Form DEFN14A Files Dataset

The Form DEFN14A Files Dataset is an archive of definitive proxy statements filed on EDGAR by non-management soliciting parties — activist funds, dissident shareholder slates, insurgent groups, and other third parties soliciting votes against or independently of a registrant's management. One record is a single DEFN14A submission, identified by its EDGAR accession number and materialized on disk as a self-contained accession folder containing a synthesized metadata.json plus the original primary proxy statement and any exhibits or additional soliciting materials. The dataset spans EDGAR DEFN14A filings from March 1995 to the present, capturing three decades of contested director elections, opposed mergers, and independent shareholder campaigns at U.S. operating companies and registered investment companies. It is delivered as a hierarchy of monthly ZIP containers in TXT, JSON, HTML, and PDF formats, mirroring the SEC's per-document SGML structure for downstream parsing.

Dataset APIs

Dataset Files

What This Dataset Contains

The dataset packages every EDGAR submission of Form DEFN14A — a definitive proxy statement filed under Section 14(a) of the Securities Exchange Act of 1934 and Schedule 14A of Regulation 14A by a person other than the registrant. Where management's definitive proxy statement is filed as DEF 14A, a DEFN14A is filed by a soliciting party that is independent of, and frequently in opposition to, the issuer's management — most commonly an activist investor, a dissident shareholder slate, or a private fund pursuing a contested election of directors, a merger vote, a bylaw amendment, or another shareholder-meeting matter. "Definitive" signals that the document is the final, mailed version of the proxy materials, distinct from preliminary versions filed as PREN14A. Schedule 14A prescribes the disclosure architecture: identification of the soliciting party and its participants, description of their interests, the proposals to be voted on, beneficial-ownership tables, voting mechanics, and the form of proxy.

The dataset is delivered as a hierarchy of monthly ZIP archives whose path mirrors the SEC API URL pattern <datasetIdInUrl>/<YYYY>/<YYYY-MM>.zip. Each ZIP contains a single top-level folder named <YYYY-MM>/, under which sits one folder per accession number. File types found in the dataset are TXT, JSON, HTML, and PDF; modern records are dominated by HTM/HTML for the proxy statement and exhibits plus the synthesized metadata.json. The dataset is sparse: DEFN14A is a relatively rare form because it is filed only by non-management proxy solicitors, so monthly containers are small and large-scale analyses should plan for low record counts rather than dense panels.

Who Files Form DEFN14A and What Triggers the Filing

The filer is the soliciting party itself, not the issuer — an outside party soliciting the registrant's shareholders, typically against management's slate, recommendation, or transaction. This is the mirror of DEF 14A, which is filed by the registrant for management's own solicitation. The "N" suffix is the operative marker that the filer is non-management; this dataset covers only that variant.

Who files DEFN14A

Common populations of filers include:

• Activist funds and hedge funds running a board-seat contest at a portfolio company.
• Dissident shareholders opposing a management-sponsored merger, sale, or recapitalization.
• Insurgent groups advancing an alternative director slate.
• Strategic or financial bidders in hostile or competing M&A situations.
• Founders, former executives, or large individual holders seeking to replace directors.
• Ad hoc shareholder committees, ESOP trustees, or labor-affiliated funds running independent governance, compensation, or social campaigns.
• Section 13(d) groups filing jointly through a lead member.

Schedule 14A, Item 4 and Instruction 3 to Item 4 require the filing to identify every "participant" in the solicitation (affiliates, nominees, certain employees and advisers). Participants are disclosed in the filing but are distinct from the EDGAR filer of record, which is the soliciting entity (or group lead).

What regulatory regime governs it

DEFN14A is compelled by Section 14(a) of the Exchange Act, which prohibits any person from soliciting proxies in respect of Section 12 registered securities except in compliance with SEC rules. The operative rules:

• Rule 14a-3 — no solicitation may occur unless each person solicited is furnished a publicly filed proxy statement meeting Schedule 14A.
• Rule 14a-6 — filing mechanics for preliminary and definitive proxy materials, including the 10-calendar-day preliminary review window and the routine-meeting carve-out from preliminary filing.
• Rule 14a-12 — permits soliciting communications before furnishing a definitive proxy statement, provided participants are identified and the materials are filed no later than the date of first use. Most non-management public campaigns begin under 14a-12 and culminate in the DEFN14A.
• Rule 14a-19 (universal proxy) — applies to non-exempt director-election contests at U.S. operating companies for meetings held after August 31, 2022. Requires a dissident to (i) deliver notice and nominee names to the registrant by the later of 60 calendar days before the anniversary of the prior annual meeting (or, where the meeting date has shifted by more than 30 days, the tenth calendar day after public announcement of the new meeting date), (ii) use a universal proxy card listing both slates, and (iii) solicit holders of at least 67% of the voting power.
• Schedule 14A — the substantive disclosure schedule: identity and background of participants, security holdings, interests in matters to be voted on, voting procedures, and (in contests) detailed Item 5(b) participant disclosures.

