The chart below tells you all you need to know about the effect of the housing market on the US economy.

Housing starts have averaged 1.487 million units annually for the past 50 years.  They have normally recovered quite quickly after recessions.  We have so far looked at 3 possible reasons to why housing starts are dragging along the bottom about 60% lower than the average annual rate of the last fifty years.

First we saw that banks are damaged and their capital ratios are being scrutinized.  Because of balance sheet impairment the availability of credit to purchase a house is limited.

Second we looked at the household sector balance sheet and income statement to find whether citizens had the ability to purchase real estate.  We concluded that households in aggregate were not stressed currently.  They do have the ability to purchase a new home but perhaps not the willingness.

This week let’s explore the existence of a disconnect between supply and demand in the housing market.  Where are inventory levels relative to the past?  Is there pent up demand because multiple households are consolidating under one roof?

Housing is both an expense and an investment.  As we saw last week, real estate is the second largest asset on household balance sheets.  Housing is also a basic need (food, water and shelter) for human existence.  The only reason financial assets are worth more than houses is that our economic success has given us a standard of living above basic sustenance.  We have been afforded the luxury to invest our hard earned dollars in areas beyond just feeding and housing our families.  This exercise is based on the premise that shelter is senior to financial assets.  People will sell financial assets in favor of housing their family.

We therefore have only 2 scenarios.  The inventory situation may be such that shelter is ubiquitous.  The downward trajectory of prices will still be intact as holders of vacant real estate continue to sell until a clearing price is reached.  This will further damage balance sheets and deleveraging will continue.  In scenario 2, supply and demand may have reached equilibrium levels and perhaps demand may begin to outweigh supply.  We would expect to see some capital come out of financial assets in order to invest in shelter/housing.  A clearing price for excess inventory will have been reached; prices will bottom and may turn north.  Production can return to normal levels which means new construction spending increases and jobs are produced.  We would expect sales to increase in this scenario, but an impaired banking system would become a hindrance here.

We’ll first look at population.  Population growth is the main feeder of demand for housing.  Let’s also peek at household formation.

Looking at the above graph I find a few things to note.  The increase in household formation in 1981 can probably be attributed to banking deregulation passed in 1980 which allowed more freedom for banks to lend.  Additionally, Reagan’s campaign and likely election may have provided some positive sentiment.

The spike in working age population during the late 90’s is most likely an influx of workers, possibly foreign, working in our booming internet economy.  The labor participation rate hit an all-time high in 2000. The combination of the .com bust and the attacks of 9/11/2001 sure look to have affected household formation in 2002.  No one was buying houses and soldiers were being shipped off to Afghanistan and Iraq.  Household formation definitely correlates with working population.  As our country and baby boomers age, working population growth rates are declining at a rate of 2.07% per year.  Household formation has followed.

Household formation has actually done worse; the linear trend line shows fresh demand for houses declining at a rate of 2.74%.  A few cautions here.  This data can be deceiving as recent softness in demand as a result of recent events may be adversely affecting the trend line which may flatten during the coming years.  Regardless, the trend is definitely lower.  Altering the slope of household formation to match that of working age population growth would add about 193,000 households / year, a level of 793,000.  This is still well below the mean of 1,194,000 for the last 40 years.

Is construction in line with these demographic trends?  Let’s overlay housing starts onto our graph.  It does appear that the construction sector has been on a very optimistic trend and has potentially been over building since the early 1990’s.  The most likely reality is that the capital destruction of the past decade has severely hurt the country.  This not only impairs banks ability to lend, but also citizens ability to buy.  The culprits are leveraged buying of inflated dot com stocks, losses of physical property and morale from 9/11, the expense of waging 2 wars, wasteful government spending, losses of equity and credit on over-priced houses, subsequent loses on financial assets and the continuous dilution of our saved earnings.

The CEO of any private company would be ousted immediately for such poor stewardship.  In many eyes the transgressions here of our leaders are so erroneous, they border on criminal.

From this view, I do not see construction returning to the mean anytime soon.  I suppose we may trend back to the 800,000 unit level; but it is doubtful we’ll see 1.2mln+ anytime in the next three years.

