Analyzing Time-Series Models for Predicting the Performance of IT Companies in Stock Market

The study compares the stock market’s predictive power to analyze forecasts, long-term pay projections, and financial commentary from small businesses. The goal of the proposed work is to evaluate how well time-series models can forecast the stock prices of six companies with strong earnings, namely the top Indian enterprises HCL, Infosys, Tech Mahindra, and Wipro. The monthly average stock prices over a 5-year period, from January 2016 to December 2020 are analyzed using designed model. The historical stock data for the four businesses (HCL, Infosys, Tech Mahindra, and Wipro) is tracked from NSE website. Each data series is differentiated with determination of the stationary followed by fitting of ARIMA models. The ideal ARIMA model is examined based on best fit parameters k, using appropriate values of AIC and BIC for autoregressive order (p), differencing order (d), and moving average order (q). ARIMA models are developed for goodness based on intermediate levels, along with considerations such as the esthetic appeal of relevant figures and the percentage error rates. Cut offs from ACF and PACF demonstrate that Infosys appears the best fit model based on AIC and BIC values.