What it does
MarketPsych produces financial sentiment datasets and tools that classify a firm's own text. Its MarketPsych Analytics feed tracks news and social media, while Text Tagging Engine returns entity, event and sentiment labels for supplied documents. Data subscriptions are arranged through LSEG, with account access for the Data App.
MarketPsych offers two useful starting points. If you need a prepared time series, MarketPsych Analytics supplies sentiment, attention and theme scores from news and social media. If you already have filings, transcripts or research text, Text Tagging Engine processes those documents through manual entry, a software connection (API) or an on-premises installation.
The tagging service returns JSON, a structured format for software, with labels attached to individual sentences. It identifies entities, topics, events and emotions as well as positive or negative tone. Its historical entity recognition is designed to match company identities to the period of the document. Keep those identifiers and dates when joining text features to market observations.
The prepared Analytics feed has minute, hourly and daily updates, with history described from 1998. Asset-specific fields differ: commodity coverage includes production and supply themes, while company fields include earnings forecasts and management sentiment. Mentions count the attention an asset receives; Sentiment measures tone. Comparing both can distinguish a widely discussed event from a small amount of strongly worded coverage.
Request A Demo opens a form for your name, email, firm, country and comments. The Data App's detailed features are reserved for data subscribers through LSEG, generally at institution level. Delivery includes API, SFTP file transfer and web access. Academic users seeking archive-only data can ask about a discounted subscription. A model-generated sentiment score is an input to analysis, not a company-reported financial measure.
Strengths
- Processes a firm's own documents or supplies prepared sentiment time series
- Returns sentence-level tags in JSON through the text service
- Distinguishes attention, tone and asset-specific financial themes
Watch out for
- Detailed Data App features require a LSEG data subscription
Best for
- Research teams building financial-text classification pipelines
- Institutional analysts comparing sentiment and attention across markets
Sources
- MarketPsych financial text and sentiment products (Oct 27, 2026)
- Financial Document Tagging and Analytics (Oct 27, 2026)
- LSEG MarketPsych Analytics: dataset structure (Oct 27, 2026)
- MarketPsych contact and Data App access FAQ (Oct 27, 2026)