The Wall Street Journal [subscription required] compares the estimated size of the subprime mortgage crisis with other recent financial crises. About to be taken over by Rupert Murdoch, it looks determined to underestimate subprime's magnitude in order to make those responsible look less incompetent. I've estimated the subprime crisis could cost $4 trillion -- the Journal limits it to $400 billion. But the analysis can't paper over an important point -- the subprime crisis is the first to be fueled by securitization.
There are two things I like about this analysis. First, it highlights the predictable decade long cycles in the U.S. economy based on the relationship between asset prices and borrowing. When asset prices rise -- possibly based on their makers' ability to charge a higher price for them than it costs them to make -- the bankers start to lend money using those assets as collateral. But the lending drives up the asset prices beyond what they're really worth as more and more bankers pile in. Eventually, the borrowers can't pay back the loans and the banks implode along with the borrowers.
The second thing I like about the analysis is the comparison of the magnitude and costs of five crises over the last 25 years:
- 1982 Lesser Developed Country (LDC) lending crisis peaked at 1.7% of GDP and cost $55 billion in losses
- 1986 -1995 Savings & Loan (S&L) crisis peaked at 3.2% of GDP and cost $189 billion in losses
- 1992 - 2003 Japanese Bank Loan crisis peaked at 7% of Japanese GDP and cost $263 billion in losses
- 2000 - 2003 Tech-bust bond defaults peaked at 0.9% of GDP and cost $93 billion in losses
- 2007- Subprime crisis ranged between 1% and 3% of GDP which the Journal estimates will cost between $150 billion and $400 billion
As I've pointed out earlier, the S&L crisis is the template for Bush's subprime mortgage intervention. That's because the first President Bush was the architect of the S&L bailout which involved government funds. And the current president is determined to avoid following in his father's footsteps by not spending any government money to solve the subprime problem.
I also believe that the Journal's estimate of the magnitude of the subprime crisis should include the loss of value of the housing stock that is a direct result of the credit crunch precipitated by lending to people who can't pay back the money. To its credit, the Journal estimates that loss at between $1 trillion and $6 trillion. Furthermore, I suspect that the problems will extend to higher quality mortgages as well.
Despite the comparison to previous crises, the current one is the first one that's magnified by securitization. Investors bought bundles of mortgages which were so complex that nobody could put a price on them. Investors relied on ratings agencies to AAA-rate them. But the investment banks paid the ratings agencies that gave the highest rating.
Nobody seems to know what these mortgage bundles are worth, who owns them, how much has been borrowed using the bundles as collateral, how deeply their value has declined, and how much bank capital the resulting write-downs will cost.
And this is just the beginning of securitization's problems -- there are hundreds of billions of securities backed by credit card receivables, auto loans, and leveraged buyout loans whose problems have yet to surface. Moreover, as I posted here, securitization's fallout could damage other investors such as life insurance companies.
When this is all over, the government will need to step in and decide whether securitization should continue. If so, there'll need to be a huge leap in transparency. And that leap may be too much to achieve.