Paper research · advisory · not live

Liquidation-NAV by size

What we'd actually realize on exit, at size — not the marketing $1.00. For each market and each ticket size: the conservative lower-bound net proceeds, haircut and time-to-exit, against that market's own on-chain depth.

What this number is (and is not)

It is a conservative lower bound on forced-unwind proceeds (constant-product L/(L+S)) — not a realized exit and not a precise execution model. Near peg it understates real proceeds; it is published only as a defensible floor.
Depth is single-market, never aggregated. A forced unwind cannot route one market's exit through another's pool. The aggregate depth is disclosure-only.
Holes are shown as holes. If a market lacks enough contemporaneous depth to clear a ticket, the row is flagged no clearing exit — never backfilled with a fabricated figure.
Advisory — moves no capital. Every row carries a proof_hash reproducible with the standalone verifier.

model: constant-product AMM, conservative LOWER BOUND · validation: docs/RATES_DESK_VALIDATION.md#exit-liquidity

Portfolio-wide exit schedule — market × ticket

source: static / offline · —

Every open position and every priced market is run against its own contemporaneous depth. Thin markets honestly show holes. From the portfolio section of /api/rates-desk/exit-nav.

Markets

—

Open positions

—

Aggregate depth (disclosure)

—

Tickets

—

Loading the portfolio exit schedule…

Each market's schedule is a conservative lower bound against its single-market contemporaneous depth; depths are never aggregated. Holes are published, never filled.

Our live book (honestly thin)

—

The desk's actual open paper book — it is small, and we don't hide it: tickets larger than its depth are honestly flagged no clearing exit.

Loading the live book…

Illustrative — a hypothetical book on a real deep market

—

Illustrative only — a hypothetical book on a real deep market's contemporaneous depth, to demonstrate how the conservative model behaves at size. This is not our live position.

Loading the illustrative schedule…

Verify it yourself

Every row's proof_hash (live + illustrative + portfolio) is recomputable from that row's published inputs by the zero-dependency standalone verifier — with none of our code.

python3 verify_spa.py exit_nav.json

spec: docs/PROOF_CHAIN_SPEC.md §6 · scripts/verify_spa.py

Go deeper

Verify it yourself

Pull the verifier, re-derive this NAV

Refusal log

What the desk refused, and why

Honest fundability

Where the edge actually is — and isn't