House Slack — 2026-08-10

desk: house-slack date: 2026-08-10 forecasts: slack: p_trade: 0.02 direction: none conviction: low

Desk Plan — House Slack — 2026-08-10

Briefing reference: /Users/dadbot/Desktop/ClaudeBod/projects/dadbrain/Analysis/briefings/2026-08-10-pm.md PM: SLACK

Shared Market Read

• Event risk today: HIGH (Bucket A convention) — no Tier-1 today, but CPI is Wed Aug 12, 2 sessions out, the week's macro fork and the only remaining hawkish trigger after Friday's −23K NFP parked the September-hike debate. Second live layer: Hormuz/oil — crude ~+1.5% on Trump's "economic pressure" pivot and Iran's hardened terms; any de-escalation headline is the day's fastest reversal trigger. No tier direction defaulted this morning (Blindspot 7 — the deal-hope/Dow leg has been fragile twice; oil's up-move keeps it the at-risk leg). Firm-wide: GO (no Tier-1 today, no economic data scheduled, VIX LOW).
• Session character expected: Quiet, positioning-driven Monday — no same-day catalyst, bill auctions only; modest ranges, structure may not form until a Hormuz headline or afternoon CPI-eve positioning. Base case per the EOD calibration directive: modest follow-through/range drift off Friday's record close, not churn and not a clean trend day. RKLB/SPG/ACHR earnings after the close set tomorrow's gap map.
• VIX regime: 15.15 — LOW (<18). Structurally irrelevant to SLACK's gate: the backtest validates profitability across all VIX regimes; the binding constraint is the 5-day return threshold, not the vol level. The only cohort effect VIX has on this desk is none.
• Key levels (IWM): The only binding level is the ±4.97% 5-day close-to-close return threshold (85th-percentile dev calibration, held fixed) — and today's measurement is already in the books (see below). Narrative context from the stack's own near-miss telemetry: $291.6 — Aug 3 close (today's window base); ~$294.2 — derived Aug 6 close; $300 — the pre-rout range ceiling, now the near-term overhead (IWM has reclaimed the $292 floor and worked up into the $294–300 zone); $288.6 — Jul 28/31 closes, the old range floor.
• Macro backdrop: Post-NFP relief regime — record S&P close Friday (+0.62%), soft print collapsed hike odds, bonds rallied, VIX LOW; leadership rotated to rate/credit-sensitive and space; the persistent weak legs are software/adtech and memory; oil is the live cross-asset variable into CPI Wednesday.

SLACK — IWM (5-day fixed hold swing)

Position state: Flat. position_ledger.json: {"open_position": null}. No open position, no scheduled exit — the state machine can only open a position or log a no-trade row today. SPRT CONTINUE (n=0, W0/L0, LLR +0.000). 30 consecutive no-signal sessions logged; zero Phase 3 trades.

Today's decision input is already in the books — and it is categorically a no-signal. The locked module shifts its flag one session forward (capacity_ratio_swing.py): sig.loc[2026-08-10] is computed from the run through Aug 7's close — window {Aug 3, 4, 5, 6, 7}, base = Aug 3's close (≈ $291.6). Friday's close is complete; nothing in today's tape can change the measurement. Reconstructing from the stack's own near-miss telemetry: the window's base rolled forward from Jul 31 (≈ $288.57, the +1.95% reading's base) to Aug 3 (≈ $291.6) — a ~+1.05% higher base that mechanically compresses the reading — and IWM closed the week near its highs (~$294–299, small-cap laggard on a mega-cap record day). Bounded estimate: the run through Aug 7 close lands ≈ +1.0% to +2.0% (central ≈ +1.5%), roughly 3–4pp short of ±4.97%. A flag would require Friday's close ≥ ~$306 or ≤ ~$277 against the $291.6 base — off the table. The script's 1:30 PM run remains the arbiter of the exact print, but the outcome is a foregone no_trade:no_signal (blocked=run_magnitude).

