Research

Ethereum needs capital that remains available for swaps when trading conditions deteriorate. Long-term ETH holders could supply part of that capital, provided the investment earns an acceptable return and the liquidity actually improves execution.
This is a treasury allocation problem. A holder can have a long investment horizon while the markets around that holder depend on capital that is actively managed, narrowly positioned, or free to leave. A deliberate allocation to onchain liquidity could connect those different horizons.
The case needs to clear two tests. Additional capital must make ETH more useful to trade and use as collateral. Its providers must receive enough compensation for the risks they take. An allocation that helps the ecosystem but consistently misses its investment hurdle needs an explicit funding rationale.
Measure the liquidity that users can actually access
A large ETH market capitalisation says little about the cost of executing a particular trade. Pool balances also provide an incomplete picture. In concentrated liquidity systems, the distribution of capital across prices determines how much is available near the current price. Uniswap v3 and v4 positions stop earning trading fees when the market leaves their chosen range.
The evidence also requires care with sweeping claims about onchain markets. Uniswap’s 2022 study found deeper ETH/USD liquidity on v3 than on several major centralised exchanges over its sample period. That comparison is historical, and its main depth analysis excluded transaction fees and gas. It demonstrates why a claim about today’s liquidity must be measured at a defined size, time and venue.
For a treasury, the relevant measure is the total cost of completing an order. That includes the quoted price, fees and execution conditions. A useful assessment should compare both directions of trade across several order sizes, then repeat the exercise when volatility rises. Routing matters too. CoW Protocol can combine different sources of liquidity, while UniswapX fillers can use their own inventory or external venues. An order that settles on Ethereum does not necessarily pay fees to a particular ETH pool. Any proposal to supply that pool needs evidence that its capital will attract useful flow.
There is therefore no defensible shortcut from “ETH is valuable”
to “this pool needs a large deposit.” We need to identify where execution is expensive or fragile, then test whether additional liquidity changes it.
Why durable liquidity can benefit more than its providers
Traders directly benefit when they can execute with less price impact. Lending protocols can also depend on executable markets when collateral must be sold. Aave’s liquidation process allows liquidators to repay debt in exchange for collateral; the economics of disposing of that collateral matter to the participant taking it.
The hypothesis is that reliable liquidity can support activity beyond the swaps that pay an LP. A protocol may be more comfortable accepting an asset as collateral when it can be sold under stress. A treasury may be more comfortable holding an asset when it has a credible exit route. Those effects could benefit ETH holders who never provide liquidity themselves. That is the public benefit we are sharing. It does not make every LP position a pure public good. A pool charges fees, each trade changes its reserves, and its capital remains exposed to loss. The useful claim is narrower: some benefits of accessible liquidity may accrue to people who do not pay the provider directly. This distinction affects funding. If fees cover the investment hurdle, a commercial allocation may be sufficient. If a measurable ecosystem benefit remains uncompensated, a foundation or other beneficiary could consider funding that service explicitly. The amount, duration and expected benefit should be visible.
Long holding periods alone do not create durable liquidity. A provider can retain an LP position while its concentrated range becomes inactive. A capital commitment can also contain withdrawal rights that matter during stress. Durability needs a defined service, an operating policy and conditions under which capital can leave.
The investment case has changed
KPK’s updated research examines passive, full-range WETH/USDC liquidity on Uniswap v2. It is a useful baseline because the position requires no active range management. It is only one venue and one strategy.
The study sampled 10,000 positions lasting between one and four years. Of these, 8,030 had completed their sampled holding period by the data cutoff, 24 September 2026. Historically, 89.9% of those completed positions beat a portfolio that initially held equal dollar amounts of USDC earning Aave yield and ETH earning Lido staking yield. The median difference in annualised return was 5.3 percentage points. The historical median dollar CAGR was 17.3%.Those results depend on the fees earned during the historical period. The same research shows a much weaker case when historical price paths are replayed with recent fee and benchmark yield assumptions.
