Wednesday plan: hold growth leaders while testing oil-driven macro stress
Keep the live book unchanged for Wednesday. Technology leadership remains constructive enough to retain the growth concentration, while Tuesday's underperformance, weak financials, and escalating oil-supply risk argue against adding exposure before the open confirms that AI leadership is holding.
Accountability — MIXED: Tuesday's plan was directionally right to reduce JPM and initiate META, as JPM fell 1.43% while META fell 0.53%; however, the prior book declined 1.2% versus SPY's 0.55% decline, so the growth concentration did not protect relative performance.
Since Last Session
Tuesday was negative for every holding. JPM fell 1.43%, MSFT fell 1.15%, AMZN fell 0.6%, NVDA fell 2.01%, and META fell 0.53%. The prior book declined 1.2% versus SPY's 0.55% decline, trailing the benchmark by 0.65 percentage points.
Regime Check
Regime: R1 — constructive but fragile risk-on conditions, with technology leadership intact but breadth and macro confirmation incomplete.
Supporting: QQQ declined only 0.08% Tuesday and retains positive 20-day momentum of 0.14% and positive 60-day momentum of 0.68%. XLK gained 0.32%, with positive 20-day momentum of 1.13% and positive 60-day momentum of 2.45%. VIX is 15.72. META, NVDA, and MSFT retain positive 60-day momentum, while reported AI infrastructure deals reinforce the link between compute spending and potential earnings power.
Contradicting: SPY fell 0.55% and has negative 20-day momentum of 0.36%. IWM has negative 20-day momentum of 1.32% and negative 60-day momentum of 0.67%, while XLF fell 1.38% and has negative 20-day momentum of 0.93%. Brent nearing $100 amid U.S.-Iran escalation raises supply, inflation, and rate-path risk; TLT's negative 60-day momentum of 1.96% offers no confirmation from duration.
Plan for Wednesday
Keep all five positions and weights unchanged. Tuesday supplied no company-specific thesis break, and the recent META initiation should be allowed to develop rather than trigger another round of turnover after one weak session. The open must confirm that QQQ and XLK continue to show relative resilience, that the oil shock is not producing a broader disorderly risk reduction, and that NVDA and META retain leadership within growth. If those conditions fail, subsequent caution should be expressed by reducing the weakest growth thesis rather than adding cash or defensive holdings.
Positioning
JPM — thesis state: WEAKENED. Action: Hold at 20%; do not rebuild the trimmed weight. Evidence: JPM fell 1.43%, XLF fell 1.38%, and both have negative 20-day momentum, although their 60-day momentum remains positive.
MSFT — thesis state: INTACT. Action: Hold at 29%. Evidence: The 1.15% Tuesday decline and slightly negative 20-day momentum are cautionary, but 60-day momentum remains strongly positive at 13.53%, with no company-specific invalidation in the packet.
AMZN — thesis state: INTACT. Action: Hold at 19%. Evidence: Tuesday's 0.6% decline and negative 20-day momentum are offset by positive 60-day momentum and the reported AI infrastructure agreement involving Qualcomm and Amazon.
NVDA — thesis state: INTACT. Action: Hold at 20%; require renewed leadership before considering an increase. Evidence: Tuesday's 2.01% decline was the book's weakest move, but 20-day and 60-day momentum remain positive and AI infrastructure news remains supportive.
META — thesis state: STRENGTHENED. Action: Hold the new 12% position at measured conviction while awaiting further execution evidence. Evidence: META had the smallest Tuesday decline among the holdings and retains the strongest positive 20-day momentum in the candidate universe, while its personal-AI-agent push expands both opportunity and execution risk.
MSFT, AMZN, NVDA, and META represent 80% of the book, keeping the portfolio deliberately concentrated in durable AI, cloud, compute, and digital-platform growth rather than broad diversification.
Institutional Signals
Institutional filings remain slow-moving context. Pershing Square reported substantial positions in AMZN and MSFT, supporting the durability of those franchises as concentrated holdings. Bridgewater reported exposure to NVDA and AVGO alongside broad index positions, signaling participation in AI infrastructure without relying exclusively on it. Berkshire's concentrated portfolio reinforces the case for owning a limited number of high-conviction businesses, but its disclosed top holdings do not directly validate this book. Scion's older filing included NVDA, but its age and concentration make it weak evidence for Wednesday's decision.
What Could Break It
- A broad loss of relative resilience by QQQ and XLK while SPY, IWM, and financials weaken further would contradict the R1 classification.
- Further oil-supply escalation that materially intensifies inflation and rate pressure would threaten long-duration growth multiples.
- Public hyperscaler capex cuts, weaker AI monetization, or deteriorating infrastructure economics would weaken NVDA, MSFT, AMZN, and META together.
- Rising provisions, weaker net-interest-income guidance, or materially deteriorating credit conditions would break the JPM thesis.
- Privacy or safety constraints that impair adoption of META's personal AI products would reduce confidence in the newest position.
- The drawdown remains 1.87% with the protocol at normal; a transition out of that stage would require reducing risk according to the portfolio protocol.