The reachable-window episode of this stretch is over — peak-and-recede, on record. The run through Aug 4 close logged +2.87% (08-05), through Aug 5 close +3.92% (08-06 — the closest measurement of all of Phase 3, 1.05pp short, mechanically inflated by the Jul 29 rout-low base), through Aug 6 close +1.95% (08-07), and now ~+1–2% through Aug 7 close. The rout-base inflation has rolled out of the window; the flag never threatened threshold and is now receding. This matches TEMPER's 2026-07-24 retrospective exactly (range-bound-to-breaking IWM = structurally unreachable threshold = expected behavior, not edge degradation; SPRT n=0 is correct behavior, not a false flag).

Direction (conditional on a flag — moot today): short. The run through Aug 7 close is positive, and the mechanism is contrarian by construction. The downside flag is impossible today for the same reason the upside flag is: the measurement is ±1–2% and already fixed. Frontmatter direction: none is the honest field for a categorically no-signal day — there is no directional lean to forecast when P(flag) ≈ 0.

Sizing vs. event risk (CPI Wednesday as a horizon event): N/A today — flat, nothing to size. Standard 15% fixed notional (~$3,750 on $25K), single position, no stacking applies unchanged on any future signal; the COVID-tail-calibrated sizing (−19.66% worst trade ≈ −3% of account equity) is what covers multi-day event risk, and a hold spanning CPI needs no adjustment — the fixed 5-day hold has no event-skip by design, and that is inherent to the validated mechanism, not a revisitable choice. Forward awareness only: a flag at Tue's open (run through Aug 10 close) would hold through CPI Wed Aug 12 as day 2 of the hold; a flag at Wed's open (run through Aug 11 close) would make CPI day 1. Moot in the base case — P(any flag this week) is low, because reaching ±4.97% now requires a ~$14–15 IWM move in five days from the ~$294–300 zone, and the two-tier tape has kept every measurement this stretch 2–4pp short.

Session-character fit: Structurally N/A — SLACK has no intraday view; the gate is a completed multi-day measurement decided by Friday's close, not today's tape. The quiet Monday, the Hormuz headline risk, and the AMC earnings set (RKLB/SPG/ACHR) cannot change the entry decision; they only shape the forward path of a hold that does not exist in the base case.

What I'm watching (early-warning, not entry signals):

1. The script's logged near-miss at today's run (through Aug 7 close) — expect ≈ +1.0–2.0%, 3–4pp short. A print in that band closes out the reachable-window episode on record; anything materially outside it would be a data surprise worth flagging (the briefing quotes no IWM prices; local caches end 2024-12-31, so the script is the only arbiter).
2. IWM vs $300 (pre-rout ceiling) / $292 (reclaimed floor) — the structure feeding next week's windows. A decisive $300 break with small-cap catch-up is the only fast path to a flag in the coming sessions; continued lag keeps every measurement 3+pp short.
3. CPI Wednesday Aug 12 as a horizon event — if a flag fires Mon/Tue, the hold spans the print (sized for it, no adjustment); the print itself also sets the small-cap path for the rest of the week (hot core → small-cap risk-off; dovish-in-line → small-cap catch-up, the fast path to a flag).

Setup evaluation: No trade expected — categorically. The day's decision input is fixed by Friday's close at ~+1–2% vs ±4.97%; the flag cannot fire today. This is the structural-impossibility baseline, not a hedged maybe — p_trade 0.02 (the desk's honest allowance for pipeline noise on a no-signal day, matching the 08-04 convention at a comparable 3.9pp gap), direction none, conviction low. The mechanism is doing exactly what it was validated to do: waiting for slack water. Today is not the day the tide turns.

Plan Filed

• Filed: 2026-08-10 10:40 ET (post-open filing — decision input was fixed by Friday's close; first House Slack plan since 08-05, the 08-06/08-07 pipeline gap being on record in the 08-07 reflection)
• Frontmatter forecasts complete for every active member: yes
• Active members: slack (IWM)
• Position state: flat ({"open_position": null}); SPRT CONTINUE n=0, LLR +0.000
• Firm-wide status: GO (event risk HIGH per Bucket A — CPI Wed Aug 12 in 2 sessions; Hormuz/oil headline risk live; no Tier-1 today)
• Expected action (base case, ~98%): no trade — paper_trade.py logs a no_trade:no_signal row (blocked=run_magnitude, near-miss telemetry = run through Aug 7 close, est. ≈ +1.0–2.0% vs ±4.97%). No contingency: the entry-decision input is already in the books and cannot reach threshold today.

No trades taken.