Skip to end of metadata
Go to start of metadata

B.1.4.Conclusions and Open Questions

14.119.        This online chapter attempts to provide useful information for compilers in capturing merchanting and manufacturing services transactions. Although significant progress has been made since methodological revisions were introduced in BPM6, significant challenges remain, not least concerning emerging questions on FGPs. As an online guide, new guidance and experiences across all areas described in this chapter will be incorporated as they become available.

14.120.        These significant challenges involved in the compilation and measurement of merchanting and manufacturing services for compilers, in this section, have been addressed as 'Open questions' (and areas of required investigation). To this end, a number of the open questions have been discussed in international fora in recent years, and the outcome of these discussions, where appropriate and helpful, have been included below.

Open questions

When is a change substantive enough to change the nature of the product, in other words, when should a transaction be recorded as a manufacturing service and not merchanting? This will require investigating the scope to develop guidelines, consistent with the accounting framework, to determine when transformation (as opposed to merchanting) of a good occurs.

14.121.        The suggestion, earlier in this chapter, to use a change at the six digit HS level to differentiate transactions (between merchanting and manufacturing services) is broadly seen as a workable solution, however countries still considered that in complex transactions a case-by-case approach would be needed. In addition, there is general agreement that any change or update to international recommendations on this topic has to ensure consistency across accounts, this is considered crucial (and something that must, for example, be taken in account for the next round of updates to international statistics manuals).

What is the scope to identify FGPs? Investigate sharing techniques and methods used by countries to identify Factoryless Goods Producers.

14.122.        Given that there is currently no specific ISIC code for identifying FGPs, most compilers consider this a major hurdle in their ability to not only identify, but also track and integrate these enterprises into their compilation production systems (for example, to do so may require a special code in the business register). However, there are a limited number of countries, where FGPs play a larger role in the economy, that do identify FGPs. In general, and what appears at present to be the most common method, is by building on the work of "large cases units" involved in profiling large and complex enterprises (e.g. units focused mostly on multinational enterprises).

14.123.        Looking more medium term, countries indicate that there seems to be scope to investigate identifying FGPs, through, for example, international trade in services surveys (e.g. via R&D or intellectual property transactions), Trade in Enterprise Characteristics (TEC) data (e.g. enterprises involved and linkages to trade statistics), and closely engaging with enterprises involved in merchanting and manufacturing services. In addition, at the international level, Eurostat has established a Task Force on FGPs.

14.124.        Compilers should note that not all countries can expect to have FGPs in their economy. Given the high level of inter-connectedness and specific nature of these types of transactions, and as stated by some countries in international meetings, there is good evidence that there are no FGPs involved in the economies of some countries.

14.125.        It is worth noting recent work undertaken by Statistics Finland on global production arrangements in the country and the identification of factoryless goods producers. The push behind the work at Statistics Finland is the implementation of the UNECE Guide to Measuring Global Production recommendations.

14.126.        Statistics Finland pinpoints a number of primary and supporting sources that can help compilers identify global production arrangements, and in particular FGPs. These sources are reconciled to look for inconsistencies in reporting of global production arrangements; the enterprise is sometimes involved in the investigation through their Large Cases Unit. Once identified, Statistics Finland explains to the enterprise how they should be reporting their global production data in surveys to ensure overall consistency. The sources are outlined in the following table.

Table Add.2 Statistics Finland - Primary and supporting sources in identifying FGPs



Financial statements/annual reports of   enterprises

-Turnover from trade margin (not   deliveries of products).

-No raw material costs, high costs on   R&D, marketing, subcontracting.

-Possible information about outsourcing   production abroad.

Monthly business survey (Sales inquiry)

High share of turnover from sales abroad   to abroad

International trade in services survey

Records sales and purchases of   factory-less production

Business register

Manufacturing industry


Production - no product information

International merchandise trade   statistics

Small share of goods exports compared to   turnover

Employment statistics

Major share of company workers are   experts e.g. engineers, designers, etc.

Statistics Office's Large Cases Unit

Company visits that ask specific   questions


14.127.        In Finland at present, with a population of around 70 enterprises involved in global production, Statistics Finland estimates that there are around 15 factoryless goods producers (compared for example with around 10 enterprises undertaking merchanting transactions). There is a need to separate FGP enterprises from merchants because of the high IPP content in their products; FGPs do more than buy and sell. In implementing the recommendations the major change has been to move the recording of the FGP's margin in use tables from being a service to being an export of goods and in the supply tables reallocate the margin to 90% as the product of the good and a 10% trade margin (previously the allocation was 70% R&D services; 20% services of head offices; and 10% agency services).

What other approaches can be used to accurately record merchanting flows and what methods can be used to assure that bilateral flows are consistent? Is there a case for bilateral sharing of customs data? Exchange practices and develop guidance on merchanting transactions where goods do not physically cross borders (e.g., a French merchanter buying goods in Germany for resale in Germany).

14.128.        As previously stated in this chapter, the use of surveys was seen as the best and most common approach to capturing merchanting flows. The use of ITRS combined with surveys and direct reporting was also seen as efficient and for some countries the logical approach, but as noted, there are known issues with ITRS. However, ensuring that bilateral flows are consistent, given the methods being used, isn’t seen as something simple to implement. This is especially so due to a lack of detail being asked for in surveys around merchanting, for example a majority of OECD countries currently can't provide breakdowns of the gross flows of goods underlying merchanting transactions by product and/or (both for importing and exporting) partner country.

14.129.        The sharing of customs data is seen as possible by countries, but, and this isn't new, there are major issues around access, completeness, confidentiality and legal regulations. In addition, the sharing of customs data won't overcome the situation where the merchanter buys and sells the goods in the same non-resident country (as in the example given above).

Develop guidance on the country allocation of merchanting transactions - as such information is not currently included in relevant manuals.

14.130.        With merchanting moving from international trade in services in BPM5 into the goods account in BPM6 there is a need to consider what country the actual 'net export of goods under merchanting' transaction is allocated to (i.e. the merchant's fee). While in BPM5 when this was a service the allocation would have been the actual merchanter's country (who was providing the 'service'); with the transaction now happening in the goods account of the balance of payments this no longer makes sense (considering the goods have never entered the merchants country). So the merchant's fee needs to be allocated either to country physically exporting the goods or the country physically importing them.

Investigate asymmetries that might be caused by differences in the recording of processing transactions – for example, valuation (such as transfer pricing).

14.131.        As can be seen in the manufacturing services case study covering Costa Rica (in the Annex following), transfer pricing by enterprises can distort the values of imports and exports of goods, causing in some cases for these enterprises imports to exceed their exports. There is a need to investigate these transactions to see if such distortions are impacting on asymmetries.

14.132.        Next steps: The international community, including the OECD, will continue to examine and investigate these open questions in the search to provide good practice for compilers of manufacturing services and merchanting transactions.


Back to Implementation guide drafted by the Informal Reflection Group of the OECD Working Party on International Trade in Goods and Trade in Services Statistics (WPTGS) on Merchanting and Manufacturing Services