Statistical Data and Metadata Exchange (SDMX)

What is SDMX?

SDMX (www.sdmx.org) is an international standard (ISO 17369:2013) for describing and exchanging data and metadata between organisations.

SDMX initially focused on statistical macro aggregates, but has evolved to also support micro data as well as a number of unstructured data formats. While designed with official statistics in mind, SDMX is flexible enough to support many different domains.

The standard is sponsored by the BIS, the ECB, Eurostat, the IMF, the OECD, the World Bank and the UN, which collectively provide governance, set the strategic direction and fund developments of both the standard itself and some of the software tools that support it. It was also supported by the G20 Data Gaps Initiative (DGI) established after the Great Financial Crisis of 2007-09 to foster the standardised transmission of data as a first step. In addition, the SDMX standard will support the recommendations of the new five-year phase of the DGI, backed by the G20 Leaders in November 2022, which aim at developing access to private and administrative sources of information and data-sharing.

The SDMX standard consists of:

Taken together, these can support enhanced business processes and the harmonisation and standardisation of statistical metadata.

Furthermore, users are supported by the SDMX User Forum, an online community to connect and share knowledge on SDMX, where users may ask questions, find or provide answers and connect with beginners as well as experienced users such as data modellers and system developers

The SDMX User Forum provides support on a variety of topics, like modelling data with SDMX, best practices on using SDMX, SDMX software tools, etc. It is designed as a space to generate ideas, receive feedback and access information on case studies and implementation scenarios.

SDMX at the BIS

In addition to acting as a sponsor, the BIS uses SDMX as the standard for the collection, production and dissemination of its own statistics. Those processes include validating, transforming and mapping data sets based on a metadata-driven environment.