Jump in housing starts points to recovery

[bmichaud]

So let's get down to brass tacks shall we?

 

The numbers on housing completions per year (in thousands): 

 

2001 1,570.8

2002 1,648.4

2003 1,678.7

2004 1,841.9

2005 1,931.4

2006 1,979.4

2007 1,502.8

2008 1,119.7

2009 794.4

2010 651.7

 

source: http://www.census.gov/construction/nrc/pdf/compann.pdf

 

14,713 built across 10 years.

Then add in 600k estimated for 2011, and you have 15,313 built in past 11 years.

 

That's 1.39 million housing units completed per year.

 

How does anyone justify the claim that we have an oversupply because we built too many?  It looks to me like we merely have an unemployment problem (and Buffett thinks it's because of the unsustainable slump in homebuilding).

 

From 2001 through 2007, completions averaged 1736. Based on historical numbers, let's call the long run average is 1500 per year. So we over built by 236 per year for seven years, for total overbuild of 1653. From 2008 through 2011e, we built 3166 in total, or 792 per year. For those four years we've underbuilt by 2832 cumulatively, which would indicate we've more than made up for the excess 1653. If we assume the long run average is 1350 per year, we overbuilt by 2702 over those seven years, and have under built by 2232 over the past four years. Assuming the second scenario, we are not quite there yet.

 

More just thinking out loud for my own sake. Household formations and a more normal level of home ownership will affect all of this, obviously.

[ERICOPOLY]

If we assume the long run average is 1350 per year, we overbuilt by 2702 over those seven years, and have under built by 2232 over the past four years. Assuming the second scenario, we are not quite there yet.

 

Where did you pull the number 1350 from?

[bmichaud]

If we assume the long run average is 1350 per year, we overbuilt by 2702 over those seven years, and have under built by 2232 over the past four years. Assuming the second scenario, we are not quite there yet.

 

Where did you pull the number 1350 from?

 

The average from 1990 to 2000.

[ERICOPOLY]

My last post claimed that 1.3m net new households were formed per year over the 10 years ending in March 2010 (despite the ongoing joblessness associated with depressed home-building). 

 

We have (over 10 years ending in Dec 2009):

1.562m housing units built per year

1.3m households formed per year

 

That creates an excess of 262k per year before accounting for destruction of housing units.

 

I've read before that destruction averages about 300k per year.  262k per year is probably close enough to average that I don't want to get too precise about it, but if anything it argues that maybe we've moved towards an actual housing deficit during the 10 year period of 2000-2009.

[bmichaud]

Perhaps we are in a deficit, but you could also say home ownership rates are just now getting back to historical norms (66% now versus 69% at the peak). If formations get back to a normalized rate, then a near term bottom would make sense. The cynic in me says markets overshoot both on the upside and downside, and given the hideous nature of deleveraging environments, perhaps household ownership rates overshoot down to let's say 63%. If I was investing in ownership rates, I'd say we're at fair value.

[ERICOPOLY]

What's the importance of ownership rates? 

 

A household needs a place to live.  It's merely paperwork whether they own it themselves or rent it from an investor.

[Guest misterstockwell]

What about >90 day delinquent and foreclosed homes--those are eventually "supply". That number is closer to 4 million.

 

Unless they are all going to crowd into a tent, then we are either going to see:

 

a)  4 million new single family home renters  (just rearranging the deck chairs isn't it?)

b)  4 million new apartments built  (lots of new jobs!)

c)  A mix of both  (some deck chairs shuffled, some new jobs too)

 

I'm guessing it's c.

 

Unfortunately, we don't know what percent of those are occupied. Some will be failed second homes, specs, etc. Other folks will extend the family in a single home, move back with parents, move back to their homeland. It definitely won't be one for one. As bmichaud points out, our long term average of 65% home ownership went wonky, fueled by easy money. The 4% added homeowners aren't coming back. Their homes won't get occupied. Building apartments is unfortunately not a sign of strength.

[bmichaud]

IMO there is a big economic difference between living in a house and renting an apartment. It's not as simple as a household either rents or owns a house. The "stuff" required to fill a house is much greater, as is the energy required to run a house, for example. Yes more apartments need to be built, but even that is not the economic equivalent to building a like number of houses to house the households seeking shelter. In theory, three households could live in a 3000sf multi-family unit broken down into three apartment units.

 

So if home ownership rates fall to 63%, that's 3% generating less than normal economic housing activity.

 

Let's not forget either that the government is running 10% deficits right now. Thats a massive boost to the economy. Think about where we'd be if we'd have run 5% deficits for the past three years let alone a balanced budget. even if housing gets back to normal, contractionary fiscal policy will be working hard to offset that.

