Just getting around to reading the Financial Times' nice piece on David Li's formula for correlating risk, the fool-proof method of mathematical certainty that in fact allowed the financial sector to underestimate risk and melt itself down.
I've written about this before, mostly because I find it fascinating that this little string of mathematical terms that cause so much damage to the world originated with the actuarial science of broken heart: people often die sooner when their spouse dies; the relationship between such seemingly unrelated events can be quantified; and Li borrowed this idea for his formula to correlate vast, much more complicated relationships, like entire markets, which are not so easily correlated. Broken hearts indeed!
The Financial Times piece does of a nice job of really explaining this -- better than the Wired piece from a few months back. First off, this one starts with Johnny Cash and June Carter. First went June and Johnny died of a broken heart four months later. Just like in a country song. And kudos to sneaking a lede like that into a financial news daily.
The hefty story also explains the context: how Wall Street had been primed for such a colossal mistake since the mid-80s when the keys were turned over to quantitative whizzes who claimed intellectual superiority and infallibility as they converting the arcane world of finance into an even more arcane world of pure abstraction. No one else could understand what they were doing enough to say: sounds fishy. And yet it was fishy. Case in point: Long Term Capital Management, the massive hedge fund run by "quant" whiz kids from MIT and elsewhere that almost punched a $2 trillion hole in the economy and had to be -- you guessed it -- bailed out.
Then, along comes well-meaning David Li, and his thoughts on "life, death, and love." His formula, and the feeling of quantitative safety it seemed to wrap around risk, single-handedly created a huge, complex, unregulated, and (it turned out) highly unstable market. That would be the derivatives full of sliced and diced mortgage backed securities and the credit default swaps that bet -- er "insured" -- them.
But the knock out punch, the article points out, was when Li's formula was incorporated into the ratings agencies' formulas for grading those fancy financial instruments. Talk about fox in the hen house. Hmmm -- let's use the same methodology as the thing we're evaluating so as to ensure we'll all be on the same page. That sounds like a super solid approach to rating!
And the rest is...well, you know. To be fair to David Li, it wasn't his fault. Back in 2005, long before the end was nigh, Li told Wall Street Journal that his model was running amok. People were using it improperly. They didn't understand it. The assumed it meant more than it did. Li's heart is the one that's broken now. He left finance and moved back to China.