House Vesper — 2026-07-22

desk: house-vesper date: 2026-07-22 forecasts: vesper: p_trade: 1.0 direction: long conviction: high

Desk Plan — House Vesper — 2026-07-22

Briefing reference: /Users/dadbot/Desktop/ClaudeBod/projects/dadbrain/Analysis/briefings/2026-07-22-pm.md PM: VESPER

Shared Market Read

• Event risk today: MEDIUM — no macro releases of Tier-1 significance, but GOOGL and TSLA report after close, making this the week's heaviest single-day earnings catalyst. FOMC blackout continues through July 29. The macro calendar is clean (MBA Mortgage Applications, State JOLTS — both LOW impact).
• Session character expected: Consolidation / pre-earnings positioning. The session is a holding pattern for GOOGL/TSLA after the close. Pre-market futures are modestly negative (S&P 500 -0.14%, Nasdaq -0.53%). The chip bounce from yesterday stalls in Asia (Nikkei -0.25%). Oil is spiking (Brent +2.78% to $93.54) on Hormuz disruption fears, creating a sector-level cross-current in energy. Expect below-ATR ranges for both SPY and QQQ — low conviction intraday.
• VIX regime: 17.44 (LOW, < 18) — VIX ticked up slightly from yesterday's 17.05 but remains in LOW territory. The fear premium from the prior week's chip rout has fully unwound. LOW regime is VESPER's operational baseline; no regime-based concern for the overnight hold.
• Key levels: SPY prior close 748.15; prior-day range 744.25–749.03; SMA20 744.94 (above — reclaimed after 2 days below); ATR(14) 7.26. For VESPER, the only level that matters is the 16:00 auction-inclusive close, which sets tonight's entry anchor.
• Macro backdrop: Mega-cap earnings positioning dominates. The two-tier structure persists (SPY above SMA20, QQQ below for 6th day). The oil spike adds a sector rotation cross-current. The FOMC blackout (July 18–30) removes Fed-speak as a variable. Today's session determines whether the chip bounce recovery has legs or was a dead-cat bounce ahead of GOOGL/TSLA.

VESPER — SPY, unconditional overnight hold

Position status from ledger: Long SPY at $748.15 (16:00 ET MOC anchor, 2026-07-21). 30 shares approximately (90% of ~$24,900 equity at entry). Exit at 09:31 ET today.

Overnight P&L estimate: The position entered at $748.15 (yesterday's close). The exit anchor at 09:31 will be the first post-open print. Against the prior close of $748.15, the overnight drift is unrealized until the 09:31 bar prints. No estimate is meaningful at this point — the backtest daily net bps σ is 88.6. Single-session noise is expected.

Tonight's entry: Unconditional MOC buy at 16:00 ET. Standard sizing: 90% of projected equity at the 16:00 auction-inclusive close. No session-level conditions can override this — the strategy is unconditional by design. Any conditioning must first beat the unconditional baseline in a Research Bench study without destroying the ~250 obs/yr validation velocity.

Session character is irrelevant to VESPER. The desk is never in the market during the day. The pre-earnings consolidation, the oil spike, the GOOGL/TSLA positioning — these are intraday narratives that do not affect the overnight hold mechanism. The only thing that matters is the 16:00 anchor price. The overnight premium is a structural edge paid by the session boundary, not a directional bet on the day's tape.

Event risk and the overnight hold: VESPER holds through ALL overnight sessions — including GOOGL/TSLA earnings nights. The backtest's 2023–24 holdout (5.57 bps/day net) includes earnings nights, FOMC nights, CPI nights, and gap-risk nights. Event risk is not a filter; it is the premium I am paid for holding the hours nobody else wants. The -26.52% COVID drawdown (2020) is the validated in-sample worst case — tonight's GOOGL/TSLA earnings are noise on that scale.

Mechanism watch status: The decisive edge monitor (trailing-250/500 SPY overnight executable-anchor gross mean) is well above the 0.655 bps/day floor — the premium is intact. As of the last cache edge: trailing-250 +8.277 bps, trailing-500 +4.995 bps. The execution-cost Gaussian SPRT (μ0=1.0/μ1=2.0) shows CONTINUE — but this is a simulation artifact (the ledger prices at anchors, so realized shortfall ≡ 0). Decision-relevant data begins when broker fills replace simulated anchors.

Pre-registered kill distances: No trigger even close. The mechanism watch's trailing-250/500 floor (0.655 bps/day) is the binding death check, and it is not threatened.

Plan Filed

• Filed: 2026-07-22 07:15 ET
• Frontmatter forecasts complete for every active member: yes