Q: Was Societe Generale’s Meltdown Preventable?
A: Yes.
Hedge Funds, banks and other financial institutions seem to meltdown on an all too regular basis. The available evidence on Societe Generale's meltdown is fragmentary and, probably, wrong in part. That said, the evidence that is available now implies that Societe Generale could have avoided the meltdown. Here is what we think we know.
1. Calling them “rogue traders” does not help. “Rogue trader” gives the impression that if we could just avoid a few bad guys, everything would be OK. It gives the impression that if anyone is at fault, besides the rogue traders, it is those screw-ups in the Human Relations department. Unfortunately, some trading approaches require hustling and then a rogue trader is exactly what you want. In Marcia Stigum’s the Money Market one of her informants says, “Trading is an art form which I could not succeed in teaching my peers who had come through the system as I did. I would have done better to take on some kid hustling on the streets of Marrakesh.”
Just so we understand each other: I am not suggesting Mr. Kerviel does not deserve jail for his sins, if, indeed, he did what the papers report. I am not condoning lawbreaking. I am not suggesting that most traders are larcenous. To the contrary, some approaches, those based on science, for example, demand ruthless honesty. I am only suggesting that for some trading approaches, and based on fragmentary evidence, that I think this includes Mr. Kerviel, a little larceny in the soul, a little carefully controlled larceny, is part of the job description. In which case, from a purely reptilian point of view, you hire the trader, grit your teeth and radically upgrade your risk management systems. Or you don’t hire the trader. And you don’t blame HR. It’s your fault if things go bad.
2. Risk management procedures should be transparent, that is, everyone should know what they are. Several articles on Jerome Kerviel noted that Mr. Kerviel was familiar with Societe Generale’s risk management procedures. The obvious--and incorrect--conclusion is that that was a mistake. But it wasn’t. The more eyes you have looking over a set of procedures, the more minds you have thinking about the issues, the better your procedures are likely to be. This is why open source software is generally more secure than proprietary software is.
Tere are a number of exceptions, of course, but less than you might think. One of the few areas where a strong counterargument can be made is security: password protection means nothing if the passwords are not kept secret. This is true, but unimportant. As long as it is possible to write passwords on post-it notes, as long as it is possible to use the name of your children as your passwords, passwords will be of only limited value.
Secure systems do not have to involve the biometric hardware you see in the movies. One device about the size of a nail clipper generates a series of semi-random numbers that are synchronized with a series produced by the software in a particular computer. When you log on the computer, the software asks you to enter the number showing on the device, the number changes every twenty seconds or so, and if your number matches the number the computer has generated, you are given access. Each user has his own device, of course, and each device generates its own sequence of numbers. If the reports I have read are accurate, Societe Generale might well be several billion dollars richer if it had used this system.
3. Management needs to put its own money at risk. According to recent reports, several layers of Jerome Kerviel’s bosses have been fired. This is right and proper, but it does not go far enough. With a seven billion dollar loss, I would say that Societe Generale’s CEO Daniel Bouton should have his head handed to him, not necessarily in one piece.
Unfortunately, while that threat might work with you or me, it does not necessarily work with CEOs. The problem here is that it is often in a CEO’s best interest to put his firm at risk. Merrill Lynch fired Stanley O'Neal, but they still let him walk away with tens of millions of dollars. An investor’s only hope is a strong board of directors, a board of directors tough enough to insist on clawback or lock down provisions in the CEO’s contract.
4. Best practices work. In 1999, Nick Leeson was able to defraud Barings because he both traded and controlled accounting. In 2001, John Rusack was able to defraud Allied Irish Banks because he traded and controlled accounting and because the management of Allied Irish Banks couldn’t be bothered to apply best practices, they couldn’t be bothered to learn from Barings’ example.
Best practices documents are written by humans, unfortunately, and they are sometimes written for bad reasons; to provide cover for incompetent, corrupt or dishonest investment mangers, for example. And sometimes there are no best practices. But when there are best practices, it is foolish to ignore them.
|