Data Discovery customers, did you know that signing in gives you access to even more features?

Sign in

Analytics Data | Quantitative Analytics

StarMine Text Mining Credit Risk Model

Text mining analytics dataset offering credit risk model signals for quantitative investment and risk assessment.

Key facts for StarMine Text Mining Credit Risk Model

Coverage metric

Geography:
North America, Asia / Pacific, EMEA, Latin America and the Caribbean
History:
From 1998
Coverage count:
Legal entities - 39000 Public companies
Asset Class
Ordinary Shares, Equities

Delivery metadata

Data Frequency
Daily
Language
English
Delivery methods:
Excel, Deployed/Onsite Servers, API, Web Service, SFTP, Cloud, Desktop, FTP, Bulk, Snowflake, Website, RSS Feed
Data formats:
PDF, GZIP, XML, JSON, CSV, SQL, Text, Python, HTML, Bitmap, MPEG, User Interface, PCAP
Minimum service frequency
Daily

Overview of StarMine Text Mining Credit Risk Model

DETECT FINANCIAL DISTRESS

Uncover Hidden Credit Risk

  • Identifies language associated with financial health and distress.
  • Helps surface credit deterioration beyond traditional financial measures.
  • Provides differentiated signals with modest correlation to other credit risk models.

TEXT-POWERED CREDIT RISK

Credit Risk Beyond Numbers

  • Assesses company default risk using text analysis of news, transcripts, filings, and broker research.
  • Provides probability of default, letter grade and relative risk ranking.
  • Offers distinct insights to complement traditional credit risk models.
DECODE FINANCIAL LANGUAGE

Intelligent Text Analysis

  • Applies source-specific dictionaries to evaluate language across news, transcripts, filings, and broker research.
  • Focuses on income, debt structure, legal obligations and external events.
  • Converts textual signals into an empirically calibrated PD.

Description of the dataset

  • LSEG provides the StarMine Text Mining Credit Risk Model to assess corporate financial distress risk using language signals from trusted textual sources. We evaluate content from Reuters News, StreetEvents conference call transcripts, corporate filings and selected broker research reports.
  • Our model applies a bag-of-words text mining methodology that analyses the frequency of words and phrases linked to observed credit outcomes. We transform unstructured language into a systematic credit health signal designed to identify companies more likely to weaken or thrive.
  • LSEG ranks publicly traded companies on a 1-100 percentile scale, with 100 representing the healthiest companies. We support credit surveillance, equity risk assessment and early-warning workflows by incorporating information that may not be fully captured in traditional financial statement data.

Accessing the dataset

This dataset can be used by the following products. Talk to us to learn more about different packages and offerings.

Workspace for Investment Bankers, Consultant Edition

Workspace For Investment Bankers Consultants Edition delivers a client-centric workflow solution powered by best-in-class market and referential information. Workspace for Investment Bankers - Consultants edition addresses a specific customer segment, consulting firms, within our investment banking business.