The Form SF-1 Files Dataset packages every EDGAR submission of Form SF-1 and its pre-effective amendment SF-1/A — the general-purpose, non-shelf registration statement filed by asset-backed securities issuers under the Securities Act of 1933 and prescribed by 17 CFR 239.44. Each record corresponds to a single accession number and contains the full HTML prospectus, every non-image exhibit (underwriting agreement, indenture, counsel opinions, charter documents, material contracts, consents, powers of attorney, and where applicable Reg AB II asset-data exhibits), and a metadata.json capturing the EDGAR header and dual-filer (depositor and sponsor) entity block. Coverage begins with the earliest SF-1 accessions in January 2016 — the operative window of Regulation AB II — and continues through the present, distributed as monthly ZIP containers of HTML and JSON. Buy-side ABS analysts, rating-agency teams, securitization counsel, structuring desks, Reg AB II compliance groups, RegTech engineers, and academic researchers use the dataset to read prospectus content, run precedent searches across legality and tax opinions, reconstruct registration histories from SF-1/A amendment chains, and train extraction pipelines on a stable, formulaic securitization corpus.
Programmatically retrieve the full list of dataset archive files, download URLs and dataset metadata.
Dataset Index JSON API
Download the entire dataset as a single archive file.
Download Entire Dataset:
Download a single container file (e.g. monthly archive) from the dataset.
Download Single Container:
Form SF-1 is the general ABS registration form created by the SEC's 2014 overhaul of asset-backed securities regulation, adopted in Release No. 33-9638 ("Asset-Backed Securities Disclosure and Registration") and effective in late 2014. The first SF-1 and Form SF-1/A accessions appear in EDGAR from January 2016 onward, which defines the lower bound of the dataset. SF-1 replaced the prior practice of registering ABS offerings on Form S-1 by introducing an ABS-specific form whose Items, prospectus content requirements, and exhibit table are tailored to securitization structures. The dataset captures every SF-1 and every SF-1/A pre-effective amendment filed on EDGAR over that window, including the full registration statement and exhibits filed with each accession.
SF-1 is used when the offering is not eligible for shelf registration on Form SF-3, which itself requires, among other conditions, a CEO certification, an asset-review provision under Rule 15Ga-2, an asset-representations-and-warranties dispute mechanism, and the investor-communication provisions imposed by Regulation AB II. Pre-effective amendments to an SF-1 are filed as SF-1/A; both form types share the same content architecture and both appear in the dataset.
Structurally, an SF-1 is a Securities Act registration statement built around a prospectus that describes the asset-backed securities being offered and the underlying asset pool. Its disclosure content is dictated principally by Regulation S-K Items 1100 through 1125 (the Regulation AB and Reg AB II disclosure regime), supplemented by the general Securities Act registration requirements of Form S-1 and Items 501–512 of Regulation S-K. Compared with a corporate S-1, the SF-1 places far more weight on the asset pool, the cash-flow waterfall, the transaction parties, and the legal structure of the special-purpose vehicle, and far less weight on operating-company financial statements. The dataset is distributed as monthly ZIP containers; the file types inside are HTML and JSON.
A defining feature of SF-1 is the dual-filer structure. The submission is made jointly by an operating-company sponsor and the single-purpose depositor / issuing entity that will issue the securities. Both appear as co-registrants under a paired Commission file-number scheme — typically 333-XXXXXX for the parent registration and 333-XXXXXX-01 for the linked co-filer. Both are listed on the cover page and both appear as (Filer) entries inside entities in the record's metadata.
One record corresponds to a single EDGAR submission of Form SF-1 or its pre-effective amendment SF-1/A — i.e., one accession number filed by an asset-backed securities issuer under the Securities Act of 1933 registration regime codified at 17 CFR 239.44. Physically, the record is a per-accession folder placed directly under a YYYY-MM/ top-level directory inside a monthly ZIP container. The folder name is the raw 18-digit EDGAR accession number with dashes stripped (e.g. 000119312525121765). Inside the folder are exactly one metadata.json describing the submission header plus every non-image document that was part of the original EDGAR submission, each preserved as the SGML-wrapped HTML file EDGAR stores natively. The unit of the record is therefore the filing (all documents that share an accession), not an individual document; a typical SF-1 accession resolves to roughly ten to forty HTML files plus the metadata sidecar.
