Sunday, February 24, 2008

One of my favorites lines from Bugs Bunny

is, "Daddy, you're back from Peru."

My kids will actually be able to say this to me on Wednesday night--I am going down to Lima to give a lecture at the Peruvian Central Bank. I am really looking forward to it.

Saturday, February 23, 2008

Alan Blinder and Mark Thoma want to bring back the HOLC

Mark's post is here.

I am myself a fan of the HOLC, and have said so in articles I wrote with Susan Wachter for Journal of Economic Perspectives and for the Jackson Hole conference last summer, as well as a comment I just wrote for Housing Policy Debate. Yet I am not sure it is alone the medicine for the current crisis.

When the Home Owners Loan Corporation was invented, it was in response to an economic tsunami that swamped lenders and homeowners. Moral hazard was not much of an issue, as loans were stringently underwritten (typical LTVs were 50 percent at origination). But loans had short terms, and therefore were vulnerable when people were forced to refinance in the teeth of the great depression. The HOLC allowed for massive loan modification and helped get incentives for borrowers and lenders aligned correctly.

Now, however, we are in the midst of a crisis that has arisen in part because of agency problems throughout the lending chain. To bail out lenders through some sort of HOLC setup could very well encourage excessive risk taking in the future, which is of course problematic.

I think if we are going to go the HOLC route, it needs to be accompanied by a regulatory structure that will prevent the sort of bad practices that led to the current crisis going forward. As I have noted before, such regulatory changes would require greater transparency, a requirement that everyone who touches a mortgage be subject to federal supervision, and a requirement that everyone who touches a mortgage have some capital at risk.

10 percent of Homeowners Upside Down???!!!

That was the headline in yesterday's NYT, based on calculations from Because the article doesn't describe how the calculations were done, it is not possible to agree or disagree with them, but I don't find the number on its face implausible.

Let us say for the sake of argument that the number is correct. It means that for 10 percent of owners, the put option contained in their mortgage (i.e., the option to put their house back to the lender) is in the money. If all these households default, the mortgage market will be cooked: Fannie and Freddie, who put 45 basis points of capital behind each mortgage they guarantee, would likely not be able to withstand so many defaults. Moreover, if ten percent of the market becomes REO, inventories will swell further, putting more downward pressure on house prices, which in turn will lead to more borrowers being underwater.

It is unlikely, however, that all such households will default. Past work by people such as Bob Van Order and Yongheng Deng has shown that people in the past usually needed to face a trigger event, such as divorce, job loss or sickness, before they default. The question is whether things are different this time: that because people are so aware of the declining housing market, they will understand more quickly that their house value is less than their mortgage balance, and will therefore default. After all, in principle, if one observes that her house is less valuable than her mortgage balance, she could default one day and buy the house back the next, and be better off.

But default doesn't actually work that way. When a borrower defaults, she trashes her credit score, and therefore reduces her options for either owning or renting going forward: people like having a roof over their heads. Beyond that, houses are idiosyncratic and can be more valuable to their dwellers than they are to the market. Houses become associated with memories, neighbors, gardens, strange paint colors, and other characteristics that make them valuable to their occupants in a way that cannot be reflected in the market. These things prevent people from defaulting (or selling when it makes more sense financially to rent). Finally, many households still feel an obligation to pay their bills. The fact that so many mortgage borrowers in Louisiana continued to make payments for homes Katrina destroyed is quite remarkable, and quite germane to the situation in which we find ourselves now.

How much these phenomena will prevent defaults in the next few years is an open and interesting question. I suspect we will know the answer pretty soon.

Wednesday, February 20, 2008

Testing Micky's Kaus' Election Hypothesis

I don't care much for Kaus' stuff, but he puts forward an interesting hypothesis: that every time Hillary Clinton does well in a primary, there is a backlash in the following primary, and that when she does badly, she generates sympathy and does better in the following primary.

I took her primary percentages each week since Iowa (using this handy dandy web site.)

