Yield & Revenue

Every dollar of yield is a fee someone paid for a service

No emissions, no token printing, no IOUs. Lender APY is manufactured from borrower activity and market movement — and the protocol's own revenue attaches to the same flows. Here is the full accounting.

The six lender flows

Where lender yield comes from

originationEvery loan and every rollover pays a duration- and volatility-priced fee upfront — the option premium for the gap risk lenders write. Pays even in flat markets.
skim shareWhen positions appreciate past their high-water mark, the realized gain is split by the utilization curve: lenders take ~20% in the healthy mid-range, up to 95% when vault capital is scarce — scarcity prices itself. The rest pays the borrower's own debt down.
premiumsEvery band-triggered settlement tranche pays lenders a premium out of borrower equity — compensation for bearing the risk period.
reversal feesA borrower whose pending position restores buys it back — paying for the assignment that almost happened.
swap feesPhase 3: un-lent bucket USDC doubles as live pool bids; organic flow trading through them pays swap fees. Post-routing upside, not the base case.
assignmentPut-writer economics: bids fill at the lender's own quoted discount below reference — the compensated risk itself, priced by ladder depth.
The one-line model: lenders are short realized variance and long fees. Every direction of movement, plus time itself, produces a cash event. Only stasis at zero utilization pays nothing — which is why loans have terms and origination fees.
Who gets which flow

Payment follows the tick

Lenders don't share one pot. Each lender's deposit sits at a bid — the tick is simultaneously their lending offer and their standing buy order (there is no lender ask side; the ask is just borrowers' escrowed collateral). Rewards are not distributed equally — they're distributed by where you stood:

top-down fundingLoans draw cash from the top of the book downward: the $0.95 bucket lends first, the $0.60 bucket last — usually never. Credit-side flows (origination, skims, premiums, reversal and roll fees) stream back to the buckets whose cash actually funded the loan, pro-rata to each bucket's share of the principal.
the gradientThe lender at $0.95 earns the fattest yield and is first to be converted to tokens when a move persists. The lender at $0.60 earns almost nothing and will almost never own tokens. Risk and reward climb the ladder together, tick by tick — the yield curve across the book is the risk curve. Aggressive quotes aren't a problem to prevent; they're a service the book pays for.
within one tickUniform: everyone in the $0.80 bucket splits that bucket's flows pro-rata by deposit share. The resting (un-lent) portion of any bucket earns swap fees only, when trades cross its tick.
delegated curationThe "same rate for everyone" experience is a sleeve, not a separate product. Nobody owns the passive vault: anyone can register as a curator by posting the junior stake, and passive LPs deposit once and delegate to a curator — that curator's ladder is painted with the capital that follows them. Yield groups by sleeve: every follower of one curator earns that ladder's blended outcome pro-rata, minus that curator's market-set fee. Never across sleeves — a reckless curator's conversions land on their own followers only, and the track record that builds is what LPs shop on. Switching curators is an internal reallocation (respecting the same utilization gates as withdrawal). A pro quoting his own ticks is simply a curator with one client — himself.
the class splitOne layer above all of this, the utilization curve (next section) decides how each fee event divides between lenders as a class and the borrower's own debt paydown. That split breathes with utilization; the lenders' share then falls down the ladder by tick as above.
Three tiers of the same answer: across the book — by tick, top pays most; within a tick — pro-rata, identical; within a curator's sleeve — one blended rate for every follower. One doctrine at every level: payment follows the judgment that placed it.
Yield by market regime

Different weather, different flows

Indicative composition of lender APY in one mid-tier market (numbers illustrative — the Monte-Carlo work calibrates the real ranges):

origination + rolloversskim sharepremiums + reversal feesswap fees
Indicative APY
~4%
Dominant flow
origination
Utilization
moderate
Nothing moves, so time does the earning: origination and rollover fees carry the book. This is the floor — and it exists because loans have terms. A perpetual-loan design would pay ~0% here.
The utilization curve

Scarcity prices itself

0–10% utillender share of skims ramps ~5→20% — a sleepy vault takes a small cut.
10–70%flat 20% — the healthy operating range.
70–90%20→40% — capital getting scarce, lenders' cut climbs.
90–100%40→95% — near-full utilization pays lenders almost everything, recruiting deposits exactly when the market needs them.
This is the deposit-recruiting engine: stress raises utilization, utilization raises the lender share of every skim, and yield spikes during the event that needs fresh USDC. The curve does the marketing.
Why liquidity stays

Rented liquidity leaves. Bound liquidity compounds.

