November 19, 2003

Good News on Residential Construction

Good news on residential construction:

Housing starts explode, set fastest pace since '86 - Nov. 19, 2003: New home construction in the United States accelerated in October to the fastest pace in nearly 18 years, the government said Wednesday, defying Wall Street forecasts for a slowdown. And the ever-rising flood of new homes appeared to be supported by demand, some economists said, indicating home prices were unlikely to be artificially inflated into a dangerous "bubble."

The Commerce Department said the pace of housing starts rose 2.9 percent to a seasonally adjusted annual rate of 1.96 million units, after rising a revised 4 percent to 1.91 million units in September. Economists, on average, expected housing starts to fall to a 1.85 million-unit pace, according to

Posted by DeLong at November 19, 2003 08:06 AM | TrackBack


Massive starts, and permits running even faster, so November starts (which benefit from a bigger lift from seasonal factors) will probably be pretty impressive, too. Meanwhile, weekly mortgage applications for purchase rose by 13.5% in the latest week, and housing completions were revised higher in September, more fuel for an upward revision to Q3 GDP. And me without a house in fee simple.

Not to be snotty, but for those who have grown suspicious of the regular pattern of upward revisions to weekly jobless claims data, it is worth noting that the permits data have routinely been revised upward in recent months. In addition, starts for September were revised higher. I realize that employment data are far higher profile than housing data, but I would also point out that monthly employment data - far higher profile than weekly claims - were also most recently revised higher. If there is hanky-panky in the first release of the data, it seems isolated in the weekly claims - an odd choice.

Posted by: K Harris on November 19, 2003 08:49 AM


CNN/Money is giving a logical explanation for the rise in housing demand - a recent fall in mortgage rates. But how do we explain interest rates declining?

Posted by: Hal McClure on November 19, 2003 09:23 AM


CNN/Money is giving a logical explanation for the rise in housing demand - a recent fall in mortgage rates. But how do we explain interest rates declining?

Posted by: Hal McClure on November 19, 2003 09:26 AM


Housing activity continues to surprise on the upside. I don't get it, I would have expected the housing market to slow significantly after rates shot back up in July/August. But it's several months later now and there isn't much sign of a let up yet. A correction seems inevitable after the runup over the last five years or so.

Posted by: Joe Blog on November 19, 2003 10:21 AM


Are the housing market revisions enough to point to an upward Q3 GDP revision? The revision to September starts is +0.9% (+17,000 units, annualized), and the revisions increase the 3-month average for Q3 by +0.4% (+7,333 units, annualized; there was a trivial upward adjustment for August as well). How does this effect compare to that of, say, the downward revision to September retail sales? (I did take the confidence interval on the preliminary October retail sales change as a mild antidepressant.)

The MBA purchase index seems to be subject to fairly substantial week-to-week swings, seasonal adjustment notwithstanding, so I'm not that impressed by the one-week change. Their Q4 data don't seem to suggest much if any of a pickup over Q3 so far. I also wonder about the increase in the ARM share. It isn't in early-2000 territory, admittedly, but neither is the probable direction of change in mortgage rates.

So the level of activity impresses, but sustaining growth from these levels, if it is to be achieved, will seem to require a fine balancing act involving increasing employment and incomes and not so rapidly increasing mortgage rates.

Posted by: Tom Bozzo on November 19, 2003 12:08 PM


Hal, the answer is savings over investment :).

I've tried twice in the last week to plug a theory by Richard Koo which seems to have some explanatory value. I've elaborated his theory most thoroughly here:

Both times there have been no takers on slay the theory. Third times a charm?

Posted by: Stan on November 19, 2003 02:19 PM



Reference to this Koo thesis? Maybe I'm just an old fashion perfect markets (as in Modigliani and Miller) type but I have always looked at the decision to invest in new capital as pure expected return v. risk considerations without regard to how much of the assets were financed by debt. After all, these companies choose their debt levels. But then I know there are challenges to perfect capital market thinking. So I'd love to read the Koo thesis.

Posted by: Hal McClure on November 19, 2003 02:47 PM



The starts data won't have much to do with GDP. You need to look at completions. They were revised higher for September in today's release, probably not enough for a 0.1% impact on GDP. However, there were already upward revisions to prior months' construction spending data, and September came in at +1.3%, so construction spending will give a lift to the GDP revision, with today's revisions completions just nailing down the case. As regards whether MBA data point to further gains - I would not worry to much about gains now. We are about to go into the "big seasonal" season for construction data, when small changes to actually activity are magnified by seasonal adjustment factors. The point is that forward-looking housing data (permits and mortgage applications) remain very strong.

Posted by: K Harris on November 19, 2003 02:59 PM


K Harris, thanks for the additional discussion of the data. I largely agree with your main point -- the question in my mind is how far forward the good news applies and can be sustained.

Hal, the Fed expectations post sort of covers my response to your question about interest rates. The bond market's (apparent) responses to the Fed's deflation watch and subsequent tea leaf reading as to how long the Fed will stay on it seem to explain at least the broad timing and directions of the last year's mortgage rate fluctuations reasonably well.

I have a couple of objections to the balance sheet recession story that Stan describes. I'm skeptical about the assumed change in the behavioral motivation for firms. I don't see a necessary contradiction between profit maximization (or the appropriate optimizing behavior) and firms repairing their balance sheets in down times. I also don't see why accommodative monetary policy wouldn't tend to facilitate balance sheet repairs among other effects. I also note this post:
where an implication of Brad's comments is that it matters, at least for the long haul, that while some bubble-era investments were valued delusionally, the assets are not actually technically obsolete.

Posted by: Tom Bozzo on November 19, 2003 10:09 PM


Hal, I agree scepticism is healthy. In this instance, Koo's defense in Japan rests a lot on a bank survey that supposedly shows a general willingness to lend through most of the last 10 years. I have no idea how accurate that survey is.

Posted by: Stan on November 20, 2003 09:14 AM


Tom, it has to do with the shape of the demand curve for investment. Under the basic profit maximizer assumption, investment demand increases as interest rates decline. Under this scenario, investment demand elasticity is very low. The elasticity would be low until enough companies balance sheets had improved.

Such a condition wouldn't make accomodative monetary policy a bad idea, it would merely make it ineffective at stimulating demand for a time. Assuming high unemployment, fiscal policy would be the only effective mechanism to increase short term demand.

Posted by: Stan on November 20, 2003 02:37 PM


In his errors a man is true to type. Observe the errors and you will know the man.

Posted by: Stern Caroline on December 10, 2003 04:04 AM


Reality is not affected by our apprehension of it.

Posted by: Bumbray Rashida on December 20, 2003 03:40 PM


To be poor without bitterness is easy; to be rich without arrogance is hard.

Posted by: Lowther Tori on January 9, 2004 03:47 AM


Post a comment