oec
oec copied to clipboard
alexandersimoes
HS Data discrepancies
Greetings,
Although HS data is said to be sourced from the UN Comtrade db and cleaned by a third party, there are wide and unexplained discrepancies (≥50% in some instances) between OEC and Comtrade data. One possibility might be explained by the difference between CIF and FOB values. Can anyone please confirm if OEC values are CIF or FOB?
Kind thanks in advance
When one creates a treemap to check the imports from China or Argentina, for example, the countries concerned as importers of them self. That is, a China imports from China and an Argentine from Argentina. China is even more intriguing because almost 10% of its imports come from China itself. Discrepancies... for sure.
Avellarjr, thanks for your comments. On an aggregated macro basis, that makes sense, However, when we look at HS codes at the 4 and 6 digit level, we can see both imports and exports. Even if you assume 100% of exports are re-imported, it would not explain the discrepancies between the same HS codes being reported from the same data set.
To explain, OEC uses data from Comtrade for all data sets from 2001 to 2014. If we look at HS 8703 for the country of Ghana in 2013, for example, OEC reports the total value of imports as 651.24M. However, Comtrade reports imports as 1.093B. From OEC, total exports from Ghana of HS 8703 in 2013 totals 2.05M.
For the sake of examining the possibility of re-imports explaining the difference between data reported by OEC vs. Comtrade, and assuming that all HS 8703 exports were re-imported in 2013, it totals 653.29M. That would leave an unexplained difference of 439.71M. Given the difference, re-importing is not a plausible explanation of the discrepancy.
As posted in my original post, one and perhaps more likely explanation is that Comtrade uses two different determinants, which are FOB and CIF values. The difference between the two explains shipping, insurance, and freight costs. FOB is the actual cost of the goods themselves delivered to the port of departure. All trade is listed in current dollars in Comtrade, so discrepancies between real and current valuations of trade can likely be ruled out.
While we do know that OEC uses data that has been "cleaned" by a third-party, what we don't know is if "cleaning" means that the FOB values were stripped from CIF values to derive the OEC stated value? That would be the question, and if so, could explain the discrepancy. Any thoughts on this are appreciated. Someone familiar with the data must know the answer I assume.
Cheers
Hi
Up to this day the data you see in the OEC is provided by UN Comtrade and cleaned by BACI.
The working paper on their cleaning methodology can be found here: http://www.cepii.fr/PDF_PUB/wp/2010/wp2010-23.pdf
Hi Pachamaltese,
I suspected this was the issue. It would be helpful if OEC posted that these were estimated FOB values.
Kind thanks