Neil Dutta on How Economists Are Missing the Macro Impact of AI

QTS’s data-center complex in Fayetteville, Georgia is expected to consume as much electricity as about 1 million US households, forcing Georgia Power to rush infrastructure buildout to meet the demand. The piece highlights the scale of power requirements tied to data-center expansion, with implications for utilities and local infrastructure. No financial results or policy changes are reported.

Analysis

BX is not trading a one-off project here; it is effectively underwriting the buildout of a power-constrained digital infrastructure corridor. The second-order winner is anyone controlling grid interconnects, transformers, switchgear, and gas-backed generation capacity, because the bottleneck is no longer land or entitlement but electrons delivered on schedule. That creates a multi-year revenue annuity for the infrastructure supply chain, while also raising the cost of capital for new entrants that cannot self-fund power procurement or secure utility priority.

The market is likely underappreciating the timing mismatch between data-center demand and utility response. Construction revenues can show up quickly, but the true earnings leverage comes later when tenants turn on loads and power pricing resets higher; that means the immediate trade is more about infrastructure inflation than hyperscale cloud demand. In the near term, local utilities and regulated grid equipment vendors should benefit, but over 6-18 months the bigger impact may be margin pressure on adjacent industrial users facing higher curtailment risk and less favorable interconnection queues.

For BX, this is mildly positive but not a clean direct catalyst: the asset is attractive if power availability becomes a scarce option value, yet any slip in permitting, transmission upgrades, or utility capex can push cash conversion further out by quarters. The contrarian miss is that “data-center scarcity” can be overbought if financing costs stay elevated—higher rates hurt the long-duration infrastructure multiple more than the near-term build narrative, especially if tenants demand better economics before signing take-or-pay contracts.

The key reversal catalyst is a policy or utility shock: slower interconnect approvals, stricter local permitting, or evidence that power costs erase the expected IRR uplift. Over 3-12 months, watch for commentary from utilities and grid suppliers on backlog growth, because that will validate whether this is a one-site headline or the start of a broader capex cycle.

Key Decisions for Investors

• Long BX on a 3-6 month horizon as a leveraged beneficiary of infrastructure scarcity; target is modest upside driven by sentiment and fundraising, but keep size small because the fundamental re-rating is indirect and rate-sensitive.
• Pair trade: long infrastructure enablers (utility grid equipment / electrical capex beneficiaries) vs. short a basket of rate-sensitive REITs with exposed financing costs; expect 6-12 month divergence if data-center power buildout keeps accelerating.
• Buy call spreads on industrial electrical suppliers with backlog leverage for 6-9 months; risk/reward improves if utility capex guidance inflects upward and interconnect queues remain tight.
• Avoid chasing the obvious hyperscaler trade; the cleaner expression is the power bottleneck, not the data-center operator itself. If power-delivery timelines slip, rotate out of infrastructure beta quickly because the market will discount the build cycle by quarters.