Data Center Spending: The Inflation Flow and Market Reaction

The monetary link from data center spending to inflation is now quantified. The scale of investment is staggering, with data center deals cresting above $61 billion in 2025. This massive capex surge is directly pressuring electricity markets and, by extension, consumer prices.

Goldman Sachs forecasts a sharp jump in consumer electricity inflation, predicting it will jump 6% from 2026 to 2027. The bank warns this will translate into broader economic pressure, as higher business power costs get passed down to consumers. This creates a measurable inflationary channel, with Goldman estimating the impact will boost core inflation by 0.1% in both 2026 and 2027.

This isn't just a theoretical model. A new study from the Federal Reserve Bank of Dallas confirms the real-world effect, finding that data center expansion could increase electricity prices and contribute modestly to inflation. The research estimates this could raise annual PCE inflation by between 0.04 and 0.13 percentage points by 2030. The mechanism is clear: the sudden, massive demand from data centers forces utilities to build costly new infrastructure, a cost that gets baked into electricity rates for everyone.

The Spending Numbers: Scale and Market Skepticism

The scale of the AI buildout is now undeniable, with spending consistently outpacing even upwardly revised estimates. Analyst consensus for 2025 AI capital expenditure has been repeatedly underestimated, with actual spending on the ground exceeding 50% growth against a projected 20%. This divergence highlights the difficulty in forecasting the true pace of the infrastructure race.

The financial commitment is set to accelerate further. A forecast shows that four mega-cap tech firms are slated to spend a combined $625 billion or more on AI infrastructure this year. This massive outlay underscores the feverish competition to dominate the AI market, but it also sets up a critical test for profitability and investor patience.

Market skepticism is already evident. The sheer size of the spending plans has not translated into universal confidence. Just last week, shares of Microsoft plummeted 11% in a day, its steepest single-day drop since 2020. The catalyst was a report of slowing revenue growth from its key Azure cloud unit, which reflects AI demand, juxtaposed against accelerated data center spending plans. This sharp reaction illustrates that investors are now demanding a clearer, sooner link between this colossal capex and tangible financial returns.

Catalysts and Risks: What to Watch

The path forward hinges on three key signals that will confirm or challenge the inflationary and financial sustainability narratives.

First, watch the EIA data on retail electricity rates, which have risen by more than 5% in the past year. This is the primary transmission channel for data center inflation. Confirming a sustained climb, especially in major tech hubs, will validate the Goldman Sachs forecast of a 6% jump in consumer power costs. Any deviation from this trend would be a major positive surprise for the broader economy.

Second, monitor the divergence in AI stock performance. The market is already rotating, as seen in Microsoft's sharp drop last week. The catalyst is clear: investors are penalizing companies with massive, debt-funded capex plans while rewarding those that benefit from the productivity gains. A sustained rotation away from pure infrastructure spenders toward software and services beneficiaries will signal that the financial sustainability of the buildout is under scrutiny.

Third, note the Fed's stance, as signaled by Powell's comments. His acknowledgment that the data center boom is "probably pushing inflation up at the margin" is a key risk. If inflation persists, it could delay the rate cuts that markets are pricing in, despite AI's long-term productivity promise. The Fed's patience will be tested as it balances near-term price pressure against the potential for a future supply-side boom.