September Housing Data

Round two of the credit crunch has most definitely made itself known.

Prices

The size-adjusted median prices were pretty solidy down, with single family homes showing a 2% decline and condos sporting a 3% decline just from the prior month. From the 2005 peak, the size-adjusted median price was down 11% for single family homes and 15% for condos.

Incidentally, you may have noticed that the background of the above graph is slightly different than in prior graphs. Last night, a couple Microsoft Excel jockey friends of mine were ruthlessly deriding me because I have always used the default "light gray" background color for my charts. Such was the extent of their mockery that I was inspired to develop the bold new "light gray with the middle bit fading into slightly lighter gray" look that you see before you.

I think you’ll agree that fading the light gray into slightly lighter gray makes the charts far more informative. In retrospect, I have no idea how anyone made heads or tails of the old graphs with their distracting solid gray backgrounds. I can really understand why my friends were so worked up about this topic.

Now, admittedly I have just added a gradiant fade to that same default shade of gray. I’m not some kind of wildman who would add a gradient AND change the color at the same time. I feel that would just be reckless.

By the way, en route to designing my exciting new chart look, I discovered that Excel users have many options when setting chart backgrounds.

You can apply a texture, giving your chart an attractive faux-wood finish:

Or you can even upload your own photos, so that your chart’s background can feature this adorably contrite puppy:

Or magician Doug Henning:

The options are without limit, and I sincerely thank my friends for removing the scales from my eyes and helping me to unleash the mighty power of Microsoft Excel.

Moving on. Here’s the year-over-year size-adjusted median. (Henningless — for now).

The interesting action was actually in the plain vanilla median price, which — not unexpectedly — got smacked around. Condos weren’t actually hit that hard, but the median single family home price was crushed for 7% in a single month.

As I said, this was not unexpected. Median prices had been artificially pushed up for months because the subprime credit crunch impacted buyers of inexpensive homes more than those of expensive homes, thus shifting the buying mix towards pricier properties. As of August, however, the credit crunch hit everyone else and (starting when those sales closed in September) that artificially high median started to adjust back down closer to a level that actually represents home price movements. The fact that single family homes took the brunt of the second, non-subprime credit crunch also makes perfect sense, because that’s where most of the more creditworthy borrowers were doing the buying.

This next chart, which compares the median-based indicators with the far superior Case-Shiller HPI, demonstrates pretty clearly that the median indicators were overstating price strength at least through July. (If you haven’t done so yet, please read my long screed on the pros and cons of the various home price metrics).

The plain vanilla median, especially, has been a poor price indicator of late. Yet despite the glaring evidence presented in the above chart, and further enhanced by an awesome background gradient, the vast majority of analysts and industry talking heads have continued to remain laser focused on the median price. Many of them have allowed the recent strength in the median to fool them into believing that home prices are stable.

So what happens now that the median has gotten whacked? Will these analysts clue into what’s really happening in the market? Will some of them suddenly downplay the importance of the median as an indicator of price movements? If so, will anyone notice that this second group of analysts was fine with the median’s inaccuracy when it understated price declines but just not so much when it started to overstate them (making it look like single family home prices plummeted 7% in a month)?

This month’s round of housing data punditry should be entertaining.

Supply and Demand

Inventory followed the same pattern as in recent months, with more single family homes on the market than last year, fewer condos, and about the same number of homes overall.

The following graph is something new that I will be doing for data series, such as inventory and sales, that have heavy seasonal influences. Thanks to user WaitingToExhale, from whom I stole the idea after s/he used it in this thread. The idea is to chart 2 calendar years worth of data, with each year as its own line to see how the years stacked up for each month.

As you can see, 2007 inventory has acted a lot like it did in 2006. The one notable difference is that while 2006 inventory peaked in August, it kept growing into September 2007 and even picked up the pace from the prior month’s growth.

This new charting method is more useful, in my opinion. However, the old technique does offer some different info by separating the property types and showing the percent change from the same month last year, so I will update both charts from here on out.

The supply situation may not have changed much since last year, but round two of the crunch has rendered weak demand even weaker, resulting in some pretty poor year-over-year comparisons.

The declining sales volume against slightly rising inventory led to a big spike up in the months-of-inventory figure. As of last month there were 12 months of inventory listed for sale on the MLS. That is, in a word, awful, and that’s without even considering how much of that inventory is of the must-sell variety resulting from the record-shattering pace of new foreclosures.

The two-year version shows that while an increase in months of inventory can be expected in September, this year’s increase was substantially larger than last year’s. As of August 2007, the months-of-inventory figure was 21% higher than it had been in August 2006. The September 2007 figure, by contrast, was up by 38% over September 2006.

Conclusion

By pretty much every measure, whether lagging (prices) or leading (sales, inventory, and foreclosures), things looked a lot worse in September 2007 than they had one year or even one month prior. We’ll see how the pundits try to spin things this month, but the reality is that the housing bust is far from over and looks ready to pick up the pace in the months ahead.