Mortgage Rates Hold Flat on Thursday Despite Lower Weekly Average

For the average lender, top-tier 30yr fixed mortgage rates were perfectly unchanged compared to yesterday. This keeps them right in line with the lowest levels in more than 3 years. That said, if we’re splitting hairs, better rates were available 4 days in the past month and a half (Jan 9, Jan 12, Feb 13, Feb 17). So why is it that there are news headlines today claiming that rates hit their lowest levels in more than 3 years? Simply put, those stories are based on weekly survey data from Freddie Mac. Freddie isn’t technically wrong, but you have to understand their methodology. Freddie’s survey is an average of the rates available from last Thursday through yesterday. Indeed, if you use the numbers from our daily rate index on those days, the average is the lowest in 3 years, even if today’s rates are a hair higher than several recent days.

Prudent AI’s Income, Correspondent, LOS Tools; The Fed’s Bank Mortgage Proposal; STRATMOR on LOs; Elliot Eisenberg Interview

It’s hard to believe that the COVID pandemic was six years ago with its increase in deaths. For our biz, we switched to “work from home,” among other changes. Along those lines, I received this note: “Rob, are you hearing about a concern among lenders about the liquidity of the non-QM sector? As in a repeat of March 2020 when many non-QM lenders suspended their programs due to volatile market conditions caused by the COVID-19 pandemic, hurting the ability to submit, underwrite, lock, or fund non-QM loans?” Personally, I don’t foresee that, but liquidity should always be discussion topic. In my 40+ capital markets years, I’ve never seen anyone hedge credit risk. Hedging a non-QM pipeline is a complicated question. If you think about the risk-free baseline, Treasury securities come to mind. For Agency loans, let’s say QM loans, there’s a spread above Treasury securities based on credit risk and prepayment risk. For non-QM loans, there’s another spread above QM loans, primarily based on credit but also loan size, geography, etc. Be careful out there! (Today’s podcast can be found here and this week’s ‘casts are sponsored by Optimal Blue. The only end-to-end capital markets platform built to power performance, precision, and profitability. Modern. Proven. Optimal Blue. Hear an interview with economist Elliot Eisenberg on trends seen across economic data and how high the bar has become for the Fed to justify easing if headline resilience persists.) Products, Services, and Software for Brokers and Lenders

Accidental Clairvoyance (Not Really…)

Yesterday’s end-of-day recap bore the title “Half-Hearted Correction Continues.” It turns out that should have been the title for this morning’s commentary as it’s a better description of this morning’s trading (yesterday, we actually had a few fundamentals to justify the weakness). In today’s case, bonds are weaker “just because.” Jobless Claims data (206k vs 225k f’cast) didn’t help though–especially considering a higher claims reading helped kick off the big rally 2 weeks ago. Bonds seem more interested in trading technicals at the moment based on the very linear selling so far this week.  Now that yields have tagged 4.10, it will be interesting to see the next move and it would be a surprise to see a lack of commitment until we get Friday AM econ data.

Half-Hearted Correction Continues

Half-Hearted Correction Continues

Don’t call it a correction.  Yields may have rallied 30bps in just over a week and apparently bounced almost perfectly at 4% (10yr), but they haven’t exactly done much over the past 2 days. It’s just as fair to say things are “sideways and waiting for guidance,” but the cautious approach is to respect the almost-too-obvious technical patterns in play since July. Specifically, bonds have rallied 25-30bps and then consolidated toward slightly higher levels on multiple occasions. This could be the beginning of another similar pattern, but anything is possible if data is gloomy over the next 2 days (especially Friday). Today’s data was modestly unfriendly with core durable goods at 0.6 vs 0.4 and a lackluster 20yr bond auction (not typically a market mover, but added some pressure today).

Econ Data / Events

MBA Purchase Index (Feb)/13

157.1 vs — f’cast, 161.5 prev

MBA Refi Index (Feb)/13

1375.9 vs — f’cast, 1284.6 prev

Core Durable Goods (Dec)

0.6% vs 0.4% f’cast, 0.7% prev

Durable goods (Dec)

-1.4% vs -2% f’cast, 5.3% prev

Housing starts number mm (Dec)

1.404M vs 1.33M f’cast, — prev

Housing starts number mm (Nov)

1.322M vs — f’cast, 1.246M prev

Market Movement Recap

08:33 AM Roughly unchanged overnight and no major reaction to 830am data. MBS unchanged and 10yr up half a bp at 4.064

01:21 PM Slightly weaker after 20yr auction.  10yr yields up 2.2bps at 4.082. MBS unchanged after being up 3 ticks (.09).

02:09 PM zero reaction to Fed Minutes. MBS unchanged and 10yr up 2bps at 4.081

Residential Construction Finds Footing in December

What goes down must come up? Definitely not always the case, but true this time for residential construction numbers. The Census Bureau’s latest report showed a rebound in December, with both housing starts and building permits moving higher after softer readings in prior months. Privately owned housing starts rose 6.2% to a seasonally adjusted annual rate of 1.404 million , up from November’s revised 1.322 million pace. Despite the monthly gain, starts were 7.3% lower than December 2024 levels. Single-family starts increased 4.1% to 981k, while multifamily starts (buildings with five units or more) came in at 402k. On the permitting side, activity also strengthened. Total building permits climbed 4.3% to an annual rate of 1.448 million , though that figure remains 2.2% below year-ago levels. Single-family permits slipped 1.7% to 881k, while multifamily authorizations rose to 515k, driving the overall monthly increase. For the full year, an estimated 1.36 million housing units were started in 2025, down 0.6% from 2024. Permits totaled approximately 1.43 million , representing a 3.6% annual decline. The year-end data suggest a construction sector that regained some footing in December but remained modestly below last year’s pace overall.