If hyperscaler AI spending really approaches $1 trillion next year, the revenues flowing to the chipmakers and cloud giants at the center of that buildout keep growing. At the same time, large pools of money like pensions are finding that traditional dive

Verdict: The $1 Trillion AI Bid Is Real Enough to Watch, Not Yet to Trade

The thesis has real fuel. NVIDIA grew revenue 73.2% year over year in its latest quarter with a 65.0% operating margin, Microsoft posted 17.7% growth at a 45.1% margin, and per the CNBC interview of September 21, 2026, potential hyperscaler spending sits near $1 trillion — so the demand tailwind shows up in reported numbers, not just narrative. But the strongest counterweight is that filed disclosures for the period ended June 30, 2026 show net open-market insider selling at all four names, roughly $565 million at NVIDIA, and NVIDIA's free cash flow fell 56% quarter over quarter to $21.4B while its debt-to-equity ratio jumped from 0.04 to 0.15. The completed 9-month backtest is encouraging but short: a 12.1% return over 7 trades in AMZN with a 57.1% win rate and a 7.9% maximum drawdown, and because exits were filled on daily bars, stop quality may be overstated. No robust parameter setup was established, so the rule set rests on a single market regime. Critically, no entry is live today: every ticker's RSI sits above 45 but not on a fresh upward cross, AMZN trades below its 50-day EMA ($254.43 versus $255.2), and MSFT's ADX of 0.12 signals no trend at all. The verdict is wait. The bid may be real, but the framework's own rules say the entry has not arrived, and the freshest ownership and cash-flow evidence leans against the "forced buying" half of the story.

Trade now: No entry is live yet — all four legs are waiting on a fresh momentum trigger

Nothing in this four-name setup (AMZN, GOOGL, MSFT, NVDA) is tradeable today, and that is the honest read. The strategy requires three conditions to align before entry: price above the 50-day EMA, a fresh upward cross of the 14-day RSI above 45, and a 14-day ADX above 20. Right now the RSI condition is the blocker on all four tickers — AMZN's RSI sits at 49.8, GOOGL's at 62.5, MSFT's at 53.4, and NVDA's at 65.3, all above 45 but not on a fresh upward cross, which is what the rules demand. "Wait" means doing nothing until the RSI dips and then crosses back above 45 while the other two conditions hold. The rest of the checklist is close but incomplete. AMZN (last close $254.43) is furthest behind: it trades $0.76 below its 50-day EMA of $255.2, and its ADX of 15.7 is short of the 20 threshold, so two of three conditions are unmet. GOOGL ($352.30) and NVDA ($229.46) have price and ADX in place — GOOGL's ADX is a strong 50.4, NVDA's just cleared 20.1 — while MSFT ($498.86) has price well above its EMA of $472.21 but an ADX of just 0.12, signaling no trend strength at all. The completed 9-month backtest on this rule set returned 12.1% over 7 trades with a 57.1% win rate and a 7.9% maximum drawdown, so the machinery has been tested; the market just has not handed us the entry yet. When an entry does trigger, the risk math is fixed: each position uses a 2.4% stop loss and a 4.8% take profit — a 2:1 reward-to-risk ratio — with no single name above 25% of the book. Exits also fire if price closes below the 50-day EMA alongside an RSI cross below 45 after 90 bars held; that is the sentiment break the thesis warns about. Discipline is the product here: the near-trigger RSI readings on GOOGL and NVDA mean those setups could arm within days of a pullback-and-recover sequence, but chasing early is exactly what the rules are built to prevent.

The fundamentals behind the $1 trillion spending thesis are real, and the rules traded them profitably

The idea's core premise — that hyperscaler AI spending approaching $1 trillion (a figure Jamie Dimon floated in the CNBC interview of September 21, 2026) keeps revenue flowing to the chipmakers and cloud giants — is visible in the reported numbers, not just asserted. NVIDIA grew revenue 73.2% year over year in its latest quarter, placing it in the 92nd percentile of its 897-company sector cohort, with a 65.0% operating margin and $34.9B in free cash flow. Microsoft grew revenue 17.7% year over year to $90.0B with a 45.1% operating margin in the top 2% of its sector, and Alphabet posted 18.0% year-over-year revenue growth alongside a 31.6% operating margin. These are the balance-sheet fingerprints of a genuine capex cycle, not a narrative one. The Bloomberg piece from September 20, 2026 on diversification breakdown points at the second leg of the thesis: index-dominant mega-caps pull passive and pension flows into the same four names. The idea's argument is that this forced concentration creates a persistent bid. The reported earnings trajectory supports the mechanism — Alphabet's quarterly net income rose to $112.2B in the June 2026 quarter from $62.6B the prior quarter, and its net margin reached 93.7%, while Microsoft's quarterly revenue rose 8.6% sequentially to $90.0B in the quarter ended June 30, 2026. When the earnings leaders keep compounding, the index-weight feedback loop has something to feed on. The strategy evidence matches the direction. Over a 9-month, 184-bar backtest on daily data, the momentum rules produced 7 trades in AMZN with a 57.1% win rate, a 12.1% total return, and a maximum drawdown of 7.9%. The equity curve shows the drawdown hitting early — the portfolio sat at roughly -5.8% in mid-March 2026 — and then recovering to +12.2% by September. That is exactly the shape the thesis wants: take the pain on sentiment wobbles, ride the trend when the bid returns. The risk-control layer is a genuine part of the bull case. Each position risks about 2.4% of equity per trade, positions are capped at 25%, and a hard stop sits at -2.4% per position with a take-profit at +4.8%. Combined with exits that trigger when price falls below the 50-day average or RSI crosses under 45 after 90 bars, the framework means a sudden AI-sentiment break — the thesis's own stated risk — mechanically cuts exposure rather than averaging down. One caveat the bull case must carry honestly: the completed backtest covers a 9-month window; longer 12-, 24-, and 60-month evaluations did not complete due to a data-coverage gap in AMZN daily bars, so the track record is short and concentrated in one regime. But within…

Scores

• Conviction score breakdown: 53
• Thesis support: 68
• Trade readiness: 35
• Risk quality: 55
• Backtest evidence: 48
• Fundamentals trend: 58

Watch items

• GOOGL — RSI (14)
• NVDA — RSI (14)
• AMZN — Price vs 50-day EMA
• MSFT — ADX (14)
• NVDA — Insider net open-market selling (June 30, 2026 filing)
• NVDA — Free cash flow
• AMZN — Close vs 50-day EMA (exit proximity)