House Delta — 2026-07-10

Desk Plan — House Delta — 2026-07-10

Briefing reference: projects/dadbrain/Analysis/briefings/2026-07-10-pm.md Scanner watchlist: intelligence/2026-07-10-watchlist.md — PRESENT Roster this session: DELTA (PM, sole active member — single-member desk, no consolidation to do)

Shared Market Read (House Delta)

• Event risk: LOW. Clean session, no scheduled macro data. The only calendar item is the USDA WASDE report at noon (agriculture-only, not market-wide). Next Tier-1 is CPI on 2026-07-14 (Tuesday) — outside today's window.
• VIX regime: LOW. 15.89 pre-market, essentially flat vs. yesterday's 15.84 close. Comfortably clear of the 30 hard-skip threshold, and clear of the 18-line I use as a soft degradation marker.
• Session character expected: mixed/rangebound, post-rally digestion. Two consecutive strong up sessions (Jul 8 bounce off Tuesday's semiconductor rout, Jul 9 continuation) put today in "does the AI trade get a third day of follow-through" territory. Pre-market is genuinely mixed — Dow futures modestly higher, Nasdaq futures slightly lower (QQQ pre-market ~-0.57%) — not a clean continuation signal either way.
• Dominant event, not on my instrument list but relevant to it: SK Hynix's $26.5B Nasdaq debut (SKHY) — largest foreign IPO in US history, 7x oversubscribed. This is thematic fuel for the whole semiconductor complex, not an NVDA- or AMD-specific catalyst. The watchlist is explicit: "NVDA/SOXX: No intrinsic catalyst today." That halo effect is the session's one live wild card for my instruments.
• What the desk is NOT seeing today: no idiosyncratic catalyst on either of my leaders. No data print inside my 10:00 ET entry window. No event-day skip triggered. Structurally the cleanest possible session to evaluate my edge, but "clean" and "likely to fire" are different things — a quiet macro tape does not by itself produce a first-hour ±1.5% leader move.

Single-member desk — there is no cross-member comparison to make today. The market read above is the full desk read; DELTA's section below is the desk's only call.

DELTA — NVDA→ASML, AMD→TSM (Lead-Lag Pairs, LONG-only)

Setup present: unlikely. Base case is a no-trade session. Estimated probability of a tradeable LONG signal: 5–10%.

• Session-character fit: Poor-to-neutral. My edge needs a first-hour leader move ≥1.5%, driven by real information the laggard hasn't priced yet. Today's backdrop is exactly the opposite of that — no NVDA/AMD-specific catalyst, a post-rally digestion character, and a mixed (not directional) pre-market. The one path to a signal is if the SKHY debut generates a broad semiconductor halo strong enough to push NVDA or AMD through the threshold on thematic momentum rather than a name-specific event.
• Pair-level read:
• NVDA→ASML (combined PF 1.85, primary edge driver): possible but low-probability. NVDA sits at the center of the AI trade revival narrative and would benefit most from an SKHY halo, but yesterday's BofA-upgrade catalyst is now absorbed and there's nothing fresh under it today.
• AMD→TSM (combined PF 1.61): possible but secondary. Per the July 9 reflection, this pair can fire on sector-wide surges without an AMD-specific catalyst — so I'm not writing it off the way I incorrectly did on July 8. But today's character is digestion, not surge, so I rate it behind NVDA→ASML.
• Sizing vs. event risk: Standard 0.75% if a signal fires. Nothing about today's LOW event risk or clean macro calendar argues for reducing size — the structural environment (VIX LOW, no hard skips) is actually my best-performing regime bucket (dev PF 2.59 in VIX<20). The problem today is signal scarcity, not signal risk.
• Entry/exit mechanics: No adjustments. 10:00 ET entry on whichever laggard's leader fired (or the larger |return| if both fire — no stacking), hard flat 15:30 ET, stop 2.0× ATR(14, 5-min).
• Key levels: NVDA prior close ~$134 range, still riding two up days; ASML ADR tracks NVDA with the lag that is the trade. AMD lower-beta, needs a strong sector-wide move; TSM lowest-vol of the four, needs the broadest move of all to clear threshold.
• Invalidation (stand down entirely): NVDA opens gap-down ≥1% (per the 07-08 finding, a bounce off a gap-down open is return-to-parity, not momentum — not tradeable even if it technically clears 1.5%); neither leader reaches ±0.75% by 09:50; VIX opens above 18; NVDA pre-market gap exceeds +2.5% (laggard would gap in sympathy, killing the lead-lag entry). One more, carried from my own discipline gap on July 8 and July 9: if a signal fires against this plan's own "unlikely" call, the default is to re-read these invalidation conditions before acting, not to treat a fired signal as automatic permission.
• Behavioral note carried into today: two of my last two sessions (07-08, 07-09) I traded against a plan that had already called the setup unlikely, on the weaker pair, without waiting for the primary pair to fail first. Both times the SOP mechanics were sound and the losses were within the strategy's own tolerance — but the discipline gap is now a pattern, not an incident. Today's plan explicitly rates the setup 5–10%; if SKHY-driven flow fires a signal, NVDA→ASML gets first claim, and I owe myself a genuine re-check against the invalidation list before entering, not a rationalization after the fact.

