How the Money Flows
We just saw the trader run the funding loop and keep the funding for themselves - so the obvious question is: if the trader pockets the funding, how do the LPs and the protocol ever get paid?
The answer is one number we've already met in passing: the borrow rate. Everything in this chapter hangs off it.
The borrow rate is the bridge
To run the loop, the trader has to borrow. They borrow USDC against their tokens, and they pay interest on it. That interest is the toll Own collects for supplying the cheap, leveraged long leg. The trader is happy to pay it, because they're funding it out of the much larger funding income they collect on the outside venue.
Here's the trick: the borrow rate isn't a single made-up number. It's built in two layers.
Funding the trader collects (~13%)
|
| they happily pay borrow interest out of it...
v
Borrow rate they pay (~7%) = Aave base cost (~4%) + Own premium (~3%)
|
v
This premium flows to LPs + Market Maker + protocol
The first layer is the Aave base cost - roughly 4%. The collateral that backs the system isn't sitting idle; it's parked in Aave, the big onchain money market, where it already earns a baseline yield. That's the floor.
The second layer is the Own premium - roughly 3% on top. This is the part Own adds for itself, and it's the piece that pays everyone in the marketplace: the LPs who supplied the collateral, the Market Maker who keeps prices fair, and the protocol treasury.
So the ~7% the trader pays splits cleanly: about 4% is the Aave floor, about 3% is Own's premium that flows back into the marketplace.
And here's the part that makes the whole machine self-tuning: when funding is hot, borrowing rises, so LP yield rises automatically. Hot markets mean more traders want to run the loop, which fills the loan book fuller, which pushes the premium up - exactly when demand is hottest. Nobody has to turn a dial. The rule of thumb is simply to keep Own's borrow rate (~7%) comfortably below perp funding (~13%), so the trade stays profitable and the book stays full.
What LPs actually earn
So what does an LP actually take home? Their yield stacks in three layers, and the result depends on which collateral vault they deposited into.
| Income layer | USDC vault | wstETH vault | cbBTC vault |
|---|---|---|---|
| Base yield (earned elsewhere) | 3.12% (Aave) | 2.55% (Lido staking) | ~0.2% (idle BTC ≈ 0%) |
| + Lending premium (borrower interest) | +0.8% | +0.6% | +0.7% |
| + Trading spread share | +0.6% | +0.6% | +0.6% |
| = Organic LP yield | ~4.5% | ~3.8% | ~1.5% |
| vs. its own benchmark | +44% over Aave | +49% over stETH | vs ~0% idle BTC |
| With launch incentives | ~6.5-7% | ~5.5-6% | ~3.5% |
Layer one is the base yield the collateral already earns just by existing - Aave interest for USDC, Lido staking rewards for wstETH, roughly nothing for idle Bitcoin. Layer two is the lending premium - the borrower-interest share we just traced through the bridge. Layer three is a slice of the trading spread on every mint and redeem (we'll split that one up in Chapter 12).
The important habit here is to read each vault against its own benchmark, not against the others:
- The USDC vault competes with Aave and Morpho (3-5%). At ~4.5% organic, it wins on yield outright.
- The wstETH vault competes with almost nothing - Aave pays wstETH roughly 0% on Base, so this is one of the only places to earn an extra ~3.8% on top of staked ETH.
- The cbBTC vault competes with idle Bitcoin, which earns 0%. Even 1.5% is "something from nothing." Its real job isn't to top the yield chart - it's to diversify the collateral mix.
Across the board that's 30-50% above each benchmark on organic yield alone, and ~6-7% once launch incentives are layered on.
The negative-funding bonus
There's one more source of money, and it kicks in exactly when the main engine would otherwise stall.
Funding isn't always positive. On some names - TSLA recently, around −2.6% - funding is negative, meaning shorts pay longs instead of the other way around. The trader's loop doesn't work on those names, because the income that powered it has flipped to a cost. So who profits there?
The Market Maker does - and it shares the proceeds with LPs.
Negative funding = shorts pay longs
|
v
The Market Maker hedges the tokens it issued by going LONG the perp
|
| on a negative-funding name, the MM is PAID to hold that long
v
MM collects the funding --> shares it with LPs via the vault
|
v
Adds up to ~+2% to LP yield on those names
Remember from Chapter 6 that the Market Maker always hedges the tokens it issues so it doesn't care which way the price moves. On a negative-funding name, that hedge happens to be a long position - and a long that gets paid to exist. The MM collects that funding and routes a share of it back into the vault, lifting LP yield by up to ~2% on those names.
So the two regimes mirror each other:
- Positive-funding names (most of the time): funding goes to traders; LPs earn through the borrow premium.
- Negative-funding names: funding goes to the Market Maker, who passes a share straight to LPs.
Either way, the funding cycle pays LPs. We treat the negative-funding bonus as upside, never a promise - it depends on the market regime and can disappear when funding flips back positive.
What just happened
- The trader keeps the funding, but pays a borrow rate (~7%) to run the loop - and that rate is how LPs and the protocol get paid.
- The borrow rate = the Aave base cost (~4%, the floor the collateral already earns) + the Own premium (~3%, which flows to LPs, the Market Maker, and the protocol).
- When funding runs hot, more traders borrow, the premium rises, and LP yield climbs automatically - no manual adjustment needed.
- LP yield stacks three layers (base + lending premium + spread share): ~4.5% USDC, ~3.8% wstETH, ~1.5% cbBTC - each judged against its own benchmark, 30-50% above it, and ~6-7% with incentives.
- On negative-funding names the loop stalls, but the Market Maker's hedge gets paid to be long, and it shares that funding with LPs - up to ~+2% as regime-dependent upside.