June 18, 2005
Babble about a housing bubble
There's been much discussion recently of whether the U.S. is experiencing a speculative bubble in house prices. Like previous historical bubble sightings, this one only seems to pop up in situations where the fundamentals on their own might justify significant price increases.
And right now housing bubbles seem to be popping up all over the place. Calculated Risk, writing at Angry Bear, finds one in Miami. Tyler Cowen thinks he maybe sees one in D.C. David Altig expects to hear about one whenever Robert Shiller is on the radio. And Brad Setser is now upping the ante, looking for bubbles in France and all around the globe.
The most recent discussion of housing bubbles was spurred in part by the release of a report from the Office of Federal Housing Enterprise Oversight showing that U.S. house prices have gone up 50% over the last five years, clearly an unsustainable pace. Of course, conventional 30-year mortgage rates also dropped nearly 300 basis points over this period, and as shown most recently by Richard Rosen of the Federal Reserve Bank of Chicago (hat tip: Calculated Risk), the two are certainly related.
But the national bubble seems to have neglected places like Jackson, TN, where house prices have failed to keep up with inflation. According to the standard theory of a bubble, speculators bid up the price only because they expect the next person to pay an even higher price. If those expectations have no foundation in reality, there's no reason why a bubble couldn't appear any old where. Yet, as the map of the OFHEO data below shows, house prices have gone up four times as fast over the last five years in places like California and Nevada compared to Ohio and Tennessee.
To think about how we might try to explain these differences across states in terms of the underlying fundamentals, consider a community in which total real income is growing at 2% per year while the stock of housing is only increasing 1% per year, the latter due to a combination of local government restrictions and the physical reality that you can't grow more land in the most desired locations. If renters are willing to spend a constant fraction of their income on housing, rising incomes and population will result in bidding up the rental rate on any given piece of property by 1% per year relative to inflation.
Let's look at a particular property in such a community whose rental income nets the owner $4,000 per year after expenses. If the property price is determined by the condition that the owner must earn a 3% real return, the property would have a fair market value of $200,000. One arrives at this fundamentals-based price through the following calculation. If the current price were $200,000, next year the owner could sell it for $202,000 (in real dollars). The rental income of $4,000 plus the capital gain of $2,000 provide the requisite 3% return on capital of $200,000. Next year's $202,000 price is likewise warranted by next year's $4,040 rental income and prospect of 1% capital gain on $202,000, and so on. There's no bubble because both the price-to-rent ratio and the rent-to-income ratio remain constant over time.
Now suppose instead that population and income gains are expected to produce economic growth for the community of 3% each year rather than 2%. In this case, if the stock of available housing is still only growing at 1%, the rental rate would now grow at 2% each year. The analogous calculation to the one above implies that the same piece of property earning the same $4,000 rental would now be worth $400,000 rather than $200,000. The owner is now rationally expecting to earn $4,080 in rent next year and to be able to sell the property for $408,000. The $8,000 capital gain plus the $4,000 current rental income again justify the property value in terms of giving the owner a 3% real return.
This simple calculation helps us to understand that if a community experiences a change in its growth rate, property values can increase a great deal over a short time. For the above example, going from 2% to 3% growth would cause the property values to double overnight. It's noteworthy that over the last 5 years, the three states with the highest population growth rates as reported by the Census Bureau-- Nevada, Arizona, and Florida-- have also been among the locations that saw the biggest increase in home prices. Forces such as these, rather than a random distribution of irrational exuberance, seem a more natural explanation for why some communities got bubbled and others didn't.
Another argument that's sometimes raised (for example, by Kash at Angry Bear) is that some of the buyers of homes at current prices aren't going to be able to repay their loans. One could ask the obvious question here, why would banks be making unsound loans, to which there is a subtle possible answer. It has long been understood that federal deposit insurance makes the government part underwriter of some loans, because if there is a real estate crash and the bank goes broke, the government rather than the bank's owners end up picking up the tab. The result is that a lending institution can end up making loans that are riskier than it would undertake if all the capital at risk were its own.
If this moral hazard is understood to be the ultimate factor driving the housing bubble, however, the alleged bubble becomes something that the regulatory agencies can and should be addressing quite directly by making sure that lending institutions hold sufficient equity, so that it's their capital, not the taxpayers', that is at risk, and also making sure that there is accurate public disclosure of all risks. Economist's View has some further policy suggestions along these lines.