A record is organized as three nested layers of content:
metadata.json, a structured representation of the EDGAR SGML header (form type, accession number, filing timestamp, registrants, document inventory, related-party block, and provider-assigned identifier).<DOCUMENT> envelope carrying <TYPE>, <SEQUENCE>, <FILENAME>, <DESCRIPTION>, and <TEXT> markers; the body is legacy HTML 4. The <TYPE> value mirrors the exhibit code (SF-1, SF-1/A, EX-1.1, EX-4.1, EX-5.1, EX-24.2, …) and matches the type field on the corresponding documentFormatFiles entry.metadata.json fieldsmetadata.json is a single JSON object inventorying the submission. The relevant fields are:
formType — "SF-1" or "SF-1/A".accessionNo — dash-formatted EDGAR accession (e.g. "0001193125-25-121765"); its dashless form is the folder name.filedAt — ISO-8601 timestamp with timezone offset, recording the EDGAR acceptance time.description — EDGAR descriptor such as "Form SF-1 - [Cover]" or "Form SF-1/A - [Amend]".linkToFilingDetails, linkToHtml, linkToTxt, linkToXbrl — URLs pointing, respectively, to the primary document on EDGAR, the filing-index HTML page, the combined SGML .txt submission, and the XBRL instance if any. linkToXbrl is generally empty for SF-1.documentFormatFiles — array of every document in the original EDGAR submission. Each entry carries sequence, description, documentUrl, type (the exhibit code), and size (bytes, as a string). The trailing entry is the combined .txt submission, recorded with a blank sequence and blank type. GRAPHIC entries remain in this list even though the binary image files themselves are omitted from the ZIP.entities — array of related parties, one entry per co-filer or related registrant. Per-entity fields include companyName, cik, fileNo, type, act, sic, stateOfIncorporation, fiscalYearEnd, and filmNo. For a typical SF-1 this array carries both the sponsor and the depositor / issuing entity, with their paired 333-XXXXXX and 333-XXXXXX-01 file numbers.dataFiles — array of XBRL or other structured data exhibits; usually empty for SF-1 and populated only when Reg AB II asset-level data exhibits (EX-102/103/104) are attached.seriesAndClassesContractsInformation — array surfaced for investment-company–style series/class structures; generally empty for SF-1.id — opaque 32-character hex identifier assigned by the data provider.<TYPE>SF-1 or <TYPE>SF-1/A)The primary document is the SF-1 or SF-1/A registration statement itself. It is by far the largest file in a record — frequently one megabyte or more of HTML — and is named under the EDGAR filer-agent convention d<jobId>dsf1.htm or d<jobId>dsf1a.htm. Its internal order is:
<TABLE> listing each co-registrant side-by-side. For each registrant the cover gives the exact name, state or jurisdiction of incorporation or organization, primary Standard Industrial Classification code, Commission file number under the paired 333-XXXXXX / 333-XXXXXX-01 scheme, IRS Employer Identification Number, principal executive office address and telephone number, and agent for service of process. The cover also carries the title and proposed maximum aggregate offering amount of the asset-backed securities being registered, the calculation-of-registration-fee table required by Rule 457, and the approximate date of commencement of the proposed sale.<FONT COLOR> markup); the table of contents; and the forward-looking-statements disclaimer.Operating-company financial statements as required by Regulation S-X are generally not present for the depositor / issuing entity, which is a newly formed bankruptcy-remote SPV with no operating history; financial information instead takes the form of pool-level statistical tables and, where required, sponsor-level financial disclosures by incorporation or summary.
