## Appendix 1.1 — GZTT consultation on the CIF‑FOB valuation of imports and exports (support-document-draft-guidance-note-g1)

## Source details

**Canonical URL:** [Appendix 1.1 — GZTT consultation on the CIF‑FOB valuation of imports and exports (support-document-draft-guidance-note-g1)](https://www.imf.org/-/media/files/data/statistics/bpm6/gztt/support-document-draft-guidance-note-g1.pdf)

## Other formats

- [Markdown version](/-/media/files/data/statistics/bpm6/gztt/support-document-draft-guidance-note-g1.pdf.md)
- [Structured JSON version](/-/media/files/data/statistics/bpm6/gztt/support-document-draft-guidance-note-g1.pdf.json)

---

### Introduction, mandate, and scope
- Task Force: Task Force on the SNA Research Agenda – Globalization Task Team (GZTT) prepared a compendium to support the Guidance Note on CIF‑FOB valuation of imports and exports.
- Drafting and review:
  - Primary drafter: Ms. Margarida Martins.
  - Reviewers (IMF secretariat – GZTT): Ms. Padma Hurree‑Gobin and Ms. Jennifer Ribarsky.
  - Co‑chairs: Michael Connolly (Central Statistics Office Ireland) and Paul Roberts (Australian Bureau of Statistics).
- Objective: examine internal consistency of valuation of trade in goods within the SNA and harmonization with BPM and IMTS; address tension between 2008 SNA valuation principles (output at basic prices; transaction/observed prices) and BPM6 uniform FOB valuation at the exporting country’s customs frontier.

### Historical context and conceptual contrast
- Historical notes:
  - Issue surfaced in 1996 (EU trade asymmetries); invoice/transaction value approach proposed in 1999 EU technical report but not pursued earlier due to inconsistency with BPM5.
  - Recent revival: Haan and Hiemstra (2018), Walter (2018); Anne Harrison raised issue at IMF BOPCOM in 2013.
- Key conceptual contrast:
  - 2008 SNA (e.g., 2008 SNA 14.68) records cross‑border trade at amounts specified between buyer and seller and treats transport depending on responsibility.
  - BPM6 (10.30–10.31) prescribes a uniform FOB valuation at the exporter’s customs frontier irrespective of varying amounts of distribution costs included in transaction prices.
  - Merchanting (BPM6 10.44) uses transaction prices as agreed by parties; freight/insurance associated with merchanting may be embedded or separately accounted.

### Empirical observations cited
- Harrison analysis: freight debits to imports FOB ratios stable; ratio of recorded freight debits to freight credits averages about 140 percent a year over 17 years examined; the 40 percent excess represents about one percent of imports FOB. Eliminating this excess could reduce asymmetries between world exports and imports from one to two percent to one percent or less.

### Options for addressing CIF‑FOB valuation tension (Anne Harrison)
- Option 1: Change SNA recommendation on domestic transportation back to pre‑1993 treatment (transportation always treated as a service and never integrated with the value of the good).
- Option 2: Change SNA recommendation to be strictly consistent with BPM6 for transactions with non‑residents.
- Option 3: Keep SNA and BPM as they are but explain inconsistencies via a supplementary table showing CIF→FOB conversion and decomposition (rerouting vs reclassification from goods to services).
- Option 4: Amend BPM guidelines to adopt valuation on a true transaction or actually observed price basis (invoice/transaction values).

### Consultation questions posed to GZTT members
- Views requested on:
  - (i) substance of the issue note;
  - (ii) preferred option among Options 1–4 and reasons for rejecting others;
  - (iii) related data sources and country/regional practices and feasibility of implementation (questionnaire in Appendix 1.1).

---

### Aggregate consultation participation and preferences
- Participation:
  - 15 out of the 17 GZTT members (excluding international organizations) provided feedback.
  - Additionally, two international organizations provided feedback.
- Aggregate option preferences (countries; totals preserved as presented):
  - Option 4 selected by 9 (in a total of 15 answers from countries).
  - Option 1 selected by 1.
  - Option 2 selected by 1.
  - Option 3 selected by 1.
  - NA 6 - 3 (as reported in Table 1).
- International organizations:
  - Both considered Option 4 preferred; one agency indicated Option 1 could also be a solution.

### Reasons offered in favor of Option 4 (transaction/invoice valuation)
- Conceptual advantages:
  - Considered conceptually sound and closest to the basic price concept.
  - Avoids compiling CIF→FOB adjustment often based on assumptions and lacking source data.
  - Expected to reduce bilateral asymmetries by unifying valuation between importers and exporters.
  - Would facilitate compilation of Supply and Use Tables (SUT).
  - Respondents noted both BPM and SNA guidelines should be amended and harmonized based on transaction cost.
- Caveats and practical concerns:
  - Need for new data sources and legal/regulatory changes.
  - Difficulty obtaining accurate transaction values even when source data exist.
  - Necessity to harmonize with IMTS and WTO recommendations.
  - Potential for implementation difficulty where contract values and customs values diverge because of taxation or confidentiality.

### Views and critiques of Options 1–3 (summary)
- Option 1 critiques:
  - Does not solve CIF→FOB valuation problems or SNA–BPM inconsistencies; requires estimating transport services without available information, leading to imputations.
- Option 2 critiques:
  - Would create internal SNA inconsistencies and imputations; impractical for implementation in many contexts.
- Option 3 critiques:
  - Seen as pragmatic by some but would not solve compilation problems, could perpetuate asymmetries, and might increase compiler workload; some regarded Option 3 as a medium‑term transition.

---

### Empirical reconciliation exercises and stocktaking surveys

### IMF‑OECD reconciliation exercise (pilot)
- Participation: eighteen countries agreed to participate in a reconciliation exercise comparing international trade and invoice data.
- Selected country results (as summarized from the BOPCOM paper):
  - Albania (2016–18): annual average ratios of invoice values to BOP trade estimated between 5.5 and 5.9 percent.
  - Belgium (intra‑EU trade, 2015): invoice value to CIF adjustment was 0.04 and 0.16 percent of total trade for imports and exports respectively; CIF→FOB adjustment was -1.67 percent of total trade for imports.
  - Indonesia (2014–18): comparison of FOB exports from customs data and bank records of export proceeds revealed a difference of 13 to 14 percent (administrative fees, discounts, subcontracting processing cited).
  - Kosovo (2018): customs vs survey of main trade corporations showed small differences generally; customs may not provide reliable estimates for freight and insurance in specific cases.
  - Moldova (Q1 2019): average ratios of invoice values to BOP trade around 5 percent with significant fluctuations across trade partner groups.

### Survey and compendium findings
- 2019 IMF‑OECD WPTGS Stocktaking Questionnaire:
  - Around 65 countries participated.
  - Findings: around 50 percent unfavorable to using invoice values, 20 percent favorable, 30 percent unsure.
  - Initial reconciliation exercises showed small differences between FOB and invoice values in pilot cases; many countries expressed concerns about practical implementation.
- UNSD 2016 Decennial IMTS survey:
  - 68 in 102 economies maintained invoice price as one valuation in basic merchandise trade statistics.
- Country reported CIF→FOB ratios (selected examples from institutional studies):
  - Cambodia: 8 percent.
  - PR China: 5 percent.
  - India: 5 percent.
  - Mongolia: 27 percent.
- Norway (detailed study): overall merchandise exports difference around 2.4 percent in 2018; product variations: fish (CPA 03) ~6 percent; crude oil and natural gas ~15 percent.

---

### Practical considerations, data availability, and quality (follow‑up GZTT consultation results)

### Second GZTT consultation (summary participation)
- Respondents: 22 national agencies and 2 international organizations.
- Access to invoice data from customs (Question 1): Yes 17, No 3, Other 2.
- Access by compiling agencies to transaction values (Question 2): Yes 13, No 7, Other 2.
- Availability from other data sources (Question 3): Yes 3, No 17, Other 2.
- Perceived accuracy of transaction data (Question 4): Yes 16, No 3, Other 3.
- Would transaction values reduce asymmetries (Question 5): Yes 16, No 4, Other 2.
- Would transaction values improve national accounts accuracy (Question 6): Yes 12, No 7, Other 3.
- Have agencies analyzed invoice vs CIF‑FOB differences (Question 7): Yes 7, No 15.

### Key practical findings and hurdles
- Invoice value availability:
  - Many countries report invoice values available in customs documentation, especially intra‑EU trade; extra‑EU availability varies.
  - 13 of 22 respondents confirm compiling agencies have access to transaction values but often with limitations (aggregates only, confidentiality restrictions, IT constraints).
  - Alternative sources (business surveys, revenue authority data) limited; 3 respondents indicated other sources.
- Data quality concerns:
  - 16 respondents consider transaction values accurate for their country; some report concerns at detailed tariff‑code level or for Extra‑EU trade.
  - Divergences between corporate invoices and customs declarations cited: under‑declaration, transfer pricing, incomplete information, zero‑cost warranty replacements.
- Impact on asymmetries and NA accuracy:
  - 16 respondents expect adoption of transaction values likely to reduce asymmetries; 12 expect significant improvement in national accounts accuracy (SUT balancing), but several respondents cautioned persistent issues (classification, timing, residence determination) may remain.
- Empirical studies limited:
  - 15 of 22 countries reported no study on invoice vs CIF‑FOB differences (reasons: lack of access; resource constraints).
  - Only one respondent quantified the difference across exports at around 2 percent in 2018 (Norway).

### IMTS and customs framework observations (EU focus)
- IMTS 2010 and WTO Customs Valuation:
  - IMTS recommends statistical value: exports FOB‑type, imports CIF‑type; WTO Agreement adopts transaction value as customs value when conditions met.
  - Transaction value may include/exclude transport, loading/unloading, insurance depending on delivery terms and national customs practice; compilers must adjust invoice price by delivery terms to derive FOB or CIF statistical values.
- Intrastat and single market implications:
  - Intrastat collects invoice values for intra‑EU trade; extra‑EU customs declarations and SAD forms may allow invoice values but availability and consistent invoice currency vary by country.

---

### Conceptual and compilation implications of adopting transaction/invoice values

### Conceptual shifts and accounting effects
- Redefinition of goods vs services borders:
  - Mixed composition of goods and services in invoice values would require redefining distinction between goods and transport/insurance services; potential to collapse goods and merchandise transport into a single aggregate considered.
- Time‑series and reclassification effects:
  - Delivery‑term changes could cause merchandise and transport series to fluctuate across periods (example: an import CIF in year t becoming import FOB in year t+1 could alter series for merchandise and transport between t and t+1).
- Residence‑of‑provider issues:
  - Customs declarations often do not identify residence of carrier; options include exceptions to carrier residence concepts for trade statistics or new data collection from carriers/manufacturers.
- Valuation and risk differences:
  - Transaction values would measure different risks and costs by delivery terms; point‑of‑uniform‑valuation concept would no longer apply and services estimation may still be required.

### Data‑collection, legal, IT, institutional and resource implications
- Overhaul required for merchandise and transport cost data collection:
  - New surveys among manufacturers and transport/insurance corporations or extension of existing surveys likely needed, particularly to cover debit side freight/invoice data.
  - Resource implications: cost and political resistance ("in most cases would be strongly resented by policy makers"); high respondent burden; likely high non‑response and delays in reporting affecting timeliness.
- Legal/regulatory:
  - New legal requirements and regulatory support needed to collect microdata from private corporations, notably non‑financial corporations; regional legal acts (e.g., EU) may require updates.
- IT and capacity:
  - Many countries lack capacity to modernize systems to build alternative sources (e.g., business financial statements); where invoice values already available in Customs, only minor compilation changes may be necessary.
- Institutional cooperation:
  - Enhanced interagency cooperation (Customs, NSO, Central Bank) essential to secure access and reduce duplication; confidentiality and access restrictions common.

### Quality control and series continuity
- Quality issues:
  - Lack of standard invoice formats and divergences between invoices and customs declarations complicate data use.
  - Many countries perform limited quality control on Customs variables.
- Series continuity:
  - Adoption of invoice values would likely cause breaks in series; retrospective estimation "highly unlikely", implying breaks in time series and costs for data users.

---

### Recommendations, staged approaches and next steps identified by respondents and GZTT
- Investigate practical aspects before any change:
  - Need to study availability and access to invoice data across regions, feasibility in developing economies, and legal/IT constraints.
  - Pilot studies recommended:
    - One respondent recommended a pilot for non‑EU member countries.
    - Several respondents noted EU feasibility due to intra‑EU transaction data but cautioned guidance must reflect all regions.
- Suggested medium‑term transition options:
  - Option 3 (supplementary CIF→FOB reconciliation table) proposed by some as pragmatic medium‑term measure while investigating Option 4 practicality.
- Reconciliation and bilateral exercises:
  - Foster bilateral reconciliation exercises with main trade partners and explore Unique Consignment Reference (WCO) to reduce asymmetries.
- Guidance and standards work:
  - Further coordination between SNA, BPM, and IMTS communities; consider updates to 2008 SNA paragraphs (e.g., 3.149, 14.61–14.77, 26.19) and BPM6 guidance in light of GZTT findings, subject to ISWGNA/AEG/BOPCOM consultation.
- Data improvement actions:
  - Improve CIF→FOB compilation methods and update coefficients frequently at detailed product, mode, and partner level where CIF→FOB retained.
  - Consider stepwise or bottom‑up approaches starting with source improvements in Customs and IMTS.

