For the purpose of calculating the effect on welfare, it does not fundamentally matter who runs the validator. The staking service provider will pass on its costs to the delegator in the form of a staking fee. The delegator will make the decision to delegate stake based on the full cost (staking fee and, e.g., assigned risks), and the principle applies just the same. The welfare loss is the aggregate costs, which must be borne regardless of who runs the validator, and ultimately degrades Ethereum. You can for example think of it as printing new tokens but just using them to compensate for/pay the cost of illiquidity, hardware, additional risks, etc. In this case it is inflation that nobody derives a surplus from.
If you have not studied the topic before or something still is unclear, you could for example have a look at Figure 1 here where the full idea was first outlined (it was also implied earlier in the MVI thread). Or review the answer in the FAQ dealing with the topic.
For example, concerning your comment about “full costs”: The full cost of operating stake is best approximated as the reservation yield associated with the stake, that is to say the lowest yield at which that ETH would be supplied as stake. The supply curve thus implies the marginal cost of staking, and the aggregate cost of staking becomes the integral of the supply curve.
I tend to refer to ETH holders as such, but we should not separate ETH holders and Ethereum, as if we could decrease welfare for ETH holders without also hurting Ethereum. The native token of Ethereum (and any other decentralized cryptocurrency for that matter) is a central part of the architecture and we must always ensure that users have access to a sound trustless asset: ETH.
As for the composition of the staking set, it is discussed in Section 2.3 with references to the FAQ and the longer answer on this topic stating my position. As directly specified at the start of Section 2:
You must thus study the FAQ for an accounting of the effect on the composition of the staking set. Including that analysis in the post would have made it too long.
This feels like an unfair accusation since I am actively working on making it possible to stake with less than 32 ETH: the vision of Orbit SSF is to allow for 1-ETH validators. Consider for example my recent write-up on Vorbit SSF or for that matter Section 2.3 where I presented new research around incentivization to this end.
I understand your comment relates to your business interests. Let’s reflect on the economics here.
It should be an objective for Ethereum to make it possible to run a 32-ETH validator at a profit. If the quantity of stake grows very large this objective must be weighed against the many other objectives that we have, such as ensuring a trustless sound money within the ecosystem (ETH). In the extremely unlikely scenario where the supply curve becomes very low, both objectives could be slightly strained, such that it might be necessary to have a very cost-effective setup or stake say 48 ETH to remain profitable, and the quantity of stake will then also grow a bit beyond what we are comfortable with. This is unfortunate but part of balancing different trade-offs under an upward-sloping supply curve, as emphasized in the post.
It should not be an objective for Ethereum to ensure that someone staking 1 ETH can do so profitably, if the associated cost calculation involves paying for hardware and broadband, etc. That would imply that someone running a 32-ETH validator can take home a profit approaching 32 times their costs and would push us towards a regime where all ETH is staked. Note that we still wish to allow people with less wealth to run their own staking setup, which is the background to pushing for Orbit SSF with 1-ETH validators. But to profit from that, they would reasonably need to use existing hardware or broadband, or some other more effective setup. Staking with 1 ETH can still be an interesting technical experience, or a way to start the staking journey at a smaller scale.
We can further delineate two different staking regimes among those staking with less than 32 ETH:
- Stakers operating 32 ETH or more without owning that much ETH, say by running mini-pools.
- Stakers operating less than 32 ETH through some specialized setup, for example using distributed validator technology.
The first of these two regimes will remain profitable longer than the second, given that the staker derives income on the full staked amount, although must share those profits according to relevant capital costs to delegators.
If a staker operates less than 32 ETH as part of a squad-staking setup, then there is a possibility that it may not be able to remain profitable. However, the costs you suggest would be covered by the staking yield even under the extremely low supply curve and full MEV burn. The equilibrium with the red reward for the extremely low reward curve in Figure 5 (red square) is situated above 0.625%, and we find:
\frac{500}{2500\times32}=0.625\%.
At an ETH price of $2500 and a staking yield of 1%, a 32-ETH validator already generates $800/year. It is in my opinion unlikely that we reach an equilibrium yield below 1% under the practical endgame, even after implementing MEV burn. I would further find 0.5% extremely far-fetched. It is important to use realistic assumptions about the supply curve in the debate.
- Inflationary tokenomics does not inherently make Ethereum cheaper, other than by driving people away from the ecosystem, thus lowering transaction demand.
- The fiat-denominated value of each ETH token may decrease in an inflationary regime relative to what it would be in a deflationary regime, but this does not make transactions cheaper. If transaction demand (and, depending on interpretation, Ethereum market cap) stays the same, then the proportion of the circulating supply that must be paid to transact will stay roughly the same. The fiat denominated transaction price will not change and the ETH denominated transaction price will increase.
- What would happen the user is in a way the opposite of what you suggest—Ethereum would get more expensive to transact on. Consider the ETH token holder that gets it proportion of the circulating supply inflated away under a high issuance regime and identical transaction demand and market cap. They would now afford to make fewer transactions. Why do you presuppose that people would want to continue using Ethereum if we design issuance policy to make them poorer? Surely, some new cryptocurrency would eventually come along that strives to minimize costs to users and fulfill the promise of sound money that this technology holds, as discussed here.
