Sen. Tim Scott wants to hear from Warsh on data centers and artificial intelligence

Senate Banking Committee Chair Tim Scott said he wants Fed Chair Kevin Warsh to address AI and data centers during Wednesday’s appearance, focusing on how they can “pay their own way” given electricity and water use. Scott noted that data center moratoriums are spreading as voters react to higher utility bills, and he framed the issue as part of a broader U.S.-China competition for AI leadership. The news is more policy/agenda-setting than a direct rate or earnings change, but it could influence regulatory and cost outlooks for data-center-heavy AI buildouts.

Analysis

This is less a thesis on AI demand and more a thesis on where the rent-seeking sits in the stack. If local politics forces data centers to internalize electricity and water costs, the immediate loser is the marginal build: colocation developers, data center REITs, and any hyperscaler relying on cheap grid access in constrained markets. The first-order hit will show up as slower leasing, longer payback periods, and a higher hurdle rate for new capacity in states where permits are already politically fragile.

The more durable winner set is the behind-the-meter ecosystem: power distribution gear, backup generation, thermal management, and grid hardening. If moratoriums spread, capex migrates from square footage to infrastructure, which is better for industrials with pricing power than for landlords with fixed-rent models. The second-order effect is that utilities may still see load growth, but their upside depends on whether regulators allow timely recovery; otherwise they become the political buffer between voters and AI economics.

The consensus may be missing that this does not necessarily cap AI adoption, it reallocates the economics. Large tech can absorb higher site-specific costs or shift deployment to more permissive jurisdictions, so the national AI spend curve may survive even if local buildouts pause. The real catalyst is policy text: if state commissions or legislatures hard-code cost allocation, this becomes a months-long valuation reset; if not, the headline fade is likely fast.

Key Decisions for Investors

• Long ETN / short DLR as a 3-6 month pair trade: benefit from a shift in spending toward electrical infrastructure while data center lease growth faces permitting friction; stop if DLR reports no slowdown in pipeline conversion or if utilities rapidly approve interconnects.
• Long GEV or CMI on a 1-3 month horizon: any move toward on-site generation, backup power, and microgrids should lift order books; risk/reward improves if state moratorium headlines proliferate beyond one region.
• Avoid initiating fresh long exposure in EQIX/DLR until after the next round of state policy decisions: the tradeable risk is regulatory sequencing, not AI demand destruction, and a quick policy compromise could reverse the move within weeks.
• Watch-list only: if XLU weakens while power-equipment names outperform, that confirms the market is pricing political backlash rather than pure load growth; use that spread as a confirmation signal before adding size.