President Trump Sold 2 Popular Artificial Intelligence (AI) Stocks. Wall Street Says Investors Should Buy Both.

President Trump’s third-party-managed brokerage accounts sold $414,000-$785,000 of Palantir and $687,000-$4.2 million of Meta stock in June, though the article argues these trades have limited current relevance. Palantir reported Q2 revenue growth of 93% to $1.9 billion and non-GAAP EPS growth of 215% to $0.41, while analysts’ $205 median target implies 20% upside from $170 despite a 142x earnings valuation. Meta’s revenue rose 28% to $60.8 billion but GAAP EPS fell 13% to $6.18 amid severance and legal costs; its $745 median target implies 21% upside from $613 despite an $18 billion settlement with 48 states.

Analysis

PLTR’s key debate is no longer enterprise-AI demand but durability of growth at a valuation that leaves little room for execution slippage. At roughly 142x forward adjusted earnings, a deceleration in U.S. commercial customer growth, government award timing, or net-revenue retention could drive a 20-30% multiple reset even if results remain fundamentally strong. The more attractive second-order beneficiary of sustained Palantir deployment is NVDA: model inference and expanded operational use cases raise compute consumption, whereas PLTR bears the greater duration-risk exposure.

META offers the cleaner risk-adjusted AI exposure because its ad auction can monetize improved recommendation, creative generation, and agent-driven commerce without requiring enterprise IT budget cycles. The legal settlement removes a large discrete uncertainty, but youth-engagement controls could create a less-visible revenue headwind if they reduce teen inventory or prompt analogous restrictions in Europe; this is likely a 6-18 month issue rather than an immediate earnings event. The relevant test is whether AI-driven ad pricing offsets incremental compliance, content-safety, and infrastructure expense.

Consensus is too willing to treat PLTR’s analyst target dispersion as validation of upside; targets are generally poor anchors when the investment case rests on a sustained premium multiple. Conversely, META’s multiple can expand if management demonstrates that agent products produce measurable conversion or business-messaging revenue, rather than merely incremental engagement. The June portfolio activity has no informational value for either security and should not influence positioning.

Key Decisions for Investors

• Favor long META / short PLTR as a 3-6 month relative-value trade: META has more diversified cash earnings and a clearer path to AI monetization, while PLTR is exposed to a high-duration valuation reset. Target 15-20% relative outperformance; exit if PLTR sustains commercial-growth acceleration while META reports ad-price deceleration or materially raises capex guidance without monetization evidence.
• Add META on post-earnings weakness if the drawdown is driven by one-time legal/compliance costs rather than ad-demand deterioration; use a 12-month horizon. Risk is a material regulatory remedy that constrains personalization or a capex step-up that cuts forward free-cash-flow estimates by more than 10%.
• Do not initiate a directional PLTR long at current valuation; maintain a watch alert for a 25%+ correction or evidence that commercial growth, remaining deal value, and operating margin can support upward 2027 earnings revisions. A long is justified only if estimate revisions outpace the multiple compression risk.
• For AI-beta exposure paired with lower application-layer valuation risk, prefer NVDA over PLTR on a 6-12 month horizon, contingent on hyperscaler capex guidance remaining intact. Falsifier: broad cloud-provider inference spending cuts or a sequential decline in data-center order visibility.