Stock Selection Insights Using Earnings Call Transcripts
6
Motivation - Historical Performance Comparison of Strategies
7
Earnings Call - Introduction
8
Text Preprocessing prior to Signal Construction
9
Two Main Categories of Signals
10
Construction of Sentiment-Based Signals: Sentiment Level
11
Construction of Sentiment-Based Signals (continued): Change in Level and Change in Trend
12
Construction of Behavioral-Based Signals
13
Description of Empirical Results
14
Sentiment-Based Signals Empirical Results
15
Sentiment Based Signals - Entire Transcripts
16
Controlling for Risk and Alpha Signals
17
Natural Tilts of Sentiment-Based Signals
18
Results after control for risk and alpha factors
19
Correlations
20
Takeaways - Empirical Results
Description:
Explore natural language processing techniques for stock selection using earnings call transcripts in this 22-minute video from Open Data Science. Learn how to decipher sentiment- and behavioral-based signals that have demonstrated historical stock selection power in the U.S. market. Discover the ABCs of NLP, understand its importance in finance, and delve into the general steps of NLP analysis. Examine the process of text preprocessing, signal construction for both sentiment-based and behavioral-based indicators, and review empirical results. Gain insights into controlling for risk and alpha factors, natural tilts of sentiment-based signals, and correlation analysis. Enhance your understanding of how unstructured data can be leveraged to differentiate sources of alpha in investment strategies.
Natural Language Processing - Deciphering the Message Within the Message Stock Selection