A French court has sentenced Jerome Kerviel, the SocGen trader who caused the company to lose 5 billion Euros, to five years in prison and a fine of 4.9 billion Euros.
I think we can safely assume that Mr. Kerviel is now bankrupt for life. Reading about the case did raise a question in my mind. How can one person cause this much damage and how did the damage remain undetected until too late? I know that people have pointed to the asymmetric reward structure (where huge bonuses are paid if you make large profits and you really don't share in the losses) at banks as a culprit, but I think there are three "behavioral" factors that contribute to disasters such as this one.
1. Selection bias: Experimental economists have been exploring differences in risk aversion across sub-groups of people and their conclusions are fairly strong. For the most part, their studies find that younger people tend to be less risk averse than older people, single people are less risk averse than married folks and men are less risk averse than women (especially young). Now think about your typical trading room in any investment bank. It is over populated with 25-35 year old males, selected primarily because they had the right "trading" instincts. In fact, I think you can safely assume that if you were picking the least risk averse group in a population, it would look a lot like that trading room.
2. House money effect: You tend to be much less careful when taking risks with "house" money than with your own. Traders almost always are playing with house money, especially when they are doing proprietary trading, and not surprisingly are more cavalier about taking risk than they would be with their own money.
3. Break even effect: Here is a phenomenon that every casino owner knows. When a gambler loses money, he tries to make it back, and the deeper he gets in the hole, the rasher he becomes in his risk taking. So, a trader who loses $ 100 million will try to win it back with big bets (even if those bets don't make much economic sense); the more he loses, the wilder his risk taking will become.
One of the problems with the risk management systems that we have is that they deal with the symptoms and not the causes of erratic and bad risk taking. If we want to reduce the likelihood of more Jerome Kerviels in the future, here are some things we should do:
a. Create more diverse trading rooms: I am not a fan of diversity for the sake of diversity, but I think that opening up trading rooms to a wider range of people will dampen some of the excess risk taking. Maybe we should hire every trader's mother or grandmother to trade side by side with him; in fact, I would give her the next desk. Seriously, though, this will require investment banks to revamp their hiring processes and look more kindly on those "not cool" kids on campus who right now would not make the cut.
b. Restrict or eliminate proprietary trading: I know that proprietary trading is viewed as too lucrative to let go by banks, but I have my doubts as to whether it actually generates long term profits for any of its architects. One side cost of the increase in proprietary trading at banks has been the increase in house money that traders have to play with.
c. Information systems: To stop the "break even" effect from kicking in, we have to intervene much earlier when traders start losing money to prevent them from accelerating the cycle. To intervene, we need to know how much traders are making or losing in real time and have automated or computerized systems step in and stop trading at a defined loss point. Traders will try to devise ways around the system, but the system has to be responsive enough to adapt.