by Calculated Risk on 7/29/2007 05:35:00 PM

This is an update to my 2007 housing forecast.

Probably the most important housing numbers are sales (new and existing homes), prices, and housing starts.

Starting with Existing Homes, my original forecast was for sales of 5.6 to 5.8 million units in 2007. To date, through June, the NAR has reported sales of 2.929 million units.

Lending standards are being tightened again, and this will most likely impact sales in the 2nd half of 2007. Since existing home sales are reported at the close of escrow, many of the July and August sales were in process before the most recent round of tightening - so I've assumed only a minor impact on reported sales for July and August (as compared to 2005 and 2006). I've assumed a greater impact on sales for the last four months of the year.

Click on graph for larger image.

This graph shows my current month by month forecast for existing home sales this year (red is actual for 2007). It is possible that sales will fall off a cliff later this year, but that kind of decline is extremely difficult to predict, and I think a more gradual decline is likely.

My current forecast is for 5.6 million units; at the bottom end of my original forecast range.

For anyone interested, you can enter our contest: Forecast 2007 Existing Home Sales. PLEASE enter your forecast in the comments to the contest thread (not in the comments here) so I know where to look in January. Thank you!

For New Home sales, my original forecast was that sales would "surprise to the downside" compared to Fannie Mae's forecast of 975 thousand units. Through June, the Census Bureau has reported sales of 458 thousand units.

Sales are reported for New Homes when the contract is signed, so the recent tightening of lending standards will probably impact reported New Home activity sooner than reported existing home activity. My current forecast for New Home sales is for a total of 830 to 850 thousand units.

For existing home prices, my forecast was for prices to decline "nominally by 1% to 3% nationwide by all measures (OFHEO, NAR)." So far prices are up slightly according to the NAR:

The national median existing-home price for all housing types was $230,100 in June, up 0.3 percent from June 2006 when the median was $229,300.

Down slightly according to the FHFB.

The FHFB reported the decline in the average price was $501, or 0.16 percent, from $306,759 in October 2005 to $306,258 in October 2006. This is the first decline in the MIRS since 1992-93.

And up in the first quarter according to the OFHEO (the second quarter OFHEO House Price Index is scheduled to be released on August 30).

The OFHEO House Price Index (HPI), which is based on data from sales and refinance transactions, was 0.5 percent higher in the first quarter than in the fourth quarter of 2006.

All of these methods have flaws and are based on nationwide price changes. I'll stick with my forecast of a 1% to 3% price decline nationwide in 2007.

Housing Starts. In my 2007 housing forecast, I didn't forecast starts. However, last year I calculated that completions might fall to 1.2 million units per year, and starts might fall even further to 1.1 million units per year. It now appears starts will be in the 1.4+ million range for 2007. What I overlooked was that the builders would get caught in a self destructive Nash equilibrium: the homebuilders have to keep building, because they can't sell the land. And they builders can't gain by changing their own behavior, because the industry is very fragmented and the other builders will keep on building. In the aggregate, this behavior is self destructive, but it makes sense for each individual builder.

I still think starts will decline significantly from the current level to work off the record levels of inventory. And it's the total inventory (new and existing homes) that matters: as an example, see this story from the WaPo yesterday: Glut of Condos Pits Private Sellers Against Developers.

I'll have more on starts, inventory levels, and the excess housing stock in the near future.