---

### Key numerical and factual highlights (preserved as presented)
- GZTT response counts: 15 of 17 members (countries) provided feedback; 2 international organizations provided feedback.
- Option 4 preference among countries: 9 out of 15 country responses.
- Second GZTT consultation respondent totals: 22 national agencies and 2 international organizations.
- Access to invoice values (Question 1): Yes 17, No 3, Other 2.
- Access by compilers (Question 2): Yes 13, No 7, Other 2.
- Perceived accuracy (Question 4): Yes 16, No 3, Other 3.
- Likely reduction in asymmetries with transaction values (Question 5): Yes 16, No 4, Other 2.
- Improvement in national accounts accuracy (Question 6): Yes 12, No 7, Other 3.
- Analysis of invoice vs CIF‑FOB performed (Question 7): Yes 7, No 15.
- IMF‑OECD stocktaking participation: around 65 countries.
- UNSD IMTS 2016: 68 in 102 economies maintained invoice price as one valuation.
- Pilot reconciliation country example results:
  - Albania: invoice/BOP ratios 5.5–5.9 percent (2016–18).
  - Belgium (intra‑EU 2015): invoice→CIF adj 0.04% (imports) and 0.16% (exports); CIF→FOB adjustment -1.67% of total trade (imports).
  - Indonesia: FOB exports vs bank records difference 13 to 14 percent (2014–18).
  - Moldova (Q1 2019): invoice/BOP ratios around 5 percent.
- Country CIF→FOB sample ratios (ADB compendium): Cambodia 8 percent; PR China 5 percent; India 5 percent; Mongolia 27 percent.
- Norway study: overall exports difference ~2.4 percent in 2018; product examples: fish ~6 percent; crude oil and natural gas ~15 percent.
- Harrison historical freight ratio finding: freight debit to freight credit ratio averages about 140 percent over 17 years; the 40 percent excess ≈ about one percent of imports FOB.

*Source: Task Force on the SNA Research Agenda – Globalization Task Team (GZTT), Supporting Document to the Guidance Note on Valuation of imports and exports of goods in the international standards (CIF to FOB adjustment).*

### Appendix 1.1 GZTT consultation on the issue note on CIF-FOB valuation of imports and exports:

### Appendix 1.1 GZTT consultation on the issue note on CIF-FOB valuation of imports and exports

### Introduction and mandate
- The Task Force on the SNA Research Agenda – Globalization Task Team (GZTT) prepared a compendium to support the Guidance Note on CIF-FOB valuation of imports and exports.
- The document was prepared by Ms. Margarida Martins (primary drafter) and reviewed by Ms. Padma Hurree-Gobin and Ms. Jennifer Ribarsky (IMF secretariat – GZTT), drawing on inputs from GZTT members and consultations with the Current Account Task Team (CATT) under IMF’s Committee on Balance of Payments Statistics (BOPCOM).
- The GZTT was co-chaired by Michael Connolly (Central Statistics Office Ireland) and Paul Roberts (Australian Bureau of Statistics).

### Scope of the issue
- The 2008 SNA recommends recording imports and exports of goods at free-on-board (FOB) values; this is consistent with BPM6 valuation principles.
- A noted tension exists between 2008 SNA valuation principles (output at basic prices and transaction or observed prices) and the BPM6 uniform FOB valuation at the exporting country’s customs frontier.
- The GZTT was tasked to address internal consistency of valuation of trade in goods within the SNA and harmonization with BPM and IMTS.

### Historical and recent developments
- The CIF-FOB valuation issue first arose in 1996 in the context of EU trade asymmetries; a technical EU group proposed an invoice approach in a 1999 report.
- The invoice/transaction value approach was not pursued earlier because of inconsistency with BPM5.
- Recent contributions include Haan and Hiemstra (2018) and Walter (2018), who revived the invoice/transaction value approach in forums including AEG 2017 and OECD Working Parties 2018.
- Anne Harrison (editor of 2008 SNA) raised the issue at the IMF BOPCOM meeting held in 2013; at that time IMF Statistics Department did not favor changes to BPM6.

### Empirical observations cited
- Analysis by Anne Harrison observed two points:
  - Ratios of freight debits to imports FOB are quite stable, suggesting the CIF-FOB adjustment is based on long-established proportions.
  - The ratio of recorded freight debits to freight credits averages about 140 percent a year over the 17 years examined, with the 40 percent excess representing about one percent of imports FOB.
  - Eliminating this excess could reduce asymmetries between world exports and imports from one to two percent to one percent or less.

### Conceptual contrasts between SNA and BPM6
- 2008 SNA (2008 SNA 14.68) states whether the value of goods covers transport depends on whether exporter or importer is responsible for transport; cross-border trade should be recorded at amounts specified between buyer and seller.
- BPM6 (BPM6 10.30–10.31) prescribes market value of goods at the exporting country’s customs frontier (FOB) as the uniform valuation, even if transaction prices include varying amounts of distribution costs (wholesaling, transport, insurance, taxes).
- Merchanting entries in the Balance of Payments are valued at transaction prices as agreed by the parties, not FOB (BPM6 10.44); freight and insurance associated with merchanting may be embedded in sale value or accounted for separately depending on delivery terms.
- 2008 SNA recognizes customs declarations are not necessarily ideal for national accounts or BOP (2008 SNA 14.69) and notes differences in recording when the importer contracts delivery from non-resident carriers (2008 SNA 14.72).

### Practical concerns raised
- Haan and Hiemstra (2018) and Walter (2018) argue that constructed CIF-FOB adjustments for exports and imports can be inconsistent with data collected on international transport services; without information on residency of the carrier, imputations may be improper and the CIF to FOB adjustment diverging from actual transaction value can easily lead to mistakes in the trade balance.
- Developments in global value chains and containerization mean business practices in the transport industry may have changed since BPM’s original assumptions; seamless transport via containers may have reduced the prevalence of unloading at the exporter’s border and reloading onto non-resident cargoes.

### Options identified by Anne Harrison to improve consistency between BPM and SNA
- Option 1: Change SNA recommendation on domestic transportation back to pre-1993 treatment—transportation always treated as a service and never integrated with the value of the good.
- Option 2: Change SNA recommendation to be strictly consistent with BPM6 when transactions with non-residents are concerned.
- Option 3: Keep SNA and BPM as they are but explain inconsistencies via a supplementary table showing how imports CIF are converted to imports FOB, including decomposition of rerouting and reclassification from goods to services.
- Option 4: Amend BPM guidelines to adopt valuation on a true transaction or actually observed price basis (invoice/transaction values).

### Consultation questions for GZTT members
- Views were solicited on the four options put forward by Anne Harrison.
- Members were asked whether they support valuation of goods on a true transaction or actually observed price basis for international trade.
- Members were asked whether practical aspects of the transaction/observed price approach should be investigated, such as availability of source data.

### Documentation and structure of the supporting compendium
- The supporting document is organized in four Sections:
  - Section 1: Draft issue note and GZTT consultation from a conceptual standpoint; balance of payments perspectives included.
  - Section 2: Follow-up GZTT consultation focusing on practical considerations, survey of country practices, comparisons across consultations; practical BOP considerations included.
  - Section 3: Summary of ongoing discussion since the last SNA and BPM update; review of current recommendations in 2008 SNA and BPM6.
  - Section 4: Identification of 2008 SNA paragraphs to be updated according to GZTT recommendations.

_Source: Task Force on the SNA Research Agenda – Globalization Task Team (GZTT), Supporting Document to the Guidance Note on Valuation of imports and exports of goods in the international standards (CIF to FOB adjustment)._

### 19. A consultation of the GZTT sought the members’ views on the issue paper. The views of the

### support-document-draft-guidance-note-g1 - 19. A consultation of the GZTT sought the members’ views on the issue paper. The views of the

### Summary of consultation process and questions posed
- The GZTT consultation requested members’ views on:
  - (i) the substance of the issue note,
  - (ii) options put forward for consideration – their preferred option as well as reasons for rejecting the others, including any other option that might have been missed in the paper,
  - (iii) related data sources and country/regional practices (the questionnaire is included in Appendix 1.1).
- Four options were proposed for improving consistency between BPM and SNA on the treatment of freight transport (and insurance costs) and the valuation of traded goods:
  - Option 1: The SNA could change its recommendation on the treatment of domestic transportation back to what it was before the 1993 revision. Under the past treatment, transportation is always treated as a service and never integrated with the value of the good.
  - Option 2: The SNA recommendation could, if necessary, be changed to be strictly consistent with BPM6 when transactions with non-residents were concerned.
  - Option 3: The SNA and BPM could stay as they are but with the inconsistencies explained by a supplementary table showing how imports CIF are converted to imports FOB. Such a table should explain how much of the difference is rerouting and how much is a reclassification from goods to services.
  - Option 4: It can be considered to amend the BPM guidelines and adopt the valuation on a true transaction or actually observed price basis.

### Participation and overall preference
- Participation:
  - 15 out of the 17 GZTT members (excluding the international organizations) provided feedback to the consultation.
  - Additionally, two international organizations provided feedback.
- Aggregate option preferences (countries; totals preserved as presented):
  - Most respondents: Option 4 selected by 9 (in a total of 15 answers from countries).
  - Option 1 selected by 1.
  - Option 2 selected by 1.
  - Option 3 selected by 1.
  - NA 6 - 3 (as reported in Table 1).
- Two international organizations provided feedback; both considered Option 4 as preferred, and for one agency Option 1 could also be a solution.

### Reasons given by respondents in favor of Option 4 (valuation on a true transaction/observed price basis)
- Considered conceptually sound.
- Would avoid compiling the CIF to FOB adjustment, often based on assumptions and lacking source data.
- Expected to reduce asymmetries in data by unifying transaction value assessment between importers and exporters.
- Would facilitate compilation of Supply and Use Tables (SUT).
- Respondents noted both BPM and SNA guidelines should be amended and harmonized based on transaction cost.
- Respondents expect an impact on balance of payments data.
- Caveats and practical concerns cited:
  - Need for new data sources.
  - Difficulty in obtaining accurate transaction values even when data sources exist.
  - Need to harmonize with other statistical manuals (notably international merchandise trade statistics and WTO recommendations).
  - One respondent viewed the solution as more relevant in the EU; another recommended a pilot study for non-EU member countries.
  - Implementation difficulty when contract and customs values diverge significantly because of taxation.

### Views and critiques of Options 1–3
- Option 1 (SNA revert to pre-1993 treatment of domestic transportation):
  - Selected by 1 respondent.
  - Critics: Does not solve current CIF-FOB valuation problems or the inconsistency between BPM and SNA; would require estimating transport services without available information, leading to data imputations.
- Option 2 (SNA strictly consistent with BPM6 for non-residents):
  - Selected by 1 respondent.
  - Critics: Does not solve CIF to FOB adjustment problems, leads to imputations, impractical to implement; some view BPM concepts may also need updating.
- Option 3 (retain SNA and BPM but add supplementary CIF-to-FOB conversion table):
  - Selected by 1 respondent.
  - Supporters view it as pragmatic; note challenges collecting invoice data with product detail due to increasing use of containers.
  - Critics: While useful to explain differences between frameworks, it would not solve CIF-to-FOB compilation problems, could perpetuate asymmetries, and might increase work burden for compilers. Two respondents suggested Option 3 could be a medium-term solution.

### Balance of payments perspective: conceptual considerations and evidence
- Rationale for current FOB-type values:
  - FOB-type values provide a uniform valuation basis at the exporting country's border and have been standard practice since the 1950s.
  - Customs records have been the major source for IMTS and related aggregates in balance of payments and national accounts.
- Drivers for review:
  - Increasing use of non-customs data sources, particularly in customs unions (e.g., EU).
  - Practical challenges in reconciling IMTS (cross-border registration) with FOB (change in ownership) resulting in persistent asymmetries.
  - Globalization and bundled goods/services transactions raising need for additional clarifications.
  - Evolving user needs for more detailed and timely data.
- Findings from the 2019 joint IMF-OECD initiative:
  - Around 65 countries participated in a WPTGS stocktaking questionnaire that included invoice values in balance of payments.
  - Initial views showed little global support for using invoice values and concerns about practical implementation; many advocated maintaining the current FOB valuation while supporting collaboration between balance of payments and national accounts communities.
  - Many respondents questioned whether CIF to FOB adjustment is the primary source of large asymmetries; other contributors include recording imports by country of consignment rather than origin, merchandise not crossing borders, shuttle trade, and illegal activities.
  - Reconciliation exercises among voluntary pilot countries showed small differences between FOB and invoice values; invoice values may not necessarily depart significantly from corresponding FOB values.
- Practical hurdles to adopting invoice/transaction values:
  - Difficulty replacing the rich, frequent, and timely customs data source in the short to medium term.
  - Alternative data sources would likely be costlier for compiling agencies and reporters.
  - Use of different source data by partner countries could increase asymmetries and raise data quality concerns.
  - Legal, organizational, and institutional aspects require careful consideration.

