Predicting Stock Market using Regression Technique

Abstract

We use two and half year data set of 50 companies of Nifty along with Nifty from 1^st Jan 2009 to 28^th June 2011 and apply multivariate technique for data reduction, namely Factor Analysis. Using Factor analysis we reduce these 50 companies’ data (50 variables) into the most significant 4 FACTORS. These four significant factors are then used to predict the Nifty using Multiple linear regression. We observed that the model is good fitted and it explained 90 % of the total variance.

Keywords: Nifty, Factor Analysis, Multiple Linear Regression, Data reduction