Arnold Kling  

My Fantasy Testimony

An Education Debate... Monetary Institutions...

This weekend I typed up "draft testimony" on what caused the mortgage/financial crisis. Like anyone wants to hear it. Anyway, I'll paste it in below the fold.

UPDATE: If you follow the links in James Hamilton's post, during his talk (starts around minute 8), he really spells out the difference between what I call Method A and what I call Method B.

Draft Testimony
Arnold Kling
October 5, 2008

Forty years ago, depository institutions handled mortgage credit risk very differently than they do today. Back then, the depository institution, which was typically a savings and loan association, held mortgages that were underwritten by its own employees, given to borrowers and backed by homes in its own community. These were almost always 30-year, fixed-rate loans, with borrowers having made a significant down payment, often 20 percent of the price of the home. Call this approach to mortgage lending "Method A."

Today, mortgage loans held by depository institutions are often in the form of securities. These securities are backed by loans originated in distant communities by unknown borrowers, underwritten by mortgage brokers or other personnel not employed by the depository institution. The loans are often not 30-year fixed-rate loans, and the borrowers have typically made down payments of 5 percent or less, including loans with no down payment at all. Call this approach to mortgage lending "Method B."

If you compare the two methods using common sense, then Method B does not pass a simple sanity check. In fact, the current financial crisis consists of banks that are up to their necks in Method B.

Method A suffered a breakdown in the 1970's, because inflation was allowed to get out of control. The 6 percent mortgage interest rates that were commonly charged by savings and loans became untenable when inflation and interest rates soared to double-digit levels. The savings and loan industry went out of business. Whether Method B could survive a similar shock is unclear. The right lesson to learn from the 1970's was not that we should use Method B. The right lesson to learn is that we should not let inflation get out of hand.

Many articles have been written by economists from academia and Wall Street extolling the benefits of Method B. They speak of the wonderful innovation of mortgage securities and the supposed efficiency of the secondary mortgage market. In fact, if Method B were to compete fairly with Method A in a market test, Method B would fail. To survive against Method A, Method B requires government favors and subsidies. It always has, and it always will.

The secondary mortgage market began in 1968, when the United States formed the Government National Mortgage Association (GNMA). GNMA pooled loans originated under programs by the Federal Housing Administration (FHA) and the Veterans Administration (VA) and sold these pools to investors. The purpose of this, as with the quasi-privatization of the Federal National Mortgage Association (Fannie Mae) that took place that year, was to take Federally guaranteed mortgage loans off of the books. President Johnson, fighting an unpopular war in Vietnam, wanted to save himself the embarrassment of having to come to Congress to ask for larger and larger increases in the ceiling on the national debt.

Thus, the first steps toward mortgage securitization were taken in order to disguise financial reality using accounting gimmicks. It has been the same ever since.

In the early 1980's, the savings and loan industry was imploding. The savings and loans held many mortgages that had lost value, due to inflation and high interest rates. Under the accounting rules prevailing at the time, they could record the mortgages at their original book values--as long as they did not sell them. The S&L's were stuck. If they did not sell mortgages, they would lack the cash to pay depositors. If they did sell mortgages, they would have to recognize losses, and regulators would shut them down.

The solution was a program called Guarantor at Freddie Mac and Swap at Fannie Mae. Under this program,, an S&L paid Fannie or Freddie a fee to pool loans into securities, which were retained by the S&L. The S&L could then use the securities as collateral to obtain loans. The key to the whole operation was an accounting ruling that allowed the S&L's to maintain the fictional book values of the mortgages held as securities,, even though lenders were using much lower market values when providing the collateralized loans to the S&L's. This accounting ruling came from fierce lobbying by Wall Street, which wanted to broker the loans to the S&L's.

Freddie Mac, Fannie Mae, and Wall Street extracted large fees from this program. Ultimately, most of the S&L's failed, with losses borne by taxpayers. The Guarantor and Swap programs increased the cost, both directly by transferring wealth out of the S&L's and indirectly by allowing defunct S&L's to remain in business and continue gambling with other people's money.

Following the demise of the S&L's, regulators established capital requirements and other rules that made Method A lending very expensive relative to Method B lending. Freddie Mac and Fannie Mae were given freedom to leverage their guarantee at much higher ratios of assets to capital than were Method A lenders. In fact, the capital regulations even told banks that holding private mortgage securities under Method B was less risky and therefore required less capital than Method A lending. The result was that Method B came to dominate the American mortgage market. Had the playing field been level, Method A would have remained in place, and Method B likely would never have gotten started.

