Industry
How to Read China Vehicle Export Data: Shipments, Registrations and Local Assembly
A practical guide to comparing China vehicle export data across customs records, manufacturer shipments, overseas registrations and local assembly.

China vehicle export data can answer several useful questions—but not with one universal number. A customs total can show how many China-built vehicles cleared the border. A manufacturer shipment series can indicate factory dispatches and wholesale activity. Registration data can show vehicles entering use in a destination market. Overseas assembly data can reveal a production footprint that may not appear in vehicle-export totals at all.
The practical rule is simple: before comparing figures, identify what is counted, where it was produced, when it was recorded, and whether it was a complete vehicle or an assembly kit.
This matters for importers assessing supply exposure, fleet operators tracking available models, suppliers evaluating production demand, and analysts estimating Chinese automakers’ real position in overseas markets. A shipment number should not automatically be treated as retail demand; a registration total should not be treated as a measure of exports from China; and a brand’s overseas production may sit partly outside both.
Start by matching the number to the question
The first question is not, “What is China’s vehicle export total?” It is, “What exactly am I trying to measure?”
Different questions require different datasets:
| Reader question | Most relevant measure | What it does not prove by itself |
|---|---|---|
| How many vehicles physically left China? | Customs export declarations | Retail sales or registrations overseas |
| How many vehicles did manufacturers report as exports? | Industry shipment data | Customs timing or final-market demand |
| How many vehicles entered service in a country? | Registration or delivery data | China production origin in every case |
| How large is a brand’s overseas manufacturing footprint? | Plant, assembly and production disclosures | China-built export volume |
A useful staged framework runs from export clearance to market entry:
- A vehicle is declared for export from China.
- A manufacturer or industry body records an export shipment or dispatch.
- The vehicle arrives, is distributed, or enters inventory in the destination market.
- It is registered, delivered, or sold to a customer or fleet.
Those stages may happen weeks or months apart. They can also be recorded under different definitions. China Vehicles’ guide to shipments, destinations and overseas sales explains why treating those stages as interchangeable creates misleading comparisons.
Before using a figure, check five points:
- Measurement point: export declaration, factory shipment, arrival, registration, delivery or retail sale.
- Product scope: passenger cars, commercial vehicles, buses, trucks, chassis, complete vehicles, or parts.
- Powertrain scope: all vehicles, internal-combustion vehicles, battery-electric vehicles, plug-in hybrids, or another subset.
- Production origin: built in China, built abroad by a Chinese-owned company, or assembled locally from imported kits.
- Period basis: calendar month, quarter, year-to-date period, model year, or registration month.
A number becomes useful only after these definitions are established.
Customs export data: vehicles that cleared China’s border
Customs data is generally the best starting point when the question is about physical trade flows from China. It is based on export declarations and commodity classifications: in other words, goods that were reported as leaving China under specified trade codes.
For analysts of logistics, trade exposure and country-to-country flows, that is a meaningful distinction. Customs records can help show:
- the value and volume of declared vehicle exports;
- the destination markets receiving China-built vehicles;
- shifts between vehicle categories where tables provide that detail;
- changes in the trade mix over time; and
- the relationship between vehicle exports and parts exports.
However, customs data is not a direct measure of overseas consumer sales. A vehicle may be in transit, awaiting customs clearance in the destination market, stored at a port, held by an importer, or sitting in dealer inventory. None of those conditions necessarily means the vehicle has reached an end user.
The commodity definitions also matter. A table labelled “automobiles” may not have the same scope as a table limited to passenger vehicles or complete road vehicles. Depending on the classification and release, data can distinguish between passenger vehicles, commercial vehicles, buses, trucks, chassis, complete vehicles and components. It may also separate volume from customs value.
This is one reason customs and industry-association totals can differ. As CNN Business noted in its comparison of Chinese and Japanese export data, customs and industry figures may cover different categories, with customs reporting potentially including a broader range of vehicle or parts classifications. The conclusion is not that one source is wrong; it is that the figures may be measuring different baskets of goods.
For a trade-flow analysis, use customs data—but document the commodity codes, unit of measurement and whether the series covers complete vehicles only. Do not use the resulting total as a shortcut for overseas retail demand.
