Accuracy of ARIMA Model for Individual Stocks of Nifty 50 – Sector Wise

Abstract

The Autoregressive Integrated Moving Average (ARIMA) model stands out as a robust tool for stock price prediction within the Nifty 50 index, a pivotal gauge of the Indian stock market. This research delves into the efficacy of the ARIMA model within the Nifty 50 index framework. Specifically, an exhaustive analysis is conducted on historical price dynamics exhibited by constituent equities of the index over a four-year period from January 1, 2018, to February 28, 2023. Our findings indicate that ARIMA offers the most precise predictions for shorter timeframes, notably the 30-day forecast, followed by the 60-day forecast, and subsequently the 180-day forecast. Moreover, an evaluation of ARIMA model accuracy across various industries is carried out through metrics such as root mean square error (RMSE), Mean Absolute Error (MAE), and mean absolute percentage error (MAPE). In summary, this study suggests the viability of employing ARIMA models to forecast future values of Nifty50 stocks within specific sectors, especially for short-term predictions. It also underscores the need for further exploration in tailoring forecasting methodologies to suit the distinctive attributes of diverse industries.