If you believe that there is a housing bubble-- and more to the point, even if you don't-- such measures make a lot more sense than trying to drive interest rates high enough to prick the bubble, as suggested recently by John Makin of the American Enterprise Institute (hat tip: Qando). We tried that little gambit in 1929, and things didn't turn out so hot that time around.
Posted by econbrowser at June 18, 2005 11:23 AMdigg this | reddit
Listed below are links to weblogs that reference Babble about a housing bubble:
Tracked on June 19, 2005 10:52 AM
» Catching my eye: morning A through Z from The Glittering Eye
Here's what's caught my eye this morning: Ronni Bennett of A Sense of Place is learning that not only can't you go home again but sometimes even when you've never left it runs away from you. Sage advice on housing... [Read More]
Tracked on June 19, 2005 12:02 PM
Tracked on June 20, 2005 01:25 PM
» Housing Bubbles? A Fact and a Thought from EconLog
James Hamilton's got my attention on housing bubbles. One of his many interesting observations is that population growth may explain... [Read More]
Tracked on June 25, 2005 10:31 AM
Professor, thanks for the mention! Very interesting post. The regulatory agencies have started addressing recent poor lending practices with this new guidance (and apparently more to come later this year):
Agencies Issue Credit Risk Management Guidance for Home Equity Lending
Posted by: CalculatedRisk at June 19, 2005 10:31 AM
I think if you look at the availability of land you will find a much better explanation of why the biggest increases in home prices are occuring in the places they are and not nationwide.
In places like San Diego, Los Angeles, Miami, New York, Boston that there is no land to be built on within a reasonable commute of the city center. Consequently home prics rise much more rapidly in those areas then in area where there is farmland with in an hour or so commute that can be converted to home sites. It is also why the big nationwide building chains have driven the small homebuilder out of the market. They have the resources to "bank" land and wait for development to sprawl out to the land they own and the small homebuilder does not have the capital to do this.
Posted by: spencer at June 19, 2005 12:49 PM
The question of why lend to people when price seems way above some sort of normal value needs to be split into two distinct questions: 1) why originate a mortgage, and 2) why hold such a mortgage. The mortgage originators aren't holding those mortgages. We've managed to institutionalize the greater fool theory with our mortgage-backed securities industry, which *always* allows an originator to pass even mediocre-quality mortgages of to... to... who knows who. Then, the risk gets amortized to death so no individual mortgage risk is ever seen by a single investor.
Who's holding all this securitized mortgages, and have *they* been adequately informed of the aggregate risk that's building up?
Or, maybe they *do* have a clear measure of the risk and it's *not* as big and out of control as "the bears" (angry or not) would have your believe.
Or, maybe they do see real risk, but have some keen insight as to how to *hedge* that risk.
If so, has anybody assessed the potential for default on the part of the *counterparties* for those brilliant hedges (anybody remember LTCM)?
Are any of those hedges "dynamic" (remember 1987)?
Does the ever-more-brilliant development of hedges have the moral hazard of eliminating any significant feedback to to final holders of the securitized mortgages? Are those holders (whoever they are) effectively being told "Don't worry, w're *fully* hedged?
Do any of these hedges have the effect of putting downwards pressure on long-term interest rates?
Anybody have the "good" dirt on what's going on with risk assessment in the MBS sector?
-- Jack Krupansky
Posted by: Jack Krupansky at June 19, 2005 03:52 PM
You ask "why would banks be making unsound loans"
Umm - have you ever heard of Fannie and Freddie? MBS's?
These banks don't carry the loans on their balance sheets.
So, yes, I think moral hazard is helping to drive the bubble in the bubble-zones.
It's great to have an economist of your stature in the blogosphere. However, it would be nice to have more informed discussion of the institutions - it can be a long way from Chicago to the real world.
Keep up the blogging - I look forward to reading your posts.
Posted by: Vancouver Housing Blogger at June 19, 2005 10:55 PM
Nicely argued. The idea of raising short term rates significantly to get at the supposed housing bubble is a decidedly poor idea :)
Posted by: anne at June 20, 2005 07:18 AM
http://www.markarkleiman.com/archives/macroeconomic_policy_/2005/05/speculative_selling_in_the_la_housing_market.php describes econ prof Mark A. R. Kleinman's attempt to beat the bubble by selling his house now, near the peak (he hopes!). I'm skeptical that he will benefit from this but we'll see.