### Conceptual and practical implications of adopting transaction values for balance of payments
- Redefinition of borders between goods and services:
  - Mixed composition of goods and services in import/export values would require redefining the distinction between goods and transport/insurance services.
  - Time series effects: delivery-term changes could cause merchandise and transport series to fluctuate across periods (example: an import CIF in year t becoming import FOB in year t+1 could alter time series for merchandise and transport between t and t+1).
  - Possible consideration to collapse goods and merchandise transport into a single aggregate.
- Residence-of-provider challenges:
  - Customs records may not identify the residence of freight/insurance providers.
  - Options: introduce exceptions to carrier residence concepts for trade statistics or collect additional information from domestic carriers and manufacturers (e.g., sample surveys with breakdown by country of residence), potentially increasing bilateral asymmetries.
- Valuation and risk differences:
  - Merchandise and freight data would measure different risks and costs by delivery terms; the point-of-uniform-valuation concept would no longer apply.
  - Services costs (freight, insurance) would still require estimation even if transaction values are used, possibly necessitating centrally provided uniform guidance on freight and insurance rates to reduce asymmetries.
- Data quality and institutional considerations:
  - Lack of standard invoice formats across corporations would complicate invoice data collection; standard formats might be required.
  - Divergences between corporate invoices and Customs declarations can be caused by under-declaration to avoid taxes, transfer pricing, incomplete information, zero-cost warranty replacements, and other inaccuracies.
  - Many countries perform no quality control on Customs variables.
  - Moving away from the established system would require changes across statistical domains to ensure consistency among business statistics, IMTS, balance of payments, and national accounts.
  - Better interagency cooperation in data collection, sharing, and validation could reduce duplication, minimize costs, and improve trade estimate quality.
  - A bottom-up approach has been suggested by some countries, starting changes at the source (Customs records and IMTS). If balance of payments and national accounts adopt invoice values while IMTS continue using CIF/FOB, analogous asymmetry issues may persist with a different adjustment replacing CIF to FOB.

### Requests and next steps indicated by respondents
- Further consultations needed on practical considerations related to Option 4, specifically:
  - (i) need for new data sources,
  - (ii) difficulty obtaining accurate transaction values even when sources exist,
  - (iii) need to harmonize with other manuals (IMTS, WTO).
- Suggested pilot studies:
  - One respondent recommended a pilot study for non-EU member countries.
  - Several respondents noted feasibility in the EU due to availability of intra-EU transaction data, but cautioned guidance must reflect all regions.
- Emphasis on investigating availability of source data in developing economies and countries with lower statistical capacity.

*Source: Consultation of the GZTT and supporting document to the IMF Guidance Note on valuation of traded goods and related services (IMF).*

### Appendix 1.1 GZTT consultation on the issue note on CIF-FOB valuation of imports and exports:

### Appendix 1.1 GZTT consultation on the issue note on CIF-FOB valuation of imports and exports

### Purpose and reviewer role
- The GZTT consultation asked members to review the issue note and consider the recommendations below.
- Reviewers were asked to bring their perspective as either National Accounts or Balance of Payments compilers to ensure both communities are considered.

### Directives to reviewers
- Provide comments on the substance and content of the issue note.
- Do not provide edits to the text. The issue note will serve as input into a draft guidance note that will describe the issue and consider data sources and country practices. The draft guidance note will reflect the opinions of the Task Team. The secretariat and co-chairs will conduct the editorial revision of the draft guidance note at a later stage.

### Considerations when evaluating the four options in the "Points of discussion" section
- Provide your opinion on the most preferred option specifying the reasons why.
- Provide your opinion on why the other options should not be considered.
- Describe other options that are missing from the note but in your opinion should be considered.
- Distinguish between what in your opinion is conceptually correct versus what is feasible to implement. If these two considerations do not coincide (i.e., what you think is conceptually correct but not feasible to implement), then is there a second-best solution?

*Appendix 1.1 GZTT consultation on the issue note on CIF-FOB valuation of imports and exports*

### Appendix 1.2 GZTT consultation on the issue note on the CIF-FOB valuation of imports and exports: detailed responses

### Appendix 1.2 GZTT consultation on the issue note on the CIF-FOB valuation of imports and exports: detailed responses

### Overview of consultation outcomes
- Respondents from national statistical offices, central banks, and international institutions evaluated four options (Options 1–4) for addressing CIF-FOB valuation issues in balance of payments (BOP) and national accounts (SNA).
- A plurality of respondents expressed support for Option 4 (valuation of goods on a true transaction or actually observed price basis, often described as "invoice values" or "transaction prices").
- Several respondents recommended investigation of practical aspects, including availability of source data and implications for BOP, SNA, and national compilation systems.

### Most preferred option by respondent (selected responses)
- AUSTRALIA — Mr. Paul Roberts (ABS): Option 4
  - "We would support work on this (Option 4) as our preference."
- BRAZIL — Mr. Roberto Ramos (IBGE, Retired): Option 1 (with conditional support for Option 3 as transition)
  - Preference for keeping services identified without incorporation into exports/imports (basic prices).
  - Concern about "changing the rules of the game"; suggests medium-term implementation and possible transition recommendation supporting Option 3 with supplementary tables if definitions are unclear.
- CANADA — Ms. Jennifer Withington (STATCAN): Option 2
  - Transport should be considered a service and measured separately; customs documents mainly show FOB values; exports valued FOB place of exit; import "Value for duty" close to FOB point of exit.
- CHINA — Ms. Yang Can (SAFE): Option 4
  - Supports invoice value for merchandise trade to reduce data asymmetry and estimation from CIF to FOB; pragmatic recognition of expected impacts on transportation service treatment in BOP.
- GERMANY — Mr. Jens Walter (Deutsche Bundesbank): Option 4
  - Supports valuation of goods on a true transaction or actually observed price basis; current methods create bilateral asymmetries.
- IRELAND — Mr. Michael Connolly (Central Statistics Office): Option 4
  - Agrees with proposals; notes recent papers highlighting inconsistencies and data issues with CIF-to-FOB adjustments.
- LUXEMBOURG — Mr. Paul Feuvrier (Central Bank of Luxemburg): NA (Neutral)
- MALAYSIA — Ms. Norhayati Razi (Central Bank of Malaysia): NA (commented)
  - Sees merit in further study, recommends pilot studies for non-EU countries if Option 4 pursued.
- MOROCCO — Ms. Lamia Laabar (Haut‑Commissariat au Plan): Option 3
  - Views Option 3 as most pragmatic; notes 2008 SNA integrates imports CIF but requires adjustment to FOB coherent with BPM6; highlights costly CIF-to-FOB adjustment and data source limitations.
- NETHERLANDS — Mr. Mark de Haan (Statistics Netherlands): Option 4
  - Considers Option 4 best; notes SNA guidance allowing CIF and FOB valuation would also need amendment.
- NORWAY — Ms. Ann Lisbet Brathaug (Statistics Norway): Option 4
  - Prefers Option 4 using "invoiced values" to avoid asymmetries and weak assumptions in CIF-to-FOB adjustments.
- RUSSIA — Mr. Andrey Tatarinov (ROSSTAT): Option 4
  - Supports invoice-based approach; unified assessment reduces international inconsistencies; flags implementation difficulties when contract price and customs value differ.
- UNITED STATES — Mr. Dylan Rassier (BEA): Option 4 (contingent)
  - Views Option 4 as best conceptually, contingent on consultation with U.S. Census Bureau regarding customs invoice data availability.
- URUGUAY — Ms. Lourdes Erro (Central Bank of Uruguay, Retired): Option 4
  - Supports amendment to adopt invoice values for merchandises and freight transport for conceptual consistency with SNA and BPM; notes need for customs data access and confidentiality arrangements.
- ECB — Mr. Celestino Giron: Option 1 or 4
  - Prefers full consistency within the overall system; reduces options to 1 (amend SNA) or 4 (amend BPM).
- EUROSTAT — Mr. Steinar Todsen: Option 4
  - Supports transaction prices as conceptually correct and aligned with SNA 2008 valuation principles; notes SNA and BPM amendments required.

### Reasons given against other options (summary of key arguments)
- General objections to Option 1 (treat transport as separate service and keep goods at basic price)
  - Leads to measurement issues because separately invoiced transport services are not universal; requires many imputations; may misrepresent transport industry output; can generate negative values for imports of transport services in input-output tables (Australia).
  - In some national contexts (Ireland, Norway), Option 1 is not viable due to harmonization with domestic treatment and impractical imputations.
- General objections to Option 2 (change SNA treatment to align border valuations)
  - Would introduce inconsistencies between domestic and cross-border accounting treatments; creates imputations and is not aligned with current SNA conceptual recommendations (United States, Netherlands, Norway).
- General objections to Option 3 (supplementary reconciliation tables)
  - Does not resolve core valuation inconsistencies or measurement problems in CIF-to-FOB adjustment; may increase burden on compilers and leave bilateral asymmetries (Germany, Netherlands, United States).
- Specific methodological concerns
  - CIF-to-FOB adjustments often rely on weak or missing source data, producing measurement errors (Eurostat, Netherlands).
  - Existing problems obtaining correct CIF and FOB values and computing CIF-to-FOB adjustments cause imbalances within National Accounts and trade asymmetries between countries (Netherlands).

### Support for valuation on a true transaction / actually observed price basis (transaction prices)
- Many respondents explicitly support transaction-price/invoice-value valuation, subject to practical considerations:
  - AUSTRALIA: "We would support work on this."
  - BRAZIL: "Support transaction price as a valuation recommendation." (also requests clarification on contract vs. price actually paid)
  - CHINA: Strong conceptual support but notes practical implications for transportation service recording; highlights potential need for new data sources and difficulty obtaining invoice values for freight.
  - IRELAND: CSO "is in favour of the proposal to record merchandise trade at transaction value"; expects transition feasible and limited disruption to time series.
  - MALAYSIA: Observed price basis "is more aligned with BPM6" (referring to merchanting entries).
  - MOROCCO: Prefers actually observed price basis reflecting what importer pays for product and transport service.
  - NETHERLANDS and RUSSIA: Support development and step forward using actual invoice prices.
  - UNITED STATES: Inclined to support conceptually "contingent on an investigation of the practical aspects."
  - ECB: Prefers SNA current treatment conceptually but emphasizes practical considerations should guide decision.

### Investigation of practical aspects (data availability, implementation)
- Strong consensus on need to investigate practical aspects before adoption:
  - AUSTRALIA: Supports work and suggests engagement with World Customs Organisation.
  - BRAZIL: Emphasizes continuing investigation of practical aspects.
  - GERMANY: Recommends further investigations based on OECD and IMF stocktaking on use of invoice values.
  - IRELAND, MALAYSIA, MOROCCO, NETHERLANDS, UNITED STATES, ECB: Support investigations on data sources, feasibility, and effort required.
  - Netherlands: Notes an ongoing study using transaction values under a Eurostat IGA Grant (duration till December 2020).
  - China: Lists specific practical challenges:
    - New data sources may be needed for transportation services if BPM7 adopts invoice approach.
    - Invoice values for freight costs may be hard to obtain; invoices for goods more accessible via Customs, but transport-service invoices particularly difficult for debit side.
    - Harmonization among statistical manuals would be required if BPM7 adopts invoice approach.

### Additional substantive comments from respondents
- Asymmetries and primary causes:
  - Canada: In Canadian experience, FOB/CIF differences account for only part of asymmetries; transshipments are often the largest source (example: 2018 reconciliation with China).
  - Netherlands: Concept differences are less problematic than lack of correct CIF/FOB values and inability to compute proper CIF-to-FOB adjustments.
- Harmonization concerns:
  - Uruguay: Even if BPM adopts invoice values, International Merchandise Trade Statistics (UN, 2010) and WTO Customs Valuation Agreement practices could leave inconsistencies with customs valuation practices that mostly apply FOB for exports and CIF for imports.
  - Eurostat and others: A change to transaction values would require amendments to both SNA and BPM and consideration of practical compilation and backward revision implications.
- National compilation practices:
  - Norway: Uses SNA methodology domestically and shows BPM-adjusted data (CIF-to-FOB) for international reporting; domestic publications present the adjustment as a memorandum item.
  - Russia: Urgent need to harmonize SNA and BOP valuation methods because BOP-derived exports/imports are used in GDP assessment by final uses.
- Documentation and clarity:
  - Germany: Finds the issue note satisfactory, readable, and clear.
  - Several respondents recommend clearer mention in the paper that SNA records imports and exports at FOB values or CIF for detailed breakdowns (2008 SNA 3.149), and that practical aspects for compilation and time-series revision should be discussed.

*Appendix 1.2 GZTT consultation on the issue note on the CIF-FOB valuation of imports and exports: detailed responses.*

### 2.    Practical considerations on valuation of imports and exports of goods

### support-document-draft-guidance-note-g1 - 2.    Practical considerations on valuation of imports and exports of goods

### Overview
- GZTT co-chairs and secretariat investigated the feasibility of using invoice (transaction) values for the valuation of imports and exports of goods following joint 2019 IMF-OECD survey results and the 2019 BOPCOM discussion.
- Some BOPCOM members emphasized the conceptual importance of separating services (transportation and insurance) from the value of goods to support existing standards.
- Initial IMF-OECD reconciliation results suggest differences between invoice values and FOB valuation are small; Harrison (2012) indicates CIF to FOB adjustment is generally based on long established ratios.
- The document presents: (i) a GZTT follow-up consultation (Section 2.1); (ii) a review of country practices and compilation methods including CIF to FOB adjustments (Section 2.2); (iii) a comparison of survey results and way forward (Section 2.3); and (iv) practical aspects from the balance of payments perspective provided by the CATT secretariat (Section 2.4).