Industry shipment data: manufacturer-reported exports and wholesale activity
Industry datasets are useful because they can provide a timely view of manufacturer output, export dispatches and changes in the vehicle mix. In China, readers often encounter export data from the China Association of Automobile Manufacturers (CAAM), as well as passenger-vehicle datasets that may have narrower coverage.
The key concept is wholesale activity. The International Energy Agency, when presenting charts based on CAAM data, describes wholesale sales as factory-level shipments by automakers to dealers and other distribution channels, including exports. That makes such data particularly useful for understanding the supply side of the market: what manufacturers report shipping, rather than what final customers have registered.
Industry shipment figures can be valuable for tracking:
- manufacturer export momentum;
- broad trends in battery-electric, plug-in hybrid and combustion-vehicle exports where reported;
- all-vehicle versus passenger-vehicle activity, depending on the dataset; and
- the timing of changes in factory dispatches.
Yet a manufacturer-reported export figure may not match a customs total in the same month or year. Several factors can create a gap:
- reporting may occur at factory dispatch rather than border clearance;
- the industry series may use a different vehicle scope;
- passenger-vehicle-only reporting will not match an all-vehicle series that includes commercial vehicles;
- the treatment of chassis, kits or other non-complete-vehicle items may differ; and
- calendar cut-offs and revisions may not align.
A comparison should therefore specify the reporting body and scope. “China exported X vehicles” is incomplete unless the reader knows whether X means customs-cleared vehicles, industry-reported manufacturer exports, or a passenger-vehicle subset.
This distinction is especially important in battery-electric vehicle analysis. A series may accurately show a rise in manufacturer-reported electric-vehicle exports while still differing from customs data because the product classification, reporting date or vehicle coverage is not identical.
Overseas registrations: the best available signal of market entry
Registration data is usually closer to actual market entry than a shipment figure. When a vehicle is registered for road use, it has normally passed beyond the export, shipping and wholesale stages. For country-level demand analysis, registrations can therefore be more informative than export declarations.
But registrations are not a perfect global sales total.
Registration systems differ substantially by market. Some data reflects new registrations; some includes fleet registrations; some providers report deliveries or retail sales instead; and some markets have incomplete public coverage. Timing also varies. A vehicle exported from China in December may arrive, clear local processes and be registered in January or later.
The gap between exports and registrations can be explained by normal commercial conditions:
- ocean transit and port handling;
- importer and distributor inventory;
- dealer stock before retail delivery;
- local homologation, inspection or registration processes;
- fleet delivery schedules; and
- differing reporting calendars.
Brand identity can complicate the comparison as well. A vehicle produced in China may be sold abroad under another marque, through a joint-venture structure, or under a market-specific branding arrangement. As a result, a destination-market brand registration series may not cleanly map to a China customs export total.
The appropriate conclusion from a registration dataset is usually narrow: it indicates reported vehicles entering the market under that jurisdiction’s rules and that source’s coverage. It should not be presented as a definitive count of every China-built vehicle sold worldwide.
For destination-market analysis, prioritize the relevant registration authority or a market-data provider that clearly explains its coverage. Then state whether the data represents registrations, deliveries, retail sales or another proxy. China Vehicles’ earlier explainer on Chinese vehicle exports, shipments and overseas sales provides additional context on this distinction.
CBU, SKD and CKD: why local assembly changes the comparison
A further complication is that not every overseas vehicle sold by a Chinese automaker was exported from China as a finished vehicle.
Three terms are widely used in international automotive trade:
- CBU (completely built unit): a finished vehicle shipped for import.
- SKD (semi-knocked-down): a partly assembled vehicle kit intended for final assembly in another market.
- CKD (completely knocked-down): a more fully disassembled set of components shipped for local assembly.
The practical implication is significant. A vehicle assembled in a destination country from kits sourced in China may contribute to a Chinese company’s overseas presence, but it may not appear in a China-built complete-vehicle export total.
Depending on the customs classification and reporting system, kits may be recorded in parts trade, in a separate kit category, or under another treatment. The answer is not universal. Customs rules, tariff structures, local-content requirements and assembly definitions differ by importing market.
This matters for several types of comparison:
Comparing brands
A brand that relies heavily on local assembly may have a larger overseas market footprint than its China CBU export volume suggests. Conversely, a brand shipping finished vehicles from China may appear more prominent in customs data even if it has fewer overseas manufacturing operations.
