Arnold Kling  

A Question for Brad DeLong

PRINT
My Debate with Ian Fletcher... Swagel on Mortgage Finance...

Tyler Cowen links to Brad DeLong who writes,


Arnold Kling's response is simply not good. It is silly enough to make me think he has not thought the issues through. a 7% delinquency rate on a mortgage portfolio is horrible in normal times, but is actually very good if you are in a depression--ever our Lesser Depression. For an investment with a 15-year duration that's a cost of less than 50 basis points in a "black swan" near worst case scenario. A portfolio that does that well under such conditions is a solid gold one.

I have a question for Brad. Does he know the fee that Freddie Mac and Fannie Mae charge for mortgage guarantees?

I really do not like to get involved in personal catfights. I think they make both people look bad. But I am not the one whose thinking about what constitutes a high mortgage default rate is silly.


Comments and Sharing






COMMENTS (14 to date)
coyote writes:

First, I have no clue what a "reasonable" default rate is in a black swan event, and my guess is that, almost by definition, no one else does either.

However, it strikes me that DeLong's argument is a bit off. If mortgage default rates went up in an economic crisis that was wholly unrelated to mortgages, ie due to an oil shock or something, that would be one thing. But in this case, the black swan is in large part due to the mortgages issued. I guess it is sort of a chicken and egg problem, but the mortgages started defaulting before the depression, not the other way around, and helped precipitate the depression.

As a result, I don't think you say that "7% is reasonable in a depression" because the depression is not an independent event that in turn caused the defaults.

mark writes:

"very good if you're in a depression"? what kind of standard is that? An endless stream of parodies suggest themselves.

Fwiw, his quantitative exercise seems to confuse delinquency with loss and of course ignores the losses on loans outside the prime sector.

Bob Murphy writes:

Prof. Kling, I urge you to write more on this. Krugman and DeLong are making much of this "revisionism" saying that Fannie and Freddie didn't make reckless loans.

I am fairly up-to-speed on this issue, and I am predisposed to side with you over DeLong, yet your quick line here baffles me. I.e. I don't know what point you are making, so I think you should please spell it out for us.

Let me summarize it this way: I often blame the Fed and GSEs for the housing bubble, but because of Krugman and DeLong's constant referral to these types of statistics, I'm not sure I should be doing so. If you think they're wrong, please spell it out.

Joseph K writes:

I think what Delong fails to see is that an economy with a low savings rate and lots of borrowing is precarious. You can imagine this easily on an individual level. A person who has lots of savings can easily hand financial difficulties. Even if they lose their job, they can live off their savings while between jobs; the more their savings of that person, the longer they can be without a job without their lifestyle being affected. Someone who's up to their ears in debt and living paycheck to paycheck but getting by, on the other hand, is very precarious, with something as simple as their labor hours being cut or having their health care premiums raised suddenly unravelling their whole financial situation when they can't make their payments for their house and their credit card and their car and so on.

The same thing applies economy-wide. During financial downturns, an economy with lots of saving can weather economic hard times more easily. As an individual, if your spending is unaffected by a job loss, then all the businesses that depend on your dollar will be unaffected. If you have low savings and much debt, then your job loss doesn't just affect you, but it affects a bunch of other people that do business with you and lend to you.

Ultimately there is a trade off, since the more people save and the less they lend and borrow, the slower growth will be for the economy, just as an individual can makes lots of money really quickly by taking big risks. So a balance will normally be struck, within an economy, between saving too much and taking on too much risk.

As soon as enter some sort of agencies that encourage borrowing, then that balance is heavily weighed in one direction. In our case we had GSEs encouraging people to borrow money for homes by lowering lending standards, which is sort of a double whammy, since you're both increasing the number of people borrowing, and decreasing the average income of the average borrower. It caused a boom in the housing market and lots of economic growth, but it also made the economy more precarious. Once you push leveraging so far, even the smallest economic shock, can cause the whole thing to tumble, and that's what happens. How exactly we want to distribute blame is a difficult question, but to exculpate the Fannie and Freddy is very difficult.