I live in an area with a supercharged housing market. 10%, 15%, 20% gains per year for years and years. I bought 10 years ago and my house value has tripled. Median housing prices for this city are over 1 million dollars. Affordability indexes are at all time lows.
And it's been this way for years. Five years ago I met an attorney at a party who made the same bet then that Kleinman is making today. He was sure the overheated housing market couldn't continue. He was selling his house and was going to rent for a few years, so he could step back into the housing market and make a killing.
So what happened? The stock market crashed and the housing market soared. House prices have nearly doubled since then. I don't know what happened to the guy, I never saw him again, but his play did not turn out to be a winning bet.
It's weird about bubbles. It's so hard to know when you are in one, but so obvious in retrospect. I almost think that there is no fact of the matter about whether you are in a bubble or not, that it depends on indeterminate future events which are fundamentally unpredictable.
Posted by: Hal at June 20, 2005 10:30 AM
Although your reasoning is interesting, I believe that you are comparing apples to oranges in your analysis; may be my lack of formal training as an economist is deterring my ability to fully comprehend your argument. Here is my comment:
In your example, you say: "consider a community in which total real income is growing at 2% per year while the stock of housing is only increasing 1% per year" (I have underlined the words to draw your attention to them). My understanding is that a community's total real income can grow in three ways: (1) increase in the working population's real income, keeping the working population constant; (2) increase in the working population, keeping the working population's real income constant; or (3) a combination of (1) and (2). From what I have read from various sources, I gather that the real income of the working population in the US has remained constant (or the growth has been very minimal), including in areas like Arizona, Nevada, and Florida. This makes me conclude that the total real income growth in the community is due to increase in population, and not due to an increase in real income (which you casually mention in your blog: "It's noteworthy that over the last 5 years, the three states with the highest population growth rates as reported by the Census Bureau-- Nevada, Arizona, and Florida -- have also been among the locations that saw the biggest increase in home prices.")
If the above mentioned premise is true -- increase in community's total real income is due to an increase in the population and not due to an increase in the population's real income, then an increase in the [real] value of the property from $200,000 to $400,000, in your example, will not attract any buyers (buyers cant afford to pay for the property's new value since the real income of the individual buyer has not increased). This will result in the short-term property value growing at the previously existing rate (or increase a little but not too much), unlike what you have purported in your blog; your analysis does not taking into account the demand factors. The only reason for the property value to increase multi-folds, given a minimal individual real income growth, is a low borrowing rate (as argued by Richard Rosen of the Federal Reserve Bank of Chicago).
Based on my reasoning, I conclude that your argument, though seemingly valid on the surface, is really not valid. Please note that I am not arguing for or against the existence of a housing bubble, but am just commenting on logic of your argument. Please correct me if I have erred in my reasoning.
Posted by: Madan Manoharan at June 20, 2005 10:59 AM
Thanks for the question, Madan.
Total income growth corresponds to total GDP growth, which has certainly gone up faster than the population.
Even if the economic growth were only due to population, however, if the housing supply is fixed and each individual's income is fixed, the rental rate would still be bid up. The way you'd see this is that, as the city becomes more crowded, people can only afford to live in smaller apartments and existing units get subdivided. But another way of saying the same thing is that the rental rate on a given apartment of fixed size would go up.
In the example I discuss, the rental rate is only going up 2% per year rather than 1% per year as a result of the faster population and economic growth, so it is quite affordable for everybody to continue renting, they just have to pay a little more each year if they wanted to get the same thing.
Perhaps where you're getting confused is that the price of purchasing a house as opposed to renting has gone up by much more than 2%, in my example, the price in fact doubles if a community changes from a 2% to a 3% permanent growth rate. But that does not mean that people need twice the income to be able to afford to buy the house as opposed to rent under the new conditions. The reason is that the true user cost of owning a home takes into account not just the cash you pay to buy it but also the capital gain you're earning each year from the fact that house prices are going up. In a community with faster economic growth, people owning a home are getting a greater capital gain each year, and that is in fact what justified the price they paid for the home in the first place.
Posted by: JDH at June 20, 2005 11:29 AM
BTW another way to hedge if you think you are in a housing bubble is via hedgestreet.com. This lets you bet on the future direction of housing prices (among other things) in a sort of futures market. At this point they are covering the following cities: Chicago, Los Angeles, Miami, New York, San Francisco, San Diego. The contracts only go out six months so you would have to roll them over if you thought a bubble was likely to burst a bit farther out.
