President Donald Trump Just Admonished the Fed for "Doing the Wrong Thing" on Interest Rates, but Ignores the 2 Biggest Inflation Catalysts

The Fed raised the federal-funds target range 25bps to 3.75%-4.00% on Sept. 16, its fourth tightening cycle this century, and major U.S. equity indexes declined following the decision. The article attributes persistent inflation—now above the Fed's 2% target for 66 consecutive months—to renewed tariffs on more than 80 countries, energy disruption from the Iran war and Strait of Hormuz closure, and AI hardware shortages that have increased downstream pricing pressure. These entrenched inflation drivers reduce the likelihood that the FOMC will accommodate President Trump's calls for rates of 1% or lower.

Analysis

The investable implication is not simply "long AI": a higher-for-longer discount rate selectively transfers value from AI infrastructure owners to component vendors with near-term pricing power and low capital intensity. NVDA, MU and AVGO can protect gross margins while hyperscalers and data-center operators absorb accelerated depreciation, power costs and financing expense; the latter risk is most acute for highly levered expansion models such as DLR and EQIX. Over the next 1-3 months, any upward revision to capex budgets without a matching monetization signal should widen this supplier/customer performance gap.

The second-order beneficiary is dispatchable power rather than generic utilities. CEG and VST have greater earnings sensitivity to tight power markets than regulated utilities, while electrical equipment suppliers ETN and VRT can retain backlog-driven pricing even if server lead times normalize. The structural risk over 6-18 months is that sustained rates constrain data-center project financing and force customers to rationalize GPU orders, converting today’s supply scarcity into an inventory correction; this would hit NVDA and MU multiples before reported revenue.

Consensus may be overstating the direct inflationary effect of AI hardware. Hardware purchases are largely corporate capex, and their durable-goods price effect is smaller than their indirect effect through electricity, construction labor and credit demand. A softer payroll or core-services inflation print could therefore produce a sharp duration-led relief rally in software and hyperscalers, even if AI supply remains constrained; the cleaner expression is relative rather than an outright broad-market short.

Key Decisions for Investors

• Initiate a 1-3 month pair: long NVDA and MU equally weighted versus short DLR and EQIX equally weighted. Thesis is margin capture by chip suppliers versus financing and power-cost pressure at infrastructure owners; target 10-15% relative return, exit if hyperscaler capex guidance is cut or 10-year Treasury yields fall below the pre-hike level for two consecutive weeks.
• Buy CEG or VST on pullbacks for a 6-12 month power-tightness exposure, paired with a smaller short in XLU to isolate merchant-power upside from rate sensitivity. Reduce if forward power prices retreat more than 15% or announced data-center load commitments fail to convert into signed interconnection contracts.
• Avoid chasing a broad AI beta rally through QQQ. Prefer ETN and VRT for 6-18 month exposure to grid and cooling bottlenecks, but require backlog growth and stable gross-margin guidance at the next earnings prints; a material backlog cancellation rate or working-capital build would invalidate the supply-constraint thesis.
• Use 3-month QQQ put spreads only as a hedge around inflation and labor releases, not as a standalone bearish position. The risk/reward turns favorable if real yields re-accelerate, but a benign core-services print would likely trigger rapid multiple expansion in long-duration technology.