Morgan Stanley's Andrew Sheets warns that surging AI infrastructure spending is breaking the US economy's traditional response to price increases.

Copper prices are up 40% year-over-year, gas turbines have climbed 50%, and memory chips have surged as much as 300%. Yet demand keeps rising, a pattern Sheets describes as "uniquely price insensitive."

The firm now projects $800 billion in AI-related capital expenditure from major US tech firms in 2026, nearly double previous estimates and triple 2024 levels. The 2027 forecast climbs to $1.1 trillion.

This is a "classic good news, bad news story," Sheets says. The massive spending props up GDP growth and employment. But when a huge sector ignores price signals, inflation becomes harder to tame, and credit markets face pressure from heavy debt issuance.

For fixed income investors, the inflationary impulse could keep interest rates elevated longer than expected, potentially breaking the Fed's traditional transmission mechanism.