GDP & Housing Revisited: I was Wrong

Last week I presented the GDP and housing data. I said that we could ignore housing data because it registered existing home sales and that selling/buying an existing home was not really producing/consuming and could be ignored.


This was wrong.


The factor I was looking at registers sales of existing homes only when they are government owned (think VA). Otherwise it registers new home sales and the investment made in those homes. That would also include upgrades to an existing home (new kitchens, for example).


As such, it absolutely reflect something produced and needs to be judged on those terms.

I apologize and would like to revisit the data and see what it tells us.

For the period Q1 03 ~ Q1 06, residential investment was additive to GDP. Generally adding 0.3% to the GDP. Also note that GDP has been bouncing around 3% for 3 years (Q4 05 being an anomaly caused by Hurricane Katrina).

The additive nature of housing reversed course Q2 06 and started to reduce GDP and at an accelerated rate:
Q2 06 -0.1%
Q3 06 -0.6%

Q4 06 -1%

Q1 07 -1.1%

At the same time, the rest of the economy has accelerated. It raced up from 2.8% to 4.1% in Q4 2006. It has eased a bit this last quarter to 3.2%, but that pace is still higher than much of 2004 & 2005.


To summarize: while housing is decelerating greally fast, the rest of the economy is accelerating.

Or is it? Before we break out the champagne, it is important to note that a lot of that growth is really due to higher spending on oil/electricity, food and medical expenses. All other items tend to be flattening out.

So we are facing an economy where residential investment has worsened far beyond a recession and the rest of the economy is growing at the expense of discretionary spending. Because food, energy and medical expenditures are necessities in the real scheme of things.

First Housing. The housing factor is a simple measure of price and volume. The media is filled with news reports that median home prices are flat across the country and that is misleading. The statistic does not weight different cities according to population: New York's housing price is factored as equal to some small Mid West town with a population of 50,000. And that's simply not right.
If we are interested in understanding what the majority of Americans will see, lets look at the major urban areas. For Q4 2006, according to NAR http://www.realtor.org/Research.nsf/files/MSAPRICESF.pdf/$FILE/MSAPRICESF.pdf :
Atlanta -2%
Boston -2.4%
Chicago 0.9%
Detroit - 1%
Dallas/FW -3.9%
Denver -0.9%
Honolulu 0%
Houston 1.6%
Kansas City - 2%
Las Vegas - 0.8%
Memphis - 1.6%
Miami - 6.2%
New Orleans -9.3%
New York 2.3%
Phoenix -2.3%
Reno -8.9%
San Francisco 2%
San Jose 2.7%
San Diego -4.5%
OC -1.3%
LA 3.2%

New York, LA and the Bay Area seem to be the only major urban areas that did not see price drops in Q4 2006. Meanwhile the Q1 2007 report won't be released until May 15th, but we know things are worse.
I look at the NAR report for directional information because the actual degree of drop is probably more severe. For example, home builders are reporting cancellation rates as high as 40% but the NAR data reports the original sale. Also, home sales reports lag current market conditions by at least 3 months. the Q4 2006 data actually reports on late summer and early Fall sales.

Whatever the case, prices were heading down before the subprime collapse of the last 4 months. Prices will continue the downward trend.

Returning to the GDP, if price is not the main cause of the drop in the residential figure, then it must be volume. We have heard from homebuilders that sales volume is down 30% on average. The current figures show only a 20% drop in value. So if prices are flat to falling and volume is dropping 30% for several quarters, that implies that the GDP figure has a lot further to fall. In fact, the numbers imply that we are heading for a figure around $300B. that's a drop of over 50%.

A further 20% drop will create a recession. I expect the first negative GDP figures to be in Q4 2006, which we won't know about until January/february 2008.

Further acclerating the drop will be massive blue collar layoffs. In Jan 2003, 6.7M workers were doing construction. http://data.bls.gov/cgi-bin/surveymost?ce That number rose to 7.72M in August 2006 and currently sits at 7.68M. About 1M construction jobs were added in 4 years. Given that we are at 2003 construction levels, that implies around 1M workers will be laid off.

I think manufacturing jobs are less exposed for two reasons. First, they have been dropping regardless of housing. Manufacturing did not grow over this time but shrank by 800K workers. Since 1999, the US has shed some 3.5M manufacturing jobs. Second, the layoffs for home supply manufacturing started last summer.

For anyone who thinks that jobs will erode slowly, I point to this week's layoffs by New Century: 2,000 employees fired without pay.

1 million people not spending money will be felt in various places. A lot of vacations won't happen (Disneyland, Carribean cruises). A lot of toys won't be bought (Harley Davidsons, Jet skis, boats, trucks, TVs). A lot of remodeling won't happen (Home Depot, Lowes). A lot more meals will be eaten at home (Safeway will be looking strong, TGI Fridays and other spots will get hit).

Another major trend will be home price erosion. I think that white collar professionals are about to get stung. I mean the financially overstretched investor who speculated on homes. And a lot of folks who bought more house than they could really afford. I think some 100,000 people will be affected and unable to spend as freely.

Every recession is different. This one will be a blue collar recession coupled with consumer spending slowdown. Against the background of a looming recession, the Fed must contend with inflation. Remember that food and energy prices are rising.

When will interest rates come down? Put differently, is the Fed interested in stimulating the economy or in keeping prices predictable? A drop in interest rates fuels inflation but stimulates the economy. I would imagine that they will make a token drop of ~0.25% sometime before September. That will be too little for the housing market.

Lastly, where will home prices end up? An indication is the recent sale by New Century of their remaining mortgage portfolio valued at $170M but purchased for $58M. http://www.latimes.com/business/investing/la-fi-newcentury5may05,1,6951879.story?coll=la-headlines-business-invest&ctrack=1&cset=true

That's a 66% price drop.

Even if they hope for a 100% gain on their investment, that's a 30% haircut in value. A deep pocket hedge fund can go years. The average Joe can't. So I expect a deeper cut in home prices.