost update caused timeouts), fixing outages (e.g. Nethermind bug).
Capital requirement of holding all the ETH in the contract, while not technically an accounting cost it’s a major decision factor on whether to continue staking. Sure it’s easy to withdraw, but by deciding not to withdraw it’s a capital allocation that for some is an extremely large and non-diversified risk.
Electricity, is minor in my opinion.
Upgrades due to changes like EIP 4844
You could say these are more like ongoing commitments rather than just costs, but I’m just highlighting this because these are the factors that play into the constant decision to stay as a home staker. Not once do I think about my upfront cost paid 2+ years ago.
I agree and I think it’s useful to conceptually split the LST considerations into the two participants:
The LST node operator, whose costs are rather fixed and will earn a percentage of what depositors earn.
The person who swaps their ETH for an LST, or deposits it into an LST contract in exchange for an LST yield bearing token
Participant #1 would likely take steps to reduce costs and increase risk if there was less yield, e.g. for instead of running 20 nodes they may cut down to 3 nodes and increase the number of validators per node. It’s an extremely efficient business, to run many validators without much of your own capital deposited. Unfortunately I don’t see this proposal as targeting these participants in a stronger manner than it would target solo stakers, because due to economies of scale with potentially thousands of validators they would still be ahead of solo stakers even if only taking 5% of the earnings (which they can choose to vary if yield changes).
Participant #2 really has zero cost, and can deposit every ETH they own since they only pay for the LST operator out of the yield they earn. Even if yield was 0.1% and they paid 0.01% commission then economically they still don’t feel the cost like a solo staker would. In reality they would have to consider smart contract and slashing risk, and the usefulness of their LST receipt token in DEFI (even though most people ignore this and throw money around like crazy).
Both of these participants I see being much more resilient to the potential changes than the home/solo staker.
Dilution prevention is something that is great for all holders of Ethereum. However when observing participants of the Ethstaker reddit and other social media comments, the dilution prevention of other changes were not appreciated. EIP 1559 burning of gas to make Ethereum deflationary and benefit all ETH holders, was also seen as a redistribution from stakers to ETH holders. I fear that dilution prevention is not seen as an incentive to be a solo/home staker, as staking is not required to receive that benefit. Personally I can understand the logic, but the public has a history of misunderstanding the many Ethereum proposals and upgrades.
I agree with the problem that has been detailed, but I think the solution needs some more input.
The primary difference I see between LST operators and Home/Solo stakers is that an LST operator runs a large number of validators from a single beacon chain client (and a single node), whereas most home stakers have just 1 validator or very few.
If there was a way to have variable issuance based on how many validators are attached to an instance of a beacon chain client it could reduce the economy of scale benefits achieved by LST operators, encourage more unique nodes and incentivize the decentralization we want to encourage.
I suspect however there would be ways to work around such solutions, and some sort arms race between the spec and LST configurations. Unfortunately I don’t have a solution, I just believe there may be other solutions yet to explore.
I agree with the decision to re-evaluate staking economics given we now have staking data spanning multiple years. One of the main unknowns at this point in time is how restaking will impact the ecosystem, however I believe that we should design the optimal staking economics for Ethereum, irrelevant of restaking. Any deviations away from this ‘optimal’ due to restaking should be seen as restaking impacting the security of the network and should be avoided.
Whilst I agree that targeting a fixed staked supply solves issues associated with a super majority of staked ETH, it does not solve the winner takes all dynamic. Applying a yield curve that limits stake to say 30% of total ETH supply will lock in Lido as the winner of ETH staking. The only way in which smaller/newer staking protocols can realistically compete is to encourage capital away from Lido using incentives, something that Lido can easily combat with its own purchasing power. This entire ecosystem is already a Moloch trap where staking protocols are required to spend significant sums to attract capital and capping the total ETH stake will likely increase this problem. The worst part, however, is that real competition (outside of incentives) will likely be squashed, reducing innovation and technological advances that would otherwise lead to a safer staking ecosystem for both stakers and the network as a whole.
We know that yield is always going to be the primary driver of where/how stakers stake. The only way I can see us safely aligning yield with ecosystem security is introducing staking economics that fundamentally rewards stakers that take steps to improve network security, such as using minority clients, home staking, etc… Admittedly, this would get complicated very quickly and introduce technical challenges (i.e. how can you prove someone is a home staker or truly using X client), but I do not see how we can meaningfully take steps to improve the current dynamic with a simple shift in the yield curve.
One of the main arguments against solo staking at the moment is the lack of economic incentives to do so, mostly due to them missing access the benefits available to liquid stakers: A) being able to use staked capital in DeFi/CeFi, and B) economies of scale that large staking pools receive due to MEV. It is important to highlight here that this doesn’t have to be the case. Whilst it is not yet widely known, solo stakers are able to liquid stake against their own node via StakeWise V3. StakeWise DAO provides the liquidity and integrations for its LST, osETH, which solo stakers are permissionlessly allowed to utilise in DeFi. This opens up the door for solo stakers to borrow against their own nodes, leverage stake, and even restake by depositing osETH on Eigenlayer. I am of the opinion that native restaking will not be available to home stakers due to the expected competition across operators to be in the active set for AVSs (in the very least the most lucrative AVSs will not be available to solo stakers, further increasing the economic disparity). The ability to restake an LST minted from a home node goes some way to solving this. Alongside an LST, StakeWise V3 enables stakers to access a smoothing pool with zero costs and goes some way to solve economies of scale problem too.
