AI money is spreading beyond Nvidia to the second tier — buy AMD as the chip rotation leader
The AI trade so far has been concentrated in one company, but the story is widening: Chinese AI labs are raising billions more than planned and heading toward huge IPOs, showing the buildout is global and accelerating. Meanwhile, the shift toward personal AI agents favors more affordable, high-volume chips — exactly what AMD and Intel make — and both have already started outperforming their megacap peers. When money starts rotating from a single crowded winner into the next tier of beneficiaries, the lagging names often have the most room to run. Buying AMD (and to a lesser degree Intel) captures this broadening without paying Nvidia's premium.
Idea
The AI trade so far has been concentrated in one company, but the story is widening: Chinese AI labs are raising billions more than planned and heading toward huge IPOs, showing the buildout is global and accelerating. Meanwhile, the shift toward personal AI agents favors more affordable, high-volume chips — exactly what AMD and Intel make — and both have already started outperforming their megacap peers. When money starts rotating from a single crowded winner into the next tier of beneficiaries, the lagging names often have the most room to run. Buying AMD (and to a lesser degree Intel) captures this broadening without paying Nvidia's premium.
Advanced Analysis — institutional-depth research report
Verdict: the rotation thesis is credible, but AMD's own entry signal is miles away
The strongest case for this idea is fundamental: AMD's Q2 2026 (period ended June 27, 2026) showed revenue up 12.5% quarter over quarter to $11.5B, net income up 66.1% to $2.3B, a 19.9% net margin, and a peer-set ranking at the 85th percentile for revenue growth and the 99th for free cash flow, while Intel's free cash flow swung from -$2.5B to +$4.5B. The strongest case against is that the trade cannot be taken today and the ownership tape argues against front-running: AMD's entry conditions are all reading far (price $93.94 above its 50-day EMA, RSI at 72.5 versus the required 45), and provisional insider filings as of September 30, 2026 show net open-market selling of roughly -$176.2M at AMD and -$6.5M at Intel. The completed 60-month backtest offers some reassurance — 30.9% total return over 9 trades with an 18.7% maximum drawdown — but the 44.4% win rate rests on a thin sample, no shorter windows completed, and no robust parameter setup was established in the sensitivity evaluation. Meanwhile, Intel's Q2 net loss of $11.0B on a -68.4% net margin and a 42.8% jump in debt-to-equity means roughly half this basket's risk is a leveraged turnaround, not a clean beneficiary. Verdict: wait — the setup is waiting for a pullback that has not begun, and AMD's RSI is closer to the strategy's exit threshold (75) than to its entry (45). The verdict flips to actionable only if a sector pullback drives AMD's RSI to 45 or below with price under $574.01 and then under $537.81, accompanied by a next quarterly print that sustains the margin and cash-flow gains.
Trade now: the setup is bullish, but the entry is not live — here is exactly what has to change
**Nothing to buy today.** The strategy is a mean-reversion long on AMD (primary) and INTC (secondary), and both names are far from their entry conditions. AMD closed at $631.75 — that is $93.94 above its 50-day EMA of $537.81 and $57.74 above the middle Bollinger Band at $574.01, so the price conditions alone are a long way from triggering. Its RSI (14) is at 72.5 versus a required read at or below 45, and the stochastic cross condition is also far from firing. Every entry leg on AMD reads "far." INTC at $116.19 is closer but still not there: it is $9.41 above its 50-day EMA of $106.78 and only $3.47 above its middle Bollinger Band at $112.72 (the nearest condition), with RSI at 51.4 versus the at-or-below-45 requirement and the stochastic cross far away. "Wait" here means a concrete pullback: for INTC, a dip of roughly 3–9% plus RSI cooling six points; for AMD, a materially deeper retracement of 9–15% from the current close before any leg even approaches its threshold. **The reward:risk is defined by the strategy, not by hope.** Once a position is open, the hard exits cap risk at a 2.4% stop (plus a support-based stop) and a 4.7% take-profit — roughly 2:1 — with an RSI-overbought exit above 75 and a 60-bar time exit also active. The completed 60-month backtest on the AMD pair produced a 30.9% total return across 9 trades with a 44.4% win rate and an 18.7% maximum drawdown — a modest win rate carried by the asymmetric exits, which is exactly why chasing now, before the conditions fire, breaks the math. Position sizing is fixed-risk at 2.35% with a 25% maximum position size. **One caution if you are tempted to front-run.** No robust parameter setup was established in the sensitivity evaluation (it exceeded its time budget), so the published thresholds are the ones to trade — not improvisations. And the near-term momentum picture argues for patience: AMD's RSI at 72.5 is only about 2.5 points from the 75 exit threshold, meaning the strategy's own exit signal is closer than its entry signal. If AMD strengthens further from here, the honest outcome is that no trade happens. The 12- and 24-month backtest windows returned engine errors, so the 60-month AMD pair is the only completed evidence base — judge the idea on that window, not on extrapolation.
The numbers behind the rotation are real: AMD's margins and Intel's cash flow inflection
The numbers behind the 'rotation to the next tier' argument are stronger than the thesis states them. AMD's Q2 (period ended June 27, 2026) showed revenue of $11.5B, up 12.5% quarter over quarter, with net income jumping 66.1% to $2.3B and net margin expanding from 13.5% to 19.9%. Gross margin improved nearly a full point to 53.8%, and operating margin rose 2.9 points to 17.3%. Against a peer set of roughly 900 Information Technology companies, AMD sits in the 85th percentile for year-over-year revenue growth, the 87th percentile for operating margin, and the 99th percentile for free cash flow ($2.4B in the latest reported quarter). That is exactly the profile of a company that can absorb a high-volume, lower-priced chip cycle without margin collapse. The balance sheet supports the bull case too. AMD's debt-to-equity fell again to about 0.048 in Q2 2026 — essentially unlevered — while the full-year 2025 free cash flow of $6.7B dwarfs the $1.1B generated in 2023. There is no dilution pressure: shares outstanding are nearly flat at 1.63B over four years. Intel, the second leg of the idea, is itself inflecting: Q2 2026 revenue rose 18.8% QoQ to $16.1B, operating cash flow quintupled to $7.0B, and free cash flow swung from -$2.5B in Q1 to +$4.5B. For a strategy explicitly hunting for lagging names with room to run, a cash-flow-positive Intel with gross margin recovering toward 40.4% is the setup the idea describes. The cited news reinforces the demand half of the thesis rather than merely asserting it. Bloomberg reported on October 6, 2026 that DeepSeek will raise at least $12B in Tencent-backed funding, and that Moonshot is eyeing an early-2027 IPO at a $50B valuation — evidence the AI buildout is globalizing beyond a single customer base. CNBC's same-day piece on Meta's Muse launch frames the shift…
Scores
- Conviction score breakdown: 52
- Thesis support: 68
- Trade readiness: 22
- Risk quality: 55
- Backtest evidence: 45
- Fundamentals trend: 72
Watch items
- AMD — RSI (14)
- AMD — Close vs 50-day EMA
- AMD — Close vs middle Bollinger Band
- AMD — Stochastic (14) vs Stochastic (3)
- AMD — RSI (14) exit threshold
- INTC — RSI (14)
- INTC — Close vs middle Bollinger Band
- INTC — Close vs 50-day EMA
- AMD — Insider net open-market value (Sept 2026 filing)
- INTC — Insider net open-market value (Sept 2026 filing)
- INTC — Quarterly net margin