Friday plan: restore growth leadership while retaining JPM as the macro anchor
Friday plan: trim JPM to 31% and raise NVDA to 20%, while holding MSFT and AMZN. The target remains fully invested and growth-led, conditional on the open confirming technology participation without a renewed bond sell-off.
Accountability grade: B- — the prior report's caution was too guarded because all four holdings advanced Thursday and the prior book gained 1.95% versus SPY's 1.05%, outperforming by 0.90 percentage points; the specific shift toward JPM was not validated by relative performance because NVDA gained 1.8% while JPM gained 1.64%.
Since Last Session
Thursday finished positively across the prior book. JPM gained 1.64%, MSFT gained 2.68%, AMZN gained 1.54%, and NVDA gained 1.8%. The prior book gained 1.95% versus SPY's 1.05%, an advantage of 0.90 percentage points. NAV stood at 108131 versus a peak of 109899, leaving drawdown at -1.61% and the drawdown protocol at normal.
Regime Check
Regime: R1 — constructive risk-on conditions have re-emerged, led by low volatility, positive broad-market momentum, and renewed participation from technology and financials.
Supporting: VIX was 14.32. SPY gained 1.05% with 60-day momentum of 2.6%, QQQ gained 1.19%, XLK gained 1.29%, and XLF gained 1.56%. MSFT, NVDA, and AMZN all advanced, showing that Thursday's strength included the portfolio's principal growth exposures.
Contradicting: IWM gained only 0.4% and retained negative 20-day and 60-day momentum, so participation was not uniformly broad. TLT's 60-day momentum was -2.24%, and the packet reported an ongoing global bond sell-off. That rates pressure remains a valuation risk for long-duration growth and prevents an unqualified risk-on call.
Plan for Friday
The intended move is to trim JPM from 35% to 31% and redeploy the four points into NVDA, raising it from 16% to 20%. MSFT will remain at 29% and AMZN at 20%. This preserves every existing thesis while shifting the book back toward the growth leadership evident Thursday and keeping NVDA within the permitted weight band for a conviction-7 holding.
The open must confirm continued participation from technology and semiconductors without a renewed acceleration in bond selling. If technology participation narrows immediately or duration weakens sharply, the resize should not be chased; the existing holdings should remain the preferred exposures until the evidence becomes clearer.
Positioning
JPM — thesis state: INTACT. Action: Trim to 31%. Evidence: JPM gained 1.64% and XLF gained 1.56%, preserving the financials thesis, but renewed growth participation reduces the need for a 35% allocation.
MSFT — thesis state: INTACT. Action: Hold at 29%. Evidence: MSFT gained 2.68% and retained positive 20-day and 60-day momentum, while no new company-specific evidence changed the Azure and AI monetization thesis.
AMZN — thesis state: INTACT. Action: Hold at 20%. Evidence: AMZN gained 1.54% and retained positive 60-day momentum, but the packet supplied no new AWS or retail-margin evidence warranting a resize.
NVDA — thesis state: INTACT. Action: Raise to 20% if the open confirms technology participation without renewed bond stress. Evidence: NVDA gained 1.8%, retained positive 20-day and 60-day momentum, and remains directly exposed to AI infrastructure demand.
The resulting book is 69% allocated to MSFT, AMZN, and NVDA, with JPM's 31% allocation retaining meaningful exposure to rates, liquidity, and capital-markets durability.
Institutional Signals
Institutional filings provide slow-moving confirmation of concentrated exposure rather than a trading signal. Bridgewater's 2026-08-14 filing included NVDA at 3.2% alongside large broad-market ETF positions, supporting AI participation within a diversified institutional book. Pershing Square's older 2026-05-15 filing carried AMZN at 17.4% and MSFT at 15.3%, consistent with concentrated ownership of durable cloud and software franchises. Berkshire's filing reinforces the broader practice of concentrating in a limited number of high-conviction businesses, but it does not directly validate today's individual holdings. Scion's 2025-11-03 filing is too stale and idiosyncratically concentrated to drive the allocation.
What Could Break It
A renewed global bond sell-off that materially intensifies duration pressure would weaken the R1 classification and challenge growth multiples. Technology leadership that narrows while small-cap momentum remains negative would suggest rotation rather than a durable broadening of risk appetite. At the company level, consecutive Azure deceleration with rising capex, slowing AWS growth with higher capex, lower NVIDIA gross-margin guidance, hyperscaler spending cuts, or weaker JPM net interest income alongside rising provisions would invalidate the relevant position theses.