NL Analytics docs
  • Welcome
  • Our platform
    • Overview
    • Keyword Tool
    • Risk Tool
      • Search earnings calls
      • Search history
      • Merge searches
    • Snippet Tool
  • Research and publications
    • Recent research
    • Academic foundations
  • Help & FAQ
    • Get Help
    • FAQ: General
      • How often do you update the earnings calls data?
      • What languages are covered in your earnings call data?
      • What is the geographic coverage of your earnings call data?
      • How is capitalization handled?
      • How are plural nouns handled?
    • FAQ: Risk Tool
      • Why do I have zero entries in my data?
      • How are Exposure, Risk, and Sentiment defined?
      • What is the "Overall risk and sentiment" file in the output of the Risk Tool?
      • How long do you retain searches?
      • What is the crosswalk to GVKey in the output of the Risk Tool?
      • How can I optimize the speed of the Risk Tool?
      • How are the various variables defined?
    • FAQ: Keyword Tool
      • Does the Keyword Tool take into account feedback by way of rejecting a term?
Powered by GitBook
On this page
  1. Help & FAQ
  2. FAQ: Risk Tool

What is the "Overall risk and sentiment" file in the output of the Risk Tool?

The "Overall risk and sentiment" CSV file contains the unconstrained counts of risk and sentiment, which can be merged with any search result using the variable earningscallID.

In particular, Overall Risk and Overall Sentiment are defined as follows:

Overall Risk=∑s1(at least one risk synonym in s)\textit{Overall Risk}= \sum_{s} \mathbb{1}(\text{at least one risk synonym in } s)Overall Risk=∑s​1(at least one risk synonym in s)

Overall Positive Sentiment=∑s1(at least one positive sentiment word in s)\textit{Overall Positive Sentiment}= \sum_{s} \mathbb{1}(\text{at least one positive sentiment word in } s)Overall Positive Sentiment=∑s​1(at least one positive sentiment word in s)

Overall Negative Sentiment=∑s1(at least one negative sentiment word in s)\textit{Overall Negative Sentiment}= \sum_{s} \mathbb{1}(\text{at least one negative sentiment word in } s)Overall Negative Sentiment=∑s​1(at least one negative sentiment word in s)

Overall Sentiment=Overall Positive Sentiment−Overall Negative Sentiment\textit{Overall Sentiment} = \textit{Overall Positive Sentiment} - \textit{Overall Negative Sentiment}Overall Sentiment=Overall Positive Sentiment−Overall Negative Sentiment

Our clients have found these additional metrics helpful because they can be used to calculate the share of a given topic in risk or sentiment mentions. For example, the Risk from a search for inflation-related terms could be divided by Overall Risk to obtain the fraction of risk-related sentences with inflation mentions in the total number of risk-related sentences.

Last updated 1 year ago