[ERICOPOLY]

IMO there is a big economic difference between living in a house and renting an apartment.

 

I agree.  Note that we're agreeing on a completely obvious thing, which is a good start  :D

 

It's not as simple as a household either rents or owns a house.  The "stuff" required to fill a house is much greater, as is the energy required to run a house, for example.

 

It's not that simple?  Oh yes it is.  I own my own house and I have it filled with furniture and pay the heating bill.

 

So does my neighbor (has it filled), except that he is renting the house.  Both houses are the same size.  Their house is every bit as full of furniture as mine.

 

Don't you reckon the houses across America going into foreclosure are being bought by investors (if not owner-occupiers) who wish to rent them out?  I presume that's what they intend to do with them.  If not, then what?

 

 

 

[ERICOPOLY]

So if home ownership rates fall to 63%, that's 3% generating less than normal economic housing activity.

 

100% of homes are owned. 

 

That's my point.

 

[Guest misterstockwell]

Foreclosure starts are 3X foreclosure sales, so, no, people are not buying houses and renting them out. Banks own them. Mortgage pools own them. In some cases, nobody has a clue who owns them.

[ERICOPOLY]

Foreclosure starts are 3X foreclosure sales, so, no, people are not buying houses and renting them out. Banks own them. Mortgage pools own them. In some cases, nobody has a clue who owns them.

 

BofA reports that investors are briskly snapping them up:

 

http://online.wsj.com/article/SB10001424052970204323904577039930179775556.html

 

[treasurehunt]

What about >90 day delinquent and foreclosed homes--those are eventually "supply". That number is closer to 4 million.

 

Unless they are all going to crowd into a tent, then we are either going to see:

 

a)  4 million new single family home renters  (just rearranging the deck chairs isn't it?)

b)  4 million new apartments built  (lots of new jobs!)

c)  A mix of both  (some deck chairs shuffled, some new jobs too)

 

I'm guessing it's c.

 

Unfortunately, we don't know what percent of those are occupied. Some will be failed second homes, specs, etc. Other folks will extend the family in a single home, move back with parents, move back to their homeland. It definitely won't be one for one. As bmichaud points out, our long term average of 65% home ownership went wonky, fueled by easy money. The 4% added homeowners aren't coming back. Their homes won't get occupied. Building apartments is unfortunately not a sign of strength.

 

The percent that is not occupied is already counted in the supply stats, so why is it important to know this percentage?

 

Homeownership isn't directly relevant to the total oversupply either. Assuming that owners and renters typically live in different kinds of residences, it might mean that the oversupply is more in detached homes and less in apartments.

 

Btw, one of Tom Lawler's articles that I linked to in a previous post indicates that the homeownership rate is back to 1990 levels (he compares the data from the 1990 census and the 2010 census; most articles in the press are still using faulty CPS/HVS data for homeownership). In fact, Lawler says that if you look at homeownership by age group, homeownership for most age groups was lower in 2010 than in 1990. So I don't see why it should decline too much from here.

 

I think folks are way too pessimistic about the future of residential construction in the US. It seems to me that there is a lot of data to support Buffett's view that a turn in residential construction and related employment could happen soon.

 

 

[bmichaud]

IMO there is a big economic difference between living in a house and renting an apartment.

 

I agree.  Note that we're agreeing on a completely obvious thing, which is a good start  :D

 

It's not as simple as a household either rents or owns a house.  The "stuff" required to fill a house is much greater, as is the energy required to run a house, for example.

 

It's not that simple?  Oh yes it is.  I own my own house and I have it filled with furniture and pay the heating bill.

 

So does my neighbor (has it filled), except that he is renting the house.  Both houses are the same size.  Their house is every bit as full of furniture as mine.

 

Don't you reckon the houses across America going into foreclosure are being bought by investors (if not owner-occupiers) who wish to rent them out?  I presume that's what they intend to do with them.  If not, then what?

 

Those households who make up the excess home ownership rate, I don't think the majority are renting out homes. Perhaps they are....

[Guest misterstockwell]

Foreclosure starts are 3X foreclosure sales, so, no, people are not buying houses and renting them out. Banks own them. Mortgage pools own them. In some cases, nobody has a clue who owns them.

 

BofA reports that investors are briskly snapping them up:

 

http://online.wsj.com/article/SB10001424052970204323904577039930179775556.html

 

LPS, an unbiased source, says foreclosure starts are 3X sales.That leaves a great many unsnapped.