Rule 14a-2(b) exempt solicitations (no proxy authority sought) fall outside Schedule 14A and are filed on PX14A6G/PX14A6N rather than DEFN14A.

When it is filed (trigger and timing)

The trigger is first dissemination of definitive proxy soliciting materials by a non-management party. The filing must occur no later than the date the materials are first sent or given to security holders (Rule 14a-6(b)).

The underlying corporate event is almost always one of:

• a contested director election at an annual or special meeting,
• a vote on a merger, sale, or other extraordinary transaction,
• a shareholder proposal or governance change advanced independently of management, or
• a director-removal or written-consent solicitation outside the annual cycle.

Adjacent timing rules:

• 10-day preliminary window: Where Rule 14a-6(a) does not exempt the matter (i.e., it goes beyond uncontested election of directors, auditor ratification, and certain shareholder-proposal-only items), the soliciting party must first file a preliminary proxy statement PREN14A and wait at least 10 calendar days before filing the DEFN14A. Most director contests, merger fights, and charter amendments fall outside the carve-out.
• Rule 14a-19 60-day notice: In a U.S. operating-company director contest, the dissident must deliver advance notice to the registrant by the deadline above; the DEFN14A must then carry the universal-proxy disclosures and card design.
• Post-definitive changes: Substantive revisions are filed as DFRN14A; supplemental fight materials that do not amend the proxy statement are filed as DFAN14A through the meeting date.

Why and how the obligation arises

Section 14(a) makes the obligation strict-liability in form: any person other than the issuer who furnishes a proxy card to holders of Section 12 securities is "soliciting" and must comply with Schedule 14A. The DEFN14A is therefore not optional once the soliciting party crosses from 14a-12 communications into actually distributing a proxy card. The "N" filing exists because a non-management solicitor cannot piggyback on the issuer's DEF 14A; it must produce its own complete Schedule 14A disclosure, on its own filer identity, before or contemporaneously with mailing.

Relationship to surrounding non-management forms

The DEFN14A sits in a small family of non-management Schedule 14A submission types, distinguished by stage and amendment status:

• PREN14A — preliminary non-management proxy statement; required at least 10 calendar days before the DEFN14A unless the matter qualifies for the Rule 14a-6(a) routine-meeting exemption.
• DEFN14A — the definitive non-management proxy statement; filed at or before first dissemination.
• DFRN14A — revised definitive non-management proxy statement; functionally an amendment to a prior DEFN14A after first dissemination.
• DFAN14A — additional definitive soliciting material from the non-management party (press releases, fight letters, investor decks, supplemental mailings) filed after the DEFN14A.
• PX14A6G/PX14A6N — Rule 14a-2(b) exempt solicitations where no proxy authority is sought; full Schedule 14A is not used and these are not in this dataset.

A complete contest record on EDGAR typically reads as: 14a-12 communications (often filed as DFAN14A) → PREN14A (where required) → DEFN14A (this dataset) → optional DFRN14A revisions → a stream of DFAN14A supplemental materials through the meeting.

Important distinctions and edge cases

• DEF 14A vs DEFN14A: same meeting, two filers. DEF 14A comes from the issuer; DEFN14A comes from the dissident. A single contested meeting commonly produces both.
• Filer of record vs participants: the EDGAR filer is the lead soliciting entity (or 13(d) group lead). All other participants — affiliates, director nominees, advisers — are disclosed inside the filing under Item 5(b) but are not separate EDGAR filers of the DEFN14A.
• Foreign private issuers: FPIs are exempt from Section 14(a) under Rule 3a12-3(b). Contests at FPIs do not generally produce DEFN14A filings and are outside this dataset.
• Registered investment companies: closed-end funds, BDCs, and other RICs are subject to Section 14(a), so non-management solicitations at these issuers do appear here — a notable subpopulation given persistent activist activity in closed-end funds.
• Exempt solicitations: Rule 14a-2(b) campaigns appear on PX14A6G/PX14A6N, not DEFN14A.
• EDGAR coverage: mandatory electronic filing for most domestic registrants and their proxy contestants phased in under Regulation S-T from 1993 and reached substantially full coverage by mid-1996; the dataset's earliest sample date of March 1995 reflects the start of meaningful EDGAR coverage of this form type, not the start of the underlying Section 14(a) obligation, which dates to 1934.

Content Structure of a Single DEFN14A Record

One record is a single EDGAR DEFN14A submission, identified by its accession number and materialized on disk as a self-contained accession folder. The folder bundles a synthesized metadata.json with the original EDGAR documents that constitute the definitive proxy statement filed by a non-management party — the proxy statement itself, any additional soliciting materials, and any narrative or tabular exhibits — minus image attachments and the consolidated complete-submission .txt wrapper. The unit is the filing, not the document, the proposal, or the participant. One accession number maps to one folder; one folder is one record.

