Discussion topic for Native Ethereum Delegation (NED).
This is an idea-stage proposal, not yet an EIP. I’m posting it to get feedback on the underlying direction and, if there is enough merit, hopefully develop a stronger community-driven specification before anything formal is proposed.
I’m approaching this mostly from the staking incentive and economic design side rather than as a consensus-layer researcher, so I’m particularly interested in corrections where my assumptions about the protocol are wrong.
Update Log
- 2026-08-07: Initial idea draft.
- 2026-08-07: Added split-invariance analysis and a possible bond-backed delegation capacity model.
External Reviews
None as of 2026-08-07.
Outstanding Issues
- 2026-08-07: Whether native delegation should exist at the protocol level.
- 2026-08-07: Whether a persistent delegation domain is the right economic unit.
- 2026-08-07: Whether direct concentration-sensitive rewards can be made resistant to economic identity splitting without real-world identity.
- 2026-08-07: Whether bond-backed delegation capacity is a better primitive, or one component of a broader concentration mechanism.
- 2026-08-07: How to avoid operator bonding disproportionately favouring already capital-rich operators.
- 2026-08-07: How delegated ETH should interact with slashing, exits and validator effective balance.
- 2026-08-07: What should happen to rewards reduced by a concentration-sensitive mechanism.
- 2026-08-07: How native delegation should interact with existing liquid staking protocols and DeFi.
The basic idea
Ethereum already has delegation economically.
An ETH holder who wants staking exposure without running infrastructure will generally use a staking provider, liquid staking protocol, exchange or similar intermediary.
So in practice we already have something like:
ETH holder → delegation/aggregation layer → validators
The protocol just doesn’t provide that delegation relationship itself.
That means Ethereum gets many of the economic consequences of delegation, including aggregation and concentration, while much of the mechanism determining how stake is allocated exists outside the protocol.
The question I’m interested in is:
If delegation has become a fundamental part of Ethereum staking anyway, should Ethereum provide a neutral native delegation primitive and let providers compete above it?
Native Ethereum Delegation
Under NED, an ETH holder could delegate through the protocol while retaining ownership of the ETH, subject to whatever lock and slashing conditions are necessary for that stake to secure consensus.
Conceptually:
ETH holder → NED delegation domain → validators
The delegator chooses where the ETH goes.
The operator does not take custody of the delegated ETH and cannot redirect it.
Existing staking companies and protocols would not need to disappear. They could still provide things like:
- liquid staking tokens
- DeFi integrations
- insurance
- institutional services
- compliance tooling
- interfaces
- analytics
- tax tooling
- validator discovery
Ethereum would provide the delegation primitive underneath those services rather than requiring the delegation relationship itself to be created externally.
Delegation domains
I think the relevant economic object may be a delegation domain, rather than an individual validator.
One delegation domain could operate one validator or thousands of validators, but the delegation economics would apply to the domain as a whole rather than resetting for every validator key.
A domain could contain or accumulate things like:
- delegated ETH
- operator economic commitment
- validator set
- performance history
- slashing history
- maturity
This separates validator count from economic delegation identity without requiring Ethereum to know the real-world identity of the operator.
That distinction seems important because a purely per-validator concentration mechanism creates an obvious incentive to split across additional validator keys.
Why concentration-sensitive economics?
Native delegation by itself could recreate the same winner-take-most dynamics that exist today.
Delegators may reasonably prefer established operators because of reputation, reliability, operational history, perceived safety or institutional familiarity.
I don’t think the goal should be to prevent large operators.
Instead, the protocol could introduce an opposing economic force:
As a delegation domain captures a larger share of native delegated ETH, additional delegation to that domain becomes progressively less attractive.
So there are two forces:
more reputation → more delegation
and:
more concentration → lower marginal delegation economics → more opportunity for competitors
There is no hard cap.
If users value a large provider enough, it can remain large.
The difference is that being the largest provider would not necessarily make becoming even larger the economically dominant outcome.
Illustrative reward model
One possible family of concentration-sensitive reward functions is:
$$
M(S)=\frac{1}{(1+kS)^n}
$$
where:
- $S$ is a delegation domain’s share of total native delegated ETH
- $k$ controls concentration sensitivity
- $n$ controls the steepness of the curve
For illustration only, using $k=10$ and $n=2$:
| Delegated Market Share $S$ | Reward Multiplier $M(S)$ | Reduction vs. Base |
|---|---|---|
| 0.1% | ≈98.0% | ≈2.0% |
| 1.0% | ≈82.6% | ≈17.4% |
| 5.0% | ≈44.4% | ≈55.6% |
| 10.0% | 25.0% | 75.0% |
| 20.0% | ≈11.1% | ≈88.9% |
| 30.0% | ≈6.25% | ≈93.75% |
These values are illustrative only.
I’m not suggesting this is the correct function or that these are appropriate parameters.
The point is the structure:
Concentration remains possible, but additional concentration becomes progressively less economically efficient.
