OpenAI files for a mega IPO — buy the chipmakers powering the boom

Verdict: real earnings, reluctant insiders — wait for Broadcom's confirmation

The thesis has real earnings behind it — NVIDIA's last full year grew revenue 73.2% to $215.9B with a 75.0% gross margin, and Broadcom's fiscal year produced $26.9B of free cash flow — so this is not a hype-only trade. The strongest single argument for it is Broadcom's setup: at $368.43 with RSI at 61.9, it is the only name already inside the entry band, with its support tag near $363.83 one confirmed close away. The strongest argument against is who is not buying: ownership reports for the period ended June 30, 2026 show net open-market insider selling of roughly $565M at NVIDIA, $283M at Broadcom and $153M at AMD, while NVIDIA's latest quarter saw free cash flow fall 56.0% to $21.4B even as revenue grew 17.9% to $96.2B. The strategy's rule set could not be evaluated on the supplied 4-hour data, so this verdict rests on live levels and fundamentals, not realized trade statistics. Verdict: wait — take Broadcom only after a confirmed 4-hour close back above VWAP with support holding near $363.83, and flip the verdict if the next quarterly filings confirm free-cash-flow conversion or a new dated IPO milestone makes the catalyst schedulable.

Trade now: two names on hold, one on watch, one already past the window

Nothing is actionable at this moment, and that is the honest headline. The idea is a long on the AI chip complex — NVDA, AMD and AVGO on the 4-hour chart — waiting for a close crossing above the 20-period VWAP with the 14-period RSI between 50 and 70, a tag of the nearest support, and a hold above that support. This rule set has not produced an evaluable historical window (the supplied price history for the 4-hour timeframe was insufficient), so today's plan is built entirely on live levels, not on historical trade statistics. Where each name stands: NVDA closed at $225.77 with RSI at 37.1 — the RSI is 12.9 points below the 50 floor, so the entry band is not in reach; the low-of-entry condition also needs a touch of the nearest support near $220.92 first. AVGO is the closest to ready: at $368.43 with RSI at 61.9, squarely inside the 50–70 band, but the VWAP value is not currently available, so the crossover leg cannot be confirmed — treat AVGO as on watch, not on trigger. AMD is actually past the window: its RSI printed 77.1, above the 70 ceiling, meaning the entry conditions are gone and, if a position existed, the overbought exit would already be armed. Risk is defined mechanically, not discretionarily: each position uses a fixed stop at a 2.5% loss and a fixed take profit at a 5.0% gain — an effective 2:1 reward-to-risk — capped at 25% of the book per name. The paper tracker is live with zero trades and no realized profit or loss since it started, which is what a waiting strategy should look like. 'Wait' here means: do nothing on NVDA until RSI recovers above 50 and price tags and holds support near $220.92; monitor AVGO for the VWAP crossover confirmation; and do not chase AMD while its RSI sits above 70. One caution on the AMD reading: it is the only name of the three whose RSI has overshot the band, and its latest quarter (June period) showed revenue up 12.5% sequentially to $11.5B with free cash flow down 39.3% to $1.6B — momentum without the same cash conversion, per the SEC XBRL facts. That is context for patience, not a reason to override the rules.

The fundamentals behind the 'picks and shovels' trade

The idea argues that OpenAI's confidential IPO filing (per CNBC, June 8, 2026) validates AI spending and makes the chipmakers a safer way to own the boom than the unproven IPO itself. On the latest full fiscal-year data, that argument has real earnings behind it, not just narrative. NVIDIA's most recent reported year shows $215.9B of revenue growing 73.2% year over year, a 75.0% gross margin, and $34.9B of free cash flow in the final quarter — placing it in the top 1.5% of its sector on cash generation and the 98.9th percentile on operating margin. Broadcom is not far behind: its latest fiscal year produced $63.9B of revenue, a 39.1% operating margin (98th percentile among 854 IT-sector peers), and $26.9B of free cash flow. These are exactly the 'real, massive earnings' the thesis leans on. The most recent quarter-to-quarter fundamentals continue to support the demand story. AMD grew revenue 12.5% sequentially to $11.5B in the quarter ended June 27, 2026, with net income up 66.1% to $2.3B and net margin expanding from 13.5% to 19.9%. Broadcom's quarter ended May 3, 2026 showed revenue up 14.9% sequentially to $22.2B with free cash flow up 28.1% to $10.3B. NVIDIA's quarter ended July 26, 2026 delivered $96.2B of revenue, up 17.9% sequentially, with a 66.2% operating margin. The Yahoo Finance piece from June 9, 2026 specifically frames Nasdaq's chipmakers as powering a further rebound after the OpenAI filing, which aligns with the idea's front-running logic. Meanwhile AMD still trades at mid-pack quality — 54.6th percentile gross margin despite an 87th-percentile operating margin — meaning even modest AI-driven share gains leave room for multiple expansion, unlike the more fully priced leaders. Balance-sheet risk is not the threat here. AMD's debt-to-equity sits at a trivial 0.03 and is falling; Broadcom cut leverage from 0.80 to 0.71 over the latest quarter while still growing free cash flow sharply; and AMD's free cash flow for the full fiscal year ended December 2025 reached $6.7B, up from $2.4B the prior year. One scope note: the strategy's rule set could not be evaluated on the supplied 4-hour data (the compiled strategy could not produce an evaluable backtest window), so this analysis scores the fundamentals, ownership, and news dimensions rather than any realized trading statistics. The unresolved question for the thesis: can sequential growth of this magnitude persist long enough for the IPO catalyst to matter, or is the market already discounting it?

Insiders are selling, and the cash-flow optics are wobbling

The clearest red flag in the data is who is NOT buying this thesis: the people closest to these companies. As of the ownership…

Scores

• Conviction score breakdown: 54
• Thesis support: 60
• Trade readiness: 40
• Risk quality: 45
• Fundamentals trend: 72

Watch items

• NVDA — RSI (14), 4h
• NVDA — Price vs nearest support
• AVGO — RSI (14), 4h
• AVGO — 4h close vs VWAP
• AMD — RSI (14), 4h
• NVDA — Insider net open-market activity (Q2 2026 filings)