1 in100 doesn’t mean it won’t happen next year.
Wired has an interesting article about how mathematician David Li’s “risk” formula revolutionized financial markets in at the turn of the century by “calculating” risk on things formerly thought un-calculable.
If risk can be quantified it can be sold, and risk was sold and sold and sold in the following years largely using a financial instrument called “Credit Default Swaps”. By bundling multiple types of debt (mortgages, bonds, loans, mortgage backed securities) you could create “known risk” instruments called Collateralized Debt Obligations (CBO).
The notion was that Li’s formula had taken the “risk” out of calculating risk.
Ooops… as mathematicians and market contrarians were quick to note, that wasn’t what the formula did. All it did was tie risks together that weren’t formerly tied together. The Wired story uses the analogy of two students that sit by each other. A default of one of their parents house, isn’t likely to affect the other. But if one gets the flu, the other is likely to get it. And this sneaker company comparison:
"The relationship between two assets can never be captured by a single scalar quantity," Wilmott says. For instance, consider the share prices of two sneaker manufacturers: When the market for sneakers is growing, both companies do well and the correlation between them is high. But when one company gets a lot of celebrity endorsements and starts stealing market share from the other, the stock prices diverge and the correlation between them turns negative. And when the nation morphs into a land of flip-flop-wearing couch potatoes, both companies decline and the correlation becomes positive again. It’s impossible to sum up such a history in one correlation number, but CDOs were invariably sold on the premise that correlation was more of a constant than a variable.
Li’s formula was 99% right, but the problem is when the 1% hit, the risk was centralized, connected, and catastrophic.
Much of the financial community decided to move to a 100 year flood plain, unfortunately, we all got flooded with them.
March 5th, 2009 at 2:50 pm
One of the large problems with risk assessment is an advocate’s explaining it to a decision maker who doesn’t understand the underlying calculations, as Salmon describes. This whole debacle has exposed us to a new term – fat tail – where a tiny probability multiplies a huge consequence with feedback to and from other systems. We have come to understand and adapt to the errors in hundred-year flood calculations, but we are still groping with financial risk assessment and pricing where there are incentives to misstate the assumptions. We should also always be apprehensive when “everybody is doing it”, whatever “it” is.
March 5th, 2009 at 2:52 pm
Ultimately, though, the governments role in backing speculation was the killer.
March 5th, 2009 at 3:57 pm
My, you are certainly dedicated to the principle that government caused the problem. Since government was only an enabler, what share of the culpability do you assign to the financiers? If you leave a door open and burglars rob your house, what share of the blame do you assign to the burglars?
March 5th, 2009 at 4:41 pm
They would have failed, and just them. The problem we are facing is trillions larger than it should be because the government – the Fed, Fannie Mae, Congress, provided backing for unsound business.