[ERICOPOLY]

Foreclosure starts are 3X foreclosure sales, so, no, people are not buying houses and renting them out. Banks own them. Mortgage pools own them. In some cases, nobody has a clue who owns them.

 

BofA reports that investors are briskly snapping them up:

 

http://online.wsj.com/article/SB10001424052970204323904577039930179775556.html

 

LPS, an unbiased source, says foreclosure starts are 3X sales.That leaves a great many unsnapped.

 

You have to read close as to what the two sources say.

 

BofA:

"Where you've had the ability to get ahold of properties ... get them cleaned up, back on the market, they have moved," he said. "It moves as fast now as it's ever moved."

 

In states like Florida, "where that process has been slower, it's going to take a lot longer," he said.

 

 

LPS:

. The slowdown is most pronounced in judicial foreclosure states, which maintain a foreclosure and seriously delinquent pipeline that is more than three times as long as non-judicial states.

 

 

 

[Guest misterstockwell]

No matter how you slice it, the results are the same.

[ERICOPOLY]

No matter how you slice it, the results are the same.

 

We aren't seeing eye to eye on what the results really are.

 

[bmichaud]

Interesting chart on page seven here: https://www.wellsfargo.com/downloads/pdf/com/research/market_strategy/2012-Economic-Outlook_12072011.pdf

 

Shows median home size versus average household size.

[ERICOPOLY]

Interesting chart on page seven here: https://www.wellsfargo.com/downloads/pdf/com/research/market_strategy/2012-Economic-Outlook_12072011.pdf

 

Shows median home size versus average household size.

 

Yeah, household sizes spike during recessions.

 

We're currently building smaller single family homes and apartments -- I'd say the adjustment has been made already.  Homebuilders aren't turning out homes that don't sell -> look at the record low inventories of new construction.

 

[shalab]

And population is increasing. It is up by 2.2 million from July 2010 to June 2011. At this rate, U.S will add one Australia and 2/3 of Canada this decade. It will equate to a GDP of one trillion dollars.

 

http://blogs.wsj.com/economics/2011/12/22/vital-signs-slowing-u-s-population-growth/

 

 

[ERICOPOLY]

And population is increasing. It is up by 2.2 million from July 2010 to June 2011. At this rate, U.S will add one Australia and 2/3 of Canada this decade. It will equate to a GDP of one trillion dollars.

 

http://blogs.wsj.com/economics/2011/12/22/vital-signs-slowing-u-s-population-growth/

 

That pace suggests 846,000 new households formed per annum if there are 2.6 per household. 

 

Very bad news -- time to let in some immigrants perhaps.

[ERICOPOLY]

Is there an official count anywhere (Census data?) of the number of homes destroyed each year?

 

CBO estimates it at 250,000:

http://www.cbo.gov/ftpdocs/98xx/doc9885/11-17-HousingStarts.pdf

 

John Mauldin reports it's about 300,000:

 

About 300,000 of these units are offset each year by dilapidated houses that are torn down, houses converted to nonresidential purposes and other factors that remove them from the housing stock.

 

Some other website said 360,000:

 

http://www.financialsensearchive.com/fsu/editorials/jain/2005/0731.html

[Rabbitisrich]

Ericopoly, I don't think that destroyed units are surveyed. Calculated Risk compares year to year changes in housing stock surveys (ACS or CPS/HVS) to Census surveys of new home completions. The difference is the implied units destroyed.

 

http://www.calculatedriskblog.com/2011/12/american-community-survey-and-total.html

 

The link above discusses a problem with the ACS survey (CR and Tom Lawler have posted about problems with the CPS/HVS):

 

However, in April 2010, the decennial Census showed significantly more housing units than the ACS had captured (obviously a negative 1.15 million homes weren't demolished in early 2010!)  The decennial Census data itself seems a little off since it suggests only about 645 thousand housing units were demolished during the decade (that would be very low).  Most estimates are demolitions are in the 200 to 300 thousand per year range (so the ACS seemed reasonable through the first 9 year of the decade).

 

 

 

[ERICOPOLY]

As bmichaud points out, our long term average of 65% home ownership went wonky, fueled by easy money. The 4% added homeowners aren't coming back. Their homes won't get occupied. Building apartments is unfortunately not a sign of strength.

 

Read em and weep boys.

 

As the boomer generation ages, it pushes up the ownership rate.  This is such a no brainer I should have thought of checking for this sooner.

 

The long term average of 65% would need to be adjusted for this.  I mean, people over age 65 have an 81% ownership rate.

 

Take a look at the census report, page 7:

 

http://www.census.gov/hhes/www/housing/hvs/qtr111/files/q111press.pdf