‘The real effects of financial reporting on pay and incentives’ – a practitioner view

I am a partner at Mercer, in the executive reward team. My day job is to advise companies and boards on everything that is mentioned in John Core’s paper, in particular how to make incentives work for a company, and how to make incentives work for shareholders. Core (2020) mentions that we need good-quality earnings to help ensure incentives are effective. I wholeheartedly agree with that. I would add a few more characteristics of a good incentive plan.

In no particular order but certainly at the top, simple; an incentive plan needs to be simple in order to work effectively. What does simple mean? It normally means only one measure. But if you have only one measure, then it can be very sensitive to external uncontrollable factors. So, typically most companies would use more than one measure.

It should certainly reinforce key objectives, and those objectives should be within the role and responsibilities of the individual that is participating in the incentive plan. For example, you would not necessarily use what we term a capital measure for someone three levels below the decision-making where capex is made.

An incentive should not be unduly sensitive to external factors. For a CEO, where Core’s focus is on, you could argue that it should be somewhat subject to external factors because that is ultimately what shareholders are exposed to, but for people lower down the organisation you would try to dial out some of that noise from external factors, and I will come on to how companies do that shortly.

Incentive outcomes are not unduly volatile. This is very important. Incentives are most motivational and trusted when they have a lack of volatility in their pay-out. This very much depends on the stage of growth the company is in, so for a small, start-up company you might expect some volatile outcomes but for a large, mature company like Shell, ideally your incentive outcomes should not be unduly volatile. Most companies would set targets such that for example their long-term incentive plan would pay out 70 per cent of the time and 20–30 per cent of the time it might not pay out at all.

Lastly, the accounting charge should reflect the value delivered. This is something that Core focuses on in his study. Whilst this is important very much to boards and CEOs, it is less important to the participant. All the participant cares about is: Is it controllable? Is it motivational? But certainly boards will take into account whether they are actually delivering the same sort of value reflected in the P&L and, as Core mentions, 10, 15 years ago, when options were all the rage, that was not necessarily the case. It is still the case today with many companies, many listed companies, for example, using relative total shareholder return-type measures, and those measures can be very inefficient from the P&L perspective because you take the charge on day one of the expected outcome of those incentives, and you do not true up, and therefore it can lead to a great disparity between the eventual outcome which is delivered to the participants and that which you reflect through your P&L.

Therefore, we need to be realistic. There is no perfect incentive measure, and therefore we need to take into account that most incentive measures will be affected by noise and unpredictability. We need to be able to reduce the impact of that noise and unpredictability. There are many approaches companies can use to reduce sensitivity to noise and unpredictability. The most common is for companies to neutralise the impact of Foreign Exchange (FX), to neutralise the impact of commodity prices, especially in their short-term incentive plans – less so in the long-term incentive plans, as managers are expected to manage that noise, but certainly in the short term it is very common to use, say, budgeted FX for incentive outcomes.

Wider performance ranges. This is absolutely critical. A good incentive plan has a performance range which captures all realistic incentive outcomes, the theory being that you want to incentivise continuous out-performance of where you are today, and the only way you do that is to have a wide performance range where you are always going to be on that slope, that sliding scale slope. So whatever performance you bank today, you want to have an incentive to out-perform that particular performance of today. And if you have a very narrow range, you are either going to be massively above it or potentially massively below it, and there is very little incentive effect in that regard.

Performance periods. Again, this comes down to the noise in the system reduces as you extend the performance period. Most companies would use different measures depending on the performance period which is captured in the incentive. One-year annual bonus plans commonly use strategic non-financial objectives, maybe revenue; three-year plans typically use an earnings-type measure. Then you have Total Shareholder Return (TSR)-based plans. They might have a five-year performance period, because the longer you measure that performance period, the less that outcome is affected by the noise in the system; it is more correlated with the management performance.

Incorporating an incentive bank. If all else fails and you can’t deliver an incentive plan which is highly insensitive to this noise and unpredictability, then accept that and use an incentive bank whereby you have these volatile outcomes or incentives, but they are paid into this notional – what we call incentive bank-, which then has an account balance for an individual, which then gets released each year, say a third of the account balance gets released each year, so there is this frequency of pay-outs to the individual. It helps to reduce that feeling for the participants in the incentive plan that they have no control and it is a bit of a lottery.

