A headline says exports grew 21%. Does that mean 21% more goods were produced, or did prices and the mix of products change? In Singapore, did the headline count domestic exports, re-exports or both?
The gold standard of trade statistics is a reproducible description of cross-border economic flows with correct definitions, valuation, time and uncertainty. This 23-chapter guide follows Singapore’s official classifications, WTO measures and original worked examples.
Quick information: trade measures are not interchangeable
| Term | What it describes | Misleading shortcut |
|---|---|---|
| Domestic exports | Qualifying Singapore-origin goods | All are entirely locally sourced |
| Re-exports | Imported goods exported without qualifying transformation | Same as domestic manufacturing |
| Total merchandise exports | Domestic exports plus re-exports | Domestic GDP or gross value added |
| Merchandise balance | Exports minus imports under set definitions | National wealth lost or gained |
| Trade volume | Volume index or comparable physical measure | Nominal dollar value |
| Services trade | Cross-border service transactions under rules | Container port throughput |
Worked example: Singapore-style export decomposition
| Fictional annual goods value | S$ million |
|---|---|
| Domestic exports | 120 |
| Re-exports | 80 |
| Total merchandise exports | 200 |
| Merchandise imports | 230 |
| Exports minus imports | −30 |
The −S$30 million merchandise balance is not a claim about lost wealth or the broader balance of payments. The invented figures are merely a study of statistical identities.
Worked example: nominal value and price-adjusted volume
| Hypothetical index | Base year | Later period |
|---|---|---|
| Merchandise export value | 100 | 121 |
| Compatible price index | 100 | 105 |
| Simplified volume index | 100 | 121/1.05 ≈ 115.24 |
Under simplifying assumptions, nominal value rises 21% while the modelled comparable volume measure rises about 15.24%. Real statistical offices publish method-specific indices; these are not real official values.
Changed-context example: percentage base matters
| Movement | Absolute difference | Percentage change |
|---|---|---|
| From 80 to 100 | +20 | +25% |
| From 100 to 80 | −20 | −20% |
Identical absolute changes create different percentages because the denominator changes. The same principle applies to annual trade comparisons and policy headlines.
All 23 trade-statistics controls
Merchandise and classification
Chapters 1–6: Trade statistics begin with a definition of what crossed · Distinguish merchandise from services trade · Define domestic exports and re-exports in Singapore · Worked Singapore-style decomposition of exports · Interpret gross imports without implying all are consumed · Know the merchandise trade balance arithmetic.
Valuation and time-series interpretation
Chapters 7–12: Account for FOB and CIF valuation differences · Distinguish merchandise and balance-of-payments goods · Keep nominal value and physical volume separate · Worked value-versus-volume illustration · Percentage changes must retain their base · Use comparable time periods and seasonal context.
Partners, services and production
Chapters 13–18: Separate product classification from country partner · Re-exports and transshipment are not synonyms · Check revisions before claiming a latest figure · Classify commercial services appropriately · Partner figures may need reconciliation · Watch currency conversions and their time basis.
Policy evidence and reporting
Chapters 19–23: Commodity-price shifts can dominate an aggregate · Gross exports are not domestic production value added · Use concentration statistics without inventing welfare · Separate descriptive data from trade-policy effects · Publish an auditable statistical release.
- Trade statistics begin with a definition of what crossed
- Distinguish merchandise from services trade
- Define domestic exports and re-exports in Singapore
- Worked Singapore-style decomposition of exports
- Interpret gross imports without implying all are consumed
- Know the merchandise trade balance arithmetic
- Account for FOB and CIF valuation differences
- Distinguish merchandise and balance-of-payments goods
- Keep nominal value and physical volume separate
- Worked value-versus-volume illustration
- Percentage changes must retain their base
- Use comparable time periods and seasonal context
- Separate product classification from country partner
- Re-exports and transshipment are not synonyms
- Check revisions before claiming a latest figure
- Classify commercial services appropriately
- Partner figures may need reconciliation
- Watch currency conversions and their time basis
- Commodity-price shifts can dominate an aggregate
- Gross exports are not domestic production value added
- Use concentration statistics without inventing welfare
- Separate descriptive data from trade-policy effects
- Publish an auditable statistical release
CONTROL 1 OF 23 · All controls
1. Trade statistics begin with a definition of what crossed
Trade statistics describe economic exchanges and movements using formal recording rules. The first question is not which economy exported the most, but which goods, services and transaction boundaries the dataset counts. Different sources can publish different-looking values while applying legitimate but distinct statistical systems.
The gold standard is traceability: identified products, reporting economy, partner, currency, time period, valuation and whether values are preliminary or revised. Trade Economics discusses mechanisms and consequences; this guide establishes what the recorded figures can and cannot prove.
