To enable institutional access to rETH, we need custody support from major providers like Anchorage Digital, Bitgo, and Coinbase.
We’re actively engaging with these custodians, and a clear pattern has emerged: they typically require an implementation fee (upfront cost) to onboard new assets, plus some form of ongoing revenue share arrangement. While they show flexibility on implementation fees, revenue sharing models appear to be standard expectations.
This isn’t our first discussion around referral partnerships, but I wanted to open this up for community input.
Is there support for offering custodians a revenue share model? And if so, how should we structure and implement it?
We will continue to negotiate and connect with providers but in the meantime, I thought I would raise this.
I can see this revenue share model implemented similarly to the reward tree where a portion of a protocol revenue stream split with UARS is distributed among rETH holders in accordance to some rules/factors (eg, time and size of holding).
I cannot see how the protocol can afford an implementation fee though.
@langers What’s the expected cut for the revenue share arrangement? What is the typical order of magnitude for the implementation fee? Could this be structured as a higher initial revenue share for a fixed period, reverting to a lower ongoing rate once the implementation fee is fully paid?
Could we clarify a little what institutional access looks like?
Are we saying “Institution A holds rETH” and “Coinbase runs the nodes for Institution A”? Or is it more like the nodes stay as is and it’s “Coinbase holds rETH on behalf of Insitution A (which owns the position)”? Or is it something else I’m not grokking?
The initial implementation fee magnitudes were, frankly far beyond what the DAO can afford. Although, speaking to the teams they are less interested in the implementation fee than in having an ongoing commercial model.
At this stage, I think we should bring what we think is sustainable/sensible and negotiate from there. As a protocol, we have limitations and that is what I have expressed.
A custodian does make money from their clients and there is growing demand for liquid staking products so I think we have a certain amount of leverage to negotiate. Although clearly we want to set the stage for a long term relationship.
At this stage, there is no requirement for others to run nodes for the institution. I did raise this. However, we are still in discovery mode and their compliance teams are figuring this stuff out. We are focusing on what we can control now.
It would be best to only lock ourselves into a custodian deal, if we have an institution lined up to take advantage of it.
What do you think custodians would think about the STAR system? Is this what they are looking for - a commission on getting rETH minted - and it’s just a negotiation around how large that commission is? Or are they looking for different terms or a specific type of contract that would make the STAR system unappealing?
STAR also requires that the newly minted rETH be held for a certain period of time, so it is basically a percentage of rewards during the maturation period. I would expect that what the GMC really cares about here is that rETH is minted and held; making it a percent of rewards (instead of a flat amount) and making it continuous (instead of ending after 12 months) would not be dealbreakers.
The GMC originally allocated ~$90k for STAR, which is enough to incentivize ~25k minted rETH for a year (although paying that out right now would be tough). However, the plan was always to find more funding if STAR proved popular; the GMC views rETH demand as crucial to protocol success, and would go to the pDAO for more funding if STAR were able to improve the situation.
What I was really trying to ask was whether the GMC was an acceptable intermediary here. The GMC is not an entity that can sign contracts, nor are we creating a smart contract that can automatically split rewards off to a custodian. While the GMC could certainly fund a grant to allocate pDAO funds for this integration, I’m trying to figure out the specifics of how to make everything work.
Direct RPL incentives — the current Star incentive model
This approach requires a six-month lock-up, with rewards distributed only after the lock period. The timeline is relatively long, and the yield is not particularly attractive. More importantly, this model is likely unsustainable — especially if the RPL price continues to decline, which creates significant budget pressure for the GMC.
From my personal perspective, I wouldn’t recommend continuing to use RPL as the primary incentive or revenue-sharing mechanism.
Exploring more sustainable incentive models
Looking at commission structures across major LST/LRT protocols:
Lido: 90% to stakers, 10% to DAO / CSM: 90% to stakers, 3.5–6% to NOs
EtherFi: 90% to stakers, 10% to DAO / Instant unstake: 0.3% to DAO
Rocket Pool: 86% to stakers, 14% to NOs / 86% to stakers, 10% to NOs, 4% to RPL stakers
Using a portion of ETH staking rewards as incentives appears to be more sustainable in the long run. One possible path is for the GMC or pDAO to stake RPL and use the 4% RPL staking yield as incentives or shared revenue.
When this idea was discussed previously, @haloooloolo raised an important question around which nodes the RPL should be staked to. This likely implies the need for institution-operated nodes (which may increase node count), or alternatively identifying a trusted node set to stake RPL and direct the resulting 4% yield toward incentive sharing.
Another idea: supporting 2048 ETH staking nodes, where institutions directly stake 2048 ETH into nodes they operate themselves. The balance between rETH and ETH backing would need further discussion, and this would likely require protocol-level changes.
DeFi Vault
One alternative is to stake rETH into the RockSolid vault and use the additional yield as a revenue-sharing mechanism. However, from a security and risk perspective, this may be difficult for institutions to fully buy into.
At some point it starts to become attractive for megapools to go to a 10% commission structured as:
0% node_share
9% voter_share
1% pdao_share where pdao_share is for protocol+institutional incentives
I guess this goes back to basically requiring RPL to get commission (just for megapools) but that helps RPL price, too. I suppose this could stunt migration, but even with low migration of mostly people with large RPL stakes who want the ETH fee, the 4 ETH pools can make up for a lot of no-migrate minipools. People who don’t want RPL can stay on minis. New people don’t get an option, but we really don’t need new NOs.
edit: I suppose the two issues I see with this are 1. low enough migration and this doesn’t help rETH a lot since too many minipools are getting the 14% commission and 2. Dropping the node_share might make it take a lot of RPL to make up enough of that loss to make migration worth it.
Coming up with some numbers on what sustainable rates would be:
From an RPL holder’s perspective, rETH minting and holding results in 0.027 (solo APR) * 0.09 (voter share) = 0.243% yearly flow to RPL stakers, assuming we have enough node operators to meet demand (very likely). This gives us our higher bound - if we incentivized at this rate, all possible value capture would be going to the custodian (with nothing left to pay for development, infrastructure, liquidity, etc.).
If we assume that RPL needs a 10% APR for people to hold and stake it, this means 0.243 / 0.1 *0.015 = 0.036% of minted rETH value will flow to the pDAO every year. This gives us a break-even rate at which the GMC / pDAO would receive as much revenue as it spends on incentives. Since this relies on the market’s price for RPL, this can only be an estimate based on what the market might value RPL at.
Recap:
0.243% APR is a higher bound, where RPL holders would likely lose money (9% of solo staking APR)
0.036% APR is a break-even, where the pDAO receives back as much as it spends (1.3% of solo APR)
So from here, how might we decide how much is an acceptable amount for a custodian?