### Second GZTT consultation: summary outcome
- Respondents:
  - Total respondents: 22 from national agencies and 2 international organizations.
  - Appendix references: questionnaire in Appendix 2.2; detailed outcome in Appendix 2.3.

- A. Access to invoice data
  - Availability from customs documentation:
    - Question 1 table result: Yes 17, No 3, Other 2.
    - Respondents from all regions reported availability from customs documents.
    - EU member countries: invoice values available for Intra-EU trade; availability for Extra-EU trade may differ by country.
    - Latin American countries: in some cases customs data not used for national accounts or balance of payments.
    - One respondent: invoice values available for imports but not for exports.
  - Access by compiling agencies:
    - Question 2 table result: Yes 13, No 7, Other 2.
    - Thirteen respondents confirm National Statistics Office, Central Bank, or other compiling agency have access to transaction values.
    - Limitations noted: access restricted to aggregates, microdata restricted by law in one case, access restricted to specific units within agency in one case.
    - Seven respondents report compiling agencies do not have access; two answered undefined.
    - Reasons for no access include: field not reported, not accurate, confidential, IT systems not prepared, differences between imports and exports, consolidated data only, collection on case-by-case basis due to response burden, lack of standard invoice form, differences between customs and invoice values.
  - Other data sources:
    - Question 3 table result: Yes 3, No 17, Other 2.
    - In general (17 of 22 country respondents) transaction data are not available from other sources.
    - Two respondents note aggregated business survey data available; three respondents state other sources exist including business surveys and revenue authority data.

- B. Quality of invoice data
  - Perceived accuracy:
    - Question 4 table result: Yes 16, No 3, Other 3.
    - Sixteen respondents consider transaction value data in their country accurate.
    - For one European country accuracy unknown for Extra-EU trade.
    - Two respondents state the question needs further analysis since invoice data not used in compilation.
    - Three respondents consider invoice data not accurate due to inconsistency with customs values, customs not aiming to collect detailed data, and need for corrections after analysis.

- C. Impact of the use of invoice data
  - Effect on asymmetries:
    - Question 5 table result: Yes 16, No 4, Other 2.
    - Sixteen respondents consider adoption of transaction values likely to reduce asymmetries in international trade data, citing harmonized valuation across countries and reduction of uncertain CIF/FOB estimations.
    - Practical challenges that may remain: harmonizing valuation across corporations, different trade regimes, timing of reporting due to transport time, classification differences, confidentiality and regulation differences, differing dissemination and revision policies.
    - Respondents note asymmetries may not be solved in balance of payments statistics that use a 'point of uniform valuation' principle; impact on trade asymmetries might not be significant.
    - Four respondents consider invoice values not likely to reduce asymmetries, citing potential creation of other inconsistencies (e.g., freight transport estimates) and need for further analysis.
  - Effect on national accounts accuracy:
    - Question 6 table result: Yes 12, No 7, Other 3.
    - Twelve respondents consider adoption of transaction data would significantly improve accuracy of national accounts estimates, facilitating Supply and Use Tables (SUT) balancing at detailed product level and reducing adjustments in SUT compilation.
    - Seven respondents consider accuracy would not improve significantly due to unchanged access, persistent identification problems for transport services, inconsistent data across sources, incomplete coverage of invoice values in some countries, and need for additional studies.
  - Analysis of invoice vs CIF-FOB differences:
    - Question 7 table result: Yes 7, No 15.
    - Fifteen respondents report no study conducted on invoice vs CIF-FOB differences (reasons: lack of access to detailed data, other compilation priorities, lack of resources).
    - Seven respondents report such a study; only one respondent quantified the difference: for overall exports it was around 2 percent in 2018. Other country studies are ongoing.

### Selected surveys and country practices (summary)
- A. 2016 UNSD Decennial National Compilation and Dissemination Practices Survey on IMTS
  - Results: Most respondents (68 in 102 economies) maintained invoice price as one valuation in basic merchandise trade statistics.
  - Some EU countries maintained invoice price only for intra-EU trade; other countries did not disseminate these data.

- B. 2019 Joint IMF-OECD WPTGS Stocktaking Questionnaire
  - Participation: around 65 countries.
  - Findings on invoice values and balance of payments:
    - Around 50 percent of respondents were unfavorable to using invoice values.
    - 20 percent were favorable to using invoice values.
    - 30 percent were unsure.
  - BOPCOM 2019 discussion:
    - Noted inconsistencies between BPM6 (FOB for goods regardless of delivery terms) and 2008 SNA (uses actually observed contractual price).
    - Initial country views showed little global support for invoice values and concerns about practical implementation.
    - Proposed way forward:
      - Maintain FOB valuation.
      - Encourage countries to re-examine methodologies to estimate trade and associated transportation costs to reach BPM/SNA consistency.
      - Improve inter-agency collaboration between national accounts and balance of payments compilers.
      - Investigate bilateral discrepancies with main cross-border trade partners.
      - Adopt unique consignment reference for customs purposes recommended by WCO to uniquely identify individual international trade transactions.
    - BOPCOM agreed further research needed before moving away from FOB valuation and supported gathering additional country information on feasibility of using invoice values.
    - Committee noted inconsistencies from different data sources are harder to solve than methodological inconsistencies and supported bilateral reconciliation exercises with possible WCO involvement.

- C. 2017 UNECE Workshop (Albania and Norway cases)
  - Both countries use detailed data from several sources (customs declarations, transportation cost information, nationality of transporters) to calculate average coefficients by product group and to compensate for missing values and errors.
  - Exchanging questionnaires, methods, and software between countries considered very useful.
  - Complexity factors: multiple carriers, split between national and international transporters, implicit adjustments for errors or the non-observed economy — raising the need for international coordination to derive consistent coefficients and avoid asymmetries.

- Box: Alternative view of Input-Output Tables (IOTs) — Australian Bureau of Statistics (ABS)
  - IOTs compiled from balanced SUTs may deviate from 2008 SNA recommendations for analytical purposes; Australian deviations include the 1968 SNA transport margin adjustment and the CIF to FOB adjustment.
  - ABS practice:
    - Each imported good in IOTs is valued CIF (equivalent to basic price of domestic product); total imports are valued FOB in accordance with BPM and SNA.
    - Two necessary operations:
      - Reconcile detailed CIF values with total imports FOB.
      - Avoid double counting of resident services when transport and insurance services on imports are performed by residents.
    - Presented equation:
      - Transport and insurance services rendered by residents =
      - = (imports FOB - imports CIF) + (transport and insurance rendered by non-residents)
    - UN Handbook recommends presenting CIF to FOB adjustment as separate item in IOTs; ABS adds adjustment to transport and insurance services rendered by non-residents and allocates components to non-margin water transport and non-margin air freight.
    - The total adjustment is, by construction, negative; ABS also compiles an alternate view by reallocating negative adjustment on imports to a positive adjustment on exports (increasing both imports and exports by the same amount) to suit analytical models.

- D. 2019 Joint IMF-OECD WPTGS Stocktaking: reconciliation exercise
  - Initial reconciliation work and surveys point to limited global support for adopting invoice values and the need for further country-level investigation and methodological coordination.

*Source: support-document-draft-guidance-note-g1 - 2. Practical considerations on valuation of imports and exports of goods (support-document-draft-guidance-note-g1). PDF chapter.*

### 70. As part of the IMF-OECD joint initiative, a reconciliation exercise was undertaken with the

### support-document-draft-guidance-note-g1 - 70. As part of the IMF-OECD joint initiative, a reconciliation exercise was undertaken with the

### IMF-OECD joint initiative: reconciliation exercise results
- Eighteen countries agreed to participate in a reconciliation exercise comparing international trade and invoice data.
- Results for five countries (as summarized from the BOPCOM paper) include:
  - Albania: for 2016-18, the annual average ratios of invoice values to BOP trade were estimated between 5.5 and 5.9 percent.
  - Belgium: for intra-EU trade in 2015, the invoice value to CIF adjustment was 0.04 and 0.16 percent of total trade, for imports and exports, respectively; the CIF to FOB adjustment was -1.67 percent of total trade for imports.
  - Indonesia: for 2014-18, the comparison of FOB exports from customs data and bank records of export proceeds revealed a difference of 13 to 14 percent, due partly to administrative fees, discounts, or subcontracting processing.
  - Kosovo: for 2018, comparison of customs data to a survey of main trade corporations revealed small differences in general, but customs data may not provide reliable estimates for some specific cases, including freight and insurance.
  - Moldova: for Q1 2019, the average ratios of invoice values to BOP trade were estimated around 5 percent, with significant fluctuations between different groups of trade partners.

### Practices and methodologies reported in other institutional studies
- 2017 Asian Development Bank Compendium of Supply and Use Tables:
  - Describes diverse practices for compiling CIF to FOB adjustments and data sources for trade margins.
  - Examples of CIF to FOB ratios reported by economies: 8 percent for Cambodia, 5 percent for PR China and India, 27 percent for Mongolia.
  - Some economies report imports already available on FOB valuation (e.g., Bangladesh and Bhutan) and thus no CIF to FOB adjustment needed.
  - Use of Business Census, SUT surveys, or freight and insurance surveys for transport margins at benchmark SUT is reported.
- 2020 UNECE Meeting of the Group of Experts on National Accounts (Serbia example):
  - Serbia moved from a fixed 3.1 percent CIF-to-FOB ratio to a customs-declarations-based method modeled on Statistics Norway; the previous fixed-ratio approach had disadvantages: same coefficient for all transport modes, commodity groups, and distance; not reflecting real costs dependent on external factors (e.g., fuel prices); and fixed ratio for resident and non-resident carriers.
  - New method (imports/exports) key steps:
    - Split by delivery terms: use FOB-type transactions to fill missing freight/insurance for CIF-type items by replacing with FOB-type freights and insurance from transactions with same characteristics (mode of transport, commodity group, country of consignment); if unavailable, use less strict replacements.
    - Test on extremes to ensure estimates are not affected by outliers.
    - Final adjustments for neighbor countries, special commodities (e.g., oil and electricity), and declarations representing imports of a single good.
    - For exports, compile freight and insurance estimates similarly but using CIF-type forms.
    - Data processing separates resident and non-resident service providers and transport versus insurance; transport costs by residents and non-residents are detailed by mode of transport and processed according to terms of delivery.
- 2017 OECD Statistics Working Paper (Miao and Fortanier):
  - Using a gravity-type model with explicit transportation and insurance cost data, the bilateral CIF to FOB trade weighted margin was around 6 percent for 1995-2004.
  - The study highlights relevant variation across countries (distance and geographical barriers) and products.

### Comparison of survey results: views on use of invoice/transaction values
- GZTT consultation (November 2019):
  - Among four options presented, most respondents (10 in 16 answers from countries) selected Option 4: "It can be considered to amend the SNA and BPM guidelines and adopt the valuation on a true transaction or actually observed price basis."
  - Respondents noted conceptual soundness of Option 4, potential to avoid compiling CIF to FOB adjustments, and to reduce asymmetries; they also noted the need for changes in both the 2008 SNA and BPM6 for consistency.
  - Practical concerns raised: (i) need for new data sources; (ii) difficulty obtaining accurate transaction values even when sources exist; (iii) need to harmonize with other statistical manuals (notably IMTS).
- IMF-OECD WPTGS Stocktaking Questionnaire (early 2019):
  - Around 65 countries participated.
  - Initial positions on invoice values showed apparent divergence from GZTT: approximately 50 percent of respondents were unfavorable to the proposal, 20 percent favorable, and 30 percent unsure.
- Differences in responses:
  - Four countries supported Option 4 in the GZTT consultation but did not favor use of transaction values in the IMF-OECD survey: Australia, China, Russian Federation, and USA.
- Reported advantages of using invoice values (IMF-OECD survey responses):
  - Valuation closer to true market values.
  - Elimination of need to estimate CIF to FOB adjustment.
  - Reduction of asymmetries.
- Reported disadvantages:
  - Lack of access to invoice values or corporations’ records (or detailed data).
  - Need for data to estimate freight transport and insurance services (possibly requiring direct surveys).
  - Difficulties in classification of goods and services and in establishing residence of corporations.
  - Increased work or respondent burden.

### Comparison of survey results: availability and access to invoice/transaction values
- GZTT follow-up questionnaire (second round):
  - Most respondents (17 of 22) reported that actual transaction (invoice) values for IMTS are available from customs documents.
  - Thirteen of 22 respondents confirmed that NSO, Central Bank, or other agencies compiling national accounts or BOP have access to transaction (invoice) values in customs declarations; in some cases access is limited to aggregates or restricted by law or internal constraints.
- IMF-OECD survey view on practical feasibility:
  - Most respondents considered it not practically feasible to develop BOP statistics for trade in goods and freight using invoice values in the medium term, due to: (i) lack of complete and accurate available data; (ii) inability to use corporations’ records; (iii) need for collaboration between Central Bank, NSO, and Customs Authority to change customs documents or provide access; (iv) lack of detailed data on transport services.
- UNSD 2016 National Compilation and Dissemination Practices survey on IMTS:
  - Most respondent countries (68 of 102) maintained the invoice price as one valuation in basic merchandise trade statistics.
  - Some EU members maintained invoice valuation only for intra-EU trade; other countries noted these data were not disseminated.
- Eurostat summary for intra-EU trade in 2016:
  - The ratio of total statistical valuation (FOB for exports and CIF for imports) to total invoice value is below or equal to 1.2 percent in absolute values, for both imports and exports, based on data transmitted by countries.

