Fed’s Tom Barkin, a former McKinsey CFO, says the ‘AI Apocalypse’ hasn’t arrived

Richmond Fed President Tom Barkin said the Fed raised interest rates for the first time since mid-2023 last week as inflation has remained persistent. He cited tariff costs, higher gasoline prices and AI-driven demand for technology equipment as continuing price pressures, implying rates may need to stay restrictive for longer. Barkin also said the AI investment boom is supporting productivity and corporate earnings while making firms more cautious on hiring, though widespread AI-driven layoffs have not yet materialized.

Analysis

The investable implication is a higher-for-longer term-premium regime rather than a one-meeting policy shock. Persistent capex-driven demand and tariff/input-cost pass-through raise the risk that nominal growth stays firm while disinflation stalls; that combination is unfavorable for long-duration software and unprofitable AI infrastructure beneficiaries, whose valuations require falling discount rates. The more resilient AI exposures are cash-generative hyperscalers (MSFT, GOOGL, AMZN) and equipment vendors with contracted demand, but even these face multiple compression if the 10-year Treasury yield reprices higher over the next 1-3 months.

AI's near-term labor effect is more likely margin expansion through slower hiring and attrition than an immediate unemployment spike. This favors large enterprises with high SG&A intensity and implementation capacity—particularly IT services buyers, insurers, and selected industrial distributors—while creating a medium-term headwind for labor-arbitrage business models such as call-center outsourcing and lower-value IT services (GENPACT, WIT, INFY). The key distinction is that productivity gains may initially sustain output and capex rather than reduce payroll, limiting the near-term recessionary case that would normally support duration.

USAR has narrative leverage to domestic supply-chain spending, but the monetary-policy signal is not a fundamental catalyst for its cash flows. Higher real rates are especially punitive to pre-scale, capital-intensive critical-minerals projects because financing costs and equity dilution rise before operating leverage materializes. Treat any rate-driven strength in USAR as an opportunity to reduce exposure unless it is accompanied by independently verifiable milestones: binding offtake, project financing, permits, and a credible path to commercial production.

Key Decisions for Investors

• Initiate a 1-3 month relative-value hedge: long XLF / short IGV in equal dollar amounts. Banks and insurers benefit from nominal-rate persistence and reinvestment yields, while software duration remains vulnerable; exit if the 10-year Treasury yield falls below its pre-hike level or if core inflation prints materially softer for two consecutive months.
• Reduce or hedge speculative AI and pre-revenue critical-minerals exposure, including USAR, rather than shorting profitable semiconductor leaders outright. For USAR, require disclosed financing terms and contracted revenue/offtake before adding; a dilutive capital raise or delayed project milestone is the principal downside catalyst over 6-18 months.
• Prefer MSFT and GOOGL over labor-arbitrage IT services (short basket WIT/INFY or ETF proxy) over 6-12 months. The thesis is operating leverage from internal deployment versus pricing and seat-growth pressure at service vendors; invalidate if service-company bookings accelerate while enterprise AI capex decelerates.
• Use TLT puts or a modest short IEF as a 1-3 month portfolio hedge against another upward repricing of policy expectations. Risk/reward deteriorates if payrolls weaken sharply or a downside inflation surprise restores a credible easing path; size as protection, not a standalone conviction short.