1. Peer Groups

Building off the definition of the competitive market in the compensation philosophy, most compensation committees use a peer group to benchmark pay and/or performance levels for the most senior executives in the company (CEO, CFO, and three other highest-paid executive officers). The peer group that the company selects is intended to reflect the competitors for talent for senior executive roles and also considers who the company competes with for customers and capital.

Most companies select peers based on comparability in terms of industry and/or other business characteristics, along with business scope:

Industry Comparability: Committees may look for peers that operate in the same industry narrowly defined (e.g., auto manufacturing) or more broadly defined (e.g., durable goods manufacturing). Global Industry Classification (GIC) codes and Standard Industrial Classification (SIC) codes are frequently used to help committees screen for peer companies.
Business Characteristics: When companies struggle to find direct industry peers, or are in a highly concentrated industry, they may expand the search criteria to include companies that share key business characteristics (e.g., consumer goods companies, global companies, professional service firms, asset-intensive businesses, companies that are highly cyclical or sensitive to commodity prices, highly regulated entities, etc.).
Company Scale: Typically, peer companies are targeted to be similar to the company in size at median. A general rule is to look for peer companies with revenues (or assets in the case of financial services companies) from 0.5x to 2.0x that of the company. Companies may expand beyond this range to include peers that are viewed as particularly important competitors for talent. Alternative and/or secondary criteria may also be used (e.g., market cap or profit margin) to supplement revenue.

The peer group is important in that it serves as the basis for establishing target executive compensation levels. It can also be used as a way to test the pay-for-performance relationship. If a company wants to calibrate the pay-for-performance relationship correctly, it is ideal to use a single peer group for pay and performance comparisons. That way, the company can demonstrate that the percentile positioning of its compensation levels is aligned with the percentile positioning of its performance. A key challenge is that sometimes peer companies that are relevant for pay comparisons may be less relevant for performance comparisons. In cases where the use of multiple peer groups is required, the rationale for the use of more than one peer group and how each peer group is used should be clearly disclosed in the company’s CD&A.

Companies are often criticized for “gaming” the makeup of their peer group to try to increase target pay levels. This criticism often occurs where companies include much larger companies in the peer group. At times, this criticism is well-founded. Pay levels systematically increase with company size. When developing a peer group, a key principle to keep in mind is that the median size (in terms of revenue, assets, market cap) among the peer group should be close to the size of the company.

For companies that follow premium pay positioning, targeting pay levels above the median, it is also important that the range of size among the peers not be too large. Inclusion of peers that are larger than 2x the size of the company may have a limited impact on the median revenue and median compensation levels for the peer group as a whole (particularly if the peer group also includes some companies that are significantly smaller than the company). However, it may have a more significant impact on the 75th percentile compensation levels.

Many times, management will feel that there are significantly larger companies in the industry (e.g., more than 3x the company’s revenue) that should be included in the peer group as they may be a source of talent for the company or may potentially recruit the company’s talent away. In these cases, it may be appropriate to consider these companies for purposes of assessing pay practices, but it can be problematic to use them to benchmark compensation levels (though in isolated cases/industries, there may not be a better alternative). While it is reasonable to be aware of the compensation practices of larger companies, including them in the calculation of summary statistics can skew the findings and may make the company a target for external criticism.

Another important perspective to understand on peer groups is that the two largest and most influential shareholder advisory firms (ISS and Glass Lewis) each have their own approaches to defining peer groups. They use the peer groups that they develop to conduct their own assessments of the competitiveness of your company’s pay levels and the alignment of the CEO’s pay levels with the company’s performance on both an absolute and relative basis.

ISS peer group development continues to evolve from year to year, and their current approach uses the subject company’s self-defined peer group as a key input in their selection criteria. They tend to take a strict approach to excluding any companies from the peer group that are less than 0.4x the revenue (or assets for financial services companies) of the company or more than 2.5x the revenue (or assets) of the company. ISS typically selects a peer group that contains between 14 and up to a maximum of 24 companies, with 12 as the absolute minimum.

Glass Lewis’s peer group methodology also begins with the subject company’s self-disclosed peers. They then apply multiple criteria including peers of peers and size criteria. Based on their methodology, Glass Lewis considers each of the criteria on a weighted basis to form the final peer group.

ISS and Glass Lewis’s peer groups will each overlap with your company’s self-identified peer group to some degree, but there likely will be differences. It is helpful to familiarize yourself with these differences in order to anticipate where the shareholder advisory firms may identify misalignments between pay and performance.

Below is an illustration of developing a peer group:

Key Questions for Committee Members to Ask:

Is our peer group comparable to us in terms of revenue? Is the answer substantively different for other measures of size (e.g., market cap, net income, EBITDA, etc.)?
Is the range in size among our peers wide or narrow (e.g., what is the difference between the median and 75th percentile revenue?)?
Are there any large or small peers that may be skewing the competitive findings? If so, how can we address this issue?
Are all peers equally relevant for financial performance comparisons? Would it be helpful to look at a subset of our most direct competitors for financial performance comparisons?
What is the degree of overlap between our self-defined peer group and the peer groups used by ISS and Glass Lewis?