Morning Brief
Mortgage Market Snapshot
The 10-year Treasury broke through 4.75% this morning, its highest level since January 2025, and that is the number that matters more than anything else on your screen today. It closed Friday near 4.67%, so we are looking at an eight basis point move in a single session, driven almost entirely by crude. Oil has been grinding higher on the Iran supply disruption and the market is now pricing a real possibility that the Fed hikes rather than cuts. The 30-year fixed responded by ticking up to 6.68% from 6.66%, but notice the divergence underneath: the 15-year dropped to 5.92% and the 5/1 ARM fell to 6.26%. When the long end sells off harder than the short end, the fixed-to-ARM spread compresses and then reopens, and right now the ARM is buying your borrower 42 basis points against the 30-year. That is a materially wider discount than the 21 point spread we saw collapse to nothing earlier this month.
The macro setup going into September is genuinely unusual. July core PCE printed at 3.3% year over year with headline PCE at 3.7%, both stuck well above target and both stubbornly driven by services rather than the energy line, which means the Fed cannot dismiss it as a transitory oil pass-through. Fed funds futures have swung violently on this all month, from roughly 82% odds of a September hike at the peak of the oil scare down to about 40% after the cooler mid-August CPI, and today's crude move is pushing that probability back up. Meanwhile the labor market is doing the opposite thing entirely. July payrolls came in at negative 23,000 against an expected gain of 80,000, June was revised down to a 20,000 gain, and consensus for Friday's August print is a soft 90,000. A central bank staring at 3.3% core inflation and a stalling labor market has no clean move, and Chair Warsh made clear at Jackson Hole Friday that he is not going to pre-commit.
For brokers, the practical read is that the "wait for rates to drop" script your borrowers have been running since spring is now actively working against them. Nobody is forecasting a cut before December, and the hike scenario is live. On a $500,000 loan, the difference between locking a 30-year at 6.68% and locking a 5/1 ARM at 6.26% is roughly $138 a month, about $8,300 over the first five years. For a borrower who genuinely expects to sell or refinance inside that window, and especially for a self-employed or investor client who is already outside agency guidelines, the ARM math has never been cleaner. Non-QM pricing has held up better than agency through this selloff because the private-credit bid for those loans has stayed firm even as the Treasury curve backed up. That is a real competitive edge, and most loan officers have not caught up to it yet.