The 10-year Treasury yield reached 4.76% intraday, its highest level against any close since 2025-01-14 — but by one basis point, and the reasoning first offered for it was wrong.
The long end is the discount rate for the AI buildout, and it is taking two pushes at once from sources this map treats separately: an oil shock raises the inflation path, a hawkish Fed repricing raises the policy path, and both land on the same tenor. Hormuz reaches datacenter economics through the bond market, not the electricity price.
Scenarios 2
The move fails at the July level and mean-reverts, because it is an oil-driven inflation scare rather than a term-premium repricing. Buildout financing costs are unaffected.
The reading the data currently favours, and cheaply falsifiable. The 10-year closed at exactly 4.75% on 2026-07-31 and has now printed 4.76% intraday - one basis point over a level tested five weeks ago, not a breakout. Watch whether the official close confirms the print, and whether the 30-year spread widens or holds near 49bp.
Precedent — The 10-year's own 2025 path: 4.79% on 2025-01-13, then the rest of the year well below it, the level never becoming a floor. A high that did not hold is the direct precedent for reading this as a retest.
The two pressures compound rather than cancel, the long end breaks above the January-2025 range, and it prices into new AI-infrastructure debt before it prices into equities.
Today's divergence is the tell: equities sat frozen for four hours while the long end kept moving and chips outperformed a down tape. That is a rate event, not an equity-risk event. Falsifiable through the coupon on a named raise - this map already holds IREN's 9.0% tranche against its own 6.0% facility as the comparator.
Precedent — The 1979-81 oil-and-policy overlap, when an energy shock and a hawkish Fed pushed the same tenor together and long rates broke a multi-year range instead of mean-reverting. Same shape, different magnitude.
Context
Read against capital-context.yaml's standing picture, in which AI capex is increasingly long-end debt-financed. Caveat: this is an intraday quote, Treasury's official 08-31 close is not out, and this thread already revised one intraday snapshot (08-28) two days later.
This is generated reasoning, not a sourced fact — the mechanism and scenarios above are the model's read on what this item could mean for capital flows, tagged with its own confidence rather than stated as settled. Back to Global Capital.