To confirm or disprove our pessimism we need to look at inventory levels.  I will disregard claims of unaccounted shadow inventory withheld from the market  as there is offsetting shadow demand from out of work 20-30 year olds still living with mom and dad.

Housing is either vacant or occupied.

Historic rates of vacancy run about 10.98%.  Current vacancy is sitting at 14.26%, too high.  Were inventory levels relative to occupancy at long run average rates, total housing units would need to be 126,345,000, lower by 4.83mln units.  Alternatively, you could assume occupancy to be too low.  Equilibrium puts vacancy at 14.4mln meaning excess inventory of 4.3mln units.   The US has formed about 571,000 households per year for the past decade.  At this rate and no additional construction, inventory would clear in 7.5-8.5 years.  Looking at the past 40 years, we have formed 1,194,000 households annually.  A return to this level of household formation and building no new houses, would clear inventory in 3.6-4 years.  Let’s hope much of this inventory is ready for the bulldozer.

Vacant housing is either vacant year round or classified as seasonal (2nd homes).

Inventory of 2nd homes is currently within 15 basis points of average for the last 40 years.  Of the 4.3-4.8mln units of excess inventory about 24.4% or 1.1mln have become 2nd homes by default.  Whether these properties re-emerge on the market is anyone’s guess.  If these properties are passed in estates or the market for seasonal properties strengthens they could emerge as for sale properties.  For now they remain on household balance sheet as assets.

We are left with an excess of 3.2-3.7mln properties classified as vacant year round.  Components of year round vacant property are For Sale, For Rent, Sale Pending or Held off Market.

In this graph we get some interesting macroeconomic insights.  Look at the trends of for sale and pending transaction properties.  The selection of houses available to purchase has been slowly rising.  Economically, this could be attributed to product differentiation; perhaps the housing industry thought they were being innovative; perhaps the developers thought Americans were becoming more mobile and sales volume would rise, justifying adding to selection.  But as the trend in pending transactions indicates, this has not come to fruition.  Internet communication has grounded the relocation process.  High real estate transaction fees make it imprudent to move often.

Next Rent v. Buy:   We can see households forming in rental properties over the past 7 qtrs.  This shift in housing preferences follows the stock market plunge of late ’08, early ’09.  Since this time, householders have eaten up about 14.75% of the inventory of rental properties.   We can also see the trends in a rent v. own graph.

During the Clinton administration, ownership rates rose from 64 to 67%, mainly attributed to lending mandates on Fannie Mae and Freddie Mac under the Community Reinvestment Act in an effort to reform inner cities.  The trend continued, hitting 69% during the Bush years as US economic strength carried the day in the late 90’s followed by no lending standards and easy money in the early part of the new century.  Seemingly fruitful for close to 1.5 decades these policies were unsustainable and we have returned to within 1% of the historic 65/35 ratio.

Beginning to conclude, if housing were at historic average levels of inventory we would expect the following.

Excess inventory of 4.3-4.8mln units falls into the following categories.




Inspecting the Held off the market inventory we have 3 components.  Occasional Use and Usual Residence Elsewhere properties would be offices in residential space, storage properties and dilapidated housing.  These components don’t seem to be straying from the norm.  Alas, the final category is “other”.  Foreclosed upon homes literally sitting on bank, GSE and personal balance sheets, rotting.  At median price levels, this is $450bln in wasted capital classified by our government as “other.”

If household formation does not pick up these 2mln units just depreciate with no current income to replace the decaying capital investment.  To put this in perspective, the combined equity capital of our 5 biggest banks at the end of last quarter was about $789bln.  Warren Buffet just gave Bank of America, the country’s biggest bank, $5 bln.  Leveraged investment gone bad destroys one’s net worth.  With current excess inventory running at 4x long-run, annual demand, I can’t help but be a bit pessimistic.

Unless we see a release of pent up demand take household formation back above 800,000 I would not expect construction to rebound.  Current levels of employment are not supportive of this scenario.  Bank problems are every day’s headline.  Moderated levels of portfolio risk are the conclusions.

If you’d like to receive this free weekly commentary by email, click here.

Print Friendly, PDF & Email