AI Is 'Huge Factor' In Success of US Economy, Says Stephen Moore

Former Trump economic adviser Stephen Moore said AI investment and the data-center construction boom are significant contributors to the US economy. He warned that elevated oil and diesel prices can lift broader costs and inflation, but remained bullish on the overall US economic outlook.

Analysis

The investable issue is not broad AI optimism but whether data-center capex remains sufficiently concentrated to offset rising power, diesel, and construction-input costs. Hyperscalers can absorb higher energy bills, while leveraged data-center developers and smaller cloud/GPU renters face a margin squeeze if utility interconnection delays force reliance on expensive temporary generation. Near term, this favors the AI infrastructure stack with contracted demand and pass-through economics—power equipment, grid hardware, cooling, and select utilities—over application software whose multiples already assume durable enterprise monetization.

Over the next 1-3 months, inflation persistence from energy and freight is more likely to pressure long-duration technology valuations than AI capex budgets. A higher-for-longer rate repricing would create a useful dispersion trade: companies funding capacity internally retain strategic flexibility, while highly levered digital-infrastructure names become vulnerable to refinancing and project-return compression. The key falsifier is a meaningful downward revision to hyperscaler capex guidance or evidence that power constraints defer server deployments rather than merely shift spending into grid equipment.

The consensus may underappreciate that power availability—not chip supply—is becoming the binding constraint over the next 6-18 months. That shifts incremental economics from GPU vendors toward electrical equipment suppliers and regulated utilities in data-center corridors; it also makes diesel-price spikes a second-order positive for grid modernization, while reducing the attractiveness of diesel-dependent backup-power economics. This is a structural theme, but the cited commentary itself is not a stand-alone catalyst and does not justify chasing broad AI beta.

Key Decisions for Investors

• Prefer a 6-12 month basket long ETN / PWR / VRT versus short IGV: electrification and thermal-management spending has more visible data-center capex linkage, while software is more exposed to rate-driven multiple compression. Reassess if hyperscaler capex guidance falls by more than 10% year over year or 10-year Treasury yields decline materially enough to reverse duration pressure.
• Maintain a selective long in CEG or VST only on pullbacks, sized for regulatory and power-price volatility: data-center load growth can tighten regional capacity, but valuation discipline matters after AI-related rerating. A state-level rate-case setback, new nuclear outage, or sustained decline in forward power curves would invalidate the setup.
• Avoid adding to leveraged data-center operators solely on AI demand headlines until project-level power contracts, lease precommitments, and funding sources are verified. Set an alert around quarterly interest expense and development yield guidance; widening debt spreads or reduced stabilized-yield targets would be a short/watch signal rather than a long entry.
• Use elevated oil/diesel prices as a macro hedge trigger rather than an outright AI signal: modestly overweight XLE versus QQQ for the next 1-3 months if inflation breakevens and freight rates accelerate together. Exit the hedge if crude rolls over alongside easing core inflation, since that would restore the relative-duration bid for technology.