Arnold Kling  

Time to Panic?

PRINT
Freeman Dyson on Global Warmin... Gov is not Great: How Democrat...

Lawrence Kudlow writes,


There are a number of deflationary factors behind my campaign to get the Fed to permanently inject new cash into the banking system and deal with the dysfunctional commercial paper market — as well as the general credit freeze-up.

There’s housing price deflation: The Case/Shiller home-price index is off 3.5 percent over the past year.

There’s commodity deflation: Gold prices are off nearly 15 percent from their 2006 highs. Stock prices for materials are off nearly 13 percent since July 19, while metal and mining shares are off 16 percent.

There’s the deflation of loan values, both CDOs and CLOs.

And there’s the deflation of the Treasury bill rate from 5 percent to 4 percent.

The Fed needs to stop this deflation by pouring in new cash.


I'm sorry, but I measure deflation the old-fashioned way, using broad price indexes for the economy, such as the CPI or the GDP deflator. When those start falling, let me know.

Meanwhile, Robert Reich writes,


For the financial market to work well -- to ensure fair dealing and to prevent speculative excess -- government must oversee it. This mess occurred because nobody was watching. The Fed and other central banks now have to clean it up. But regulators in America, Europe and Asia have to make sure it stays clean. Hedge funds have to be more transparent. Credit-rating agencies must not have any relationship with underwriters. Banks and mortgage lenders should be better supervised. Finance is too important to be left to the speculators.

Again, I'm sorry, but I don't think that the regulators were any more on top of this than the speculators. Was Robert Reich out there two years ago complaining that the finance industry was making housing too affordable by coming up with too many innovative approaches to supplying credit to first-time homebuyers? I must have missed that column. Yes, I'm sure that some government officials were on top of the problem, but so were some of the speculators. The reality is that unless you want to stifle innovation completely, you are going to have to expect occasional excesses and mistakes.

By the way, thanks to Mark Thoma for the pointer.

Finally, Zimran Ahmed writes,


I think one inevitable requirement of unwinding the housing bubble market is that housing prices have to come down to fall in line with historic trends. In some areas this means very dramatic decreases -- maybe 40%+ in real terms? I'm not sure what a "deflated housing bubble" would look like if it did not bring prices back to historic norms.

I'm sorry, but unless by "some areas" you mean areas the size of a 9-digit zip code, we're not going to see 40 percent declines in house prices.

Think of the equilibrium house price as the equilibrium rent times the equilibrium price/rent ratio. I do not see rents falling in real terms. Housing starts are below the level needed to keep supply in line with new household formation, especially in an economy with relatively low unemployment. So, if anything, rents are going to be drifting up.

The price/rent ratio is a bit harder to judge. I use a 4 percent real interest rate to arrive at an equilibrium ratio of 25 times annual rent. This in turn means that the price ought to be 300 times the monthly rent for equivalent rental properties. I call this the "rule of 300." I think that price/rent ratios are below that in most of the U.S. Real estate is, if anything, cheap today. In the short run, it may get a bit cheaper in some places, but I would not want to be sitting on a big short position in house price index futures.

Even if the price/rent ratio needs to fall, one limiting factor is that as house prices fall, builders supply fewer new homes, and this tends to push up rents. If homes do get cheaper over the next year, the subsequent bounceback in prices will be even stronger.


Comments and Sharing


CATEGORIES:



COMMENTS (18 to date)
Mike D writes:

AUTHOR: Mike D
EMAIL: cr2005econ@yahoo.com
IP: 207.62.205.207
URL:
DATE: 08/29/2007 05:21:50 PM

Independent George writes:

Something doesn't add up. Either my condo was the steal of the century (purchased two months ago - post bubble, but pre-meltdown), or 300 x monthly rent is way too high a ratio.

Lab Rat writes:

His ratio ignores maintenance, vacancy, management overhead, the need to pay down principal, and the profit motive. In other words, classic economic theory. This ratio is only workable in a wildly overheated bubble market if you're flipping rentals like the pancake cook at IHOP. I have to admit, it takes a pretty tall ivory tower to believe that your paper napkin calculations are more correct than 50+ years of market data!

He also apparently hasn't noticed that the housing market has imploded like a steel drum at the bottom of the Mariana Trench, now that the Germans and Chinese are no longer interested in buying NINJA subprime toxic waste repackaged into dangerously opaque securities.

Now is a great time to be short on housing futures.

dcpi writes:

When I first moved to NYC in 1990, the "old landlord pros" used a price to rent ratio of 18 .. though some were willing to push to 20-21 times annual rent (so 216 to 252 times monthly rent). That translates into a 5.6 expected real return before taxes. Those that survived long term as pro landlords rarely went over that ratio. I take that a sort of Darwinian answer to the question.

Looking at the question analytically rather than experiencally, using 4 percent real return seems too dangerous. Why would you want to take on the onus of being a landlord for so little excess return in a market where it is easy to find dividend paying large caps price at 12-16 times earnings that literally allow me to sleep at night (rather than take calls from renters)?

I would expect a 5.5 percent real return at least from rental property. Or roughly a 218 multiple, 200 to make the math easy and factor in some extra protection.

Since disagreements make markets ... would you be willing to buy some property from me at 300 times annual rents? I will sell. My bet ... how can we make one? ... would be that a 300 ratio "bakes in" little prospect for near-time appreciation above inflation and a large chance of at least some depreciation.

Arnold Kling writes:

I would not want to be a landlord using the rule of 300. As a pure investment, and with all the headaches of being a landlord, I'd want a larger cushion.