### Comparison of survey results: interest and experience with pilot studies
- GZTT responses on studies of invoice value vs CIF-FOB:
  - 15 of 22 respondents reported no study had been conducted on the difference between invoice value and CIF-FOB valuation, usually due to lack of access to detailed data, other priorities, or lack of resources.
  - Seven respondents reported that their agency conducted a study on the difference; only one reported the size of the difference: Norway reported a difference of around 2 percent for overall exports in 2018, with significant fluctuations across products.
- As part of the IMF-OECD joint initiative, a reconciliation exercise was undertaken with agreed participation of eighteen countries (see IMF-OECD results summarized above).

### Balance of payments perspective: practical considerations for changing valuation principle
- A change from uniform FOB valuation to transaction (invoice) values for imports and exports would have implications for compiling agencies, respondents, and data users.
- Main practical issues identified:
  - Need for new or expanded data sources and improved data access arrangements between Customs, NSO, and Central Bank.
  - Legal or administrative restrictions may limit access to microdata in some countries.
  - Requirement for detailed data on transport and insurance services if invoice values are used, potentially necessitating new surveys or data collections.
  - Increased classification and residence determination challenges for goods and services when using transaction values.
  - Potential respondent burden and additional workload for compilers.

*Source: IMF (support-document-draft-guidance-note-g1).*

### 90. Need of an overhauled data collection for both merchandise and transport costs would

### Need of an overhauled data collection for both merchandise and transport costs

### Resource implications and data-collection challenges
- The guidance note (paras. 90–93) states that overhauling data collection would have resource implications, including significant cost implications that "in most cases would be strongly resented by policy makers."
- Main compilation concern: potential unavailability of invoice values leading to asymmetric recordings of imports and exports (corporation versus customs records), "notably in some large trading countries."
- Recommended data-collection responses:
  - Introduce new surveys among manufacturers and transport (and insurance) corporations, or adapt/extend existing surveys to collect invoice values and freight and insurance services data, particularly to cover the debit side.
- Key methodological and practical issues to weigh:
  - (i) surveying a sample versus a census (customs records);
  - (ii) access to invoice data (questions related to invoice values are considered to be more sensitive by the respondents);
  - (iii) a likely high non-response rate with resulting biases;
  - (iv) delays in reporting affecting the timeliness of production.
- Specific risks for services data: complexity of global arrangements may introduce larger errors, including overvaluation of freight due to many intermediaries, or undervaluation because of shifting values to the merchandise goods.

### Legal, regulatory, IT, and institutional requirements
- Legal/regulatory:
  - The guidance note (para. 91) states that new legal requirements and regulatory support to data collection from private corporations would be needed, "in particular for non-financial corporations."
  - In regional unions (such as in EU), legal acts governing the external trade of goods would have to be updated together with national regulations.
- IT and capacity:
  - The guidance note (para. 92) warns that such changes would take time to implement and require an update of IT systems in a "very demanding business setting."
  - Many countries "would not have the capacity to modernize their systems to build adequate alternative data sources for foreign trade statistics (e.g., using business financial statements)."
  - In cases where invoice values are already available from Customs records, "the change would only require minor adjustments to the compilation system."
- Organizational/institutional arrangements:
  - The guidance note (para. 93) calls for enhanced interagency cooperation (between compiling agencies and with Customs) to secure access to more detail.
  - Examples of access issues:
    - In many European countries, access to invoice values for Extra EU exports (customs declarations) is given at aggregated levels (e.g., for the entire delivery) or the information is missing.
    - In some countries, customs data are shared with the national statistical office only, so "the collaboration of the balance of payments compiling agency should be strengthened to ensure access to invoice values."

### Reporting units and respondent burden
- The guidance note (para. 94) states that additional information needs "would increase the response burden, especially for merchandise trading and transport and insurance corporations."
- Practical burden notes:
  - Considering the large amount of transactions in a month, reporting "would be a time-consuming operation involving a significant amount of resources."

### Data users, series continuity, and retrospective limits
- The guidance note (para. 95) highlights costs for data users to adapt to the new data and notes that "retrospective estimations are highly unlikely, leading thus to breaks in series."
- Specifics:
  - Currently, invoice values "may not be collected in the customs declarations."
  - Even if collection becomes mandatory, foreign trade figures based on invoice values "cannot be calculated retroactively, implying breaks in series."

### Follow-up consultation instrument (Appendix 2.1): questionnaire themes
- The guidance note includes a second GZTT consultation questionnaire (Appendix 2.1) to investigate feasibility of using transaction (invoice) values. Themes and question foci include:
  - A Access to invoice data:
    - Q1: Are actual transaction (invoice) values available on custom documents? (YES/NO)
    - Q2: Does the NSO, Central Bank, or other compiling agency have access to transaction values?
    - Q2.1: If no, describe why.
    - Q3: Are these data available from any other data source?
    - Q3.1: If yes, describe.
  - B Quality of invoice data:
    - Q4: Are the data on transaction values considered accurate?
    - Q4.1: If not, describe main shortcomings.
  - C Impact of the use of invoice data:
    - Q5: Would adoption of transaction values reduce asymmetries in international trade data?
    - Q5.1: Please describe.
    - Q6: Would adoption of transaction values significantly improve accuracy of national accounts (e.g., balances at product level using trade statistics vs. firm reports)?
    - Q6.1: Please describe.
    - Q7: Has your agency analyzed/quantified the difference between invoice value and CIF-FOB valuation?
    - Q7.1: Please describe.

*Source: support-document-draft-guidance-note-g1*

### Appendix 2.2: Second GZTT consultation: detailed responses

### Appendix 2.2: Second GZTT consultation — detailed responses

### Availability of actual transaction (invoice) values on customs documents (Question 1)
- Australia (ABS): Yes for imports; No for exports.  
  - Imports: Customs Value (transaction value), CIF and FOB collected at consignment level; Customs Value collected at total and for each commodity; imports statistics based on Customs Value.  
  - Exports: Declarations require FOB value by commodity; exports statistics published on FOB basis.
- Brazil: Yes.
- Canada (STATCAN): No.
- China (SAFE): Yes.
- Germany (Deutsche Bundesbank): Yes.
- Ireland (Central Statistics Office): Yes. Comment: Article 70 of REGULATION (EU) No 952/2013 — “The primary basis for the customs value of goods shall be the transaction value, that is the price actually paid”.
- Multiple COMESA responses:
  - Kenya: Yes.
  - Kingdom of Eswatini: Yes.
  - Madagascar: Yes.
  - Malawi: No.
  - Mauritius: Yes.
  - Rwanda: Yes.
  - Zambia / Zimbabwe entries appear in other tables.
- Malaysia (Central Bank of Malaysia): Yes.
- Morocco (Haut-Commisariat au Plan): Yes.
- Netherlands (Statistics Netherlands): Yes.
- Norway (Statistics Norway): Yes.
- Russia (ROSSTAT): Yes.
- United Kingdom (ONS): YES/NO — Intrastat and EU trade: invoice values collected; Customs declarations and non-EU trade: Box 42 ‘Item price customs value’ can contain either invoice price (in invoice currency) or customs value; lack of consistent invoice currency complicates use; significant resource implications to change.

### Access by NSO, Central Bank, or other compilers to transaction (invoice) values (Question 2)
- Australia (ABS): "Yes for imports; No for exports". ABS has access to Customs Value, CIF and FOB at consignment level for imports; FOB for exports.
- Uruguay: Yes.
- United States (BEA): No.
- Zambia / Zimbabwe (COMESA responses): Zambia Yes; Zimbabwe Yes.
- ECB: Yes. Comment: Invoice value availability would not be (is not) a problem for euro area countries.
- Eurostat: Yes. Comment: For EU countries, invoice values are available for intra-EU trade; extra-EU availability may differ.
- Country-specific notes:
  - Brazil: partial yes; M and X data recorded by SRF; IBGE and BCB receive restricted data for SNA and BoP; legal/fiscal confidentiality impedes micro-data access.
  - Canada (STATCAN): Yes but only for a small subset of transactions for data confrontation.
  - China (SAFE): No.
  - Ireland (CSO): Yes for intra-EU trade; IT systems not set up to receive invoice item.
  - Malaysia (Central Bank): No.
  - Mauritius: No (data confidential, managed by Customs).
  - Netherlands: No (actual transaction values available on custom documents, but large part of trade is intra-EU with no customs documents; surveys used for that part).
  - Russia (ROSSTAT): Yes (aggregated data provided to Bank of Russia; no access to detailed customs declarations database).
  - United Kingdom (ONS): Yes.

### Availability from other data sources (Question 3)
- United States (BEA): Yes. U.S. Census Bureau collects invoice data case-by-case for a limited number of trades; cannot compel firms; invoice, customs, and shipping manifest values often diverge.
- Zambia (COMESA): No. NSO has access to invoice values; Central Bank receives aggregated data.
- Zimbabwe (COMESA): Yes.
- ECB: Yes.
- Eurostat: Comment — invoice values available for intra-EU trade; extra-EU availability may differ.
- Many national replies: Australia No; Brazil No (annual surveys provide limited aggregate data); Canada No; Ireland No (only total imports/exports from enterprise surveys); several COMESA countries No.

### Perceived accuracy of transaction (invoice) values (Question 4)
- Varied responses:
  - Considered accurate by: Australia (ABS) — Yes; Brazil — Yes; Canada — Yes; China (SAFE) — Yes; Germany — Yes; Ireland (CSO) — Yes for Intra-EU invoice value; Morocco — Yes (but reporting gaps); Mauritius — Yes in some replies; Rwanda — Yes; Uruguay — Yes.
  - Considered not accurate or concerns raised by: Madagascar — No; Malaysia — No; Mauritius (one reply) — No; Netherlands — No / under investigation; Norway — No (not routinely checked; accuracy uncertain at detailed level); Russia (ROSSTAT) — No; United Kingdom (ONS) — No; United States (BEA) — No; Zambia — No; Eurostat — No.
- Notable detail (Norway): invoice often basis for customs declaration so transaction value for entire transaction should be accurate, but not necessarily at detailed tariff-code level in multi-good shipments.
- Notable detail (Malawi): TRANSACTIONS ARE CONSOLIDATED BY DATA PROVIDERS LIKE RESERVE BANK OR MINISTRY OF FINANCE AND SHARED WITH NSO; quality concerns noted where corrections are often required.
- Notable detail (Netherlands): study under Eurostat Grant due end date December 2020 to investigate inexplicable differences between CIF values and transaction values.

### Would adopting transaction values reduce asymmetries in international trade data? (Question 5)
- Many respondents: Yes.
  - Australia (ABS): Yes — adoption could alleviate some asymmetries, though many other factors contribute (timing, classification differences, confidentiality/suppression, policy/regulatory differences, release/revision windows). ABS notes that CIF/FOB values currently contain domestic costs (transport, packaging, loading) that if excluded and counted separately could improve comparability.
  - China (SAFE): Yes.
  - Germany (Deutsche Bundesbank): Yes.
  - Ireland (CSO): Yes in practical terms but not necessarily significant for BoP universal valuation.
  - Malaysia (Central Bank): Yes in principle — importance of quantifying differences.
  - Mauritius (COMESA): Yes — fair, uniform, neutral; WCO GATT Valuation Agreement reference.
  - Morocco: Yes for import level in FOB contracts; CIF contracts still pose transport registration issues.
  - Netherlands: Yes — using actual transaction values by both sides should lower asymmetries.
  - Norway: Yes in principle.
  - United States (BEA): Yes.
  - Uruguay: Yes (likely adequate; controlled by tax collection agency).
- Some respondents: No or qualified:
  - Brazil: No (maybe reduce, but not much); databases and country-specific compilation options vary.
  - Canada (STATCAN): No — strong data sharing with US already.
  - Kenya (COMESA): No — more work required to study inconsistencies, multiple entries per item.
  - Russia (ROSSTAT): No — could increase asymmetry, make price indices incomparable, distort trade-in-services for freight, and geography of flows.
  - United Kingdom (ONS): YES/NO — invoice data quality high for some uses; adopting transaction values would not solve many other asymmetry sources; converting CIF to FOB statistically might aid comparisons but would generate different differences.
  - Eurostat: Difficult to say for EU as a whole.

### Would adopting transaction values significantly improve national accounts accuracy? (Question 6)
- Many respondents: Yes (see country list and comments):
  - Australia (ABS): Yes — unified transaction valuation would reduce asymmetry and improve NA accuracy.
  - United States (BEA): Yes — harmonized valuation likely reduces asymmetries; practical implementation challenging.
  - Zambia / Zimbabwe (COMESA): Zambia Yes; Zimbabwe mixed responses (Yes in some replies; one response: No — impact not clear).
  - ECB / Eurostat: Yes — would avoid CIF-FOB estimates based on uncertain information; improve consistency with output values and exports.
  - Malaysia: Yes in principle.
  - Mauritius: Yes — more disaggregated product-level compilation improves estimates.
  - Morocco: Yes — considerable effect on NA consistency between business declarations and customs; issues with transport intermediaries and bulk/containered goods noted.
  - Netherlands: Yes — expect fewer supply-demand differences in SUT and better linkage between trade in goods and services.
  - Norway: Using transaction data improves comparisons with company reports.
  - Uruguay: Yes — expected to improve harmonization between BoP and NA and reduce need for adjustments in SUT; 2017-2018 LAC research showed 9 of 15 surveyed countries make adjustments to BoP Goods and Services Account within SUT.
- Some respondents: No or uncertain:
  - Brazil: No — not expected to "significantly improve"; legal/access issues and SUT implications need clearer specification.
  - Canada (STATCAN): No — BoP based on customs data and product aggregations; current system adequate.
  - Ireland (CSO): No — not a significant improvement; differences usually not explained by valuation.
  - Kenya (COMESA): No — studies required.
  - United Kingdom (ONS): No — possibly for some periods/products; invoice not collected for all trade.
  - Russia (ROSSTAT): No — potential to increase asymmetries and distortions.

