The Research Accelerator provides 12-months of access to four MarketPsych data feeds, delivered alongside structured research support, quant team access, and ready-to-use Python notebooks. It is designed for institutions building NLP-based quantitative strategies who need deep historical data and hands-on expertise to derive value quickly in back-testing environments.
MarketPsych Research Accelerator
Research infrastructure, data access, and expert support — purpose-built for NLP-based financial research.
Overview
Included Datasets
MarketPsych Analytics
News & Social Media Sentiment
ESG Analytics
Media-Based ESG Signals
Transcript Analytics
Earnings Call NLP
StarMine MMS Model
30-Day Equity Ranking
MarketPsych Analytics
Sentiment and thematic indicators derived from 4,000+ global news and social media sources, covering 140,000+ entities across equities (140,000+) and equity indices (150+), commodities (50+), currencies (40+), cryptocurrencies (1,000+), and sovereign bonds (60+). History from 1998 with 60-second, hourly, and daily formats available. Separate feeds available for news and social media.
MarketPsych ESG Analytics
External, media-derived ESG signals for 140,000+ public and private companies and 252 countries. Covers environmental, social, governance, controversy, and buzz metrics across 4,000+ dimensions. Built for quantitative investing, greenwashing detection, and risk profiling. History from 1998 with 60-second, hourly, and daily formats available. Separate feeds available for news and social media.
MarketPsych Transcript Analytics
NLP-processed earnings calls and corporate presentations for 18,000+ global companies, point-in-time, from 2001. Provides sentence-level scoring, speaker identification, 13 emotional tone dimensions, and 1,000+ topic tags for model building and fundamental research.
StarMine MarketPsych Media Sentiment Model
A co-branded predictive equity ranking model scoring 16,000+ global stocks on a 1-100 scale, forecasting 30-day relative price performance. Orthogonal to traditional factors; live since January 2020 with a Sharpe ratio above 1 for U.S. equities' decile spread (top-bottom ranked deciles).
Research Support & Infrastructure
Research Team Access
Direct engagement with MarketPsych quant researchers to design studies and interpret results.
Python Notebooks
Pre-built notebooks for each dataset. Load data, run analyses, and adapt strategies with straightforward code.
Backtesting Guidance
Structured support for signal construction, factor testing, and model validation.
Quant Workflow Integration
Hands-on help mapping data to existing research pipelines and identifying high-value use cases.
Python Research Notebooks
Each dataset comes with user interface access, free Python research notebooks, and academic citations (over 500 total). Notebooks cover data loading, exploratory analysis, signal construction, and backtesting templates, enabling clients to move from access to insight without building infrastructure from scratch.
- Pre-configured for each of the four included datasets
- Covers data ingestion, cleaning, visualization, and factor construction
- Backtesting templates aligned to common quant research workflows
- Maintained and updated by the MarketPsych research team
Delivery
REST API
Streaming, bulk & request-response
Snowflake
Native data sharing & SQL analytics
Databricks
Spark-native lakehouse integration
At a Glance
| Feature | Details |
|---|---|
| Minimum Term | 12 months |
| Historical Coverage | 20+ years (from 1998 for news/social; 2001 for transcripts) |
| Datasets Included | MarketPsych Analytics, ESG Analytics, Transcript Analytics, StarMine MMS |
| Delivery | REST API, Snowflake, Databricks |
| Notebooks | Free Python notebooks provided for each dataset |
| Support | Direct quant researcher access; structured onboarding and research guidance |
| Data Timing | Delayed access; 1 to 3 months per dataset |
Contact your LSEG account manager or reach us at [email protected] to arrange a trial or demonstration.