Tuesday, July 28, 2009

Rant #3,847b

Sometimes you read shit that is so wildly and utterly stupid that you simply can't help being baffled that those charged with its production keep breathing, much less retain their jobs at MIT.  Today's rant is brought to you by William C Wheaton, who has a brief note over at VoxEU.  The subtitle really says everything you need to know about it -- Is the US due for a protracted decline in housing prices? This column argues that fundamentals on the supply side of the housing market imply that US housing prices are about to bottom out. They should resume rising soon. 

The article is short enough to read quickly, but just to summarize, he points out that

  1. A house is a physical asset that people need to live in, and not simply a financial asset
  2. The population of the US is still growing and new households are being formed (albeit at a diminished rate currently) so ...
  3. We will thus need more houses, and consequently ...
  4. The price of houses will have to rise so that we encourage people to build more of them.  
At first, this argument seems relatively coherent.  I mean if we need more houses, won't that eventually drive the price of them up?  This sounds convincing only until you realize that it would have been exactly as true 100 years ago, and yet, somehow, real home prices have not really trended up at all over this time period, as you can see in the graphic below, which ends near the peak of the bubble in 2005 (the national index has now fallen to something like 140).

What a mystery, you say.  How could this happen, you wonder?  Um ... well ... let me see if I can put this succinctly ... when we need more houses, we build them.  There's no real bottleneck in housing outside of a few specific places.  If more people want houses then we build more houses, and the fact that you can build them just about freaking anywhere and still be within spitting distance of a Wal-Mart means that the price never really has a chance to rise much above replacement cost.  This is what is formally know as a "market", which apparently our dotty MIT prof is not familiar with.  Replacement cost doesn't set a floor on pricing, it sets a ceiling -- why on earth would you ever buy an old house for more than what you could build it for brand new?

The article also contends that most homes are at or below replacement cost today.  I haven't seen any data about this, but I am wildly dubious.  Look again at the chart.  The inflation adjusted building index in 2005 was at 161.  Given that this was the peak of a speculative mania when every conceivable resource for building a home was in heavy demand, I would guess that the index is lower now.  I'm also not sure how big a component land is in that building cost index, or even if it is counted there at all, but land is a particularly volatile component of replacement cost in the short-term.  Certainly, any of the other commodities that go into a building are more than 10% lower today than they were in 2005, excluding, perhaps, labor.

If you blur your eyes you might conclude that the house price index (140) and the building cost index (say 145) have more or less converged, making the case for the fact that we must be somewhere near the bottom in housing.  This isn't actually far off, and the best guess (based on things like affordability, price to income, price to rent etc ...) is that we are maybe 10% from the bottom in terms of inflation adjusted prices.  So, in the end, this totally flawed article is like the proverbial blind squirrel who finds (half) an acorn.  The other half, that is the notion that prices will start rising again because of a rising population, is completely ludicrous.  Housing in general does not appreciate in real terms.  Period.  There will always be exceptional markets, but overall, there's a lot of space in the world, and home prices will rise at the rate of inflation.

No comments: