Parameter bounds

Every governable parameter in XPower Banq has two kinds of bounds: absolute bounds (hard limits enforced by the supervisory contract) and per-cycle multiplicative bounds (the lethargic constraint).

Absolute bounds

These are encoded into the supervisory contracts (PoolSupervised, PositionSupervised, OracleSupervised, VaultSupervised) and cannot be changed by governance. They define the universe of valid parameter values.

Parameter	Min	Max	Default	Notes
WEIGHT_SUPPLY	0	255	170	Supply-side weight (per token); 170/255 ≈ 66.67% LTV
WEIGHT_BORROW	0	255	255	Borrow-side weight (per token)
SPREAD	0%	50%	10%	Half-width interest spread
UTIL	0%	<100%	90%	Target utilisation
RATE	0%	<100%	10%	Target rate at UTIL
LOCK_BONUS	0%	SPREAD	10%	Max supply bonus on full lock
LOCK_MALUS	0%	SPREAD	10%	Max borrow malus on full lock
FEE_ENTRY	0%	50%	0.1%	One-shot deposit fee
FEE_EXIT	0%	50%	1%	One-shot redemption fee
DECAY	0.5	1.0	0.944	Oracle EWMA decay (α)
LIMIT	1 second	1 day	1 hour	Oracle refresh interval
DELAY	1 week	3 months	14 days	Feed-enlist delay

The bounds for LOCK_BONUS and LOCK_MALUS against SPREAD are dynamic — the supervisor refuses any new SPREAD value below either the bonus or the malus.

Per-cycle multiplicative bounds

Within a single governance cycle, each parameter can change by at most a factor of 2× (and at least a factor of 0.5×). Repeated cycles can compound, but each individual change is bounded.

The general form, where θ₀ is the prior parameter value and θ₁ the proposed:

$$\frac{1}{2}\cdot {\theta }_0\le {\theta }_1\le 2\cdot {\theta }_0$$

So:

• WEIGHT_SUPPLY = 170 can be moved to anywhere between 85 and 255 (capped at the absolute max) in one cycle — i.e. effective LTV between roughly 33% and 100%.
• SPREAD = 10% can be moved to anywhere between 5% and 20% in one cycle.
• A larger move requires multiple cycles.

How long a "big" change takes

Multiplier	Cycles needed	Time at 1-month cycles
2×	1	1 month
4×	2	2 months
10×	4	4 months
32×	5	5 months
64×	6	6 months

The arithmetic: ⌈log₂(M)⌉ cycles to compound to a multiplier M. The supervisor verifies these bounds inside the contract, so a malicious or compromised governor cannot bypass them.

Where to go next

• Transition curves — how each approved change is phased in
• Parameter catalog — the full list of governable parameters
• Change-rate constraints — the implementation of these bounds