Executive Compensation Analysis: Rebuilding the Committee's Work from Public Filings

The package you will see only once

Your renewal is on the calendar for next quarter. Somewhere in the next few weeks the compensation committee will receive a package from its consultant: a peer group, a table of disclosed pay for each peer, and a chart showing where your total compensation falls against it. You will see that package once, probably as a summary, and the conversation will move quickly from there.

The analysis is already finished when you walk in, while your own case is usually a handful of figures from memory and a few consumer pay lookups that cannot say where they came from.

Every input to the committee's work, however, is drawn from filings anyone can read. The disclosure rules, the tables and the peer proxies are public. What separates the two sides of the table is the method, and the method can be reproduced.

By the end of this article you will know what a committee's executive compensation analysis contains, which public filing supplies each input, and how to rebuild the sequence yourself with a citation behind every figure.

What a committee's executive compensation analysis contains

Strip away the formatting and a committee analysis has the same few components in nearly every company. It defines a peer group. It collects disclosed pay for the executives in that group. It puts the pay on a common footing so that unlike packages can be compared. It computes where each pay element falls across the group, usually at the 25th, 50th, 75th and 90th percentiles. And it records the reasoning for the decisions it reaches.

The sources behind the work matter as much as the steps. Firms that sell executive-compensation data and surveys, such as Equilar, sell primarily to boards and compensation consultants, and do not publish self-serve public pricing. Chief Executive Group publishes a compensation report built for companies setting executive pay, including private companies. Neither is an individual-facing benchmarking tool.

What you can do is rebuild the public-company portion from the same underlying filings. That is the premise of the rest of this guide. For a broader orientation to benchmarking your own role, see how to benchmark your own executive salary.

Step one: define the peer group before you look at any pay

The most consequential decision in any executive compensation analysis is made before a single figure is pulled: who is in the comparison. Change the peers and the percentile moves, even if no one's pay has changed.

Committees typically build a peer group from a few filters. Company size, commonly measured by revenue, is one. Sector is another, and SIC and NAICS are the standard sector-classification systems used to define comparable peer groups. Region and ownership type (public, nonprofit, sponsor-backed) round out the usual set. A peer group should also be large enough to support a percentile at all; a handful of companies cannot meaningfully produce a 75th or 90th.

Two habits make your version more defensible than most. First, write down the rule for each inclusion and exclusion at the time you make it, rather than reconstructing the logic after you see the result. Second, resist the pull to add flattering peers.

The Peer-Set Construction Playbook is built for exactly this stage: it helps you build a defensible executive peer group by size, sector, region, and ownership, with a documented-rationale log that stands up to a comp committee.

Step two: pull the disclosed pay from proxy statements

For public companies, the primary source is the proxy statement. Proxy statements must generally disclose compensation for the top five most highly paid executive officers, the named executive officers (NEOs), including the CEO and CFO, in the Summary Compensation Table (Meridian Compensation Partners, 2025). That table is where salary, bonus, stock awards, option awards, non-equity incentive pay and other components appear by year. Our guide to the Summary Compensation Table covers how each column is built.

The filing is free to retrieve. EDGAR provides free public access to filings, and SEC EDGAR guidance is public domain (SEC.gov, 2025). So the raw material for an executive pay analysis from proxy statements is open to anyone willing to read it.

Two details deserve attention at this stage. The first is that the table shows grant-date accounting values for equity, not what an executive eventually realizes; read the footnotes before treating any number as take-home pay. The second is which executive you are comparing to. The CEO line suits a CEO. If you hold another C-suite role, the NEO lines are your comparison, and the scale shifts: all non-CEO NEOs' total compensation was 38% of CEO pay in the Russell 3000 and 31% in the S&P 500 in 2024 (Harvard Law School Forum on Corporate Governance, 2025, FY2024).

Step three: cross-check against Pay Ratio and Pay Versus Performance

Two further disclosures give you a second view of the same peers, and they are structured differently from the Summary Compensation Table.

The CEO Pay Ratio rule, Item 402(u) of Regulation S-K, was adopted on August 5, 2015 under Section 953(b) of Dodd-Frank. It requires annual disclosure of CEO total compensation relative to that of the median employee (Harvard Law School Forum on Corporate Governance, 2015). For a sense of scale, the S&P 500 median CEO-to-median-worker pay ratio rose to 196:1 in 2023, and the median S&P 500 employee earned $81,467 that year (Equilar / Associated Press CEO Pay Study, 2024, FY2023).

Pay Versus Performance, Item 402(v), was adopted on August 25, 2022 and became effective for 2023 proxies. It mandates Inline XBRL tagging (Mintz, 2022). Inline XBRL is a structured data language that produces a single human- and machine-readable document (SEC.gov, 2024). One sourcing point trips up many people: Item 402(v) is required in proxy and information statements, not in the Form 10-K, so the machine-readable pay data lives in the proxy (Greenberg Traurig, 2023). Our explainer on the Pay Versus Performance disclosure walks through the table.

