Marvell raising its long-term targets is a direct signal that the chip side of the AI buildout still has room to run, beyond the usual mega-cap names. Evercore's upgrade of Amazon on AI-driven shopping growth shows the demand side of AI — not just the inf

Verdict: dual-confirmation thesis is sound, but neither leg is live yet — wait

The idea's two-sided confirmation is real but lopsided. Amazon's latest quarter (ended June 30, 2026) showed revenue up 10.5% sequentially to $200.6B with net margin jumping from roughly 16.7% to 31.2%, while the Wall Street Journal reported on August 27, 2026 that Marvell raised its long-term targets on AI demand. But Marvell's own most recent filing (quarter ended May 2, 2026) shows revenue down 70.5% sequentially to $2.4B and net margin compressed to 1.4% from the mid-30s, so the chip-side thesis leans heavily on the next earnings print in late August or early September 2026. Trade readiness is the binding constraint. AMZN sits just $0.50 above its 50-day EMA of $255.76 with RSI at 38.6 needing a cross above 45; MRVL at $244.2 needs a roughly 9% pullback to reach its $222–223 thresholds. The completed 24-month backtest returned 59.1% on a single trade with an 18.3% maximum drawdown, but the 60-month window produced no triggers at all, and no robust nearby-parameter setup was established — so the historical evidence is thin. One further caution: Amazon's latest ownership filing, for the period ended June 30, 2026, showed about $16.0M of net open-market insider selling across 18 holders — a dated but notable signal from people closest to the company. Verdict: the setup is well-framed but not yet live, so we wait for the entry checklist to complete rather than chase.

Trade now: both legs are waiting — AMZN needs an RSI turn, MRVL needs a pullback

Neither leg is live yet, so the instruction today is wait, not buy. For AMZN (last close $256.26), three of four entry conditions are close or met: price sits just $0.50 above its 50-day EMA of $255.76 (needs to be below), price is below the 20-day Bollinger midline at $267.56 (met), ADX at 44.8 is above 20 (met), but RSI at 38.6 has not crossed above 45 (needs a rise of about 6.4 points). For MRVL (last close $244.2), the situation is inverted: RSI at 60.5 is already above 45 and ADX at 49.4 is above 20, but price is $22.64 above its 50-day EMA of $221.56 and $21.21 above the Bollinger midline of $222.98 — a pullback of roughly 9% would be needed to reach those thresholds. Once an entry triggers, the risk plan is mechanical: a 2.4% stop loss and a 4.8% take profit apply to both legs, implying an effective reward-to-risk of about 2:1. Harder levels sit underneath: AMZN's second support zone is around $250 with first resistance near $260, while MRVL's second support is around $240 with second resistance near $260. The completed backtest on this exact rule set returned 59.1% over 24 months with one trade and an 18.3% maximum drawdown, which is the evidence read supporting the setup. One honest caveat on tuning: the parameter-sensitivity evaluation exceeded its time budget, so no robust nearby setup was established — you are trading the parameters as published, nothing optimized. Position sizing is fixed-risk at 2.4% with a 25% maximum position size. If AMZN's RSI crosses above 45 while price dips under the EMA, or MRVL pulls back to the $222–223 zone, both leg checklists can complete within a session or two; until then, patience is the trade.

Two independent AI confirmations, and the fundamentals actually back them up

The idea's core claim — that AI momentum is being confirmed from both the supply side (Marvell's chips) and the demand side (Amazon's AI-driven shopping) — is supported by two fresh, dated news events. Per the Wall Street Journal piece published August 27, 2026, Marvell raised its long-term targets as AI demand boosted second-quarter profit and revenue. One day later, per the CNBC report of August 28, 2026, Evercore upgraded Amazon explicitly because AI is driving more purchases on the platform. Two different companies, two different analyst conclusions, inside 24 hours — that is exactly the dual-confirmation structure the thesis requires. The fundamentals give the news substance. Amazon's latest quarter (ended June 30, 2026) shows revenue up 10.5% sequentially to $200.6B, net income up 107.1% to $62.6B, net margin jumping from 16.7% to 31.2%, and ROE nearly doubling to 11.4%. Operating cash flow rose 74.4% to $45.4B. On a full-year basis (ended December 31, 2025), Amazon grew revenue 12.4% to $716.9B with diluted EPS up nearly 30% year over year. Marvell's full-year picture (ended January 31, 2026) is the stronger growth story: revenue up 42.1% year over year to $8.2B, net income of $2.7B, gross margin at 51.0%, and a return on equity of 18.7% — placing it in roughly the top fifth of its IT-sector peer group on operating margin and ROE. The thesis that AI demand is flowing beyond mega-cap names into a supplier like Marvell is consistent with those numbers. On the quant side, the completed backtest over the most recent 24-month window produced a 59.1% return…

Scores

• Conviction score breakdown: 52
• Thesis support: 65
• Trade readiness: 35
• Risk quality: 60
• Backtest evidence: 45
• Fundamentals trend: 55

Watch items

• MRVL — MRVL price vs 20-day Bollinger midline
• MRVL — MRVL price vs 50-day EMA
• MRVL — MRVL RSI (14)
• AMZN — AMZN RSI (14)
• AMZN — AMZN price vs 50-day EMA
• AMZN — AMZN insider net open-market activity
• MRVL — MRVL free cash flow
• AMZN — Price below EMA (50)
• AMZN — Price below Bollinger (20)
• AMZN — RSI (14) crossed above 45
• AMZN — ADX (14) above 20