Amazon and Microsoft just proved that the billions being spent on AI are translating into real cloud growth, separating them from competitors like Meta and Alphabet who missed their numbers. Yet Nvidia, which sells the chips to all of them, is stuck in a

Verdict: strong thesis, but the entry mechanism has never fired

The thesis that AI spending is concentrating in proven cloud winners has genuine fundamental support — per the Yahoo Finance piece, Amazon and Microsoft separated themselves from Alphabet and Meta on cloud growth, and Microsoft's 45.6% operating margin (98.7th percentile among IT peers) and $71.6B in free cash flow demonstrate real profitability. The strongest point against the trade is not the fundamentals but the mechanics: the entry rules produced zero triggers across 1,787 evaluated daily bars over 60 months, and the research author's authorized expanded parameter search could not establish any robust alternative setup before exceeding its time budget. Amazon's most recent quarter (ending March 2026) showed negative free cash flow of $18.2B against $131.8B in capital expenditures, a sharp reminder that the capex burden required to maintain cloud dominance is consuming cash at an accelerating rate. With both names currently at overbought extremes — Microsoft's RSI at 91.3 and Amazon's at 74.1 — and prices sitting 13–15% above their Bollinger Band entry zones, this is a watch-list setup waiting for a pullback that may be structural rather than cyclical. A sustained sector-wide de-rating, potentially triggered by capex-fatigue questions or a laggard contagion event like the Meta plunge cited by Investor's Business Daily, would be the most direct path to the oversold conditions this strategy needs. **Conviction Breakdown** - **Thesis support (70):** The concentration-of-winners narrative is well-supported by the cited August 3 earnings results and Microsoft's top-tier profitability metrics. - **Trade readiness (15):** Zero historical triggers over 1,787 bars, no validated parameter alternative, and prices 13–15% from entry zones make this unactionable today. - **Risk quality (45):** The risk-parity basket offers a 1.37 diversification ratio and balanced 52.5/47.5 weights, but an expected maximum drawdown of 32% and fat-tailed kurtosis readings (13.7 for AMZN, 16.4 for MSFT) temper enthusiasm. - **Trigger proximity (10):** The primary Bollinger-band and RSI conditions are classified as far from current levels; only ADX (above 20) is met. - **Fundamentals trend (60):** Microsoft's trend is uniformly strong, but Amazon's deteriorating free cash flow — from $32.9B in FY2024 to negative $18.2B in Q1 2026 — introduces real uncertainty about the sustainability of its capex path.

Trade now

This setup is currently on the shelf. The thesis argues that Amazon and Microsoft deserve a premium for proving out AI-driven cloud revenue, but the strategy's entry rules are waiting for a pullback that simply is not happening today. Amazon last closed at $277.42 and Microsoft at $492.81 — both well above their 20-period Bollinger Bands of $245.32 and $419.47, respectively. For the primary entry to activate, Amazon would need to drop roughly $32 from current levels to reach its Bollinger trigger, while Microsoft would need to fall about $73. Neither name is close to triggering the RSI crossover condition either: Amazon's 14-day RSI sits at 74.1 (the rule needs a cross back above 35 from below) and Microsoft's at 91.3, both firmly in overbought territory rather than the oversold reversal state the strategy is built to catch. The one entry condition already satisfied is trend strength. Amazon's ADX (14) reads 75.5 and Microsoft's 95.5, both well above the 20 threshold — so the market is trending hard, just in the opposite direction of what the setup wants. The MACD line is also coiled near its signal on both tickers, meaning a momentum cross could arrive soon, but without the price and RSI components falling into place first, that cross alone won't fire an entry. "Wait" here means exactly that: no position, no partials, no anticipation buys. The strategy's hard stop is a 2.4% loss and its fixed take-profit is 4.9%, for an effective reward-to-risk of roughly 2:1. Those levels only matter once price actually retreats to the Bollinger zone with RSI recovering from a deeply oversold reading. The rules were evaluated across 1,787 daily bars over five years without triggering once, which tells us the current parameter stack — four overlapping conditions on a single bar — is extremely restrictive. The research author has flagged this and authorized an expanded parameter search to find a workable entry definition, but until that produces a validated alternative, the live levels below are what you monitor. If you want to track readiness, the key levels to watch are Amazon's Bollinger band at $245.32 (about 13% below current price) and Microsoft's at $419.47 (about 15% below). A daily close at or below those bands, coinciding with RSI climbing back through 35, would put the setup on alert.

The Cloud Monopoly Moat Is Real — And the Numbers Prove It

The central claim — that AI spending is concentrating in the proven cloud winners rather than distributing across the whole sector — is directly supported by the most recent earnings data. Per…

Scores

• Conviction score breakdown: 40
• Thesis support: 70
• Trade readiness: 15
• Risk quality: 45
• Trigger proximity: 10
• Fundamentals trend: 60

Watch items

• AMZN — Price vs Bollinger Band (20)
• AMZN — RSI (14)
• AMZN — ADX (14)
• MSFT — Price vs Bollinger Band (20)
• MSFT — RSI (14)
• AMZN — RSI (14) — exit signal
• MSFT — RSI (14) — exit signal
• AMZN — Price
• AMZN — RSI (14) crossed above 35
• AMZN — MACD (12,26,9) crossed above MACD (12,26,9)
• AMZN — ADX (14) above 20