Exhibits are filed as separate HTML files in the same accession folder and indexed by <TYPE> codes that match the SEC exhibit table for SF-1 (Item 601 of Regulation S-K as modified by Reg AB II). Each exhibit document opens with a right-aligned <B>Exhibit X.Y</B> heading followed by counsel letterhead, contract recitals, or a signature block. The typical exhibit set is:
Every .htm file in a record is wrapped in the legacy EDGAR <DOCUMENT> envelope. The preamble carries five header fields — <TYPE> (the exhibit code, identical to documentFormatFiles[].type), <SEQUENCE> (matching documentFormatFiles[].sequence), <FILENAME> (matching the on-disk filename), <DESCRIPTION> (a human-readable label), and <TEXT> (which opens the actual HTML body). The body itself is <HTML><HEAD><TITLE>...</TITLE></HEAD><BODY>...</BODY></HTML>, and the document closes with </TEXT></DOCUMENT>. The same wrapper applies uniformly to the primary registration statement and to every exhibit, so extraction tooling can rely on a consistent SGML boundary regardless of exhibit type.
Document filenames follow EDGAR's filer-agent convention d<jobId>d<exhibitTag>.htm, where <jobId> is a numeric job identifier shared by every document in a single accession and <exhibitTag> is a short code encoding the exhibit (sf1, sf1a, ex11, ex33, ex41, ex51, ex81, ex82, ex101, ex242, ex992, ex993, and so on). Image companions, if any, would follow g<jobId>g<imageTag>.jpg but are not stored in the dataset.
A record packages:
metadata.json capturing the EDGAR header, the complete documentFormatFiles inventory (including the entries for graphics and the combined submission text), and the structured entity block for every co-registrant.A record does not include:
.jpg, .gif) referenced as GRAPHIC entries in documentFormatFiles. The metadata still lists them and exposes their documentUrl, so they remain retrievable from EDGAR but are not stored on disk in this dataset..txt submission file. It is referenced in documentFormatFiles (with a blank type and blank sequence) and via linkToTxt, but only the per-document HTML files are physically extracted.Form SF-1 has had a relatively stable disclosure architecture since its creation, but several developments materially shape what appears inside a record:
linkToXbrl is typically empty and dataFiles is typically empty in the metadata.Because SF-1 was created only in 2014 and the dataset begins in 2016, every record sits well past the SEC's ASCII-only and plain-HTML transitions. The file-types found in the dataset are HTML and JSON: the HTML documents cover every non-image exhibit in the submission, each retaining its SGML <DOCUMENT> envelope, and the JSON file carries the structured header. The presentation has been consistent throughout — HTML 4 with table-driven layout, inline STYLE attributes, <FONT> tags, non-breaking spaces, and named or numeric HTML entities for smart quotes and special characters (“, ”, , and the like). Exhibits that were historically filed as scanned PDFs in corporate registrations are, in SF-1 practice, almost always converted to HTML before filing, so the dataset's exclusion of image files has minimal effect on the readability of the legal text.
entities as a single transactional unit. The fileNo pairing (333-XXXXXX and 333-XXXXXX-01) is the most reliable link between sponsor and depositor; SIC codes (commonly 49xx for utility ABS, 61xx for finance-company ABS) together with stateOfIncorporation identify which entity is the operating sponsor and which is the bankruptcy-remote SPV.333-XXXXXX file number as the original SF-1 but receives a new accession number. To reconstruct the full registration history of an offering, consumers should join records on fileNo rather than on accessionNo. The description field marks amendments explicitly ("... - [Amend]").333-XXXXXX registration family. The documentFormatFiles inventory captures only the exhibits filed with the current accession; cross-accession references must be resolved externally against EDGAR.documentFormatFiles[].type rather than filename heuristics to identify exhibit categories. The <TYPE> value inside each SGML wrapper matches the metadata field exactly and is the canonical exhibit-code source. Filename <exhibitTag> codes are filer-agent conventions and not authoritative.dataFiles and EX-102/103/104 entries is not a defect; many SF-1 offerings are utility rate-recovery, energy-transition, or other ABS structures that fall outside the asset classes for which Reg AB II mandates asset-level data. When such exhibits do appear they indicate a residential-mortgage, commercial-mortgage, auto loan/lease, debt-security, or resecuritization offering.<TABLE>-driven layout, and inline styling mean that semantic tagging is absent; extraction of pool statistics, waterfall mechanics, transaction-party background blocks, or signature data generally requires either table-structure heuristics or targeted parsing keyed off the prospectus section headings and exhibit-letterhead patterns.Each record is a single EDGAR accession for an initial Form SF-1 or a pre-effective Form SF-1/A amendment registering a public offering of asset-backed securities (ABS) under the Securities Act of 1933.
