Unbelievably Great Starts Data May be Just That: Unbelievable
Starts rose sequentially by 17.2%, per the U.S. Census’s latest release, which no one can decipher. Big Builder’s report on the release is here. At 532,000, seasonally adjusted, starts beat the Street by 47,000, or 10%. Permits, at 518,000, also eclipsed Wall Street analysts’ expectations by 10,000, or about 2%.
Evidently, the Street–and its gaggle of “consensus” economists–have neither visibility nor acumen into what to expect from new residential construction.
Or it means that government data culling is suspect. As Raymond James VP for equity research Buck Horne notes in his comment on May starts, permits, and completions data, the margins for error in the government data are all important. “Material downward revisions to the May housing starts estimates are more likely than not,” he says.
Here, from HousingWire, is the nub of the starts and permits data:
The good month for housing starts comes after the volume dived 12.9% the month before. A 62% increase in new multifamily construction drove the month-over-month gain, while single-family home starts rose 7.5% to an annual rate of 401,000 units. Single-family building permits — an indicator of future starts — rose 7.9% in the month to an annual rate of 408,000 permits.
Which do you believe? The Street’s analysts are full of it or the government’s math is off.
Either could be the case, and both are true. Economists are quick to say that a big difference between The Great Depression and now is that President Franklin Roosevelt lacked a good economist to advise him on ways to steer toward a quicker recovery during the 1930s. Still, as smart as a lot of those guys are, how much are they helping us know anything before it actually goes down, and even when it does, how helpful is their commentary?
On the other hand, it could be that in this case, economists’s estimates are smarter than the data flow from the Census. But once the headlines get a hold of the data, it’s really too late to worry about what’s really correct or not.
RaymondJames’ Horne dives into the government release and turns up a number of self-cancelling and contradictory figures that lead him to his conclusion that, when we see adjusted numbers in 30 days, may vindicate The Street’s more guarded estimates after all.
Here’s a doozy of a finding from the Buck Horne analysis.
Shrinking number of homes actively under construction directly contradicts starts data: However, the most strikingly peculiar aspect of this morning’s data was found in the little-noticed Table 4 of the full release, representing the Census Bureau’s estimate for housing units actively under construction. Under normal circumstances, we would think that if single-family starts had troughed and were actually rising materially – particularly in the seasonal low point for new home deliveries – there should be a coincident increase in the number of homes actively in construction. Oddly, however, the Census data here (which carries +/- 1.4% confidence interval) shows that homes under construction actually fell 3.9% versus April seasonally adjusted and dropped 1.3% month/month on an absolute basis. In the housing recovery of 1991, we note that this statistic indeed showed a sharply accelerating increase in the number of homes under construction beginning in April 1991 coincident with the recovery pattern.
Point is, at HousingCrisis.com, we do feel that bigger, well-capitalized companies may actually roll the dice on going vertical with more homes in the next couple of months. Why? Because, in this market environment, about a third-to-half of willing and able buyers want to settle and move in quickly. That makes specs almost inevitable.
Also, as deadlines approach on both state and federal tax credit programs for home buyers, builders who can grab capital to offer ready-to-move-in new homes are prepared to gamble on their ability to nail the shrunken, moving target of buyer demand.
As we posted yesterday, we believe that home builders are hormonally range-bound between unflappable and optimistic, and they’ll look at every mixed signal as a sure sign that it’s time to start moving the dirt and opening the new models.
Ultimately, It occurs to us that as the downturn sputters and runs out of gas, it’s likely will see a knifing up and down of starts before they settle on a conviction in their direction. Two months up, one month down; two months down, one month up, and so on.
Whatever the case, somebody’s wrong about today’s starts numbers. It looks for the moment like the The Street is, but stay tuned.
The takeaway: Clip the postive headline for the scrapbook, but wait a couple of months before you paste it in as a keeper.
Comments
Leave a Reply