Saturday, December 15, 2007
I Have the Flu, so it's time for another thought-experiment
by Ken Houghton
Don't get me wrong. There are still areas where rents are rising significantly, though many of those are traditionally and well-identified "bubble" areas. (It seems not untoward to suspect that the renters in Miami who are paying that 7.3% will find better deals when their lease is up, should they choose to look.)
But, looking at the broader picture, the numbers that have been thrown around as the decline range from 15% (GS) to 30% (Krugman).*
Speaking as a long-time owner (10 years in less than two weeks) in a "bubble" area that has some intrinsics going for it, I look at the 15% figure and say "Yours!" to the extent that, if you guaranteed this house wouldn't drop in price more than 15%, I would be happy. (30% is more worrisome, but we're still talking something resembling a positive NPV, I think, maybe.) So I don't think I'm the marginal case in this decline.
But I'm not sanguine about either number. As Barry Ritholtz noted here, that OER ratio is moving slowly back toward any convergence.
Some of that is due to sticky prices (or, as those who speak Business-English refer to it, "the inability or unwillingness to sell due to negative equity"). But it still looks as if the graph in Ritholtz's piece needs to drop another 21-23% just to get back to the previous high area of ca. 2.5. And it's fairly clear that the graph doesn't move in a 1:1 proportion with pricing changes.
So the question is: can anyone build an equilibrium model in which (1) house prices decline by at least 15% but no more than 30% [UPDATE: This is too restrictive; let's just make it less than 30%], (2) OER returns to reasonably normal levels (you can set the level higher than 2.5, but you have to explain why it should be so, and (3) median rental rates do not exceed 25% (or, at least, all in housing costs, including utilities, does not exceed 33%) of median Gross Domestic Income?**
For a closing paragraph, I return to Dean Baker, who put an estimate on the breadth of the convergence required:
*CR at CR discusses the posisble results of those projections here.
**Again, if you want to vary from those assumptions, feel free to do so, but please explain why you choose to relax the constraint, and what people will sacrifice (or have to pay less for on an absolute basis) in a basic income model for more housing.
Let's take some disparate pieces of data and see if someone can build a model out of it:
- The price-rent ratio should be approximately 1.00 (the tax gains from owning a house are mitigated by the cost of maintenance—trust me on this one). Yet the ratio was running around 1.6 recently, meaning housing prices were significantly higher, to the extent that Robert Poole and Randal Verbrugge attempted to explain the divergence, which appeared to level off but not decline coming into 2007.
- There are two ways for convergence to occur. Either the housing price declines, or the rental price rises (link to Dec 2006 data). (The third possibility is that both move at once; this is, of course, the most likely in a rising-asset environment.)
- We know that housing prices are declining, and Dean Baker noted back in July that the OER differential was declining.
- There are also signs that rental prices are being pressured in some areas by increased supply, whether people who moved to Denver and held onto their California-based not-a-quick-asset-except-in-David-Malpass's-delusions albatross or other, even less encouraging, reasons.
- Home ownership is at roughly the same level as it was when "subprime" lending took off [PDF link] (h/t Dr. Krugman). So all things being equal, demand for demographic reasons shouldn't be changing much in the near term.
Don't get me wrong. There are still areas where rents are rising significantly, though many of those are traditionally and well-identified "bubble" areas. (It seems not untoward to suspect that the renters in Miami who are paying that 7.3% will find better deals when their lease is up, should they choose to look.)
But, looking at the broader picture, the numbers that have been thrown around as the decline range from 15% (GS) to 30% (Krugman).*
Speaking as a long-time owner (10 years in less than two weeks) in a "bubble" area that has some intrinsics going for it, I look at the 15% figure and say "Yours!" to the extent that, if you guaranteed this house wouldn't drop in price more than 15%, I would be happy. (30% is more worrisome, but we're still talking something resembling a positive NPV, I think, maybe.) So I don't think I'm the marginal case in this decline.
But I'm not sanguine about either number. As Barry Ritholtz noted here, that OER ratio is moving slowly back toward any convergence.
Some of that is due to sticky prices (or, as those who speak Business-English refer to it, "the inability or unwillingness to sell due to negative equity"). But it still looks as if the graph in Ritholtz's piece needs to drop another 21-23% just to get back to the previous high area of ca. 2.5. And it's fairly clear that the graph doesn't move in a 1:1 proportion with pricing changes.
So the question is: can anyone build an equilibrium model in which (1) house prices decline by at least 15% but no more than 30% [UPDATE: This is too restrictive; let's just make it less than 30%], (2) OER returns to reasonably normal levels (you can set the level higher than 2.5, but you have to explain why it should be so, and (3) median rental rates do not exceed 25% (or, at least, all in housing costs, including utilities, does not exceed 33%) of median Gross Domestic Income?**
For a closing paragraph, I return to Dean Baker, who put an estimate on the breadth of the convergence required:
In fact, while real house prices rose by more than 70 percent, real rents rose by less than 10 percent. And now rents are falling in real terms, making the divergence even greater (at least until house prices start plunging). This fact warrants some attention from the media -- even busy reporters should have time for $8 trillion in housing bubble wealth.
*CR at CR discusses the posisble results of those projections here.
**Again, if you want to vary from those assumptions, feel free to do so, but please explain why you choose to relax the constraint, and what people will sacrifice (or have to pay less for on an absolute basis) in a basic income model for more housing.
Labels: convergence theory, equilibrium, Housing Bubble, mortgage
Comments:
<< Home
It seems that part of the rise in the ration from 1 to 1.6 would be accounted for by the fall in interest rates in the 2000s. I don't see that this is taken into account in the ratios.
Thus, if you are at a ratio of 1 with 30 yr rates at 7%, then you would expect house prices to rise to a new stable level if rates fell to say, 5.8%...thus resulting in a ratio of >1.
It seems to me that if you are relating in the cost of renting versus buying an asset, you have to include the cost of money.
Thus, if you are at a ratio of 1 with 30 yr rates at 7%, then you would expect house prices to rise to a new stable level if rates fell to say, 5.8%...thus resulting in a ratio of >1.
It seems to me that if you are relating in the cost of renting versus buying an asset, you have to include the cost of money.
The shift in monthly payments isn't linear. If the interest rate drops from 7% to 5.375% (down 30%), the 30-year fixed monthly payment only drops 18.8%.
It's a start, but you're not going to come close to Ritzholtz's graphic (and note that the climb continues as the Fed raises rates into 2005).
It's a start, but you're not going to come close to Ritzholtz's graphic (and note that the climb continues as the Fed raises rates into 2005).
I think Robb is getting at a different problem -- in his example, you drop the discounting rate ~17% and the PV of a constant stream of owner's equivalent rent increases by ~19%. This is probably more than the price should rise to the extent that the decrease in the interest rate is the product of countercyclical cheap money vs. a whipping of permanent inflation expectations.
Post a Comment
<< Home
