House Vesper — 2026-07-23

desk: house-vesper date: 2026-07-23 forecasts: vesper: p_trade: 1.0 direction: long conviction: high

Desk Plan — House Vesper — 2026-07-23

Briefing reference: /Users/dadbot/Desktop/ClaudeBod/projects/dadbrain/Analysis/briefings/2026-07-23-pm.md PM: VESPER

Shared Market Read

• Event risk today: HIGH — ECB rate decision at 7:45 AM ET, post-GOOGL/TSLA earnings digestion, Intel after close. The briefing rates this HIGH. None of which matters to VESPER — the overnight hold is unconditional.
• Session character expected: Mixed / news-driven — gap-down bias on tech (GOOGL -3.74%, TSLA -5.56%), two-tier divergence (SPY above SMA20, QQQ below for 7th day). ECB at 7:45 sets cross-asset tone. Oil continues to rally (Brent $95.17, +1.17%). The intraday soup is rich but irrelevant to my desk.
• VIX regime: 17.57 (LOW <18). VESPER has no VIX filter — holds through all regimes. The backtest's 2023–24 holdout (5.57 bps/day net) covered every VIX regime.
• Key levels: SPY prior close 747.49, pre-market 745.96 (-0.19%), SMA20 745.63, ATR(14) 7.02. The only level that matters for VESPER: tonight's 16:00 auction-inclusive close, which sets the entry anchor.
• Macro backdrop: Two-tier market persists — SPY resilient above SMA20, QQQ in multi-week correction. Oil-driven stagflation undertone. GOOGL/TSLA post-earnings gap-down sets the tech tone. ECB decision at 7:45. VESPER holds through all of it — the overnight premium is paid for bearing these gaps.

VESPER — SPY (unconditional overnight hold)

p_trade: 1.0 | direction: long | conviction: high

Summary: The strategy is unconditional — every session, every night, no exceptions. The only content worth stating is the position status, kill/watch distances, and the realized-cost check.

Position status (from ledger): Long SPY at $748.14 — entered 2026-07-22 at 16:00 ET MOC anchor. 29.96 shares, $22,416.48 notional (90% of $24,907.20 equity at entry). Held overnight, awaiting 09:31 ET exit.

Exit today (09:31 ET): Marketable-limit sell at 09:30:00 ET, flat by 09:31. The entry was priced at $748.14; the exit will be the 09:31 bar open. Single-session drift noise is expected — the backtest daily net bps σ is 88.6. This cycle's P&L is a single observation, not a signal.

Entry tonight (16:00 ET): Unconditional MOC buy SPY at 16:00 ET. Standard sizing: 90% of projected equity. The overnight premium does not care about today's intraday tape — the clientele segmentation edge is about the session boundary, not the GOOGL/TSLA reaction or ECB decision.

High event risk tonight: ECB at 7:45, GOOGL/TSLA after-hours repricing, and Intel after close create a high-event-risk session. VESPER holds through all of them — the backtest's 2023–24 holdout (5.57 bps/day net) includes every earnings night, FOMC night, and gap-risk night in the sample. The COVID -26.52% drawdown is the validated in-sample worst. Tonight's events are noise on that scale.

Mechanism watch status: OK — trailing-250 +8.277 bps, trailing-500 +4.995 bps (as of last cache edge). Both well above the 0.655 bps/day kill floor. No action warranted.

Execution-cost SPRT: CONTINUE — n=5, LLR +1.875. In simulation the ledger prices at the anchors, so shortfall ≡ 0. The monitor is decision-relevant only against real broker fills. The budgeted round-trip cost of ≤ 1.0 bps/day remains the binding operational constraint.

What I'm watching: (1) 09:31 exit fill quality — normal conditions on a modestly negative pre-market morning; (2) the 16:00 close anchor price — the premium earned tonight is (tomorrow's 09:31 open) minus (today's 16:00 close); (3) mechanism watch update after tonight's cycle closes — the trailing means will tick one session forward, expected to stay well above the floor.

Invalidation: VESPER does not invalidate — the strategy is unconditional. The only invalidation path is a Mechanism Watch trigger (trailing-250 AND trailing-500 < 0.655 bps/day, confirmed by two-window P12 protocol), which would initiate a TEMPER review. No such conditions are present.

Plan Filed

• Filed: 2026-07-23 07:15 ET
• Frontmatter forecasts complete for every active member: yes