by Calculated Risk on 3/22/2012 04:35:00 PM

In a post yesterday I mentioned some bearish comments from Professor Kenneth Rosen back in 2006. That reminded me of some comments I made back in 2005 and 2006 when I argued the sequence for housing would be:

1) A surge in inventories as sellers try to get out at today's [2006] high prices.

2) followed by a drop in orders as buyers become leery of buying at the top. Historically house prices tend to be sticky as sellers want prices close to those of recent sales in their neighborhood. And buyers want a discount from recent sales. The result is a drop in orders.

3) Then prices start falling as some sellers (speculators and homeowners in distress) need to get out.

Sure enough inventory started rising in the second half of 2005, and then activity started to decline - and then prices eventually started to fall - and finally fell off a cliff.

Now inventory is declining, activity is picking up gradually, and I think it is time to look for prices to stop falling. (I don't expect prices to rise quickly, but if prices just stopped falling, then people would become more confident in the real estate market). Note: by fundamental measures (real prices, price-to-income), prices are probably close to a bottom, so I think it is OK to start looking for a bottom.

As I noted yesterday one of indicators I'm looking at is the year-over-year change in house prices. If we are at the house price bottom on a national basis, then year-over-year price changes should start to get smaller soon - and eventually turn positive in early 2013.

In addition to Case-Shiller, CoreLogic, and LPS, I'm also watching the FNC and RadarLogic indexes.

Click on graph for larger image.

The first graph is based on the FNC index (four composites) through January 2012. The FNC indexes are hedonic price indexes using a blend of sold homes and real-time appraisals.

The indexes are generally showing less of a year-over-year decline in January (I think prices will fall seasonally through the March report).

Also RadarLogic released their January report today.

According to the January 2012 RPX Monthly Housing Market Report released today by Radar Logic Incorporated, the RPX Composite price, which tracks home prices in 25 major US metropolitan areas, declined 5.42 percent during the year ending January 19 to $169.75 per square foot. The last time the RPX Composite was this low was in July 2002.

The year-over-year rate of decline in the RPX Composite price slowed in December and January after reaching its fastest pace since 2009 in early December, 2011. While the slowing rate of decline is encouraging, it is still too early to tell whether it will lead to lasting stability in home prices any time soon.

"Frankly, I don't think we've reached the bottom in housing prices." said Quinn Eddins, Director of Research at Radar Logic Incorporated. "The fact is there is still too much supply in the housing market for the current level of demand, particularly if you consider homes in the foreclosure process and those under water. At very least the excess supply will delay the recovery in housing prices, and could well push prices lower."

This graph shows the year-over-year decline for the RadarLogic composite index.

From RadarLogic:

While the slowing rate of price decline is promising, it is too early to say yet whether housing prices will find a bottom soon. After all, we saw price declines slow in 2009, only to see them start accelerating again in 2010.

The third graph shows the RPX futures for house prices.

From RadarLogic:

Exhibit 8 shows historical RPX Composite prices plotted with RPX futures prices. The historical RPX prices are plotted according to their publication date (Radar Logic publishes its daily prices 63 days after the last day in the transaction period) and RPX futures prices are plotted according to their settlement date. The term structure of RPX futures prices indicates that home prices are expected to increasing at an accelerating rate from 2012 through 2015.

Investors think prices will bottom soon, but any increase will be sluggish.