House Delta — 2026-09-04

Desk Plan — House Delta — 2026-09-04

Briefing reference: /Users/dadbot/Desktop/ClaudeBod/projects/dadbrain/Analysis/briefings/2026-09-04-pm.md PM: DELTA

Shared Market Read

• Event risk today: HIGH (Tier-1) — the August NFP (Employment Situation) prints at 8:30 AM ET. Consensus clusters ~+55K (+53–58K; MUFG +70K) vs. prior −23K (July, a hard miss), U/R 4.1% cons, AHE +0.3% m/m. The print is the read-ahead into the Sept 15–16 FOMC, which the market prices at a genuine ~50/50 hike/hold coin flip after Warsh (Jackson Hole hawkish) vs. Waller (hold-lean) swung it both ways this week. Firm-wide status: NO-GO. VIX 14.18 (LOW). After-close single-name reporters gap today (LULU −~18% premkt on a third FY guidance cut, CIEN beat, IOT) — layered single-name risk on top of the macro. Labor Day Mon Sep 7 closed.
• Session character expected: direction unknown until 8:30, then trending. Not a contained pre-positioning day — the briefing is explicit (EOD directive #2): a data surprise plus a live rate signal can break the tape at 1.6× ATR or more. The binary: soft NFP → locks the FOMC hold, extends the Waller relief leg (rate-sensitive growth leads); hot NFP / hot wages → re-ignites the Warsh hike leg, reverses Thursday's tape. Structure will not form until after the print.
• VIX regime: LOW (14.18) — DELTA's best-performing regime (dev PF 2.59 in VIX < 20), and well under the 30 hard-skip. But regime being favorable is moot today: the hard skip closes the session before any vol question is even asked.
• Key levels: DELTA's only load-bearing number is the leader's first-hour return measured off the 09:30 cash open, checked at 09:55. That window straddles and follows the 8:30 print — a 09:30–09:55 leader move is by construction post-NFP macro repricing, not an idiosyncratic supply-chain cascade. Prior-close refs (NVDA ≈ 225–230, ASML ~1625–1800, TSM ~418–435, AMD ~426–465) are stale and irrelevant to a no-trade day.
• Macro backdrop: A binary Fed/rates NFP resolving a ~50/50 Sept-FOMC coin flip, layered over a live US–Iran/Hormuz oil clock entering a holiday weekend — the single most directionally-binary macro day of the week.

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

Call: No trade expected — NFP is a hard-skip day for DELTA, and the macro shape is structurally incompatible with a supply-chain lead-lag signal. p_trade 0.0, direction none, conviction high. Flat by pre-registered rule; this is a deterministic stand-down, not a hedged maybe.

• Primary reason — the hard skip. NFP is in DELTA's hard-skip set (FOMC/CPI/NFP/PCE) from strategy_v2.md, enforced via paper_trade.py off data/event_calendar/. Today's 8:30 AM NFP closes the session before any signal logic runs. No leader impulse is evaluated; there is no LONG to consider and no branch to scan. This is the correct, designed behavior — the 2026-08-26 PCE skip is the standing precedent (p_trade 0.0, Brier 0.0).
• Secondary reason — even absent the hard skip, the mechanism and the day are structurally incompatible. DELTA's edge is an idiosyncratic first-hour information cascade: a design-company (NVDA/AMD) move repricing its supplier (ASML/TSM) later in-session as supply-chain consequences propagate. A binary macro print does the opposite of that: it reprices both leader and laggard simultaneously and in one direction (systematic beta). The 09:30–09:55 measurement window lands entirely after the 8:30 print, so any first-hour leader impulse on an NFP day is macro-repricing by construction — the wrong shape for the lead-lag alpha the mechanism was cleared on. This mirrors the same structural stand-down reasoning DELTA applied through the oil-shock mornings (08-31 → 09-02): a broad one-way macro move is not a supply-chain cascade.
• Sizing: n/a — no position, standard 0.75% irrelevant on a closed session. Zero exposure at the open, at the print, and into the close.
• Session-character fit: A post-print trending tape (soft → risk-on extension, hot → hike-reprice) is a momentum day, not a pairs-convergence day. DELTA does not call SPY/QQQ direction and takes no view on which way the NFP resolves — direction-refusal applies to DELTA as to every desk (the briefing's discipline stands 9-for-9).
• What this is not: Not a missed signal. A no-trade hard-skip day is a valid, expected outcome — the same shape as the PCE skip and the pre-print sessions — and forecasting it at p_trade 0.0 with high conviction is the honestly calibrated call. The dry-up regime-monitoring flag DELTA raised on 09-03 (up-branch structurally near-absent ~3 weeks against an easing-vol backdrop) is not advanced by today: an NFP hard skip contributes no signal evidence either way, so the flag stays queued and unformalized.
• Book state carries forward untouched: Cumulative −$442.33 across 10 trades (SPRT CONTINUE, n=10, LLR −0.159). AMD→TSM sub-book n=8, WR 25%, PF 0.191 stays queued for TEMPER's calibration conversation. NVDA→ASML clean cell 2W/0L remains the quality cell for the next tradeable session. Open event-day diagnostic run (skip disabled) remains moderate priority — today's hard skip is exactly the data point that run exists to interrogate, but DELTA does not run it unilaterally.
• Watch items (informational only — no action today): the NFP resolution as a read-ahead into next week's FOMC and CPI Sep 11; whether any leader makes a post-print first-hour move ≥ 1.5% (would be macro repricing, but tells DELTA how far off the up-branch the tape is); and whether Thursday's relief-leg leadership in the semis survives the print into the next tradeable window.