### Has analysis/quantification been done on invoice value vs CIF-FOB valuation (Question 7)
- Positive/ongoing analyses:
  - Germany (Deutsche Bundesbank): Yes — planning to present first results at OECD WP ITS and UNECE NA Expert Group meetings.
  - Ireland (CSO): Yes — updating CIF to FOB adjustment using intra-EU comparisons of invoice vs statistical values.
  - Netherlands (Statistics Netherlands): Yes (partly) — conducting a study (end date December 2020).
  - Norway (Statistics Norway): Yes — analyzed exports differences: fish (CPA 03) difference around 6 percent; crude oil and natural gas difference 15 percent; other product groups smaller differences; overall merchandise exports difference 2.4 percent in 2018.
  - Mauritius (COMESA): Yes — real time access to CIF-FOB values available for analysis.
  - Rwanda (COMESA): Yes — survey to adjust CIF from first EAC point of entry to Rwanda where necessary.
  - Kingdom of Eswatini: Yes — real-time data exchange with SARS mirrors transactions on both customs systems.
- No / not done:
  - Australia: No.
  - Brazil: No.
  - Canada (STATCAN): No — insufficient invoice data subset.
  - China (SAFE): No — would require major compilation system work.
  - Kenya / Madagascar / Malawi / Morocco / Zambia / Zimbabwe: No or not applicable; Malawi notes lack of invoice values impedes such analysis.
  - United Kingdom (ONS): No — regular reviews on asymmetries but not specifically invoice vs CIF/FOB differences.
  - United States (BEA): No — U.S. work focuses on measuring freight differently; BEA does not derive CIF to FOB adjustment.
  - Eurostat: No (response in table).

*Source: Appendix 2.2: Second GZTT consultation — detailed responses*

### 3.    Existing materials

### support-document-draft-guidance-note-g1 - 3.    Existing materials

### Ongoing discussion since the last update of SNA and BPM: a summary
- At the end of 2011, Anne Harrison, Editor of the 2008 SNA, highlighted inconsistencies in guidance on how to record freight and insurance costs between national and international economic accounts; the issue was first discussed in a 2012 BOPCOM paper.
- IMF Statistics Department initially did not favor changes to BPM6 because the balance of payments treatment is long-established.
- The Intersecretariat Working Group on National Accounts (ISWGNA) and the Advisory Expert Group (AEG) were asked to clarify the issue; the 2013 AEG concluded that both 2008 SNA and BPM6 recommend recording imports and exports of goods at FOB values but that the 2008 SNA does not fully reconcile the FOB valuation principle with output valuation at basic prices.
- The 2013 AEG recommended that, in the longer term, the agreed change of ownership principle be applied consistently across the SNA, BPM, and IMTS.

### Key papers, proposals, and meeting outcomes (2012–2019)
- 2012 BOPCOM (Anne Harrison):
  - Noted BPM6 uses a concept of uniform valuation driven by customs documentation and identified special cases needing further guidance (e.g., transport of goods for processing and merchanted goods).
  - Identified three diagnostic questions for recording flows: (i) residence of the transport-service provider; (ii) residence of the transport-service requester; (iii) whether the goods provider charges the purchaser explicitly for transport.
  - Presented four options to align SNA and BPM (transcribed as options i–iv in the source).
    - Option (i): revert SNA domestic treatment of transportation to always treat it as a service (pre-1993 practice) — considered unlikely and legally constrained for EU/ESA95.
    - Option (ii): change SNA to be strictly consistent with BPM6 for transactions with non-residents — would create inconsistency within SNA and likely face opposition.
    - Option (iii): retain both systems but explain inconsistencies via a supplementary table converting imports CIF to imports FOB, requiring extra clarification on goods for processing and merchanted goods.
    - Option (iv): amend BPM guidance to codify a conventional classification (2a/2b/2c) but explore exceptions given global imbalances and containerisation effects.
  - Raised concerns that CIF to FOB adjustments often rely on outdated long-established proportions and may be inaccurate given containerization.

- 2013 AEG:
  - Recognized 2008 SNA and BPM6 alignment on FOB valuation but that SNA is an exception to the change of ownership accrual principle used elsewhere.
  - Agreed that change of ownership accrual should be applied across SNA, BPM, and Foreign Trade Statistics with full partner involvement.

- 2017 AEG (Hiemstra and de Haan; Statistics Netherlands):
  - Proposed recording imports and exports of goods in national accounts and SUTs based on observed transaction values due to data limitations.
  - Argued that converting invoice values to FOB and estimating CIF-to-FOB adjustments is often low-quality or unavailable and that FOB/CIF valuation can lead to inaccuracies and asymmetries.
  - Noted that information on carrier residency and terms of delivery is often unavailable, complicating trade-in-services adjustments.
  - Recommended adopting transaction values for recording exports and imports in national accounts and balance of payments statistics; this would abandon a uniform valuation principle but avoid detailed data modifications that reduce quality.
  - 2017 AEG conclusions: use of transaction prices is not consistent with current SNA recommendations but recommended ISWGNA assess country experiences in CIF-FOB recording and develop a guidance note as part of the globalization research agenda.

- 2018 OECD WPTGS (Walter):
  - Argued valuation concepts need updating to reflect containerized transport and reduced customs controls, making separation of transport costs (export border / between borders / import country) less natural.
  - Described current two-stage adjustment methodology: (1) adjust transaction values to CIF/FOB using multiple data sources and (2) split transport services between resident and non-resident providers using survey information — often based on estimates and assumptions, possibly producing asymmetries.
  - Proposed an invoice-values approach: record imports and exports using invoice values and record transport services only if a market transaction between a resident and a non-resident occurs.
    - Claimed advantages of the invoice approach include: (i) elimination of estimates that cause asymmetries; (ii) direct data extraction from company accounts; (iii) reduced compiler data requirements; (iv) disconnection of freight compilation from IMTS weights and delivery terms; (v) avoidance of difficult geographical allocation problems; (vi) easier combination of freight with merchandise data by country of origin/consignment; (vii) better fit with processing and merchanting components; (viii) consistency with 2008 SNA recommendations for valuation of goods and related transport services inside the economic territory; (ix) consistency with 2008 SNA Chapter 14 when customs data are unavailable; (x) fostered consistency between SNA and BPM; (xi) better reflection of economic reality for users.
    - Main disadvantage noted: breaks in time series from adopting invoice approach.

- 2019 IMF-OECD Stocktaking Survey and BOPCOM:
  - Joint IMF-OECD survey showed majority reluctance to change to transaction values, mainly for practical reasons (details in Section 2.2 of the source).
  - A reconciliation exercise in pilot countries found the difference between FOB and transaction values not significant.
  - The paper recommended: analyze and improve current CIF-to-FOB compilation methods; foster inter-agency and country cooperation; adopt Unique Consignment Reference to reduce trade asymmetries.

### Valuation of imports and exports of goods in the national accounts (2008 SNA recommendations)
- Time of recording and valuation basis:
  - The 2008 SNA recommends recording acquisitions of goods at the moment when economic ownership changes; imports and exports should be recorded when change of ownership occurs.
  - In absence of data on change of ownership date, goods crossing frontiers close to that moment may be used as an approximation; customs-based trade statistics may be used as an approximation.
  - Services are recorded when provided. (2008 SNA, paragraphs 3.169-3.170).

- Market prices and alternative valuation methods:
  - Market prices are the basic reference for valuation: amounts of money that willing buyers pay to acquire products from willing sellers in exchanges between independent parties, after rebates/refunds/adjustments.
  - In absence of market transactions, valuation is by costs incurred or by reference to market prices of analogous goods and services; transactions valued at prices agreed by economic units. (2008 SNA, paragraphs 2.59 and 3.119-3.121).

- Output and use valuation (basic, producer, purchaser prices):
  - 2008 SNA recommends recording output at basic prices (or producers’ prices if basic prices not feasible) and uses at purchasers’ prices.
  - Producers and users perceive product value differently due to taxes/subsidies, transport costs, trade margins; SNA records uses at purchasers’ prices (including these elements) but excludes them from output value.
  - Basic and producer prices exclude transport charges invoiced separately by the producer; purchaser price includes transport charges paid separately by the purchaser. (2008 SNA, paragraphs 3.145-3.148).

- Valuation tension highlighted:
  - Chapters 3 and 26 of 2008 SNA state recording standard for international trade in goods is FOB, while Chapter 14 can give the impression of valuation at transaction prices when change of ownership occurs — creating a need for clarification on applying the change of ownership accrual principle regardless of residency of transactors, especially given customs unions, complex supply chains, and the use of basic prices for output valuation including non-invoiced transport and insurance.

*Support document: Draft guidance note G1 — Section 3: Existing materials*

### 118. Imports and exports are recorded in the national accounts at the border values, and

### Valuation of imports and exports and guidance for Supply and Use Tables (SUT)

### Valuation principles and SNA guidance
- The 2008 SNA recommends:
  - Output be measured at basic prices.
  - Imports and exports of goods be valued at FOB (value at the exporter’s customs frontier).
- The 2008 SNA recognizes:
  - Imports of goods may be recorded at CIF (at the importer’s customs frontier) when FOB values are not obtainable for detailed product breakdowns, supplemented with global adjustments to FOB values (i.e., insurance and freight charges between exporter’s and importer’s frontiers).
  - Invoice values may differ from both FOB and CIF. (2008 SNA, paragraph 3.149).
- Valuation consistency:
  - Valuation principles are the same in the SNA and the international accounts, using market values (or nominal values if market values are not observable).
  - Exports and imports of goods in the international accounts use a uniform FOB-type valuation at the customs frontier of the exporting economy to provide consistent measurement between exporter and importer. (2008 SNA, paragraphs 26.19-26.20).

### Rest of the World accounts, IMTS, and adjustments
- Main data source for exports and imports of goods: IMTS data.
  - IMTS typically use CIF-type valuation for imports.
  - BPM6 recommends an FOB-type valuation for both imports and exports.
- Practical implications and necessary adjustments:
  - Freight and insurance costs incurred between customs frontiers must be excluded (CIF-to-FOB adjustment) when transforming IMTS CIF imports into BPM6-consistent FOB imports.
  - Some freight and insurance costs may need to be rerouted due to variations between FOB-type valuation and contractual arrangements.
  - Adjustments to IMTS may also be needed for differences in time of reporting and special cases: goods sent on consignment, merchanting, nonmonetary gold, goods entering or leaving the territory illegally, goods procured in ports by carriers, goods moving physically without change of ownership. (2008 SNA, paragraphs 26.49-26.57).

### SUT framework: supply, use, and valuation reconciliation
- SUT accounting principle:
  - Product available for use = domestic production + imports.
  - Uses: intermediate consumption, final consumption, capital formation (including changes in inventories), or exports.
- Typical data valuation mixes and implications:
  - Production and output often valued at basic prices.
  - Intermediate consumption and final uses usually valued at purchaser’s prices.
  - Detailed imports often available valued CIF.
  - Exports often available valued FOB.
  - Because production is at basic prices and uses are at purchaser’s prices, trade and transport margins and taxes less subsidies must be estimated so supply and use are expressed in purchaser’s prices to balance.
- Valuation of imports in the SUT:
  - Imports should be recorded in the supply table at basic prices with taxes and margins added subsequently.
  - Recommendations when customs documentation is the data source:
    - If exporter meets transport costs, basic price should include transport; CIF approximates basic price.
    - If importer meets transport costs, basic price should exclude transport; FOB approximates basic price.
    - If it is not possible to determine responsibility for transport costs from customs declarations, detailed CIF figures may be the only disaggregated source; an adjustment column (CIF-to-FOB) is inserted into the supply table to avoid double counting of services and to make a global import adjustment. (2008 SNA, paragraphs 14.76-14.77).
  - Valuation of exports in the use table is less challenging because trade statistics uniformly value exports FOB, though valuation at the border may not coincide with change of ownership. (2008 SNA, paragraph 14.114).

### Transport margins and allocation issues
- Treatment of transport margins depends on contractual arrangements:
  - Domestic transport: if producer delivers without explicit charge, delivery cost is included in basic price; if purchaser is invoiced explicitly for delivery, transportation margin is part of purchaser’s price. Change of ownership timing depends on who provides transport (producer vs purchaser). (2008 SNA, paragraphs 14.54-14.60).
  - International transport: allocation information is often not readily available from surveys because some carriers or providers are non-resident. Customs data are primary but do not capture all flows (e.g., goods in single customs areas, deliveries to offshore establishments, high-value small-volume goods carried by people, ships and aircraft). (2008 SNA, paragraphs 14.61-14.62).
  - For products not covered by customs documentation, survey and other sources are used and recorded at transaction prices. Treatment parallels domestic transactions: exporter- or importer-responsibility for transport determines whether transport costs are included in exports/imports; if a third party is hired, the residence of the carrier determines whether transport is recorded as a domestic transaction, export of services, or import of services (Box 3B; 2008 SNA, paragraphs 14.63-14.68).