The regulations that favored Method B were heavily influenced by the lobbying activities of Freddie Mac, Fannie Mae, and Wall Street investment bankers. The secondary mortgage market lobby made large campaign contributions to key legislators. Their influence tilted regulations to favor Method B. Thanks to the regulatory advantages given to Method B, Freddie, Fannie, and the investment banks earned large profits. It was a triangular trade: contributions for favors for profits.

The recent "rescue plan" can be viewed in this context. It attempts to revive the otherwise moribund mortgage securities market. It is another large favor being granted to Method B.

There has been a flight to safety in credit markets in recent weeks. What is probably a necessary and significant consolidation in the financial sector has turned into a rout.

Under the circumstances, there are two policy imperatives. First, regulators must sort out the banks that are sound from those that are insolvent. The insolvent institutions need to be merged or closed as expeditiously as possible. Sound banks should be encouraged to make loans to qualified borrowers. Some forbearance of capital requirements may be appropriate in order to ensure that good borrowers do not get turned down. Finally, there may be some banks that are neither clearly insolvent nor clearly sound, in part because of questions concerning the values of their mortgage securities. These banks should be allowed to continue operating, under close supervision, perhaps with loans from the Federal Reserve.

Under no circumstances is it justified to attempt to revive the mortgage securities markets. Resumption of active trading in those markets is neither necessary nor sufficient to address tightness in credit markets caused by the flight to safety.

Next, I wish to turn to examine the state of the housing market. One of the characteristics of Method B lending has been loans with low down payments. This caused a wave of speculation, leading to a sharp rise in house prices followed by a sudden decline. In fact, this is to be expected when most homes are bought with little or nothing down. When prices or rising, anyyone can afford a home. When prices are falling, on one can.

I wish to emphasize that I support the effort to increase the rate of home ownership among minorities and people with low incomes. The problem with Method B lending is not the color of anyone's skin or the content of anyone's credit report. The problem is the absence of a reasonable down payment. In my view, assistance with saving for a down payment would be better than a system that subsidizes indebtedness. I think of buying a home with little or no money down as home borrowership, not home ownership.

The housing market today is distorted and out of balance. The slogan "Let's keep borrowers in their homes" is rather useless, considering that many borrowers never occupied their homes to begin with. The Washington Post of December 10, 2007, contained a very illuminating story of people with very modest incomes who bought two and three homes during the speculative frenzy. (

The plural of anecdote is data. In an article published in the Federal Reserve Bulletin of December 2007, economists Robert B. Avery, Kenneth P. Brevoort, and Glenn B. Canner reported that the share of mortgage loans going for non-owner-occupied housing went from 6.4 percent in 1996 to 17.3 percent and 16.5 percent in 2005 and 2006, respectively. (

The United States today has an unusually large inventory of unoccupied housing units, which speculators have been unable to sell or rent. Given the unoccupied housing glut, attempts by policymakers to try to boost house prices are likely to prove futile. As long as the housing market remains out of balance, the valuation of mortgage securities will be problematic. As difficult and painful it may seem in the short run, it probably would be better to work through the foreclosure process and let prices adjust to levels that bring supply and demand into balance in the housing market than to prolong the state of imbalance and uncertainty by trying to prevent foreclosures. This is particularly true for the many homes that were bought for speculation, not as primary residences.

Finally, I wish to call your attention to a communication gap that exists between the financial engineers who design and value mortgage securities and the executives and regulators who make important decisions about how they are used. I call this the problem of the "suits" (the executives) and the "geeks" (the financial engineers). In many cases, the suits at banks and other financial institutions are putting these securities in their portfolios without appreciating their risk characteristics. Regulators, too, seem unaware of the threats posed by these complex instruments.

Charles Duhigg of the New York Times has written two stories, one about Freddie Mac on August 5th ( and one about Fannie Mae on October 4th (, that illustrate the suits vs. geeks divide. Each story cites former mid-level executives of the companies who issued warnings about risk, mis-pricing, and under-capitalization. But the geeks were ignored by the suits.

Robert Merton, a Professor of finance at Harvard and a Nobel Laureate, suggests that the decisionmakers who I call "suits" need better training in modern mathematical finance. That is one solution. Another solution would be to try to limit the ways in which under-educated "suits" can expose their firms and our economy to risk. As I indicated earlier, I do not believe that the mortgage securities market would have emerged without regulatory favoritism. If such favoritism were taken away, and the mortgage securities market were to fade away as a result, then the communication gap between the suits and the geeks would cease to be a problem.

