“Question: Help me understand...Why would any mortgage company/bank make a loan that they knew would not be paid back? Only a law like CRA would make this possible? No one goes into business to lose money.”
Argent, Ameriquest, and the similar firms that they had spawned didn’t hold on to the mortgages that they wrote. So the mortgage writers weren’t exposed to the default risk. The mortgages were bundled and sold off to investors as CDOs, CDOs Squared and a variety of similar products.
The CDO buyers weren’t able to gauge the quality of the mortgages inside the CDOs without examining them individually and they weren’t going to spend the time to do it. You’re talking about thousands of mortgages in each CDO. In many cases they only “owned” slices of a particular mortgage anyway, making the whole deal more complex.
CDO investors were accustomed to the ‘conforming loan’ paper required by Fannie and Freddie. Conforming loan paper has a very good record of safety. But the new mortgage paper that the investors were buying from private bundlers wasn’t conforming paper. It was exotic and high risk with no proven track record.
CDO buyers trusted the ratings given this paper by the big rating agencies. Moodys, S&P, that sort. The problem is the rating agencies were rating the paper as AAA when it turned out to be junk.
One big reason that the entire investment industry blundered into this extremely risky world is because they were all relying on a risk formula that few of them appeared to understand. This is David X Li’s gaussian copula function which he had borrowed from the life insurance industry. The financiers grabbed on to this formula because it appeared to give them a way to calculate the risk of these hard to value products. But unless they understood the math behind it they didn’t appreciate its limitations.
Another more sinister reason these risky mortgages were written involves credit default swaps (CDS), a sort of insurance product for bonds. It was possible to buy CDSs that would pay off if a CDO bundle of mortgages defaulted. And you could buy as many CDSs against a particular CDO as you wanted. In essence it was like buying a bunch of life insurance policies on your neighbor’s life. You might make a bundle if you had inside knowledge that your neighbor was very sick.
By the end of the bubble the demand for high risk mortgage paper was driving the writing of the mortgages, a tail wags the dog situation. CDO makers wanted all the high risk paper that they could get so that they could write derivatives against the CDOs. They appeared to want the riskiest paper possible. If someone knew that the mortgages inside these CDOs were doomed to fail then that someone could buy a load of credit default swaps against the mortgage paper and be guaranteed a huge windfall when the mortgages blew up. There are some hedge funds that made billions of dollars in just this fashion when the bubble popped. No one has been prosecuted so apparently no laws were broken.
All symptoms... Who assumed the risk? Not the banks? Why? Freddie and Fannie purchased the Mortgages. They may have used those other organizations to do their dirty work...
Government intervention into the free market (getting into the mortgage business just like healthcare) could cause such catastrophic results. The free market quickly exposes bad business practices. Governments create perverse incentives that last a long time...
A person who is allowed to place bets which he might not be able to cover will, as a result of such ability, be able to achieve a higher payoff expectation than someone without that ability. In some cases, a such person may be able to combine multiple negative-expectation bets in such a way as to have a net positive outcome for the person placing them (and a huge negative outcome for someone else who will be left holding the bag).
I don't think companies that issued credit default swaps far in excess of their net worth failed to understand the dangers of correlated risks. More likely, they understood that the optimal way to monetize their ability to place bets they couldn't cover is to concentrate all the risk into one negative outcome, so all other outcomes would be positive. If the most one would stand to actually lose would be $10M, having a 1/1000 chance of being "obligated" (but unable) to pay $1B dollars is less "risky" than a 1/100 chance of having to pay $10M even though the former "expectation" is $1M and the latter expectation is only $100K. Accepting the latter risk in exchange for $80K cash would be a losing proposition, but "accepting" the former for $50K would be a winning proposition (since it really represents a 1/1000 chance of losing not $1B, but only $10M).
If one couldn't avoid one's obligations to pay, accepting $50K for a promise to make a 1B payment for a 1/1000 event would be a grossly losing proposition. On the other hand, if one isn't going to be paying anyhow, any premium one gets beyond what would be required to justify one's actual risk exposure is pure gravy.