Jerome Powell says you’re right to blame data centers for making your bills more expensive: ‘probably pushing inflation up’

Fed Chair Jerome Powell said the AI-fueled data center buildout is adding upward pressure on inflation and likely raises the neutral interest rate in the near term, even as the Fed held rates steady. Officials revised long-run growth up from 1.8% to 2.0%; Powell said disinflationary benefits of AI remain theoretical. Goldman Sachs warned consumer electricity prices could rise ~6% from 2026-27 and utilities have sought a record $31B in 2025 rate increases; Wood Mackenzie found only ~1/3 of data-center pipeline is in active development due to grid limits.

Analysis

Rapid buildout of AI capacity creates concentrated upstream demand for a handful of long‑lead items — high‑voltage transformers, cryogenic cooling systems, copper conductors and heavy civil work — which transmits into a sectoral inflation pulse rather than broad consumer disinflation. Expect price passthrough and margin pressure to show up in supplier P&Ls within 6–18 months (lead times) and in regulated rate filings over 12–36 months as utilities push to socialize grid upgrade costs.

Regulatory responses will reallocate returns: well‑structured rate cases that expand regulated asset bases are a tailwind for utilities and their credit profiles, while regions with slow permitting or constrained RTO capacity will see project cancellations and stranded capex for colo operators. That bifurcation magnifies counterparty and regional concentration risk in corporate procurement chains and in muni issuance used to fund transmission builds.

On policy, a sustained mismatch between demand for physical capacity and the pace of supply expansion is likely to raise neutral rates and term premia, not lower them. The practical consequence: higher real yields compress long‑duration tech valuations and increase the cost of financing for growth capex — a 50–75bp upward revision to terminal rate expectations would meaningfully widen equity dispersion across capital intensive vs. asset‑light models.

Actionability hinges on three signals to watch: transformer/cable lead times and spot premiums, utility rate‑case outcomes in state commissions, and regional RTO congestion prices. Use pair trades and duration hedges to isolate exposure — liquidity and basis in the power and equipment suppliers will widen in stressed scenarios, so size positions with optionality rather than outright leverage.

Key Decisions for Investors

• Long Eaton (ETN) — 12–18 month horizon. Rationale: exposure to power distribution and transformer demand; target +30% if grid capex accelerates. Positioning: buy shares or a 12‑month call spread (buy 1x ATM, sell 1x+25%); downside risk ~‑15% if capex is deferred.
• Long ABB ADR (ABB) or Siemens ADR (SIEGY) — 9–12 month horizon. Rationale: makers of high‑voltage gear and substation equipment benefiting from multi‑year replacement cycles. Positioning: size as core long with 6–9 month covered calls to fund carry; expect 20–35% upside if win rates on large T&D contracts rise.
• Pair trade — Long AES (AES) / Short Equinix (EQIX) — 6–12 month horizon. Rationale: AES gains from storage and behind‑the‑meter solutions as colo customers seek resilience; EQIX faces higher incremental power cost risk and project delays. Positioning: equal notional sizes, target asymmetric payoff ~2:1; hedge sector beta with XLK or a cloud infra ETF.
• Macro hedge — Short long‑duration risk via TLT or enter a 2s10s steepener (futures/swaps) — 3–12 month horizon. Rationale: if neutral rate repricing persists, long‑duration assets will reprice negatively. Positioning: small tactical size to cap drawdown if Fed pivots; take profits if 10y yields fall >50bp from entry.