Article

Relative total shareholder return survey: an opportunity to compare, assess and fine-tune plans

July 12, 2018

By Stephen Zwicker, Bill Turner and Mark Daniels

Relative total shareholder return (TSR) remains the most popular performance metric for long-term incentive awards. These awards compare investment returns for a company against peer companies over a specific time period. Multiple surveys have shown that TSR is used by more than half of companies offering such performance-based awards, either as a single performance metric or as part of a suite of metrics. As such, our clients are interested in any plan design information and trends related to these plans.

A valuable tool

TSR’s popularity as a benchmark reflects dual demands faced by companies: the need to offer the proper incentives and rewards to covetable talent, and the growing demand among shareholders to align long-term compensation with performance. It offers:

  • A transparent approach to tie pay to actual performance
  • An early alert to compensation levels that may either hurt say-on-pay support or reinforce the appropriateness of rewards
  • A way to measure whether the right talent is in place to achieve articulated strategies and goals

What we found

This article provides insights into relative TSR design practices based on a recent survey of our clients’ plan provisions. It also presents an opportunity to evaluate how your company’s remuneration framework compares with others and whether any changes warrant consideration.

Figure 1: Form of settlement

graph showing that most plans are stock-settled at 81.3%

Not surprisingly, most plans are stock-settled (Figure 1). This allows for fixed accounting. The fair value is determined as of the grant date using a pricing model, such as Monte Carlo simulation, and the resulting expense is recognized over the required service period. It is not updated for actual performance. Cash-settled plans are marked to market.

Figure 2: Threshold performance and vesting

graph showing the minimum performance required to warrant a payout is most commonly set at 25th percentile Graph showing that 59.6% set their threshold vesting or payout at 50%

Threshold performance (Figure 2a), the minimum performance required to warrant a payout, is most commonly set at the 25th percentile (i.e., 25% of peers have worse TSR performance), with a mix of companies above and below that level. Threshold vesting or payout (Figure 2b) is most commonly set at 50% of target. Other vesting percentages are almost entirely below this level.

Figure 3: Target performance and vesting

Graph showing that the 50th percentile performance level is most popular at 85% Graph showing target payout is almost always 100%

There is little variation in target performance levels and vesting (Figure 3a). They are almost always set at 50th percentile performance with target vesting at 100% (Figure 3b). A small number of companies require above median performance in order to earn target vesting. This can satisfy shareholder requirements that the company be better than average to earn target vesting.

Figure 4: Maximum performance and vesting

Graph showing that the most common maximum performance level is the 75th percentile Graph showing that maximum vesting is 200% of target

The most common maximum performance level is the 75th percentile (Figure 4a), but a significant number of plans establish higher levels. Maximum vesting is 200% of target (Figure 4b) about two-thirds of the time and 150% about a quarter of the time, with little variation.

Figure 5: Capped vesting and payouts

A little more than a quarter of plans cap vesting at target if the company’s absolute TSR is negative during the measurement period Graph showing that overall payout caps are using by only 3.9% of our clients

A little more than a quarter of plans (Figure 5a) cap vesting at target (or something less) if the company’s absolute TSR is negative during the measurement period (or below some other threshold). Even if performance is above target relative to peers (i.e., negative TSR is better than the negative TSRs of peers), vesting is limited if shareholders don’t earn a positive return. In addition to satisfying shareholder concerns about negative returns, this feature will lower the fair value of the grant.

Overall payout caps (Figure 5b) are used by only 4% of our clients. These caps are used to limit the total amount that a participant can earn from the plan, including share price appreciation and vesting. For example, if the cap level is set at three times target value, the payout would be limited if the share price doubles over the performance period and 200% vesting is earned (for a total of four times target value). In this case, the number of shares delivered would be reduced. In addition to satisfying shareholder concerns about outsized payouts, this feature can significantly lower a grant’s fair value.

Figure 6: TSR measurement starting date and stock price averaging

Graph showing that 93% of plans start measuring TSR at the beginning of the year. Graph showing that 20-30 days is the most common period to find the average stock price of a company to avoid short-term spikes or dips.

More than 90% of plans start measuring TSR at the beginning of the year (Figure 6a), either fiscal year or calendar year. However, the grant date for most plans is somewhere between one and three months into the year, when the board approves the awards. This requires that actual TSR performance between the beginning of the year and the grant date be reflected in the fair value. If the company is off to a good start relative to the peers, the fair value is higher and vice versa. This can add volatility from year to year in the fair value. Some companies have begun to manage this effect by starting the TSR measurement at the grant date.

Roughly 85% of plans use some kind of stock price averaging mechanism for calculating TSR (Figure 6b). This helps avoid short-term spikes or dips in the company’s stock price or peer company stock prices that could unduly impact the vesting result. A mix of averaging periods is used, but the most common is between 20 and 30 trading days (including one-calendar-month periods).

Figure 7: Peer company selection

Graph showing an almost even split between setting the peer group using a custom selected group of companies and using index constituents Graph showing a wide range of companies used as peers but majority use between 10 and 25 companies. Graph showing that 16.4% include peer companies whose main listing is outside the U.S.

It is almost an even split between setting the peer group using a custom selected group of companies and using index constituents (Figure 7a). The index constituent category includes plans that use an industry subset of the index. Broad-based indices are often used when it is difficult to identify a custom peer group.

There is a wide range in the number of companies used as peers, but almost half of plans use between 10 and 25 companies (Figure 7b). There is a balance between percentile ranking issues with too few peers and the administrative effort for too many peers. The 10 to 25 range provides a good balance. Larger groups often indicate the use of an index.

About one in six plans include peer companies whose main listing is outside the U.S. (Figure 7c). Non-U.S. peers create issues about how to measure TSR. Alternatives are to use American depositary receipts, if available; measure TSR in a peer’s home currency; or measure TSR using a peer’s main listing with conversion to U.S. dollars. These issues should be vetted at the outset of the plan to avoid any measurement uncertainty at the conclusion of the plan.

Figure 8: Percentile ranking methodology

Graph showing an almost even split of companies that include themselves with peers when calculating percentile rank and companies that exclude themselves

It is a pretty even split of companies that include themselves with the peers when calculating percentile rank (discrete) and companies that exclude themselves (continuous) (Figure 8). The discrete methodology slots the company into one of the fixed percentile outcomes regardless of how much they exceed the peer below them or trail the peer above them. The continuous methodology places the company in between two peers based on how much they exceed the peer below them or trail the peer above them. Differences between the two methodologies tend to shrink as the peer group grows.

Figure 9: TSR as a modifier

Graph showing 8.6% of plans use TSR to modify the vested outcome from another performance measure

About 10% of plans use TSR to modify the vesting outcome from another performance measure (Figure 9). Vesting is initially determined based on the other criteria and then adjusted upward or downward based on relative TSR performance. For these plans, the TSR modifier condition is subject to a fair value measurement (fixed if stock-settled), and expense is only updated for the vesting outcome related to the nonmarket performance condition.

Figure 10: Dividend treatment

Graph showing 62.4% of clients offer dividend protections for their TSR-based performance awards.

More than half of our clients offer dividend protection for their TSR-based performance awards. This is generally in the form of dividend equivalents (Figure 10) that are accrued during the performance period and paid at the end based on the vesting outcome.

A fresh look

Whether a company has sponsored a relative TSR plan for a number of years or is just thinking about implementing one, a robust review of the plan features is encouraged to ensure a full understanding of the proper pay-for-performance alignment and accounting valuation implications. Many of the core features of these plans have not changed much over the years, but there are new features that have started to gain varying degrees of traction and may add value to companies’ plans.