Container and folder layout

Accession folders are named in the flat 18-digit form of the accession number with no dashes (for example 000110465925104369), even though the dashed form 0001104659-25-104369 appears inside metadata.json. Joining records to other EDGAR datasets therefore requires choosing the appropriate representation: the folder name strips dashes, the metadata preserves them.

Inside each accession folder there is always a metadata.json file alongside the original EDGAR documents from the submission, with two categories of files deliberately removed:

• Image files. GRAPHIC entries with .jpg, .gif, or .png extensions are stripped, even though they remain enumerated in documentFormatFiles.
• The consolidated <accession>.txt wrapper. EDGAR's umbrella complete-submission text file, which concatenates every component document of a submission into one stream, is omitted because the per-document SGML envelope on each individual file already preserves the same metadata at finer granularity.

In practice, modern records are dominated by HTM/HTML for the proxy statement and exhibits plus the synthesized metadata.json. PDFs surface when a soliciting party attaches scanned letters, glossy mailers, or supplemental decks. Standalone TXT documents appear in older filings and as occasional press-release exhibits. The original filer's filenames are preserved (for example tm2528518-4_defn14a.htm); only metadata.json is a synthesized constant filename.

metadata.json contents

metadata.json is a single flat JSON object describing the submission at the filing level.

Top-level scalar and link fields:

• formType — always "DEFN14A" for this dataset.
• accessionNo — dashed accession number, e.g. "0001104659-25-104369".
• filedAt — full ISO 8601 timestamp with timezone offset (e.g. "2025-10-30T16:05:51-04:00"), capturing the EDGAR acceptance time rather than just the calendar date.
• periodOfReport — typically the date of the shareholder meeting that the proxy concerns, which is later than filedAt and should not be confused with it.
• description — short human-readable label, typically "Form DEFN14A - Definitive proxy statement filed by non-management".
• linkToFilingDetails, linkToTxt, linkToHtml — EDGAR Archives URLs pointing at the primary document, the full-submission text file, and the *-index.htm filing index page respectively.
• linkToXbrl — present in the schema but typically empty for DEFN14A.
• id — a 32-character internal hexadecimal identifier for the record.
• seriesAndClassesContractsInformation, dataFiles — array fields that exist in the schema but are usually empty for proxy filings.

The documentFormatFiles array contains one entry per document in the original EDGAR submission, including a trailing entry for the complete-submission text file (whose sequence and type fields are a single space " "). Each entry carries:

• sequence — a string preserving the EDGAR column verbatim ("1", "2", …, or " " for the trailing umbrella entry).
• size — a string containing a byte count.
• documentUrl — EDGAR Archives URL of the document.
• description — for example "DEFN14A", "GRAPHIC", "EX-99.1", or "Complete submission text file".
• type — the EDGAR document type, mirroring description for most entries.

Because this array reflects the original submission, it enumerates the image and .txt files that have been excluded from the ZIP. Diffing the array against the folder contents identifies exactly which artifacts are absent for any given record, providing a deterministic audit of what was stripped.

The entities array carries one object per filing entity. The two canonical roles for DEFN14A are encoded inside companyName via the suffixes (Subject) and (Filed by) rather than a dedicated role field:

• The (Subject) entity is the registrant whose securities the proxy concerns.
• The (Filed by) entity is the non-management soliciting party.

Subject entities carry the richer identifier set — cik, companyName, type, fiscalYearEnd, stateOfIncorporation, act (typically "34"), fileNo, irsNo, sic (with both code and label, e.g. "6792 Oil Royalty Traders"), filmNo, and tickers (an array of trading symbols). Filer entities are leaner, typically exposing only cik, companyName, type, fiscalYearEnd, stateOfIncorporation, and irsNo. Every record is expected to surface at least one of each role, and contests with multi-member dissident groups produce multiple (Filed by) entries — consumers extracting "the soliciting party" must accept a one-to-many relationship.

The primary proxy statement document

The primary document — the one with sequence: "1" and type: "DEFN14A" in documentFormatFiles — is the proxy statement itself. Its filename follows the filer's own naming convention (for example tm2528518-4_defn14a.htm).

The file is not a plain HTML document. It is wrapped in EDGAR's per-document SGML envelope:

          1
          <DOCUMENT>
        

          2
          <TYPE>DEFN14A
        

          3
          <SEQUENCE>1
        

          4
          <FILENAME>tm2528518-4_defn14a.htm
        

          5
          <DESCRIPTION>DEFN14A
        

          6
          <TEXT>
        

          7
          <html>...</html>
        

          8
          </TEXT>
        

          9
          </DOCUMENT>
        

Downstream consumers must strip this <DOCUMENT>/<TYPE>/<SEQUENCE>/<FILENAME>/<DESCRIPTION>/<TEXT> wrapper (and matching close tags) before feeding the body to an HTML parser. Inside the HTML, layout typically relies on pixel- or point-sized inline styles rather than semantic heading tags or CSS classes; section breaks are commonly implemented via page-break-after:always divs, one per printed page, so the document carries an implicit pagination that mirrors the mailed booklet. References to the excluded GRAPHIC files are preserved as <img> tags whose alt attribute is rewritten to "[MISSING IMAGE: <filename>]", providing a deliberate, machine-detectable marker that an image was present in the original EDGAR submission but is absent from the dataset (rather than missing due to corruption).