When evaluating the economics between liquid staking and solo/home staking, it is vital to consider the ability for solo stakers to liquid stake in this manner as it fundamentally changes the dynamics. It is also vital to increase the awareness for solo/home stakers to liquid stake in this manner to further encourage the participation of home stakers.
Thank you for the in depth post. One question that I have is how can the “Real ETH Yield” be negative? With the inclusion of the EIP-1556 and the burning of the base fee the amount of ETH has decreased since the merge. As modeled by https://ultrasound.money/. Hence a holder of ETH who does not stake is not being diluted. I don’t think the model can ignore the effects of 1556 and simply assume that unstaked ETH is being diluted by the Beacon chain issuance.
I believe this footnote addresses your question, indeed the burn may shift up the real yield, but that shift is constant across the whole real curves, so does not change the relative quantities and thus the arguments of the post.
I’d like to share my perspective as an average Home staker to provide a more down-to-earth example :
If issuance approaches zero or goes negative, I’ll continue staking out of conviction and also because I don’t want to make the “effort” to withdraw my validator. Most home stakers, like me, want to avoid unnecessary entries or exits of their validators. This benefits the network.
However, if negative issuance persists, I’ll be forced to withdraw because it’s not sustainable for me in the long term (I will no longer have ETH at the end so it’s better to exit beforehand)…
This creates a re-entry barrier, necessitating the redeposit and setup of my validator again, an effort I won’t undertake if the low APR remains. I’m not a machine; I won’t cyclically enter and exit, as it demands effort and entails risks. I’ve sketched my stance on a graph to illustrate the gap between the my APR thresholds for exiting and re-entering. (The figures are quick approximations and vary among home stakers.)
This simple graph demonstrates that if the APR falls significantly low, I will not consider re-entering if it persists in that range, effectively establishing a ‘no-return buffer’.
It suggests that pushing issuance toward zero or negative risks driving away home stakers not out of conviction but due to re-entry barriers.
However, I believe there’s an equilibrium APR before reaching 100% stake. Solutions like MEV burn and reducing issuance seem simpler and quicker to achieve this equilibrium at a lower ETH staking ratio.
I don’t think it’s necessary to implement a complex targeting system, as I don’t believe (or hope) we’ll reach 100% of ETH staked. However, if implemented, I think it could cause side effects for home stakers.
First of all, great job. I read most of this proposal very carefully, and I agree with most of the points brought up in it. I also took to social media as to support the proposal and to address some of the popular criticisms:
There is a class of criticism that entirely misses the point as it assumes the SEC would monitor the internet to draw conclusions on whether ETH is a security based on what is written in this forum: https://twitter.com/JimmyRagosa/status/1774817554445299717 Even if this was the case, why would it matter when engineering the protocol and addressing its challenges outlined above? The SEC is not the primary customer we’re working towards to build the protocol.
Everyone who holds ETH should be more aware of how incredibly valuable it is for an asset to have a monetary premium. Educate yourselves here, for example On Monetary Premia | AIER. Everyone who has to pull 17 magic tricks a quarter to preserve their wealth’s value over time is intuitively aware of how valuable it is to JUST hold a true store of value. That should be the vision for being an ETH holder. Not to imitate the existing financial system where I have to cast 23 different spells on my fiat money such that it doesn’t insta depreciate, but where I can buy ETH and know that the network’s productivity (that goes far beyond just staking) takes care of preserving the value.
Then, this aside, I picked two quotes from the article:
To the authors, I would emphasize immensely this part of the proposal, which I think is counterintuitive for many ETH and LST holders. If I understood correctly, then the punch line here is, in reality, that with more staked ETH receiving yield, ETH inflates more in absolute terms, which means holding ETH becomes even less economical.
I think printing a curve showing the number of staked ETH vs. how high absolute inflation is could help to explain this or further understanding. E.g., I’d be interested in how much ETH is being printed when, e.g., 60M ETH is staked at 2% vs., e.g., 30M ETH at 3%.
I actually think a part of the requirement here should also be that “holding raw ETH is simply like holding money,” as to say that it should be straightforward to hold money and not require me to do 17 tricks to hold the actual money and not a derivative that is actually inflated away by someone who more sophisticated than me.
This, in turn, also means that we should see staking rewards merely as the counter good for providing a service and paying expenses for that service. And in that line of arguing, I’d like to say: Motivationally, I’m pretty sure that most validators already hold ETH for its money-ness or as an investment. ETH has also been a great investment, especially in a fiat currency environment where all value globally is rapidly inflated. So, we can hardly argue that we should pay ETH validators for the risk of holding >= 32 ETH. So then we actually pay them for an active internet connection, a computer, the amount of time they spend on maintenance, and the risk of participating in validating. And what is that cost? I doubt that it is, in reality, so high that barely anyone does it economically. I personally know of people whose non-technical family members also got a computer to stake ETH because it was apparently such easy money! WTF
If you think about how you have to put your fiat money into an ETF that then spends it on 500 US-based companies and has this incredible black-box complexity, all just to preserve the fiat money’s value over time, then this is the inverse of what we should be aiming for, and sadly, this is the direction we’ve been going towards. So I support this proposal! Make holding ETH economical and preserve its property of being money.
GEB controller seems robust enough, is doing a great job in RAI and related forks. It intakes a target and sets the rates accordingly. Plus, it’s very well documented and in-prod tested, with community tweakings.
