Goldman Sachs pushes Fed rate-cut call to 2027 on strong US jobs data

The bigger implication is not “higher for longer” in rates, but a regime shift in cross-asset pricing: if the Fed stays pinned through 2026, duration stops being the marginal macro hedge and inflation-sensitive cash flows regain a bid. That is a headwind for long-duration equities and a tailwind for banks, insurers, and value sectors that can live with sticky front-end yields, while rate-sensitive growth multiple expansion becomes harder to justify unless earnings revisions re-accelerate materially.

The AI angle matters because it creates a second-order squeeze on the equity beta complex: if a meaningful share of capex and cloud spend is being priced on assumed perpetual demand, any slowdown in that incremental spend path can hit semis, data-center supply chain, and the “picks-and-shovels” beneficiaries before it shows up in headline AI revenues. In other words, the market is vulnerable not just to a higher discount rate, but to a slower terminal growth assumption for the AI trade; that combination tends to compress the highest-multiple names first.

Geopolitically, the market may be underestimating how much oil-driven inflation can be tolerated before policy reaction becomes the dominant shock absorber. The more relevant risk window is 1-3 months: if energy prices stay elevated into summer, breakevens and wage expectations can re-tighten even with stable growth, forcing the Fed to lean harder against easing expectations and keeping real rates elevated. That is usually constructive for the dollar and negative for EM beta, especially where external balances are fragile.

The contrarian read is that a prolonged pause is already partly consensus, but the market still treats any eventual cut as a near-term rescue for risk assets. If cuts are pushed out to 2027, the more important question is whether earnings need to do all the work in 2025-26; absent broadening profit growth, index leadership should narrow further and crowded secular growth names could de-rate faster than macro data deteriorates.