Comments and Sharing


TRACKBACKS (1 to date)
TrackBack URL:
The author at PrestoPundit in a related article titled ARNOLD KLING HAS writes:
    a must read on what went wrong in the housing mortgage industry.... [Tracked on October 7, 2008 1:32 AM]
COMMENTS (24 to date)
Jim writes:

I guess I'm missing the point.

Let's assume Bailey Brothers Building & Loan holds mortgages on 500 homes in Bedford Falls, said loans being its sole asset class. When Bedford Falls' principal employer closes its doors due to Chinese competition, would a run on the Building and Loan be more or less likely if, instead of the individual mortgages, it held a portfolio of MBS?

Nick writes:

Thank God for this. I've read probably a hundred articles on the mortgage crisis, and this is the only one that is written clearly enough to make me believe that the author is thinking clearly.

Francis writes:

I have a question.

Method A has always been the one used in Canada. It still is. Yet it survived the inflation of the 70s and 80s, which was the same here as in the U.S. Why?

It is still the same banks (really the same: Royal Bank, Bank of Montreal, etc.) that supply loans for mortgages. So why didn't they go through the same troubles?

DWAnderson writes:

I have found your posts incredibly illuminating and your apearance on discussing the same subject was excellent as well.

However, I think you need to better explain why you think the mortgage securitization market would not exist absent government favors. Absent that explanation I would think such a market would allow finacial firms to pursue what they are best at, e.g. firms that are great at originating loans (including evaluating default risk) could do that and firms that are best at aggregating investment funds or deposits could focus on that. I would think that sort of specialization would be made possible by a market for mortgage securitization and would normally be a good thing.

BTW, you made a strong case that government favors created the market and indeed crowded out the alternatives, I just don't see why the market would so clearly not exist (albeit in a lesser form) absent those favors.

Reprinted at

sohaib writes:

I've been reading this blog for around a year now but have never commented. I feel compelled to on this entry. It is simply excellent. Thank you.

Steve Roth writes:


On suits and geeks: just to point out some excellent reporting from Bloomberg suggesting that the same was happening at the rating agencies. (Though the reporting suggest at least borderline criminality by the suits there.)

It also suggests that your previous comments on the ratings agencies' minimal effects were perhaps...sanguine.

cputter writes:

A very clear and concise analysis can be found at:

It clearly shows the government's influence in promoting method B, and differentiates well between A and B. It lays greater fault on ARMs then lack of down payments though.

It also makes the case that there is/was no real difference between the rate of home ownership among minorities and the rest of the public.

Its quite thorough and worth the read.

Ami writes:

Thanks for the elaborate, yet readable history to the present problem in mortgage securities market. Very informative.

Frejus writes:

Interesting analysis from a person who was arguing for years that housing prices could very well be priced just right.

E.g. you said: "For housing, I would rate the probability of a bubble at about 20 percent."


Rick Stewart writes:

Arnold supports 'the effort to increase the rate of home ownership among minorities and people with low incomes.' Why?

Only 43% of Germans own their own homes. About 90% of Mexicans own their own homes. The US is half way between, and you want us to move in which direction?

I suggest the government has no business caring how many Americans own their own homes, which Americans those are, nor how many homes presidential candidates own.

Nick Rowe writes:

Francis above asks why Canadian banks, which use method A, did not fail in the 70's and 80's.

Jim above points out that method A fails if the banks have all their eggs in one local housing basket.

I think that might be the key? Canadian banks are bigger (no rules against interstate banking) and so more regionally diversified. And more diversified outside the mortgage market as well.

Thomas Dinsmore writes:

One historical point. The traditional building and loan did not make thirty-year fixed rate loans; the typical mortgage prior to the Depression was a five year interest-only balloon note with no prepayment penalty. The homeowner rolled over the balance into a new loan at a new rate at the end of the term.

Many employed and creditworthy homeowners defaulted during the Depression because their note came due during the financial crisis and it was impossible to refinance the balance due.

Fixed rate long term mortgages are primarily a phenomenon of the US market and are rare in Canada and elsewhere.

Paul writes:


You don't say anything about the pressure brought to bear by government to force banks to make home loans to people with low incomes and bad credit. This began in earnest in the 1990s. Do you not see that as a significant factor?

Arnold Kling writes:

I think of the pressure to lend in those markets was part, but only part, of the push toward Method B lending. I think that a lot of the minority and low-income borrowers were actually good risks. But lending to anyone without a substantial down payment, particularly for speculating on a house that is not a primary residence, is not safe.

Francis writes:


Thanks for your answer. Diversification may be the key, you are right, but as for size, I don't think it matters because the big inflation and subsequent disinflation were nation-wide.


