Lawler on Publicly-Traded Home Builder Results, "Good Chance for Downward Revision in New Home Sales"

From housing economist Tom Lawler:

Several large, publicly-traded home builders have released earnings for the quarter ended June 30th, 2013, and the operating results varied wildly.  Below are some stats for net orders, settlements, and average sales prices on settlements.

All results in the table below include “discontinued operations,” and Standard Pacific’s results exclude JVs.

Ryland’s results included 177 net orders and 17 settlements associated with its June acquisitions of the Dallas operation of LionsGate homes. Standard Pacific completed its acquisitions of “select homebuilding assets” from Centerline Homes and affiliates in June, which included about 30 current and future communities, five of which are active with 119 homes under contract.

Of the results released so far, “most surprising” was the weakness in net order growth at the two largest US home builders, D.R. Horton (up just 12.2% YOY) and PulteGroup (DOWN 12.4% YOY). D.R. Horton’s sales cancellation rate was 24%, up from 19% in the previous quarter and 23% a year ago. An official cited rising mortgage rates as contributing to the rise in cancellations. Last quarter’s sales cancellation rate was down from a year ago, however, at Meritage, M/I, NVR, and Ryland. D.R. Horton, of course, relies more on the first-time home buyer market than these other builders. Pulte does not report its sales cancellation rate in its press release, though an official said that cancellation rates were “little changed.” In Pulte’s press release the company said that “the recent rise in interest rates has had little effect on overall activity,” despite its incredibly disappointing orders. Pulte attributed the slowdown in net home orders to a 16% reduction in community count, as well as “the company’s decision to purposely slow sales in a number of communities,” particularly in Arizona, Nevada, and Southern California, where an official said Pulte had sold homes “too quickly” (presumably meaning at “too low” a price) in 2012. Stated another way, Pulte increased prices in a fashion that slowed sales.

	Net Orders			Settlements			Average Closing Price
Qtr. Ended:	3/31/13	3/31/12	% Chg	3/31/13	3/31/12	% Chg	3/31/13	3/31/12	% Chg
D.R. Horton	6,822	6,079	12.2%	6,464	4,957	30.4%	$252,290	$224,975	12.1%
Pulte Group	4,885	5,578	-12.4%	4,152	3,816	8.8%	$294,000	$268,000	9.7%
NVR	3,278	2,614	25.4%	2,878	2,475	16.3%	$344,700	$305,100	13.0%
The Ryland Group	2,191	1,415	54.8%	1,659	1,149	44.4%	$287,000	$253,000	13.4%
Standard Pacific	1,516	1,108	36.8%	1,095	815	34.4%	$397,000	$337,000	17.8%
Meritage Homes	1,637	1,353	21.0%	1,321	1,042	26.8%	$330,000	$270,000	22.2%
M/I Homes	1,078	826	30.5%	788	625	26.1%	$281,000	$259,000	8.5%
Total	21,407	18,973	12.8%	18,357	14,879	23.4%	$294,805	$262,220	12.4%

On the pricing front, all builders reported significant increases in average sales prices, with the increases generally attributed both to overall price increases and a shift in mix, with several builders saying there were focused more on the “move-up” market. Demand from first-time home buyers was perceived at many builders as being “weak.”

While overall results varied a lot, net home orders for the above seven home builders combined were up just 12.8% from a year ago – and a bit less after adjusting for acquisitions – which combined was SUBSTANTIALLY below consensus. In the quarter ended March 31, 2013, net orders for these seven builders combined were up 24.0% from the comparable quarter of 2012.

Given the net home orders reported so far by publicly-traded builders, the Census estimates for new SF home sales for Q2/2013 seem surprisingly strong – 134,000 on an unadjusted basis, up over 30% from the second quarter of 2012. The estimated YOY gain in sales Q1/2013 was about 25% .

Of course, comparing home builder reports with Census SF home sales estimates is “challenging.” First, Census does not treat sales cancellations the same as do home builders. Second, historical data suggest that the timing of the recognition of a “sale” by Census lags that of home builders. Third, there can be substantial quarterly swings in market share. And fourth, preliminary Census estimates are subject to substantial revisions, partly because preliminary estimates include “imputed” data, because the survey data used to estimate sales are based on a permit being issued. Many homes may have a sales contract signed prior to a permit being issued, and Census must “guesstimate” such sales using some “historical trends” model (that has often been changed).

Nevertheless, I have found builder reports to be somewhat useful in projecting revisions in Census’ estimates of new SF home sales. Based on the builder reports so far, there is a better than even chance that Census’ new SF home sales estimates for last quarter will be revised downward, and I’d guess that the bulk of the downward revision will be in June sales. It’s worth noting that while Census’ estimated new SF home sales last quarter were up by over 30% YOY (on an unadjusted basis), estimated sales of homes not yet started were up 53%, compared to 24% for sales of homes under construction, and 20% for sales of completed homes. For June Census estimated that there were 18,000 sales of new SF homes not yet started, up from 11,000 last June, and the highest level since June 2007.