The actual curve, parameter selection and even whether a curve like this is the best mechanism at all should be treated as open questions.
Why the naive concentration curve can be split
There is an important problem with applying a declining reward curve independently to pseudonymous domains.
Suppose a domain with delegated stake $D$ earns:
$$
R(D)=D \cdot a(D)
$$
where $a(D)$ is the reward per unit of delegated ETH and decreases as the domain becomes larger.
If the operator splits the same stake equally across $m$ independent domains, total reward becomes:
$$
mR(D/m)=D \cdot a(D/m)
$$
Since $D/m < D$, any genuinely decreasing per-domain reward function gives:
$$
a(D/m)>a(D)
$$
and therefore:
$$
mR(D/m)>R(D)
$$
So the problem isn’t specific to the illustrative curve above.
If smaller anonymous domains simply receive better per-ETH economics, splitting a large domain into smaller anonymous domains will improve the operator’s economics unless the protocol has some other scarce or persistent economic variable to work with.
This seems like an important constraint on the design.
A purely per-validator curve doesn’t solve it, and simply replacing “validator” with “delegation domain” doesn’t completely solve it either if domains themselves can be created cheaply.
Bond-backed delegation capacity
One possible way around that problem is to make attractive delegation capacity depend on operator economic commitment rather than identity count.
Let:
- $D_i$ represent delegated ETH
- $B_i$ represent slashable operator bond or committed capital
- $\lambda$ represent the amount of delegation efficiently supported by each unit of bond
A domain’s efficient delegation capacity could then be:
$$
C_i=\lambda B_i
$$
and its utilization:
$$
U_i=\frac{D_i}{C_i}
$$
One simple illustrative reward function could be:
$$
M(U_i)=
\begin{cases}
1, & U_i\leq1
\frac{1}{(1+k(U_i-1))^n}, & U_i>1
\end{cases}
$$
The exact function isn’t important here. The property I’m interested in is what happens when an operator splits.
Suppose an operator divides both its delegation and its bond proportionally between several domains.
For each new domain $j$:
$$
D_j=\alpha_jD
$$
and:
$$
B_j=\alpha_jB
$$
Then:
$$
\frac{D_j}{B_j}
\frac{\alpha_jD}{\alpha_jB}
\frac{D}{B}
$$
So the utilization ratio is unchanged.
If an operator has 1,000 ETH delegated against a 100 ETH bond, splitting that into ten domains with 100 ETH delegated and 10 ETH bonded in each does not improve the ratio.
In other words:
Creating more identities does not create more bond-backed delegation capacity.
To support more delegation at the efficient rate, the operator has to commit more slashable capital rather than generate more validator keys or domain identities.
That gives a useful split-invariance property that the original market-share curve doesn’t have.
Adding maturity to bond-backed capacity
Maturity could then be added as another economic friction rather than being expected to solve splitting by itself.
Let $A_i$ be a maturity factor between 0 and 1.
Instead of using the full bond immediately, define an effective bond:
$$
\widetilde{B_i}=B_iA_i
$$
and therefore:
$$
C_i=\lambda B_iA_i
$$
A simple illustrative maturity function could be:
$$
A_i(t)=1-e^{-t/\tau}
$$
where $t$ represents the age of an active delegation history and $\tau$ controls how quickly the domain matures.
The important part would be that maturity accrues from actual economic activity, not just the age of an empty domain.
Otherwise an operator could create hundreds of unused domains in advance, wait for them to mature and activate them later.
Now consider a mature operator that splits into new domains.
The original domain may have:
$$
A=1
$$
while the newly created domains begin with:
$$
A_i<1
$$
Their effective bond-backed capacity is therefore initially lower:
$$
C_i=\lambda B_iA_i
$$
and their utilization is higher:
$$
U_i=\frac{D_i}{\lambda B_iA_i}
$$
So splitting into new domains initially makes the economics worse.
As those domains establish genuine operating and delegation history, $A_i$ approaches 1.
But at full maturity, a proportional split only gets the operator back to the same $D/B$ economics it had before.
It doesn’t create the permanent yield increase produced by the original per-domain concentration curve.
So the progression becomes:
split → temporarily worse economics → eventually neutral
rather than:
split → temporarily worse economics → eventually higher economics
That seems like a much more useful role for maturity.
What this does and doesn’t solve
I think this is an important distinction.
Bond-backed capacity gives a fairly clean mathematical answer to the identity-splitting problem because the scarce variable is economic capital rather than the number of domains.
It does not, by itself, reproduce the original idea of directly reducing rewards simply because one operator controls a large percentage of the total delegation market.
A sufficiently capitalized operator could still grow very large while maintaining the required bond/delegation ratio.
So there may be two different ideas here:
-
Direct concentration pricing, where market share itself affects rewards. This directly targets concentration, but appears difficult to make Sybil-resistant without some credible aggregation mechanism.