Fixed 30-year mortgages were not rare in Canada. They were no longer available in the 80's and 90's, but they resurfaced in the 2000's. In 1991, I heard an old person tell me that one of his relatives was still having one from the 60's.

i think this crisis told us to never invest more than our capacity, because every investment has a risk.

Alex J. writes:

In the commodities markets, it is routine for someone who buys corn from a farmer (which is a long position) to short corn through the CBOT. If a bank makes a 30 year fixed loan, that's a bet against inflation. Why can't the bank buy gold or short 30 bonds or make some other counter-balancing investment to hedge their interest rate risk?

Thomas Dinsmore writes:


Suppose it depends on the meaning of "rare". Am unable to quickly locate a source on Canadian mortgage originations; however, brief review of mortgage broker websites in Canada shows none offering terms greater than ten years.

Feel free to post a dispositive source.

Francis writes:


You are right, it all depends whether 'rare' means 'rarer than in the U.S.' or not. Yet they are available. Find here a fixed-rate 25-year mortgage:
(I guess you won't argue for the 5-year difference, will you?)


You are also right that bankers here will often talk about their 'appariement' (French for 'matching'), meaning that they must match somehow the number of X-year mortgages they sell with the number of X-year closed deposits they get. I don't know, though, if this is a regulated practice or not.

Thomas Dinsmore writes:


The Canadian government thinks that long-term fixed rate mortgages are sufficiently rare that this primer states that fixed terms are available "up to ten years".

Kevin writes:

Maybe I have no common sense, but why is Method A better than Method B? It seems that Method A concentrates risk for lenders and exposes borrowers to the unique preferences of a particular lender in shopping for rates. Sure some of the products are riskier, but I take the "Method" to mean securitization of mortgages generally.

Don't forget that 30-year fixed mortgages with 20% down are securitized too. Are you saying that Method B pooling and securitization of Method A style mortgages is inefficient in and of itself because of the value destroyed by originator moral hazard? If not, why blame Method B? Why not just point the finger at the riskier products?

Rich writes:

"These were almost always 30-year, fixed-rate loans, with borrowers having made a significant down payment, often 20 percent of the price of the home."

I am new to the mortgage debate, but this claim seems inconsistent with the data in the Statistical Abstract. The SA provides the percent of loans with adjustable rates beginning in 1982, and they only provide it for "Conventional First Mortgage Loans for Purchase of Single-Family Homes." Still, the data show a fairly clear downward trend in the percentage of loans with an adjustable rate, and adjustable rates were extremely common in the early 1980s. In 1984 they accounted for over sixty percent of all conventional loans to purchase existing homes. Again according to the Statistical Abstract, S&Ls still dominated mortgage lending at the time - accounting for 43% of all outstanding home mortgages in 1982 and 40$ in 1984; Agency and government-sponsored enterprises-backed mortgage pools accounted for just 16% in 1982 and 21% in 1984. Is the problem that the data excludes subprime loans? That it excludes refinancing?

Richard Boltuck writes:

Questions for Arnold:

First, regulatory favoritism results from rent-seeking/lobbying activities. How do you propose to limit this?

Second, with securitized mortgages, is there any way to anticipate and manage systemic risk, or is it an inherent and destructive externality?

Third, you do not expressly discuss moral hazard and arising from deposit insurance and "too-big-to-fail" insurance. If not offset/controlled by appropriate regulation (and recall that regulation is subject to industry capture), moral hazard leads to excessive risk-taking (what you call in a related context, gambling with other peoples' money) for short-term profit, a rational strategy. Is there any solution within a democracy to the pernicious combination of ineffective regulation and the time inconsistency that assures that government will always bailout too-big-to-fail institutions (or too-many-to-fail smaller institutions)?

Fourth, are other innovations in mortgages, such as recourse lending that turns the current one-way borrower bet on rising home prices into a legitimate two-way bet? American mortgages are currently non-recourse. (A financial columnist at the Australian wrote the other day in the WSJ that mortgages in Australia are recourse loans, so underwater borrowers are still pursued for the full amount due). Granted, 20% down goes part way toward this goal, but changing the legal terms would also move the goal posts. Would it be prudent to do so?

Rich writes:

I think I found the answer to my own question - S&Ls were not allowed to issue adjustable rate mortgages until passage of the Garn-St Germain Depository Institutions Act in 1982. However, it seems difficult to attribute the rise of adjustable rate mortgages to regulations that purportedly encourage them when your comparison is to a time when they were effectively not allowed.

Comments for this entry have been closed
Return to top