What happened. Five INTX+IB positions closed this week, all triggered by the same condition: funding spiking to an emergency threshold (−51% to −60% annualized) and forcing an immediate perp exit. In every case, the IB hedge leg was closed at the next market open, 2–23 hours later.
| Symbol | Funding earned | Net P&L (strategy) | True P&L (all-in) | Naked-leg gap |
|---|---|---|---|---|
| INTC | +$8.84 | +$535.96 | +$375.17 | 14.0h |
| SPY | +$29.40 | +$17.90 | +$5.11 | 20.0h |
| AMD | +$23.73 | +$12.23 | −$6.96 | 1.5h |
| SNDK | +$27.38 | +$15.88 | −$109.30 | 22.8h |
| MU | +$14.63 | +$3.13 | +$11.68 | 23.3h |
Why the two P&L columns diverge. Net P&L is the strategy's own accounting — funding minus fees plus only the naked-window price move. True P&L reconciles the entire hold, both legs, in full. For three of these five (AMD, SNDK, MU), the naked-window price snapshot used to book Net P&L came out to exactly $0.00 (the 2026-09-01 stale-quote fix correctly deferred the close by one tick rather than locking in a stale price) — but that doesn't mean nothing happened to the hedge leg's price over its full hold. SNDK's hedge leg moved against the position by enough to turn a nominal +$15.88 strategy result into a −$109.30 all-in one.
On the naked-leg outcomes specifically: of the five, INTC and MU had favorable naked-window price moves, SPY was roughly neutral, and AMD and SNDK had adverse ones. That's close to a coin flip, which is what unhedged directional exposure over a short window should look like — see When funding profit differs from total trade profit for why counting wins/losses this way can be misleading, and What happens when the perp closes before the hedge for the mechanism itself.
Open going into next week: TSM, GOOGL, QQQ, AAPL, MU, and NATGAS remain open, all fully hedged (no naked legs currently outstanding). See the Lab Portfolio for live figures.