XAVIER SCHOOL OF MANAGEMENT

Home Faculty & Research Research & Publications

Research & Publications

Details of Faculty ResearchBack

Use of Uni-variate Time Series Models for Forecasting Cement Productions in India

The Article, Published by Purna Chandra Padhan in International Research Journal of Finance and Economics, 83, 2012, 167-179, 2012

In recent years, there has been a great deal of discussion on applications of various time series forecasting models and their performance in forecasting business activities. This paper discussed few of these univariate time series forecasting models and their application for forecasting cement productions in India. Applying monthly data spreading over April 1993 to March 2011, on productions of cement in lakh tons, the forecasting performance of various competing models is evaluated through forecast accuracy criteria such as mean absolute percent error (MAPE), mean squared deviation (MSD) etc. Among these models Seasonal Autoregressive Integrated Moving Average (SARIMA) model performs better than other competing models in forecasting cement productions. It provides lowest MAPE value. In fact, the model has also an advantage over other models as it explicates autoregressive and moving average process the data along with seasonality. Therefore, as a policy implication, SARIMA model can be used for forecasting cement productions in India.