Simultaneous dynamic demand estimation models for major seaports in India

Abstract

Prediction of cargo volumes is important for authorities such as port administrations and port operators, as cargo throughput affects port planning and operation. The forecasting results help planners and policy-makers to take decisions on issues such as port investment, port development, berth location selection, port operation, and freight rate. The success of the port operation policy depends on the accuracy of the cargo forecast. This paper presents an improved model known as dynamic regression (DR) model to forecast cargo demand simultaneously for ports within a port system; the model is applied to forecast cargo demand at major seaports in India. Past cargo flow data from years 1980–1981 to 2013–2014 at 11 major ports in India are used for estimating the proposed model. Average prediction error is found to be within 10% for most of the ports. The DR model is used to produce forecasts up to 2019–2020 for all the ports studied. The model forecast is compared with the projections of the Ministry of Shipping, Government of India. This study is intended to provide guidance to planners in taking decisions on issues related to port infrastructure development, such as construction of new terminals and improvement of access roads to ports for the major ports in India. The study will also be beneficial to shipping agencies for their investment strategies in the Indian ports sector.

Keywords:

• Dynamic regression model
• demand forecasting
• Indian seaport
• prediction error

Acknowledgment

The authors wish to thank the Ministry of Shipping, Mumbai Port Trust authority, Paradip Port Trust authority, and Center for Monitoring Indian Economy (CMIE) for providing the data used in this study.