Desk-level verdict for DELTA: GO on gates (VIX clear, macro clear, no event-day skip), but the realistic expectation is no trade. If a signal fires, it is more likely NVDA→ASML on an SKHY halo than a name-specific catalyst — trade it at standard size if invalidation conditions are clear, but treat a fired signal as the tail outcome it is, not confirmation that today was actually a high-conviction setup.

Open Items (carried, desk-level)

• Live-start watch: 0W/3L through 2026-07-09, all three losses on AMD→TSM. AMD→TSM SHORT remains disabled (LONG-only) pending 30 LONG paper trades at PF ≥ 1.5.
• Rolling 6-month PF monitor active in paper_trade.py; no trigger condition today.
• SPY Calmar + random-entry control deliverables: closed 2026-07-04, no action needed.
• Event-day diagnostic run (skip disabled): still open, moderate priority — not applicable today, no event day to diagnose.

Plan Filed

• Filed: 2026-07-10 07:30 ET
• Event risk confirmed: LOW
• VIX gate: CLEAR (15.89, LOW regime — best-performing bucket)
• Macro hard-skip gate: CLEAR (no Tier-1/Tier-2 data; USDA WASDE is agriculture-only)
• Scanner watchlist: PRESENT — semiconductors flagged active via SK Hynix debut, no instrument-specific catalyst for NVDA/AMD
• DELTA: Phase 3 active, LONG-only — setup rated unlikely (5–10% probability), standard sizing if it fires, NVDA→ASML has first claim over AMD→TSM

Desk Reflection — House Delta — 2026-07-10

Desk plan reference: desks/house-delta/plans/2026-07-10-plan.md EOD briefing: projects/dadbrain/Analysis/briefings/2026-07-10-eod.md Roster reflected: DELTA (PM, sole active member)

Desk-Level Summary

One member, one trade, one candid verdict — no averaging needed on a desk of one. The session played out almost exactly as the desk plan described: LOW event risk, LOW VIX, quiet post-rally digestion (QQQ 0.60× ATR, SPY 0.78× ATR, flat into the close). The one live wild card flagged in the plan — the SK Hynix debut creating a semiconductor halo strong enough to fire a signal despite no NVDA-specific catalyst — is exactly what happened. The plan rated this a 5–10% probability outcome and it landed.

DELTA — NVDA→ASML, AMD→TSM

What happened: One trade — ASML LONG, 12 shares, entered 1791.07, exited 1797.71 at HARD_FLAT 15:30 ET. Net P&L: +$79.74, DELTA's first winning paper trade (cumulative now -$350.78 over 4 trades, 25.0% WR).

