Most economic conversations about artificial intelligence centre on its potential to transform productivity or on doubts about the durability of its growth. Less attention is paid to the sheer scale of investment now pouring into AI infrastructure, and the economic effects that follow.  

Greg Ip of the Wall Street Journal points out that the AI goldrush is transforming tech companies’ business model. In recent years, tech firms have been notably ‘asset light’, driving profits on intellectual property and software.  On this model, “revenue required little in the way of more buildings and equipment, making them cash-generating machines.”

By contrast, tech companies are now spending massively in AI hardware, including cutting-edge GPU chips and servers. This spending has helped boost the economy as consumer demand softened.

The obvious challenge is generating sufficient profits to sustain such a high level of investment, at least in the medium term. As Jason Thomas, head of research at Carlyle Group, told Ip: “All this capital spending may prove productive beyond their wildest dreams, but beyond the relevant time horizon for their shareholders”.

The cost of capital

But, as Ip points out, all that spending also has consequences for capital markets.

Because tech companies were generating so much cash, and spending relatively little on investment, that extra cash went back into the financial system, helping moderate long-term interest rates.

By contrast, companies now face the added cost of tariffs plus pressure to invest heavily in AI infrastructure. All while, “government deficits are even larger, inflation is above 2% and the Fed has been shrinking its bondholdings.”

In his recent Carlyle Compass note, Thomas expands on another tension: housing. Higher mortgage rates have produced a “rate-induced affordability crisis.” A rate cut could help relieve housing, but AI spending itself absorbs “hundreds of billions of dollars of capital each quarter, compounding at an annualized rate of 40–60%, while deficits consume 1.5× as much private savings as in the 2010s”. Against this backdrop, Thomas asks provocatively: “How much crowding out is necessary to meet price-stability targets”?

Now wonder that, as Reuters’ Mike Dolan suggests, “tension between the Federal Reserve’s jobs and inflation mandates may be less worrisome than a dilemma over whether to focus on the spluttering housing market or rocketing tech infrastructure spending.”

All opinions, news, research, analysis, prices or other information is provided as general market commentary and not as investment advice and all potential results discussed are not guaranteed to be achieved. The information may have been derived from publicly available sources, company reports, personal research, or surveys. Past performance is not indicative of future performance. Trading carries risk of capital loss. Service available to professional clients only.