The narrative content follows the disclosure architecture mandated by Schedule 14A for a non-management solicitation. The major sections, in roughly the order they appear, are:

• Cover sheet. The Schedule 14A cover, identifying the SEC, the schedule, the registrant whose securities are involved, and the soliciting party. Includes the standard fee-table check boxes prescribed by Rule 14a-6.
• Notice of meeting. A formal block such as SPECIAL MEETING OF HOLDERS OF UNITS OF BENEFICIAL INTEREST or ANNUAL MEETING OF SHAREHOLDERS, stating the meeting type, date, time, location, and record date.
• Proxy statement opening. A PROXY STATEMENT heading followed by a question-and-answer or narrative introduction explaining who is soliciting the proxy, what is being asked, and why.
• Background and reasons for the solicitation. The narrative core of a contested filing — a chronological account of interactions between the soliciting party and the registrant, the dispute giving rise to the contest, and the soliciting party's case for change.
• Description of the proposals. Item-by-item discussion of each matter to be voted on: typically election of an alternative slate of directors or trustees, but also bylaw amendments, removal of incumbent directors, advisory votes, or merger/transaction-related proposals.
• Information about the participants. Disclosure required by Schedule 14A Items 4 and 5 for each "participant" in the solicitation — biographical information, ownership of the registrant's securities, transactions in those securities over the preceding two years, contracts and arrangements with respect to the securities, and any other interest in the matters to be acted upon.
• Security ownership tables. Beneficial-ownership disclosure for 5%-plus holders, directors, and named executive officers of the registrant, typically rendered as HTML tables.
• Voting and proxy procedures. Mechanics of voting: record date, quorum, vote required for each proposal, revocation of proxies, treatment of abstentions and broker non-votes, and solicitation costs.
• Other matters / additional information. Catch-all sections covering matters not separately itemized, cross-references to the registrant's own filings (often incorporated by reference), where to obtain copies, and householding notices.
• Form of proxy / voting card. A closing block reproducing the proxy card itself, with the typical Vote by Internet, Vote by Telephone, and Vote by Mail instruction columns.

Signature blocks and dates are rendered as plain HTML text rather than as structured fields. Beneficial-ownership and participant tables are HTML — not structured data — and require table-aware extraction.

Additional documents in the record

When the underlying EDGAR submission contains exhibits or additional soliciting materials, they appear as additional files in the accession folder, each carrying its own SGML envelope and an EDGAR type such as EX-99.1, EX-99.2, etc. Common attachments include letters to shareholders, white papers, press releases, presentation decks (occasionally as PDFs), consent forms for nominee directors, and biographical exhibits for the dissident slate. Each such file is enumerated as a separate documentFormatFiles entry with its own sequence number. Image files referenced by these exhibits are subject to the same exclusion rule as the primary document.

What is included in the dataset record

Each record contains the synthesized metadata.json, the SGML-wrapped primary DEFN14A document, and any non-image, non-.txt-umbrella documents that were part of the original EDGAR submission — exhibits, additional soliciting materials, narrative attachments, and PDFs.

What is excluded

Three categories of content are omitted:

• Image files. GRAPHIC documents (logos, scanned proxy-card images, photographs, charts saved as bitmaps) are not included even though they are enumerated in documentFormatFiles. References to these images are retained inside the HTML as <img> tags with alt="[MISSING IMAGE: ...]" markers.
• The complete-submission .txt wrapper. The umbrella <accession>.txt file that EDGAR uses to concatenate every document of a submission is omitted because the per-document SGML wrapper on each individual file already preserves the same metadata at finer granularity.
• XBRL data files. DEFN14A is not subject to XBRL tagging, so dataFiles and linkToXbrl are typically empty; there are no instance documents, schemas, or label linkbases to include.

Coverage and historical evolution of content

The dataset spans EDGAR DEFN14A filings from March 1995 to the present. Schedule 14A's substantive disclosure framework has been stable across that period — the participant-information, ownership, voting-mechanics, and form-of-proxy backbone visible in current filings is essentially the same architecture imposed by the original Rule 14a-6 and the long-standing Items of Schedule 14A. Several rule changes have nonetheless modulated what appears inside a record:

• Rule 14a-12 (2000). The 1999/2000 proxy-rule amendments liberalized pre-definitive solicitation, increasing the volume of supplemental soliciting materials filed under separate form types but also shaping the cross-references that appear inside the DEFN14A's "additional information" section.
• E-proxy / notice-and-access (2007). The Rule 14a-16 notice-and-access regime altered the form-of-proxy and voting-mechanics sections, adding internet-availability notices and online-voting instructions; the Vote by Internet block in modern DEFN14A filings reflects this rule.
• Universal proxy (Rule 14a-19, effective for meetings after August 31, 2022). For contested director elections, dissident filers must include a universal proxy card listing both the registrant's and the soliciting party's nominees and must comply with notice and consent requirements for nominees. Records filed for director-contest meetings after that date include the universal-card layout and the additional Rule 14a-19 representations; pre-2022 contests show the older one-sided dissident-only card.
• Pay-versus-performance, clawback, and other registrant-side rules. Most recent registrant-disclosure rule changes (pay ratio, pay-versus-performance, Rule 10D-1 clawback policies, climate disclosure) primarily affect management's DEF 14A rather than DEFN14A, so they appear in DEFN14A records mainly through references to, or critiques of, the registrant's own proxy.

Coverage and historical evolution of format

EDGAR mandated electronic filing for proxy statements during the phase-in period that culminated in 1996, so the earliest records (1995–1996) are predominantly plain ASCII text wrapped in SGML, with hand-laid-out tables built from spaces and dashes. Through the late 1990s and early 2000s, filings transitioned to HTML while retaining the per-document SGML envelope; tables migrated to <table> markup, and proxy cards became HTML grids. From the mid-2000s onward, HTML with extensive inline styles and page-break-after:always div pagination became the dominant form, often produced by financial-printer toolchains, and remains the modern default. PDF attachments — primarily scanned letters and graphical mailers — appear sporadically across the entire era. Across this evolution, the per-document SGML envelope is the one structural constant.

Interpretation notes

Several nuances matter for downstream use:

• SGML stripping is mandatory. Every primary and exhibit document is wrapped in the per-document EDGAR SGML envelope; HTML parsers must strip it.
• documentFormatFiles is the authoritative inventory. It will list image files and the umbrella .txt that are intentionally absent from the folder, so diffing the array against the directory listing identifies the omitted artifacts deterministically.
• Roles live in the company-name suffix. Identifying the subject registrant versus the soliciting party requires parsing (Subject) and (Filed by) out of companyName rather than reading a structured role field. Multi-member dissident groups produce multiple (Filed by) entries.
• periodOfReport is the meeting date, not the filing date — chronologically later than filedAt.
• [MISSING IMAGE: …] is a deliberate marker. It signals an intentional dataset-level exclusion, not data corruption, and lists the original GRAPHIC filename so it can be reconciled with documentFormatFiles.
• Two accession-number forms coexist. The folder name is the flat 18-digit form; metadata.json.accessionNo is the dashed form. Cross-dataset joins must normalize.

How This Dataset Differs From Similar Datasets or Filings

DEFN14A is the definitive proxy statement filed by a non-management soliciting party (activist, dissident slate, or other third party) for a contested or independent solicitation under Section 14(a). The closest comparisons sit in the same proxy family, in exempt-solicitation filings, and in two non-proxy disclosures common to activist campaigns.

Closest related filing types

DEF 14A (issuer definitive proxy). Same statute and "definitive" status, but filed by the registrant's management for its own slate and proposals at every meeting. DEFN14A appears only in contests; its content mirrors DEF 14A in structure but presents the challenger's narrative, nominees, and case against management.

PREN14A (non-management preliminary proxy). Same filer and same campaign as DEFN14A, but the earlier preliminary version subject to SEC staff review. PREN14A may contain blanks or pre-comment language; DEFN14A is the final version actually mailed and voted on.

DFAN14A (additional non-management soliciting materials). Same soliciting party, filed after DEFN14A, but contains supplemental campaign communications (press releases, decks, shareholder letters, ads). DFAN14A is filed repeatedly throughout a contest; DEFN14A is the single canonical proxy statement it supports.

DFRN14A (revised non-management definitive proxy). A revised/amended DEFN14A by the same filer, triggered when material changes (revised slate, corrected disclosures, settlement) require re-circulation. Treat DEFN14A plus any DFRN14A as a versioned pair.

PX14A6G / PX14A6N (Rule 14a-2(b) exempt solicitations). Filed by shareholders soliciting votes without seeking proxy authority; PX14A6G is the written communication, PX14A6N the notice of intent. No proxy card, no full proxy statement, far lighter disclosure. DEFN14A applies when the campaign is a formal contest with a competing card; PX14A6G/N applies to "vote no" or recommendation campaigns.

Schedule 13D. Triggered by crossing 5% beneficial ownership with non-passive intent; discloses stake and Item 4 purpose, not voting matters, nominees, or a proxy card. Covers the ownership and intent phase that often precedes a DEFN14A solicitation.

Form 8-K Item 5.07. Filed by the issuer after the meeting to report actual vote tallies. DEFN14A discloses what the dissident proposed; Item 5.07 is the only authoritative source for whether it passed.

