Spend an afternoon mapping where onchain yield actually comes from, and a quiet problem emerges. The dashboards show hundreds of pools, dozens of protocols, an entire vocabulary of strategies: vaults, lending markets, liquid staking, restaking, points programs, basis venues. Trace each one back to its source, and the apparent variety collapses into a handful of underlying engines. Most of what calls itself "yield" is the same few risks, repackaged.
This is the on-chain yield monoculture. It is among the most underappreciated risks in the space, and it is the reason most institutional capital has stayed on the sidelines.
A handful of engines, wearing many costumes
Strip away the branding and almost every on-chain return resolves to one of a small set of sources.
- Lending interest. Someone borrows your asset and pays for it, with the rate set by utilization. Aave, Morpho, and Compound are the canonical venues.
- Token emission. A protocol prints its own governance token to subsidize deposits. The arithmetic is closer to a transfer from future holders to present ones, and the yield persists only as long as the emissions schedule does.
- Trading fees. Users pay to trade, and that payment flows to whoever takes the other side. Two distinct streams sit inside this category. Funding rate is what leveraged traders pay to hold a position. Trading fees are what every trader pays on volume, spot or leveraged. The receive side is anyone supplying capital to the venue: Uniswap and Curve LPs on the AMM side, GMX and Hyperliquid LPs on the perp side, and basis-trade desks intermediating funding.
- Staking and validation reward. The base issuance a network pays to secure itself, increasingly re-wrapped through layers of liquid staking and restaking derivatives.
- Arbitrage and MEV. Capital that captures price differences across venues, or extracts value from transaction ordering. Often invisible to the end depositor. Lido, for example, routes its validators' MEV through MEV-Boost and folds the proceeds into stETH's reported yield, so an LST holder earns MEV revenue without ever opting into it.
- Underwriting and insurance. Depositors stake capital as a backstop against protocol losses and are paid a premium for sitting in the loss-absorbing tranche. Aave's Umbrella module is the live example: stakers of stkETH or stkUSDC underwrite Aave bad debt and earn a slice of protocol revenue in exchange.
That is essentially the whole menu. The thousands of pools on offer are recombinations and re-wrappings of these six. A "delta-neutral vault" usually blends lending with trading fees. A "high-yield stable pool" is often emissions dressed as organic return. A restaking position can stack validation reward, lending, emissions, and MEV into a single token whose risk profile no depositor fully sees.
And every engine on that list lives inside the same closed loop. The collateral is on-chain. The fees are paid by on-chain users. The tokens are the protocols' own. The only meaningful TradFi yield that has crossed onto chains so far is tokenized treasuries, a single thin straw drawn from one market. The aggregate effect is a yield universe that looks vast and is, in source terms, narrow.
This is where Jetstream sits. Global Basis+ taps a yield source from outside that list entirely: the carry available in global equity and commodity markets when spot is hedged with the corresponding futures. The trade is run in size in TradFi and delivered to on-chain capital through a regulated wrapper. The premium reflects the structural work of building and operating that bridge, and the fact that very few people are doing it well. The monoculture problem sits downstream of this gap.
Why monoculture is fragile
A monoculture is efficient until the thing it depends on fails. Then it fails everywhere at once, because everything was the same thing underneath. On-chain yield shows this property in three distinct ways.
Correlation hiding as diversification. A depositor who spreads capital across a lending protocol, a liquid-staking token, and a restaking vault feels diversified. In reality all three may sit on the same underlying collateral asset, the same oracle, or the same bridge. When that shared dependency breaks, the "diversified" book moves as one. The KelpDAO rsETH exploit on April 18, 2026 was a clean illustration. A single-verifier flaw in the LayerZero bridge let an attacker mint roughly $292 million of unbacked rsETH on Ethereum, and within minutes that synthetic supply was already posted as collateral on Aave to borrow WETH against itself. Lending markets that looked unrelated to a staking-derivative bridge ended up directly absorbing its failure.
Reflexive collateral. A great deal of on-chain yield is earned by posting one yield-bearing token as collateral, borrowing against it, and earning another. Each layer looks individually sound. Stacked, they form a chain where a wobble at the base, a depeg or a reprice or a paused market, transmits upward with leverage. The monoculture problem compounds here. The engines are few, and they are plumbed into each other.
Mercenary capital. Because so much yield is emissions-driven, much of the capital chasing it is transient by design. It arrives for the incentive and leaves the moment the incentive fades or fear appears. On-chain liquidity becomes structurally flighty. Pools that look deep in calm markets can empty in hours, precisely when depositors most need the exit.
The cohort that the monoculture leaves out
There is a particular kind of capital that finds almost nothing it can use in this landscape, and it is the cohort Jetstream was built to serve.
A regulated fund, a corporate treasury, a family office operating under a compliance mandate, any institution that answers to auditors: such an allocator needs to know its counterparty, satisfy KYC and AML obligations, document the provenance of every position, and explain to an investment committee, a regulator, or an auditor exactly where a return came from and what could impair it. It needs operational controls: defined custody, segregation of duties, a paper trail.
Almost none of the prevailing on-chain yield is built for this. It is pseudonymous by default, permissionless by design, and opaque about its true risk source by accident or convenience. An allocator with a fiduciary duty cannot deploy into a pool whose counterparties are unknown, whose yield is an undisclosed blend of six engines, and whose liquidity is mercenary. The return can be excellent. The position still cannot survive a diligence process.
So this capital sits out, or confines itself to the narrowest possible sliver of custodial stablecoin products and tokenized treasury bills. It accepts a thin, single-source return because the rest of the opportunity set is structurally inaccessible. The monoculture does more than concentrate risk for the participants already inside it. It actively excludes the most risk-disciplined capital in the world.
What would have to be different
The cohort that has been left out needs a different kind of yield product, defined less by the headline rate than by its structure. Four properties:
- Knowable counterparties. Identity and KYC built into the venue so an institution can satisfy its obligations rather than route around them.
- A legible, single-named risk. A return whose source is one clearly articulated thing the allocator can underwrite. No undisclosed composites. No token-emission subsidy presented as organic yield.
- Real custody and segregation. Assets held where an auditor can see them, with operational controls that resemble traditional finance.
- Genuine, non-reflexive backing. Returns that do not depend on stacking yield-bearing collateral into a reflexive tower, and that do not evaporate the moment incentives or sentiment turn.
None of these are exotic. They are the baseline assumptions of every other asset class an institution touches. Their near-total absence on-chain is the gap.
This is the design problem Jetstream is built around. Global Basis+ was built against those four properties. The fund operates under Cayman law with CIMA oversight. Custody runs through Fireblocks, with segregated assets and a defined paper trail. The risk has a single name: the basis between spot and futures in global equity and commodity markets, with directional price exposure hedged out. An investment committee can read the structure end to end and understand what it owns. And because the engine itself lives outside the on-chain silo, the return profile is structurally uncorrelated with the failure modes that travel through it.
The maturation of on-chain finance will be marked by the arrival of yield that a serious risk manager can actually say yes to, and by the arrival of yield sources that do not all share the same underlying engines. Those are the design problems worth solving next.
Jetstream Research publishes on market structure and risk in tokenized markets. This note is for informational purposes only and is not investment advice.