[RFC] ETHplus Rebalance Proposal Q2 2026

[pete]

Hi @Ham, appreciate your responses as always:

@ham

@pete I don’t think those excerpts fully represented my position from the earlier discussion, particularly as the caveats immediately following them were omitted.

Fair criticism. I should have included the caveats - I highlighted the quotes I did because your stance on the quotes were different from your statements above. The broader context also includes your April liquidity analysis report mentioning OETH having strongest redemption profile in the test set, and @robtg4 separately calling it a natural replacement on that thread - they both point toward inclusion rather than against it.

AMO

@ham

While the AMO has been tested, e.g during the Curve Vyper exploit, it has never been tested during an incident to which OETH has direct exposure.

There is no other place the OETH Curve AMO has direct exposure - are you suggesting that the OETH AMO won’t be fully tested until Curve experiences another attack?

I am not sure what would convince you or any others having this same thought. Maybe an analogy? The airbags in my car haven’t been tested, and I’m exposed to the risks of driving on the road daily, yet I trust that they’re going to work when needed because Kia has tested them endlessly in their cars and adheres to certain safety standards. Origin has been running AMOs since 2022 across multiple chains and tokens aside from OETH: OUSD on mainnet, OS on Sonic, superOETHb on Base, and superOETHp on Plume (see AMO audits on OUSD, Sonic, Base, Plume).

The protocol hasn’t depegged under stress, which is what the AMO design is structurally engineered for: 100% collateralization is enforced mechanically, circulating OETH is backed 1:1 by Beacon Chain staking, and unbacked OETH in the AMO never enters circulation without being collateralized first.

@ham

Given the centralised nature of OETH liquidity, this is what concerns me rather than incidents where OETH itself has no direct exposure.

The centralization of the OETH AMO is limited to the Guardians’ ability to move liquidity out of the AMO into the OETH vault, where it can then be withdrawn by OETH holders. The OETH contract dependencies chart has always displayed this:

The Guardians move very quickly to respond to any bug bounty submissions, maintaining an average response time of 7 hours, placing Origin as the 4th fastest to respond when sorting by resolution time. Origin still maintains a $1,000,000 max bounty for critical findings.

@ham

However, your comments regarding the frxETH AMO are warranted. As we’ve both mentioned ETHplus already has exposure to an AMO via frxETH and this proposal aims to double their allocation. I just want to highlight that while the frxETH AMO can and has diversified via ezETH it currently only amounts to 0.7% of the AMO’s total ETH supply and is currently isolated to the Fraxtal chain. There is currently no exposure to pzETH (Frax Facts).

You are not wrong that the frxETH AMO only has 0.7% of the AMO’s total ETH supply allocated to ezETH and that it currently has no exposure to pzETH, but the proposal that brought ezETH/pzETH into the AMO sets caps at up to 20% combined for those assets, and up to 100% for wstETH/stETH, allocated by the Frax team “if deemed advantageous.” That’s a discretionary expansion that could be activated at any time without further ETH+ governance - Is that not a centralization vector that could cause issues in times of stress?

Worst case scenarios at scale:

• If the Frax team exercised the full 20% ezETH/pzETH cap, ETH+ would inherit roughly 4% basket exposure to these LRTs, if frxETH were 20% of the basket, without further ETH+ governance involved
• If the Frax team exercised the full 100% wstETH/stETH cap, and stETH depegged, ETH+'s stETH exposure at 20% frxETH would be ~70% of the basket, without further ETH+ governance involved

I’m not arguing the Frax team would make these moves - they’ve operated responsibly. But the same “centralized discretion + untested in stress” concern you’ve raised about the OETH AMO applies to the frxETH AMO at magnitudes here. At the current 10% frxETH, this is a smaller risk. At the proposed 20% frxETH, the discretionary surface is multiplied. The OETH AMO has no reliance on any additional LSTs or LRTs - it is structurally bound to ETH/WETH.

