Sentiment analysis of Q1FY20 earnings call transcript

The analysts are always looking a newer ways to analyse the results of the company and overall sentiment of the market. Machine learning and artificial intelligence has given analyst a new arrow in their quiver to analyse the overall market.  Natural language processing (NLP) is one of the ways in which you can analyse the results and expectation of the management from what they say. NLP in simple terms refers to the use of algorithm to process sentences/text in a natural language such as English. The objective here is to extract information from unstructured or semi-structured data found in what management speaks. To enable this NLP makes use of artificial intelligence, computational linguistics, and computer science. You can use NLP to models hundreds of text documents to ascertain the sentiment.

To analyse the sentiments of the India Inc, we consolidated latest earnings call transcript of 161 companies, which covers almost 70% of the market cap and across 12 sectors. On this we run the model, implementing the “Bag-of-words” approach to sentiment analysis. The process identifies positive and negative words (or a string of words) within management commentary. For this, it makes use of a large dictionary which contains words that carry sentiment.  Each word in this dictionary can be assigned a weight. The sum of the positive and negative words is the final sentiment score generated by the model.

The sentiment analysis shows that overall the India Inc management has a positive outlook; however, it is quite measured one. Against a positive score of 288 the negative score stands at 271. Hence, it is marginally positive.

These are the most commonly used words in their con-call.  You can see that many of them are expecting growth in coming quarters while are also concerned with cost and debt.  

 These are the most frequently used words in their commentary.