SUMMARY
This paper proposes an early warning model for global logistics system based on principal component regression (PCR) that predicts a country’s global logistics system risk, identifies risk sources with probabilities, and suggests ways of risk mitigation. Various quantitative and qualitative global logistics indicators are utilized for monitoring the global logistics system. The Enabling Trade Index is employed to represent the risk level of a country’s global logistics system. Principal component analysis is applied to identify a small set of global logistics indicators that account for a large portion of the total variance in the original set. An empirical study is carried out to validate the predictive ability of PCR using datasets of years 2010 and 2012 published by the World Economic Forum. Furthermore, the superiority of PCR is evaluated by comparing its performance with that of a neural network with respect to the correlation coefficient and coincident rate. Finally, a real-life example of the South Korean global logistics system is presented.