Plan Filed

• Filed: 2026-09-04 04:33 PDT (pre-open)
• Frontmatter forecasts complete for every active member: yes (delta — p_trade 0.0, direction none, conviction high)

No trades taken.

Reflection — House Delta (DELTA) — Weekly — 2026-09-04

Period Performance (since last reflection 2026-08-28)

Single-member desk. DELTA took zero trades across all five sessions since the 08-28 reflection (08-31 → 09-04, all flat). Net P&L $0.00 for the period; cumulative remains -$442.33 across 10 paper trades. SPRT sequential monitor CONTINUE (n=10, LLR -0.159). AMD→TSM sub-book unchanged: n=8, WR 25.0%, PF 0.191 — not triggered, and gained no new LONG evidence this period (AMD never cleared +1.5% up).

The period's five sessions broke down as: 08-31 no_signal (AMD +1.20% — 0.30pp short of the +1.5% floor, the week's first up-approach), 09-01 no_signal (AMD +0.22% / NVDA +0.10% — leaders inside floors), 09-02 no_signal (NVDA +1.23% — 0.27pp short of the up-floor on the primary NVDA→ASML cell, the closest that cell has come in over a month), 09-03 no_signal (NVDA +1.02% up, AMD -1.33% down-side grazing the short floor from 0.17pp but structurally untradeable under LONG-only), 09-04 NFP hard-skip (today: hot print +162K vs +55K cons, resolved contained two-tier, VIX 14.53 LOW). Net zero, Brier mean ~0.02 for the window — no forecast busts.

What Happened (load-bearing)

The load-bearing gate question — regime or calibration? Regime for the LONG gate this period, but with a materially warming up-branch that I am now actively watching. Across 08-31 → 09-04 the tradeable up-direction (leader first-hour ≥ +1.5% LONG) never cleared — all four tradeable sessions were sub-threshold no_signal and today (09-04) was the NFP hard-skip. Yet the near-misses are no longer the distant, mostly-down texture of the prior two reflection periods (08-14 → 08-28, where the up-floor was touched near once and the disqualifications were almost all the down-side). This week the up-side came close, repeatedly, and on the primary edge-carrying cell: NVDA +1.23% (09-02, 0.27pp short of floor), NVDA +1.02% (09-03), AMD +1.20% (08-31, 0.30pp short) — three up-approaches inside ~0.5pp across four tradeable sessions. The one down-side disqualifier (AMD -1.33% on 09-03) stopped 0.17pp short of even the short floor.

Why is this still regime and not yet calibration? Because (a) no LONG fired — the floor was approached, never cleared; (b) the three near-misses span two different cells at slightly different distances (NVDA +1.23/+1.02, AMD +1.20), not repeated grazing of a single threshold by a single cell; and (c) the binding cause is the same LOW-VIX (14.2–15.4 all week) contained two-tier tape that keeps first-hour leader moves inside ~1.2% — realized vol too low to produce a clean ≥1.5% up-impulse. That is the structural-absence signature, so no fix is warranted and no unilateral tweak was made. But the warming texture is the honest watch item of this period: if the post-Labor-Day / post-NFP tape next week produces another NVDA up-graze inside ~0.3pp (a third close miss since 08-25), that shifts from "structurally starved" toward a genuine calibration question — is the +1.5% first-hour floor right for the current low-vol regime? — which belongs to TEMPER, never to me. Pre-registered threshold; if it becomes calibration evidence I raise it, I do not move the gate.

What Worked

• The 09-04 NFP hard-skip was a clean, correct no-trade. I forecast p_trade 0.0 / conviction high and the engine's Tier-1 filter closed the session before any entry logic ran. Brier 0.0. This is the discipline operating exactly as pre-registered — no position at risk into a hot +107K-surprise print that re-opened the September-hike door, into the pre-weekend tape, or into the Labor Day holiday. The hard-skip is the week's cleanest no-trade.
• Forecast calibration was honest and correctly low across the window. Brier for the five sessions: 0.0225 / 0.0225 / 0.0144 / 0.04 / 0.0, mean ~0.02. The up-branch was genuinely starved-to-warming and I priced it low every session; no forecast busted — a clean window after 08-27's 0.36 blemish last period.
• The down-branch was correctly recognized as structurally untradeable, not chased. On 09-03 AMD reached -1.33% (0.17pp from the ±1.5% short floor) — under LONG-only that is the disabled-SHORT side, no trade by rule. A leader heading toward the short floor is not a missed LONG; it is the branch this book is closed out of by design. (AMD did not actually clear even the short floor this week — unlike the prior period's outright down-breaks.)
• Event handling was clean the whole window. The only Tier-1 (NFP, 09-04) was hard-skipped; the MEDIUM days resolved as no_signal with no event-contaminated entry. My 15:30 hard flat carried zero overnight exposure into the print, the weekend, and the 3-day holiday (with the live Iran/Hormuz clock riding it).

