Was David Li the guy who 'blew up Wall Street?'

Mike Hornbrook on the Canadian connection in the great Wall Street meltdown.

Described as both "modest" and "outgoing" by his former professors at the University of Waterloo, David Li is a Canadian math whiz who, some now say, developed the risk formula that destroyed Wall Street.

When the full history of the Great Meltdown is written — presumably sometime after all the financial pain has eased — Li may occupy a special place.

David Li in an undated photo.

The formula he devised was widely used and helped some of America's biggest financial institutions determine how to market the controversial instruments known as credit derivatives, including what risk they faced, what strategies they needed to minimize the risk and what return they should demand.

These days Li, 45, is in China, the country where he was born, keeping a low profile. He heads the risk-management department of China International Capital Corporation in Beijing.

The company won't make him available for interviews. But according to friends, Li is said to be "sheepish" about all the trouble he has caused. 

Rise of the quants

David Xiang Li obtained his Canadian citizenship in the 1990s and after completing his education here went to work on Wall Street as a quant.

Mike Hornbrook has been the economics correspondent for CBC Radio News since 2004, covering economic and financial trends as well as significant business stories. Before that, Mike spent a decade abroad as a foreign correspondent for CBC radio and television.

Quants are "quantitative analysts" — math wizards who design computer models to assess risk, price the latest financial instruments and predict market movements.

Quants grew to prominence on Wall Street in the bull markets of the late-1980s and '90s with the rise of ever more complex financial innovations.

In 1997, Li landed a job on the New York trading floor of CIBC World Markets, which was then a pioneer in the emerging credit derivatives market.

In those days, investment banks were seeking ways to pool corporate bonds and then re-sell these in chunks, as they were doing with mortgages. The bond pools were known as collateralized debt obligations or CDOs.

CDOs were supposed to make investing less risky because the exposure was diffused.

If you bet the farm on the bonds of a single company, for example, you could lose it all if that firm failed. Invest in the bonds of hundreds of companies, on the other hand — all packaged together and sliced into portions reflecting differing risk — and it was thought unlikely that you would suffer a big loss if a few go bust.

At least that was the theory.

A woman strolls through Toronto's financial district in September 2008 when the markets began their big slide. (Canadian Press) ((Nathan Denette/Canadian Press))

Love and death

Twelve years ago, Li began thinking about CDOs as insurance actuaries do, in terms of love and death. Actuaries know that, statistically, many people die soon after their spouses.

So Li applied this insight to a computer-driven financial model that predicted the probability of a given set of corporations defaulting on their bonds in quick succession. He published his ideas in the peer-reviewed Journal of Fixed Income in March 2000.

Dr. Harry Panjer, a professor emeritus at the University of Waterloo and an actuary who supervised Li's PhD work, says his former pupil's great strength lay in synthesizing diverse concepts.

"He was very good at capturing ideas from different areas and bringing them together. This paper was a manifestation of that ability."

In the statistical community, Li's financial model became known as the Gaussian copula function applied to default correlations and was named after the 19th century German statistician Carl Friedrich Gauss.

Copulas are a concept in quantitative analysis that helps to predict the likelihood of different events occurring when those events depend upon and are affected by one another. They had not been applied to credit derivatives before and Li's formula was quickly adopted on Wall Street.

They seemed to work fine as long as the housing and bond markets continued to rise. But once they faltered, it was a different matter.


Writing in February in Wired magazine, Felix Salmon suggested that Li must share blame for the financial collapse that began in August 2007. 

"For five years, Li's formula looked like an unambiguously positive breakthrough, a piece of financial technology that allowed hugely complex risks to be modeled with more ease and accuracy than ever before.

"With his brilliant spark of mathematical legerdemain, Li made it possible for traders to sell vast quantities of new securities, expanding financial markets to unimaginable levels."

One example of this rise would be credit default swaps, or bets on how well bonds would perform. These swaps went from $920 billion in 2001 to $62 trillion in 2007.

Everyone hitched a ride on the gravy train: bond investors, regulators, Wall Street banks and rating agencies. Buyers and sellers of credit derivatives could agree on price because they worked off the same formula. The big ratings agencies — Moody's, Standard & Poor's and Barclay's — assigned AAA ratings the same way.

Math fame: German statistician Carl Friedrich Gauss on the 10 deutschmark bill.

Salmon says the formula became so deeply entrenched on Wall Street, where people were making so much money, that warnings about its limitations "were largely ignored."

Indeed, Li himself cautioned that his model relied on limited historic data so it could not capture all possibilities. The chilling truth of this became too apparent as defaults began to rise among borrowers with the so-called subprime mortgages and the house of cards that was built on leveraged debt began to collapse.

Is Li really to blame?

"Models are only as good as their assumptions," says Steve Brown, associate dean at the faculty of mathematic at Waterloo. "If you ignore the assumptions that go into the model, [it] can produce things it was never intended to produce."

Some experts warn that even with the trillions in bailouts the U.S. financial system is effectively bankrupt, so the butcher's bill for David Li's financial model is still being added up.

But is Li really to blame? Tagging him with responsibility for the meltdown seems like blaming Albert Einstein for the nuclear disaster at Chernobyl. Both developed a brilliant theory; others used it for their own purposes.

As the scale of the global financial meltdown unfolds, a journalist would like to report David Li's reaction to what happened and his part in it. But Li is not talking.

"He's a little bit sheepish about the whole thing and surprised by the response to it," says Waterloo's Panjer, who still keeps in touch.

Panjer signed Li's Canadian passport renewal papers last year and defends his former student. "He has been singled out."

But, says Panjer, "you don't need to be a scientist to realize that common sense didn't prevail in the U.S. People who were putting these financial products together were ignoring reality. It resulted in an abuse of these models."