Tuesday, September 21, 2010

Residential construction still in a bottoming formation


August housing starts came in a bit higher than expected, and remain consistent with a view that residential construction bottomed in the second quarter of last year, but has since made little progress. (Housing starts are the green line in the above chart.) Always on the lookout for market-based indicators that may lead the numbers put out by government agencies, I note that the Bloomberg index of 17 major homebuilder stocks (red line in the above chart) may fit the bill. It bottomed in the first quarter of last year and has spent about 18 months consolidating. It also predicted the modest slump in housing starts in May-July, and the modest upturn in August. Millions of investors on the ground and crunching numbers may prove better at divining the course of the housing industry than the folks at the U.S. Census Bureau who put out the housing starts number each month.

At the very least, I would venture to say that with housing starts and homebuilders' stocks failing to reach new lows after hitting bottom well over a year ago, one can say with some degree of confidence that we have seen the worst of the housing recession. More and more the issue becomes the timing and the strength of the recovery.

UPDATE: Here is a long-term chart of housing starts. Note how the past two years have seen the most severe drop in residential construction in recorded history, and the weakness has persisted far longer than in any prior recession. The good news is that once the excess inventory of homes is depleted, the rebound in construction—which does not appear imminent, but should be starting within the next 6-9 months—should be fairly dramatic.

14 comments:

Benjamin said...

Both commercial and residential real estate are in a depression. I am surprised this has not gotten more ink.
With core inflation trending towards negative territory, the Fed has a wide-open playing field to be far more stimulative.

Evil Capitalista said...

Sorry Scott, 6-9 Months is WAAAAAYYY too optimistic. Many areas have current delinquency rates in excess of 25%, plus there's a lot of people that will not be qualified to purchase a home for years to come.

It's going to be more like 6 years before the previous "normal range".

I wish it were different, I really do.

Scott Grannis said...

I don't necessarily disagree with you. But getting back to "normal" in six years would still equate to some pretty impressive growth rates. I think the change on the margin (e.g., growth rates) is more significant than the level of starts.

brodero said...

The single unit housing starts to
nonfarm payrolls ratio has averaged
around .90% for the last 20 years...using .90% as an equilibrium point single unit housing starts should be 1.172 million but we are .430...we have
been underneath the equilibrium point for well over 3 years. By
the 1st quarter of 2011 we will have wiped out the overbuilding starting from the late 1990's.

brodero said...

The single unit housing starts to
nonfarm payrolls ratio has averaged
around .90% for the last 20 years...using .90% as an equilibrium point single unit housing starts should be 1.172 million but we are .430...we have
been underneath the equilibrium point for well over 3 years. By
the 1st quarter of 2011 we will have wiped out the overbuilding starting from the late 1990's.

brodero said...

Redundant don't you think??

marmico said...

Scott The Stopped Clock Grannis. The system is fixed, huh.

The starts data reverses itself next month when the "out-sized" multi-family starts this month falls back to trend next month. Look at Table 3.

I'm really surprised you aren't talking up the NSA data. LOL.

I think the change on the margin (e.g., growth rates) is more significant than the level of starts.

2010Q3 residential investment will be negative. You better set your clock forward another 3 months.

brodero said...

Marmico the 3rd qtr Resi number
will be negative because the second
quarter number was too high....
Residential investment is now only 2.5% of GDP.....damage has already been done...

Benjamin said...

BTW, Zuckerman has op-ed in today's WSJ re housing.

John said...

If you believe Scott's post that the worst is behind us in housing, and if you believe Bob Toll's view that pent up demand is increasing, and you believe the recovery in housing is a matter of timing and degree, one way to profit from being right about those things is to buy the bank stocks. I think they will move upward before we see positive housing numbers (they will lead them). Credit quality continues to improve, the banks are well capitalized (generally...some better than others) and we should see loan demand pick up in coming quarters.

Two suggestions...but DO YOU OWN HOMEWORK. First is XLF, the broadly diversified Spyder Financials ETF. Largest holdings are JP Morgan, Bank of America, Wells Fargo, etc. Second is PJB, the powershares financial ETF comprised of more regional banks like US Bancorp, PNC Financial, etc but still holding JP Morgan and Wells Fargo.

The banks are lagging the overall market but they should play catchup at some time. So if you feel like you've missed the move, this is a sector you might want to consider.

Of course, if you think housing has many years before it recovers, or the bus is heading for the proverbial cliff, or deflation is going to soon increase, ignore this and stay in your zero yielding MM accounts where you will be safe.

XLF - 14.93
PJB - 11.94

marmico said...

the second quarter number was too high

I'm not following you. Q2 is Q2, Q3 is Q3. Too high, too low. I don't get it.

Now on to your model. If nonfarm payrolls are stagnant, shouldn't housing starts be stagnant? Payrolls in 2010 are not materially different than payrolls in 2000. Shouldn't your model correlate the growth rate?

Housing starts were 554k in 2009 and are estimated to be 600k in 2010. I would forecast that housing starts will be higher again in 2011. So what. What matters for the GDP residential investment account is construction put in place. If the value of labor and materials is flat even though starts rise, housing does not contribute to GDP growth. I'm telling you, home builders are building smaller houses which require less labor and materials even though they are building more houses.

As to the so-called inventory over-hang, I would point to you Census data. There are still too many vacant for sale only units. My guesstimate is 1-1.5 million excess inventory units relative to trend. That's alot of inventory to clear before equilibrium.

brodero said...

"If nonfarm payrolls are stagnant, shouldn't housing starts be stagnant? Payrolls in 2010 are not materially different than payrolls in 2000. Shouldn't your model correlate the growth rate?"

I wish i could show you a chart but here goes...Equilibrium is .90%
of nonfarm payrolls...Non farm payrolls is 130,311..so .90% is 1.172 million today we are .430.
An example of extreme overbuilding
would be May 2006...nonfarm payrolls were 135.910...so .90% was
1.223 but we had single unit housing starts at 1.570 and more importantly we were cumulatively way
over equilibrium point for years.
Today we have way way under the equilibrium point....we will stay
under this point for several more years as we work off the excess foreclosed( not rented) inventory.
BUT we will actually see gradual improvement in housing starts because some areas will demand and need new homes and some areas will
still have excess inventory ( i.e. there will be cheap condos in Las Vegas for years but some other more desirable areas will have a shortage of homes). As far as residential investment look at the nominal numbers for Resi and for GDP and follow them quarter to quarter

brodero said...

Peak overbuilding was January 2006
at 1.823 million when equilibrium
( .90% times nonfarm payrolls) was
1.215 million.

Charles said...

We should see housing starts growing rapidly off of an extremely low base. This means a very small boost to the economy.

Housing is an segmented market. There are places where supply and demand are in balance and as the economy creeps forward more and more regions will see a resumption of normal activity. The unemployed construction workers and the firms that would employ them are not necessarily located in the places that will see a resumption of construction.

Over time the economy heals if people are left alone to solve their problems.