- You seem to conflate monetary policy of nation states with the issuance policy of a cryptocurrency. Take the example of when central banks lower their policy interest rates. It reflects a commitment to buy government securities (like bonds), which floods the markets with new money. This money is put to work by banks and financial institutions, being lent to consumers and businesses, spurring demand in the economy (economic activity). In contrast, when a cryptocurrency holds the issuance rate high, its users are encouraged to lock up its tokens in staking, which ultimately lowers economic activity, making the platform less attractive for building apps that users could spend their tokens transacting on.
- We should not attempt to compel users to transact more by making the native token gradually lose value. It is not the way to build a flourishing economy. We do not have the power to retain users under inflation in the way that a nation state has the power to retain users of their fiat currency. And even if we had that power, we should not try to leverage it to this end. Furthermore, the inflation would come to benefit staking service providers but not users.
Ethereum will get cheaper to use by implementing the scaling roadmap, not by making its users poorer.
I find it disheartening seeing people argue for making regular users poorer for “the greater good”. In particular when it is the new financial institutions close to the source of new money in the economy doing it, essentially replaying a centuries-old playbook, but in crypto instead. It is in my opinion a bit immoral. Once again, read Foundations of minimum viable issuance for an accessible outline of what we are trying to achieve here.
This is a list of the ideas we are already pursuing. The practical endgame does not require any of these ideas being implemented. If the reward curve would have involved negative or zero issuance, the situation would have been different.
If you do not support reducing issuance, you can state your arguments for it already today. There is no need for a delay of several years. There is also no need for the excessive issuance stipulated by the current reward curve; this is already clear today and to delay the reduction by 4-5 years will not serve Ethereum well. What we can do is to cap the issuance rate at 0.5% at this point, solidifying it as a ceiling, and then return to the conversation in a few years. This is discussed in Section 3.4. I do believe even a 0.5% is too much at high quantities of stake, but it is at least not extremely excessive.
We cannot make the correlated attestation penalties too aggressive because that opens up for, e.g., discouragement attacks. I support correlated attestation penalties, but this is a nuanced issue and there is this idea that they will somehow substantially alter conditions for solo staking—they won’t.
Let’s first recap basic economics from point 4 of the previous answer: central banks lower their policy interest rate → by buying government bonds → to flood the financial system with money → that will be lent to businesses and consumers → spurring economic activity. We might not think that this strategy is appropriate, but there is at least some logic to it.
Now let’s review your assumptions: Ethereum should keep issuance high → to encourage people to… lock up their ETH staking it → precluding them from spending it on gas → making it harder for prospective app developers on Ethereum to find willing customers → …and somehow this is supposed to spur economic activity? There is no logic to it. While derivatives can abstract away some of the barriers, a high issuance is fundamentally a step in the opposite direction of spurring economic activity. Remember, Ethereum activity exploded under proof of work when there was no staking at all. There was just a lot of uncommitted capital with app developers trying to attract it.
Ethereum is fundamentally for its users. We should never assume control over our users. There are two meanings to that statement:
- We should not assume control over regular people’s intertemporal choice, trying to coerce them to transact more or less than what they want to. Inflating away our users’ savings with this goal in mind will only hurt them.
- We should not assume that such a control is viable in a digitalizing and globalizing world. If we degrade the user experience, the assumption must be that we will get outcompeted by a cryptocurrency that does not. Nation states have a different relationship to their “token users” than Ethereum, but the frictions they rely on might also be eroded over time.
Bitcoin’s failure was to not adapt smart contracts; let’s not conflate things. Repeating myself a little because it is important: We do not own Ethereum’s users. We cannot slap an inflation tax on them and expect them to stick around if a competitor with a native token without an inflation tax shows up.
It does not assume that stakers operate their own nodes and it applies to both solo stakers and delegating stakers. Concerning the statement: “Operators providing validation as a service do not suffer from inflation”, Ethereum does not exist to enrich staking service providers.
It should not be the objective of Ethereum to promote separation of labor and capital. We thrive on home stakers staking their own ETH, but they can of course stake other people’s capital as well. The argument that decentralized stakers do not stake their own ETH seems very convoluted. It seems more like an attempt to promote your company, fitting also with your advocacy for an inflationary issuance policy.
As for the distribution of decentralized stakers, my position on this very complex issue can be found here.
My post argues against a Category 4 change and in favor of a practical endgame. The full quote is
Your selective quoting with the associated comment makes it appear as if I find it okay to institute a stake cap and then reverse it, when it is clear that this is not the meaning in context. This is an inappropriate debate strategy and should be avoided.
In conclusion, we should not attempt to compel users to:
- transact by making the native token gradually lose value, or
- assume staking costs when unnecessary by stipulating excessive issuance.
We are building with users’ best interests in mind, nothing more and nothing less. The Practical endgame reflects this commitment to our users.