Friday plan: hold the AI leaders and demand stabilization at the open
Keep the existing four-stock growth book unchanged for Friday. Thursday's AI selloff hurt relative performance, but MSFT, NVDA, and META retain positive medium-term momentum, while the available evidence doesn't show an earnings or capital-spending break.
Prior plan grade: C. Keeping the concentrated AI growth book preserved thesis continuity, but the portfolio fell 1.57% on Thursday versus a 0.42% decline for SPY.
Since Last Session
Thursday was a difficult session for the prior book. MSFT fell 1.35%, AMZN lost 2.25%, NVDA declined 2.94%, and META slipped 0.06%. The weighted portfolio fell 1.57% versus a 0.42% decline for SPY.
Regime Check
Regime: R2. Selective growth leadership remains intact, but Thursday exposed the book's sensitivity to doubts about AI monetization and high rates.
Supporting: SPY retained 0.87% 20-day momentum and 1.48% 60-day momentum. QQQ's corresponding readings were 1.61% and 4.15%. MSFT, NVDA, and META also retained positive momentum over both periods, with META at 13.8% over 60 days. VIX at 15.41 doesn't signal broad market stress.
Contradicting: QQQ fell 1.34% Thursday as AI-linked shares came under pressure following the OpenAI revenue report. TLT still has negative 20-day and 60-day momentum, showing that duration conditions remain difficult. IWM and XLF also have negative momentum over both periods, while stronger energy prices and tariff-linked inflation could keep rate pressure alive.
Plan for Friday
Keep MSFT at 29%, AMZN at 15%, NVDA at 28%, and META at 28%. No rebalance is planned before Friday's open. One weak session, even a painful one, isn't enough to discard earnings-backed growth theses when most of the book still has positive medium-term momentum.
The open must confirm that Thursday's AI selling isn't accelerating into a broader reassessment. Orderly trading in MSFT and NVDA, continued relative stability in META, and firmer behavior from AMZN would support holding the current weights. Another broad AI selloff combined with renewed Treasury yield pressure would rule out adding exposure. It would prompt a fresh thesis review rather than an automatic price-based exit.
Positioning
| Holding | Weight | Thesis state | Action | Evidence |
|---|---|---|---|---|
| MSFT | 29% | INTACT. | Hold. | Thursday's 1.35% decline weakened the immediate setup, but 2.85% 20-day momentum and 8.65% 60-day momentum remain positive. |
| AMZN | 15% | WEAKENED. | Hold at the book's smallest weight. | Thursday's 2.25% decline and negative 0.51% 60-day momentum require better relative behavior before any increase. |
| NVDA | 28% | INTACT. | Hold, but don't add before stabilization. | Thursday's 2.94% decline followed pressure on AI-linked shares, while 1.9% 20-day momentum and 5.55% 60-day momentum remain positive. |
| META | 28% | INTACT. | Hold. | Thursday's 0.06% decline showed relative resilience, and 13.8% 60-day momentum remains strong. |
The book remains fully invested in high-quality growth. MSFT and NVDA account for 57% of the portfolio, reflecting the strongest conviction in enterprise AI, cloud software, and semiconductor infrastructure. AMZN's smaller weight contains the risk from its weaker medium-term price evidence.
Institutional Signals
Institutional filings support concentration as a portfolio style, but they don't settle Friday's decision. Pershing Square reported AMZN at 17.4% and MSFT at 15.3% in its May filing. Bridgewater's August filing included NVDA at 3.2% and AVGO at 2%, alongside large broad-market ETF positions. Berkshire's August filing also remained concentrated in a small set of holdings. These disclosures reinforce long-horizon conviction in durable franchises, while their reporting lag makes them unsuitable as trading signals.
What Could Break It
-
Evidence that major hyperscalers are cutting or deferring 2027 capital spending would directly weaken NVDA and the broader AI infrastructure thesis.
-
Azure or AWS growth slowing while capital-spending guidance continues to rise would show that investment is outrunning monetization.
-
Failure by META's AI investment to support advertising or subscription growth would weaken the case for its 28% weight.
-
Renewed rate pressure, persistent tariff-linked inflation, and further oil supply disruption could compress growth valuations even if company earnings remain sound.
-
A sustained deterioration in relative leadership across MSFT, NVDA, and META, combined with weaker earnings evidence, would move the regime away from selective growth leadership.