In terms of day-over-day changes, today's mortgage rate movement was forgettable. The average borrower wouldn't see much of a difference from yesterday's rates at the average lender. In both cases, those rates would be at or near the highest levels since 2001. The underlying bond market experienced a bit more drama. The early morning hours actually pointed to slightly lower rates, but those dreams were shattered by lunchtime. Bonds had already lost a decent amount of ground by the time mortgage lenders released rates. This kept lenders in a more conservative mindset, but several lenders were still forced to bump rates slightly later in the day as bonds lost more ground. Best case scenario rates remain near 7.5%, but many scenarios are seeing rates closer to 8%.
Stronger Start, Weaker Finish, No Real Reason
Bonds improved in the overnight session, but reversed course the moment that the US bond market hit the 8:20am CME open. The selling wasn't intense, but it was consistent from that point on. 10yr yields ultimate set new long term highs just over 4.56 and MBS lost an eighth to a quarter. The most interesting market mover of the day was the complete absence of compelling market movers. This fits the "repricing" narrative discussed in the morning commentary. The silver lining is that the day-over-day losses were far smaller than those seen yesterday, but it's too soon to conclude bonds are settling in for next week's big ticket data.
Econ Data / Events
Case Shiller Home Prices y/y
+0.1 vs -0.3 f'cast, -1.2 prev
FHFA Home Prices y/y
+4.6 vs 3.2 prev
Market Movement Recap
09:24 AM Initially weaker in Asia, but stronger in Europe. Giving back some overnight gains. 10yr down 1.4bps at 4.517 and MBS unchanged to a few ticks lower.
12:18 PM Selling continued into PM hours. 10yr now up 1.3bps at 4.544. MBS down just over an eighth.
02:27 PM Treasuries continue to weakest levels with 10s up 2.9bps at 4.56. MBS are down 3/8ths, but more than half of that weakness is due to illiquidity.
August sales of newly constructed single-family homes failed to match the robust numbers from July but were significantly better than those a year earlier. The U.S. Census Bureau and Department of Housing and Urban Development said last month’s sales were at a seasonally adjusted annual rate of 675,000, the lowest since March and an 8.7 percent decline from July’s revised estimate (from 714,000) of 739,000 units. The August results were 5.8 percent higher than the 638,000-unit rate in August 2022. The August results did not meet the consensus estimates from either Econoday (699,000 annual units) or Trading Economics (700,000). Robert Dietz, chief economist for the National Association of Home Builders said of the report, “Builders continue to grapple with supply-side concerns in a market with poor levels of housing affordability. Higher interest rates (the average was over 7 percent) price out demand, as seen in August, but also increase the cost of financing for builder and developer loans, adding another hurdle for building.” On an unadjusted basis, there were 54,000 homes sold during the month, down from 61,000 in July. Over the first eight months of 2023, sales of new homes have totaled 474,000 compared to 466,000 at the same point last year. Sale prices have fallen slightly in the last 12 months. The median price in August was $430,300, $10,000 lower year-over-year. The average price has dropped from $530,800 to $514,000.
The seemingly never-ending sea of red continues. And for good measure, this one began with a bit of hope in the form of an overnight rally that got 10yr yields back below 4.50%. Not one moment after the official end of the overnight session, domestic traders started selling, pausing only briefly for the 9:30am NYSE open before taking yields to new multi-year highs in the 11am hour. .
Like yesterday, there are no new or significant root causes for the weakness. It's the same old story of broad, large-scale "repricing." But what even is "repricing" in this context? It's a term we've only dusted off a few times over the past 15 years. It occurred in 2013 surrounding the taper tantrum and in 2016 after the presidential election. Those were examples of rapid repricing to higher yields.
There is a more gradual version as well, like the pervasive rally that took place in 2014 as the market repriced expectations for European QE and in 2019 as global growth concerns collided with the trade war and an oversold bond market.
In all examples, "repricing" refers to a sustained move in one direction that requires no new surprises from typical fundamentals. In other words, econ data didn't justify the scope or pervasive nature of these moves. There was some deeper, underlying x factor that caused traders to identify a destination and then to proceed methodically in that direction. The was abundantly clear in the first half of 2022--probably the single best example of a "repricing" event in modern economic history.
But here's the kicker: that repricing has been underway since August of 2020. That's when bonds first began to lift off from the covid lows and when market participants first contemplated a future where trading levels weren't tied to covid case counts. It's not as if the world knew yields needed to be over 4% at the time, but it was clear they needed to be higher. In late 2021, we could similarly see that yields needed to be MUCH higher.
Now in late 2023, with 10yr yields already around 4.5%, I wouldn't agree there's as compelling a case for a repricing event, but that is nonetheless how the market is trading it. It could be that it's a short-term affair in reaction to last week's Fed announcement, but only if next week's data is week. Otherwise, it's just the market's way of getting in position to finally not be surprised by surprisingly resilient big-ticket data.
Home prices have resumed their upward climb despite mortgage rates that have doubled post-COVID. According to Craig J. Lazzara, Managing Director at S&P CoreLogic Case-Shiller Indices, the National Index for July hit an all-time high. That index, which covers all nine U.S. census divisions, rose 1.0 percent from the previous July, after posting zero change on an annual basis in June. The 10-City Composite showed an increase of 0.9 percent after a 0.5 percent loss the previous month and the 20-City Composite was up 0.1 percent, improving from an annual loss of 1.2 percent. Chicago, Cleveland, and New York led the way for the third consecutive month reporting the highest year-over-year gains among the 20 cities in July. Chicago remained in the top spot with a 4.4 percent increase, with Cleveland (which has long vied with Detroit for the low spot in Case-Shiller’s numbers) was second, with a 4.0 percent annual gain. New York held down the third spot with a 3.8 percent increase. Eight of 20 cities reported lower prices and 12 of 20 reported higher prices in the year ending July 2023 compared to prior annual numbers. Eighteen of the 20 cities accelerated at a higher rate than in June. Lazzara said, “We have previously noted that home prices peaked in June 2022 and fell through January 2023, declining by 5.0 percent in those seven months. The increase in prices that began in January has now erased the earlier decline, so that July represents a new all-time high for the National Composite. Moreover, this recovery in home prices is broadly based. As was the case last month, 10 of the 20 cities in our sample have reached all-time high levels. In July, prices rose in all 20 cities after seasonal adjustment and in 19 of them before adjustment.
As if lenders and vendors don’t have enough other stuff to worry about, the budgetary standoff in the U.S. doesn't look like it will abate soon, raising the likelihood of the first government shutdown since 2019. Current funding for federal operations will end on October 1 unless a deal is reached or the proverbial can kicked down the road. Thousands of federal workers might be furloughed without pay. Sure it will be temporary, and its wider impact will likely be limited, but still even talking about it is lousy. According to Morgan Stanley, the last 20 government shutdowns that occurred since 1976 "appear to have had limited impact on the economy." As for bond prices, a shutdown may cause some "temporary instability", but this is not a given. There is talk of a short-term Continuing Resolution (CR) providing funding until later this year, but federal agencies, including HUD and Treasury, will cease to function normally. The National Flood Insurance Program (NFIP) authorities also expire on October 1st. The Mortgage Bankers Association created a guide outlining how HUD (including FHA and Ginnie Mae), VA, and USDA would be directly affected by the furlough of government employees and the curtailment of agency operations. (Today’s podcast can be found here and this week’s is sponsored by Built. Built is powering smarter and faster money movement for the entire construction and real estate ecosystem, all while reducing risk. Hear an interview with Servbank’s Bryan Crofford on how companies can best invest in employees, promoting longevity and success.)
