Today is genuinely quiet on the mortgage news front — a Sunday with no prints, no agency bulletins, and a thin corporate calendar. The one thing worth your reading time is Logan Mohtashami's weekend piece at HousingWire arguing that mortgage spreads, now running near 2.01%, are the reason the retail 30-year is still carrying a 6-handle. His point is structural: at the 10-year's current level, the spreads of 2023 and 2024 would have put the national 30-year above 7% months ago. Instead, weekly pending sales eked out a small year-over-year gain for the second straight week, and his purchase-application data softens only once rates clear 6.64%. The national 30-year is 6.76% today, unchanged from Saturday.
Saturday's brief made the observation that Friday's contracting payroll report — down 23,000 against an 80,000 consensus, another 103,000 revised away, wage growth cooling to 3.2% — bought roughly three basis points in the 10-year and a quarter point in MBS, and that the retail rate sheet gave the improvement straight back. Nothing over the weekend changed that. The 30-year is 6.76% for a second day, three basis points higher on the week and twenty higher on the month. The 10-year's last published close remains August 6 at 4.69%; Friday's session has not landed in the series yet.
Put those two observations next to each other and you have the actual state of the market. The bond market paid almost nothing for a labor contraction, which says the bid is not being set by growth expectations. The retail rate then failed to follow even that small improvement, which says the retail level is not being set by the 10-year either — it is being set by the spread. That is a comfortable place to be right up until it isn't: spread compression is what is currently buying your borrowers roughly half a point of rate, and it can widen back out on nothing more than volatility. We are also already twelve basis points above the 6.64% level where Mohtashami's purchase applications start to soften, which is worth remembering the next time a purchase file stalls and everyone blames the listing.
For the week ahead, CPI is the only thing that matters. The release window runs August 10 through 15. Jobless claims land August 13, Freddie Mac's weekly survey the same day, and housing starts and permits August 16 through 18. The July FOMC meeting on the 28th and 29th is behind us; the next one is September 15–16 and it carries a Summary of Economic Projections, so between now and mid-September the market gets one inflation print, one jobs report, and then a dot plot to reprice against. On positioning: at 6.76% you are six basis points under the 90-day high of 6.82% and thirty-one above the 90-day low of 6.45%, with a 30-day average of 6.70%. That is the rich end of the range, not the cheap end, and it argues for a lock bias on anything closing inside thirty days. The gov-loan spread is where the real optionality sits right now — FHA at 6.28% and VA at 6.30% against conventional at 6.76%, with the VA number worth about $121 a month on a $400,000 note and carrying no monthly mortgage insurance. Jumbo-to-conforming is nine basis points, unusually tight.
On the industry and regulatory side: HUD published a second notice for its non-vacant loan sale HNVLS 2026-1, covering HECM loans secured by occupied properties, with an estimated bid date of September 1, 2026 after the original January-announced sale was delayed. Treasury finalized its backup-withholding regulations for third-party network transactions, which touches you only when you are documenting 1099-K income on a self-employed file. RE/MAX reported second-quarter revenue down 5.8% with U.S. agent count at a two-year low, ahead of the August 14 shareholder vote on the sale to The Real Brokerage. Mortgage Daily ran a genuinely useful desk-level explainer on how FHA, VA, and conventional each treat a $0 income-driven student-loan payment in DTI — that difference decides borderline pre-approvals more often than most originators realize. And fast-moving wildfires through northwest Spokane over the weekend destroyed hundreds of buildings; if you write in eastern Washington, expect insurance-binder and appraisal-condition problems on files in that corridor this week.
Things you may have missed this week: Treasury published a final rule on August 3 rescinding portions of its Title VI regulations to eliminate disparate-impact liability, the companion action to HUD's supplemental proposed rule on the Fair Housing Act disparate-impact standard. Both actions narrow the regulatory codification of that theory; neither changes anything in your origination or disclosure workflow today, and the Fair Housing Act and ECOA apply to your files exactly as they did last month. Zillow's second quarter showed revenue up 18% to $772 million with mortgage revenue up 75% to $84 million — their mortgage arm is growing four times faster than the rest of the company, which is a competitive fact rather than a headline. A House subcommittee is seeking details on the Compass and MRED private-listing arrangement. And a brokerage AI-adoption report found the adoption gap between mid-size and large firms has closed entirely. A quick 72-hour recap for anyone who was off this weekend: Friday's negative payroll print was the week's only real market event and it moved the tape about three basis points, Freddie Mac's weekly survey printed a 2026 high of 6.69%, and HUD's supplemental proposed rule on disparate impact hit the Federal Register. That is the whole board.
pull every in-flight file closing in the next thirty days, confirm lock status on each one, and flag any VA-eligible borrower who was quoted conventional — you are six basis points off the 90-day high with an inflation print landing inside the next five business days.