-
Bond-backed delegation capacity, where delegation efficiency depends on operator economic commitment. This is naturally split-neutral, but targets leverage and economic alignment rather than market share itself.
It’s possible the second is a better primitive.
It’s also possible it should be one part of a broader mechanism rather than a replacement for concentration-sensitive economics.
I think figuring out that distinction is one of the more interesting questions here.
Operational concentration
There may also be a separate way to think about concentration that doesn’t require the protocol to identify real-world operators.
Economic identity and operational dependence aren’t necessarily the same problem.
If many validators are supposedly independent but fail together because they share infrastructure, clients, networking or some other common dependency, correlated behaviour itself can potentially become an economic signal.
That suggests a possible broader structure where NED economics consider more than one dimension:
$$
\text{delegation reward}
\text{base reward}
\times
\text{economic commitment factor}
\times
\text{operational correlation factor}
$$
The interesting property is that creating another pseudonymous identity would not necessarily help if the underlying operation still behaves as one system.
And if a large operator responds by genuinely separating infrastructure enough that the validators stop behaving correlatively, then it has actually reduced at least one form of concentration risk rather than merely hiding its identity.
I don’t know whether this belongs in NED itself, but it seems worth considering alongside the economic identity problem.
Why I think this could improve competition
The current staking market has strong positive network effects.
A large provider can benefit from:
reputation → users → liquidity → integrations → more reputation → more users
NED could introduce a counterweight without banning successful providers.
A smaller operator could compete through:
- comparable infrastructure
- good performance
- meaningful economic commitment
- potentially better delegation economics
The protocol still doesn’t decide where stake goes.
Individual ETH holders do.
That is the part of the idea I find most interesting.
Rather than trying to force decentralization, the protocol creates better conditions for competition.
Liquid staking and DeFi
NED would not need to create a native Ethereum liquid staking token.
Existing protocols could potentially use native delegation underneath their products and continue providing pooled liquidity, DeFi integrations, insurance and other services.
Conceptually:
Ethereum → native delegation
with protocols competing above it on:
liquidity → UX → insurance → DeFi → institutional products
The intention would be to standardize delegation, not the financial products built on top of it.
There is a legitimate concern that validator or domain-specific delegation positions could be less fungible than something like stETH.
I would expect application-layer protocols to remain useful specifically because they could aggregate diversified NED positions into fungible products.
Whether those wrappers would simply recreate concentration at a different layer is something that would need to be studied.
Existing staking providers
I don’t see NED as necessarily being anti-Lido, anti-Coinbase or anti-staking-provider.
The point is not to make those businesses disappear.
It is to move the base delegation relationship into a neutral protocol primitive.
Providers could continue competing through:
- reputation
- liquidity
- integrations
- insurance
- institutional relationships
- compliance
- interfaces
- other services
A successful provider could still become very large.
The question is whether the protocol can make that success depend more on genuine economic commitment and operating quality, and less on the ability of scale itself to continuously reinforce more scale.
Questions I’d especially like feedback on
- Is native delegation desirable at the protocol level in the first place?
- Is a persistent delegation domain a sensible primitive, or is there a better aggregation unit?
- Is there any identity-free mechanism that can make direct market-share-based rewards genuinely split-resistant?
- Is bond-backed delegation capacity a better way of creating economic competition than directly pricing market share?
- Can the bond/capacity relationship be designed without giving an excessive advantage to already capital-rich operators?
- Is the split-invariance argument above missing an important attack or edge case?
- Is maturity useful as a temporary anti-splitting cost if long-run splitting is already neutral?
- Should correlated validator behaviour have any role in measuring operational concentration?
- What should happen to rewards reduced by any concentration or utilization mechanism?
- How should delegated ETH inherit slashing while remaining owned by the delegator?
- How should delegation interact with Ethereum’s current effective-balance and validator model?
- How should delegation activation, redelegation, exit and validator exit interact?
- How would existing LST providers and DeFi systems interact with native delegation?
- Is there a simpler construction that achieves the same objectives?
What I’m hoping to accomplish
I’m not trying to prescribe a finished implementation here.
The idea is to make delegation permissionless, keep validator/operator choice with the ETH holder, and see whether the protocol can create better economics for a competitive staking market without hard limits or real-world identity requirements.
Put another way:
If Ethereum is going to have delegation economically anyway, can the protocol provide the neutral delegation primitive and make competition more attractive by construction?
The original concentration curve is one possible direction, but the splitting problem makes it clear that the actual mechanism probably needs another scarce economic variable.
Bond-backed delegation capacity looks interesting to me because proportional splitting is mathematically neutral, and adding non-transferable active maturity can make newly created domains temporarily less efficient without giving them a higher long-run return.
Whether that is actually the right construction is exactly the sort of thing I’m hoping to get feedback on.
If the direction has merit, I’d much rather see the specification develop through discussion with people already working on Ethereum staking and mechanism design than try to design the entire thing myself.
The working name is Native Ethereum Delegation (NED). Informally, the Flanders Protocol.