While we’re comparing AMOs, it is probably worth mentioning the POL accounting. The Frax proposal states “most of frxETH POL is on Curve”, with the AMO able to be moved up to 100% across asset types. Is frxETH’s POL included in the TVL figure being used for ETH+'s 10% dependency calculation? If it’s excluded the same way we’ve discussed for OETH, then disregard this accounting question.

Yield

@ham

I appreciate that OETH yield is calculated incorrectly on DeFiLlama and suggest we use the 30d MA on the OETH landing page instead.

Using other verified data sources is stated as an option within the methodology, so this we can agree on.

@ham

It’s futile trying to compare OETH yield over the last 6 months to the proposed baskets blended yield given the broad decline of ETH staking yields over the past year.

ETH staking yields have declined across all ETH LSTs, this isn’t specific to OETH. It is the addition of OETH’s AMO, and the bonus yield coming from those opted-out from yield, that allow OETH to earn yields higher than standard staking rates.

@ham

it’s my opinion that while OETH has a yield profile that matches stETH alongside a superior liquidity profile under normal market conditions, the yield profile underperforms frxETH

larger gains can likely be achieved by increasing the allocation to frxETH instead.

Every LST/LRT being considered for this basket has a yield profile that underperforms frxETH. The goal isn’t ‘beat frxETH on yield’ - that framing would push ETH+ toward a yield-maximization mandate closer to dgnETH’s original purpose than ETH+'s current safety-and-liquidity-first methodology.

On the actual numbers: OETH’s 30d yield is currently 2.42% per the landing page, at parity with stETH and above rETH at 2.11%. Adding OETH would lift the blended basket yield from where it sits today, before counting the liquidity and diversification gains.

Liquidity

@ham

the circulating supply of OETH is small relative to the total supply of ETHplus and would therefore need to be sized appropriately. At that scale, the allocation would likely not materially improve either the yield or liquidity characteristics of the basket

From the methodology itself:

@ham

• Curate a basket that seeks safety and liquidity over maximised yield or diversification profiles.

• Where safety and liquidity isn’t compromised, the collateral basket can be optimized for favourable diversification and yield profiles.

• Champion a deeply liquid collateral backing with slippage for minting and redeeming ETHplus less than or equal to 0.5% for sizes up to 20% of total ETHplus supply.

OETH can handle size multiple times larger than the amount I suggested in the RFC before reaching 0.5% slippage:

Versus the others in the basket at the same size:

Other LRT/LSTs in the basket:

The amount tested above at 12,010 ETH is roughly 4x the size of the proposed OETH allocation. At the actual allocation size, slippage on OETH is practically zero. A 2,700 - 3,000 ETH allocation lands OETH at 12.86% - 14.29% of the ETH+ basket at 21,000 ETH supply - that is comparable in size to rETH, frxETH, and ETHx in the current basket. That is real basket weight, not a marginal slot.

@ham

simplifying basket monitoring and management alongside reducing some governance complexity.

I don’t think these were goals of ETH+ within the methodology, but the operational cost of one extra constituent seems small relative to the cost of either a methodology breach on frxETH or leaving the strongest exit profile in the basket unallocated. The current ETH+ basket already runs at 5 constituents, so adding OETH wouldn’t introduce new operational complexity, it would just preserve the diversification profile.

Going forward

The OETH being discussed in this rebalance is materially different from the one in the original IP and RFC. We’re not discussing 2024 OETH anymore, especially considering OETH is the only major LST using cryptographic proof of reserves. OETH at 12.86% - 14.29% of the basket stays within methodology on every dependency metric, captures the strongest exit profile in the test set per your April analysis, and lets governors/holders compare the trade-offs directly. My ask: when the end-of-month liquidity analysis goes out, would you be open to modeling a 5-constituent alternative alongside the proposed basket? Happy to provide any data inputs that would help the modeling.