Natural Language Processing for Financial Analysis
Purpose-built for financial analysis, our NLP tools and AI-ready data simplify the extraction of insights from complex, unstructured documents like transcripts and filings.
Transform Text Into Insight
Analyzing textual data from earnings call transcripts, research reports, and other financial documents can be time-consuming and costly. Our NLP-ready datasets eliminate the need for manual data preparation, allowing you to move straight into analysis, enhanced by rich metadata tagging and seamless integration with core datasets like Financials and Estimates.
And with ProntoNLP, our advanced natural language processing platform, you can go even further. Purpose-built for financial analysis and powered by proprietary large language models (LLMs), ProntoNLP transforms unstructured corporate disclosures into structured, actionable intelligence. It identifies and scores key performance indicators (KPIs), monitors both macro and micro risks, and surfaces forward-looking strategic signals in near real-time.
ProntoNLP: The No-Code Auto-NLP Platform for the Investment Community
ProntoNLP makes it simple to turn complex financial text into actionable insight. With its intuitive, no-code platform, analysts can uncover trends, test ideas, track sentiment, and detect inflection points across earnings calls, filings, and other corporate disclosures, all without writing a single line of code.
By cutting through the noise and surfacing the signals that matter, ProntoNLP helps investors move faster, validate strategies with confidence, and stay ahead of market shifts. Now part of S&P Global Market Intelligence, it brings together advanced AI and trusted data to deliver sharper, forward-looking perspectives on company performance.
What is Natural Language Processing?
Natural Language Processing (NLP) is a field of artificial intelligence that enables computers to interpret and generate human language. By combining linguistics with machine learning, NLP makes it possible to analyze large volumes of unstructured text, such as financial disclosures, earnings call transcripts, and research reports, with speed and consistency. It helps uncover patterns, extract meaning, and translate language into data that can be used for research, forecasting, and decision-making.
NLP has become increasingly important in the financial world, where language often reflects strategic intent, risk exposure, or market sentiment. Analysts and researchers use NLP to track executive tone over time, monitor emerging topics, classify business events, and quantify signals that were once considered qualitative. As the complexity and volume of financial text continue to grow, so does the need for scalable, language-aware tools that can keep up.
At S&P Global, we’ve invested in NLP capabilities for over a decade, from machine-readable datasets to platforms like ProntoNLP, which apply large language models to financial text. Our work aims to support the broader adoption of NLP across finance, academic research, and machine learning, helping users explore what language reveals about markets, companies, and the economy.
NLP Services
-
Kensho LLM API
The Kensho LLM-ready API allows you to access essential S&P Global datasets via any GenAI application. The solution integrates seamlessly with many of the leading Large Language Models and can also be integrated with proprietary LLMs. You can leverage natural language queries to explore S&P Global’s tabular data with ease.
Now powered by enhanced support for Model Context Protocol (MCP), the API also includes a built-in Model Context Protocol (MCP) server for customers using MCP-compatible systems.
What datasets are included?
- CIQ Financials
- Market Data
- Business Relationships
- Earnings Call Transcripts
- Company Intelligence
- Transactions
-
ProntoNLP
ProntoNLP is an advanced natural language processing platform that transforms Filings and Transcripts into actionable intelligence. Leveraging a proprietary large language model, it extracts and scores key performance indicators (KPIs) from corporate disclosures—helping users identify strategic inflection points, validate financial statements, and monitor risks such as inflation or supply chain disruptions. The Transcript Analytics dataset surfaces insights from earnings calls, while the Filings Analytics dataset processes both SEC and non-SEC filings, offering hourly signal updates and low correlation to earnings data, providing fresh, forward-looking perspectives on corporate performance.
-
Script Insight
Script Insight is a revolutionary AI-powered tool designed for issuers and their investor relations teams. The tool allows corporate issuers to analyze their earnings call transcripts before their earnings calls take place.
