A point and a question

Doh.

Here is the "nice piece of work" I mentioned.

I agree with much of the analysis – at least up to a point. I have to disagree with where Liebowitz places the blame, though.

He largely points to the government, claiming the government chose to deregulate because they wanted to enable low income and minority Americans to own homes. The clamor for deregulation came from the financial industry, who sold the idea as a way to increase home ownership.

Also, I think the connection between the early steps identified (1-4) and the rest is a little weak. What happened in between bears discussion.

Home mortgages are rarely held by the lender for their entire lifetime. Most are sold quite rapidly. The buyers bundle together multiple mortgages into a single trading instrument. The theory assumes that mortgage failures (foreclosures) are independent events. If one mortgage has a 5% chance of failing, and I own 100 identical mortgages, then instead of a 5% chance of losing everything, I have 100, independent, 5% chances of losing 1% of my total investment.

If I take that bundled collection of 100 mortgages and sell 1% each to different investors, then they’re getting the same interest rate as on an individual mortgage, but all 100 of them have to fail for them to lose their whole investment. That’s a much nicer risk profile.

So we’ve taken a relatively risky investment – a single mortgage – and "magically" turned it into a fairly low-risk investment. Our insurance is cheaper, and it’s a big hit with the investors.

What happens next? Well, everybody wants in on the action – high yield, low risk investments are exactly what the market wants. You can sell these by the truckload.

Except for one thing. You’re limited by the number of mortgages being sold each year.

But you’ve got a zillion investors willing to buy – the demand is there, if only we could increase the supply to meet it. But increasing the supply means selling more mortgages. But the financial incentives are really already there – almost anybody who can possibly afford a house buys one just for the tax breaks. Where, oh where, can we find more mortgages?

There’s the ticket – if "anyone who can possibly afford a house" will buy one, let’s make it possible for more people to afford one. Of course, that means going a little deeper into the "risk" pool – one of the reasons those people couldn’t afford a house is because nobody would risk lending them the money.

So the government gets a ton of pressure to fiddle with the regulations, ostensibly to enable more Americans to own their own homes, but the result was to jack up the supply of mortgages to bundle into these magic investment instruments.

Unfortunately, mortgage defaults aren’t entirely independent. When demand goes down, and prices drop, those borderline loans start to fail. The investment needs to make up the difference by bumping the variable interest rates, which starts the snowball – whole groups of loans all fail together. That bundled investment wasn’t quite as risk-free as it seemed.

No problem – we insured the investment. We’ll just file a claim. But, we insured them with another financial institution, who also puts together these little investment bundles, we even underwrote a few of theirs – after all, they’re low risk investments, we’re not likely to have to pay off, anyway. Now we’re claiming our insurance, and they’re claiming theirs, and nobody’s getting paid.

That’s a lot of how we got here. It’s not housing speculators – they’re an effect, not a cause. People started speculating because home loans were easy to get. The financial institutions are at least as culpable as the government in this, too.

And finally, the whole concept of insurance is about carefully assessing risks. The underlying failure here was that the underwriters used incorrect risk assessment models to set prices. If they hadn’t, the insurance would have been enough to cover the failures, which would have pushed up the price of the bundled instruments, which would have lowered their demand (ultimately to zero).

Greenspan testified to Congress recently and admitted he was wrong in some of his assumptions. Forecasting isn’t an exact science, anyway, and being wrong happens. He also said that they didn’t have a clear picture as to how far astray things had gotten – an information problem. His conclusion was that the banks that were insuring these investments should not have been permitted to sell the entire risk – they should have been required to keep some of it in house, which would have made it in their own best interests to ensure that the risks were modeled accurately.

That seems like a fine start.

On the other hand, these same financial organizations are now the ones receiving well over a half a trillion dollars in "bail out" money. So maybe they did model the risks accurately after all – they just didn’t mention they were counting on free government "insurance" payoffs to cover the coordinated failure case.

Funny you should ask. I recently wrote a post about that – Stop credit card junk mail.