I have illustrated here an example of a target-setting approach which we often use. For an effective incentive plan it is critical not just to get an effective and robust incentive measure; the performance range you use also needs to be robust. This is an approach we use to help our clients set realistic performance ranges which are stretching but achievable.

The approach uses a number of different reference points both internal and external. The most important reference, and the reference point most companies use when setting incentive ranges is the strategic plan or the budget. Many companies would set the bonus range around their budget, but that is very sensitive to the assumptions in that budget, clearly. We advise our clients to think more holistically around the perspectives on performance, so we would capture many different internal perspectives on performance and many different external perspectives on performance.

Other reference points used are all about what is expected in the sector of the particular client, how they expect to perform over the next one year or three years, and it helps to set targets which capture this wider perspective on performance, hence a wider performance range typically is delivered by this target-setting approach which – I will go back to my earlier comments – helps to ensure that the incentive itself is more trusted, because you are not going to be too much above the scale or too much below the scale; always on the scale.

I would argue that a good incentive plan needs a combination of formula and discretion. Discretion is a bad word for listed companies. Most institutional shareholders prefer remuneration committee (remco) discretion not to be used, or, when it is used, only to have a downward effect on incentives. But I think that is wrong. I think discretion is required for incentives, especially for the more senior folk, where we tend to use measures which capture all aspects of performance at the bottom of the P&L, maybe balance sheet measures, maybe external factors such as TSR. Some discretion is therefore required, but most remuneration committees, most boards, find the application of discretion quite uncomfortable because they feel that there is not enough information for them to use that discretion. Discretion is best applied when you have, as on the left-hand side here, a set of principles which the remco can assess for each and every adjustment which they might consider in assessing an incentive outcome. They use those principles to assess whether that adjustment is valid.

You can use a sanity check for an incentive outcome. Whilst your standard bonus may be based on just an earnings measure, a single measure, the sanity check could be a scorecard of other measures relevant to your organisation which the remco can use to assess whether the outcome on your regular bonus plan is appropriate. For example, you could use eight different measures capturing all aspects of performance, and it is not formally linked to the outcome of the incentive, but if that scorecard there is showing lots of greens, but your incentive plan based solely on earnings has not vested, then a remco might use that as an opportunity to look back at the targets which were originally set for the earnings measure for that year to assess whether that was a valid earnings range. Of course, vice versa, if it shows all reds there on the scorecard, on this sanity check scorecard, but your incentive plan has paid out fully, there is some disparity there. There is something wrong with the incentive – so it can help a remco determine whether the actual outcome, the formulaic outcome, is truly warranted.

Core also talks about restricted stock and the use of restricted stock incentives. This is very much an emerging field in the UK plc at the moment, and it is being driven by two main factors. The first one is the volatility of pay outcomes, and from a participant’s perspective many executives, even senior executives, are quite uncomfortable with the fact that their long-term incentives in particular which capture three-year performance can be very volatile, very uncontrollable. There is this ground-swell of opinion that the traditional earnings-based, our typical relative TSR-based plan, no longer works for participants, so that is a key driver for the movement from these performance-based, performance vesting long-term incentives to a time-based restricted stock plan.

Shareholders are also strongly supporting this – many institutional shareholders are – for the largest companies, for the Shells of this world, top end of the FTSE, mainly because the pay numbers being generated for chief executives in particular are becoming quite uncomfortable. Numbers in line with the many millions for the CEO that Core shows for Shell are becoming quite uncomfortable for many remuneration committees of listed firms now. So they are looking for an opportunity to avoid these headlines in the national newspapers around their CEO pay, and restricted stock is one way of doing that because it reduces the potential value being delivered ultimately through the long-term incentive plans because it is less risky. Therefore the headline number, the face value granted at the original award date, is much, much lower – typically half of that which you would grant under a PV long-term incentive plan.

Disclosure statement

No potential conflict of interest was reported by the author(s).