Authors : Sigo Marxiaoli, Murugesan Selvam, Kasilingam Lingaraja and Vinayagamoorthi Vasanth
Abstract: Prediction of financial markets, especially prediction of highly volatile stochastic stock market indices, plays a crucial role in identifying profitable investment avenues by the financial investors at large. The investing community encompasses retail investors, financial institutions, investment banks and Foreign Institutional Investors who look for the creation of wealth in the form of capital appreciation and earning the title of ownership of business enterprises by investing in the securities market, through buying and selling of shares of stock exchange listed corporate entities. The forecasting of dynamic financial market movements is one of the scientific endeavours which demands a great deal of market intelligence, financial acumen and domain knowledge of the characteristics of behavioural finance in a wider spectrum. This paper aims to discuss the non-linear movement pattern/trend of the most active two stock indices of India, namely, the Sensex and Nifty, during the study period from 2009-2015 by applying the traditional logistic regression method and one of the neural network tools, namely, k-nearest neighbourhood algorithm. This study would help the investors to streamline their investment patterns and strategies in order to take well informed investment decisions and optimize their stock returns by using the relevant market information.
Sigo Marxiaoli, Murugesan Selvam, Kasilingam Lingaraja and Vinayagamoorthi Vasanth, 2017. Forecasting the Stock Index Movements of India: Application of Neural Networks. International Journal of Soft Computing, 12: 120-131.