How on earth should risk bonuses be calculated?

eFC logo

Such is the conundrum faced by some of the biggest minds in the world of compensation theory: how do you work out a formula for paying risk managers in investment banks?

"It's one of the major priorities for banks at the moment," says Jon Terry, head of the human resources reward practice at PricewaterhouseCoopers. "Paying risk managers from the same bonus pools as the people they have oversight over is no longer acceptable."

The difficulty in devising a reward system for risk managers is that they need to be paid for preventing own goals instead of scoring goals of their own. And this can be difficult to quantify.

In the quest for a solution, Terry says everything is on the table. This includes setting up separate bonus pools for risk professionals, giving risk professionals preferential access to traders' bonus pools, and paying them very big salaries and very, very small bonuses.

Given that risk bonuses have never been exactly huge, some might argue that the latter has always been the case.

Long time horizons required

Assuming, however, that risk managers do require some kind of incentivisation, the prevailing wisdom seems to be that risk managers should be paid for keeping banks out of trouble, retrospectively defined.

"What I am going to want to see is that during periods of difficulty, good risk management ensures that my bank stays in the middle of the pack and is not an outlier in terms of losses," says Christopher Whalen of Institutional Risk Analytics. "However, this assumes that risk managers are actually able to affect performance," he adds.

Linking bonuses in risk management to performance in a down cycle would require very long time horizons. Jan Simon, a former Goldman Sachs trader and professor of finance at Spanish business school IESE, says this may not matter because risk professionals tend not to change jobs very much anyway.

But Terry favours paying risk peeps based on 'objective' performance targets. "You can't put metrics around how many time you put spot a problem and solve it, but you can look at things like team work, and time spent turning around issues," he says. "You can come up with a series of indicators and assess against them."

How do you think risk bonuses should be calculated? Help solve the conundrum with (constructive) suggestions below.