XBRL 2.0

Digital financial reporting that actually works

THE XBRL FILES: QUALITY SCORE

To document issues related to financial data filed using XBRL, AsReported.com has developed a metric - QUALITY SCORE - that measures the usability of XBRL filings from the standpoint of serious investors, analysts and researchers who demand a high level of standardization. In the process of building AsReported.com’s standardization engine (which corrects all of the problems discussed here), there were 7 categories of ‘errors’ that made full standardization of the primary statements difficult. I use the term ‘error’ loosely because the lack of definitive filing rules in certain of these categories allows companies wide discretion to make choices that are often inconsistent with the goal of comparability with other companies. 

Here is the scoring summary by filing software for 10K and 10Q filings made in 2017. (***note that this is a revised version of the original table, adding almost 2,000 scored filings and correcting an issue related to element relationship disclosures)


Here are the components of QUALITY SCORE:

1. TAGGING ERRORS - TAXONOMY. The US GAAP taxonomy consists of 5 separate sector-based taxonomies: commercial, deposit-based, securities-based, insurance-based, real estate and REIT. Each has its own primary statement structures and elements. Some overlap, but many are specific to the taxonomy. Often companies in one sector will use elements from another sector in its filing, rather than an equivalent element in its own taxonomy. This requires creation of a link to an element from the correct taxonomy.

      2. TAGGING ERRORS - SECTION. Breaking down the primary statements into sections and blocks is very useful in standardization. Examples of sections are current assets, non-current assets, current liabilities, etc. When a company item is tagged to an element from the wrong section, it requires mapping back to the correct section.

      3. TAGGING ERRORS – PERIOD TYPE. Elements are characterized as either ‘static’ or ‘duration’. When an element with the wrong period type is used, it requires conversion.

      4. INVALID TOTALS.  An important feature of XBRL is the requirement that relationships between elements be specified. This allows statements to be easily audited and quickly reveals discrepancies. And yet, filings are made with incorrect totals, or incorrect relationships, or missing elements, or all of the above. These errors could and should be revealed by the filing software, or by SEC enforcement testing.

      5. MISSING RELATIONSHIPS. One reason for invalid totals is that element relationships (called calculation arcs) are often simply not specified.

      6. MISSING ROLLUPS. When an element is broken down into segments by product or geography or another company-specific breakdown, it’s done with dimensions. Long-term debt might be shown by issuances in this way. Dimensions provide great flexibility in financial reporting, but it’s good practice to roll-up the components into a total value for the element, ensuring comparability.

      7. SIGNAGE ERRORS. If a specified element relationship is inconsistent with the relationship expressed in the US GAAP taxonomy, then its sign will be inconsistent and require changing. It’s not possible to compare companies using dissimilar signage in their statements.

Exceptions under these 7 tests are totaled to Total Quality Issues and divided by the number of primary line items to arrive at QUALITY SCORE. These issues are within the control of the filer and the filing software.

Example Quality Score detail:


Note that an additional measure is shown - USABILITY SCORE - that takes into account the number of extensions, or custom tags, which inhibit full standardization (and hence comparability) of reported financial results.