Whereas a taxonomy defines XBRL concepts and their relationship to other concepts, the instance document is a report containing the actual data. There is a tight relationship between taxonomies and instance documents. After a taxonomy is created, you can use its definitions and their relationships to produce an XBRL report. In addition to taxonomy references, instance documents also contain the following information:
XBRL Context—Provides information about the reporting (business) entity, a time-frame, and other optional details such as scenarios and dimensions.
XBRL Unit—Describes what the numeric data represents. Examples of units are: “US Dollars,” “Euros,” and “shares.”
Data—Instance documents contain numeric and/or textual data that reside within a Microsoft Office document, Financial Reporting grid, and an Oracle Hyperion data source. The generic “document data” term can mean one cell in Excel, one word or entire paragraph in Microsoft Word or a cell in a Financial Reporting grid. This term is used throughout to mean data that can be mapped by the Disclosure Management Mapping Tool. Additionally, numeric data can be scaled and have references to footnotes.
The instance document is similar to an HTML Web page, but instead of the report language being HTML that can be read by a browser, the language is XML read by a variety of XBRL applications that consume and analyze instance documents.
The XBRL filing consists of the XBRL taxonomy and the instance document. The XBRL taxonomy explains the metadata behind a company's disclosure, and the instance document shows how facts are mapped to the taxonomy. Validation verifies semantic relationships between concepts, confirming that the correct facts have been mapped to the correct fact field in the base taxonomy. For example, validation verifies that the facts filed for “Assets” equals the facts filed for “Liabilities” and the “Owner's equities”. XBRL instance document generation is the last step of generating the XBRL-compliant disclosures. To ensure the accuracy of the XBRL data that is submitted in a filing, Disclosure Management validates your taxonomy against XBRL taxonomy specifications before creating the instance document.
Validation is a three step process. First you validate the taxonomy. The next step is the generation of the instance document, which creates an XML file associated with the instance document. XBRL is an XML-based framework and relies on XML syntax to declare semantic meaning such as XLink and XML Schema. The last step is the creation of the instance document, which can be exchanged with other business entities or filed with a regulatory agency.