Glossary

Comext monthly trade files

Comext is Eurostat's domain of intra/extra-EU trade-in-goods statistics, published as one complete file per dataset per reference month in CSV and TSV, running from January 1988 to the latest reference month. It is reachable through the Eurostat bulk download facility and queryable through the data browser and public APIs.

What is Comext?

Comext is the trade-in-goods domain inside Eurostat's dissemination system: monthly and annual files covering flows between EU member states (intra-EU) and between the EU and non-EU partners (extra-EU), classified by product code, reporter and partner country.

The cataloged Eurostat Database - Bulk Download (Eurobase) record describes it as a dedicated domain "of monthly and annual CSV trade files back to 1988", sitting alongside the main database that refreshes twice daily. Its file convention is unusual and useful: "Complete multi-dimensional dataset files, one file per dataset per period for Comext" - each reference month arrives complete, so archival pulls never need stitching.

The same domain surfaces in the Eurostat Data Browser record (quality score 10) covering apparel, textiles and trade topics, with Comext trade files dated "from January 1988 to the latest reference month". Formats there span JSON, SDMX, TSV, CSV and XLSX under commercial delivery terms 4.0 with no registration.

Why does Comext matter when choosing a dataset?

History depth and completeness are the two properties buyers usually compromise on elsewhere. Most trade products start in the 2000s and ship slices; Comext's files run back to January 1988 and land whole months at a time, which makes longitudinal and reproducible-pipeline work straightforward.

The buyer consequence is access-model fit. Bulk files suit warehouse loads - one complete file per dataset per period - while the Eurostat Data Browser and APIs suit interactive filtering; the cataloged records document both under the same commercial delivery terms 4.0 attribution terms with no key required, confirmed by unauthenticated queries during cataloging. Choosing the wrong route shows up as either oversized downloads or needless request volume.

There is also a topical-fit consequence: Comext covers goods trade only. Production volumes need PRODCOM and consumer prices need HICP, which the catalog pairs in the same browser topic - mixing them up wastes a build cycle.

How do you evaluate Comext-based sources?

Check the following before adopting a Comext-derived product:

  1. Verify the 1988 baseline. Files from January 1988 to the latest reference month is the cataloged standard; shorter histories should be justified.
  1. Confirm the complete-file-per-month convention. It is what distinguishes Comext bulk distribution from partial extracts.
  1. Check format spread. CSV, TSV, SDMX and JSON coverage in the cataloged records means fewer conversion steps.
  1. Test anonymous access. No registration or integration key is required per the cataloged terms - a gated mirror is adding friction, not value.
  1. Mind the topical boundary. Goods trade here; PRODCOM production and HICP prices live in sibling datasets of the same browser topic.

BACI trade cube is the OEC-harmonized bilateral alternative built on HS revisions rather than Comext's EU reporting framework.

HS code and PRODCOM are the product classifications flanking Comext on the trade and production sides.

Industries where the catalog applies this term: apparel-accessories-luxury-goods data and home-improvement-retail data.

Frequently asked questions

How far back do Comext files go?

To January 1988 per both cataloged Eurostat records, running to the latest reference month, with monthly and annual CSV/TSV files each containing the complete dataset for that period.

Is Comext data accessible on demand?

Yes. The cataloged records apply commercial delivery terms 4.0 under Eurostat's free re-use policy, with attribution and no registration or integration key required - confirmed by unauthenticated API queries during cataloging.

Datasets containing this field

Datasets containing Comext monthly trade files

6 datasets carry comext monthly trade files in the catalog. Open one, count the fields, judge for yourself.

Apparel, Accessories & Luxury Goods Global web-sourced shop and consumer… · Image level, one record per…

DeepFashion Dataset (CUHK MMLab)

Apparel, Accessories & Luxury Goods

World Bank DataBank & API — World Development Indicators

indicator.id · indicator.value · country.id …+7 more

Apparel, Accessories & Luxury Goods India · Product level, one row per SKU

Fashion Product Images Small — Hugging Face (ashraq)

gender · masterCategory · subCategory …+4 more

Apparel, Accessories & Luxury Goods

Apparel Images Dataset — Kaggle (trolukovich)

black_pants · red_dress · white_shorts

Every listing shows the field dictionary, sample rows, and coverage before you commit. API, files, or your warehouse. Daily, weekly, or hourly.

Get sample rows