House Echo — 2026-09-16

Desk Plan — House Echo — 2026-09-16

Briefing reference: /Users/dadbot/Desktop/ClaudeBod/projects/dadbrain/Analysis/briefings/2026-09-16-pm.md PM: ECHO

Shared Market Read

• Event risk today: HIGH — and this is the binary itself, live, same-day. The FOMC decision + SEP/dot plot + Warsh presser lands TODAY at 2:00 PM ET (two-day meeting concluded). This is the first rate move of 2026 — a 25bp hike to 3.75–4.00% is ~88–92% priced (Polymarket ~88%, CME FedWatch ~92%), so the headline is nearly free information. The tradable variable is the median 2026–27 dot (one-and-done vs cycle-start) and Warsh's presser — he has deliberately refused forward guidance, which amplifies repricing risk. A fresh Reuters poll has economists expecting the hike plus at least one more. Firm-wide status is NO-GO (FOMC); the watchlist marks every directional catalyst N/A and states agents must not trade. Both of my stacks carry FOMC as a hard skip in their cleared gate logic — this is not a discretionary stand-down, it is deterministic: the engine will log no_trade:hard_event:FOMC for ECHO and no_trade:event:FOMC for SURGE regardless of what the tape does (both have honored this event class all cycle: ECHO 07-29, SURGE 07-29).
• Session character expected: event-driven, two-sided, low-conviction morning — direction genuinely undefined until 2:00 PM ET. A contained, low-conviction morning/mid-day (most desks get nothing tradeable before the print), then volatility expansion on the statement at 2:00 and a second expansion on the presser at 2:30. The briefing's discipline is explicit and it is the established record (five straight sessions where the AM either refused or leaned, and both explicit leans were wrong side of a coin-flip): do NOT pre-assign index direction. Two live tails, roughly symmetric: a hawkish dot plot (more hikes) into a 10-yr at a 19-year high and $100+ oil → renewed discount-rate downside; a dovish-relative surprise (hike but dot plot unchanged / presser frames it as insurance) → a squeeze that relieves the beaten-down semi complex. Structure should not be expected to form cleanly until the dot plot is on the tape. This is a repricing-to-conclusion day held in suspension, not a trend day — until it abruptly becomes one.
• VIX regime: 16.97 — LOW (<18), vs prior close 17.20. LOW is the weakest cohort for both my stacks (ECHO's vol_ratio ≥ 1.2 / exhaustion_score ≥ 0.80 need genuine participation; SURGE's PF is 1.36 in VIX < 20 vs 2.18 at 20–30). It is also, today, irrelevant to my execution — the skip is calendar-anchored, not vol-anchored. Log the bucket even on a no-trade day (TEMPER disaggregates against it).
• Key levels: SPY prior close 757.42 (prior range 756.18–760.32; SMA20 765.30 — below; ATR14 5.82). QQQ prior close 704.60 (prior range 703.63–709.52; SMA20 713.34 — below; ATR14 7.70). Pre-market points to a modestly HIGHER, tech-led open (futures Dow +0.1%, S&P +0.2%, NDX +0.4%; semis tentatively bid — INTC +3.35%, NVDA +0.5%) — but direction, not magnitude, is the reliable read at 07:00 ET; the exact gap is not (confirm at cash open). Rates anchor: the 10-yr ~4.99–5.00%, a hair below Monday's 19-year high above 5.00% — the pause that permits the pre-market bid, and the transmission check on whatever the dot plot says (watch hold-or-reject of 5.00%). Energy anchor: Brent ~$107.5 (−~1%) / WTI ~$104 after a surprise US crude build, but Saudi East-West Hormuz bypass still shut, Yanbu suspended, Libya disrupted — supply risk intact; XLE's EIA report (10:30 AM ET) is a live datapoint today, though displaced by the 2:00 PM binary.
• Macro backdrop: a stagflation-tilted, FOMC-gated tape — the Fed about to raise the discount rate on purpose into a 10-yr at a 19-year high and $100+ oil, with a BOJ tightening and a failed crypto-regulatory vote tightening the liquidity backdrop on two continents. The market's question has hardened from "will they spend?" to "can they profit, and at what discount rate?"
• Desk-level consequence: no member has a live path today. Both cleared strategies hard-skip FOMC; the watchlist forbids trading; the briefing explicitly refuses to pre-assign direction. The plan below is therefore a pre-commitment on a structurally empty session — I am forecasting it as such: p_trade 0.0 for both stacks, direction none, conviction high. A 0.0 is not a hedge and not a default; it is the honest read of a deterministic calendar gate. And unlike 09-15 (where ECHO's plan/engine gap cost a driver-fade loss on a live session), today there is no gate-clearing path at all — the skip preempts gate evaluation, so the engine has nothing to offer even if it ignores a plan-level stand-down.

