Measuring dynamic competitiveness among container ports: an autoregressive distributed lag approach

This paper proposes a method for applying the autoregressive distributed lag (ARDL) approach to the analysis of port competitiveness to overcome the methodological limitations of the commonly used time series approach, which may cause inconsistent orders of integration or large standard errors. An empirical analysis based on the ARDL approach was also conducted to investigate the competitive relationships among the top five worldwide container ports: Shanghai, Singapore, Hong Kong, Shenzhen, and Busan. Their long-term relationships were first evaluated by the bounds testing procedures with three alternative models. Then the long-term ARDL models were developed to understand their long-term equilibriums. In addition, we estimated short-run dynamics by the ARDL model's error correction form. Our approach can provide maritime transport practitioners a practical tool for reliable port competition analysis and container demand forecasting.