When Banks Don't Manage Their Bankers' Risks

Dear John Thain has a thought-provoking piece up today about the way that risks and revenues are managed at investment banks. Think of banks as being divided in two: on the one side are the bankers, who generate fee income and need little if any capital; and on the side are the traders, who make money from risk and who do need capital.

Generally speaking, when bankers generate fee income, they also take on a certain amount of risk. But the minute they do that, the risk is shipped over to the trading side of things, whose entire job is to manage risk and make a profit doing so.

The problem is that the trading side isn't particularly good at managing risk it has been bequeathed by the banking side. Sometimes, as in the case of leveraged super-senior CDO tranches, the trading side was all but unaware that the bank had taken on vast quantities of risk in the first place. And in any case it's really hard to hedge that kind of risk. As DJT noted yesterday, Goldman Sachs came out of 2007 smelling like roses largely because it was lucky enough not to have much ordure in its stables in the first place:

Goldman wasn't naturally long sub-prime in the same way everyone else was. If you're not long, it's easier to be short. Do you think Merrill could have hedged their exposure by being short a mere few billion of ABX? No.

What this all means is that chief risk officers need to spend much more time away from the trading desk, hanging out with bankers, and worrying about all the tail risk which bankers are particularly bad at worrying about. After all, the more bankers worry about such things, the harder it is for them to make their fee income. Somebody needs to save them from themselves.