Portfolio Risk & Liquidation Cockpit

How to read it. Today's speed is how hard a normal down day hits: a 1% S&P drop costs about that many % of NLV. It is measured at today's prices, so it ignores the hedges' cap. Effective at -10% ... -40% is the share of NLV the scenario grid below loses, divided by the drop, with every hedge paying and every fortress stopping at its Max Loss: it falls as the drop deepens because the hedge puts take over. Max Loss / NLV is the absolute worst case: every fortress at its Max Loss (Repair's forward figure); above 1.00x means the account could end below zero. NLV if every stock goes to $0 also counts plain shares and unhedged short puts, which no Max Loss caps; the second figure assumes the rented hedges were not renewed. Max Loss and the $0 floors are from the Repair run of 2026-10-30 16:16.

Read across a row: at that S&P drop each name trades at the price shown (its own beta times the drop, never below 2% of today), and each account's NLV is what every position it holds is worth at those prices, hedges paying and fortresses stopping at their Max Loss. NLV under $250k is red. The same NLVs, with excess liquidity and the floor crossings, are in the grid below.

Where each NLV comes from (under the number): IBKR = the account's own Risk Navigator export (beta-weighted, volatility unchanged), used while no position has changed since it was taken; model, matched to IBKR = the model's loss scaled to that export (Repair's own fit, the same numbers behind its card), for drops the export does not price (it stops at -20%%: deeper, IBKR prints N/A) and for any account that traded since; model = no usable export. Excess at a drop is that NLV less the model's margin there (TIMS-style scan, calibrated to today's maintenance).

Moves are beta-weighted: an S&P -20% is a ~-55% move in the crypto/miner names at their betas. The grid and the first figure of each range hold volatility still, as IBKR Risk Navigator's what-if does (the yardstick since 21 Oct: it matched the model within 2-11% on all four accounts, and the crash-vol ramp was the optimistic end); crash-vol is shown second for reference only. "Forced selling" is the first move where excess liquidity hits zero; the PM floor is where NLV drops below ~$100k and IBKR reverts the account to Reg-T, the effective liquidation point for the thin accounts. Margin risk is the Repair card's verdict, computed by the same code: the worse of the model and the account's IBKR check (trusted 45 days), At risk at a 30% drop or less, Not today to 45%. * = deeper than the 30% IBKR's check reaches: model only, and unstable, because near a 32% S&P drop the high-beta names reach the model's price floor and the curve flattens, so a small input change moves the crossing many points.

Each name falls by its own beta times the S&P drop (IBKR's measured betas; never below 2% of today's price), and every leg the account holds on it is repriced there with volatility held still, matched to the account's IBKR reading like the grid above. A name whose loss stops growing between columns is capped by its hedge put; one that keeps growing has shares or an unhedged short put. Biggest = the largest loss at -20%; the grid's account totals include every name, and at -10% / -20% they are IBKR's own numbers where exported.

A deep-in-the-money put with less time value than the interest on its strike (cushion < 1.0x) is rational for the holder to exercise early; you would be assigned the stock at the strike and must fund it in cash (red = the account cannot, so it goes on a margin loan). Roll it down or buy it back to remove the overhang on your timing.

Selling stock frees roughly its margin interval times its value, and the position carries no convexity, so it is the cleanest maintenance lever. But the valve shrinks in the drawdown where you would pull it (a stock worth half as much frees half as much margin); it is a cushion, not a rescue, and it does nothing for the PM floor since it converts stock to cash without moving NLV.

Instant reval (if the name gapped to that price today), Black-Scholes at each leg's implied vol, all accounts, shares plus every option leg. "Breakeven vs today" is the price at which holding neither gains nor loses from here; it drifts up over time as near-ATM long calls burn time value. "Vs basis" adds back the unrealized loss already on the book (approximate; uses IBKR average cost).