The depositor is the Securities Act registrant of record. Under Regulation AB / Regulation AB II, the depositor receives the asset pool from the sponsor and transfers it to the issuing entity (the trust or SPV that issues the ABS to investors). For Securities Act purposes the depositor is treated as the "issuer." Depositors are almost always:
The sponsor is not the registrant but plays a substantive role. The sponsor originates or acquires the assets, supplies the underwriting, origination, and static pool data disclosure, and bears the risk retention obligation under Regulation RR (17 CFR Part 246, implementing Dodd-Frank Section 941 / Exchange Act Section 15G), generally holding at least 5% of the credit risk through itself or a majority-owned affiliate. Under the Regulation AB II certification regime (Release No. 33-9638), the depositor's CEO signs a prospectus-level certification; that officer is frequently also an officer of the sponsor.
SF-1 is therefore a dual-entity filing. On EDGAR this commonly appears as paired CIKs and a shared registration number in the form 333-XXXXXX (depositor) and 333-XXXXXX-01 (sponsor or affiliated co-registrant), reflecting the sponsor's status as a co-registrant or guarantor on the registration statement even though the depositor remains the formal issuer.
The trigger is the registrant's decision to register a public ABS offering when Form SF-3 (the ABS shelf form, 17 CFR 239.45) is not available. Form SF-1 is prescribed by 17 CFR 239.44 as the general ABS registration form for "all issuers of asset-backed securities for which no other form is authorized or prescribed" — the residual, transaction-specific (non-shelf) ABS form.
Shelf eligibility on SF-3, established by Regulation AB II in Release No. 33-9638 (adopted August 2014), requires CEO certification at each takedown, transaction-document provisions for an asset representations reviewer, dispute resolution for repurchase requests, investor communication mechanics, and timely Exchange Act filer status for the depositor and all affiliates that have offered the same asset class. SF-1 is filed when those conditions cannot be met. Common situations:
Unlike SF-3, an SF-1 is not automatically effective. The lifecycle is:
A single initial SF-1 typically generates multiple SF-1/A accession numbers, each captured as a distinct record.
SF-1 vs. S-1. SF-1 is exclusively for ABS issuers post-Regulation AB II. Operating companies continue to use S-1. An entity filing SF-1 is, by definition, an ABS depositor.
SF-1 vs. SF-3. SF-3 is the shelf form for ABS issuers meeting the Regulation AB II shelf-eligibility conditions. SF-1 is the transaction-specific form for those that do not. Programs can move between the two over time as compliance status changes.
Depositor vs. issuing trust. The registrant is the depositor entity, not the issuing trust. The trust is typically formed at or near closing and is described within the prospectus rather than serving as the filer.
Sponsor vs. registrant. The sponsor's identity, history, underwriting, and pool data are disclosed in the prospectus, and the sponsor (or affiliate) holds risk retention, but the sponsor is not the Securities Act filer of record except when it appears as a co-registrant under the 333-XXXXXX-01 numbering.
Out of scope. Private ABS issued under Section 4(a)(2) or Rule 144A generates no SF-1 record. Pre-Regulation-AB-II ABS offerings registered on S-1 / S-3 are not part of the SF-1 universe. Post-effective amendments carry different EDGAR form types and fall outside the SF-1/A pre-effective amendment class captured here.
Form SF-1 lives inside the asset-backed securities (ABS) disclosure regime created under Regulation AB and overhauled by Regulation AB II (Release No. 33-9638). The only useful comparison set is the other ABS-specific forms, the general-corporate Securities Act forms SF-1 was branched off from, and the prospectus-stage documents that flow from an SF-1 registration. Forms like 10-Q, 8-K, and proxy materials are not meaningful neighbors because ABS issuers report under a separate item set at 17 CFR 229.1100 et seq.
