With Much Blood-Letting to Come, the U.S. Housing Finance System Needs Replacing

By Martin Hutchinson
Contributing Editor

In much of the discussion about the collapse of the U.S. housing market, commentators have assumed that the massive run-up in property prices that preceded the subprime-mortgage meltdown were simply the result of a speculative frenzy that became a full-fledged market bubble.

But that's not the case at all.

You see, the bubble and subsequent crash were inevitable under the current system of housing finance. Fundamental changes must be made.

Standard and Poor's recently projected the likely future loss rate on the $650 billion of subprime-mortgage-backed securities that are still out in the marketplace. From that we can estimate the losses S&P is projecting on the actual mortgages themselves.

According to S&P, senior AAA-rated bonds will pay out about 60% of principal, junior AAA-rated bonds about 35%, AA-rated bonds about 5% and lower-rated bonds nothing at all. Since about 75% of subprime mortgage-backed securities were AAA rated, we can calculate that S&P thinks subprime mortgages will eventually return about 40% on the original principal amount.

That's a startling number.

Losses Still to Come

If you had a portfolio consisting entirely of 100% loan-to-value mortgages, on which the appraisals were accurate but a large percentage of the borrowers had poor credit, and house prices were destined to drop between 20% and 25% over the next few years, you'd expect to lose 25% - or perhaps 30% - of principal, but still manage to keep 70% to 75% of your money.

When you had a foreclosure, there would be costs involved that increased your loss. On the other hand, some of the borrowers would be able to make their mortgage payments, leaving you with no loss at all. Thus, if subprime mortgages are expected to return only 40%, almost half of them must have had some fraud involved, either by the borrower, the mortgage broker or the appraiser.

Let's now turn to actual housing prices. The S&P/Case-Shiller Home Price Indices of home prices in the Top 20 urban markets dropped a bigger-than-anticipated 12.7% in the 12 months that ended in February - the worst showing since the index debuted in 1991. What's even more alarming, however, is that the decline is accelerating. In February alone, prices dropped 2.7%  - the equivalent of a 28% decline if this rate persisted for the entire year.

That should have alarmed both homeowners with large mortgages and mortgage market participants - if prices were to drop 30% to 40%, instead of the generally expected 15% to 20%, even prime home mortgages would get in trouble and the losses would be appalling - in the range of multiple trillions of dollars.

Since the first-quarter vacancy rate in U.S. housing - owner-occupied and rental - increased to 2.9%, the highest level in 50 years, we may indeed be approaching such a bearish scenario.

However, when you look at factors like the ratio of house prices to incomes, it becomes obvious that the problem is not the current drop, but the previous rise. Since World War II, the average house price was 3.2 times the average income. By 2006, however, the average house price had jumped to 4.5 times the average income. With house prices outrunning incomes in that way, mortgage financing was bound to become more and more risky, and a substantial drop was eventually inevitable - to take prices from 4.5 times income to 3.2 times would require housing prices to plunge 29%. And that doesn't even consider the possibility that prices might overshoot on the downside. 

Story continues below...

The principal reason for the excessive rise in house prices and the high level of fraud was the housing finance system. In the modern system, the originator of a home mortgage loan is paid a fee on the origination, and never has to worry again about the credit risk on that loan, which is passed off to investors through a process known as "securitization."

Because securitization separates the mortgage originator and the actual investor, the investors - often foreigners - have no idea of the actual underlying quality of the loans that they're purchasing.

The mortgage broker's incentive is to maximize loan volume - pretty much regardless of whether or not the borrower can afford the loan. Falsification of documents, suborning of appraisers, and other similarly reprehensible machinations becomes a normal course of action in such a situation, as does turbo-charging the housing market to valuation and sales levels it cannot sustain. A system in which prices are forced up to unsustainable levels and fraud is rampant is broken, and needs to be replaced with something better.

It (Once) Was a Wonderful Life

Years ago, the United States had a superior home-financing system; it was extolled in the 1946 Jimmy Stewart movie, "It's a Wonderful Life." Home-mortgage loans were made by local institutions to borrowers whom they knew personally. The system had some inefficiencies. For example, if the housing needs in a particular area expanded rapidly, there might be a shortage of funds, so that mortgages would be unavailable. However, banking is mostly national today, so local funds shortages would be less important, although there would probably be a corresponding decline in personal knowledge of the borrowers.

It's not true that the Jimmy Stewart system of financing home mortgages was less efficient than today's: That's a myth put out by Wall Street, which has been one of the chief beneficiaries of the recent shenanigans. (Riddle me this: Do you think that "George Bailey" - Jimmy Stewart - ever got a million-dollar bonus?).

If you look at the U.S. Federal Reserve statistics on U.S. interest rates (which started recording home mortgage rates in 1972), you will discover that in 1972-78 - when the Jimmy Stewart home financing system was still mostly in place - 20-year Treasury bonds yielded an average of 7.41%, while 30-year fixed rate home mortgages yielded 8.49%, a differential of 1.08%. In 2000-06, an equivalent period that predates the recent worries about credit risk, 20-year Treasuries yielded an average of 5.28%, while home mortgages yielded 6.50%  - a differential of 1.22%. 

Thus, the "spread" of home mortgage interest costs over Treasury bond yields, the most appropriate measure of home mortgage costs, has widened by 0.14%. That may not sound like much until you realize that it's an effective cost increase of 13%. Where did that increase go?

While some lawyers made money, too - what did you expect - it's largely Wall Street that would rather you didn't think about that question.

Two Key Changes

We can move back a long way toward the Jimmy Stewart system, but to do so two things must be changed:

  • First, we need to close down Fannie Mae (FNM) and Freddie Mac (FRE), the government-sponsored enterprises that guarantee most home mortgages. By putting a quasi-government guarantee on the mortgages - they're not quite government-guaranteed, but close (another worry for the taxpayers of America) - you take away the local credit risk and make them easy to pass around the money pits of Wall Street. Most countries don't have the government-guarantee home mortgages; it's entirely unnecessary as good home mortgages are fine credit risks on their own.
  • Second, you need to enact some kind of small transfer tax on paper shuffling that makes securitization of mortgages expensive. The British have a 0.5% stamp duty on share dealing; even a 0.1% duty per transfer would probably be enough, here. It would make packaging the mortgages and turning them into mortgage bonds just a little less competitive, compared to the other alternative of having local banks or bank branches lend directly.

Once the financial incentives had been tilted against securitization, and the foolish government guarantees had been removed from the mortgage market, home-mortgage lending could revert to the Jimmy Stewart model.  (You might want to leave a small loophole for low-income housing, on which mortgages - otherwise non-creditworthy - still could be guaranteed by Ginnie Mae, part of the U.S. Department of Housing and Urban Development; it seems voters want the government to subsidise housing for the poor and help them get mortgages).

Overall, under the new model, home mortgages would be a little cheaper, you would go to a proper bank - and not a salesman - to get them, and housing prices would remain reasonable.

Of course, thousands of rich Wall Streeters would be thrown out of work. There's a pity.

News and Related Story Links: