Executive Compensation Peer Group Selection for a CFO

When the peer tab in your spreadsheet is two proxy seasons old

The compensation committee has moved your review up to March. The only peer data you have is a tab you built two proxy seasons ago. Twelve company names sit down the left side, and nothing records why they were chosen. Three of them have since been acquired. One changed its fiscal year-end. You cannot say whether a fourth ever named a CFO in its Summary Compensation Table at all.

Rebuilding the tab is not hard. The harder part is that the result is only as defensible as the decisions behind it. If you cannot say why a company is in the set, any percentile you calculate from it is easy to question, including by you.

This article treats executive compensation peer group selection as a sequence of recorded decisions. It covers what question the set must answer, how to screen candidates by revenue band, sector and region, how many companies are enough, and how to log an in-or-out reason for each one.

Decide what the peer set has to answer

A peer set is built for a question. "What do comparable CFOs receive?" is too loose to select against. Tighten it before you open EDGAR:

• Which measure? For public-company peers, the usual measure is the SCT total. It includes equity at aggregate grant-date fair value, a rule in place since the SEC's 2009 amendments. It is not realized or take-home pay. Compensation actually paid under Item 402(v) is a different measure, because it reflects changes in the value of equity awards during the year. Pick one and say which one every time a figure appears.
• Which proxy years? Item 402(c) requires the SCT to show each named executive officer's compensation for the last three completed fiscal years. A proxy filed in 2025 typically reports the fiscal year that ended in 2024. That offset breaks for off-calendar fiscal years, so record both the proxy year and the fiscal year for each row.
• Which conversation? An annual review, a succession, a new offer and an IPO-readiness hire each pull toward different peers. A succession may warrant a closer look at companies that recently changed CFOs. An offer comparison may weight companies near the hiring company's size.

Write the question in one sentence at the top of your working file. Every screening decision below is then a test against that sentence.

The measure also limits what the set can show. Grant-date equity values compare what boards decided to award, not what executives realized. A peer set can support a discussion of positioning among disclosed figures. It cannot say what any individual CFO is worth.

Executive compensation peer group selection: three screens and a reason for each

Most hand-built sets use some version of three screens: size, sector and geography. Each screen is a judgement, and each deserves a written reason.

Revenue band

Revenue is the most common size screen because it is disclosed consistently and is easy to defend. Choose a band around your own company's revenue and decide in advance how wide to make it. A narrow band gives closer comparables but fewer disclosed CFOs. A wide band gives more rows but weaker comparability.

Record the band as a rule ("revenue between X and Y in the most recent fiscal year") rather than as a feeling. Note the fiscal year the revenue figure refers to, since a company can move across a band boundary between proxy seasons. Our companion piece on CFO pay by company size goes further into how size screens behave in practice.

Revenue is not the only size measure. Market capitalisation, enterprise value and headcount are alternatives. Whichever you choose, apply it to every candidate in the same way. Mixing measures between candidates is a common source of hidden inconsistency.

Sector

Sector screens carry more judgement than size screens. A broad grouping gives more peers. A narrow one gives better comparability and risks leaving you with too few disclosed CFOs to use.

Use a stated classification, such as NAICS groupings, so the rule is repeatable. Then look at the candidates the classification returns and ask whether the CFO job at each is recognisably similar to yours. A bank, a software company and a hospital operator can all report to the same label of "large employer" while the CFO role differs a good deal. The Conference Board's analysis, republished by the Harvard Law School Forum on Corporate Governance, found that CFOs in financials earn 41% of CEO pay and described wide industry differences. That is a published, attributed observation about financials, not a rule for your sector. The detail on how CFO compensation varies by industry is worth reading before you fix a sector rule.

Region

Region is the weakest of the three screens for many CFOs, because a national public company often competes for finance talent nationally. Use it when you have a reason: a company whose finance team is concentrated in one market, or a role that is location-bound. If you apply a regional filter, state why. If you cannot give a reason, consider leaving it out and widening the pool.

Recording the reason

For every candidate that survives or fails the screens, write one line: the company, the proxy year and fiscal year of the row you used, the screen it passed or failed, and the reason. "Excluded: fiscal-year revenue outside band" is a complete entry. So is "Included: sector match, but CFO role became interim mid-year; flagged for exclusion." The log is what turns a list of names into a set you can defend.

Use an issuer's own peer group as an input, not a template

Many companies describe their own compensation peer group in the Compensation Discussion and Analysis. Item 402(b) of Regulation S-K governs the CD&A. Item 402(b)(2)(xiv) gives, as an example of material information, whether the company engaged in any benchmarking of total compensation or a material element of it, identifying the benchmark and, where applicable, its component companies (17 CFR 229.402).

That makes the CD&A a useful place to find candidates. Read it as a list of names to consider rather than as a ready-made answer, for three reasons.

First, the issuer's group was built for the issuer's purpose, usually to inform its compensation committee about its executives as a whole. It was not built to give you a CFO comparison.

Second, the group may include companies that do not name a CFO in their own SCT, or that you would screen out on your own rules.

Third, you are choosing peers for your question. The issuer's peer group answers theirs.

If you use an issuer's group as a starting list, note in your log that the candidate came from that CD&A, and note the filing. Then run each name through your own screens. There is no single correct number of peers, and the CD&A of any company you study shows how large its own group is.

How many companies, and when to stop adding them

The question people search for is how many companies belong in a compensation peer group. No single count is right. The practical answer depends on how many peers actually disclose a CFO, not on how many you can name.

That distinction matters because of how disclosure works. Under Item 402(a)(3), the named executive officers are everyone who served as principal executive officer, everyone who served as principal financial officer, and the three most highly compensated other executive officers. A non-SRC filer's CFO is therefore a required NEO whatever their pay rank. We cover the mechanics in named executive officers under Item 402.

A smaller reporting company is different. Its NEOs are the PEO and its two most highly compensated other executive officers. The principal financial officer is not automatically included, so an SRC's CFO may not appear in its SCT at all. Scaled disclosure also covers three NEOs instead of five, and two years of SCT instead of three. Since the SEC's 2018 amendments, a company is an SRC if public float is below $250 million, or if annual revenue is below $100 million and it has no public float or a float below $700 million. Our guide to smaller reporting company NEOs covers what that does to the rows you can find.

An emerging growth company can also use scaled disclosure, and EGCs are exempt from the pay-versus-performance requirement. A newly public candidate may therefore be missing both the CFO row and the 402(v) data you would use to check it.

So count disclosed CFOs, not candidates. Our design rule, applied to the app we are building, is a minimum of five distinct employers before a percentile is shown. The word "distinct" matters. Three proxy years from one company give three rows but one employer, and they should not be allowed to stand in for a wider set. A set of twelve names with four disclosed CFOs is a set of four, and the log should say so.

Keep adding candidates until you reach the minimum on your own screens. If you cannot get there without loosening a screen, loosen it one step, write down which and why, and re-count. If you cannot get there at all, the honest result is that the set is too thin for a percentile. A list of cited disclosed figures, shown without a ranking, is still useful for review preparation. It simply does not claim more than the data supports.

A proposal now before the SEC could make thin coverage more common. On May 19, 2026, the SEC proposed amendments to the public-company filer-status framework in Release No. 33-11419 (SEC fact sheet). As of October 30, 2026, the proposal has not been adopted, and nothing changes unless it is. According to Morrison Foerster, about 81% of public companies would qualify as non-accelerated filers if it were adopted as proposed. If it were adopted, a CFO who is not among a scaled filer's PEO-plus-two NEOs could drop out of that filer's SCT, and CFO peer coverage would thin further. Check the SEC's website for its current status before relying on any of this.

Clean the rows before you rank anything

Once the set has enough disclosed CFOs, the rows themselves need cleaning. This is where a hand-built set most often goes wrong.

• Identify the CFO by title. The row may read "Chief Financial Officer", "EVP & CFO" or "Principal Financial Officer". Interim and former CFOs also appear in the same table, and you must decide whether each belongs. Our design excludes former-CFO rows and flagged rows by default, and your set can follow the same rule if you record it.
• Line up the columns. The SCT has eight compensation columns plus a total: salary; bonus; stock awards; option awards; non-equity incentive plan compensation; change in pension value and non-qualified deferred compensation earnings; all other compensation; and the total. The pension and deferred-compensation column can move a total without any change in pay decisions, so note it when a figure looks out of line.
• Cross-check the figure. Item 402(v) tags the PEO and an average of the other NEOs, never the CFO individually. That average can still act as a consistency check on a keyed total. In our design, SCT CFO totals are checked against the Item 402(v) average non-PEO NEO total, and failures are flagged and excluded. It is a consistency rule, not an audit, and it cannot be applied to a filer exempt from 402(v).
• Cite every figure. Record the filing, the fiscal year and the date you pulled it. A figure you cannot trace is one you should not defend.

A short illustration, using round numbers. Worked example: in a peer set of ten disclosed CFO SCT totals, a figure with seven peers at or below it sits at the 70th percentile under the at-or-below definition. The definition is the share of peer totals at or below your figure. Change which rows are in the set and the answer changes. That is why the log matters more than the arithmetic. For more on the calculation, see how a CFO pay percentile is read.

What a worksheet can do, and what stays manual

Everything above can be done with EDGAR and a spreadsheet. EDGAR Full-Text Search covers filings submitted electronically since 2001, including exhibits. The SEC's fair-access policy limits automated access to 10 requests per second, which is not a constraint for hand work. The route is: company search, then DEF 14A, then the Summary Compensation Table, then the principal financial officer's row. For a single set, a careful CFO or their finance team can do all of it.

What the manual route lacks is structure: a place to hold the screening rules, the in-or-out log, the title classes, and a running count of how many peers disclose a CFO. The CFO Peer Set Builder and Disclosure-Coverage Worksheet lays out those steps, including the coverage check described above. It is a downloadable workbook available now.

The app that CFOCompensation is building would extend the same ideas. It is designed to let you filter peers by revenue band, sector and US region, show a filing citation and last-updated stamp beside each figure, and show a disclosure-coverage indicator for the set. It is built to withhold a percentile when a set has fewer than five distinct employers. None of this is available to use yet.

The limits are plain. The comparables use public-company disclosures only, so private-company pay is not covered. Grant-date equity is not realized pay. A PE-backed CFO who uses disclosed public peers is working with a stand-in, and should label it that way. A well-built peer set supports preparation and does not give an answer about any individual's pay.

About these figures: CFOCompensation works from disclosed figures only, shown as filed and cited to the source filing; any third-party survey or study is named and labelled as such. Nothing in this article is legal, tax or negotiation advice. CFOCompensation and Rovaryn Digital Inc. are not affiliated with the SEC or the IRS. Private-company pay is not covered.