Wednesday, April 09, 2008

Paul Krugman and Brad Delong have an argument about Rents

Krugman argues that house prices are fundamentally still too high, because CPI real rents have been flat. Delong says he is not sure what to make of CPI Rents. As I noted in a previous post, Ralph Turley, who knows more about these things than most all of us, argues that CPI rents are in fact measured incorrectly.

In a paper I did with Cutts and Chang (for which I must get around to doing the final revision), we performed hedonic regressions showing that quality adjusted rents during the 1990s in many cities did rise more rapidly than the CPI suggested. And when I look at rent to price ratios for comparable properties (i.e., condos and apartments) in, say, Los Angeles, the financial argument for owning seems pretty reasonable right now. Again, because we don't know how people are forming their expectations at the moment, it is difficult to say when a bottom is coming. But I do think equilibrium house prices are higher than Krugman thinks.

5 comments:

TStockmann said...

Looks like you're going to have to start using verification.

One question: does the paper's methodology assume a standard hedonic adjustment or is there a formula with variables dependent on locality factors that would have predictive power in rent vs. ownership?

knzn said...

"The financial argument for owning" depends partly on interest rates. I'm surprised that both Delong and Krugman seem to ignore this critical element. A large part of the reason for the housing boom was that real interest rates went very low, while overconfidence in securitization reduced the market risk premium associated with housing. The bust seems to be driven largely by positive feedback in the opposite direction: first prices overshot on the upside and had to correct; declining prices made mortgage lending appear more risky, and the securitization market collapsed, pushing up dramatically the risk premium associated with housing, and causing prices to decline further. There were also "credit channel" effects, in that it became very easy to get a mortgage during the boom, and now it has become much harder. My guess is that where the bust stops will depend mostly on when the credit market stabilizes and establishes the fundamental value of any given stream of rents.

Ken Deuel said...

I fail to see how converting from nominal to real makes any difference. You are still comparing apples to apples when you compare rents with housing prices because they both would be converted together, no? And without substitution effects (i.e. unlike how a 2008 50 in plasma is a 1958 13 in crt b&w)?

And I'm not that familiar with the LA market, but a cursory glance at neighborhoods I'm familiar with has a 1/1 condo for rent for about $1500 and the equivalent for sale asking around 400K. So its still paying a premium of at least 800/mo (over 50%) in which to break even you still need at least 10 years and a 3% per year price appreciation.

knzn said...

It's not a question of converting from nominal to real; it's a question of what has happened to nominal rents in aggregate. That's what the price index is supposed to tell us. We know (as long as you trust the Case-Shiller Index) what has happened to nominal house prices.

Anonymous said...

Rents can also get too high. Saying home prices should be unaffordable because rents are also unaffordable misses the point.

Unaffordable prices are inherently unstable. They can only be sustained by borrowing ever greater amounts of resources from overseas. Eventually, the party stops when high loans can no longer be repaid. At that point, foreign savers stop lending to us, and find somewhere else to store their deferred consumption.