Have we seen the bottom in mortgage rates for 2024 after a crazy roller coaster ride so far this year? My 2024 forecast had a mortgage rate range of 7.25%-5.75%. To get to the lower end of this range, we needed to see two things: the labor market getting softer and the mortgage spreads improving. This is the double-whammy impact, and that’s what has happened.
However, it’s still September, and we have three months to go! Can my lowest range forecast be wrong? Yes, here’s how and why.
10-year yield and mortgage rates
My 2024 forecast included:
- A range for mortgage rates between 7.25%-5.75%
- A range for the 10-year yield between 4.25%-3.21%
How rates get to the lower-end range of the forecast is critical. There are two variables: the labor data getting softer is the prime one and the second one is the spreads getting better. Again, the double whammy of lower yields and spreads. This is not about more Fed rate cuts, because the market has priced in a lot Fed rate cuts already, but they haven’t priced in a recession yet. People wonder why rates went up after the bigger than expected Fed rate cut, as shown in the chart below. I talked about this in this HousingWire Daily podcast.
With the 10-year yield at 3.74% as of Friday, we have some room left to reach the very bottom of the 2024 forecast before the year is out. However, this will need the labor and economic data to get much weaker. That’s the first variable — the second one is the spreads.
Mortgage spreads
The mortgage spread story has been positive in 2024, whereas it was negative in 2023. We have seen a big move, which has helped, and we still have some runway left to return to historical norms. This can help get mortgage rates down toward 5.75%. If we took the worst spreads from 2023 and incorporated those today, mortgage rates would be 0.68% higher. At the same time, we are far from average with the spreads, as we are still 0.85% higher today than the low levels of 2022 in the chart below.
Purchase application data
Purchase applications had another positive week, making the winning streak four weeks in a row — the longest of the year. Last week, purchase apps grew 5% weekly and fell 0.4% year over year. The slight decline year over year is the smallest decline since 2022. However, remember that last year at this time, mortgage rates were heading toward 8%, so the year-over-year comps will be easy to beat. That said, we have had a material change in data in the last 15 weeks.
This is what weekly purchase application data looked like with rising rates starting from the latter part of January:
- 14 negative prints
- 2 flat prints
- 2 positive prints
As you can see, this was shaping up to be a highly negative year with the weekly application data. Before late January when rates started to rise, we had about eight weeks of positive trending purchase apps, and then the rising rates zapped the data in a very negative curve.
This is what weekly purchase application data looks like since mortgage rates started to fall in mid-June:
- 10 positive prints
- 5 negative prints
The volume down and up this year hasn’t been much, but we can clearly see a difference in the data now.
Weekly housing inventory data
The best story for me about housing this year has been the inventory growth. Unlike some crazy people on the internet, I was never worried about a massive housing bubble crash, but since the summer of 2020 I’ve been more worried about home prices escalating out of control. With the inventory growth we’re seeing and demand firming up, this is the best we could have hoped for in 2024.
Last week, we added 11,589 houses in the effective average inventory model, but with better demand and lower mortgage rates. This is the sweet spot for housing.
- Weekly inventory change (Sept. 13-Sept. 20): Inventory rose from 713,660 to 725,249
- The same week last year (Sept. 14-Sept. 21): Inventory rose from 519,458 to 528,797
- The all-time inventory bottom was in 2022 at 240,497
- The yearly inventory peak for 2024 is 725,249
- For some context, active listings for this week in 2015 were 1,198,819
New listings data
Another positive data line this year is that new listings data has grown from the lowest levels ever recorded in history in 2023. Since most sellers are buyers, this data must return to normal before seeing real, long-lasting sales growth. However, I missed my 2024 forecast of at least 80,000 new listings per week this year during the seasonal peak months by roughly 5,000. On a positive note, we did see some good growth last week!
- 2024: 70,157
- 2023: 59,194
- 2022: 63,853
Price-cut percentage
In an average year, one-third of all homes take a price cut — this is standard housing activity. Rising mortgage rates last year and this year have created a growing level of price cuts, especially with inventory rising. This data line has slowed as rates have fallen. In my 2024 price forecast, I was on the shallow end for price growth and I would have been too low if mortgage rates hadn’t risen earlier in the year to slow down mortgage demand.
A few months ago, on the HousingWire Daily podcast, I discussed that the price-growth data would cool down in the year’s second half. The price cut percentage data is below 2022 levels and risks an earlier seasonal curve lower than 2022 and 2023. This is with more inventory than both years, too. *But remember, we had rising rates in the fall of 2023, toward 8%, and 2022 had sales crashing, too.
Here are the price-cut percentages for last week over the previous few years:
- 2024: 39.3%
- 2023: 37%
- 2022: 41%
Weekly pending sales
Below is the Altos Research weekly pending contract data to show real-time demand. We are seeing the seasonal decline in the data line but have some year-over-year growth. Demand has recently been firming up just a tad with lower mortgage rates. I wouldn’t compare it on a year-over-year basis because last year’s rates were heading toward 8% right about now, but demand has perked up a bit.
- 2024: 360,090
- 2023: 344,409
- 2022: 390,935
The week ahead: Fed speeches, home prices, new home sales and PCE inflation
This week will will have three Fed presidents talking on Monday, so watch for market reaction. We also have home price data, which should show a year-over-year cooldown in prices, which I have been talking about for months. Remember, Case-Shiller and others lag our work by months. New home sales will come out, too this week, I do expect some lower revisions to the monster beat we had last month. We also have the Fed’s preferred Inflation index, the PCE inflation data, but as we all know now, labor is more important than inflation, so keep an eye out for jobless claims.