Proxy advisors have a lot on their plates. They may not have the time, inclination or ability to calculate total shareholder returns. But nevertheless, with access to online databases, they often have enough to query whether executives should have received vesting on their total shareholder return incentive. If, or at some stage, when that happens, it would be useful for the remuneration committee chair to readily respond why vesting has occurred, and why it may differ from the rough and ready online data service version the proxy advisor receives.

Likewise, it would be useful for the Remco chair to know the difference in method that could be employed from year to year should vested interests show TSR results just squeaking over the median.

An example is the percentile rank method. For relative TSR measures based on a percentile ranking calculation, companies must also consider whether they should include themselves (target company) in the comparator group. If the comparator group is defined as the constituents of an index, the company may be included in the index, and thus the comparator group.

Inclusion vs exclusion

The target company’s percentile rank is defined as the proportion of the comparator group whose TSR is less than the target company’s TSR.

When the target company is included in the comparator group, the target company’s percentile rank is simply based on the ranking of the company TSR relative to the comparator group TSR.

When the target company is excluded from the comparator group, the target company’s TSR will instead fall between the nearest lowest and highest company TSRs. The target company’s percentile rank is then estimated based on the percentile ranks of the two nearest peer company TSRs.

There are many ways to estimate the percentile rank when the target company is excluded from the comparator group. The Excel PERCENTRANK functions uses linear interpolation to estimate the percentile rank.  To a purist this is not correct. Percentile ranks are not uniformly distributed. That is, the difference between 50^th and 60^th percentile is not the same as the difference between the 60^th and 70^th percentile. But linear interpolation is simple and easy to understand.

Example

Consider an example which highlights the main differences between the two approaches, using Excel’s PERCENTRANK methodology.

Table 1: Example relative TSR percentile rank using Excel’s PERCENTRANK

Note that when the target company is included in the comparator group, the percentile rank does not change if the company rank does not change. That is, the target company is at the median regardless of where TSR falls between 50% and 150%.

When the target company is excluded in the comparator group, the percentile rank changes on a linear scale between 50% and 150% TSR.

The difference in the methods could be the difference between nil vesting and partial vesting. Thus, in determining the relative TSR methodology, percentile rank calculation is an important consideration.

The chart below illustrates this. With the target company in the comparator group, the percentile rank moves in a step-wise manner based on each peer company’s TSR ranking. In contrast, when the target company is excluded in the comparator group, the percentile rank moves linearly with the company TSR between the two nearest peer company’s TSR.

Figure 1: Percentile ranks differences when including versus excluding target company in the peer group

Conclusion

As shown in the example, the differences between the two methods become more pronounced when there are large variations in peer company TSRs in the comparator group.

Companies using relative TSR with a small bespoke peer group or high volatility peer companies should consider excluding itself from the comparator group for more valid percentile ranking outcomes.

As we have often discussed in the past and this example illustrates, there are many factors that must be considered for a robust relative TSR methodology. It is worthwhile for the board remuneration committee to have these addressed in a written and approved relative TSR methodology (see HERE).

© Guerdon Associates 2024 Back to all articles