The Textual Data Analytics (TDA) dataset takes earnings calls transcripts one step further with sentiment and behavioral-based metrics rigorously researched and tested against frequently used quantitative strategies.

Implement signals from 800+ predictive and descriptive metrics derived from Natural Language Processing (NLP) in combination with data from professionals, ownership, and estimates datasets with differentiated and additive characteristics. Quickly assess the sentiment and transparency expressed on earnings calls of 11,600+ active companies to analyze and hone in on high-impact individual sections, speaker types, and individual components of calls or as building blocks to uncover additional signals.

Dataset includes:

• Bag-of-Words Sentiment metrics such as Net Positivity, Positivity-to-Negative, Positive Sentiment and Negative Sentiment as well as TF-IDF and Cosine Similarity weighted sentiment
• Behavioral metrics such as Language Complexity, Analyst Favoritism, Numerical Transparency, Guidance References, Exogenous Factor References, Language Similarity and Timing of Numerical References
• Measurement of Financial Performance Topic Identification and neighboring positive descriptors around market-moving topics such as "Revenue", "Bottom-Line", and "EPS"
• Objective measures of Executive and overall earnings call performance in a time series or cross-section

[Awards]

• 2020 Data Management Insight Award for Best Proposition for AI & Machine Learning