Fed’s Goolsbee: Strong demand may be adding to US inflation, ’no ambiguity’ how Fed would react

Chicago Fed President Austan Goolsbee said inflation, estimated at 3.7% in July versus the Fed's 2% target, may increasingly reflect strong demand rather than only temporary tariff, energy and commodity-price shocks. He said persistent supply shocks and AI-related investment could be boosting demand beyond the economy's capacity, leaving higher interest rates as the likely response if overheating is confirmed. The Fed raised its policy rate by 25bps last week, and Goolsbee warned that restoring 2% inflation could require slower growth and pose risks to employment.

Analysis

The market implication is a higher terminal-rate premium rather than simply one more hike: if AI-related capex is treated as demand stimulus, long-duration equities lose the benefit of being viewed as a disinflationary productivity trade. The near-term transmission runs through real yields and equity discount rates, with SMCI particularly exposed because its valuation and customer order cadence both rely on hyperscaler capex remaining aggressive despite a higher cost of capital. APP is less directly rate-sensitive operationally, but its multiple is vulnerable if advertisers tighten budgets or if high-beta software/adtech de-rates alongside semiconductors.

The non-obvious second-order issue is that persistent inflation driven by investment demand can crowd out smaller borrowers before it restrains megacap AI spending. That is unfavorable for regional-bank credit quality: FISI may initially benefit from asset-yield repricing, but any extended restrictive period raises commercial real estate and small-business loss risk faster than deposit economics improve. Watch the 2-year Treasury yield and high-yield spreads rather than headline equity indexes; a rise in both would signal the market is shifting from a valuation problem to a growth-and-credit problem.

Consensus still appears inclined to interpret AI spending as inherently margin-accretive productivity investment. The contrary view is that the investment cycle may be inflationary before it is productive, particularly where data-center power, networking, and construction capacity are constrained; that would delay any earnings payoff while lifting financing costs. This thesis is falsified by successive core-inflation downside surprises, a sustained decline in the 2-year yield, or hyperscaler guidance showing AI capex can be funded without incremental debt or margin pressure.

Key Decisions for Investors

• Trim or hedge SMCI into strength over the next 1-3 months; use a short SMCI versus long SOXX basket only if SMCI's valuation premium remains elevated. Target 10-15% relative downside if real yields continue higher; cover if hyperscaler capex guidance accelerates materially or the 2-year yield falls below its post-meeting level.
• Avoid adding to APP solely on AI narrative momentum until advertising demand and forward EBITDA revisions demonstrate resilience to tighter financial conditions. A tactical short is not warranted without evidence of estimate cuts; set an alert for deteriorating digital-ad spend commentary or a 10%+ decline in forward EBITDA consensus.
• Express the macro view through long IEF or short-duration underweight only after confirmation from the next core inflation release; absent confirmation, the cleaner hedge is a modest long IEF/short IWM pair, which benefits if restrictive policy shifts from multiple compression to domestic-demand slowing. Reassess if high-yield spreads widen more than 75 bps, which would increase recession risk and favor duration outright.
• Keep FISI neutral: do not treat higher rates as automatically positive for regional-bank earnings. Upgrade only if deposit costs stabilize and criticized-loan/CRE disclosures remain contained; reduce exposure if credit provisions rise faster than net interest income at the next earnings update.