10-21 Sept Existing House Sales…beneath the headlines

by Mark on October 21, 2012

Nobody ever looks at the NSA results and internals.  I do.  Especially when the recent trend has been toward heavy, favorable seasonal adjustments in all sorts of data releases.

The following charts illustrate clearly that consensus estimates and sentiment are far too exuberant based on the ‘muddle through” activity occuring in the resale market that I believe will carry through to the builder sooner than later.

September Existing Sales internals — and NSA volume — were lousy in the context of the consensus “recovery”.  It was actually lousy outright; down 100k units to 377k (vs 369k last year), or 21% MoM, and only up 2.2% (8k units) YoY. 

The YoY ‘gain’ was the smallest of the year and comps against last year when the housing market was in full-blown stimulus hangover (double-dip) mode.  It is very important to remember that September Existing Sales come from July & August contracts, which is still peak season. This is unlike next week’s Pending and New Sales, which are from the month of September, when I believe this year’s stimulus-induced ‘demand’ it an important inflection point.

Many think the number of days in the subject month greatly effect existing sales closings. This is mostly incorrect. The number of days in the subject month effect Existing Sales closings — contracted 30 to 60 days prior — to a far less extent than Pendings, New Home Sales and refi’s for example, which are counted at execution.

Bottom line: September Existing Sales (a result of July / Aug contacts) FELL 100k unit or 21% MoMThis is huge.  On a YoY basis they were only up 2.2%, the smallest gain in a year. Based on September Pending sampling October Existing Sales will be xxxxxxxxxxxxxxxxxxxxxx (forecast redacted). 

Sept Existing Home Sales Highlights (as I see them):

1)  NAR’s seasonal adjustments were running hot this month;  over 10% greater than in 2007, which was the last time there was 19 days in September.

2)  NSA Existing Home Sales ONLY up 8k units, or 2.2%, YoY;  the smallest gain of the year and a rounding error.

3)  Sept Existing Sales data (July & Aug contract signings) do not bode well for next week’s Pending and New Home Sales reports

4)  Investors and first-timers — who have carried the market for 2.5 years — have gone away. They have either had enough to eat or have been priced out. Foreign buyers have also shut it down.

5)  Repeat buyers turned on the heat late summer after being largely missing for 3 years due to rates and pent-up demand. But they always go away in the fall.

6)  This leaves weak support — no leadership — in housing in the fall and winter vs. 1-year ago when sales were strong due to Twist, weather and strong insti investor demand. This leaves sales volume vulnerable to much weaker YoY comps for the next full year.  In other words, serious demand ‘headwinds” will be blowing.

7)  Supply will increase into Q4 as distressed supply ticks higher, people initiate short sales to get ahead of the housing fiscal cliff, and sales volume drops sharply.  This will stabilize volume at lower YoY levels but bring down “prices paid” for houses.

8)  Next week’s Pending Sales and New Home Sales xxxxxxxxxxxxxxxxxxx (forecast redcated)

Reiterate Negative following today’s report and ahead of next’s weeks Pending and New Home Sales reports:  xxxxxxxxxxxxxxxxx (names redcated)


The Data

1)  Mo Existing Sales Volume (NSA) Muddling Through. Big declines on deck.

Only up 8k YoY, or 2.2%. And last year we were in the midst of a “double-dip” meaning in July & August when September Existing Sales were contracted the market must have been very soft.

A closer look at Sept Existing Sales (NSA) relative to 2009 – 2011. Sales are in-line with the worst years ever for housing. Sept 2012 were sharply higher than right after the 2010 tax credit expired and sales crashed (blue). That’s not saying much.


2)  Demand by cohort says it all.

Bottom line: This is a market lacking leadership and set-up for weaker YoY sales volume for the next year at least.

 – First-timer demand weak and bodes ill for builder sales.  First-timers helped to carry the market for the past 2.5 years. Without strong demand from this cohort the market will be without a huge tailwind.

– investor demand weak and down YoY again in Sept…for a second straight month.  Again, investors in large part carried resales for the past few years. A sleepy investor is not what a “durable” recovery is made of.

– All-cash investors finally went negative YoY.   I believe this is in part a result of foreigners pulling out. Yet, another headwind for this market.

– Repeat buyers kicked into gear late in the season but barely made up over 50% of total volume. Repeats will go away in Q4. Next month’s Existing Sales results will reflect this.


3)  Existing Home Sales YoY results (NSA)…September the smallest gain of the year.

A Sept 2.2% gain, or 8k units, is not what a “durable” recovery with “escape velocity” looks like especially considering last year the market was in the midst of a double-dip.

Note, the pink bars are my Oct through Dec forecasts.



4)  Number of days in the month misnomer:   With respect to “Existing Sales” the number of days in a subject month carries a lot less weight than with Pending & New Home Sales — surveyed at contract — for example.

You will hear arguments there were fewer days in September, which is why sales crashed in many regions.   Of course, this has some impact but not as much as most think.

In reality, fewer days in the subject month when it comes to ‘Existing Home Sales’ specifically does not matter as much as fewer days does to ‘Pending’ or ‘New Home Sales’ for example.  That’s because Existing Sales are generally contracted a month or so prior and there is a large percentage of buyers and sellers who benefit by closing as near to month end as possible for a number of reasons.

Bottom line:  Fewer days in a particular month as it relates to Existing Home Sales ‘closings’ specifically means people that work to close these escrows have more to close each day of the short month going into month end.   By contrast, builders selling ‘New Homes’ — which are counted at contract not closing — physically have fewer days in which to get buyers into contract, which has more of a direct impact on sales counts.


Best Regards,


{ 8 comments… read them below or add one }

Leave a Comment

Previous post:

Next post: