- On August 24, 2017
“Steps” in your incentive plan payout curve increase compensation risk unnecessarily. Smooth your curves and eliminate this risk.
The most common incentive plan payout designs provide a significant benefit for achieving a defined minimum level of performance. Viewed graphically, this “step” in the performance-payout relationship creates unnecessary business risks, and – more importantly – payout steps damage the pay for performance relationship. Small changes to incentive plan payout design can eliminate these risks and improve pay for performance.
“You get what you pay for” is accepted as a truism by compensation professionals. However, our experience has shown that compensation plans, in particular incentive plans, can at times lead to unintended results – results for which compensation committees may not wish to pay.
Compensation risk has existed ever since variable pay was introduced. Just as managers understood they could encourage behavior on targeted activities to improve performance results, employees quickly figured out how to adjust their behaviors and activities to maximize their own pay.
Poorly-designed incentive programs can even maximize employee pay while performance results are sub-optimal. At best, poorly designed incentives can result in counter-productive or dysfunctional behavior. At worst, these incentive programs can encourage poor customer outcomes and fraud. The compensation world recently witnessed such an incident with the fraudulent customer account scandal at Wells Fargo last year.
This article discusses a slightly different compensation risk – one driven not by the measure of performance used (e.g. the number of new customer accounts opened, as in the Wells Fargo customer account scandal), but rather increased risks that are a result of incentive plan design and the shape of the management incentive plan payout curve. Specifically, whether the incentive payout curve includes any “steps”, and if so, the magnitude of such steps.
What are stepped incentive payout curves?
An incentive payout curve defines the relationship between various levels of performance (the independent variable on the x-axis) and the associated payout amount (the dependent variable on the y-axis) for different levels of performance. A traditional incentive payout curve delivers different payout amounts for performance at various points between a minimum (or “threshold”) level of performance and a maximum (“superior” or “stretch”) level of performance. Usually, a “target” level of performance is defined to pay out 100% of an employee’s target incentive opportunity.
Figure 1 below illustrates a typical incentive plan payout curve. Achievement of the threshold level of performance will immediately generate an incentive payout equal to 50% of target with the payout increasing linearly to 100% of incentive target for the achievement of target performance. The payout then increases linearly between target and superior performance goals, usually with a maximum payout capped at 150% or 200% of target. This recent Harvard Law School Forum article summarizes typical payout levels for threshold and maximum payouts in long-term incentives. Among the S&P 100, 38% of the measures used in long-term incentives paid 50% of target for achievement of the threshold goal, while another 34% had some other higher or lower payout step. Only 29% of measures included no step (i.e., payout beginning at zero).
As is clear from the illustration, if performance falls short – even by one dollar — of the threshold goal (or whatever unit of performance is measured: one customer, one percentage point, or one widget), then a zero payout is earned. As the incremental dollar, customer or widget is added to results, the calculated payout immediately jumps to 50% of target. Such a stepped payout relationship increases business risk, as described in greater detail below.
History and context: the benefit of a payout step
Traditional payout curves like the one illustrated in Figure 1 have existed for as long as incentive plans have been around. When the Sheffield Barry team started our compensation consulting careers in the 1990s, the initial step at threshold was heavily influenced by the fact that Earnings Per Share (EPS) was a key and very common measure (and sometimes the only measure) in many management incentive plans.
Missing Wall Street analysts’ consensus EPS estimates – even by just a penny – can result in significant downward pressure on stock prices. So, most management incentive plans included a very significant step at threshold, recognising the sensitivity of this measure and providing a meaningful payout benefit for achieving the threshold EPS goal, which is often aligned with consensus EPS estimates. If a company missed consensus EPS estimates, there was a significant penalty in payouts for incentive plan participants: zero payout. The payout step served to align management incentives with shareholder outcomes.
With EPS as the performance measure, there was a rationale for a significant threshold payout step. However, this approach to payout curve design has been implemented far more widely than just for EPS-based incentives. It is now common to see a large threshold step payout for all financial measures and even many non-financial measures. It is also common for payout curves like those in Figure 1 to be used not only for executives but also for other employee groups, like managers and professionals in sales, marketing, operations and functional support areas.
Risks of incentive payout steps
Most companies have realized that EPS is far less than an ideal measure of financial performance (stay tuned for a future blog post discussing EPS in more detail). As EPS has faded in popularity in favor of more meaningful measures of operating performance, the original rationale for incentive plan payout steps is gone.
More importantly, the steps in incentive plan payout curves may encourage inappropriate risk-taking and behaviors. Large steps have the potential to put management teams and Compensation Committees in a compromising position when performance results are expected to fall right around the threshold payout step.
Consider the situation in which year-end performance is expected to end up just short of the threshold payout step. How would management’s behavior and decision-making change given the impact on individual incentive outcomes?
If performance results end up just above the threshold payout step, Compensation Committees and executives may wonder if short-term decisions were made simply to push performance over the payout threshold. Did mangers delay expenditures on training, development, hiring, equipment maintenance or even marketing? Did the sales and marketing team reduce prices or provide more favorable terms to customers to push volume and pull revenue forward from next year’s first quarter into this year?
If decisions on these matters are influenced by individual incentive payments rather than solid operational and/or strategic reasons, then the incentive plan perversely introduces unnecessary business risks.
Potentially even worse is the situation in which performance results fall just short of the payout step.
Consider this potential discussion with the executive team or Compensation Committee: management could reasonably argue that some sort of payout is deserved, since they could have delayed expenditures or pulled revenue forward to this period to achieve the threshold goal, but avoided doing so because it risked longer-term sustainable performance.
Even without management’s lobbying, the Compensation Committee might feel that the team deserves at least a small payout greater than zero, placing them in an awkward position relative to the original plan design and expectations, and potentially damaging the pay for performance relationship. Moreover, a non-zero payout sets a precedent that the Committee will provide some award, even for performance below threshold.
A better alternative: eliminate your payout steps
So, how to address the risk associated with incentive plan payout steps? Consider one of two alternatives: (A) eliminate payout steps entirely, or (B) significantly reduce the “height” or magnitude of your payout steps.
Alternative A. Eliminate payout steps entirely
Rather than setting a payout step for achieving the threshold performance goal (e.g., 50% of target at threshold), payout can begin begin at 0% at threshold. This approach eliminates potential gaming or inappropriate business decisions that may occur when performance is forecast to fall just short of the threshold as discussed above. An example of this approach is provided in Figure 2.
In this design, participants will earn a smaller payout at the same threshold level of performance (relative to the typical approach illustrated in Figure 1), but incremental improvements in performance are rewarded with a steeper payout curve. One way to ameliorate the impact of removing the threshold step is by easing the threshold performance goal. If, for example, the performance threshold was reduced to 60% a similar payment would be provided for 80% performance outcomes relative to the original stepped performance curve in Figure 1. See Figure 3 below.
Note that in Figure 3 there is a steeper payout curve for above target performance. This indicates that incremental performance above target is valued more highly than incremental performance between threshold and target. An increase in the slope of the payout curve is often called an accelerator. Accelerators can have similar risks to steps, so consider your payout curve carefully and adjust as needed so incremental performance is appropriately rewarded between threshold, target and stretch goals.
Alternative B. Significantly reduce magnitude of payout steps
If eliminating your payout steps seems too significant a change for your company to digest, then consider reducing the height of any steps. Instead of a payout of 50% of target for achieving threshold performance, reduce the threshold payout to 25% or 20%, see Figure 4.
While the step and associated risk still exists, the magnitude is reduced through a smaller threshold incentive payment. Many of our clients have taken such an approach after concluding that a 0% payout would have been too much of a challenge to communicate to participants.
Another important risk mitigation strategy is to ensure that all final payments are subject to Compensation Committee discretion, rather that determined purely by a formula (see Formulaic Executive Compensation: Risks and Rewards). Principles for discretion should be set in advance of any decision on payment outcomes, and can include things such as: material and unanticipated changes in the external environment (e.g., commodity pricing, forex, or regulation); changes in the internal business environment or strategy; or even adjustments for executive behaviors (e.g., risk-adjustments) and how results are achieved.
Long-term incentive steps add even more risk
All of these same risks apply to long-term incentive (LTI) programs with payout steps as well, and with potentially greater implications.
Many public companies have implemented (or are considering) LTI programs that measure total shareholder return relative to a peer group of companies (relative TSR, or rTSR). For example, a common design provides a threshold payout of 50% of target for achievement of rTSR at or above the 40th percentile of peers.
Consider the risks of the design described above.
As the company nears the end of the 36-month performance period, all participants should be able to estimate where their company’s TSR falls relative to peers (and it is always a good practice for companies to regularly communicate this information to participants anyway). If rTSR is at the 39th percentile of peers, imagine the potential decisions the management team could make to push their company’s rTSR over the threshold goal. Some of these decisions may not contribute to real, long-term, sustainable shareholder value, the intended outcome of an LTI plan.
Management may decide to communicate more aggressive internal expectations of performance for the following year; fast-track product releases or innovations; or communicate anticipated geographic growth, or discussions with potential M&A partners. Anything that increased rTSR an incremental point or two would have a significant effect on the earned compensation of the executive team.
Similar to the risk of diluting the pay for performance relationship described above, if rTSR falls just short of threshold, would the Compensation Committee feel compelled to make a discretionary award to participants?
Where LTI is a significant portion of total compensation, executives may come to undervalue this component of their reward if forecasts indicate that no payment will be made even if performance is only marginally short of the threshold. This devaluating of a whole portion of total compensation may also introduce retention or flight risk into the management cohort.
The design of incentive payout curves has not evolved significantly for quite some time. But with the additional focus on compensation risk, and the importance of aligning management to long-term sustainable shareholder value, now is the time to review incentive plans. Where there are risks associated with payout relationships that contain significant steps, Compensation Committees should eliminate those risks – or at least mitigate risk by reducing the height of those steps – by redrawing management incentive plan payout curves.
Compensation Committees should regularly review and adjust compensation plans to ensure plans continue to support the business strategy and effectively reinforce the pay for performance relationship. Eliminating or mitigating risk by addressing incentive payout steps can improve pay for performance alignment.
Sheffield Barry is an Executive Compensation and HR Consulting firm, providing customized advice to clients at an affordable price. We leverage technology to deliver data and analysis as efficiently as possible, so we can invest more time understanding our clients’ unique business issues to develop custom solutions and advice. For more information, please visit us at SheffieldBarry.com or email us at info@SheffieldBarry.com.