The practical value of the cross-check is reconciliation. If the compensation figure tagged in the structured disclosure does not match what you keyed from the Summary Compensation Table, you have found an error in your data entry or a definitional difference you need to understand before it reaches the table. Confirm current filing requirements, phase-in details and any later amendments with the SEC or the filing itself.

Step four: put pay on a common footing

A peer set of ten companies produces ten different pay structures. One weights salary, another leans on performance equity, a third paid a large one-time grant. Before any percentile means anything, the figures need to be normalized.

At minimum, align the fiscal year across peers, separate recurring compensation from one-time awards, and compare like components against like. Long-term incentives, chiefly performance equity, remain the primary driver of CEO pay (Harvard Law School Forum on Corporate Governance, 2026, FY2024), so how you treat equity affects almost everything downstream. Mixing grant-date values for one peer with realized values for another will distort the result more than any choice of peer will.

Anchor your sense of scale to sourced context rather than intuition. Median S&P 500 CEO total compensation was $17.1 million in 2024, up 9.7% year over year (Equilar / Associated Press CEO Pay Study, 2025, FY2024).

Step five: compute percentiles and read them correctly

With a normalized table, percentile positioning is arithmetic. Sort the peer values for each element and total, then find where your figure falls relative to the 25th, 50th, 75th and 90th percentiles. Our guide to CEO total compensation percentiles covers interpretation in more depth.

Here is a worked example with round, illustrative numbers rather than real company data. Suppose ten peers have normalized total compensation that, sorted from lowest to highest, runs from 3 to 12 in arbitrary units in steps of one, and your figure is 7.5. Five peers sit below you and five above, so you are near the median. If a counteroffer or target would move you to 10.5, you would sit above eight of the ten peers. A committee reasoning from a median anchor will hear that as a move from the middle of the group to the upper quartile or beyond, and it should be prepared to explain it in those terms.

Say-on-pay outcomes show how rarely boards face binding pushback: roughly 1.2% of Russell 3000 say-on-pay votes failed in 2024, below 2.1% in 2023 (Harvard Law School Forum on Corporate Governance, 2025, FY2024).

Nonprofit executives: the same method on Form 990

If you lead a nonprofit, the filing changes but the steps do not. Form 990 is the annually required, publicly accessible return for all 501(c)(3)s (Library of Congress, 2026), and Schedule J reports compensation for officers, directors, trustees, key employees and highest-compensated employees (IRS, 2024). The Taxpayer First Act, of July 1, 2019, requires e-filing of the Form 990 series, which makes named-officer compensation available as structured e-filed data (Holland & Knight, 2019).

Peer definition matters even more here, because mission area changes the range considerably. In the Candid 2025 Nonprofit Compensation Report, median executive compensation in 2023 was $68,958 in religion, $109,600 in food, agriculture and nutrition, and $202,490 in science and technology research (Candid / NonProfit PRO, 2025, FY2023). The report analyzed 217,556 compensation records from 130,794 organizations, drawn exclusively from Forms 990 and 990-EZ. It is licensed at $449 for a single user and $1,199 for an organization, per the 2025/2026 edition as reported by GrantLedger. It is a static annual PDF, priced for institutional and consultant buyers, so an individual executive will usually get further by reading the underlying filings for a defined set of peers.

Rebuilding the work in practice

Putting the sequence together: define the peer group and log your reasons, pull the Summary Compensation Table for each peer, reconcile against the Pay Ratio and Pay Versus Performance disclosures, normalize, compute percentiles, and write down the rationale next to every number. The last step is the one most executives skip, and it is the one that makes the case usable. A figure with a filing, a fiscal year and a stated peer rule behind it can be checked by the person across the table. A figure without those can only be argued with.

This is the work CEOSalary performs as a workspace. It parses public SEC DEF 14A Summary Compensation Tables, cross-validates against XBRL Pay-Versus-Performance and CEO Pay Ratio disclosures, ingests IRS Form 990 Part VII and Schedule J from bulk e-filed XML, builds peer sets by revenue band, sector, region and ownership type with a minimum peer-group-size gate, computes 25th, 50th, 75th and 90th percentile positioning with per-figure source citation, and exports a source-cited negotiation PDF. It does not cover private-company named executive pay or international comparables.

If you would rather start by hand, the Quick-Start Peer Comparison Worksheet takes peer filings in and gives your percentile out, positioning your comp against a public-company and nonprofit peer set you build. Pair it with the peer-set playbook above and you will have the same five steps in a form you can bring to the table.

This article describes how public disclosures work and is not legal, tax or financial advice. Confirm filing requirements and thresholds with the SEC, the IRS, qualified counsel or the filing itself.