CONTROL 2 OF 23 · All controls
2. Distinguish merchandise from services trade
Merchandise trade statistics generally concern goods moving across relevant territorial boundaries under a specified recording system. Services trade includes cross-border and other supply arrangements that are not adequately represented by container counts. A trade dashboard that covers goods alone cannot describe all international economic exchange.
The WTO’s statistics portal publishes separate merchandise and commercial-services datasets, with annual, quarterly and sometimes monthly information. Keep the source concept and classification explicit; do not add unrelated physical and service figures without first checking compatible reporting units and dates.
CONTROL 3 OF 23 · All controls
3. Define domestic exports and re-exports in Singapore
Singapore’s merchandise exports comprise domestic exports and re-exports in its published trade definitions. Domestic exports include goods of Singapore origin, including manufacturing or processing that transforms imported materials. Re-exports are imported goods exported in the same form without transformation under the relevant rules.
The Singapore Department of Statistics merchandise-trade glossary specifies these concepts. Repacking, sorting or splitting into lots may not constitute the transformation needed for domestic-origin classification. A high re-export value is not evidence that the entire value was manufactured domestically.
CONTROL 4 OF 23 · All controls
4. Worked Singapore-style decomposition of exports
Suppose a wholly fictional reporting economy records domestic exports worth S$120 million and re-exports worth S$80 million during a defined year. Total merchandise exports are S$120 million + S$80 million = S$200 million, with a 60% domestic-export share and 40% re-export share.
The components can matter to production analysis, but the domestic-export amount is still not identical to domestic value added. Domestic-origin products can incorporate imported intermediate inputs. Conversely, logistics and services surrounding re-exports can create local economic activity without transforming those goods.
CONTROL 5 OF 23 · All controls
5. Interpret gross imports without implying all are consumed
Merchandise imports can include inputs used in manufacturing, capital equipment, consumer goods and items subsequently re-exported. A statement that the country ‘consumed every imported dollar’ would be unsupported unless the accounting system actually measures that use.
The Enterprise Singapore trade statistics glossary describes imports and the domestic-export/re-export relationship. For analytical questions about domestic retained goods, supply chains or final demand, additional categories and reconciliations may be needed.
CONTROL 6 OF 23 · All controls
6. Know the merchandise trade balance arithmetic
A merchandise trade balance is the value of exports minus the value of imports under compatible reporting definitions. In the fictional case, suppose imports are S$230 million while total exports are S$200 million. The recorded merchandise balance is S$200 million − S$230 million = −S$30 million.
This is a net merchandise flow, not proof that the economy lost S$30 million of wealth. The broader external accounts include services and financial or income flows. The economic meaning depends on production, investment, financing and the period being studied. Economics and Trade Economics explain those contexts.
CONTROL 7 OF 23 · All controls
7. Account for FOB and CIF valuation differences
Merchandise exports are commonly recorded on a free-on-board basis while imports may include cost, insurance and freight on a CIF basis. These are valuation concepts, not just acronyms to reproduce. Comparing export and import amounts without understanding their bases may conflate freight charges and merchandise value.
Singapore’s merchandise-trade glossary identifies FOB exports and CIF imports, while the WTO technical notes describe these conventions. Actual trade datasets should be checked individually. A reader should not assume that differences in two published values represent only a quantity imbalance.
CONTROL 8 OF 23 · All controls
8. Distinguish merchandise and balance-of-payments goods
Territory-based merchandise statistics and balance-of-payments concepts can differ in coverage because their underlying frameworks serve different purposes. Changes in economic ownership and goods that cross borders do not always map one-to-one.
UNCTAD’s Handbook of Statistics 2025 highlights the distinction between international merchandise-trade recording and balance-of-payments goods. Use the appropriate framework for the question and do not casually add merchandise and financial-account concepts into one homemade total. Source metadata should accompany the figures.
CONTROL 9 OF 23 · All controls
9. Keep nominal value and physical volume separate
Trade value can rise because prices increase, because quantities increase or because the composition of goods changes. A headline nominal export value does not automatically tell whether more physical goods were shipped or more real output was produced.
The WTO publishes merchandise trade value and volume indicators separately. When analysing an index, identify its base period, price deflator and product composition. An aggregate volume index is not a simple tonnage count when products have very different economic characteristics.
CONTROL 10 OF 23 · All controls
10. Worked value-versus-volume illustration
Imagine a fictional export value index rises from 100 to 121 while the matching price index rises from 100 to 105. Under a simplified multiplicative index relationship, a comparable volume index relative to the original base is 121/1.05 ≈ 115.24, a rise of about 15.24%.
