Spin-offs in Indian Stock Market owing to Twitter Sentiments, Commodity Prices and Analyst Recommendations

Abstract

These days the most crucial and commercially valuable information is becoming increasingly available on the World Wide Web. Companies which provide financial services are also making their products available on the web. As there are various types of web financial information sources, such as News Blogs, News Articles, Financial websites and Social Media, a lot of work is being carried out in the Stock Market domain using Data Analytics. This paper tries to find out if twitter sentiments and commodity prices help in predicting actual stock prices for top 50 companies listed on NIFTY at NSE, India, by using Natural Language Processing, Sentiment Analysis and Machine Learning techniques. The results show that, Twitter sentiment gives 70% accuracy while predicting the actual stock prices and the accuracy is improved by 15% when integrated with commodity prices for making company-wise predictions. Furthermore, we check if analyst's recommendations have more impact on stock market price movements for all companies listed on NSE as compared to tweeter public sentiments. The results show that, analyst's recommendations contribute more with 9% of the increase in prediction accuracy.