Form SF-3 (shelf ABS registration). The single closest neighbor. Same Securities Act posture, same Regulation AB II item set, same dual depositor/sponsor filing pattern. The dividing line is Regulation AB II shelf eligibility: SF-3 requires a CEO disclosure certification, an asset representations review provision, a dispute-resolution provision, an investor communication mechanism, and timely Exchange Act reporting for prior ABS of the same asset class. SF-1 is the residual form for issuers that fail or skip these conditions — first-time sponsors, novel or esoteric asset classes, sponsors with a prior compliance gap, and structures with no available shelf. SF-1 cannot support serial takedowns and does not rely on a base-prospectus / supplement architecture, so each filing is larger and more self-contained.
Form ABS-EE (asset-level data exhibits). Not a registration statement. ABS-EE carries the structured XML asset-level data required by Item 1111(h) and Item 1125 (Schedule AL) for the covered asset classes — residential mortgages, commercial mortgages, auto loans, auto leases, debt securities, and resecuritizations. ABS-EE is filed at the offering and again with every Form 10-D, is tabular and loan-level, and is event-driven against the servicing cycle. SF-1 is filed once (plus SF-1/A amendments), is narrative HTML prospectus content, and is the registration umbrella that authorizes the later ABS-EE exhibits. Complementary, not substitutable: SF-1 supplies structural and contractual context; ABS-EE supplies loan-level economics.
Form S-1 (general non-ABS registration). The corporate-issuer analogue under 17 CFR 239.11 rather than 17 CFR 239.44. Same residual posture in the Securities Act hierarchy, entirely different content model. S-1 is built on Regulation S-K and centers on the operating company — MD&A, financial statements, management, going concern. SF-1 is built on Regulation AB items 1100 et seq. and centers on the pool, the transaction parties (sponsor, depositor, servicer, trustee, originator), credit enhancement, and waterfall mechanics. The issuing entity in an SF-1 is a passive SPV with no operating financials.
Form S-3 (seasoned-issuer shelf, non-ABS). Parallels SF-3 the way S-1 parallels SF-1. Two distinctions at once vs. SF-1: S-3 is Regulation S-K based and relies on incorporation by reference to Exchange Act reports; its eligibility test is reporting history and public float, not the ABS shelf conditions. SF-1, by contrast, is not shelf-eligible and does not incorporate by reference, so the body of the filing carries the full prospectus.
Forms 10-D and ABS 10-K (ongoing periodic reports). 10-D is the per-distribution-date report under Rule 13a-17 / 15d-17 covering collections, distributions, pool performance, and servicer or trustee events. ABS 10-K uses the tailored item set at Reg AB Items 1112, 1114, 1119, 1122, and 1123 — servicer compliance assessment, attestation, and the asset representations review report when triggered. These are post-issuance Exchange Act reports; SF-1 is the pre-offering Securities Act registration that sets the baseline they report against. Use 10-D and ABS-EE for performance over time; use SF-1 (or its derived 424B) for the original deal terms, eligibility criteria, and reps and warranties.
Forms 424B (prospectus and supplements). Filed under Rule 424 after a registration statement is effective. For ABS, 424Bs typically attach to an SF-3 takedown rather than SF-1, because SF-1 generally embeds the prospectus in the registration statement itself. When a 424B is filed off an SF-1, it is usually the final pricing prospectus reflecting post-amendment changes. Content overlaps with SF-1, but 424B is the point-of-sale document; SF-1 captures the full pre-effective registration history including SF-1/A amendments, exhibits, and counsel opinions.
Form FWP (free writing prospectus). Rule 433 offering communications outside the statutory prospectus — preliminary collateral term sheets, computational materials, ratings-agency decks. Same subject matter as SF-1 (same pool, same securities) but categorically different: short, marketing-stage, often issued in bursts during the marketing window, and filed by the issuer or an underwriter. FWPs typically accompany SF-3 takedowns but can also appear in an SF-1 lifecycle when collateral data is disseminated pre-pricing. Use FWP for marketing-stage materials; use SF-1 for the binding registered document.
Form SF-1 is the residual, non-shelf ABS registration statement, filed under the dual depositor-as-registrant / sponsor-as-pool-discloser structure that is distinctive to ABS and only partially mirrored on SF-3. It is governed by Regulation AB items 1100 et seq., not Regulation S-K, so the disclosure object is the pool and the transaction structure rather than an operating company. It is one-time and pre-effective (with SF-1/A amendments), not periodic. It sits upstream of the ABS-EE, 10-D, ABS 10-K, 424B, and FWP filings that operate within its framework but does not duplicate any of them.
The dataset is the right source — and the only right source — when the user needs the initial, non-shelf, registration-level ABS disclosure: the SF-3 shelf-eligibility gap is precisely what populates it. Use SF-3 for the shelf-eligible universe, ABS-EE for Schedule AL loan-level data, 10-D and ABS 10-K for ongoing performance, and 424B or FWP for offering-stage communications.
Each SF-1 submission bundles a prospectus, asset pool exhibits, transaction-party disclosures, EX-5.1 legal opinions, and dual-filer entity metadata, so users cluster tightly around ABS structuring, investing, ratings, and Regulation AB II compliance.
Analysts at asset managers, hedge funds, and insurance separate-account desks use SF-1 prospectuses to map the waterfall, credit enhancement, triggers, and rep-and-warranty framework, then pull asset pool exhibits (loan-type mix, geographic concentration, FICO bands, LTV, seasoning) into cash-flow and stress models. SF-1/A amendments are tracked for pricing changes and pool revisions. Output: tranche-level credit memos and bids on new issuance.
Buy-and-hold ABS allocations governed by NAIC designations and ALM constraints depend on sponsor and servicer descriptions, static pool data, originator underwriting standards, and risk factors covering true sale and bankruptcy-remoteness of the depositor. EX-5.1 legality opinions feed internal credit committee write-ups and capital-charge calibration.
Pre-sale and shelf-monitoring teams covering consumer, auto, equipment, esoteric ABS, and CMBS conduits read the prospectus for structural mechanics and the Regulation AB II asset-level exhibits for collateral cuts. SF-1/A amendments are reconciled against pre-sale assumptions; archived submissions support rating committee memos and post-issuance surveillance.
Issuer, underwriter, and trustee counsel run precedent searches across EX-5.1 legality opinions and EX-8 tax opinions to benchmark assumptions and qualifications. They compare risk factor sections, ERISA and Volcker disclosures, EU/UK risk retention statements, and Regulation AB II reps across peer SF-1 and SF-1/A filings to build drafting checklists and confirm alignment with the operative pooling and servicing agreement and indenture.
Structuring bankers use peer SF-1 filings to identify SEC-cleared variations on credit enhancement, prefunding, revolving period mechanics, and cleanup-call thresholds. Dual-filer entity metadata (depositor and sponsor CIKs) plus the issuing-entity description and asset pool exhibits drive pitchbooks, structural option memos, and shelf-refresh recommendations to issuer clients.
In-house compliance at sponsors, depositors, master servicers, and trustees, plus outside compliance counsel, monitor how Reg AB II requirements appear in practice: asset-level data exhibits for residential and commercial mortgages, auto loans and leases, debt securities, and resecuritizations; CEO certifications; NRSRO engagement disclosure; sub-servicer chains; and Regulation RR risk-retention statements. Output: gap analyses and updated disclosure templates for shelf refreshes.
Finance and law-and-economics researchers mine prospectuses and asset pool exhibits to build panel datasets on collateral composition, sponsor concentration, and structural innovation. SF-1/A timing data measures comment-letter responsiveness and time-to-effectiveness for empirical work on whether Regulation AB II met its disclosure objectives.
Product teams at structured-finance analytics vendors train extraction pipelines, NER models, and clause classifiers on the standardized EDGAR submission structure, the metadata.json fields (form type, filer roles, exhibit indexes), and the HTML prospectus body. RAG developers building securitization copilots use SF-1 filings as a dense, formulaic benchmark corpus.