That is different from the 21% nominal-value increase. The simplified calculation assumes suitable compatible indices and ignores methodology details; it is not an actual Singapore or WTO index. A researcher must use the official published constant-price or volume series where available rather than invent a deflator from unrelated prices.
CONTROL 11 OF 23 · All controls
11. Percentage changes must retain their base
An increase from 80 to 100 is 20/80 = 25% relative to 80. A return from 100 to 80 is a fall of 20/100 = 20%. Equal absolute movements produce unequal percentages because their denominators differ.
Trade commentaries often say a value recovered to its earlier level without noting the larger percentage rebound required after a decline. Label which month or year serves as base. Before aggregating changes across products, consider how their initial values and classifications differ.
CONTROL 12 OF 23 · All controls
12. Use comparable time periods and seasonal context
Monthly export values can vary with working days, holidays, commodity prices, shipment timing and cyclical patterns. A month-on-month change is not the same comparison as a year-on-year change. Seasonal adjustment creates another series whose methodology should be disclosed.
The WTO’s trade statistical resources provide short-term and longer-run series, with revisions. For Singapore, check whether a cited number refers to seasonally adjusted data, a raw monthly value or a three-month moving average. Do not combine the measures as though all describe the same fact.
CONTROL 13 OF 23 · All controls
13. Separate product classification from country partner
Trade statistics can be grouped by commodity or product type and by partner economy. Classification systems such as HS or SITC serve different analytical structures and can change with revisions. A code that appears stable numerically may describe a changed product category across versions.
Preserve product nomenclature, revision, partner definition and any conversion mapping. If an export passes through another economy, the recorded partner can depend on origin, destination or consignment rules, not simply a route shown on a map. Trade Compliance provides the associated regulatory distinctions.
CONTROL 14 OF 23 · All controls
14. Re-exports and transshipment are not synonyms
Re-exports describe goods recorded as previously imported and then exported without qualifying transformation under the applicable trade-statistics rules. Transshipment refers to a customs or logistics procedure for transferring cargo between means of transport under specified conditions. A port handling many containers does not imply every container represents a re-export.
Singapore’s Enterprise Singapore trade statistics FAQ explains re-exports and the general trade recording system. A dataset about port throughput counts a different event from merchandise exports, so the two totals should not be added together to represent national trade.
CONTROL 15 OF 23 · All controls
15. Check revisions before claiming a latest figure
Trade statistics can be preliminary, updated and subsequently revised when new source information arrives. The first released figure may change without the old analysis having been fabricated. A trustworthy article states the reporting period and the dataset vintage.
The WTO Global Trade Statistics portal identifies regular update cycles for merchandise and commercial services. On 8 October 2026, the WTO published revised merchandise statistics and short-term data updates in its statistics system. That date matters to a 2026 analysis; never present an undated cached figure as the current official value.
CONTROL 16 OF 23 · All controls
16. Classify commercial services appropriately
Travel, transport, financial and digitally delivered services have distinct concepts and statistical reporting methods. A tourist’s expenditure can contribute to travel-related services statistics but is not identical to the country’s entire tourism GDP or all commercial service exports.
The WTO Global Services Trade Data Hub provides datasets for commercial services, modes of supply and digitally delivered services. Their classification, geographical and temporal coverage may differ. The Tourism Data Analytics root explores the importance of matching the service measure to the decision.
CONTROL 17 OF 23 · All controls
17. Partner figures may need reconciliation
An export recorded by one reporting economy need not equal the corresponding import recorded by the partner. Timing, valuation, origin/consignment rules, statistical territory and reporting practices may differ. Such asymmetries deserve investigation rather than immediate allegations of fraud.
Compare the exact bilateral product categories, year, currency, value basis and data revision. A partner’s recorded imports may use CIF while exports are FOB, making mechanical equality inappropriate. The aim is a transparent reconciliation, not to force both datasets to match by arbitrary adjustment.
CONTROL 18 OF 23 · All controls
18. Watch currency conversions and their time basis
International statistics often convert local values into a common reporting currency. Movements in exchange rates can alter apparent US-dollar trade values even when the local-currency amount is unchanged. Conversion methods may use transaction-time rates or period averages under the reporting system.
A fictional S$100 million trade value converts to US$74.07 million at 1.35 SGD/USD or US$75.76 million at 1.32 SGD/USD. The underlying Singapore-dollar value is unchanged while the displayed US-dollar value rises about US$1.69 million. This example is not a current exchange-rate quote or real trade statistic.
