To investigate the impact of NLP variables, look into companies that have similar financial variables but differing NLP features, and hence differing ratings or changes in ratings
Provide examples of the NLP features to show why they are different (for example, short piece of text from each call, demonstrating the difference in readability, or difference in sentiment)
Steps
Create an Excel file with ticker, fixed quarter date, Rating, Change in Rating, Altman's Z (good enough as a representation of financial variables), some NLP features, and the transcript. I suggest TONE1, then Gunning-Fog Score, then numeric transparency as these are kind of unintuitive, and our explanation can also serve to explain what the features are.
Sort by Altman Z and look for big differences in rating and in the NLP variables while having similar Altman Z
Read the transcripts and find excerpts to illustrate differences in the NLP variables
For writeups and presentations we need the variables, the transcript excerpts, and an explanation of what the NLP variables mean in light of the excerpts, and how they influence the rating or change in rating
To investigate the impact of NLP variables, look into companies that have similar financial variables but differing NLP features, and hence differing ratings or changes in ratings
Provide examples of the NLP features to show why they are different (for example, short piece of text from each call, demonstrating the difference in readability, or difference in sentiment)
Steps
Supersedes #48