Financial text research
Research resources availableAnalyText.
Financial text for accounting research.
Financial textual data and research tools for studying accounting disclosures. Explore SEC filing sections, financial statement notes, calculated text metrics, word lists, and analysis applications on the dedicated AnalyText website.
A research project by Mauricio Codesso , Rani Hoitash, and Udi Hoitash.
Downloads and registration are maintained by AnalyText. Its access and reuse terms are described below.
Narrative disclosures as research evidence
Start with the text.
Define the question.
Financial disclosures contain narrative explanations alongside reported amounts. AnalyText brings extracted disclosure text together with analytical resources, providing a starting point for research on accounting language, financial reporting, and the information companies communicate.
Choose a resource according to the unit of analysis your question requires: a major filing section, an accounting note, a calculated measure, or a set of words. Read the related study and the source definitions before selecting observations, comparing firms, or interpreting a textual measure.
Resource overview
Choose the material
your question needs.
This overview helps you select a starting point. The AnalyText website owns current coverage, file formats, variable definitions, and detailed instructions.
SEC filing sections
Extracted text from major sections of annual 10-K and quarterly 10-Q reports, including business descriptions, risk factors, and Management's Discussion and Analysis. The source website identifies which filing items are available.
Research questions to consider
- Compare how firms describe their businesses or discuss risk.
- Examine changes in management's narrative alongside an accounting question.
Financial statement notes
Text from financial statement notes in 10-K and 10-Q filings, with accounting topic classifications and related measures. Notes provide a more focused unit of analysis than a complete filing section.
Research questions to consider
- Study the explanations surrounding a particular accounting topic.
- Compare disclosure language while preserving the context of the note.
Calculated text metrics
Precalculated measures accompany filing sections and notes, including measures of readability, sentiment, and text length. Consult the source descriptions for variable definitions and the text represented by each measure.
Research questions to consider
- Explore how a disclosure's language and length vary across observations.
- Check whether a measure captures the concept your analysis intends to study.
Word lists and dictionaries
Domain-specific COVID-19 and human capital word lists developed from SEC filing language. These resources support dictionary-based analysis of terms relevant to a defined disclosure topic.
Research questions to consider
- Investigate language associated with human capital or pandemic disclosures.
- Review dictionary coverage and context before interpreting word frequencies.
Text analysis applications
The downloadable tools include a Word Frequency Analyzer for dictionary-based word counts and Excel output, and a Topic Discovery Tool for latent Dirichlet allocation (LDA). AnalyText maintains application information and demonstrations.
Research questions to consider
- Explore a selected dictionary within an appropriate collection of text.
- Investigate candidate topics and evaluate them against the source documents.
Who can use these resources?
Research, study,
and critical interpretation.
The following are suggested uses. Choose resources and methods that fit your question, preparation, and applicable access conditions.
- Accounting researchers
- Explore narrative financial disclosures when developing questions about reporting, risk, accounting topics, or the language firms use. Match the selected text and measures to the research design.
- Graduate students
- Use the project and its related study to examine the relationship between source documents, text extraction, variable construction, and the interpretation of research evidence.
- Instructors and independent learners
- Consider disclosure examples for financial text analysis or research methods learning. Review the project's access requirements, file guidance, and reuse terms before assigning or adapting materials.
Access and responsible reuse
Review the source
before downloading.
Selected downloads open a registration dialog on the AnalyText website.
The dialog requests a full name and email address; institution is optional. Review its explanation of data collection before continuing.
AnalyText maintains the current resource files, variable guidance, application information, and access conditions. Follow its navigation to the relevant collection before choosing a download.
License and permissions
CC BY-NC 4.0 (opens in a new tab) · Creative Commons Attribution-NonCommercial 4.0 International
The AnalyText website states this license for its data and content. Review resource-specific notices and contact the project about commercial-use permissions.
Review access and terms on AnalyText ↗ (opens in a new tab)Research behind the repository
Read the related study.
Use the article to understand the repository's methods and the construction of its textual data. Cite the relevant study and follow the source project's guidance for the specific resources used.
Shared Python code
The public repository describes a pipeline for retrieving filings, extracting text, and calculating textual metrics. Consult its documentation when examining how the data is constructed.
View the shared Python repository ↗ (opens in a new tab)Textual Financial Data Repository for Machine Learning, Artificial Intelligence, and Textual Analyses: Major Sections from 10-K, 10-Q, and Financial Statement Notes Extracted Using Shared Python Code ↗ (opens in a new tab)
Mauricio Codesso; Rani Hoitash; Udi Hoitash. (2026). Journal of Information Systems, 40(2), 117–137.
DOI: 10.2308/ISYS-2024-084 (opens in a new tab)
See the author's selected publications →Connected through the Hub
Different resources.
A shared interest in accounting data.
AnalyText supports financial text research. For introductory accounting analytics with Excel, SQL, and Power BI, explore the free textbook. For synthetic business transactions, accounting processes, and ledger tracing, start with Charles River.
The Hub connects these resources for discovery. Each project's dedicated website maintains its materials and applicable terms.