CONTROL 19 OF 23 · All controls
19. Commodity-price shifts can dominate an aggregate
A country’s trade value may change because energy or other commodity prices move sharply. Exported tonnage and trade value can therefore move in different directions. Product mix also matters: a shift toward higher-value goods raises aggregate value without necessarily increasing physical volume.
Use product-level analysis and appropriate price/volume series where the question requires physical growth. Avoid saying that a headline trade value proves every industry expanded. The Time Series Analysis root explains how trends, shocks and relevant denominators change interpretation.
CONTROL 20 OF 23 · All controls
20. Gross exports are not domestic production value added
Goods can contain imported components and can cross borders multiple times during production. Gross export values record transactions under trade-statistics rules; domestic value added captures the domestic production contribution under appropriate national accounts or input-output frameworks.
If a hypothetical exported item sells for S$200 and contains S$120 of imported components, the difference of S$80 is not automatically the complete domestic value added because other intermediate inputs and accounting treatments may apply. The example shows why a direct subtraction can be informative but insufficient.
CONTROL 21 OF 23 · All controls
21. Use concentration statistics without inventing welfare
Trade concentration measures can describe how flows are distributed among partners or products. For a fictional exporter whose three main partner shares are 40%, 35% and 25%, the sum of squared shares is 0.40²+0.35²+0.25² = 0.345.
That index reports relative distribution under the chosen aggregation. It does not by itself establish resilience, risk, local job creation or a policy’s success. A highly concentrated export pattern may create dependencies, but actual risks depend on products, substitutes and geopolitical or logistical conditions.
CONTROL 22 OF 23 · All controls
22. Separate descriptive data from trade-policy effects
A rise in exports after a free trade agreement, customs change or campaign can coincide with price movements, economic recovery or new investment. The observed time series alone does not prove the policy caused the entire change.
A credible evaluation needs a baseline, relevant comparison, timing and analysis of confounding factors. Trade Policy considers public instruments while statistical methods establish what was observed. Do not choose the easiest before-and-after window and call it a causal assessment.
CONTROL 23 OF 23 · All controls
23. Publish an auditable statistical release
An independent reader should be able to find dataset title, official source, extraction date, economy, partner, unit, product classification, time period, valuation, missing data, revision status and calculated formulas. This is the difference between a reproducible report and an attractive headline.
Where newer official data replace an older source, update the page transparently. The immutable Secondary 1 Mathematics Tutor Clementi Small Groups Tutorials page provides a transferable learning model: name the denominator, demonstrate the arithmetic and check a changed case before drawing conclusions.
Official trade-data audit
| Check | Evidence to retain | Mistake prevented |
|---|---|---|
| Scope | Goods vs services and recording system | Different statistical universes combined |
| Origin and purpose | Domestic exports, re-exports and product classes | Re-export amount called domestic manufacture |
| Value basis | FOB, CIF, currencies and prices | Money value taken as physical volume |
| Temporal record | Period, release date, revision and seasonality | Stale preliminary figure called final |
| Reconciliation | Partner, product, transport and statistical differences | Bilateral asymmetry called automatic fraud |
| Inference | Denominator, comparison and uncertainty | Observed growth called policy causation |
Frequently asked questions
Are Singapore re-exports the same as domestic exports?
No. Domestic exports involve qualifying Singapore origin or transformation, while re-exports are imported goods exported without the relevant transformation.
Is merchandise trade identical to services trade?
No. The populations and statistical classifications differ.
Does an export value increase prove more physical goods were shipped?
No. Price changes, composition and currency may affect value; use appropriate volume measures.
Does a merchandise deficit mean a country lost that amount of wealth?
No. It is a defined net flow, not a complete welfare or external-account measure.
Are FOB exports and CIF imports identical valuation bases?
No. Freight and insurance treatment differs under their conventions.
Why can two countries’ bilateral trade statistics disagree?
Differences in valuation, timing, origin, coverage and reporting methods can cause asymmetries.
What is the first audit check for a trade headline?
Find the official series definition, period, unit, valuation, classification and revision date.
Singapore and international primary statistical references
Official sources: Singapore Department of Statistics, Merchandise Trade: Our Data Explained; Enterprise Singapore trade statistics glossary; WTO Global Trade Statistics; WTO Statistical Technical Notes; UNCTAD Handbook of Statistics 2025; and WTO October 2026 short-term trade releases. Each is used for its statistical definitions and provenance. All calculated trade values here are fictional, not reported Singapore trade outcomes.
Connected eduKate Gold Standards
Trade Economics · Trade Policy · Trade Compliance · Trade Finance · Trade Logistics · International Trade · Statistics · Data Science · Time Series Analysis · Economics · Supply Chain Management · Public Policy · Evidence-Based Thinking · Education
All 23 controls · Trade Economics · Trade Policy · Education