Q: What does "SA" stand for?
A: Sentiment Analysis Score
Q: How is the Sentiment Score calculated?
A: (P - N)/(P +N+1) P and N are the sum of all genetic sentiments and all event sentiments.
Q: What do positive and negative numbers mean?
A: Each number represents the count of positive vs. negative results relative only to the score that Amenity Analytics provides for each company. For article sentiment scores, the positive/negative numbers represent the count of positive/negative results as the scores are +1/-1 (there are no weights).
Q: Why do I see issues with the data before 2007?
A: Our transcript provider, Factset, began operation in 2007. Transcripts that were created before 2007 will have a creation date of 2007. The full Factset database of transcripts goes back to 2003.
Q: Can I use a transcript source other than Factset?
A: Amenity Analytics can supply data from S&P Global Transcripts in addition to Factset. Upon request, we can process other vendors transcript text.
Q: What is Aggregate Scoring, and how is it different from a Sentiment Analysis Score?
A: Aggregate Sentiment includes a decay factor which does not apply to insights from earnings call transcripts due to the infrequency of data.
Q: Why do event counts not add up to total Sentiment Analysis positive or negative numbers?
A: The total Sentiment Score is an events count, plus the general sentiment.
Q: Why are there multiple transcripts for the same quarter?
A: Our transcript providers will release two versions of the same script. The first version is a raw transcript, and the second is a corrected version that includes improvements to the accuracy on the transcript. There are occasional times when the provider will release an updated version of the transcript as well.
Q: Do you cover international companies?
A: Amenity Analytics will deliver acquired data on all transcripts that have been provided by our transcript supplier.
Q: Have you backtested, or have any independent research?
A: Upon request, Amenity Analytics can provide several research papers on data. Please contact us at email@example.com