NVDA→ASML fired — the primary pair (combined PF 1.85), exactly as the desk plan said would take first claim if a signal materialized. The trigger was the SK Hynix halo: SKHY closed up roughly 13% on debut, the largest foreign IPO in US history, and that thematic enthusiasm was evidently enough to push NVDA through the +1.5% first-hour threshold even with no name-specific catalyst. QQQ +0.33%, SPY +0.38–0.45% for the session — the AI trade got its third consecutive up day, just a subdued one compared to Thursday's decisive rally.

Self-critique — the trade was taken against the plan's own explicit caution, again. The desk plan called this a 5–10%-probability, no-trade-expected session, and carried forward an explicit behavioral flag from the prior two sessions (07-08, 07-09) that this exact pattern — trading a signal that fires against a plan already rated unlikely — needed a genuine re-check against invalidation conditions before acting, not a rationalization after the fact. None of the five invalidation conditions in the plan actually triggered (no gap-down open, VIX held at 15.03, no outsized pre-market gap), so this trade cannot be waved through on "the plan said pass and DELTA traded anyway" alone the way 07-08 and 07-09 could — the signal was, by the plan's own written conditions, a legitimate entry. What the plan did not anticipate, and should have, is that a first-hour move sourced from one-time IPO novelty rather than instrument-specific fundamentals is a different kind of signal than the mechanism the strategy was actually backtested on — the lead-lag thesis is about supply-chain information propagation, not about a halo effect from an unrelated listing. That is a gap in the plan's invalidation list, not a discipline failure this time, and it's worth being precise about the difference: 07-08 and 07-09 were "I had a clean no-trade read and overrode it"; today is "my no-trade read correctly flagged the one live wild card, the wild card fired, and I traded a signal my own written conditions did not tell me to reject." Those are not the same failure mode, and conflating them going forward would blur a real distinction.

Mechanically clean: standard sizing, 10:00 ET entry, hard flat at 15:30 ET, stop never threatened. No execution issues.

Desk-level verdict: GO call was correct, gates were correctly read, and the trade that resulted was a legitimate signal by the plan's own written rules — but the plan's invalidation list has a real gap around thematic/novelty-driven leader moves (SKHY-style halo effects) versus fundamentals-driven ones, and that gap should be closed before the next session where a debut, rebalance, or other one-time narrative event is live in the semiconductor complex.

Desk-Level Learnings (flag for backtest-agent/learnings.md, desk-blind)

• Thematic-novelty invalidation gap. A first-hour leader move driven by a one-time narrative catalyst (IPO debut halo, index rebalance) can clear the lead-lag signal threshold without the underlying supply-chain information-propagation mechanism the strategy is actually built on. The resulting laggard divergence may close by reversion/convergence rather than by the laggard catching up to real information, which is a structurally different — and untested — dynamic from what the backtest measured. Recommend adding an explicit invalidation condition for future plans: when the leader's first-hour move is attributable to a one-time, non-instrument-specific narrative event rather than the instrument's own fundamentals, treat the divergence as lower reliability.
• First winning trade, still 25.0% WR against a 44.5–46.4% backtest baseline. Not yet meaningful at N=4, but worth tracking whether the sample converges toward the backtest win rate as more trades log.

Reflection Filed

• Filed: 2026-07-10 18:30 ET
• Next session event risk (from EOD briefing, Monday July 13 calendar): LOW — no Tier-1 data, but the week ahead is dense: CPI (Tue Jul 14), PPI + Fed Chair Warsh testimony (Wed Jul 15), Retail Sales (Thu Jul 16). DELTA is hard-skipped Tuesday through Thursday next week; next tradeable session is Friday July 17.
• Desk cumulative: 4 trades, 25.0% WR, -$350.78 net paper P&L.