Lots of Data? Need Quality and Comparability? xBRL-CSV is here
On 2nd May, the XBRL Standards Board approved a Public Working Draft of xBRL-CSV, a new standard for representing XBRL data using CSV. At first glance, CSV might seem like a legacy technology, but it’s got a lot of things going for it. Firstly, it’s simple. In fact, it so simple that it’s supported just about everywhere. Secondly, it’s very efficient. For large data sets where size of representation is an important consideration, it’s hard to beat CSV. Until now, the downside of CSV has always been the lack of semantics, which inhibits interoperability.
We’re not the only people thinking this way. The W3C commissioned the CSV-on-the-Web Working Group to develop standard ways of associating meta-data with CSV files. This has resulted in the Tabular Metadata specification, and XBRL International’s Open Information Model (OIM) Working Group has now built upon this to produce xBRL-CSV.
To help users understand how this new format works and what it can be used for, the OIM Working Group has now produced an introductory presentation and accompanying samples. These resources, as well as the technical specification itself, can be found on the XBRL specifications site.
We see xBRL-CSV as an important new initiative, which broadens XBRL utility in the field of bulk data collection and publication, and also a mechanism which can make it easier to integrate XBRL into legacy systems.
xBRL-CSV is part of the Open Information Model, which is an incremental approach to simplifying and modernising XBRL. In practical terms, the xBRL-CSV specification allows the collection of very large quantities of data, for example, highly granular information that has to be provided to a regulator, while still being able to use XBRL’s strong validation and data definition capabilities. It can equally be used by regulators to publish information in a format that is easily consumed.