Wednesday, December 8, 2010
There have been important developments since I wrote on housing statistics last week, so yet another addendum is in order.
The Turkish Statistical Institute, or TurkStat, released the third quarter construction and occupancy permits yesterday. Apartments with construction permits increased by 34 percent during the first three quarters compared to the same period previous year. In contrast, apartments with occupancy permits decreased 20 percent. So we now have all the third quarter housing data, except for construction in GDP, which will be released on Friday. And for now, we have a strong construction, weak occupancy permits as well as weak housing sales data.
For one thing, the weak occupancy permit figures are a follow-up to the weak construction permit numbers of the previous years (you first get the latter to start building, get the former when you are finished). While I cannot do an official test because of the short time span of the data, have a look at the figures and decide for yourself (the permits are year-to-date, meaning that the 3Q figure is the sum of the three quarters, not just the third quarter; home sales are quarterly):
What is worrying me is the discrepancy between the housing sales and construction permits numbers. So in essence, more houses are being built, but less are being sold. Actually, you don't need to look at data to reach that conclusion. Anyone who has spent a couple of days in the country of late could have told you of the construction boom. But with all the housing advertisements and the bullish sentiment, it might be tough to convince the occasional visitor, let alone the skeptical Turk, that home sales are falling.
The only way statistics and common sense come together is by assuming that home sales are more and more shifting towards new homes (we don't have new home sales data in Turkey). That would explain robust loan growth, which would probably translate to the increase in the value of each dwelling sold. The billion-dollar question is how much this could go on. Construction permits for 480,000 apartments have been granted so far this year, and I am not sure there will be enough demand even for new homes when these half-a-million apartments hit the market.
Having said that, I don't think there is a bubble in the making, although there were rumors (article in Turkish) that the IMF is worried about a bubble. Let me explain why: I was exchanging emails with three guys, one economist, one finance professional and one ex-trader, on exactly the same topic on Monday, and as the economist was noting, housing loans are only 5% of GDP, their average maturity is five years and most importantly they are not securitized. In other words, we simply don't have the conditions for a housing bubble. Besides, the Central Bank has been carefully watching over the situation and taking precautions, slowly but surely for a few months now so that paranoiac people like me can tuck in worry-free at night.
But I continue to claim that housing prices are bound to fall because of a temporary supply-demand mismatch. That was the whole point of the Hurriyet column. And when that happens, the ripples will go far beyond housing. For one thing, since the builders borrow from the banks as well, the banking sector will be affected even if there aren't many delinquencies by consumers, but just unsold homes... So the argument that NPLs will not go up does not mean much. Just as a footnote, I got some insider info. on the amount of credit extended to some of the major builders by a major bank a few months ago, and I was left in awe... After all, when you owe the bank 1 million liras, it is your problem, but when you owe them 1 billion liras, it is the bank's problem.
BTW, the Fund already knows all this, so I don't think they are worried about a bubble at all; when they visited GYODER last week, I would assume they were simply on a fact-finding mission. If the Fund was worried about a bubble for every sector they looked into, I would be the first one to abandon ship...
It is also important to know that there are important regional differences, both at the inter and intra-city level. Just play with the permit and sales data, which have some regional detail, or Garanti price statistics, which go into a lot of detail for the major cities, if you have the time.
Finally, the email exchange with my three buddies also reminded me our lack of housing data. I am sure the banks are doing their own analysis for internal use, but I would have loved to know new homes sales, unsold housing stock, the average length of time a house sits unsold, etc...