Chip stocks are carrying the entire market while tech giants bleed — long Nvidia and AMD, short Microsoft and Apple
The overall stock market is having its best quarter in years thanks to chipmakers, even as the giant tech companies known as the 'Magnificent Seven' have lost trillions in value. This mismatch creates a trade: buy the companies powering the rally and sell the giants that are dragging it down.
Idea
There is a massive split happening in the stock market right now. The headlines say the S&P 500 and Nasdaq are having their best quarter in six years, powered almost entirely by chip and semiconductor companies. But at the exact same time, the seven biggest tech giants — Apple, Microsoft, Amazon, and the rest — have watched $2.3 trillion evaporate. This means the market's gains are concentrated in a narrow group of hardware companies while the software and internet giants are sinking. By betting on chipmakers like Nvidia and AMD while simultaneously betting against the struggling giants, you're positioning for the trend that's already working.
Advanced Analysis — institutional-depth research report
Verdict: the chip-versus-giants split is real, but there is nothing to trade yet
The idea's core claim is visible in the filings: NVDA just booked $96.2B in quarterly revenue, up 17.9% sequentially, while Apple's revenue slipped 1.6% to $109.4B and Microsoft's latest quarter came in at $90.0B — a genuine chip-versus-giant split. But the strongest counterpoint is on the record too: insider ownership filings covering the period ended June 30, 2026 show net open-market selling at all four names, led by roughly $375.3M at NVDA, and NVDA's free cash flow fell 56.0% to $21.4B that quarter even as revenue jumped. What should decide it for now is the strategy status: the entry rules were evaluated over 1,235 daily bars across 60 months without opening a single trade, and the parameter evaluation ended without a robust nearby setup being established — this is a watch-list thesis, not a live signal. The single most important thing to watch is NVDA's net margin, which compressed from 71.5% to 62.0% in the latest quarter; another step-down would undercut the long-chips leg even if the megacap short leg works.
Trade now: a staged pair still waiting on its triggers
Nothing to execute yet. The strategy wants a coordinated move — SMH making new 20-day highs while the Apple/Microsoft/Amazon/Google basket slides at least 1% on the week — and today's market does not line up that way. On the long side, the entry conditions for SMH are only partly in place: the 14-period trend-strength check reads 26.2, above the 25 threshold, and the 12-day rate of change is 1.9% versus the 0.5% requirement — both satisfied — but the momentum-condition checks are far from triggering. On the short side, AMD trades at $496.02 with its trend-strength reading at just 1.5 versus the required 25, nowhere near the entry condition. In short, every entry leg is still waiting. One transparency note that materially affects the plan: the compiled entry logic currently demands, on every leg, both a cross above zero and a price below negative one at the same time — conditions that cannot coexist — and the research author has requested a bounded fix that preserves the symbols, direction, exits, and risk rules. Until that re-run establishes a robust setup, there is no validated parameter set to lean on, and we state that plainly rather than guessing at thresholds. Treat everything below as a watch list. If and when a coherent setup goes live, the risk envelope is already defined: each position risks 2.5% of its value with a take-profit at 4.9% — roughly 2:1 reward-to-risk — and positions capped at 14.3% of the book. So "wait" means concretely: do nothing on SMH, NVDA, AMD, or the Mag-7 basket legs until the entry conditions fire simultaneously, and expect the fixed logic to be re-evaluated against an untouched holdout before any trade is taken. The underlying thesis is still supported by the live data — NVDA just posted $96.2B in quarterly revenue, up 17.9% quarter over quarter, and AMD posted $11.5B, up 12.5% — while the June 30 ownership reports show net open-market selling across all four names, including roughly $375 million of NVDA sales. That is the tension to monitor, not an entry signal.
Chip Fundamentals Are Doing What the Thesis Says: Accelerating While Megacaps Flatten
The core of this idea is a divergence, and the fundamental record supports it. NVIDIA's latest quarter (ended July 26, 2026) produced $96.2B in revenue, up 17.9% from the prior quarter, on a 62.0% net margin and 66.2% operating margin — near the top of its 854-peer sector grouping. AMD's quarter, ended June 27, was even sharper on the margin side: revenue of $11.5B, up 12.5% sequentially, with net income jumping 66.1% to $2.3B and net margin expanding from 13.5% to 19.9%. These are exactly the earnings profiles a rotation into semiconductors would reward. Meanwhile the giants the idea wants to short look qualitatively different. Apple's most recent quarter showed revenue of $109.4B, down 1.6% sequentially, with net margin of just 27.2% — a strong business, but one that is shrinking at the top line while its per-share payout growth (3.9% annual dividend growth) suggests a maturing capital-return story rather than a growth one. Microsoft's most recent annual results (ended June 30, 2026) still delivered $331.8B in revenue with a 46.8% operating margin, but its Q4 alone came in at $90.0B in revenue versus a $133.7B prior-period net income figure — momentum that is decelerating relative to the semiconductor cohort. The thesis's central claim, that gains are concentrated in hardware while the mega-cap software/internet complex stalls, is visible in these filings, and it is corroborated by the cited market coverage: the Wall Street Journal and Yahoo Finance both reported the S&P 500 and Nasdaq heading for their best quarter in six years on June 30, 2026, while Yahoo Finance separately reported the Magnificent Seven losing $2.3 trillion in June 2026. Two structural tailwinds reinforce the long side of the pair. First, NVIDIA's balance sheet is flexible despite its scale: free cash flow of $21.4B in the latest quarter and a gross margin holding above 74% leave room to absorb capex or volatility. Second, AMD's debt-to-equity is just 3.5% and falling, and its ROE more than doubled to 3.4% from 2.1% in a single quarter — a company with genuine operating leverage if the chip cycle extends. On the strategy itself, the entry conditions have not occurred in the evaluated window — the rules were run on 1,235 daily bars across 60 months (and again over 24- and 12-month windows) without opening a position, so this is a watch-list setup rather than an active signal. That is a conditional filter being strict, not a trust deficit: the fundamentals above are what the entry is waiting for — SMH making new 20-day highs while the mega-cap basket weakens. When those conditions align with the fundamental split documented here, the setup becomes live.
Insiders Are Selling All Four Names, and NVIDIA's Cash Flow Just Halved
The most striking adverse signal is uniform: every one of the four companies — AAPL, AMD, MSFT, and NVDA — shows net open-market insider selling in the ownership filings covering the period ended June 30, 2026. NVIDIA's is by far the largest, at roughly $375.3 million of net selling across 23 reporting holders, followed by AMD at about $153.2 million across 15 holders. Insider conviction does not appear to be backing the long…
Scores
- Conviction score breakdown: 45
- Thesis support: 70
- Trade readiness: 20
- Risk quality: 45
- Trigger proximity: 25
- Fundamentals trend: 65
Watch items
- AAPL — AAPL RSI (14)
- AMD — AMD trend strength (ADX 14)
- NVDA — NVDA RSI (14)
- NVDA — NVDA insider net open-market flow
- NVDA — NVDA net margin
Key details
Community
News sources
- Magnificent Seven stocks lose $2.3 trillion in June 2026 — Yahoo Finance
- S&P 500, Nasdaq Head for Best Quarter in Six Years — Wall Street Journal
- Nasdaq heads for best Q2 since 2020 as chip stocks surge — Yahoo Finance