President Donald Trump's Investment Team Sold Up to $25 Million in 2 Foundational AI Stocks in July

President Trump’s independently managed investment portfolio disclosed sales of $5.0M-$25.0M each in Amazon and Microsoft during July 2026, amid approximately 1,156 trades that month and 28,700 since January 2025. The sales occurred despite accelerating AI-cloud growth: AWS revenue rose 37% year over year, its fastest pace in 18 quarters, while Azure grew 43% in Microsoft’s fiscal Q4. The article attributes some activity to automated direct-indexing and profit-taking but flags elevated AI-driven equity valuations and the risk of a broader technology valuation correction.

Analysis

The disclosed selling is not a reliable fundamental signal: direct-indexing rebalances can create large gross turnover without expressing a discretionary view, and the reported value bands are too wide to infer position sizing or conviction. The more actionable read is that AMZN and MSFT have become duration-sensitive AI infrastructure proxies: their equity multiples now depend on sustained cloud growth converting into incremental operating margin, not merely continued capex. Any evidence that AI workloads are low-margin, cannibalize traditional software seats, or lengthen depreciation cycles would pressure free-cash-flow expectations simultaneously across hyperscalers.

Over the next 1-3 months, the key differentiator is monetization efficiency. MSFT is more exposed to enterprise budget scrutiny and Copilot attach-rate proof; AMZN has greater upside torque if AWS utilization absorbs fixed infrastructure costs, but also more consolidated-margin noise from retail and logistics. A broader AI de-rating would likely hit NVDA first through capex-expectation resets, then compress AMZN/MSFT multiples even if reported cloud revenue remains strong; conversely, a resilient capex cycle with improving cloud operating margins favors AMZN relative to MSFT. The contrarian view is that consensus is too focused on an eventual AI bubble while underweighting the possibility that hyperscaler scale converts inference demand into a durable oligopoly profit pool; the missing data are workload-level gross margins, customer contract duration, and depreciation-policy changes.

Key Decisions for Investors

• Do not trade AMZN or MSFT solely on the disclosed transactions; treat any sharp, flow-driven weakness over the next several sessions as a liquidity event, not a fundamental catalyst.
• Initiate a 1-3 month relative-value position: long AMZN / short MSFT in equal dollar amounts if AWS growth and margin commentary continue to improve relative to Azure. Target 8-12% relative upside; exit if Azure reaccelerates materially or AWS operating-margin expansion fails to appear in the next report.
• For portfolio AI-beta protection, buy 3-6 month QQQ put spreads or reduce NVDA exposure rather than shorting AMZN/MSFT outright. A capex-guidance reset or weaker enterprise IT spending would likely transmit first through semiconductor estimates; risk is continued AI capacity scarcity driving another leg higher.
• Set an earnings watch item: if either company raises AI infrastructure spending without a corresponding increase in cloud operating-income guidance, reduce hyperscaler exposure. That combination would signal lower near-term return on invested capital and could drive multiple compression over 6-18 months.