The real yield of hedged leverage, decomposed
Borrow against a yield-bearing asset, lever the spread, short out the price risk, and the fat headline collapses into a thin residual of four large, offsetting flows. We decompose net carry into its real drivers and show where it goes negative.
The pitch for hedged leveraged carry is seductive: borrow against a yield-bearing asset, lever the spread, short out the price risk, and collect a clean number while the market does whatever it wants. Stated that way it sounds like free money. It is not. Net carry is a thin spread between four large flows that mostly cancel, and that spread is far narrower than any headline.
Here is the claim, stated so you can falsify it: the net yield of a hedged leveraged-carry position is staking-or-lending yield times leverage, minus borrow cost, minus hedge funding, minus slippage, and each of those terms is the same order of magnitude, so the residual is thin, not the gross. Lever the carry and you lever the borrow cost in the same motion. The strategy lives or dies on whether the spread stays positive after every real cost, including a funding rate that can flip against you.
Why now
Carry only works when the asset you hold out-yields the asset you borrow, and that gap is thin and visible. Staking and money-market supply rates are real, attested, observable numbers on the venues themselves, and they are low enough that a small move in borrow cost or hedge funding eats a large fraction of the gross. That is precisely why the decomposition matters: when the inputs are small, the residual is dominated by whichever cost you forgot to count.
One honesty note up front, applied throughout. Borrow rates and perpetual funding float continuously, so any decomposition is a model against the inputs of the moment, never a promise. We will not present a modeled cost as if it were observed, and we publish the formula rather than a number so you can drop in your own inputs.
The mechanism
A Priime Loop position does four things in one structure. It (1) supplies a yield-bearing collateral, staked ETH or a lending-market deposit, that accrues a base yield; (2) borrows against it and recycles the proceeds to reach a target leverage of roughly two to three times, multiplying the base yield and the borrow cost together; (3) opens a short perpetual sized to the position's net exposure, so the directional P&L cancels and you are left holding a rate spread instead of a price bet; and (4) pays funding on that short and gas/slippage on every rebalance. What you keep is carry: the levered yield minus the levered borrow cost minus the hedge cost. Nothing about the price of the underlying is supposed to survive in the residual, only the spread.
No rates and no performance figures appear on this surface; the decomposition is published as a formula, not a number. Staking and supply rates are observable on the venues themselves; borrow rates and perpetual funding float and must be read at the moment you size a position.
- Net carry = base yield × leverage − borrow cost × (leverage − 1) − hedge funding × hedge ratio − slippage and gas
- Funding sign convention: positive funding = shorts are paid; a flip to negative funding turns the hedge from a small cost into a direct drag
- Leverage multiplies the yield and the borrow cost in the same stroke
Decompose the carry, and the spread appears
Net carry is not one number; it is the sum of one large positive and three sizeable negatives. The levered yield is the only thing the headline ever quotes. The borrow cost is levered alongside it, the hedge pays or receives funding, and every rebalance costs gas and slippage. What is left is a thin residual, and the gap between the gross a naive headline would quote and the residual a user actually keeps is the entire content of this post.
Gross is not net, and net is not guaranteed
The same decomposition, told as a waterfall from gross to what a user keeps, makes the erosion explicit. A meaningful share of the gross spread goes to financing and hedge costs before anything else, and protocol fees apply on top of what remains. The headline number and the kept number are separated by every cost in between, which is why the honest object to underwrite is the residual after every cost, not the gross.
Risk and failure modes
Hedged carry trades one set of risks for another. It removes the directional bet; it does not remove the risk that the spread inverts. Three failure modes break the thesis, and none of them is exotic.
1. Funding flips against the hedge. The hedge assumes the short either receives funding or pays only modestly. If perp funding turns sharply negative, shorts paying longs, the hedge-funding term can widen until it overwhelms the spread and net carry goes negative while every position is still solvent. This is the quiet failure: nothing liquidates, you simply pay to hold a position that no longer carries.
2. Borrow cost rises into the yield. Money-market borrow rates float with utilization. Because leverage multiplies borrow cost, a move in the borrow rate toward the supply yield compresses the spread far faster than it compresses an unlevered position: every point on the borrow rate is multiple points off the levered cost line.
3. Liquidation on a collateral or peg shock. Leverage on a lending market carries a liquidation band. A collateral depeg, an oracle lag, or a violent move that briefly outruns the hedge can push the health factor toward the threshold. The position deleverages proactively and the act band is sized for that, but no rebalancing assumption survives every market, and the protection is only as good as its assumption that rebalances execute in time.
How to verify this
The cost side of carry is a model, so the right way to check it is to substitute your own observed inputs. Read the live borrow rate and the venue's perpetual funding at the moment you size a position, drop them into the decomposition above, and the net updates directly; the model exposes every coefficient rather than hiding them inside a single headline. When a Priime Loop runs live, the position wallet and the rebalancing transactions are on-chain, so the realized borrow cost, funding paid, and slippage can be read back and reconciled against the model line by line.
The takeaway
Hedged leveraged carry is not free money and it is not a fat headline; it is a thin, leveraged spread between a real staking-or-lending yield and the cost of borrowing and hedging it. The unlevered edge is slim, and a levered residual exists only because leverage scales that slim edge while the hedge strips out the price risk. Lever the carry and you lever the borrow cost beside it; the net is whatever survives. The downside is not a dramatic liquidation but a funding flip that quietly turns the spread negative, which is exactly why the position deleverages proactively and why the honest object to underwrite is the residual after every cost, not the gross. Outcomes are not guaranteed and depend on market conditions and user and partner decisions.
Put the stack to work.
Priime Pools turns any liquidity pool into a delta-neutral position. Priime Loop runs leveraged carry, hedged every block. Self-custodial, exit any time.