Risk-Adjusting The 'Bonus Culture'

As current events in the financial markets unfold, there has been some criticism of the levels of salary and bonuses paid to senior bankers and other financiers. It is obviously difficult to justify paying senior executives substantial amounts of money when there is, to put it mildly, a question mark over the way that their companies have been run.

One obvious answer is to reward executives, not just according to the profits that they make, but according to their 'risk-adjusted' performance. Some of these arguments reminded me of a phenomenon known as the Trader's Option within the banks in the City, and which is described later.

By Risk-Adjusting an executive's remuneration, their bonus would not just depend on achieving a profitability target, but also on the risk profile that they generated in the process. High quality, low risk returns could be rewarded more generously than potentially lower quality risky returns, even if the latter might produce a higher short-term 'bottom line' profit.

Obstacles That Are Encountered In Practice

Although this would appear to be a sensible idea, and many financial institutions claim that they try to do this, there are a couple of practical snags that need to be overcome before one could have real faith in such an approach working in practice.

The first problem is that, in 'the real world' immediate profitability almost always attracts more attention than a - perhaps rather theoretical - risk analysis of what might happen at some point in the future.

The second problem is that many financiers, especially in investment banks, are natural risk takers. This is particularly true of those who have come from a financial trading background. This is not in itself a bad thing, controlled risk taking is a necessary part of the market.

Unfortunately, if we do try to 'risk adjust' performance in such a way, we hit several more problems. The most obvious being that it is not necessarily in the immediate interests of those being assessed to have their performance risk-adjusted, and they might well be expected to resist it.

Estimating 'Risk' In a Sensible Way

Another, more technical problem is that there needs to be an agreed way of measuring 'risk', and offsetting it against profits. This is not as easy as it sounds, because, as we have argued elsewhere, there is not a simple accounting definition of how to measure 'risk'. There are many different ways of quantifying and calculating it, and the better ways of doing it (in our opinion) all require a degree of mathematical modelling.

Here lies another more subtle problem which can be a real bone of contention - models need to be 'calibrated'. Where a model requires inputs which are estimated, then these inputs and other features of a risk analysis model can be 'tweaked' to tune its behaviour. This type of calibration is both an art and a science, and auditing such a process is not necessarily at all straightforward. Unfortunately, in the wrong hands, 'calibration' can sometimes become a euphemism for adjusting a model's inputs until it produces the desired result.

Risk-Reward Asymmetry - The 'Traders Option'

However, if we do not manage to find a way of rewarding executives and other staff in a risk-adjusted way, then we run the risk of creating a significantly asymmetric risk-reward profile for incentivisation, and this leads us back to the aforementioned Trader's Option.

A 'Call Option' is a type of financial derivative. A key feature of such an option is that in return for paying a premium, you effectively receive a 'one-way bet'.

There is an inherent asymmetry built into the option - if the price goes up you win, but if the price goes down you do not lose. The pricing and risk management of such contracts is based on a number of usually reasonable assumptions - including that the investor owning the contract is 'rational' and will act accordingly - if they are not, then this usually just means that they will lose out.

When a financial trader makes buy or sell decisions, and takes on risk in the process, the nature of the job is that they are aiming to make money, and will be richly rewarded if they do. If we look at this in broad risk management terms, then there is potentially a significant problem - again arising from an asymmetry in the risk / reward profile of this arrangement.

Suppose that a trader is considering whether to take on a huge position in a risky market, and for the sake of argument, let's suppose there is really a 50% chance of winning and a 50% chance of losing.

A rational trader might well come to the conclusion that such a bet was his optimum strategy, because if he or she wins, then they will make a lot of money for their employer, and be very well paid themselves; whereas if they lose, they will probably just lose their job. It can be well argued that if there were no other considerations, then because of this asymmetry, such a bet would be a rational - although unethical, immoral and possibly illegal, thing to do.

It is also worth pointing out what is hopefully the obvious: that when putting in place mechanisms to control and manage risk, there needs to be an understanding of the risk / reward profile at the key points in an organisation, and relying on 'good faith' is not enough. Most people will of course have professional integrity, loyalty to their employers and be fundamentally honest - but the management of risk is at least partly about dealing with the unusual, unexpected and unwelcome, rather than the run-of-the-mill.

Conclusion

A key lesson from all of this is that regardless of any other measures that might be in place, any business model which includes a large asymmetry between risk and reward for key participants is not one that a responsible risk manager should be especially comfortable with - at whatever level of scale we might care to consider.

It should also be remembered that whereas, with a few well-known exceptions, 'rogue' traders might 'bet the bank' figuratively; 'rogue' senior executives are quite able to do it literally.