### CIF-to-FOB adjustment methods and SUT compilation flexibility
- CIF-to-FOB adjustment approaches described in the 2008 SNA:
  - Detailed reallocation: service margins are reallocated from producing industries to an adjustment row for CIF-to-FOB; goods columns have industry-by-industry values that include service margins which are then deducted on the adjustment row so the total equals imports FOB.
  - Simpler (less consistent with BPM6): ignore the balance-of-payments split between goods and services and adjust imports of services by the amount of non-resident services included in detailed imports of goods. This ensures total imports of goods and services match balance of payments totals but will not match the breakdown of imports of goods FOB and services; freight and insurance on imports provided by residents must still be shown as an export of services. (2008 SNA, paragraphs 28.10-28.12).
- UN SUT Handbook guidance:
  - Provides step-by-step guidance on SUT compilation and treats IMTS as the main data source for imports of goods, and balance of payments or specialized statistics for imports of services.
  - Notes differences between IMTS and 2008 SNA/BPM6 concepts (valuation CIF vs FOB, time of recording, special cases such as merchanting, non-monetary gold, goods sent abroad for processing, and direct purchases abroad by residents) and recommends adjustments to reconcile these differences.
  - Describes two types of adjustments for SUT use of IMTS: data adjustments prior to entering the SUT system, and the CIF-to-FOB adjustment within the SUT framework. (UN SUT Handbook, paragraphs 5.59-5.87; 6.150-6.157).

### Step-by-step practical implications for compilers
- Data compilation priorities and actions:
  - Use IMTS for detailed imports by product, but apply CIF-to-FOB adjustments to align with BPM6/BOP FOB valuation.
  - Identify and adjust for cases where IMTS recording timing or coverage differs from SNA/BPM6 concepts (merchanting, consignment, nonmonetary gold, goods procured in ports, goods without change of ownership).
  - When responsibility for transport cannot be determined from customs data, insert a CIF-to-FOB adjustment column in the supply table and make an offsetting global adjustment to imports of goods to avoid double counting services.
  - For exports, use IMTS and BOP data, convert classifications as needed, and gross up survey-based exports when firms below reporting thresholds are excluded. (UN SUT Handbook, paragraphs 6.150-6.157).

*Source: 2008 System of National Accounts (2008 SNA); BPM6 guidance; United Nations Handbook on Supply, Use and Input-Output Tables with Extensions and Applications (UN SUT Handbook).*

### 137. Several challenges exist in the compilation of transport margins in the SUTs. According to

### support-document-draft-guidance-note-g1 - 137. Several challenges exist in the compilation of transport margins in the SUTs. According to

### Challenges in compilation of transport margins in SUTs
- Transport margins are complex to compile because data must cover transport margins according to the SNA definition; a significant part of transportation activity takes place as ancillary activity in non-transport industries and is not identified in the system; and it is not possible to assess the importance of transport margins relative to the total transport activity in the economy. (UN SUT Handbook, paragraphs 7.89-7.117).
- Data collection constraints:
  - Special surveys of purchasers can provide data but require a level of detail that represents a burden to respondents and is not widely available.
- Four options when source data are unavailable:
  - (i) consider that transport margins are insignificant and are not estimated;
  - (ii) concentrate on products where important transport services are involved (e.g., agricultural and forestry products, energy products, iron and steel products, or construction products);
  - (iii) establish a full matrix of transport margins based on general assumptions about total transport margins and their distribution by products and uses;
  - (iv) rerouting transport margins by product and by use, based on the estimates for wholesale trade (estimate for each type of transport output the share being transport margins, record this as input into wholesale trade; output of wholesale trade should be increased by the same amount). (UN SUT Handbook, paragraphs 7.89-7.117).

### Valuation and timing principles for imports and exports of goods (BPM6)
- Time of recording:
  - Transactions in goods should be recorded at the time that the change of economic ownership takes place, per BPM6 recommendations. (BPM6, paragraphs 3.44-3.45, 3.61-3.66, 10.26-10.27).
  - IMTS recording generally follows customs procedures (movements across borders) which approximate change of ownership; timing adjustments to IMTS may be necessary to remove recorded movements that do not correspond to a change in ownership or to add movements that did correspond but were not recorded. (BPM6, paragraphs 3.44-3.45, 3.61-3.66, 10.26-10.27).
- Valuation principle:
  - Market prices are the basis for valuation. Market price defined as amount a willing buyer pays to acquire something from a willing seller, after rebates, refunds, adjustments, and so on. Imports and exports of general merchandise recorded at FOB values, which take into account any export taxes payable or any tax rebates receivable. (BPM6, paragraphs 3.67-3.69).
  - Principle for valuation of general merchandise: market value at the point of uniform valuation = FOB at the customs frontier of the exporting economy. (BPM6, paragraphs 10.30-10.31).

### IMTS valuation types and CIF-to-FOB adjustments
- IMTS statistical valuations:
  - Exports: FOB-type valuations including: (i) FOB at port on the frontier of the exporting country (sea or inland waterway transport); (ii) free carrier (FCA) at terminal on the frontier of the exporting country (for means of transport to which FOB is not applicable); (iii) delivered at frontier of the exporting country (for means of transport to which FOB and FCA are not applicable).
  - Imports: CIF-type valuations including: (i) cost, insurance, and freight (CIF) at the border of the importing country; (ii) carriage and insurance paid to the border of the importing country. (BPM6, paragraphs 10.32-10.33).
- Converting CIF to FOB:
  - Deduct freight and insurance premiums incurred from the frontier of the exporting country to the border of the importing country.
  - BPM6 recommends obtaining the CIF to FOB adjustment at a detailed level because FOB/CIF relation depends on type of good, weight, scale, special needs, mode of transport, and distance traveled; CIF to FOB ratios change over time (e.g., due to fuel prices, competition and technology in transport industry, change in the proportion of types of goods, and changes in the source economy). (BPM6, paragraph 10.34).

### Treatment of freight and insurance services (BPM6)
- Freight services consequences of FOB uniform valuation:
  - All freight costs up to the customs frontier are recorded as incurred by the exporter; all freight costs beyond the customs frontier are recorded as incurred by the importer.
  - If payment arrangements differ from FOB terms, BPM6 recommends rerouting freight services; rerouting may treat resident-to-resident transactions as between residents and nonresidents and vice versa. Timing of freight services should be the period they are rendered but attributed to the importer in the period when the goods are purchased; in practice, timing adjustments may not be feasible or material given aggregated recording. (BPM6, paragraphs 10.78-10.79).
- Freight insurance premiums:
  - Insurance premiums payable before reaching the customs frontier of the exporter are included in the FOB price of the good and subsequently treated as payable by the importer; when parties arrange payment differently, partitioning and rerouting are needed, similar to freight services adjustments. (BPM6, paragraph 10.116).

### Methods to estimate freight and insurance premiums for CIF-to-FOB adjustment (BPM6 Compilation Guide)
- Described methods (BPM6 Compilation Guide, paragraphs 12.44-12.46 and table 12.2):
  - (i) Use IMTS data when both CIF and FOB values of imports are recorded, or a sample survey of customs records when both valuations are not recorded.
  - (ii) Enterprise survey to collect data from importers on freight and insurance premiums paid on imports.
  - (iii) Enterprise survey to collect data on freight from branch offices and agents of nonresident operators.
  - (iv) Analyze trade flows, freight, and insurance rates by commodity, mode of transport, and economy with data derived from IMTS and apply freight and insurance premium rates.
  - (v) Use an arbitrary ratio approach, for example assuming freight and insurance costs as percentages of the value of imports (method to be avoided).
  - (vi) Extrapolate from resident’s experiences collected through enterprise surveys.
- Practical guidance for compilers:
  - Many methods are data intensive and may not be feasible regularly; compilers might need to estimate by: (i) calculating ratios of freight and insurance premiums to total imports (or groups of imported commodities) from a detailed analysis; and (ii) extrapolating ratios for more recent periods.
  - Extrapolation should consider changes in freight and insurance rates, capacity, type of transport, and commodity composition of imports.

### Reconciliation of SUTs, IMTS, and BPM6 valuation treatments
- Consistency across frameworks:
  - Valuation, time of recording, and ownership principles are consistent between the 2008 SNA and BPM6. Uniform valuation point used is FOB at the customs frontier of the exporting economy, providing consistent measurement when contractual arrangements differ.
- Use of IMTS data in SUT compilation (UN SUT Handbook, paragraphs 5.73-5.87):
  - Two types of adjustments described:
    - (i) Data adjustment: made to balance of payments data prior to entering SUTs to start from a consistent set with imports of goods valued FOB; adjustment made at product level starting from rest of the world account with imports valued FOB; adjustment consists of services linked to difference between FOB and CIF including transportation, insurance, and other expenditures between exporter’s exit point and importer’s entry point. If non-resident producers deliver these services, imports of services must be reduced to avoid double counting; if resident producers deliver them, a balancing service export of same value must be introduced so the net adjustment is zero. Composition of services making up CIF-FOB difference usually available from balance of payments working tables or regular surveys.
    - (ii) The CIF and FOB adjustment row: an ex post adjustment made to totals for exports and imports of goods and services to derive corresponding SNA totals (goods and services account and rest of the world account); purpose is to show SUTs consistency with sequence of accounts and avoid double-counting of CIF-type services provided by residents. Adjustments mirror those made as detailed data adjustments.
- SUT special cases (2008 SNA, paragraphs 14.73-14.75):
  - Merchanting: acquisition registered as negative exports and resale as positive exports; services to transport merchanted goods should be recorded consistently.
  - Goods sent abroad for processing: goods are not registered as exports; only processing fee is registered as export of services; transport costs for both movements must be registered as transportation services and treated according to which party is responsible for transport (imports by the party whose economy bears the cost, unless carried out by that party or a co-resident).

*Source: support-document-draft-guidance-note-g1 (IMF).*

### Appendix 3.2 IMTS valuation. Data collection in the European Union (EU).

### Appendix 3.2 IMTS valuation. Data collection in the European Union (EU)

### Statistical value and customs valuation
- A statistical value should be recorded for all goods covered in IMTS, according to the IMTS 2010.
- The main source for establishing the statistical value is the customs value placed on goods by customs administrations according to national and other relevant regulation.
- IMTS 2010 recommends adoption of the WTO Agreement on Customs Valuation as the basis for valuation of international merchandise trade for statistical purposes (IMTS 2010, paragraphs 4.1-4.5; IMTS 2010-CM, paragraphs 14.6-14.7).
- The WTO Agreement adopts the transaction value (the price actually paid or payable for goods, including some adjustments) as the customs value of imported goods, provided certain conditions for a fair, uniform and neutral valuation are met.
- The transaction value may include or exclude, in whole or in part:
  - (i) the cost of transport of the imported goods to the port or place of importation;
  - (ii) loading, unloading and handling charges associated with the transport of the imported goods to the port or place of importation;
  - (iii) the cost of insurance.
- The transaction value can therefore be defined differently depending on country customs practices and may partially or completely cover the cost of services performed to deliver the goods. (IMTS 2010, paragraphs 4.1-4.5; IMTS 2010-CM, paragraphs 14.6-14.7).

### Valuation approaches: FOB-type and CIF-type
- Under the WTO Agreement, countries can choose two approaches to valuation of goods: FOB-type or CIF-type.
- FOB-type values include:
  - the transaction value of the goods; and
  - the value of services performed to deliver goods to the border of the exporting country.
- CIF-type values include:
  - the transaction value of the goods;
  - the value of services performed to deliver goods to the border of the exporting country; and
  - the value of services performed to deliver the goods from the border of the exporting country to the border of the importing country.
- IMTS 2010 recommends:
  - (i) the statistical value of exported goods to be an FOB-type value;
  - (b) the statistical value of imported goods to be a CIF-type value;
  - it also encourages countries to compile FOB-type value of imported goods as supplementary information (IMTS 2010, paragraphs 4.6 and 4.8).
- In general practice:
  - countries use FOB-type values for the valuation of exports and CIF-type values for valuation of imports.
  - FOB-type values provide a uniform basis for goods valuation (single point of valuation at the border of the exporting country’s statistical territory).
  - FOB-type values of imports are more difficult to compile and are systematically applied by a limited number of countries.
  - Countries compiling only CIF-type values for imports are encouraged to compile separate data for freight and insurance at the most detailed commodity and partner level possible (IMTS 2010, paragraphs 4.7 and 4.9).

