We are absolute financial geniuses.^*

Energy profits and scarce physical supply can protect the portfolio when inflation or geopolitics make growth stocks remember that reality exists.

Oil weakness becomes durable, energy earnings roll over, and the stocks stop outperforming even when commodity prices stabilize.

The remaining core depends on broader market participation and stable financing conditions. We removed one share on September 9 after the added breadth stage failed its price-and-credit test; the September 11 rebound has not restored the original repair levels.

The added stage has been removed. Continued weakness below $294 with deteriorating equal-weight participation and credit challenges the remaining core. Restoring the stage requires a recovery above $296 with improving breadth and financing conditions.

Industrial spending, reshoring, infrastructure, and a broader economy can reward companies that manufacture objects too heavy to become an app.

Orders and guidance weaken while the sector breaks trend and starts losing to the broad market for reasons more serious than one ugly session.

People continue aging without first obtaining permission from the Federal Reserve, giving healthcare durable demand and useful defensive ballast.

Policy, earnings, or price evidence damages the defensive thesis and the sector loses both absolute and relative trend support.

AI-infrastructure demand and semiconductor earnings support a measured sector allocation. Recent TSMC revenue growth is encouraging, but financing conditions, capital intensity and more efficient AI architectures can change the payoff. The ETF spreads company-specific risk; it does not remove technology-sector risk.

A close below $545 while Nasdaq breadth and high-yield credit both weaken, or material deterioration in AI-infrastructure demand, ends or reduces the thesis. The September 11 last regular-session trade was $568.48, above the review level; this is not the final settled exchange close.

Amazon grew sales 20%, operating income 43%, and AWS 37%—its fastest cloud growth in 18 quarters. We added a second share on the post-earnings pullback, apparently establishing the world's smallest cloud-and-commerce conglomerate.

A high-volume close below roughly $255 challenges the post-earnings breakout. A failure through the pre-earnings region near $235.50, weaker AWS demand, or impaired operating guidance would normally end the thesis.

Azure grew 43%, management guided to roughly 45% constant-currency growth, and the backlog showed that companies besides frontier AI labs are spending actual money. This is the sort of evidence that lets one share wear a very large ceremonial cape.

A heavy-volume close below roughly $410 challenges the post-earnings repricing; a return below the pre-earnings region near $390 normally ends the thesis unless the business evidence materially improves.

Western Digital grew revenue 44%, expanded non-GAAP gross margin to 54.4%, generated $1.28 billion of free cash flow, and guided the next quarter above the one it just reported. We waited through the opening plunge, then bought one share after the price reclaimed the premarket level on exceptional volume—an acquisition program of almost breathtaking moderation.

A close below $430 without a company-specific repair catalyst triggers reassessment; a decisive high-volume close below roughly $405, material estimate cuts, or evidence that margin and free-cash-flow improvement is not durable ends the starter thesis.