Risk tiers

Risk tiers translate collateral evidence into policy posture. They are qualitative by design. Lyrasing does not have deployed markets, active LTV parameters, borrow caps, liquidation parameters, supported collateral, or a numeric parameter engine in this repository.

The tiering model consumes labels from the AVS-risk methodology: observable inputs, exposure taxonomy, slashability/timing, and LTV adjustment posture. It should not duplicate that taxonomy. Its job is to say what those labels mean for a candidate LRT.

Tier model

Tier	Evidence profile	Policy posture
Reviewable	Backing, wrapper, redemption, liquidity, oracle units, AVS exposure, operator exposure, slash treatment, and insurance mapping are current and source-backed.	Eligible for future collateral review with explicit caps, refresh triggers, and stress tests.
Conservative	Core mechanics are observable, but one surface has material timing, correlation, liquidity, or discretionary uncertainty.	Lower cap, lower LTV posture, tighter looping limit, and shorter review cadence.
Compressed	Price/liquidity evidence exists, but AVS exposure, operator overlap, redemption timing, or insurance mapping is stale or incomplete.	Strong cap compression; admission only after risk-owner review and no aggressive leverage.
Delayed	A required primary input is missing, currently changing, or controlled by an unresolved governance/curator/process update.	No collateral factor yet; keep as research-only until the input stabilizes.
No admission	Slashability, backing, redemption, oracle units, or bridge/wrapper mechanics cannot be bounded from primary sources.	Do not list as collateral. Re-open only if the risk surface changes.

These tiers are not a hidden numeric ladder. A future implementation may map tiers to parameters, but that mapping would require a separate parameter model, oracle implementation, and review process.

How AVS-risk labels feed tiers

AVS-risk label	Tier pressure
Isolated direct exposure	Can remain reviewable if backing, liquidity, and timing evidence are also strong.
Known correlated exposure	Pushes toward conservative or compressed caps, especially when other listed assets share operators or vault paths.
Opaque exposure	Pushes toward delayed or no admission until the missing surface is observable.
Coupled insurance exposure	Requires capacity allocation before insurance can support tiering.
Long or ambiguous timing window	Pushes toward lower LTV posture and no looping until liquidation timing is reconciled.
Stale oracle or exchange-rate input	Pushes toward compressed or delayed, even when spot price looks healthy.

The price feed is intentionally absent from this table as a tier by itself. Price can confirm market state, but it cannot explain restaking commitments, operator concentration, vault epochs, slash processing, or insurance capacity.

Cap and LTV posture

Collateral policy should use caps before fake precision:

Policy lever	Tier use
Borrow cap	Bound total exposure when evidence is incomplete or correlated.
Supply cap	Prevent a thinly reviewed asset from dominating the collateral set.
LTV posture	Follow the tier after backing, liquidity, timing, and exposure evidence are current.
Liquidation threshold	Account for exit timing, queue stress, and price volatility together.
Looping limit	Compress faster than base LTV because recursion amplifies price, liquidity, oracle, timing, and slashing risk.
Review cadence	Shorten when operator sets, vault limits, redemption queues, or oracle units can change quickly.

No numeric values are assigned here. Numbers would imply implemented protocol logic and live risk ownership that do not exist yet.

Review transitions

Tiers should move when evidence changes:

Trigger	Expected transition
Fresh primary sources disclose a missing redemption or exposure input.	Delayed can move to compressed or conservative.
Operator overlap rises across the collateral set.	Reviewable can move to conservative even if the single-asset record is unchanged.
Withdrawal queue, vault epoch, or buffer stress lengthens exit timing.	Conservative can move to compressed.
Slash treatment or Burner/redistribution route changes.	Re-review insurance mapping before tier remains stable.
Exchange-rate/NAV source becomes stale.	Any tier can move down until unit accounting refreshes.
Bridge, wrapper, or protocol admin incident occurs.	Treat the affected variant separately and consider immediate delisting review.

Non-claim boundary

Risk tiers are review language only. They do not represent a live supported list, implemented LTV table, risk committee, oracle contract, governance vote, or market launch state.

Source notes

EigenLayer’s Operator Sets, Unique Stake, magnitudes, slashing, and safety-delay docs support the exposure and timing labels. Symbiotic’s vault epoch, opt-in, slashing, VetoSlasher, and Burner docs support the timing and settlement labels. Lido and ether.fi docs support wrapper, queue, and redemption examples used by the backing/liquidity review pages.