Bank bonuses: three ways to reform the system
Saturday, 15 March 2008
Daniel Heller
IN the search for lessons from the credit market crisis, one issue that deserves more debate is the system of bonuses in investment banking. It also attracted attention this week when Jérôme Kerviel, the 31-year-old at the centre of the Société Générale scandal, told prosecutors he wanted to seem like an exceptional trader and get a higher bonus.
The textbook version of a bonus system is fairly simple. The compensation of an employee consists of two parts: a stable base salary and a variable incentive. The size of the bonus varies from year to year and is supposed to reflect the effort and success of an employee. To a large extent, the bonus is typically paid out in cash. A smaller fraction is distributed as equity and options. Bonuses have gone up considerably during past years when profits have been high. Frequently, the sum of the bonus contributions in investment banking has exceeded the dividends paid to the shareholders.
Unfortunately, the reality is more complicated. Risk and uncertainty introduce a crucial asymmetry. At the upper end, the size of the bonus is not limited. Higher net profits generated by the employee translate into higher bonuses. At the lower end, however, the bonus is limited to zero. In other words, any losses are borne entirely by the bank and the shareholders and not by the employee. This asymmetry clearly provides an incentive for employees to take risks without being fully accountable in monetary terms.
The amount of bonus payments is less variable than one would expect. The huge losses in the subprimerelated segments of the financial market were typically caused by a small portion of the workforce. Of course, the employees responsible for these losses did not or will not receive any bonuses (or have even been fired). But competitive pressures in the labour market seem to have made it hard for the banks to cut the bonuses of the majority of their employees, who worked in profitable business units. In addition, some investment bankers were hired with "guaranteed bonuses", which have to be paid even if the employee has generated a loss.
These drawbacks are large enough to be a potential threat to financial stability. Not only do they provide incentives for substantial risk-taking; banks also appear unable to reduce costs quickly by cutting the overall bonuses. This reduces their ability to absorb losses and writedowns by prompt corrective cost reductions.
There is no easy way to improve the bonus system. No bank wants to be the first to introduce a change, given the competition it faces from the others. One promising way would be for the industry to agree guidelines. The timing for such an initiative could not be better, since most internationally active banks are struggling with the same issues. Such self-imposed guidelines could involve these elements:
First, bonus schemes need to be more long-term oriented. Instead of being primarily based on last year's performance, the bonus should take the performance of several years into account. One way is to pay out only part of the bonus in profitable years. The nondistributed part would be set aside as "reserves" that could be used to cover any losses in the future. Employees who are successful for several years will be able to reap the benefits, as under the current schemes, since the accrued bonuses would be paid out eventually. The accrued past bonuses of those employees who lost money would be used to cover the losses, at least partially. A similar scheme has been applied for a long time in the hedge fund industry, where employees have to reinvest a substantial fraction of their bonus into the fund.
Second, the level of the average bonus in investment banking should be reduced in order to take into account that potential losses are not borne by the employee. One way is to limit the maximum bonus in a bank to the bonus of the chief executive. The fact that in the past some rewards exceeded that of the chief executive illustrates that schemes were not designed in an economically efficient way. In a profit-maximising, privately owned company the chief executive is supposed to be the employee with the most skills, who adds the most value.
Third, guaranteed bonuses should be abolished. They are a contradiction in terms, since a bonus is by definition the variable component of the compensation. Guaranteeing bonuses releases the employee from the requirement to make at least a small contribution to cover the losses he may generate.
(The writer is director and head of financial systems at Swiss National Bank)
IN the search for lessons from the credit market crisis, one issue that deserves more debate is the system of bonuses in investment banking. It also attracted attention this week when Jérôme Kerviel, the 31-year-old at the centre of the Société Générale scandal, told prosecutors he wanted to seem like an exceptional trader and get a higher bonus.
The textbook version of a bonus system is fairly simple. The compensation of an employee consists of two parts: a stable base salary and a variable incentive. The size of the bonus varies from year to year and is supposed to reflect the effort and success of an employee. To a large extent, the bonus is typically paid out in cash. A smaller fraction is distributed as equity and options. Bonuses have gone up considerably during past years when profits have been high. Frequently, the sum of the bonus contributions in investment banking has exceeded the dividends paid to the shareholders.
Unfortunately, the reality is more complicated. Risk and uncertainty introduce a crucial asymmetry. At the upper end, the size of the bonus is not limited. Higher net profits generated by the employee translate into higher bonuses. At the lower end, however, the bonus is limited to zero. In other words, any losses are borne entirely by the bank and the shareholders and not by the employee. This asymmetry clearly provides an incentive for employees to take risks without being fully accountable in monetary terms.
The amount of bonus payments is less variable than one would expect. The huge losses in the subprimerelated segments of the financial market were typically caused by a small portion of the workforce. Of course, the employees responsible for these losses did not or will not receive any bonuses (or have even been fired). But competitive pressures in the labour market seem to have made it hard for the banks to cut the bonuses of the majority of their employees, who worked in profitable business units. In addition, some investment bankers were hired with "guaranteed bonuses", which have to be paid even if the employee has generated a loss.
These drawbacks are large enough to be a potential threat to financial stability. Not only do they provide incentives for substantial risk-taking; banks also appear unable to reduce costs quickly by cutting the overall bonuses. This reduces their ability to absorb losses and writedowns by prompt corrective cost reductions.
There is no easy way to improve the bonus system. No bank wants to be the first to introduce a change, given the competition it faces from the others. One promising way would be for the industry to agree guidelines. The timing for such an initiative could not be better, since most internationally active banks are struggling with the same issues. Such self-imposed guidelines could involve these elements:
First, bonus schemes need to be more long-term oriented. Instead of being primarily based on last year's performance, the bonus should take the performance of several years into account. One way is to pay out only part of the bonus in profitable years. The nondistributed part would be set aside as "reserves" that could be used to cover any losses in the future. Employees who are successful for several years will be able to reap the benefits, as under the current schemes, since the accrued bonuses would be paid out eventually. The accrued past bonuses of those employees who lost money would be used to cover the losses, at least partially. A similar scheme has been applied for a long time in the hedge fund industry, where employees have to reinvest a substantial fraction of their bonus into the fund.
Second, the level of the average bonus in investment banking should be reduced in order to take into account that potential losses are not borne by the employee. One way is to limit the maximum bonus in a bank to the bonus of the chief executive. The fact that in the past some rewards exceeded that of the chief executive illustrates that schemes were not designed in an economically efficient way. In a profit-maximising, privately owned company the chief executive is supposed to be the employee with the most skills, who adds the most value.
Third, guaranteed bonuses should be abolished. They are a contradiction in terms, since a bonus is by definition the variable component of the compensation. Guaranteeing bonuses releases the employee from the requirement to make at least a small contribution to cover the losses he may generate.
(The writer is director and head of financial systems at Swiss National Bank)