by Calculated Risk on 5/29/2012 07:55:00 PM

Economist Tom Lawler wrote today:

I put “seasonally” in quotes, as there have substantial changes in the purported “seasonal” pattern of home prices since the housing market cratered. The reason, of course, is that there is a marked “seasonal” in the distressed-sales share of home sales, which peaks in the late winter months and hits a trough in the summer months. Not coincidentally, the “shift” in the “seasonal” pattern of home prices has been one where home prices are “seasonally” much weaker than they used to be in late winter, and “seasonally” much strong than they used to be in the summer.
...
Everyone “knows” why, and since the “cause” of the apparent wider “seasonal” swings is known (and someday will go away), it’s not rightly correct to call such swings “seasonal.”

The following graph shows the change in the seasonal factor over time using the Case-Shiller National Index.

Click on graph for larger image.

Most of the wild "seasonal" swings are related to foreclosures. As Lawler noted, foreclosure sales are fairly steady throughout the year, and conventional sales have a seasonal pattern. So in the winter foreclosures are a higher percentage of sales, and that pushes down the NSA prices.

The second graph shows the year-over-year change in the seasonal factor. Clearly something started to happen around 2005.

It seems that the seasonal factors started changing in 2005 - when prices were still going up.

Also, it appears that the change in the seasonal factors has slowed, and will probably start to reverse soon.

Lawler also commented that he thinks there is a "better than even shot" the National HPI will show a year-over-year gain next quarter. That would be a 6% increase in the NSA index in Q2 (or about a 2% increase in the "seasonally adjusted" index). My guess is the index will turn positive on a year-over-year basis later this year.

Earlier on house prices:
• Case Shiller: House Prices fall to new post-bubble lows in March NSA
• Real House Prices and Price-to-Rent Ratio at late '90s Levels