Creating and assessing composite indicators: Dynamic applications for the port industry and seaborne trade

Explaining the dispersion of economy-wide fluctuations, national and global, has been long sought by economists. Short- and long-lived versions of economic fluctuations are transmitted and reflected on the port and maritime industries. Through the creation and assessment of composite indicators, utilizing non-parametric spectral domain approaches, we are able to: (a) Extend the context in which disaggregate port activity and trade volumes can be combined and used as factor-based indicators, outlining the properties of ports as regional facilitators between demand and supply, and demonstrating their capacity for reflecting national economic activity. We apply this concept in the U.S port sector, by creating Port Services Indicators based on the GDP and the IP indices, and argue for their utilization as a national port industry pulse and as timely available value-added coincident indicators. Utilizing our factor model as a forecasting device, we document favorable forecasting performance versus benchmarks. (b) Provide insight for freight rates, through the creation and assessment of composites and existing indices, assess analogies with output-related fluctuations, and reveal their temporal evolution through empirical mode decomposition and time–frequency analysis. We follow freight rates from 1741 to 2014 and explore in detail Kuznets and Kondratiev-type long-term periodic effects, business, Juglar-type investment cycles, and Kitchin inventory cycles. We assess the coherence of the empirical results with theory and maritime history.