Desk cockpit · S7 · risk
Risk & liquidation-NAV
PAPER · advisory · ~0 leverage · no real capitalThe dedicated risk screen — NAV, aggregate delta, per-venue margin health, the liquidation-NAV-by-size curve, modeled stress, and the portfolio kill-conditions. Honest by construction: this is a conservative paper / stablecoin-lending desk with ~0 leverage, so much of the classic «margin / liquidation» risk is genuinely LOW or n/a — that is the point, not a gap. The real exit risk lives in the liquidation-NAV-by-size ladder below. Live from api.earn-defi.com, fail-closed (stale shown explicitly; a missing number is "—"/UNKNOWN, never fabricated). Paper / advisory — not investment advice.
1 · NAV · aggregate delta · margin health
Portfolio-level vitals. Aggregate net delta targets ±0.5% (a market-neutral book). Margin health is n/a on a ~0-leverage paper desk — shown honestly, not fabricated.
Risk strip
Margin / liquidation health is n/a: the desk carries ~0 leverage (stablecoin lending + hedged carry). There is no maintenance-margin to breach — the conservative posture IS the answer, not a fabricated ratio.
2 · Liquidation-NAV by size — the real exit risk
THE differentiator. For a conservative desk, the real risk is not margin — it is EXIT liquidity: what you actually realise unwinding the carry books at size. Net proceeds fall and the haircut climbs as the ticket grows; a tier where contemporaneous on-chain depth does not cover the size is FLAGGED (a hole), never a fabricated fill. Conservative lower bound on forced-unwind proceeds — not a realized exit.
live book
3 · Stress scenarios (modeled)
The desk exposes no live stress endpoint, so these are MODELLED transparently from the exit-NAV ladder + the two-tier kill thresholds. Inputs are shown; nothing here is a fabricated live reading. On a ~0-leverage stablecoin book a price gap has near-zero direct NAV impact — the honest bite is on the hedged-carry exit and the funding leg.
worst-tier net (largest ticket)
—
exit-NAV ladder unavailable → stress inputs UNKNOWN (fail-closed)
worst-tier haircut
—
flagged tiers (depth-limited)
—
20% price gap (ETH/collateral)
Direct NAV impact: ~n/a — the book is stablecoin lending + delta-hedged carry (β≈0), so a spot gap is largely offset by the hedge leg. The real exposure is the DEPEG / basis residual and a thinner forced-unwind: the exit-NAV haircut widens as depth dries up. Gates: below the SOFT 5% rung → no action; a depeg-driven drawdown past 5% arms SOFT de-risk (halt new / no increase), past 10% arms HARD all-cash.
Funding flip (perp funding turns negative)
On a hedged-carry book a funding flip is a DRAG, not a liquidation: the short-perp hedge starts costing carry instead of earning it. The refusal-first rates-desk gate declines fresh entries whose fair-value carry no longer clears the tail-risk hurdle; existing legs de-risk toward the RWA floor. No margin call (~0 leverage) — the cost is opportunity, not solvency.
4 · Portfolio kill-conditions
The two-tier kill (SOFT 5% de-risk → HARD 10% all-cash, ADR-034/048) plus the sharpe / red-flags / manual conditions, each as a manometer with REAL headroom. THIN conditions (e.g. Sharpe below the min-bar gate on this early track) render UNKNOWN — never a fabricated headroom. approved=False cannot be overridden by anyone.