The average lender was already very close to multi-decade highs for 30yr fixed mortgage rates last Thursday afternoon, but a modest recovery on Friday meant that there was a chance we could have avoided printing today's headline this week. Unfortunately, the bond market started the day in rough shape and continued to lose ground throughout. At the time that most mortgage lenders released today's initial rate offerings, last Thursday was still worse. It wasn't until several lenders released negative/upward revisions to rate sheets that we officially crossed above the multidecade ceiling. For the average lender, a top tier 30yr fixed rate is now over 7.5% for the first time in at least 22 years. The average borrower (not "top tier") is seeing rates that are even higher. This assumes an adjustment for discount points. Many loans are being quoted with points currently, and in those cases, the note rate would be a bit lower. Frustratingly, there were no compelling new motivations for the bond market weakness. Negative momentum has been more of a snowball than a calculated decision over the past few business days. It may be hard for rates to muster much of a counterattack without a meaningful, negative shift in economic data.
Sell Now and Wait For Something to Convince You to Buy
Bonds began the week with another move to long-term yield highs. There was a wave of selling in the overnight session led by Europe and another when domestic traders ramped up for the day. Neither were unequivocally the product of some data or news headline although there were a few scapegoats that could be mistaken for motivation. The problem with said scapegoats is that--while they likely contributed--they were not nearly meaningful enough to justify the movement in question. Conclusion: this sort of selling is broader and more sentiment-driven. Traders are repricing "higher for longer" odds with the longer end of the yield curve. Buyers are on strike until something convinces them to buy and that will be hard to do unless next week's data is weak.
Market Movement Recap
09:55 AM Sharply weaker overnight with more selling early and now a modest bounce. 10yr up 7.9bps at 4.515. MBS down just over 3/8ths.
01:16 PM Sideways to slightly stronger into PM hours for MBS, now down 11 ticks (.34). 10yr sideways near highs, up 8.7bps at 4.523.
02:47 PM Weakest levels of the day with 10s up 11bps at 4.546. MBS are down 5/8ths but at least an eighth of that is attributable to illiquidity.
04:08 PM little-changed from the last update. MBS down half a point. 10yr up just under 11bps at 4.544.
We continue to anticipate a broadly sideways trend this week although that expectations is being put to the test in early trading. Yields in the longer end of the curve hit more long-term highs this morning, and not in response to any especially compelling data or events. Weakness was generalized in the overnight session, and not much better in early domestic trading.
European bond markets pushed the pace of weakness after the EU open. This coincided with some economic data that was only slightly stronger than expected as well as some ECB comments that were arguably not too
One could argue that Fed comments on an utter commitment to the 2% inflation target added pressure, but it's just as easy to argue that selling remained generalized at home as well. This speaks to an ongoing process of capitulation and "repricing" that could continue to test the expected sideways vibes ahead of next week's top tier data.
Are you gradually eating dinner earlier? Join the trend. Want to know another trend that isn’t so benign? Along the lines of the STRATMOR piece below, I received this email over the weekend. “Rob, I’m a processor at a lender who was doing about $80 million a month in 2020 and 2021 but is now doing $15 million. We saved money during those years and retained servicing. But now we’ve eaten through our savings, sold all of our servicing, had some Agency buybacks, and are barely breaking even. We’ve moved to entirely best-efforts execution, outsourced as much as we can, are continuing to lay people off, and are shedding unused, expensive technology. Where will we go from here?” That is a rough note, but, if it is any help, you’re not alone out there. It’s going to be a difficult autumn and winter, and lender management not doing anything is similar to an ocean swimmer being caught in the waves while everyone is telling them to swim out past the breakers or in to the beach. (Today’s podcast can be found here and this week’s is sponsored by Built. Built is powering smarter and faster money movement for the entire construction and real estate ecosystem, all while reducing risk. Hear an interview with Nationwide Processing’s Guru Amrit Khalsa on offshoring operations and reducing fixed costs for lenders.) Lender and Broker Software, Programs, and Services Since hitting the scene, SimpleNexus’ single sign-on simplicity and mobile convenience have raised the bar for what mortgage technology should be. Now an nCino company, SimpleNexus is backed by a new wealth of resources that will raise the bar even higher. nCino has already harnessed the power of AI to fuel data intelligence and automation across every aspect of banking. For an up-close view into how nCino is using AI to redefine the customer lifecycle and banking as we know it, join us via webinar tomorrow, 9/26, at 11 a.m. ET. Register now for a front-row seat to how nCino is shaping the future of financial services.