Johnson says no moratorium on AI, would give China competitive edge

U.S. House Speaker Mike Johnson opposed any moratorium on AI development, arguing that restricting the sector would risk ceding the U.S. competitive edge to China and create national-security consequences. Johnson said AI firms can self-regulate and expects AI to be discussed when President Donald Trump meets Chinese President Xi Jinping later this month, signaling a policy environment broadly supportive of continued AI investment and development.

Analysis

The policy signal marginally lowers the probability of a near-term U.S. compute-capex interruption, but it does not create new demand. APP and SMCI have materially different sensitivity: APP benefits primarily if advertisers continue funding AI-driven performance marketing, while SMCI remains a higher-beta expression of server deployment volumes, component availability and customer concentration. The more investable implication is reduced regulatory discounting for the AI infrastructure complex rather than a fundamental earnings upgrade.

Near term, this is unlikely to offset the valuation pressure created by higher real yields. SMCI's multiple is particularly vulnerable if long-duration technology derates, even with intact order growth; a 100-200bp EV/sales compression can overwhelm a strong quarterly beat. Over 1-3 months, hyperscaler capex guidance, Nvidia supply lead times, and evidence that enterprise deployments convert from pilot projects to production are the relevant catalysts—not political rhetoric.

The consensus may underweight the geopolitical asymmetry: a less restrictive domestic posture can invite more targeted China export controls rather than broad U.S. AI regulation. That outcome would favor domestic software monetizers such as APP over hardware firms with greater exposure to constrained components, cross-border server demand, and inventory-cycle risk. The thesis is falsified if U.S. hyperscalers cut 2027 capex plans, SMCI reports rising inventory or lower gross margin, or legislative language shifts from voluntary standards toward binding compute, model, or data-center restrictions.

Key Decisions for Investors

• No standalone directional trade on this policy commentary; treat it as a modest reduction in regulatory-tail-risk rather than an earnings catalyst.
• For a 1-3 month relative-value expression, prefer long APP / short SMCI in equal dollar amounts. APP has a more software-like path to AI monetization, while SMCI carries greater duration, margin and supply-chain sensitivity; reassess if SMCI raises full-year gross-margin guidance or APP's ad-revenue growth decelerates materially.
• Maintain a watch alert on hyperscaler earnings and capex revisions: a broad upward revision to Microsoft, Alphabet, Amazon and Meta data-center spending would justify reversing the pair toward long SMCI, provided SMCI inventory days and gross margin remain stable.
• For existing SMCI longs, use a defined-risk hedge rather than adding on the headline: buy 1-3 month downside puts or reduce exposure if Treasury yields continue rising and SMCI underperforms SOXX by more than 10% without a company-specific order-book update.