The finding? There is a negative correlation of about .65 in her share of the primary vote from one date to the next.

Monday, February 18, 2008

Why an Interest Rate Freeze won't solve the Subprime Problem

I. Freezing interest rates on Adjustable Rate Mortgages would have little impact on the current mortgage crisis.

As I noted in a previous post, a recent paper by Federal Reserve Bank of Boston economists Kristopher Gerardi, Adam Hale Shapiro, and Paul S. Willen take a careful look at the characteristics of 2-28 ARM mortgages. Among its findings are that interest rate resets have little if anything to do with the large number of defaults we are observing. For the vast majority of subprime loans, defaults or refinances happen before resets take place. Moreover, the size of the resets is smaller than most of us think, because the initial teaser rates are pretty high. Also, because the Federal Reserve Board has cut interest rates so aggressively, reset rates will not be as high as people have feared.

The authors also note that while ARMs have higher default rates than FRMs, putting ARM borrowers into FRMs (which is what a freeze would effectively do) will not necessarily reduce defaults. The difference in default performance between the two instruments may reveal more about the borrowers than the instruments themselves. ARM borrowers by their very nature are more prone to risk-taking that FRM borrowers.

Finally, some of the most potentially explosive mortgage products, such as Option-ARMS, have negative amortization features—that is, borrowers actually accrue interest, so their mortgage balance rises each month for awhile. These features produce default incentives independent of changes in interest rates; an interest rate freeze will not help home borrowers who are under water because early payments did not pay for all interest due.

II. Freezing Rates on all Adjustable Rate Mortgages could reduce the availability of mortgage credit in the future.

Sometimes in the search to find the source of a tragedy, people try to settle on a villain. There have been lots of villains to go around in the sub-prime tragedy: unscrupulous brokers, speculative borrowers and lax rating agencies among them. But to the extent that investors in mortgages are villains, they are already being punished pretty severely, as they have taken on billions upon billions of dollars of losses.

It is therefore not surprising that investors—some of whom include the pension funds that provide the retirement incomes for schoolteachers and firefighters—are being frightened away from mortgages. The upshot is that the subprime market has nearly vanished for the time being, and mortgage credit more generally has become more expensive and scarce in high cost areas such as the coasts.

In light of this, policy responses that add to the perceived risk of investing in mortgages—including safe mortgages – would be counter-productive. A policy whereby the government freezes interest rates is a policy in which the government changes the terms of contracts, and thereby adds both to the perceived and real risk of investing in mortgages. This will make it harder and more expensive for borrowers to obtain mortgage credit, and could spill over to the point where it becomes more difficult for all entities in the US—including the U.S. Government—to borrow.

Other countries around the globe, such as India and Korea, have experimented with major government intervention in the enforcement of mortgage terms. These have produced stunted mortgage markets. We need to remember that the United States had until quite recently the most robust mortgage market in the world. We also need to remember that the adjustable rate mortgage saved the mortgage market in the 1980s, when macroeconomic conditions were unstable. Had it not been for the adjustable rate mortgage (along with investor confidence that those mortgage rates could reset with the market), housing conditions in the US in the 1980s would have been far worse.

Sunday, February 17, 2008

What is wrong with this picture?

I read the following in today's Slate about Clinton and Obama sending representatives to the AAAS meeting this year:

About 400 science researchers, policy wonks, and journalists packed into a meeting-room to see the surrogates duke it out. In the Clinton corner was Tom Kalil, a ruddy, thick-necked bureaucrat armed with a PowerPoint presentation and a full clip of obscure facts about academe. (Did you know that the average age for receiving your first NIH R01 grant is 41?) Representing Obama was Alec Ross, a smarmy and pandering thirtysomething in shirtsleeves. “I'm one of those guys who’s deeply moved by data,” he said, and then failed to adduce a single obscure fact about academe through the course of the session.