### Invoice price, delivery terms, and adjustments
- Statistical value and invoice price are different concepts: the invoice price represents an expected direct monetary payment to the seller and may not account for other payments to be included or excluded from the transaction value.
- Invoice price is usually a starting point and is adjusted as needed. It depends on terms of delivery and may include various service components.
- IMTS compilers should have information about delivery terms to identify specific items and their costs included or not included in the invoice price to obtain the recommended statistical value (IMTS 2010-CM, paragraph 14.8-4.9).
- Compilers make appropriate adjustments to the invoice price to obtain CIF or FOB-type values when these are not available from customs or other sources.
- Terms of delivery are defined by the International Chamber of Commerce; common types include FOB, FCA, DAF, CIF, CIP, and others. Practical equivalences noted:
  - Footnote 26 explains that FOB-type encompasses FOB, FCA at port of export, or DAF when applicable (IMTS 2010, paragraph 4.10).
  - Footnote 27 explains that CIF-type encompasses CIF or CIP at port of importation when applicable (IMTS 2010, paragraph 4.10).
- If terms of delivery other than CIF or FOB-type are used (e.g., “ex works”, or “free alongside ship”), other data sources must be used to establish FOB-type or CIF-type values by adding or subtracting cost items from the invoice price (IMTS 2010, paragraphs 4.10-4.11).

### Compilation of statistical value for imports and exports
- Statistical value of imported goods (IMTS 2010-CM, paragraphs 14.10-14.22):
  - If the customs value already includes required costs (insurance and freight), the customs value is the statistical value.
  - The customs value for imports should be accepted as the statistical value if:
    - (i) the customs value was established in accordance with articles 1-8 of the WTO Agreement;
    - (ii) the terms of goods delivery are CIF or CIP, and none of the exclusions from the customs value allowed in article 8 (2) were made.
  - If terms of delivery are not CIF or CIP, customs value should be accepted as statistical value provided appropriate adjustments to the invoice price were made by customs or the trader.
  - Customs often require a declaration of the customs value identifying cost components included, depending on terms of delivery.
  - If customs value deviates from WTO Agreement requirements or there is no customs value available, IMTS compilers should derive or estimate the statistical value following WTO Agreement principles.
  - If FOB-type values are not available, they can be estimated using:
    - actual or estimated freight and insurance costs provided by traders on declarations and supplemented by freight and insurance rate information from service providers;
    - CIF to FOB adjustment factors obtained from supplementary surveys of importers;
    - cooperation with authorities in exporting countries to obtain exported values, if systems and confidentiality rules allow.
  - Adjustment factors:
    - should be derived in detail by country, product, and mode of transport;
    - should be updated frequently;
    - are usually expressed as percentages of trade values and are an approximation.
    - insurance companies that insure goods when they leave a country are possible sources for insurance costs.
- Statistical value of exported goods (IMTS 2010-CM, paragraphs 14.24-14.28):
  - IMTS 2010 recommends an FOB-type value as the statistical value of exported goods.
  - Customs generally require actual prices paid and costs of delivery to the border to be declared; an FOB-type customs value can be established.
  - In the absence of price information, customs might require substitutes such as prices of identical or similar goods.
  - The customs value for exports should be accepted as the statistical value if the transaction value was established in accordance with articles 1-8 of the WTO Agreement and the terms of delivery were FOB or FCA; otherwise appropriate adjustments should be made to the invoice value.

### IMTS, balance of payments, and national accounts valuation differences
- IMTS are based on recording all goods that add to or subtract from the stock of material resources of a country by entering (imports) or leaving (exports) its economic territory.
- Balance of payments (BPM6) records transactions based on change of ownership between residents and non-residents.
- Timing: For BPM6, timing of change of ownership is usually assumed approximately the same as time of customs recording.
- Valuation:
  - IMTS recommends CIF-type valuation for imports and FOB-type valuation for exports.
  - BPM6 requires that imports and exports be valued at a point of uniform valuation at the customs frontier of the economy from which the goods are first exported.
  - Consequently, while IMTS recommends CIF-type for imports, balance of payments and national accounts require an FOB-type valuation for imports (IMTS 2010, paragraphs F.1-F.16).
- IMTS data are the main data source on trade in goods for the balance of payments and national accounts, but conceptual differences require adjustments to IMTS data for coverage, time of recording, valuation and classification.

### IMTS data sources and collection in the EU
- Main data source: customs declarations (including electronic forms such as ASYCUDA); however, customs data may not always fulfill statistical needs due to simplified declarations, exemptions, or thresholds.
- Alternative data sources (IMTS 2010-CM, paragraphs 3.1-3.28) include:
  - (i) parcel post and letter post records;
  - (ii) aircraft and ships registers;
  - (iii) enterprise surveys;
  - (iv) foreign shipping manifests;
  - (v) currency exchange records and records of monetary authorities;
  - (vi) reports of commodity boards;
  - (vii) administrative records associated with taxation;
  - (viii) data exchanges between countries.
- Customs unions abolish customs controls between members, making customs records unavailable and requiring alternative data sources such as taxation records (VAT or sales tax) and surveys.
- The single market establishment in the EU in 1993 required new data sources and led to Intrastat:
  - Intrastat characteristics:
    - (i) direct collection from consignees and consignors via monthly summary statements of transactions;
    - (ii) close link with the VAT system relating to intra-EU trade;
    - (iii) reduction of business workload through exemption or simplification thresholds.
  - Exemption threshold objective:
    - each Member State must guarantee that at least 97 percent of the total value of dispatches and 95 percent of the total value of its arrivals (measured based on VAT declarations) is directly collected.
  - For transactions above the threshold, EU companies must provide information on intra-EU goods transactions, specifying merchandise according to the Combined Nomenclature and providing value and quantity for each goods item.
  - Major concerns and challenges:
    - asymmetries in partner reporting;
    - respondent burden vs. data requirements and availability;
    - centralized customs clearance allowing enterprises to declare goods in only one country while physical clearance occurs in any country of the customs union, requiring exchange of information among countries to obtain complete data for trade with countries outside the customs union.
- The Single Administrative Document (SAD) is used for customs declarations in the EU and several neighboring countries, reducing administrative burden and increasing standardisation; the SAD is composed of a set of eight copies with different functions (source attribution in original: United Nations, IMTS 2010-CM).

### Implications for the 2008 SNA and other statistical domains
- Section 4 of the draft guidance identifies 2008 SNA paragraphs to be updated according to GZTT recommendations; finalization depends on consultation outcomes.
- Suggested updates to 2008 SNA Chapter 3 (paragraphs 3.149 and 3.150) to reflect GZTT recommendations should state that:
  - (i) the recommended valuation for imports and exports of goods should be as close as possible to the valuation of output at basic prices;
  - (ii) the use of transaction values is the closest valuation to the basic price concept;
  - (iii) the FOB valuation is an alternative when transaction values are not available;
  - (iv) practical considerations concerning availability of data sources and country-specific situations should determine the valuation used for imports and exports of goods;
  - (v) the valuation principles used for imports and exports of goods should be carefully explained in national metadata notes, along with the adjustments needed for the data.
- The input and feedback from CATT and the global consultation are considered essential; even if AEG and BOPCOM consultations prefer not to change current recommendations on valuation of imports and exports of goods, additional clarifications in SNA and BPM would be useful to compilers.

*Source: Appendix 3.2 IMTS valuation. Data collection in the European Union (EU), support-document-draft-guidance-note-g1.*

### 150. The SUT compilation is addressed in Chapter 14 of the 2008 SNA (The supply and use

### support-document-draft-guidance-note-g1 - 150. The SUT compilation is addressed in Chapter 14 of the 2008 SNA (The supply and use

### Supply and Use Tables (SUT): valuation of imports and transport margins
- Chapter 14 of the 2008 SNA (The supply and use tables and goods and services account) addresses SUT compilation; paragraphs 14.44 to 14.80 provide recommendations on the valuation of imports in the supply table.
- Specific recommendations to be reflected in updates:
  - Paragraphs 14.53-14.77: recommendations on transport margins.
  - Paragraphs 14.61-14.75: treatment of international transport margins.
  - Paragraphs 14.76-14.77: recording of transport margins in the SUTs.
- Recommended updates for paragraphs 14.61 to 14.75 on international transport margins:
  - (i) Recognize that data sources for integrating imports of goods at a detailed product level may be collected using different valuation concepts: transaction values, FOB-type, or CIF-type valuation.
  - (ii) In each case, explain the impact and adjustments needed in the transportation charges, including numerical examples as currently done.
  - (iii) Include updated recommendations on transport of merchanted goods (paragraphs 14.73-14.74) and transport on goods sent abroad for processing (paragraph 14.75).

### Treatment of transport margins and necessary adjustments
- Paragraphs 14.76-14.77 should provide clearer explanation of adjustments needed to source data by type of source data used.
- Guidance should include:
  - Recommendations on adjustments needed to basic data.
  - The CIF to FOB adjustment (if the AEG and BOPCOM decision implies retaining this adjustment).
  - Reference to the 2018 United Nations Handbook on Supply, Use and Input-Output Tables with Extensions and Applications for methodology.
- The CIF to FOB adjustment considerations:
  - Should take note of the treatment recommended in BPM.
  - Should be obtained at the detailed level (at least for the national accounts benchmark year).
  - Rationale: the relation of FOB to CIF prices depends on the type of good, weight, scale, special needs, mode of transport, and distance traveled.
  - CIF to FOB ratios change over time due to factors such as fuel prices, competition and technology in transport industry, change in the proportion of types of goods, and changes in the source economy.
- Alternative compilation methods and data sources should be explained as done in the BPM Compilation Guide.

### Exports in the use table and SNA–BPM valuation alignment
- Paragraph 14.114 of the 2008 SNA describes the treatment of exports in the use table.
- Update to reflect GZTT recommendations:
  - Exports should be referred to as valued at transaction prices or FOB, depending on available data sources.

### Rest of the world accounts and links to BPM6
- Chapter 26 of the 2008 SNA (The rest of the world accounts and links to the balance of payments) treats the relationship between the SNA rest of the world sector and international accounts as described in BPM6.
- Valuation of imports and exports is addressed in paragraph 26.19; this paragraph should be updated to incorporate BPM6 updates concerning valuation of imports and exports.
- Relation of the SNA with the goods and services account is included in paragraphs 26.49-26.57.
  - In particular, paragraph 26.52 could be updated to refer to alternative data sources with different valuation methods.

### Input-Output Tables and margins on imports
- Chapter 28 of the 2008 SNA (Input-output and other matrix-based analyses) describes compilation of Input-Output Tables.
- Paragraphs 28.9-28.12 address treatment of margins on imports and should be updated according to updated recommendations in Chapter 14.

### BPM6 accounting principles and goods/services valuation
- Chapter 3 of BPM6 (Accounting Principles) outlines general accounting principles for balance of payments statistics.
  - Paragraphs 3.68-3.80 concern valuation of transactions and should be updated as a result of GZTT work, after consultation with the BPTT.
- Chapter 10 of BPM6 (Goods and Services Account):
  - Paragraphs 10.30-10.40: recommendations concerning valuation of goods.
  - Paragraphs 10.78-10.79: treatment of freight services.
  - Paragraph 10.116: treatment of freight insurance services.
  - These paragraphs will be updated as a result of GZTT work, after consultation with the BPTT.

### IMTS 2010 valuation and data needs
- Chapter IV of IMTS 2010 (Valuation) addresses valuation of imports and exports of goods.
- IMTS is a main data source for compilation of imports and exports in balance of payments and national accounts.
- Potential implications and recommended actions:
  - Changes proposed to BPM6 and 2008 SNA may add requirements, especially for transaction values data.
  - After consultation with the BPTT, the GZTT will provide guidelines for the IMTS 2010 update concerning data needs for balance of payments and national accounts.

### Membership and coordination (Annex highlights)
- GZTT Chairs: Mr. Michael Connolly (CSO, Ireland) and Mr. Paul Roberts (ABS, Australia), until July 2020.
- Secretariat members: Ms. Jennifer Ribarsky (IMF), Ms. Margarida Martins (IMF), Ms. Padma Sandhya Hurree Gobin (IMF).
- GZTT includes representatives from national statistical offices, central banks, international organizations, and experts, such as:
  - Mr. Roberto Luís Olinto Ramos (FGV, Brazil)
  - Mr. Andrey Tatarinov (ROSSTAT, Russia)
  - Mr. Sanjiv Mahajan (ONS, UK)
  - Mr. Branko Vitas (ABS, Australia)
  - Ms. Grace Kim (ABS, Australia)
  - Mr. Dylan Rassier (BEA, US)
  - Ms. Ann Lisbet Brathaug (Statistics Norway)
  - Ms. Jennifer Withington (Statistics Canada)
  - Mr. Jens Walter (Deutsche Bundesbank)
  - Mr. Paul Feuvrier (Central Bank of Luxembourg)
  - Ms. Norhayati Razi (Bank Negara, Malaysia)
  - Mr. Satoru Hagino (Fukuyama University, Japan)
  - Ms. Lamia Laabar (Haut-Commisariat au Plan, Morocco)
  - Ms. Yang Can (State Administration of Foreign Exchange (SAFE), China)
  - Mr. Themba Munalula (COMESA)
  - Ms. Lourdes Erro (Uruguay)
  - Mr. Celestino Giron (ECB)
  - Mr. Nadim Ahmad (OECD)
  - Mr. Filippo Gregorini (Eurostat)
  - Mr. Steiner Todsen (Eurostat)
  - Mr. Herman Smith (UNSD)
  - Mr. Tihomira Dimova (UNECE)
  - Mr. Rami Peltola (UNECE)

---


_Source: https://www.imf.org/-/media/files/data/statistics/bpm6/gztt/support-document-draft-guidance-note-g1.pdf_