Validator Staking Rewards for Ethereum Growth | Devansh Mehta - Ethereum Foundation

Next we have Devanch who will be doing a talk on validator staking rewards for Ethereum growth. Devanche leads public goods and governance at Ethereum Foundation.

Um hey guys um I think we are all in a bare market now. So all of our thinking naturally turns towards where is the money going to come from. And here I'm going to try presenting a very specific proposal for how we can actually coordinate at the protocol level and try getting more money for Ethereum growth.

Yeah. Yep. Um so yeah so this is a proposal for how validators can coordinate amongst themselves and come to consensus about what to fund and what not to fund. This is very experimental like I've never given this talk before. So uh like don't take it as something we are definitely going down but something as which we want to get feedback for.

Um so at one level we see that there's already an incentive for a lot of validators to fund Ethereum growth because by having more demand they like more block space gets consumed which leads to an increase in the ETH value which they use for staking and they can use the rewards to fund the ecosystem development. So at a very structural level we do see that validators have an incentive to make sure that the protocol stays up and many more use cases are being built on it but in practice we don't actually see this flywheel um uh actually in practice like no one is actually funding like like very few validators are choosing to put their money towards ecosystem development some exceptions like octant which is taking 100,000 ETH and you and putting the rewards for supporting Ethereum the DAO security fund recently is also taking the ETH and putting it to ecosystem development. But all of these are just one-off initiatives. But how can we do this at a more structural level and why isn't this happening organically? Um so I think it's like a classic prisoners dilemma sort of situation where you have two validators, validator A and B and right now we are stuck at this end of the quadrant which is neither of them fund.

If one of them funds then it's bad for them because their rewards are lower and it's good for everyone else. So in order to move to a situation where both are funding that it's good for A and B like first this one will fund then maybe this one and then eventually we'll come here but there'll always be free riders and that means that this dynamic is inherently unsustainable. We've seen voluntary taxation doesn't work in the real world and in a lot of cases. So this is why we are not seeing a lot of concerted efforts by the ecosystem to fund what they depend on. Um the like this part below will show maybe like the loss which we are facing as an ecosystem because of this.

Assume that there's some non-commercial infrastructure which is deliver which costs $50 to maintain. So that costs $50 and there are six projects each of whom get $10 in benefits from that existing. So the overall overall ecosystem is benefiting because $60 of value is created for these six projects and it only costs $50 to maintain but no single project really derives $50 of value from it. So as a result it doesn't get funded at all. Um so this economists call this dread weight loss and there's a lot of theory about how the economy which can minimize dread weight loss is the one that's going to eventually out compete the others and right now Ethereum does have dead weight loss because we haven't figured out ways to support the infrastructure together.

And if we ever want to out compete the traditional world and the traditional economy we have to figure out ways to minimize the dead weight loss. Um so yeah hopefully I've like I've explained why it's important to work on like in order to work on getting funding for the things that we depend on. So yeah let's move to the implementation. How would we like what is a concrete proposal for actually trying to work on this in a concerted way. So similar to how validators can put their flag up to show that they want to increase the gas limit, we are proposing that they can put the flag up to show that they are willing to put a percentage of their rewards into funding the ecosystem.

So if I as a validator put my flag up to 10%, then 10% of my rewards go to the ecosystem. Uh if 51% of other stakers also put their flag up. So if 51% put their flag up, all 100% of stakers have to part with a portion of their rewards. So it's a very similar mechanism to the gas limit where if a majority put it saying that yep, we can handle a higher throughput um then all the validators have to. So similar to that, if some of them put their flag up saying yep, we want to give some money then if a majority put it then all of them have to give the money.

So we overcome the free rider problem. We solve the prisoners dilemma with this approach. But the I think the biggest reason if you look at the past history of EIP which haven't worked which try to use pro like which try to use the protocol for coordinating funding um it's because of the distribution aspect. We don't want money from the protocol to go towards like scammers or bad projects because that impinges on Ethereum's credible neutrality and its reputation. So how should we actually like so so so how would if stakers agree on how much to give the next step is like where should that money go and we need to work on both of these in concert so I'm going to focus more on the second part because the first part is fairly easy and straightforward which is they put the flag up and that shows how much they want to give but the second part is much harder so I'll talk a little bit more about that um so there's a splitter contract which can split money between different entities or mechanisms And stakers can put the flag up and they can also put the recipient addresses that where do they want this money to go.

So they could say like you know I want to put 20% to retrop PGF. I want to put 30% to quadratic funding. They can even put some other like liquid like liquid DA which can vote on it. So they can put their own preferences and then um and then we can use a method of condor set voting to figure out uh which is the optimal distribution between all the stakers preferences. So assume that the val like in the example above the validator's ideal distribution is 20% to public key number one 0% to public key number two and 80% to number three.

