AIG Enterprise Risk Management: Tougher Than It Sounds

             Photo credit: Philip Montgomery for the Wall Street Journal, Online Edition


AIG has a new "risk czar," Peter Hancock, an economist by training I'm happy to see, who heads up the Chartis unit charged with doing "enterprise risk management."  I have yet so to meet a public company director who can describe in concrete terms what this means for their company, and especially so for financial companies.

One of the operational problems generally, and especially for financial companies, is that risk is generated in silos.  An example of a large silo would be proprietary trading, and a sub-silo might be what's called Delta-1, where the latest rogue trader has surfaced at UBS.  Every profit center will have its own system of reporting, especially for the purposes of calculating bonuses and incentive compensation.  Lots of magic takes place when financials are rolled up for the purposes of external reporting.

Since risk is being generated in large numbers of relevant silos, how can a person at one desk look at a chart, report, or dashboard and monitor risk on an enterprise-wide basis?  I don't believe its possible for an investment bank, which is why we've had rogue traders since way in the early days of MBS trading in 1987.  Howard Rubin generated $377 million in losses which were not discovered until he left Merrill Lynch, and somebody found the unreported trade confirms in his locked desk drawer; he was banned from the industry for nine months and eventually joined Bear Stearns.  Joe Jett was another prominent "profit center" for Kidder Peabody who created some fictional profits of $350 million and paid himself a $9 million bonus before being discovered.  A Japanese trader for Royal Dutch Shell lost $1 billion in unauthorized commodity trading.  Nick Leeson was a recent example who lost $1 billion for Barings, and apparently Kweku Adoboli of UBS has hired Lesson's attorney to defend him in his court action. Sounds like a prudent move for the trader.

The point is that rogue traders are nothing new, and the ones in the news are really a subset of those who are out there and undiscovered, or who had big exposures that reversed themselves before they were found out.
Bonuses, like those that Joe Jett generated for himself, are done in the profit centers themselves and on a basis that is too fast for any effective risk management, except well after the fact.

Managers in the profit silos will complain to risk czars that rigorous systems of oversight and reporting will negatively impact their ability to recruit superstar trading talent.  Guess which side is going to win this argument?  Not somebody like the man in the picture above.

Now AIG is supposedly a simpler business than before, with the demise of their specialty financial businesses.  If this is true, and it has gone back to its traditional  insurance and reinsurance businesses, then it might be possible that a risk management system could be devised and operative.  I can't conceive of how this can be done for a traditional global investment bank.

Dodd-Frank does not give any degree of comfort. Banks will have to divest their prop trading desks, but the problem is that nobody can agree on what constitutes a prop trading desk.  Witness the confused discussion on whether or not Adoboli's Delta 1 desk constituted a prop trading operation, or one that worked on behalf of UBS clients!  The more things change, the more they remain the same.