But I'm thinking about it from the standpoint of a consumer who has to live somewhere, making the rent vs. buy decision, under circumstances where you know you won't be moving for, say, 7 years or more. If the price/rent ratio is below 300 and you're paying rent, then I think you're throwing away money.

dWj writes:

It was just released that there are 9.6 months' supply of houses for sale. In the near term, rent should be soft. When some of the supply is worked off and rents pick back up, construction will increase again to the rate of household formation.

I'm not sure real estate anywhere outside of Tokyo has ever sold for 25 years' rent. I agree with the fellow who suggests you've forgotten maintenance and similar current expenses.

ed writes:

Doesn't the rent vs. buy decision depend heavily on whether you expect homes to appreciate? Shouldn't you shade your "rule" to account for that?

Neel Krishnaswami writes:

I find equilibrium arguments about housing prices hard to swallow, because it is a market that is wildly inefficient and regularly very far from equilibrium.

Take a look at the OFHEO repeat-sales price index and you see very smooth, regular changes in the price over time -- it's completely unlike a random walk. If you give Fiserv lots of money, they'll sell you ZIP-code level repeat-sales indices, which have exactly the same smooth changes. This rules out smoothing caused by aggregation, so we know for sure that the housing market isn't in equilibrium.

In fact, I have wondered if some bright grad student could reinvent Keynesian macroeconomics without the weird menu cost handwaves of the new Keynesians, just by exploiting the non-equilibrium features of the real estate market.

dcpi writes:

Ed:

Over the past century there has been very little real appreciation in real estate (that is appreciation after accounting for inflation). That means that Kling's formula does not need to account for appreciation to have value as a guide. Those pesky expenses -- new roofs, boilers, remodels, repainting, etc. -- serve to further make the bet on a rising price an unlikely winner. The big wins that people see on their housing bets have resulted from the use of leverage to magnify small gains. With five percent down you are levered 20:1. Can't calculate the leverage on the piggybacks. Of course leverage works the other way on the way down. The 1 percent decline this year translates into a 20 percent loss for the 5 percent down club.

Independent George writes:

Are you calculating based on list price, or projected total expense including interest on a 30-year mortgage?

I'm going to have to dig out my old spreadsheet, but after factoring in mortgage interest over 30 years minus tax benefits, I came up with an equilibrium point of around 15 x annual rent (which is roughly where I ended up after closing costs & credits). As much as I'd like to think I'm a genius, but I'm pretty sure I wound up paying the natural market clearing price.

Michael Sullivan writes:

But I'm thinking about it from the standpoint of a consumer who has to live somewhere, making the rent vs. buy decision, under circumstances where you know you won't be moving for, say, 7 years or more. If the price/rent ratio is below 300 and you're paying rent, then I think you're throwing away money.

I disagree. At least some of the landlord's costs are borne by homeowners as well: taxes and maintenance. The only cost that homeowners do not bear is vacancy.

Also, the risk bears harder on single homeowner than on a typical professional landlord. A landlord who is careful to avoid negative cashflows, probably will not have to sell unless the rental market is severely collapsed, and if they are a large landlord, can probably get away with selling a few properties, rather than their entire real estate position. A homeowner who loses their job is very likely to be forced to sell, and job losses correlate with bad real estate markets.

All in all, I think the homeowner needs less cushion than a landlord, but not a lot less. If a landlord should stop at 18x, a homeowner should probably stop around 20x. And this of course, assumes that housing will appreciate roughly at CPI. In depressed areas, that may not be a good bet.

Barkley Rosser writes:

In the second edition of his Irrational Exuberance, Robert Shiller estimated that we were seeing on average all-time high levels of price to rent and also price to income. Now, some people were arguing this was defensible due to low interest rates. Is that your argument? How do you reconcile your calculations with Shiller's data? Were those past price/rent ratios just way too low, or have things changed somehow?

ed writes:

"The 1 percent decline this year translates into a 20 percent loss for the 5 percent down club."

Uh, yeah exactly, that's why you should modify your rule to account for the probability that that happens. Don't you think at decline over the next three years is much more likely now than it was in, say, 1995? Like Neel said, housing prices don't look like a random walk.

Larry writes:

Even if he's right nationally, there are lots of regional markets that aren't anywhere close to equilibrium. Those where the price/rent ratio are way off their norms (for that region/area) are the places to short.

Arnold Kling writes:

Barkley,
I think there is nothing sacred about historical P/E ratios, for either stocks or houses. For stocks, the realized real returns over the past 100 years have been quite high, which suggests that traditional P/E ratios were too low. I think that the same is true for houses, although to a lesser extent.

George writes:

Journalist-turned-professor Robert Reich wrote: "Finance is too important to be left to the speculators."

And way too important to be left to the government.

dcpi writes:

Actually, you are totally wrong about real housing price appreciation over the past 60 years at least. Adjusted for inflation, the average home cost $30,600 in 1940 and $119,600 in 2000 and clearly even more today. http://www.census.gov/hhes/www/housing/census/historic/values.html

I think this is a function of several factors including the larger homes of today and the rising real income of the American consumer over the years. So, in the long run and on an inflation adjusted basis real estate values have always gone up (not to mention the amazing returns offered when you factor in inflation and the use of leverage to buy more real estate with borrowed money). So, dont bet against American real estate. It wont be down for long.

Jackson writes:

dcpi:
Actually, you are totally wrong about real housing price appreciation over the past 60 years at least. Adjusted for inflation, the average home cost $30,600 in 1940 and $119,600 in 2000 and clearly even more today. http://www.census.gov/hhes/www/housing/census/historic/values.html

I think this is a function of several factors including the larger homes of today and the rising real income of the American consumer over the years. So, in the long run and on an inflation adjusted basis real estate values have always gone up (not to mention the amazing returns offered when you factor in inflation and the use of leverage to buy more real estate with borrowed money). So, dont bet against American real estate. It wont be down for long.

Comments for this entry have been closed
Return to top