Boundary summary

DEFN14A is uniquely (a) definitive not preliminary, (b) non-management not issuer-filed, (c) a full proxy statement not exempt or supplemental material, and (d) the operative solicitation document itself, not an ownership filing or post-vote report. No adjacent dataset substitutes: DEF 14A is the opposing side, PREN14A the earlier draft, DFAN14A the ancillary messaging, DFRN14A the later revision, PX14A6G/N the lighter exempt route, Schedule 13D the stake-building stage, and 8-K Item 5.07 the outcome.

Who Uses This Dataset

DEFN14A filings are definitive proxy statements from non-management soliciting parties — typically activists, dissidents, and insurgents running campaigns against incumbent boards. The dataset serves a narrow but high-stakes audience that pulls distinct sections of each filing.

Proxy contest defense counsel

M&A and shareholder-activism partners at corporate law firms mine the "Background of the Solicitation" chronology, "Reasons for the Solicitation" rhetoric, and Schedule 14A Item 5 participant disclosures to anticipate dissident framing. Output: defensive proxy statements, advance-notice bylaw memos, universal-proxy response strategies, and disclosure-litigation pleadings.

Proxy solicitors

Solicitation firms on both sides benchmark voting-instruction language, dissident proxy-card layouts, and "Solicitation of Proxies" cost and method disclosures. Output: solicitation budgets, retail-versus-institutional turnout forecasts, and counter-solicitation scripts.

Activist investor research analysts

Analysts at activist hedge funds, event-driven funds, and family offices study dissident thesis sections, nominee biographies, and exhibit-level financial analyses to calibrate their own campaigns. Output: messaging architecture, target screens by issuer profile, and disclosure templates calibrated to proxy-advisor expectations.

Proxy advisory firm analysts

Analysts at major proxy advisors read DEFN14A filings — paired with the issuer's DEFA14A and PREC14A — to form contested-election recommendations. They focus on the case for change, nominee qualifications, soliciting-group economic interest, and analytical exhibits. Output: vote recommendations applying published frameworks for change-in-control, short-slate, and governance-only contests.

Stewardship teams at institutional investors

Stewardship and proxy-voting teams at large asset managers, public pension plans, and sovereign funds read filings independently of third-party recommendations. They weigh strategic arguments, nominee independence, and dissident funding/compensation arrangements. Output: vote decisions, rationale memos, published voting bulletins, and post-season stewardship disclosures.

Corporate development, IR, and activism-defense teams at issuers

In-house corporate development, IR, and corporate-secretary functions use the archive to model their own threat surface before a campaign starts. They study thesis structures, recurring nominees, and the escalation arc in "Background of the Solicitation" narratives. Output: vulnerability assessments, board education decks, engagement plans, and rapid-response playbooks triggered by a 13D or activist letter.

Investment bank M&A and shareholder-advisory bankers

Bankers in M&A, ECM, and dedicated shareholder-advisory groups draw on soliciting parties' economic interest, derivative positions, and proposed strategic alternatives (sale, spin-off, recap) when advising boards on contests or preemptive structural defenses. Output: activism-defense pitches, board presentations on contest dynamics, and fairness analyses when a contest converts into a transaction.

Academic researchers in corporate finance, law, and governance

Finance, accounting, and corporate-law researchers use the archive as one of the few comprehensive public sources on non-management solicitations. They extract participant disclosures, proposal types, and "Reasons for the Solicitation" text. Output: peer-reviewed work on activism returns, board-composition outcomes, settlement dynamics, and the diffusion of universal-proxy practices.

Financial journalists covering proxy fights

Reporters at business newswires, national newspapers, and corporate-governance publications use DEFN14A filings as primary source material during active campaigns. They cite the dissident narrative, nominee slates, economic-interest disclosures, and quotable exhibits (investor letters, peer-comparison decks). Output: breaking coverage, season-end recaps, and investigative pieces tracing recurring activists and their nominee networks.

LLM and AI teams working with adversarial governance text

ML engineers building retrieval, summarization, and classification systems for governance and activism use the corpus as adversarial, highly structured, multi-decade training data. Output: contested-election summarizers, classifiers separating governance-only from change-in-control campaigns, stewardship retrieval systems, and evaluation sets for argumentative legal-financial prose absent from management-side filings.

Specific Use Cases

The DEFN14A archive supports a small set of high-value, contest-specific workflows that depend on different sections of the proxy statement and different fields of metadata.json. The use cases below are the most concrete projects the dataset is built to power.

Proxy contest outcome studies

Build a panel that joins each DEFN14A record to the corresponding Form 8-K Item 5.07 vote tally for the same (Subject) CIK and meeting date. Use accessionNo, periodOfReport (the meeting date), and the (Subject) entity's cik to assemble dissident-versus-management win rates by proposal type, slate size, and SIC. The "Description of the Proposals" and "Form of Proxy" sections classify what was on the ballot; the joined 8-K supplies whether it passed. Output: regressions of contest outcomes on issuer characteristics, dissident ownership stakes, and slate composition.

