Port predictions need to evolve
Graham Cox: GDP approach “ignores the effect that different sectors of the economy have on traffic flows”
Since the downturn, the world of forecasting has suddenly become much more complicated, and the industry needs to stop 'dumbling down' port traffic predictions, according to Graham Cox of Maritime Traffic Forecasts.
Firstly, it should be noted that port figures require a different approach to forecasting at the trade route level because the dynamics are different, said Mr Cox.
Traditional port forecasts often use a solely gross domestic product approach, “which is a dramatic simplification”, he explained. Under this approach, a historic relationship between cargo and overall GDP is estimated and then applied to projections. An “extraneous subjectivity” can also blur the picture when an analyst feeds in an additional multiplier based on his own views.
Mr Cox went on to say this approach “is deficient when applied at the port level if for no other reason than it ignores the effect that different sectors of the economy have on traffic flows”. For example, one industry sector may experience a downturn in activity when the overall economy remains strong: if that sector constitutes a significant proportion of national maritime trade it can have a marked effect on cargo volumes. Just as important are sectoral developments in a country’s trading partners.
Further, there is another major problem that arises from the use of fixed relationships: i.e. that the link between cargo traffic and what drives it was always the same value and will remain so. “The implicit assumption that nothing is changing is rarely justified,” said Mr Cox.
“This effective ‘dumbing down’ of GDP forecasting in the maritime industry may have occurred because of data problems and because it has become widely used as a convenient simplification for containerised traffic, where there are limited data for trade route traffic. There is no information about the nature of the cargo within the boxes, especially at the trade route level, and so simple GDP trend projection can be considered to be a justified approach.
“Unfortunately, analysis based predominantly on GDP trends right down to the port level has become the norm throughout the industry regardless of the availability of data on the economic drivers and more accurate information about the volume and make-up of the trade; even in some cases for containerised traffic (i.e. what is in the boxes). Despite the potential for improvement, the large majority of forecasts produced for the liquid and dry bulk trades follow the same simple GDP methodology typically used for the container trade.”
A different approach is taken by a few forecasters that involves interacting with individuals and organisations from sectors that are involved in import and export activities. This personal approach can run into trouble because it is explicitly or implicitly based on third parties’ unknown assumptions about external influences on traffic and a short term outlook.
The alternative to these two approaches can best be described as multi-factoral, which is a more fundamental approach to forecasting port cargoes. It recognises the importance of relationships between cargo volumes and sectoral and industry developments plus the relationships with partner countries. This approach also appraises the impact of shocks to the economy and the effect of new relationships that will follow between GDP, each component sector and other drivers and traffic volumes.
Mr Cox completes the picture with a focus on the need to take into account how certain or likely changes in structure, infrastructure and market constraints in the transport industry will affect predictions from models; “whether for a group of large ports in an advanced country or one port in a developing country”.
Images for this article - click to enlarge
Unless otherwise stated, all images copyright © Mercator Media 2012. This does not exclude the owner's assertion of copyright over the material.