-
Document Intelligence on S&P Capital IQ Pro
Document Intelligence brings together multiple GenAI innovations in a singular interface. Users benefit from smart search and the ability to interact with documents using GenAI tools and leverage smart summarization. The solution also provides key phrases that highlight relevant and significant topics within earnings call transcripts and natural language processing (NLP)-derived sentiment scores.
-
Kensho Nerd
Kensho NERD (Named Entity Recognition and Disambiguation) is a cutting-edge machine learning system that unlocks the full potential of your textual data by linking it to existing sources of structured knowledge. Trained on millions of business-related documents, NERD is the only technology on the market specifically optimized to extract financial entity information from text documents.
-
Kensho Scribe
Kensho Scribe is a leading, on-demand transcription service for financial audio. With two complementary offerings available, Scribe AI and Scribe Human-in-the-Loop, your audio files are transcribed into human and machine readable text utilizing sophisticated deep learning models.
Scribe is purpose-built for financial audio, leveraging S&P Global's long history of providing high-quality transcripts to Wall Street, including more than 100,000 hours of domain-specific audio and associated text (e.g., earnings calls, management presentations, and acquisition announcements).
Insights from NLP algorithms
Webinars
Case Study
An Investment Firm Uses GenAI to Find Market-Moving Insights in Earnings Calls
-
About
A buy-side firm sought to enhance its equity research with advanced AI tools to extract deeper insights from earnings calls and filings.
-
Challenge
Traditional models and manual reviews were too slow and limited to keep pace with growing volumes of unstructured financial data. The firm needed a faster, more accurate way to detect signals and strategy shifts.
-
Solution
The firm adopted ProntoNLP from S&P Global Market Intelligence, a GenAI-powered platform optimized for finance. It delivers structured sentiment, event detection, and thematic insights from transcripts and filings through user-friendly interfaces and data feeds.
-
Outcome
With ProntoNLP, the analysts gained access to hourly updates across 110 key topics, streamlined workflows, and more precise signals. This strengthened their research process and gave them a competitive edge in equity investing.
Featured Content
Familiar Signal, New Context The Evolution of Earnings Call Sentiment Analysis from Lexicons to LLMs
Large language models (LLMs) have garnered widespread attention for their ability to understand natural language. Their application in equity investing, however, remains in its infancy due to novelty and cost. This study leverages LLMextracted features from earnings call transcripts and transforms them into actionable stock selection signals.
Hear from the experts in our webinar: Betting on LLM’s Intelligence: How Contextual Understanding Drives Outperformance
This webinar provides a detailed blueprint on how investors can systematically leverage LLM-based transcript features to identify future outperformers.
QRS Alpha Research
Watch our new video to find out how our Generative Al alpha research identifies key behaviors of executives during earings calls that correlate with company performance. By analyzing earings calls, we classified executives into four communication styles based on proactiveness and transparency Proactive & On Topic executives consistently outperform, generating an annualized average of 247 Bps of Alpha, while their Reactive & Off Topic peers underperform with -256 Bps of Alpha. With these insights, you can identity executive communication styles on earnings calls.
Expert Perspectives on NLP in Finance
Liam Hynes is interviewed at the Snowflake Summit on his research “Questioning the Answers: LLMs Enter the Boardroom”, where ProntoNLP and Snowflake Cortex were used to analyze 192,000 earnings calls with LLMs, revealing how executive communication patterns correlate with market performance.
Prof. Ronen Feldman explains how ProntoNLP takes unstructured textual data and finetunes it unto structured data & analytics.
Eghbal Rahimikia from Alliance Manchester Business School presents his academic research on year-specific LLMs for finance, trained on historical data—including S&P Global’s Key Developments—to avoid look-ahead bias and outperform larger models like LLaMA in trading scenarios.
Request Follow Up
Textual Data for Natural Language Processing
Our NLP-ready datasets empower you to stay ahead of the market and make informed decisions with confidence.
With our textual data suite, you can access machine-readable transcripts, filings, broker research and proprietary S&P Global credit research in structured formats for NLP analysis, so you can skip the entire process of sourcing, cleansing, and maintaining the data, and jump straight into insightful analysis.