I'm not entirely sure how to interpret their price presentation, but it appears that the market is predicting a levelling off of L.A. housing prices.
Posted by: Hal at June 20, 2005 06:27 PM
Listen up folks, whether we are in a "bubble" or not doesnt really matter. What matters is when prices will go down. Prices will go down based when one the following occurs:
There Are More Sellers Than Buyers
Then you ask the question, what will cause that? The current scenarios are:
1. Interest Rates rise (unlikely given the coming recession)
2. Demographics (not to hit fully force for another 3-7 years)
So yes property prices are sky high. But things usually take longer to unfold than people expect. Property prices may begin to top out, but to expect a plunge in prices at the moment is to suffer from analysis bias. Where is the catalyst? I think we got a few more years left. If things start to go south you can bet the Fed will lower rates to dampen the fall.
But when prices start falling, it will be real ugly. I predict you will have many banks having capital shortfalls, fannie mae will need a bailout, and pension funds and all those foreigners holding u.s. agency debt will lose a ton as debt values implode and credit spreads blow out.
BEST HEDGE AGAINST PROPERTY PRICES IS TO SHORT THE USD.
Dont buy a house now unless you are independently wealthy and can afford to lose 50%.
Posted by: John at June 21, 2005 01:24 AM
"It's weird about bubbles. It's so hard to know when you are in one, but so obvious in retrospect. I almost think that there is no fact of the matter about whether you are in a bubble or not, that it depends on indeterminate future events which are fundamentally unpredictable."
Agreed, and it is one thing to sell off technology shares or buy against them and quite another thing to sell your home.
Posted by: anne at June 21, 2005 04:10 PM
Fascinating discussion, Prof. Hamilton. I'm not sure if I follow all your logic, but it sounds to me as if your calculations are a variant on the Gordon growth equation for calculating stock prices.
Let's see...if you assume a rental "dividend" of $4,000 a year, your required rate of return is .03 and the expected growth rate of your rental income is .01, then the Gordon model says the house is worth $200,000 in Scenario 1. If everything stays the same, but the growth rate of your rental income rises to .02 in Scenario 2, then the Gordon model says the house is worth $400,000.
Dandy. But for the sake of completeness, how about looking at Scenario 3 where the real growth rate of rents falls to zero? In that case, am I right in thinking that the house is worth only $133,000? And how would you rank the probabilities of these three different scenarios across the U.S. as a whole?
What scares me about today's market is the sensitivity of home prices to relatively small changes in the difference between required rates of return and expected growth rates. Am I missing something?
Posted by: Ian at June 22, 2005 09:31 AM
You're exactly right, Ian. I plan to talk about this issue in one of my next posts.
Posted by: JDH at June 22, 2005 09:37 AM
Interesting article about the housing bubble. I'm the editor of a non-commercial blog focused on the stock market. Yesterday I posted an article about how to profit from a slowdown in the housing sector.
I hope you can take a look at it. If you want to link to it, that would be great (It's a new blog). I'm not selling anything (except maybe some free publicity).
Posted by: Jay Buster at November 9, 2005 03:34 PM
This housing market run up is much like the Japanese housing bubble cry many economists.
In reality the housing market in Japan is very different when compared to the United States and indeed to rest of the world.
The Japanese population is slowly dying out; there has been practically no growth in population in Japan. Further US housing prices are the lowest when compared to home size and purchasing power of the US consumer.
They forget to tell you that the Japanese population increased by only 4 million from 1990-2005 where as US population increased by 45 million in the same period.
Further the job growth in Japan is anemic whereas the US continues to add new jobs at a good pace.
Many of these doom and gloom gurus worry that homeowners have refinanced their homes and spent the money. A portion of the money definitely has been spent on vacations and consumer non-durables, but a larger portion would have gone for home improvements like energy efficient heating and air-conditioning, energy efficient windows and siding, roof systems etc. Further, some of this money could have been used to reduce high interest credit cards and for funding college education for kids- future consumers with higher incomes.
Is it possible these housing bubble forecasters and gurus are really stockbrokers and analysts hoping you will sell your home and invest in stocks and bonds so that they can earn a commission. Wonder what they plan to do with their bonus earned in January 2006; will they buy stocks and bonds or homes? We shall know soon.
Read the full article at www.NoFeeMLS.com
Posted by: Anonymous at January 13, 2006 04:28 PM