Enforcement and AML units track which sponsors, depositors, servicers, and originators register ABS, and how transaction-party identifications, affiliate relationships, prior securitization experience, and material legal proceedings evolve across SF-1/A amendments. The records support referrals and risk-based exam scoping.
Teams advising on acquisitions of consumer lenders, auto finance companies, equipment finance platforms, and mortgage originators read SF-1 filings where the target has acted as sponsor, depositor, or originator. Historical pool performance disclosures, repurchase obligations, servicer covenants, and related-party arrangements with securitization vehicles drive diligence reports and purchase-price adjustments tied to retained interests.
Concrete workflows the Form SF-1 Files Dataset supports across ABS structuring, investing, ratings, compliance, and research.
Structuring desks and buy-side analysts pull the prospectus body (Items 1113–1115 sections on payment priority, credit enhancement, and derivatives) and the EX-4.1 indenture across SF-1 records sharing a sponsor SIC or asset class. The output is a peer table of subordination levels, overcollateralization targets, reserve account mechanics, and trigger thresholds used to calibrate new-issue structures or to bid tranches on offered deals.
Using fileNo (the 333-XXXXXX shared by an SF-1 and its SF-1/A children) to join records and filedAt to order them, users reconstruct the pre-effective amendment history of a registration. Diffs across the prospectus summary, risk-factor section, and asset pool tables surface staff comment responses, pool revisions, and pricing changes — feeding rating-agency pre-sale reconciliation and counsel comment-letter benchmarking.
Securitization counsel extract EX-5.1 and EX-8.1/EX-8.2 documents (identified by documentFormatFiles[].type, not by filename) into an opinion library. Searching across counsel letterheads, assumption paragraphs, and qualification language produces a precedent set for drafting issuer and underwriter opinions, sizing reliance carve-outs, and aligning REMIC or partnership tax characterizations with peer transactions.
The paired 333-XXXXXX / 333-XXXXXX-01 file numbers and the two-entry entities block in metadata.json cleanly link operating-company sponsors to their bankruptcy-remote depositor SPVs across all SF-1 accessions. Compliance, M&A diligence, and surveillance teams build a sponsor-to-SPV graph keyed on CIK, SIC, and stateOfIncorporation to scope securitization-program exposure, retained-interest holdings, and affiliate-relationship disclosures.
RegTech engineers use the consistent SGML <DOCUMENT> envelope and the EDGAR HTML 4 body to train clause classifiers, NER models for transaction parties (sponsor, depositor, servicer, indenture trustee, originator), and table extractors for pool stratification grids. The fixed Reg AB II item ordering inside the prospectus makes SF-1 a high-signal benchmark for securitization-domain RAG systems and clause-tagging models.
Filtering on the presence of EX-102/EX-103/EX-104 entries in documentFormatFiles (or a non-empty dataFiles array) isolates SF-1 records covering the prescribed asset classes — residential mortgage, commercial mortgage, auto loan/lease, debt-security, and resecuritization pools. Compliance teams use the filter to gap-test Schedule AL coverage; researchers use it to separate prescribed-asset-class deals from utility rate-recovery and other esoteric structures dominant in the SF-1 population.
Academics join SF-1 and SF-1/A records on fileNo, compute the gap between the initial filedAt and the final amendment, and pair it with prospectus-level features (risk-factor length, transaction-party count, pool size). The resulting panel supports studies of comment-letter responsiveness, the cost of falling off SF-3 shelf eligibility, and whether Regulation AB II achieved its disclosure-comparability objectives.
The Form SF-1 Files Dataset is accessible through three endpoints: a JSON index for metadata and discovery, a single archive download for the full dataset, and per-container downloads for monthly slices.
Dataset Index JSON API: https://api.sec-api.io/datasets/form-sf1-files.json
This endpoint returns dataset-level metadata (name, description, last updated timestamp, earliest sample date, total records, total size, covered form types, container format, and file types), the full dataset download URL, and the list of monthly container files with their individual size, record count, updated timestamp, and download URL. Poll this endpoint to detect which containers changed in the most recent refresh and download only the updated months. This endpoint does not require an API key.
