The Overshoot

The Overshoot

The Growth Impulse from the Data Center Boom

There is much less than meets the eye, although the official numbers may also be undercounting some of what is happening.

Matthew C. Klein's avatar
Matthew C. Klein
May 02, 2026
∙ Paid

It is convenient that Amazon, Google, Microsoft, and Meta all reported their 2026Q1 results the night before the Bureau of Economic Analysis (BEA) published its initial estimate of U.S. Gross Domestic Product (GDP) in January-March 2026. Those four providers of cloud computing services, plus Oracle, which reported slightly earlier, together reported about $150 billion in capital spending in 2026Q1. That capex was equivalent to about 2% of the value of all of the goods and services produced within the U.S. in the first three months of 2026.

As recently as 2025Q1, total capex by the big five cloud companies was under $80 billion, or 1% of U.S. GDP. Put another way, the growth in their combined capex from 2025Q1 to 2026Q1 would appear to have contributed about 1 percentage point to the total 6% increase in U.S. GDP (in dollar terms) over those 12 months. Go back slightly further and the big five were spending only about $40 billion/quarter, equivalent to roughly 0.5% of U.S. GDP. With the exception of Amazon, these companies were, until recently, known for being able to generate outsize returns without needing to invest much in physical assets. Now, thanks to their contributions to America’s data center buildout, they are the five largest capital spenders in the entire S&P 500.

All financial statements data come from Koyfin. I have been using them for years to get data on markets and companies. Readers of The Overshoot are eligible for discounted subscriptions or a free trial using this link.

This shift is big enough to affect the overall U.S. macro data.

But it is far less clear how much, if at all, this investment is actually contributing to aggregate economic growth.

This note is not going to get into speculative questions about what increasingly powerful software tools running on these data centers may eventually do to the economy. The answers I have seen range from “basically nothing” to “the end of scarcity” to “the end of human civilization”. While I enjoy good science fiction, I am more concerned with (what should be) the measurable impacts of the data center capex boom on the official statistics today. The big challenge is that it is difficult to line up the companies’ numbers with what the BEA publishes.

What the Official Numbers Show

This post is for paid subscribers

Already a paid subscriber? Sign in
© 2026 Matthew C. Klein · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture