by Calculated Risk on 9/27/2007 03:03:00 PM

Professor Roubini write: I Was Way Too Optimistic on the Housing Recession...

In March of this year this author published a long paper (with Christian Menegatti) titled “The Housing Recession is Still Far from Bottoming Out” that predicted a much worsening housing recession through all of 2008. In that paper the prediction was that housing starts – that had already fallen by 38% by January of 2007 to a level of 1.4 million - would fall much further and would bottom out at 1.1 million in 2008. Indeed, by looking at previous housing recession – where the average fall in starts was 51% - it was sensible to be that pessimistic. Again those predictions were dismissed as too gloomy and pessimistic and unrealistic.

But it turned out that I was way too optimistic about housing, not too pessimistic. As recently reported housing starts have now fallen by 42% and now JP Morgan – one of the most respected research houses on Wall Street and a persistent proponent until recently of the view that the housing recession would bottom out – is predicting that housing starts will fall another 25% to a cumulative fall of 56% from peak and will bottom out at 999 thousand units some time in 2008. Compared to my initial March prediction of a bottom at 1.1 million I turn out to be an optimist. And indeed many other research firms (including Goldman Sachs, Citibank and others) are now predicting the bottom of housing starts at 1 million to 1.1 million units. So what in March was considered as borderline lunatic is now becoming conventional wisdom.
...
Right now I see housing start falling more than the 1.1 million that Menegatti and I predicted in March as the housing market has deteriorated since then much more than our then dire predictions. Given the continued fall in building permits and in home sales and the massive excess supply of new and existing homes housing starts could bottom out at level close to previous housing recession, i.e. between 800k and 900k.

To add to Roubini's comments, here is the current Goldman Sachs housing forecast (New Home sales falling to 650K units and starts falling to 1.1 million units).

Since I predicted a fall in starts to 1.1 million units a couple of years ago, people are now asking if I was also "too optimistic". First, I don't measure my predictions to the WS forecasts, but to the actual numbers. Starts have fallen to 1.331 million units (SAAR) in the latest report, so it is still too early to count coup. Second, I wasn't trying to forecast the exact number, rather the general size of the down turn - IMO there is little difference between 1.1 million starts and, say, 1.0 or 1.2 million starts. The overall downturn is about the same.

However Roubini's new bottom forecast for starts (800K to 900K) is a significant downward revision from his previous forecast. I'll update my forecast at the end of the year (the housing market moves in slow motion, so there is no need to change forecasts frequently). Right now I think my number still looks about right.