DO NEWS HEADLINES HAVE ENOUGH SENTIMENTS? ANALYSIS ON FOREX, EQUITY AND BOND MARKETS OF INDIA

Abstract

Study of Financial market is very riveting topic for investors and researchers. Financial markets are very important for the growth of country, it will help company to raise money from people, it helps savers to become an investor. People are very risk averse in nature so they started adapting analytics and machine learning and over the years they have seen that amalgamation of data analytics and finance has created wonders for investors. This study involves the sentiment analysis of the News headlines of Indian newspaper Time of India, which is the most read English newspaper in India. For the sentiment analysis Python programming language has been used, in Python Natural Language Tool kit (NLTK) has been used to know the polarity of each news headlines of the dataset. NLTK is a one of the platforms in Python, this platform go hand in hand with human language data for implementing statistical natural language processing (NLP). This study considers to do analysis on three dominated markets of India. (i.e. Forex market, Equity market and Bond market). Three different indicators have been chosen from each market; USD/INR from Forex market, Sensex From Equity Market and S&P BSE India Agency Bond Index from Bond market. If sentiment of the day and movement of particular index are in same direction it means predicted direction is correct otherwise not by this rule final received accuracy for USD/INR, Sensex and Bond Index are 71%, 53% and 58% respectively. Therefore, it can be concluded that yes news headlines have the sentiments, if retail investors or traders even ignore the technical analysis and focus only on fundamental analysis by using this algorithm they will be in Profit in each market..