Bonus and malus

Locked positions adjust your effective interest rate. Suppliers earn a bonus; borrowers pay a reduced rate (the "malus" — meaning a smaller magnitude of interest, not a penalty). Both come out of the protocol's spread.

The mechanism

The default protocol spread is s = 10%. This is the half-width applied symmetrically to the gross rate R(U) (the utilization-dependent rate — kinked curve plus base risk premium):

• Borrow rate = R(U) × (1 + s) = R(U) × 1.10
• Supply rate = U × R(U) × (1 − s) = U × R(U) × 0.90

The supply rate carries an extra utilization factor U: interest is only earned on the borrowed portion of the pool, so idle supply earns nothing (see Interest rates).

When you lock, your effective spread shrinks by an amount proportional to your lock ratio λ:

• Locked supplier effective rate ≈ U × R(U) × (1 − s + s · λ)
• Locked borrower effective rate ≈ R(U) × (1 + s − s · λ)

At λ = 1 (fully locked), the spread compresses to zero: the borrow rate equals the gross rate R(U), and the supply rate equals U × R(U). The two sides do not converge to the same number — the supply side keeps its utilization scaling. The protocol earns nothing in spread.

What is λ exactly?

λ is the time-weighted lock ratio. It accounts for both how much of your balance is locked and how long the locks remain.

The protocol uses an O(1) algorithm to compute a normalised commitment depth in token-equivalent units — see Ring-buffer locks if you want the details. For the user-facing intuition:

• A permanent lock of all your principal → λ = 1.
• A 1-quarter timed lock of all your principal → λ ≈ 1/16 (since the maximum horizon is 16 quarters).
• A 4-year timed lock of all your principal → λ ≈ 1.
• A permanent lock of half your principal → λ ≈ 0.5.

The bonus scales linearly with λ.

Ceiling

The bonus is capped by governance parameters:

• LOCK_BONUS — maximum supply bonus (default = s = 10%)
• LOCK_MALUS — maximum borrow reduction (default = s = 10%)

Both are bounded above by the spread s per parameter. The protocol enforces this on-chain (the supervisor rejects any new LOCK_BONUS/LOCK_MALUS above SPREAD, and any new SPREAD below the existing bonus or malus) — it's not possible for governance to set a bonus larger than the spread itself.

As a corollary, the combined constraint LOCK_BONUS + LOCK_MALUS ≤ 2s holds whenever the per-parameter bounds hold. This isn't enforced as a separate require, but it's a theorem about the per-parameter constraints. The practical consequence: even at full lock and maximum bonus, the protocol's spread margin can compress to zero but never go negative — if everyone locks, the protocol simply earns no spread, it doesn't lose money.

Worked example: locked supplier

Suppose:

• Pool utilization is 50%, gross rate R(U) is 5%.
• You supply 1,000 XPOW and lock it permanently.
• Default parameters: s = 10%, LOCK_BONUS = 10%.

Your effective supply rate at λ = 1:

rate = 50% × 5% × (1 - 0.10 + 0.10 × 1) = 0.5 × 5% × 1.00 = 2.50%

Compared to an unlocked supplier in the same pool:

rate = 50% × 5% × (1 - 0.10) = 0.5 × 5% × 0.90 = 2.25%

You earn an extra 25 bps annualised by locking permanently. Over a year on 1,000 XPOW, that's 2.5 XPOW — which sounds small but is meaningful at scale, especially if you're locked anyway for cascade-protection reasons. Note both rates are roughly half the gross rate R(U): the supply rate is scaled by utilization (50%), so it never approaches the full 5% regardless of lock state.

Worked example: locked borrower

Suppose:

• Pool utilization is 95%, gross rate R(U) is 50% (post-kink).
• You borrow 1,000 XPOW and lock it for 4 years.
• Default parameters: s = 10%, LOCK_MALUS = 10%.

Your effective borrow rate at λ ≈ 1:

rate = 50% × (1 + 0.10 - 0.10 × 1) = 50% × 1.00 = 50.00%

Compared to an unlocked borrower:

rate = 50% × 1.10 = 55.00%

You save 500 bps annualised. On 1,000 XPOW, that's 50 XPOW/year — substantial.

When the bonus matters most

The bonus is structurally most attractive at high utilization. From the Nash equilibrium analysis:

• Below 90% utilization: lock breakeven period is several years. Locking is marginal.
• 90–95% utilization: breakeven drops to ~3 years.
• 95–98% utilization: breakeven drops to under a year.
• Above 98%: breakeven drops to a few months.

This is by design: when the protocol most needs cascade protection (high utilization, stressed market), locking is most attractive. The mechanism is counter-cyclical.

What about the protocol margin?

At full lock adoption (everyone locked, λ = 1 for all users), the protocol earns zero spread on locked positions. This is solvent — the protocol doesn't lose money — but it earns nothing.

In practice, full adoption is unlikely. Realistic equilibria are 10–20% lock adoption in calm markets, 40–70% in stressed markets, with the protocol margin landing in the 8–18% range of base interest. See the theory paper for the analysis.

Where to go next

• Cascade protection — the systemic benefit you pay for with your lock
• Locking positions — how to lock in the app
• Position parameters — LOCK_BONUS, LOCK_MALUS, SPREAD defaults