ECHO — dynamic universe (SPY/XLE/XBI baseline), close-leg exhaustion fade at 15:30 ET

No trade expected — binding gate: hard_event:FOMC. FOMC is a hard skip in my cleared v6/v7 gate logic, and today's FOMC is a same-day Tier-1 decision at 2:00 PM ET — squarely inside my 15:30 entry window. The engine evaluates instruments only after the skip check, so no candidate can reach the 15:20 gate stack. p_trade 0.0, direction none, conviction high.

The skip is not an inconvenience I work around — it is load-bearing for my mechanism. My fade is an exhaustion-by-institutional-participation trade on a program that ran all day; on an FOMC decision day the afternoon is binary repricing on the statement and the presser, not a completing institutional program. My window (15:30 entry / 15:58 hard flat) sits exactly on top of the 2:00/2:30 repricing, so any "morning bias" I measured at 10:00 would be a pre-decision artifact, not a trend program completing — fading it into a live policy binary would be a worse version of the 08-10 driver-fade error class (−$75.89), with an event-sized tail. Standing down is the correct expression of my edge, not its suppression. Note also: the one thing that made 09-15 tradeable at all — a completed, gate-clearing sector program in energy — is today displaced: even XLE's EIA print (10:30 AM ET) will be run over by the 2:00 PM Fed, which dominates the rates/discount-rate channel regardless of the inventory print.

For the record, since the plan is also a market read, the Scanner watchlist marks every ECHO directional catalyst N/A today (firm-wide NO-GO):

• XLE — blocked (hard skip). EIA Wednesday is today (10:30 AM ET, prior week −0.391M bbl), which under normal conditions is a live XLE program — but the supply story is intact yet the sector is displaced by the Fed binary. Context, not a setup.
• XBI — blocked (hard skip). No FDA drug decision today (the Sep 16 FDA item is a PDUFA reauthorization meeting, a process event). Sub-book n=2, WR 0.0%, exclusion trigger (n≥15) far off.
• SPY / QQQ — blocked (hard skip). The catalyst is real and enormous but a coin-flip by construction; direction is not defined until 2:00 PM ET. Pre-market modestly higher, tech-led, low-conviction.
• Crypto complex (COIN, MSTR, IBIT, miners) — blocked (hard skip). Idiosyncratic policy shock (Senate Clarity Act cloture vote failed; BTC ~−2% to $75–76K) — headline-driven, not a macro leg, but the FOMC skip covers it.
• Semis (NVDA, AMD, MU, INTC, ASML, SNDK) — blocked (hard skip). The day's most-watched bounce candidate, >20% off highs in bear-market territory, but the reaction function is entirely the dot plot — not an intrinsic catalyst, and not fadeable at 15:30 into a live binary.

Sizing vs. event risk: not applicable — no sizing engages. Had the skip not existed, my answer would still be stand-down: an FOMC decision day needs a volume-cleared, non-grazed gate stack on a still-bid-vs-exhausted discrimination, and a binary-repricing tape does not offer one. The structural fact even a trade would have been safe on: my hard flat is 15:58 ET, so this stack never carries overnight exposure into a following event either — but today's event lands inside my window, which is precisely why the skip exists.

Key levels (context, not triggers): SPY 757.42 prior close / SMA20 765.30 / ATR 5.82; QQQ 704.60 / SMA20 713.34 / ATR 7.70; XLE's 10:30 EIA print as the energy datapoint; the 10-yr's hold-or-reject of 5.00% as the transmission check on the dot plot.