Example response:
1
{
2
"datasetId": "1f13365b-9ae0-69ef-acd0-48e5aa77030f",
3
"datasetDownloadUrl": "https://api.sec-api.io/datasets/form-sf1-files.zip",
4
"name": "Form SF-1 Files Dataset",
5
"updatedAt": "2026-05-02T03:05:26.209Z",
6
"earliestSampleDate": "2016-01-01",
7
"totalRecords": 858,
8
"totalSize": 40977936,
9
"formTypes": ["SF-1", "SF-1/A"],
10
"containerFormat": "ZIP",
11
"fileTypes": ["HTML", "JSON"],
12
"containers": [
13
{
14
"downloadUrl": "https://api.sec-api.io/datasets/form-sf1-files/2025/2025-05.zip",
15
"key": "2025/2025-05.zip",
16
"size": 1842736,
17
"records": 8,
18
"updatedAt": "2026-05-02T03:05:26.209Z"
19
}
20
]
21
}
Download Entire Dataset: https://api.sec-api.io/datasets/form-sf1-files.zip?token=YOUR_API_KEY
Downloads the complete dataset as a single ZIP archive containing all monthly containers from January 2016 to the present. This endpoint requires an API key.
Download Single Container: https://api.sec-api.io/datasets/form-sf1-files/2025/2025-05.zip?token=YOUR_API_KEY
Downloads one monthly container ZIP keyed by YYYY/YYYY-MM.zip. Each container holds one folder per 18-digit accession number, with a metadata.json file and all original EDGAR submission documents except image files. This endpoint requires an API key.
The dataset covers Form SF-1 (the general-purpose, non-shelf ABS registration statement prescribed by 17 CFR 239.44) and Form SF-1/A (its pre-effective amendment). Both share the same Regulation AB II content architecture and appear together in the dataset; the formType field in metadata.json distinguishes them.
One record corresponds to a single EDGAR accession — i.e., one SF-1 or SF-1/A submission identified by its 18-digit accession number. The record is delivered as a per-accession folder containing one metadata.json plus every non-image document filed with the submission, each preserved as an SGML-wrapped HTML file.
The depositor — typically a bankruptcy-remote special-purpose LLC or corporation wholly owned by the sponsor — is the Securities Act registrant of record. The sponsor commonly appears as a co-registrant under the paired 333-XXXXXX-01 file number, making SF-1 a dual-entity filing. SF-1 is filed when the offering is not eligible for shelf registration on Form SF-3, for example by first-time entrants, sponsors that have lost SF-3 eligibility, novel asset classes, or one-off offerings.
Coverage begins on January 1, 2016 — the earliest SF-1 accessions on EDGAR following Regulation AB II's phased compliance dates — and extends to the present, refreshed monthly. The form itself was created by the SEC's 2014 ABS rulemaking (Release No. 33-9638).
Each record contains HTML and JSON files only. Every non-image document from the original EDGAR submission is preserved as an SGML-wrapped HTML file (registration statement, exhibits including EX-1.1 underwriting agreement, EX-4.1 indenture, EX-5.1 legality opinion, EX-8 tax opinions, EX-10.x material contracts, and so on), and metadata.json carries the structured EDGAR header and entity block. Binary image files (.jpg, .gif) and the combined SGML .txt submission are not stored on disk, though their URLs remain in the metadata.
SF-3 is the shelf form for ABS issuers meeting Regulation AB II's shelf-eligibility conditions (CEO certification, asset representations reviewer, dispute resolution, investor communication mechanics, and timely Exchange Act reporting); SF-1 is the residual, transaction-specific form for issuers that do not meet those conditions. SF-1 cannot support serial takedowns and embeds the full prospectus in the registration statement itself, so each filing is larger and more self-contained than an SF-3 base prospectus.
Join records on fileNo (the 333-XXXXXX value shared between an initial SF-1 and its SF-1/A children) rather than on accessionNo, then order by filedAt. The description field marks amendments explicitly with "... - [Amend]", and diffs across the prospectus summary, risk-factor section, and asset pool tables surface staff-comment responses, pool revisions, and pricing changes through the pre-effective amendment cycle.