Comparing destination markets
Markets with policies encouraging local assembly can show lower finished-vehicle imports while still receiving substantial Chinese-origin content or production equipment. Looking only at CBU imports can understate the commercial relationship.
Assessing supplier opportunity
For suppliers, an SKD or CKD programme can change the demand profile. It may create local needs for assembly services, logistics, packaging, selected locally sourced components and after-sales support, even when the core vehicle architecture remains China-developed.
Avoid assuming that every kit is treated the same way. The degree of disassembly, local value addition and customs classification must be verified market by market.
A practical comparison framework for analysts, buyers and suppliers
A disciplined comparison does not force datasets to reconcile perfectly. It explains why they differ.
Use this six-step process.
1. Define the business question
Decide whether the goal is to measure trade exposure, manufacturer output, destination demand, brand penetration or overseas production capacity. One dataset rarely answers all five.
2. Select the measurement point
Use customs declarations for border trade, industry data for factory-level shipments, and registrations or clearly defined delivery data for market entry.
3. Align vehicle scope
Check whether each series covers all vehicles, passenger vehicles only, commercial vehicles, complete vehicles, chassis, or kits. Also align powertrain categories before comparing electric-vehicle figures.
4. Identify the production origin
Separate China-built CBU exports from overseas assembly and overseas manufacturing. A Chinese brand’s international sales may include both, while customs data from China captures only part of that picture.
5. Align the time period
Compare calendar months with calendar months and calendar years with calendar years. Where shipment and registration timing differs, allow for transit, inventory and registration lags before interpreting a gap as a demand change.
6. Explain residual gaps instead of hiding them
If customs exports exceed registrations, possible explanations include in-transit vehicles, inventory, incomplete registration coverage or branding differences. If industry shipments differ from customs declarations, examine reporting dates and product coverage first.
The table below provides a compact source-selection guide.
| Objective | Preferred starting source | Main caution |
|---|---|---|
| Measure China-built vehicle trade | China customs export data | Confirm codes and treatment of complete vehicles versus kits |
| Track factory export activity | CAAM or other manufacturer-reported data | Do not equate wholesale shipments with retail demand |
| Assess country-level market entry | Destination registrations or defined delivery data | Registration rules and coverage vary by market |
| Map overseas production footprint | Manufacturer plant disclosures and local production data | Do not infer production volume from export data alone |
| Assess imported-kit activity | Customs and local assembly documentation | CKD/SKD classification is jurisdiction-specific |
What to watch when reading a headline export figure
When a report cites a large China vehicle export number, ask four follow-up questions:
- Is this a customs total or a manufacturer shipment total?
- Does it include passenger and commercial vehicles, chassis, kits or parts?
- Is it measuring China-built vehicles, Chinese-brand vehicles, or both?
- Does the claim concern exports, destination registrations, or a broader measure of overseas sales?
Those questions make a major difference. A China-built vehicle can be exported under a foreign brand. A Chinese brand can sell a vehicle assembled outside China. A vehicle can be exported in one period and registered in another. And a kit can support overseas production without appearing as a finished-vehicle export.
The most reliable reading of China vehicle export data is therefore not a single headline number. It is a clearly labelled view of where the vehicle was produced, what crossed the border, when it was counted and whether it ultimately entered service in the destination market.
FAQ
Why do China customs export totals and CAAM export figures differ?
They can use different reporting points and scopes. Customs data records declared trade flows at the border under commodity classifications, while CAAM-based data reflects manufacturer-reported wholesale or export shipments. Differences in timing, vehicle categories, chassis or kit treatment, and all-vehicle versus passenger-vehicle coverage can all matter.
Does a vehicle exported from China count as an overseas sale?
Not necessarily. Export clearance shows that a vehicle left China under an export declaration. It may still be in transit, in importer inventory or awaiting retail delivery and registration. Overseas sales or registrations are later market stages.
Are CKD and SKD kits included in China vehicle export data?
It depends on the dataset and customs classification. A complete-vehicle series may exclude kits, while kit shipments may be recorded under parts or another category. The specific trade codes and importing-market rules must be checked before drawing conclusions.
Do overseas registrations include Chinese brands assembled outside China?
They may, depending on the registration dataset. Registration data generally reflects vehicles entering use in the local market rather than their production origin. To separate China-built imports from locally assembled vehicles, combine registration data with origin, plant or customs information where available.