Parameter posture

Parameter posture defines what Lyrasing would review before permitting recursive leverage. It deliberately uses qualitative bands rather than launch numbers because this repository does not contain a deployed market, oracle contract, collateral list, or parameter engine.

The parameter posture should be downstream of the LRT collateral framework, the AVS-risk methodology, and slashing-insurance design. A loop should not receive a more permissive parameter stance than the collateral and insurance evidence can support.

Parameter families

Family	What it controls	Qualitative posture
LTV posture	Borrowing power against the collateral value.	Reviewable, reduced, materially compressed, or unavailable.
Liquidation threshold posture	The health boundary before liquidation can begin.	Normal review, tighter buffer, stress-only, or no recursion.
Supply caps	Maximum collateral growth for a candidate asset or route.	Open for review, capped, cap-to-small, or closed.
Borrow caps	Maximum borrow exposure against the collateral and borrow asset pair.	Matched to liquidity, reduced for correlation, or unavailable.
Loop-depth caps	Maximum recursive turns after initial supply.	Reviewable, low-depth, single-turn only, or no-loop.
Health-buffer posture	Required distance between account health and liquidation.	Standard review, wider buffer, stress buffer, or no recursion.
Isolation treatment	Whether correlated or hard-to-unwind assets can share risk.	Shared market review, isolated pair, isolated collateral, or excluded.
Insurance-capacity coupling	Whether capacity can support the repeated slash surface.	Mapped, thin, shared, pending, depleted, or unmapped.

LTV and threshold posture

LTV posture should be less aggressive than the base collateral review when recursive exposure is allowed. Liquidation threshold posture should also account for the fact that unwind may need several route decisions, not one sale.

Evidence state	LTV posture	Liquidation threshold posture
Backing, redemption, liquidity, AVS exposure, operator exposure, and insurance capacity are current and mapped.	Reviewable.	Normal review with explicit health buffer.
One input is weaker but observable.	Reduced.	Tighter buffer and lower loop depth.
Several inputs are correlated or stale.	Materially compressed.	Stress buffer and cap compression.
Slash surface, exit timing, or capacity is unresolved.	Unavailable for recursive use.	No-loop until resolved.

The goal is not precision theater. Qualitative posture is useful when it keeps the review honest: unresolved timing or unmapped insurance should lower appetite even when a price chart looks calm.

Caps and depth

Supply caps, borrow caps, and loop-depth caps should be reviewed together:

Cap	Compression trigger
Supply cap	Candidate collateral grows faster than liquidity, redemption capacity, or risk-owner review cadence.
Borrow cap	Borrow demand depends on correlated liquidity or a borrow asset that can move with the collateral.
Loop-depth cap	Repeated exposure consumes the same liquidity route or insurance capacity multiple times.
Pair cap	The collateral/borrow pair creates a circular exit path or shared oracle dependency.

Depth should compress faster than base LTV because it multiplies the same collateral thesis. A low LTV with high loop depth can still create a stressed unwind if every turn depends on the same route.

Isolation and correlation

Isolation treatment should be used when a loop cannot be safely grouped with other collateral or borrow exposure. Inputs that push toward isolation include:

a narrow operator set or concentrated operator cluster;
a vault or middleware route with epoch-bound exits;
a borrow asset that shares the same backing or liquidity venue;
unclear oracle units across wrapped, rebasing, or share-like assets;
insurance capacity that is mapped to one surface but not reusable across all loop turns.

Isolation does not automatically make a loop acceptable. It is a way to contain risk while the review determines whether the position is reviewable, conservative, compressed, or no-loop.

Insurance-capacity coupling

Insurance capacity should influence every parameter family. If capacity is mapped to a specific slash surface and current, it can support review. If it is thin, shared, pending, depleted, or unmapped, it should reduce caps, lower LTV posture, widen health buffers, and compress loop-depth caps.

This coupling is qualitative until an implementation exists. The policy record should say which capacity band was used and why, not hide it behind a numeric parameter.