And the contract at the moment, the splitter contract which is current king of the hill has a a 50 2525 and anyone can propose a new contract saying I think this contract aligns with the validator's preferences better. And then we use something called the Manhattan distance to see which one actually fits with their preferences. So uh so so we also ran a simulation of this with like a bunch of validators and their preferences and where the money would actually go. And the good thing is that it's a stable mechanism and I think the most important feature which is maintained is that since it's the validators that are parting with their with their rewards, they should be the ones to decide. Um so we like in the next slide we'll talk about some of the problems with this approach.

But this part is like we should just remember that validators have to do two things. One is they signal the gas limit or the they also signal how much money should go from their rewards to public goods and third they should indicate a list of addresses which they want the money to go to. So these are the three things which validators would now need to do if this proposal actually gets accepted and if it gets traction. So um so yeah what are the ramifications? I've got a lot of feedback on this from people uh at the EF and outside and even amongst validators and maybe the biggest problem is ster cartilization that um if 51% of the stakers put their own address like they all collude amongst themselves and they say let's steal money from the remaining stakers and we'll put the limit up to 100 and then we'll just put our own address so that we can get their rewards as well.

So creating that sort of cartilization is the biggest issue. There are ways to solve it by making it 66% or like by changing the parameters. But I think this is the single biggest issue that we haven't solved at a design level. The original proof ofstake construction was actually very neat in solving the staker cartilization. You should look at some of Lad Zami's work where he talked about how he dealt with the problem.

But in this proposal, if a majority of stakers put the reward as being high and then they put a recipient address which is just their own address, then that would be a big issue. So I think that's the unsolved part that we are still getting design work for. The second is that we start formalizing the amount that we are overpaying an issuance. So if a validator is saying I'm willing to put 10% of my rewards to funding the ecosystem like I've heard the argument I don't accept it but I've heard the argument that this means we can reduce issuance by 10% because they're willing to put 10%. Um I don't accept it because I think validators can have a higher time preference that they can say I'm willing to go for a loss right now because increased block space in the future will increase my profits.

Um but I think the real issue is the alignment between staking operators and those putting up the stake. So 90% of Ethereum is currently through staking operators whose incentive is uh they get a percentage from issuance and the validators care about their underlying ETH like the people who are putting the money at stake they care about their principle going up. So if it's the staking operators that are going to be voting on this mechanism and deciding whether it's up or down, then ultimately how do we solve the principal agent problem as staking operators are deciding on behalf of those entrusting their ETH to them. Um and how will they just not hide behind the fact that we have to look after the fiduciary interest of our uh of of those entrusting their stake to us. So we're going to put it to zero.

Um so that's like the bigger problem of like the incentive alignment between staking operators and those giving the ETH to them. Um and I think that comes to the long-term role of validators in the network. Are they do we want to even enshrine them as some sort of governance actor or are they just simply a service? Uh we can look at some of the other blockchains which where validator governance hasn't worked out very well. So we definitely don't want them to have a say in whether EIPs get approved or not at least in my opinion.

But I think in terms of making the protocol a house for a coordination so that there is no defection of free riding possible. I think that's another question and validators are one example but another example might be if different rollups are paying the base fee like the blob fee and a majority of them say we want to pay an extra amount of blob fee and use that to fund what research is on rollups then all of them have to everyone who pays a blob fee has to. So I think the larger question is do we want the Ethereum protocol to be this venue for coordination between actors in a way that doesn't allow free riding or defection. Um and I think like there are many like second or third order effects in terms of the validator of this validator proposal. Right now if I have to decide where to stake my ETH I can either solo stake it which is about 10% or roughly of the network is what I've heard or I can choose a staking operator to do it for me.

And right now the choice on which staking operator to use is mostly financial that which one is giving me the best rate which one has the least risk. But in such a if such an EIP did pass then we can actually choose operators based on values that this operator has put the flag up to 30% and I think 30% of rewards should go to funding the ecosystem. And I also like their choices that they're putting X amount to protocol guild Y amount to quadratic funding for early stage apps. And I actually agree with their distribution. So you can actually hopefully hopefully we would get more solo stakers because when I'm a solo staking it's not just a uniff undifferiated service.

I'm also influencing where the protocol funds its growth. Uh and I might also choose operators based on this rather than just purely financial uh decisions. Um, and yeah, I think that's the last part, like how should we pursue this? Like, should we actually do the research and get this filed as an EIP, which obviously carries some controversy and it can also go horribly wrong, but it's like you're filing the EIP and you're doing the research for it, or should we start trying to get some empirical data with an optional redirect where we ask validators to optionally put this redirect, but then we again fall into the same prisoners dilemma situation. So, this is where we are.

It's a new talk. Don't take any of it as the gospel. It's a very experimental one. And if any of this interests you, just like please come and talk to me. Thanks.