Ever since the subprime mortgage collapse began, I've been thinking a lot about risk. My thoughts crystallized in January of this year, when "rogue trader" Jerome Kerviel's quasi-legal arbitrage moves left France's Societe Generale bank exposed to billions in losses. In the aftermath of SG's selloff, the issue became how Mr. Kerviel managed to circumvent SG's "risk controls," especially since he turned out not to be an evil genius; he seems to be simply a bright guy who felt underappreciated (read: underpaid), saw the flaws in the bank's controls, and proceeded to make billions for SG, before realizing that he would have to hide it all.
I found myself fascinated by that idea of risk control, and it set me to wondering how we define risk, how we presume to control it, and how we do such a good job of ignoring our past failures at doing either of those things.
To that end, my summertime reading has included Peter Bernstein's Against the Gods: the Remarkable Story of Risk, and Roger Lowenstein's book on the collapse of Long-Term Capital Management (LTCM), the perfectly named When Genius Failed. (For beach reading, I went with Jimmy Breslin's new book, The Good Rat, and some Henry James and H.P. Lovecraft, if you're curious; can't recommend that Breslin book highly enough.)
Mr. Bernstein's account, although dry, gave me some good background on how man came to understand risk, which he describes as "a humanist project, an attempt to abolish the idea of unknowable fate and replace it with the rational, quantifiable study of chance."
With Mr. Lowenstein's book, we get a blow-by-blow account of how this rational, quantifiable study can get people into real trouble. Established in 1994, LTCM was a hedge fund founded by an exec from Salomon Brothers and stocked with renowned MIT professors, Nobelists and others from outside the traditional trading floor. The LTCM team had developed innovative models and derivative bets that would (largely) insulate them from risk, but they relied on massive leverage to generate returns.
That is, their "risk-free" deals would only generate, say, five cents per $100 transaction, but if they made $1 billion in those transactions - mostly with other people's money and loaned shares and bonds - they'd be rolling in dough. They referred to this as "vacuuming up nickels," but one outside money manager contended it was more like "picking up nickels in front of a steamroller." There were returns to be made, but there was also a slight risk of getting run over.
I know you'll find it hard to believe, but after a few years of spectacular results from these humongously leveraged deals, LTCM expanded beyond its reach, extended bets into areas in which it had less expertise and found itself steamrolled when Russia defaulted on its government bonds in the late summer of 1998. LTCM's collapse threatened a number of major banks, which were tied to it through convoluted trades and swaps, prompting a multi-billion-dollar bailout. It seems that LTCM's risk controls covered all the contingencies, except for the ones that actually happened and "Long Term" turned out to mean 1994-1998.
At one point in When Genius Failed, the author calls out LTCM director and Nobel-winner Robert J. Merton's faulty theory of "continuous time" trading - the idea that share prices do not change radically, but move smoothly - by quoting Christian author G.K. Chesterton, who wrote: "[Life] looks just a little more mathematical and regular than it is; its exactitude is obvious, but its inexactitude is hidden; its wildness lies in wait."
The one exactitude I can see? Everyone underestimates the downsides.
* * *
This issue, we have an article on Global Risk Management; I'm happy to report that it covers drug-safety issues in various regions, and not the use of derivatives for pharma-financing.
P.S.: Does anyone realize that it was only 10 years ago that Russia couldn't pay its bonds? If there hadn't been a 1000% markup in oil futures, where would they have ended up? And if oil prices slide back down, where will they end up?