The Census Bureau reported new home sales for October at a seasonally adjusted annual rate of 632,000 units. Take the headline sales number with a huge grain of salt. The Census numbers are notorious for monthly revisions. It’s reasonable to expect the October sales estimate will receive a haircut.  Census estimates are obviously not capturing the spiking cancellation rates of home builders. The October estimate also does not fit with numbers builders have been reporting on actual sales in the latest quarter.

The attached chart comparing Census estimates for sales vs the October spike in mortgage interest rates is another clue that the folks at Census are missing the broader picture. You don’t get an increase in sales with median prices spiking to record highs while mortgage rates are also spiking to 7 percent. That defies logic, for good reason. They’re called estimates for a reason.

According to Census there are 1.2 months of completed supply. The inventory of new homes under construction is sitting at 5.7 months. There are a record 111 thousand homes which have not started construction. That translates to about 2.1 months of supply, or roughly double the normal level.

Here in the land of reality we’ve been watching builders ramp up the incentives and even lower prices on available new construction. The data on actual new home sales in the Dallas Fort Worth area is still unreliable in terms of year-over-year comparisons. Builders didn’t start putting their inventory back into the MLS until late spring and early summer. It could be another 6-9 months before we see some reliable MLS sales comparisons in that regard.

What we do know for certain is that publicly traded home builders have been reporting collapsing orders and spiking cancellation rates. Here’s a sample of the cancellation rates for some of the largest publicly traded builders in their latest quarter

  • D.R. Horton 32%
  • PulteGroup 24%
  • Meritage Homes 30%
  • KB Home 35%

Those are actual numbers from the builders, not some estimate. D.R. Horton’s net sales orders for the quarter ending in September fell 15 percent. It’s a pretty safe assumption their orders (the equivalent of Census “sales” estimates) were even worse in the month of October when rates jumped over 7 percent. It’s important to remember that Census counts a “sale” at the time of contract signing, not the actual closing. Census “sales” numbers would be better described and labeled as pending contracts.

A recent survey from John Burns Consulting put the cancellation rate for Texas home builders at 39 percent in October.

“The cancellation rates vary substantially by region: In the Southwest, the cancellation rate spiked to 45%. Nearly half of all contracts signed are then cancelled! This was up from a cancellation rate of 9% a year ago. In Texas, the cancellation rate spiked to 39%, up from 12% a year ago.”

As Wolf Richter points out, those high cancellation rates render many of the new home “sales” absolutely worthless. In the real world it’s not a sale if it never reaches the closing table. Census estimates don’t factor in this reality.

There are still a large number of homes under construction. The latest estimates show 472,000 new homes available in various phases of construction. That’s the largest total so far this year. The amount of completed inventory continues to grow. As it moves higher over the next several months that will put more downward pressure on new home prices. Builders are offering up all kinds of incentives to drum up demand. They should dispense with the “limited time only” nonsense, because those incentives aren’t going away anytime soon. There are a number of DFW submarkets sitting on a half year or more of new construction inventory.

The Fed still has their foot on demand’s throat, and the housing market will remain in correction mode until this cycle has run its course. If you are in the market for a new home, there is no rush. The builder sales reps would like you to think there’s no better time than the present, but the data says more price cuts are coming.

Caveat Emptor