Fed hikes aim at the inflation that keeps US nominal growth above a 5.16% yield

Take the roughly 2% of real growth out of nominal growth running well above 6%, and about 4 points remain ^[2][2]. Those 4 points are price growth. They are most of the reason the economy still outruns a 10-year yield of 5.16% ^[3], by at least 0.84 of a point ^[1]. Fed policymakers raised rates in September to rein in inflation ^[1]. That means the central bank is working to shrink the part of nominal growth that keeps the economy ahead of the Treasury's borrowing cost ^[1][2].

The debt side is simpler. On Fortune's round figures, a $2 trillion annual deficit ^[5] on $40 trillion of debt ^[4] adds about 5% a year to the stock ^[3]. With nominal GDP growing faster than 6%, the economy is outgrowing its debt by a point or more a year ^[8]. The Committee for a Responsible Federal Budget expects that to reverse. "With interest rates on new Treasury bonds and notes at around 5% and medium-term nominal economic growth expected to be closer to 4%, the U.S. is entering a debt spiral," CRFB said ^[6]. At 4% growth, 5% debt growth runs a point ahead of the economy ^[4], and that is before higher yields add to the servicing bill ^[4].

A smaller deficit would also mean less growth, on Research Affiliates' reading. The firm argued early this year that much of the borrowed money reaches consumers through entitlement payments and ends up in profits and stock valuations ^[12]. So the $2 trillion also feeds the nominal growth the debt is measured against.

The paths split on what yields do when growth cools. If AI spending holds up, nominal growth may stay high. Capex at Alphabet, Amazon, Microsoft, Meta, Oracle and SpaceX is projected at $870 billion this year against $470 billion in 2025 ^[7], an increase of $400 billion, or 85% ^[6]. S&P Global expects hyperscaler spending above $1.3 trillion in 2027 ^[8]. "The demand impulse from AI appears to be spilling over to help create demand for capex outside of tech," UBS economist Jonathan Pingle wrote ^[13]. If the Fed gets its disinflation instead, nominal growth drifts toward CRFB's 4% while yields stay near 5%. Fortune reports that other indebted governments and AI hyperscalers are competing for the same bond buyers, and that geopolitical shocks are being priced into yields ^[10]. The third path starts with an AI bust. Ruchir Sharma of Rockefeller International predicted the bubble could pop once the 10-year decisively exceeds 5%, signaling a "new era of tighter money, in which AI mega projects will be harder to fund" ^[9]. Growth would slow in that case, though Fortune notes that slower growth will not necessarily bring yields down ^[10].

I think the second path is the likeliest on this evidence. It supports treating a slowdown as the bigger fiscal risk, or rather a slowdown in nominal growth, where the disinflation the Fed wants counts as much as lost output. The counter-case is the third path, where a bust pulls yields down along with growth. It helps less than it seems. The 10-year has risen more than a point since the Iran war began ^[3], so before the war it was below about 4.16% ^[5]. Set against CRFB's 4% growth, even that yield leaves the two within a fraction of a point of each other ^[7].

The view is wrong if the 10-year falls well below its pre-war level while nominal growth holds near 4% or better, or if AI capex keeps nominal growth above 6% into 2027 ^[8]. For now, a September business-activity gauge at a five-year high ^[11] points to the second.