Two kinds of liquidity live here, and they're bound differently. Vault liquidity is the credit capital — single-asset USDC that backs loans and sets the mark. Pool liquidity is the venue depth swappers hit — and its core is the same capital, wearing a second hat. Vault side first:

can't leaveThe market-owned floor is permanently locked and ownerless — no withdrawal function exists, and fees accrete to it, so the base layer deepens with age. Sponsor seeds are factory-locked to milestones. Stickiness as a property of the bytecode.
term-committedAt healthy utilization, ~80% of deposits are out with borrowers on fixed tenors — those claims resolve on the loan schedule, not on sentiment. Only the un-lent slice is flight-capable at any instant; a stampede becomes a drainage schedule, never a stampede.
paid to stayThe utilization curve is counter-cyclical: as capital exits, the lender share of every skim climbs toward 95% — maximum fear is maximum yield, and leaving means walking away from the best APY the market ever pays. The pro-rata queue removes the run's engine (no first-mover advantage), so nobody gains by racing.
nothing to wait forMercenary capital exits when emissions end — there are no emissions to end. Origination pays in flat markets, skims in rallies, premiums in chop and stress: every regime pays something, so no calendar date is worth waiting for. Issuer streams and rebates pay only into time-locked positions — capital that would be mercenary never qualifies.
converts, not exitsWhen bids fill in a crash, that capital doesn't leave — it becomes tokens bought at chosen discounts, inventory that re-cashes the floor richer on recovery. An AMM's stress drains liquidity; this market's stress transforms it, at prices set in advance.
And the layers compound: liquidity that stays builds clean history → higher tiers → more borrow capacity per deposited dollar → more fee flow → more reason to stay. Depth becomes collectively valuable in a way every depositor shares — and none can take with them.

Now the pool itself — the depth a swapper actually hits — which inverts how every other DEX pool works:

the core is the bookThe pool's bids are the vault's un-lent cash. The venue is deep because the credit capital is bound — every vault-side layer above is automatically a pool-side layer. No other pool's depth inherits lending economics.
permanent nucleusThe locked floor's bids and the market-owned asks (reserved supply, dip-bought inventory) can never be pulled by anyone — and fees accrete to them. Burned-LP pools are "permanent" too, but xy=k drains smoothly to zero; this nucleus is discrete quoted depth that regenerates.
stress inversionA normal pool is shallowest exactly when needed most — LPs pull in panic. Here the bids can't flee below the queue rules, the floor holds by construction, and the counter-cyclical curve is recruiting fresh USDC mid-crisis. Venue depth is most reliable during stress, because the ladder is literally the crash absorber.
mercenary fringePlain two-sided LPs (door 3) can come and go freely — classic AMM positions, classic IL, swap fees only. The design doesn't try to bind them because they never matter: they back no loans and never touch the mark. Flighty liquidity is welcome garnish precisely because nothing depends on it.
The inversion in one line: a normal DEX bribes mercenary LPs to stay and prays they don't leave in a crash. Here the venue's core depth is credit capital that structurally cannot run — the pool is deep because the vault is bound, not because anyone is bribed.
How the protocol earns

Revenue attaches to the same flows

All splits below are fee-switch surfaces, placeholders pending final tokenomics — the design decision already made is where revenue can attach, not the numbers:

origination bpsA protocol share of every entry and rollover fee — scales with borrowing activity across all markets.
curation feeMarket-set, not protocol-levied. Curation is an open seat: competing curators price their own sleeves, and competition disciplines the fee. Our flagship sleeve earns its fee the same way everyone else's does — by track record — so this line scales only if we stay the best allocator on our own rails.
settlement shareA slice of deleverage premiums and reversal fees — revenue that scales with volatility, not just volume.
graduation feeLaunchpad candidate: a small percentage of curve raises at graduation — paid once, by markets being born.
RWA servicesPermissioned-market setup and operation for issuers — the B2B lane: whitelist management, NAV routes, calendar config.
What the protocol never earns from: token emissions (there is no token), the executor's ε (that pays keepers, not us), or lender deposits themselves (no spread on idle capital — un-lent USDC works as pool bids for the lenders' own benefit).
The bright lines

Rules that survive any future decision

no emissionsYield is never printed. If a number on this page can't be traced to a fee someone paid, it doesn't exist.
pricing is sacredNo incentive program — ours or an issuer's — ever discounts an entry fee, widens a cap, or shallows a ladder. Subsidies may pay on top of honest prices; they never touch them.
yield names its riskEvery flow above maps to a seat on the Risk page. APY without its risk line is marketing; with it, it's a price.
Illustrative mechanism walkthrough — numbers are pedagogical, not audited economics, not an offer. Fee splits are placeholders pending tokenomics decisions; the utilization curve reflects the existing FeeCalculator design.