Let me say that I think either candidate would make a fine President, although personally I prefer Obama. I am glad that they at least sent representatives to talk to scientists (the other party's hostility to science is an important reason why I don't support it). But I looked up both Kalil and Ross--and neither of them is a scientist! JFK and LBJ used Jerome Wiesner, who would later become President of MIT, for science policy advice. It occurs to me that there are lots of articulate, policy-minded, honest-to-goodness scientists out there. It would be good to hear from them.

Saturday, February 16, 2008

Airline Mergers and Urban Dynamics

Apparently there is a very good chance that Delta and Northwest will merge. Delta's three domestic hubs are in Atlanta, Cincinnati and Salt Lake City; Northwest's are in Minneapolis, Detroit and Memphis.

The distance between Atlanta and Memphis is only 380 miles. It is hard to see both hubs survive: Pittsburgh couldn't survive as a hub given its proximity to Philadelphia for US Airways (even though Allegheny, US Airway's predecessor) was originally a Pittsburgh company)and St. Louis couldn't survive as a hub after TWA's merger into American given its proximity to Chicago. Jan Brueckner and I have done separate papers showing that this is very bad news indeed for Memphis. My paper, moreover, shows that air cargo has little impact on local economic development, so the presence of FedEx at Memphis won't do much to soften the blow of a likely dehubbing there.

Thursday, February 14, 2008

Sometimes Harvard Really is Great

I really like this policy that was just approved by the Faculty of Arts and Sciences there.

David Wessel has a Nice Piece on House Price Indexes in today's WSJ

It is here.

Let me repeat something I wrote here last September:

The two indices have three principal differences:

OFHEO has only houses financed with conforming mortgages (those with balances of less than $417,000). CS uses all sales.

OFHEO includes appraised values for refinance loans. CS looks only at sales.

OFHEO includes houses that may have been substantially improved from one sale to the next. CS throws those houses out.

I think the last of these three points is especially important in partially explaining the difference between price indexes.

Wednesday, February 13, 2008


I like the name of the tune.

Postal Experiments

One of my students, Jocelyn Leitzinger, forwarded this to me. It is fun.

Some hope on housing from Pasadena

I was running some rent to price ratio stuff there. I basically get that rent near Cal Tach and Downtown Pasadena is about $18 per square foot per year, and condos sell for about $400 per square foot. This give a ratio of 4.5 percent, or a price to rent ratio of 22. For purchasers requiring mortgage of less than $417K (an amount that will soon go up), mortgage rates are in the mid-fives right now, so the appreciation one would need to make one indifferent between renting and buying is between 3 and 4 percent (I will share my spreadsheet that does these calculations with anyone that is interested). This is not an unreasonable expectation.


I am teaching in Hyderabad for a few weeks in March. Fortunately, my kids' spring break coincides with some of the time I will be there, so they will fly over and join me. The cost of a roundtrip ticket on Qatar Airlines (change in Doha) is $1350. The Great Circle distance between Washingon and Hyderabad is about 8500 miles, or 17,000 miles round trip. This means that the cost of getting from Washington to Hyderabad is about 8 cents a mile.

Monday, February 11, 2008

Favorite Variations

Bach, Goldberg Variations

Beethoven, Diabelli Variations

Brahms, Handel Variations

Schumann, Symphonic Etudes

Rzewski, The People United will Never be Defeated!

The last is good regardless of your politics. The recording I link to is by Stephen Drury, my good friend and piano teacher from college. It is also the best recording of the piece.

It was cold in Washington today

So it is good to think of spring in Paris.

Saturday, February 09, 2008

Is Everything we "Know" about Subprime Wrong?

Yesterday I saw a great talk by Paul Willen of the Boston Fed. An earlier version of the paper he gave is here.

I don't think I am caricaturing what he said to say when I describe the following takeaways from his talk:

(1) Falling house prices lead to defaults more than defaults lead to falling house prices. In the early
part of the decade, when delinquencies reached record levels, default rates remained extremely low, because house prices were rising (see his graph on Page 48).

