The giant cloud companies are spreading their chip orders across multiple suppliers instead of relying on one — Alphabet adding Marvell as a second source is a direct hit to Broadcom, whose custom-chip business was built on being the sole provider. Meanwh

Verdict: interesting thesis, no entry — wait

The thesis is genuinely interesting: Alphabet adding Marvell as a second custom-silicon source already shows up in Marvell's numbers — 42.1% revenue growth to $8.2B and an operating-margin swing from -10.3% in fiscal 2024 to +16.1% — while Broadcom's 23.9% growth sits at just the 56th percentile of its sector. But the short leg is the problem: Broadcom is the 99th-percentile operator on margins and free cash flow ($26.9B FCF vs Marvell's $1.4B), and shorting the strongest business in the space on a marginal share shift is a high bar, especially with the two names 55% correlated so the 'hedge' only trims volatility about 13%. Critically, this is a watch-list setup, not a signal: the four-condition entry never triggered across 1,236 daily bars over five years, and no robust loosened parameter set was established, so the trade simply is not on. **Conviction breakdown:** fundamentals trend 70 (Marvell's operating leverage is real); thesis support 60 (dual-sourcing is confirmed news but says nothing about allocation size); risk quality 45 (tight 2.36% stop against 75%-volatility names that gap on AI headlines); trade readiness 35 (no demonstrated entry in 60 months); trigger proximity 30 (Marvell sits ~3.2% above the required pullback zone with RSI mid-range).

Trade now

The long Marvell / short Broadcom pair is a watch-list setup today, not an active signal. The entry logic keys off Marvell's daily chart: price must first pull back to or below the 20-day average (currently about $222, roughly 7.5 points or 3.2% below the last close near $229.6), then see RSI cross back above 40, price close back above the average, and a MACD signal cross — all on the same bar. Right now none of that is in place: Marvell is above its 20-day average with RSI at 53.1 and MACD at 5.17, and the simultaneous four-condition entry has never fired across 1,236 daily bars over five years. That is why the research author requested bounded optimization to loosen the conjunction — but no robust alternative parameter setup was established, so treat the published rules as-is. What "wait" means concretely: do nothing until Marvell trades down to the $222 area (or the then-current 20-day average) and then shows an RSI cross above 40 together with a MACD signal cross. Buying here, above the average with RSI mid-range, is front-running the strategy's own pullback requirement. Broadcom, near $359, is already 25% below its range high and below both its 50- and 200-day averages with RSI at 27.8 — the market has partly priced the Alphabet second-sourcing story the idea argues, which reduces the edge in the short leg if entered immediately. Risk framing if the entry does trigger: the strategy caps any single position at 25% of the book, risks 2.36% per trade, and takes profit at 4.7% against a 2.36% stop (roughly 2:1 reward-to-risk), with a secondary stop below Marvell's second-rank support level and a hard exit if RSI runs above 72 or the position ages past 45 daily bars. Note the two legs are 55% correlated, so this is only a partial hedge of Nvidia-earnings risk, not a market-neutral trade.

Second-sourcing is real, and Marvell's numbers back it up

The core of the thesis — Alphabet adding Marvell as a second custom-silicon source — has a direct fundamental echo in Marvell's reported numbers. Marvell grew revenue 42.1% year over year to $8.2B in its latest fiscal year, well ahead of Broadcom's 23.9% growth on a base nearly eight times larger. Per the Yahoo Finance piece on Alphabet dual-sourcing chips from Marvell, this is not speculation about future share gains; the orders are already flowing, and Marvell's top line is compounding faster than the incumbent's. Marvell's profitability inflection also supports the long leg. The company swung from negative operating margins as recently as fiscal 2024 (-10.3%) to +16.1% in the latest fiscal year, with gross margin recovering to 51.0% from a 37.1% trough in fiscal 2024. Free cash flow hit $1.4B, placing Marvell in the 97th percentile of its Information Technology peer set. That is the signature of operating leverage kicking in as custom AI silicon volume ramps — exactly what you want to see if second-source wins are compounding. The short leg is defensible on relative, not absolute, grounds. Broadcom remains a phenomenal business — a 39.9% operating margin (99th percentile among peers), $26.9B of free cash flow…

Scores

• Conviction score breakdown: 48
• Thesis support: 60
• Trade readiness: 35
• Risk quality: 45
• Trigger proximity: 30
• Fundamentals trend: 70

Watch items

• MRVL — Close vs 20-day EMA
• MRVL — RSI (14)
• MRVL — MACD (12,26,9) signal cross
• MRVL — RSI (14) — exit
• AVGO — RSI (14)
• NVDA — Next earnings report
• GOOGL — Supplier diversification evidence (earnings / next XBRL snapshot)
• AVGO — Price
• AVGO — RSI (14) crossed above 40
• AVGO — Price above EMA (20)
• AVGO — MACD (12,26,9) crossed above MACD (12,26,9)