Activist campaign tactic analysis

Group records by (Filed by) CIK to assemble per-activist campaign histories, then segment the "Background of the Solicitation" and "Reasons for the Solicitation" sections to extract recurring tactical patterns — initial private engagement, escalation letters, board-seat demands, sale-process pressure, public-relations campaigns. Layer in any DFRN14A revisions and DFAN14A supplementals retrieved from sibling datasets to reconstruct the full timeline. Output: tactic taxonomies, escalation-arc models, and benchmarks for predicting whether a campaign settles, goes to a vote, or is withdrawn.

Universal-proxy adoption tracking

Filter records with periodOfReport after August 31, 2022 and parse the proxy-card block at the end of the primary document for the joint registrant-and-dissident nominee layout introduced by Rule 14a-19. Detect the Rule 14a-19 representation language and notice/consent disclosures in the participant section. Compare against pre-2022 single-sided cards to measure compliance, "round up" voting patterns, and the share of contests that drop or shrink slates after universal-proxy notice deadlines. Output: regulatory-adoption dashboards, slate-size trend analyses, and inputs to SEC rulemaking comment letters.

Proxy solicitor market mapping

Extract the "Solicitation of Proxies" subsection from each DEFN14A to capture the named solicitation firm, fee, expected number of solicitors, and method mix (mail, phone, internet). Aggregate by solicitor name across years and (Subject) SIC to map market share, fee inflation, and which solicitors specialize in which types of contests (short-slate governance versus change-in-control). Output: competitive briefings for solicitation firms, fee benchmarks for issuers selecting a defense solicitor, and league tables for trade press.

Nominee director background extraction

Parse the "Information about the Participants" section and any nominee-biography exhibits (often EX-99.x HTML or PDF attachments) to extract each dissident nominee's name, age, prior board seats, employment history, and stated independence. Link nominees who recur across multiple (Filed by) campaigns to surface the professional dissident-director network. Output: a structured nominee database, network graphs of activist-aligned directors, and screening inputs for issuers vetting incoming dissident candidates.

Argumentation and rhetoric corpus for LLM training

Treat the "Reasons for the Solicitation," "Background of the Solicitation," and shareholder-letter exhibits as an adversarial-governance text corpus paired by (Subject) CIK and meeting date with the issuer's DEF 14A and DEFA14A pushback. The result is a labeled, position-tagged dataset of argument-counterargument pairs across three decades. Output: fine-tuning data for contested-election summarizers, classifiers separating governance-only from change-in-control rhetoric, stance-detection benchmarks, and retrieval evaluation sets for activism-aware assistants.

Beneficial-ownership and economic-interest mining

Extract the security-ownership tables and the participant-level transaction tables (two-year purchase-and-sale history mandated by Schedule 14A Item 5) to reconstruct each soliciting group's stake-building pattern, derivative exposure, and total economic interest at the time of mailing. Cross-reference with the corresponding Schedule 13D from the same (Filed by) CIK to validate disclosed positions. Output: activist-stake league tables, synthetic-exposure detection (swaps and options disclosed only in the proxy footnotes), and inputs to "wolf pack" coordination studies.

Settlement and withdrawal detection

Identify campaigns that ended without a vote by joining DEFN14A records to subsequent DFRN14A revisions, DFAN14A withdrawal announcements, and issuer 8-K Item 5.02 board-appointment filings within the window between filedAt and periodOfReport. Records whose meeting passes with no matching 8-K vote tally, paired with a director-appointment 8-K naming a (Filed by)-affiliated nominee, mark settled contests. Output: settlement-rate time series, board-seat-yield metrics per activist, and event windows for return studies on settlement announcements.

Dataset Access

Dataset Index JSON API: https://api.sec-api.io/datasets/form-defn14a-files.json

This endpoint returns metadata describing the Form DEFN14A Files Dataset and the list of monthly ZIP containers that make up the archive. The response includes the dataset name, description, last update timestamp, earliest sample date (1995-03-01), total records, total size, covered form types (DEFN14A), container format (ZIP), and content file types (TXT, JSON, HTML, PDF). It also includes the full dataset download URL and a containers array with per-month entries listing each container's key, size, record count, last updated timestamp, and download URL. This index can be polled to detect which monthly containers were modified in the latest refresh run, so only changed containers need to be downloaded again.

This endpoint does not require an API key.