Scenarios | USDC lending + ETH staking | Median ann. return Δ |
|---|---|---|
Historical fees and benchmark yields | 89.9% | +5.3% |
Fixed fees based on the trailing 12 months | 58.5% | +0.7% |
Fixed fees based on the trailing 6 months | 0.0% | −1.4% |
In the six-month scenario, the assumed fee yield accruing to LPs was 2.96% a year. It was insufficient to clear the model’s yield-bearing benchmark at any tested ETH outcome. Scale weakens the case further when trading volume stays constant. In the twelve-month scenario, adding $5m to a pool with roughly $18.5m of capital reduced the share of positions beating the benchmark to 44.9%. The median annualised difference became negative, at −0.3 percentage points. More capital shares the same fees unless it attracts additional volume or improves the quality of that flow. A large holder’s capacity to invest is therefore different from a pool’s capacity to absorb that investment profitably. Fee allocation also matters. Uniswap’s current documentation assigns 0.25% of v2 swap notional to LPs and 0.05% to the protocol. Applying the old assumption that LPs receive the entire 0.30% fee would overstate their income under the active fee regime.
The evidence supports testing a liquidity allocation under explicit conditions. It cannot support a promise that holding a passive LP position for long enough produces a stable return.
Match the position to the treasury’s actual mandate
An ETH and stablecoin LP changes the holder’s exposure. As ETH rises, the pool sells ETH into the rally; as ETH falls, it acquires more. A treasury whose mandate is to maximise the number of ETH it owns must evaluate that tradeoff directly. In KPK’s historical sample, the LP beat a pure ETH holding in 63.1% of completed positions, and extending the holding period did not systematically improve that comparison. A result against an equal-weight ETH and stablecoin portfolio cannot be presented as the result against an all-ETH treasury.
The natural initial audience is a treasury that already wants both ETH and stablecoin exposure and can tolerate the position’s changing composition. Its benchmark should reflect the alternatives it can actually use. Staking and lending perform useful functions and provide competing sources of yield.
LP fees also compensate for adverse selection. Arbitrageurs can trade against a pool after the external market price has moved, transferring value from the LP as the pool catches up. Research on loss-versus-rebalancing formalises this cost. A high trading fee total can coexist with disappointing investment performance.
Risk depends on the assets and the implementation. Permissionless access to a pool does not remove the issuer controls of its tokens. Circle’s USDC disclosures describe blocklisting and contract controls, for example. A treasury must assess those dependencies alongside the pool contract and its own custody arrangements. The Ethereum Foundation’s treasury policy already places DeFi participation within a framework of treasury needs and risk limits. That is a relevant institutional precedent for evaluating an allocation. It is not an endorsement of this proposal.
Define the service before committing the capital
A practical first step is a bounded pilot with a published baseline. The baseline should show execution costs for ETH purchases and sales at several sizes, using the same market reference and recording the route. Ethereum mainnet and L2 venues should be reported separately. Measurements during ordinary conditions need to be accompanied by observations during volatile periods.
The strategy comparison should include full-range and managed concentrated liquidity. Full-range capital remains available across prices but spreads its usefulness over a broad range. Concentrated capital can provide more depth near the market, with an operating burden and the possibility of becoming inactive. The choice should follow the execution need and the treasury’s constraints.
Capital should enter in stages. Each stage should have a maximum allocation, a return benchmark and a review point. The provider should know who can change the strategy or withdraw funds. If someone is paying for an ecosystem benefit, that payment should be reported separately from trading income. The pilot should measure the improvement attributable to its capital, as far as the data permits. It should also report the provider’s net return, including operating costs, against the agreed alternative. Market conditions will change during the test, so a before-and-after improvement alone cannot establish causation.
A pilot should be able to conclude that a venue has too little profitable capacity, that a different design serves users better, or that the proposed service requires explicit funding. Those are useful results for the next treasury deciding whether to participate.
KPK proposes convening a small working group of treasury allocators, execution specialists and independent researchers to define that test. We invite participants who can contribute an allocation constraint, execution data or a competing strategy. The first output should be a public measurement standard and a pilot brief with conditions for proceeding. Capital commitments should follow that work.
Share this article

