metadata

dataset_info:
  features:
    - name: _id
      dtype: string
    - name: timestamp
      dtype: timestamp[s]
    - name: cik
      dtype: string
    - name: company_name
      dtype: string
    - name: form_type
      dtype: string
    - name: date_filed
      dtype: timestamp[s]
    - name: filename
      dtype: string
    - name: url
      dtype: string
    - name: quarter
      dtype: int64
    - name: year
      dtype: int64
    - name: master_metadata
      struct:
        - name: Description
          dtype: string
        - name: Last Data Received
          dtype: string
        - name: Comments
          dtype: string
        - name: Anonymous FTP
          dtype: string
        - name: Cloud HTTP
          dtype: string
    - name: idx_type
      dtype: string
  splits:
    - name: train
      num_bytes: 9818628455
      num_examples: 24943448
  download_size: 1882894445
  dataset_size: 9818628455
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
language:
  - en
tags:
  - legal
  - SEC
  - EDGAR
  - filings

Dataset Card for SEC EDGAR Filings Master Index

Dataset Details

Dataset Description

This dataset contains metadata for all submissions to the Securities and Exchange Commission (SEC) through their EDGAR system from 1994 until December 14, 2024. The data is extracted from quarterly master files and includes key information about company filings such as CIK numbers, company names, form types, and filing dates.

Curated by: Arthur ([email protected])
Language(s): English
License: Likely public domain as US government data

Dataset Sources

Repository: Based on SEC EDGAR database
Primary Source: SEC EDGAR system (https://www.sec.gov/edgar.shtml)

Uses

Direct Use

Financial research and analysis
Corporate compliance monitoring
Historical filing pattern analysis
Company information retrieval
Regulatory filing tracking
Market research and analytics

Out-of-Scope Use

Real-time filing monitoring (dataset is historical)
Analysis of filing content (contains only metadata)
Legal advice or compliance determinations
Personal identification or contact information gathering

Dataset Structure

The dataset contains the following fields:

_id: Unique identifier (36 characters)
timestamp: Timestamp of record creation
cik: Company Identifier (1-7 digits)
company_name: Name of filing company (0-150 characters)
form_type: Type of SEC form filed (734 distinct values)
date_filed: Date of filing submission
filename: Path to filing in EDGAR system (37-43 characters)
url: Direct link to filing (66-72 characters)
quarter: Filing quarter (1-4)
year: Filing year (1994-2024)
master_metadata: Dictionary containing:
- Description
- Last Data Received
- Comments
- Anonymous FTP
- Cloud HTTP
idx_type: Filing index type

Dataset Creation

Curation Rationale

The dataset was created to provide a comprehensive index of all SEC EDGAR filings, enabling research, analysis, and monitoring of corporate financial disclosures and regulatory compliance.

Source Data

Data Collection and Processing

Data is collected from SEC EDGAR quarterly master files
Processing includes structuring the data into a consistent format
URLs are standardized to SEC.gov domain
Timestamps are standardized to a consistent format
Unique identifiers are assigned to each record

Who are the source data producers?

Primary data producers are publicly traded companies and other entities required to file with the SEC
Data is collected and distributed by the U.S. Securities and Exchange Commission

Personal and Sensitive Information

The dataset contains only publicly available information from SEC filings. While it includes company names and identifiers, it does not contain personal information beyond what is already public in SEC filings.

Bias, Risks, and Limitations

Historical data may be incomplete for earlier years
Company names may change over time
CIK numbers may be reused in rare cases
Form types may be deprecated or changed over time
URLs may become invalid if SEC changes its archive structure
Coverage limited to SEC-regulated entities

Recommendations

Verify current company information through SEC's current EDGAR system
Cross-reference CIK numbers when tracking companies over time
Consider form type changes when analyzing historical patterns
Implement appropriate error handling for URL access
Use appropriate date ranges for analysis based on regulation changes
Consider regulatory context when analyzing filing patterns

Dataset Card Contact

Arthur ([email protected])

Code

https://github.com/arthrod/sec-edgar-bulker