          1
          {
        

          2
            "datasetId": "1f13365b-9ae0-69bc-9e00-7f3828b4f86f",
        

          3
            "datasetDownloadUrl": "https://api.sec-api.io/datasets/form-defn14a-files.zip",
        

          4
            "name": "Form DEFN14A Files Dataset",
        

          5
            "updatedAt": "2026-04-15T12:16:24.383Z",
        

          6
            "earliestSampleDate": "1995-03-01",
        

          7
            "totalRecords": 335,
        

          8
            "totalSize": 8488220,
        

          9
            "formTypes": ["DEFN14A"],
        

          10
            "containerFormat": "ZIP",
        

          11
            "fileTypes": ["TXT", "JSON", "HTML", "PDF"],
        

          12
            "containers": [
        

          13
              {
        

          14
                "downloadUrl": "https://api.sec-api.io/datasets/form-defn14a-files/2026/2026-03.zip",
        

          15
                "key": "2026/2026-03.zip",
        

          16
                "size": 124583,
        

          17
                "records": 4,
        

          18
                "updatedAt": "2026-04-15T12:16:24.383Z"
        

          19
              }
        

          20
            ]
        

          21
          }
        

Example curl request:

          1
          curl -O https://api.sec-api.io/datasets/form-defn14a-files.json
        

Download Entire Dataset: https://api.sec-api.io/datasets/form-defn14a-files.zip?token=YOUR_API_KEY

The full dataset is available as a single ZIP archive. Given the dataset's small footprint, downloading the complete archive in one request is practical for most workflows. This endpoint requires an API key passed via the token query parameter.

          1
          curl -O "https://api.sec-api.io/datasets/form-defn14a-files.zip?token=YOUR_API_KEY"
        

Download Single Container: https://api.sec-api.io/datasets/form-defn14a-files/2026/2026-03.zip?token=YOUR_API_KEY

Individual monthly container ZIPs can be downloaded directly using the path pattern <YYYY>/<YYYY>-<MM>.zip. This is useful for incremental updates where only recently changed months need to be retrieved. This endpoint requires an API key.

          1
          curl -O "https://api.sec-api.io/datasets/form-defn14a-files/2026/2026-03.zip?token=YOUR_API_KEY"
        

Frequently Asked Questions

What form does this dataset cover?

The dataset covers Form DEFN14A — a definitive proxy statement filed under Section 14(a) of the Securities Exchange Act of 1934 and Schedule 14A by a non-management soliciting party. The "N" suffix is the operative marker that the filer is not the registrant; management's mirror filing is DEF 14A.

What does one record in this dataset represent?

One record is a single EDGAR DEFN14A submission, identified by its accession number and materialized on disk as a self-contained accession folder. Each folder bundles a synthesized metadata.json with the original EDGAR documents — the primary proxy statement plus any exhibits or additional soliciting materials — minus image attachments and the consolidated complete-submission .txt wrapper.

Who is required to file Form DEFN14A?

The filer is the soliciting party itself, not the issuer. Typical filers include activist hedge funds, dissident shareholder slates, insurgent groups advancing alternative directors, strategic or financial bidders in hostile M&A, large individual holders, ad hoc shareholder committees, ESOP trustees, labor-affiliated funds, and Section 13(d) groups filing through a lead member. Foreign private issuers are exempt under Rule 3a12-3(b), so contests at FPIs do not generally produce DEFN14A filings.

When must a DEFN14A be filed?

The filing must occur no later than the date the definitive proxy materials are first sent or given to security holders, under Rule 14a-6(b). Where Rule 14a-6(a) does not exempt the matter, the soliciting party must first file a PREN14A and wait at least 10 calendar days before filing the DEFN14A. For director contests at U.S. operating companies with meetings after August 31, 2022, the dissident must also satisfy Rule 14a-19 universal-proxy notice requirements.

What time period does the dataset cover?

The dataset spans EDGAR DEFN14A filings from March 1995 — the start of meaningful EDGAR coverage of this form type — to the present. Schedule 14A's substantive disclosure framework has been stable across that period, though rule changes including notice-and-access (2007) and the universal-proxy regime (2022) have modulated the form-of-proxy and participant disclosures inside individual records.

What file formats are inside the dataset?

The dataset is delivered as a hierarchy of monthly ZIP containers. File types found inside the records are TXT, JSON, HTML, and PDF: HTML for the modern primary proxy statement and exhibits, JSON for the synthesized metadata.json, PDF for scanned letters and graphical mailers, and plain ASCII TXT for older 1995–1996 filings. Every primary and exhibit document is wrapped in EDGAR's per-document SGML envelope, which downstream parsers must strip before feeding the body to an HTML parser.

How does DEFN14A differ from DEF 14A?

Both are definitive proxy statements under Section 14(a), but they are filed by opposite sides of a contest. DEF 14A is filed by the registrant's management for its own slate and proposals at every meeting; DEFN14A is filed only in contests by a non-management soliciting party advancing a competing slate, opposing a transaction, or pursuing an independent shareholder proposal. A single contested meeting commonly produces both filings.

How does DEFN14A differ from PREN14A and DFAN14A?

PREN14A is the preliminary version of the same filer's proxy statement, subject to SEC staff review and filed at least 10 calendar days before the DEFN14A unless the matter qualifies for the Rule 14a-6(a) routine-meeting exemption. DFAN14A is additional definitive soliciting material — press releases, fight letters, investor decks — filed by the same non-management party after the DEFN14A and through the meeting date. DEFN14A is the single canonical proxy statement; PREN14A is its earlier draft and DFAN14A is the supplemental campaign messaging that surrounds it.