(2) Interest rate resets have little if anything to do with the large number of defaults we are observing. For the vast majority of subprime loans, defaults or refinances happen before resets take place. It is moreover the case that the size of the resets is smaller than most of us think, because the initial teaser rates are pretty high.

(3) While ARMs have higher default rates than FRMs, putting ARM borrowers into FRMs will not necessarily reduce defaults. In the first place, as people move out of ARMs into FRMs, the denominator in the default proportion ratio will decline for ARMs, while increasing for FRMs, meaning that if the number defaults stay constant, the default share in ARMs will still rise.

Beyond this last point, a working paper I have with Cutts and Ramogopal shows that ARM borrower are fundamentally more likely to be risk-takers than FRM borrowers, so the difference in loan performance between the two products may say more about the distributions of the different borrowers, rather than the products themselves.

In any case, all of this means that many of the policies being pushed at the moment (such as interest rate freezes) may not be particularly helpful in resolving the crisis.

Friday, February 08, 2008

Couldn't help but notice

It is almost certain that a sitting Senator will be elected President. Hasn't happened since JFK.

Thursday, February 07, 2008

John Dickerson in Slate has a Great Line

On Romney:

"He was a PowerPoint candidate who was all points and no power."

A thought all business school professors should take to heart.

Wednesday, February 06, 2008

How houses are like movies

Brad Delong has a nice piece on selectivity bias and the quality of movies. His basic point is that when people say "they don't make them like they used to," they are comparing the best movies from past generations with average movies from ours. There is a natural enough reason that we do this: the only movies from the past we still watch are Hitchcock/Wilder/Huston/Capra type movies; we don't watch the mediocre stuff from the past anymore. The same is true for novels: Dickens was not the representative novelist of the 19th century.

Houses are like this too. In many markets, a hedonic regression will yield a U-shaped relationship between house vintage and value. Here in Washington, houses built in the 1920s and 30s are especially valuable. An important reason for this is that the typical house is removed from the stock between age 70 and 80 (Jack Corgel did the definitive work on this some years ago); the houses that remain are those that were especially well constructed, or have unusual and desirable design features.

Tuesday, February 05, 2008


On December 11, I wrote this.

I would have mentioned it had I gotten it wrong, too.


I have not written for the last few days, because I was in Chicago with my daughter, who was visiting colleges and auditioning for a theater program. I stayed with my cousin, Jonathan, who teaches at Kellogg. Jon, Hannah and I had a nice lunch at a deli in the loop, and had a really nice time talking about his recent paper on experts.

The point of Jon (and Nabil I. Al-Najjar's) paper is that the best way to determine expertise is to compare predictions in some meaningful way. It is not sufficient to call someone an expert for choosing some mean outcome (if it rains 20 percent of the time, and a weatherman predicts a 20 percent chance of rain everyday, he will, in a sense, be right, but he adds no information, no expertise).

This got me thinking about Macroeconomic forecasting models. The Wall Street Journal will celebrate an economist's forecast if it is the best for a particular quarter. This is a strange metric: if someone predicts 6 percent GDP growth (an unusually high but not unheard of number) every quarter, in the unusual quarter where growth is that fast, he or she will be lionized in the pages of the country's leading financial newspaper.

A possibly better way to measure macroeconomic forecasts is to look at the mean squared error of forecasts over a reasonable time horizon--say 20 quarters. Perhaps someone has performed the exercise--it would be fun to know how it would shake out. My guess is that the eternal optimists and doomsayers wouldn't look so good.

Friday, February 01, 2008

My favorite cities that have lost more than 10 percent of peak population

1. Chicago (my favorite American City)
2. Duluth (incredibly nice people; incredibly nice lake shore)
3. Boston
4. Philadelphia
5. Washington
6. Minneapolis
7. Pittsburgh (great college town!)
8. Cleveland (great art museum, even greater orchestra)
9. New Orleans (authentically tawdry, as opposed to Las Vegas)

Music for 18 Musicians Piano Rehearsal 01