Risk Modelling, Regulation and the Return of Dr Doom
IT MUST be hard for people today to have any appreciation of the power that Henry Kaufman wieldedwhen in the 1970s and 1980s he was chief economist at Salomon Brothers now part of Citigroup but then fiercely independent.
Markets hung on his every word. If he was bearish, Wall Street fell; if he was optimistic which was not very often shares rose.
He long since left the public stage but still writes occasionally. In the current issue of Financial World, the magazine supported by the Centre for Financial Innovation, he lets fly with both barrels. His subject is today's financial markets and his theme is that our ability to innovate has outstripped our ability to exercise restraint and control.
He is particularly scathing about risk modelling, the computer programmes that lie at the heart of most investment banks' trading strategies.
These models have worked reasonably well in the benign markets of recent years, but Kaufman doubts they will cope well in a more hostile environment.
This is because they assume that everything which matters can be measured, whereas he quotes Einstein as saying "not everything which counts can be counted and not everything that can be counted counts".
There is a second danger born of the competitive nature of today's markets.
Risk modelling will spread, Kaufman says, and as it spreads it will become riskier. This is because the investment bank that has the most relaxed attitude to risk, and the least restrictive inputs, will by definition have the most aggressive risk-control model.
This means it will get the lion's share of the business because its riskcontrol mechanism will allow it to do stuff that other firms turn away as being too dangerous. Seeing their bonuses disappear, the rival firms will then doctor their risk models and loosen their standards so they can be competitive again.
The firms are in effect in a race to the bottom.
At the same time, qualitative judgments that draw on skill, experience and reason to worry about broader economic, political and social issues that affect market behaviour will be pushed aside, because they are neither clear-cut in timing nor have that aura of certainty which comes from the computer print-out.
Unfortunately, although he highlights the problem, Kaufman cannot see an answer. Leaving the market to sort things out will not work because most of the worst sinners are now deemed too big to fail. They will be bailed out by the central bank and they know it, so they see no reason to constrain their behaviour.
The second option would be greater supervision by the authorities, but that won't happen either. The securities firms don't want it, their political donations and clout on Capitol Hill
US legal perils in takeovers
IT'S a sign of how the business world has changed in recent years that Clifford Chance, one of the City's biggest law firms, ran a seminar today on what it calls "creeping criminalisation". …