Rate Model & Yield Policy

Where the rate comes from

Rayyan uses Compound V3's native utilization curve. Governance sets the shape, the market sets the level, and each borrower's rate is sampled once, at borrow, and locked. There is no external oracle, no interest-rate benchmark, and no AMM, so no rate slippage or MEV: the rate quoted is the rate executed at any size.

Launch curve, anchored to the global murabaha secured-lending range (4.5–8.5%) that Islamic banks charge for collateralized financing:

Parameter	Value
Base rate	1.5%/yr
Slope (below kink)	8.5%/yr per 100% utilization
Kink	80%
Slope above kink	300%/yr (safety mechanism; the 80% borrow cap normally prevents reaching it)

So: 5.75% at 50% utilization, 7.03% at 65%, 7.62% at 72%, and ≤8.3% at the borrow cap, meaning the entire reachable curve prices inside the 4.5–8.5% murabaha anchor. The curve is cheap below the band: a ~3.2% lockable rate at low utilization undercuts today's variable market, which is the bootstrap entry point.

The yield identity

LP APY = curve rate × utilization × LP share

Everything in launch yield policy is about placing the pool on the right point of that identity, without ever touching an existing borrower's locked terms (rate-lock immunity is what makes active policy side-effect-free).

1. The POL utilization band: 65–72%

The treasury's protocol-owned liquidity is actively managed by the keeper: withdraw POL when utilization drops below 65% (raising rates and LP yield), redeposit above 72% (rebuilding the liquidity buffer and keeping new-borrow rates inside the anchor). Why these numbers: the upper half of the band clears T-bill parity under the 85/15 split (parity sits near 69% utilization); 72% leaves an 8-point buffer under the 80% cap for withdrawal shocks and keeps the band-top rate (7.62%) inside the murabaha anchor. The band is a bootstrap instrument. As organic depth replaces POL, the (Phase 2) RWA floor takes over the low-utilization regime.

2. The mudarabah split: 85/15, flat from day one

LPs receive 85% of streamed markup, the steady-state mudarabah profit-sharing ratio, applied from launch. No teaser split, no step-down schedule to track: the number you see is the number the protocol runs on. On-chain parameter (lpShareBps), bounded 70–99%, and any governance change applies only forward, with accrual settling at the old split first.

3. Enforcement money

Nobody profits from a borrower's non-payment. The maturity settlement fee, 0.25% of the amount settled with a 1 USDC minimum, is the operator's wage for selling the collateral as the borrower's appointed agent, and it never reaches depositors. A liquidation credits the borrower fair value less the collateral buyer's discount and rebates the unearned markup, so it leaves nothing for the protocol. Money with no rightful owner goes to a charity fund with a public ledger. Reserves build only from the treasury's 15% share of markup. See Shariah Compliance.

The live Sepolia deployment still runs the earlier terms (a 1.5% settlement fee on collateral value, liquidation at 93% against the full contract price) until the next deployment.

Result at the band

Utilization	Rate	LP APY (85/15)
65%	7.03%	3.88%
68%	7.28%	4.21%
72%	7.62%	4.66%

Governance playbook (documented in advance)

• Reviews run monthly during bootstrap and quarterly thereafter, as single-parameter changes with no code deploys.
• Held in reserve, slopeLow 8.5% → 9.5–10%: steepening the curve raises mid-band yield quadratically in utilization (+0.44pp LP APY per +1pp slope at 68%) while barely moving the low-utilization teaser. Constraint discovered in design: the murabaha anchor ceiling couples slope to the band top (rate(band top) ≤ 8.5%), and beyond slope 10 the shrinking band costs more than the steeper curve earns. Trigger to act: utilization pinned at the band top ≥4 weeks with POL fully withdrawn. Not in the first 60 days; not while utilization struggles below 60%.
• The base rate (1.5%) does not move for yield reasons, because parity-via-base requires pricing outside the anchor and kills the cheap-entry flywheel. The one legitimate future reason is that once RWA yield is live, the base must exceed net RWA yield ÷ LP share so every new borrow is LP-accretive versus idle deployment.

Phase 2: the RWA floor

Commodity murabaha on idle capital (above a 30% liquidity floor) via Bursa Malaysia's Suq Al-Sila, the standard Islamic-bank treasury instrument. Activation economics favor a licensed partner desk (revenue share, no fixed cost base: accretive from ~$10M TVL, ≈+0.5pp at 55% utilization) over an owned entity (fixed costs make it destructive below ~$50M TVL). RWA is a low-utilization hedge, since above ~70% utilization there is no idle capital over the floor, which is exactly why it pairs with the band: the band handles the high-demand regime and the floor handles the low-demand regime.