THE PREDICTIVE ACCURACY OF ARIMA MODELS FOR STOCK MARKET INDICES: A CASE STUDY OF KSE-100

Authors

  • Shoaib Anwer Qambrani Independent Researcher, Karachi, Pakistan
  • Dr Israr Ahmed Associate Professor Ilma University, Karachi, Pakistan
  • Abdul Basit Lecturer, KASBIT, Karachi, Pakistan

DOI:

https://doi.org/10.56596/jbem.v4i2.155

Keywords:

Time Series Forecasting, Stock Market Prediction, ARIMA Model, Pakistan Stock Exchange, Financial Modeling

Abstract

This study examines the time series predictions of the KSE-100 index using the autoregressive integrated moving average (ARIMA) model. KSE-100, which represents the Stock Exchange of Pakistan. It plays a key role in reflecting the financial status of a country. Predicting stock market indices is important for investors. Policy decision makers and economists, because it is subject to inertial fluctuations and plays a role in economic planning. The study uses daily stock data from January 2012 to December 2023, using the ARIMA model to predict future trends. Two models, ARIMA (1,0,1) and ARIMA (1,1,1), are evaluated and compared according to their performance. The results suggest that although the ARIMA (1,0,1) model is better suited to short-term forecasts, the ARIMA (1,1,1) model captures the long-term uptrend more effectively. The study highlights the limitations of the ARIMA model in capturing the nonlinear nature of stock market movements. It suggests future exploration of hybrid models that combine traditional time series methods with machine learning approaches. These findings provide information on the effectiveness of the ARIMA model in financial forecasting. It has an impact on investors and policymakers who aim to reduce risk and optimize investment strategies.

Downloads

Published

2025-06-30

How to Cite

THE PREDICTIVE ACCURACY OF ARIMA MODELS FOR STOCK MARKET INDICES: A CASE STUDY OF KSE-100. (2025). Journal for Business Education and Management, 5(1), 1-22. https://doi.org/10.56596/jbem.v4i2.155