Additionally, it's not just the banks that lent that are hurt by high default rate. Delong's point that 7% isn't too bad is completely one-sided, since it really only asks if banks can, at least temporarily, manage a 7% default rate during tough times. Perhaps they can, but then there's the other side, the 7% of people actually defaulting. That's a lot of people in serious financial straits and that has consequences.

Hugh writes:

I read the whole of DeLong's post. It was poorly argued to the point of being incoherent.

Don't waste your time with this fool - he provides zero marginal value.

mdb writes:

Have to agree with Hugh, though a more thorough explanation would be nice - I always enjoy your insights.

keatssycamore writes:

Arnold Kling wrote in the comment thread of 'Mortgage Loans with Low and High Risk':

"@David Min:

If a 'normal economy' is one in which house prices do not fall, then in a normal economy there is no such thing as a high-risk loan. Worst case, the borrower sells the house rather than handing you the keys. Even a lot of fraud gets covered up by rising house prices.

So risk has to be defined relative to a scenario where prices fall. (They could fall locally, of course. They do not have to fall everywhere.) From that perspective, even a prime loan is risky if the loan-to-value ratio is above 90 percent. I would actually say that it's high when the LTV is above 80 percent. And in a cash-out refi, it's high when the LTV is above 70 percent, and even lower if the appraisal is hard to corroborate."

Trying to ascertain what's appropriately labelled high risk "relative to a scenario where prices FALL" seems as foolish as the Wall Street plan for labelling what was high risk by using a scenario where prices RISE.

In reality, prices rise AND fall and trying to label what is, or isn't, high risk utilizing only one or the other has already led to failure, no?

So I think I'm with Bob Murphy from above in wanting some further explanation of what you are claiming.

mark writes:

I'd also note that the capitalization of the GSE's was such that "less than 50 basis points over a 15 year duration" of asset impairment - if unevenly spread out - was sufficient to render them insolvent.

Bill writes:
a 7% delinquency rate on a mortgage portfolio is horrible in normal times, but is actually very good if you are in a depression

So which is cause and which is effect? I thought the argument was that the defaults caused the depression not the other way around.


For an investment with a 15-year duration that's a cost of less than 50 basis points in a "black swan" near worst case scenario.

Calling the blow up a "black swan" implies that it was hard to predict and beyond normal expectations. A fall in prices is hardly unprecedented. "The Big Short" suggest that this was hardly unexpected.

Æternitatis writes:

I hate to take DeLong's side against Kling's but:

As an attorney who was involved with drafting and reviewing a lot of the securitization documentation for Freddie and Fannie's underwriters, I can say that while the fees charged by the GSEs were generally trivial (a couple hundred grand per deal), the spread between the average coupon issued and the interest on the collateral mortgages (or mortgage pools) was typically 50 to 200 basis points. (I am not violating any confidences here--these documentation have long since been released in the public record.)

And, yes, given this large profit, I too find it puzzling that the GSEs went effectively bankrupt.

Æternitatis writes:

I do not understand the generalized contempt for cash-out refis. While we are all supposed to cheer for saving and heap scorn on spenders, but I hesitate to join the parade. Sometimes people will for good and valid reasons choose to use up savings (which is what a cash-out refi is) and I, for one, do not believe that I am in a better position to judge the wisdom of that choice than the person who makes it for themselves with their own assets.

As for riskiness, it is unclear why it matters whether a mortgage is a cash-out refi or an original purchase mortgage. All that matters is the final loan-to-value ratio. That is the cushion for price declines the lender has before the property becomes under water.

It is true that too-high ratings were given to mortgage pools with too-low LTVs. But whether they were cash-out does not enter. I for one would much rather be the lender on a 10% to 50% cash-out refi than on an 80% loan-to-value purchase mortgage.

Jacob Oost writes:

I never find it puzzling when a government-backed enterprise goes bankrupt.

ezra abrams writes:

regardless of who is right, delong is a much better writer: he takes the time, or has the skill, to write a graf that lets the uninitiated reader understand the issues, rather then cryptic for the cognoscenti only stuff.
if you care,

Michael Bishop writes:

[Comment removed awaiting response for clarification or relevance. Email the webmaster@econlib.org to request restoring this comment. --Econlib Ed.]

Comments for this entry have been closed
Return to top