Rebound in Housing Starts Is Not a Sign of a Sustainable Turnaround

by Dirk van Dijk

At last month's sales pace, it would take 13.3 months to sell all the new homes currently in inventory. This is by far an all-time record. So what happens? New housing starts jumped in February 22.2% to a seasonally adjusted annual rate (SAAR) of 583,000 from the 477,000 pace of January.

While this is still down significantly (-47.3%) from the year-ago pace of 1,107,000 it is nothing short of economic insanity. The last thing the housing market needs right now is new supply. These numbers will help the look of the first quarter GDP, since the residential investment component will be much larger than anyone expected. However, this is, quite frankly, an unmitigated disaster in terms of the long-term health of the economy.

We need for housing starts to be falling, not rising. It's not that I want to see unemployed construction workers -- I don’t. But if we are going to see the housing markets stabilize, we need to get the inventories under control.

Interestingly though, most of the rise was in apartments, not single-family structures. Single-family starts rose just 1.1% to a seasonally adjusted pace of 357,000 from January’s pace of 353,000. This was less than half the year-ago level of 722,000. Thus most of the rise was in new apartment and condo complexes? What are these developers thinking?

The rise in starts was widespread, with 3 of the 4 census regions seeing major increases. The Northeast posted a stunning 88.5% rise in overall starts, but was still running 48.8% behind a year ago, even with the spectacular rise for the month. The Midwest followed with a 58.5% rise for the month, which was still 45.5% below last year. The South also had a strong month, rising 30.2% -- it is now running 42.5% below year ago levels. The West was the only region where starts remained depressed, falling 24.6% for the month and leaving the pace 59.1% lower than a year ago.

The permits numbers were a muted version of the starts numbers. Nationwide, building permits rose 3.0%, largely due to a big 27.6% pop in the Northeast, and was helped by a 5.9% rise in the South. The Midwest was unchanged. Out West, permits fell 13.6%. On a year over year basis, permits are down 44.2% nationwide, with the regions ranging from a 29.5% decline in the Northeast to a 57.9% decline in the West. Nationwide, single family permits rose 11.0% for the month but are down 42.3% year over year.

While this might cause a short-term pop in the shares of the Homebuilders like D.R. Horton (DHI) and Ryland (RYL), I do not think this is the start of a sustainable turnaround in the housing market. Now, if next week, when the new housing sales numbers come out and show a similar pop, then it will be time for celebration. Simply adding to the pile of houses waiting to be sold is a reason for worry, not celebration.

This means that the decline in housing prices might well be deeper and more protracted than it might otherwise have been. This will cause further damage to the balance sheets of both the consumers, and of the major mortgage lenders like Bank of America (BAC), Fifth Third Bank (FITB) and Comerica (CMA).

Just remember the mantra: More new home sales -- good; more new housing starts -- bad. The first graph below (larger versions available here) shows how this hosing downturn has been steeper than any we have seen before. Yet it has to be seen in conjunction with the second graph showing the inventory overhang. Yes, a rebound in housing starts tends to lead us out of recessions, but that has to be accompanied by a reduction in housing inventories.

Comments

[Richard K:]

Is the demand for rental units being pushed up by millions of displaced former homeowners? If so, what happens to the rental market if -- someday, somehow -- they get their financial act back together and migrate back to private housing?

Mar 17 03:50 PM

[Heaven, Hel...:]

Good article. So at the end of the day, while increased new housing starts viewed in a vacuum may be interpreted as a healthy sign, if not accompanied by increased new housing sales, they will only increase the new housing inventory overhang…..and we all know what inventory overhang does to prices……so let's hope that builders are reacting to something which has not yet shown up in the numbers.

Mar 17 05:18 PM

[David White:]

There are probably a few more points to make about this apparent jump in housing starts last month. First the standard error for the housing starts number is 13.8%, so the 22% number may be highly in error. In fact the building permits number is usually looked at as a better guage of actual activity because it is a much more reliable number. It was only up 3%. Second the weather was unseasonably warm in Feb. This could have led to some projects starting earlier in the year than they might have otherwise. This may mean that we will see the reverse effect on the number next month. All the projects that started a little early, will tend to subtract from the ones that would normally have been recorded for March. Finally since a lot of the new starts had to do with apartments, they may have been anxious to start because they take a longer time to finish than simple single family homes. This too may have skewed the numbers to Feb. from some that might have been in March normally.

In sum one shouldn't read too much into this one month's worth of data. It may be nearly meaningless as a lone piece of data. I would not be at all surprised to find that next month (March) the numbers are down from Feb.'s even though one might suppose there would be more starts in a generally warmer month. If so, this might very well be because some of the starts that would have been March starts got started a little early. There might not be any increased number of overall starts.

Mar 18 01:31 AM

[DMACK:]

There's a problem with your math and it's in the geography. Yes there may be 13 months of new homes sitting on the market but the bulk of them are in 4 regions; California, Vegas, Phoenix and Florida. Unless we develop the technology to move already constructed houses from one state to another economically then we'll likely see starts up from the recent dismal numbers despite the backlog because some parts of the country have their new home inventories in line. I wouldn't be overly concerned about the jump in starts because a 20+% jump from the bottom still leaves you one step from the bottom.

Mar 19 11:48 AM