Reading the CD&A: Compensation Discussion and Analysis

Your comp committee renewal is on the calendar, and you have a stack of proxies to read

You have three weeks before a scheduled comp committee conversation, and someone on the board has already circulated two peer proxies "for context." You open the first one, skip past the Summary Compensation Table you already know how to read, and land on a section titled Compensation Discussion and Analysis — pages of prose about pay philosophy, peer groups, and performance metrics with no obvious table to scan. This is the section boards actually use to justify a number, and it's the section most executives skim past on their way to the dollar figures. By the end of this piece, you'll know exactly which parts of a CD&A to read closely, which to skip, and how to translate a peer board's stated pay philosophy into a percentile position you can bring into your own conversation.

The Compensation Discussion and Analysis, or CD&A, is the narrative section of a DEF 14A proxy statement that sits alongside the Summary Compensation Table — but it does a different job. The table shows what was paid. The CD&A explains why the board believes that amount was right. If you've already worked through how to read a DEF 14A proxy statement, the CD&A is the section that turns those disclosed numbers into a defensible argument — which is exactly what you need when you're the one being benchmarked.

What the CD&A actually discloses — and how it differs from the Summary Compensation Table

Proxy statements are generally required to disclose compensation for the five most highly paid executive officers, including the CEO and CFO, in the Summary Compensation Table, per Meridian Compensation Partners' 2025 guidance on proxy disclosure requirements. That table is backward-looking and numeric: salary, bonus, stock awards, option awards, and total compensation for the fiscal year just closed, for each of those named executive officers.

The CD&A sits in front of that table and does the interpretive work. It typically walks through the board's compensation philosophy, the peer group used to benchmark pay, the specific performance metrics tied to incentive plans, and the rationale behind any unusual pay decisions — a retention award, a one-time equity grant, a below-target bonus payout. Where the Summary Compensation Table answers "what did this person get paid," the CD&A answers "why did the board think that was the right number." If you've already reviewed how the Summary Compensation Table is built, the CD&A is the missing half of the story: the reasoning that produced those figures.

How comp committees set a pay philosophy and target percentile

Most CD&As state a target percentile explicitly — a board might disclose that it targets the market median, or the 75th percentile, for total direct compensation relative to its named peer group. This is the single most useful sentence in the entire document for benchmarking purposes, because it tells you the board's stated intent, not just the outcome.

Context matters here. Median S&P 500 CEO total compensation reached $17.1 million in 2024, up 9.7% year over year, per the Equilar / Associated Press CEO Pay Study for 2025 (fiscal year 2024). That figure describes the market center of gravity — it says nothing about where any individual board has chosen to position its own CEO relative to that center. A board explicitly targeting the 75th percentile of its stated peer group is making a different commitment than one targeting median, even if both boards cite similar market data. Reading the CD&A tells you which commitment you're actually being compared against, rather than assuming everyone targets the same point.

The gap between internal and external hires adds another layer worth watching for in the CD&A's rationale section. Externally hired S&P 500 CEOs carried median total pay of $10.99 million versus $7.76 million for internally promoted CEOs — roughly a 41.6% gap — according to Equilar's 2015 analysis. A board explaining a new CEO's pay package against this backdrop will often address hire type directly in the CD&A, because it materially changes what "market rate" means for that specific transition.

Reading the peer group section without getting lost in company names

Every CD&A that references market positioning will name or describe its compensation peer group — typically companies of similar revenue scale, sector, and sometimes ownership structure, often selected using SIC or NAICS sector-classification codes as a starting filter. This is worth reading slowly, because the peer group is the board's own definition of "comparable," and that definition can be generous or conservative.

Look for three things: how many companies are in the group, whether the group changed from the prior year's proxy (and why, if stated), and whether the company's own scale sits near the middle or an edge of that peer set. A company that sits at the small end of its stated peer group's revenue range, but targets median pay against that group, is effectively targeting a higher relative position than the raw percentile number suggests. The CD&A sometimes explains this directly; more often, you have to infer it by cross-referencing the peer list against public revenue figures yourself.

Cross-referencing the CD&A against the Pay Versus Performance table

Since Item 402(v) of Regulation S-K took effect for 2023 proxies, companies have also been required to disclose a Pay Versus Performance table, tagged in Inline XBRL, connecting compensation actually paid to specific financial and total shareholder return metrics, per Mintz's 2022 summary of the rule. This table lives in the proxy statement itself, not in the annual report, according to Greenberg Traurig's 2023 analysis — which means the CD&A's performance-metric narrative and the PvP table's structured figures sit in the same document and should tell a consistent story.

If a CD&A describes an incentive plan weighted toward revenue growth and total shareholder return, the PvP table should show compensation actually paid moving in some relationship to those metrics year over year. Where the narrative and the structured data diverge — a CD&A that emphasizes pay-for-performance language alongside a PvP table showing compensation rising while performance metrics fell — that gap is itself information. Boards rarely face a binding vote against this reasoning: only about 1.2% of Russell 3000 say-on-pay votes failed in 2024, down from 2.1% in 2023, per the Harvard Law School Forum on Corporate Governance's 2025 analysis of fiscal year 2024 votes. A low failure rate doesn't mean every CD&A's logic holds up to scrutiny — it means shareholders rarely push back regardless.

The CD&A tells you what a board believes it's paying for. The Summary Compensation Table and the Pay Versus Performance data tell you whether that belief matches the numbers.

What the CD&A won't tell you — and where a benchmarked case fills the gap

A single peer's CD&A gives you one board's stated philosophy and one peer group's structure. It does not give you a percentile position across a properly constructed set of peers matched to your own company's revenue, sector, and ownership type — that requires pulling Summary Compensation Table data across multiple filings and building the peer set yourself, or working with a tool built for that purpose.

This is the gap CEOSalary is built to close: it parses SEC DEF 14A Summary Compensation Tables, cross-validates figures against XBRL-tagged Pay Versus Performance and CEO Pay Ratio disclosures, builds a peer set by revenue band, sector, and ownership structure with a minimum group-size gate, and computes your percentile position — 25th, 50th, 75th, 90th — with every figure tracing back to its source filing. Reading one CD&A teaches you the method; a benchmarked peer set applies that method at the scale a real negotiation requires.

Turning a CD&A read into your own negotiation case

Before your next comp committee conversation, pull the CD&A from two or three named peers, note their stated target percentile, and check whether their peer group composition actually resembles your own company's scale and sector. Then go a level deeper: pull the underlying Summary Compensation Table figures yourself, or work from a structured guide to parsing a peer's Summary Compensation Table that walks through the line items methodically.

A CD&A read in isolation gives you one board's argument. A percentile position built across a proper peer set, sourced back to filings the way EDGAR's free public access was designed to support, gives you your own. See current pricing or browse the full reading guide library to build that case before you walk into the room.