Invalidation: already fully invalidated — the binding gate is satisfied on the calendar before the session begins. The only thing that could produce a trade is an out-of-cycle engine behavior, which the prior FOMC session (07-29) shows does not happen.

SURGE — SPY/QQQ, VWAP pullback resumption (10:00–12:00 ET)

No trade expected — binding gate: event:FOMC. FOMC is in my cleared hard-skip set (FOMC/CPI/NFP/PCE/QQQ mega-cap earnings), and today's FOMC is a same-day Tier-1 decision at 2:00 PM ET. The engine logs no_trade:event:FOMC before any signal evaluation — exactly as it did on 07-29. p_trade 0.0, direction none, conviction high.

Even setting the skip aside, today is close to a worst-case session for my mechanism, and I want that on the record so the no-trade is not read as merely mechanical:

1. This is my documented catalyst-gap loss class, squared. The 07-24 TEMPER learning (0/2 on catalyst-gap sessions) and the 06-17/06-18 TEMPER observations (a low-magnitude, one-directional pre-decision drift lifts ADX and trips an ADX-band — which for a floor-gated momentum agent like me is an amplifier, not a protection) both point the same way: an FOMC day is price-discovery, not clean trend establishment, and the three-phase structure (trend → measured pullback → resumption) is fundamentally invalid there.
2. LOW VIX (16.97) is my weakest cohort — PF 1.36 in VIX < 20 vs 2.18 at 20–30. My thesis lives in elevated vol; this is its floor. And my own 07-17 event-day-skip diagnostic already answered the "is this skip leaving money on the table?" question: FOMC is a consistent dog (dev 3t PF 0.0, holdout 1t PF 0.0, −$529 combined) — the one event type where the blanket skip earns its keep, unlike QQQ_EARN/PCE which it over-covers. Skipping today's FOMC is the correct, evidence-backed call.
3. The chronic binding gate is pullback_resumption. Even on a clean day the three-phase structure has denied me ≥11 trend-qualified sessions since 07-15 (including 08-13 with both legs qualified simultaneously) — on a repricing day it has no chance of forming.
4. Direction is a refused coin-flip. Neither leg has a pre-assignable trending extension into a same-day binary; the dot plot sets whatever direction exists, after 2:00 — five hours past my 12:00 window close.

Sizing vs. event risk: not applicable — no sizing engages ($187.50 / single position / hard flat 14:00 ET standard, invoked only if a full three-phase structure completes, which the skip prevents). Worth naming the asymmetry honestly: because my hard flat is 14:00 ET, even a cleared signal would carry zero exposure into the 2:00 PM decision — but the skip is doing calendar-hygiene and edge-hygiene work here, because FOMC is the one event type my own diagnostic shows is a consistent loser for me, not just an exposure risk.

Key levels (context, not triggers): SPY 757.42 prior close, range 756.18–760.32, ATR 5.82, SMA20 765.30 (below). QQQ 704.60, range 703.63–709.52, ATR 7.70, SMA20 713.34 (below). Both legs' session VWAP through the 10:00–12:00 window would have been the operative level — moot today.

Invalidation: already fully invalidated by the hard skip. If I were to reconstruct the no-trade conditions: a low-conviction, event-gated morning (pre-2:00 PM drift is pre-positioning, not a clean impulse), a 2:00 PM binary repricing that sets the day's direction after my window closes, and — even if a trend-qualified leg emerged before noon — the pullback_resumption precondition that has denied me all period. The dot plot is the single number that decides the following sessions' character; it is not a number that can make me trade today.

Plan Filed

• Filed: 2026-09-16 04:31 PDT
• Frontmatter forecasts complete for every active member: yes
• Firm-wide status: NO-GO (FOMC, same-day Tier-1 decision at 2:00 PM ET) — both members' hard skips are binding; no live path.
• Desk net forecast: ECHO p_trade 0.0 / none / high, SURGE p_trade 0.0 / none / high. Both are the honest read of a deterministic calendar gate, not a hedge — a 0.0 is the correct forecast when the engine is preempted before gate evaluation.
• Carry: neither stack carries the FOMC — ECHO flat 15:58 ET, SURGE flat 14:00 ET, and today neither is even live.

No trades taken.