What Didn't Work

• Nothing traded — the up-direction setup still never fully formed. This is the recurring story, but the shape is changing in a way worth stating plainly: the setup is no longer structurally absent the way it was through mid-August. For four consecutive prior periods the +1.5% up-floor was barely touched; this week it was grazed three times in four sessions and twice on the primary NVDA cell. The binding gate still did not fire — a graze is not a clear — but a desk that has logged $0.00 across ~five straight weeks of sessions needs to be honest that the tape is approaching its trigger, even if this window is technically flat by rule. The risk is not that I missed a signal (I didn't); it is that when the up-branch finally does break out cleanly after this long a starvation, I stay disciplined and take it at standard size rather than over-pricing a would-have-been-close day.
• AMD→TSM remains the standing loss source whenever its up-impulse fires — but it did not fire this period. AMD's 08-31 +1.20% was the one sub-book approach and it stopped sub-threshold; its down-moves (09-02/09-03) are the untradeable side. No new LONG evidence to advance or retire the queued calibration conversation — it stays queued for TEMPER exactly as it was.

Confidence Changes

• NVDA→ASML primary edge: unchanged, 2W/0L live, but the cell is now provably within reach of its trigger. The clean cell's live edge is intact and untouched (no NVDA→ASML LONG fired); this week's NVDA +1.23% (09-02) and +1.02% (09-03) show the primary cell can reach the floor on the current tape. That raises my forward watch priority, not my confidence in the mechanism.
• AMD→TSM sub-book: unchanged at its pre-registered kill-switch boundary. Still n=8, WR 25%, PF 0.191. No new LONG evidence this period → the standing TEMPER calibration conversation stays queued exactly as it was, neither strengthened nor weakened.
• Calibration of the LONG gate: regime read holds for a third consecutive reflection period, but now with a warming qualifier. The up-floor was not cleared and the misses are not single-threshold repetition on one cell — so no calibration conversation is warranted yet. However, the distance-to-floor has collapsed (three approaches inside 0.5pp this week vs. one distant touch in prior periods). I am recording this as a watch-to-escalate: a third NVDA up-graze inside ~0.3pp next week turns this from regime into a calibration question worth taking to TEMPER. No unilateral tweak either way.

New Hypotheses

Junior Agent Queue

None — DELTA is a single-member desk. Hypotheses are flagged for TEMPER's calibration conversation, not queued for junior-agent deployment.

Notes for Next Period

• The load-bearing gate question — regime or calibration? Regime for the LONG gate this period, with a warming qualifier. The up-direction never cleared (four sub-threshold no_signal + one NFP hard-skip); the misses span two cells at different distances and the binding cause is LOW-VIX structural absence — so no fix is needed, wait for the setup. But three up-grazes inside ~0.5pp this week (NVDA +1.23/+1.02, AMD +1.20) means I am now watching whether the tape is about to produce my trigger rather than assuming it stays starved. AMD→TSM sub-book remains a queued calibration question, unchanged.
• Forecasting while the floor is being grazed. Clean-tape days now warrant p_trade ~0.2–0.25 (not the 0.12–0.15 of the starved stretch), but I must forecast floor-clearing probability, not approach probability — a +1.23% day that doesn't clear is still a no_signal and should read as such (Brier 0.0144 on 09-02 was honest). Own lesson from this window.
• Forward drivers into next week: the NFP resolved hot and hawkish (yields ~4.53%, ~5% flagged as the trouble spot; hike odds back toward a coin flip). Next Tier-1 CPI 09-11 (my hard-skip), FOMC 09-16 behind it. The live Iran/Hormuz clock + 3-day Labor Day weekend is a gap risk into Tuesday's open more than an intraday driver. Watch whether the post-NFP hawkish read keeps the two-tier LOW-VIX character (which suppresses my up-branch) or whether the semis — which held leadership Friday (NVDA +1.8%) — finally produce the clean first-hour ≥1.5% up-impulse my LONG has been waiting a month for.

Reflection Filed

• Filed: 2026-09-04 (EOD reflection, desk cadence — Friday trigger)
• Next session event risk (from EOD briefing): next session Tue Sep 8 (Labor Day Mon Sep 7 closed); next Tier-1 CPI 09-11 (my hard-skip). Rates (hot-NFP hike repricing, 10-yr ~4.53%) and the weekend Iran/Hormuz gap clock are the forward character drivers; NVDA up-grazing the +1.5% floor is the setup watch.