The SEC has formalized its ongoing commitment to XBRL quality. The EDGAR filing system now supports data-quality checks based on the rules for XBRL tagging that were developed by the XBRL US Data Quality Committee. These rules are included in the EDGAR Filer Manual (Volume II), as implemented in March 2021. The Financial Accounting Standards Board (FASB) also made them part of the US GAAP Taxonomy Release 2021.

Joan Berg, chair of the XBRL US Data Quality Committee and Director of XBRL Services at Toppan Merrill, affirms the significance of the new requirement. “It is a major acknowledgement by the SEC of the critical importance of XBRL data quality and the role of the DQC in improving the overall quality of XBRL submissions,” she told Dimensions. In addition, she explained, all registrants are now on notice that these rules, developed by a group invested in the goal of achieving usable XBRL data, “should be taken seriously.”

For insights into this requirement, including some background on the Data Quality Committee, the prevention of XBRL errors, the role of the FASB, and the likely future developments, we turned to three experts:

• Campbell Pryde, CEO of XBRL US
• J. Louis Matherne, Chief of Taxonomy Development at FASB
• David Shaw, Senior Project Manager—XBRL at FASB

XBRL US: Campbell Pryde
What is the XBRL US Data Quality Committee?
The XBRL US Data Quality Committee (DQC) is a group of individuals representing software providers, data aggregators, institutional investors, the accounting profession, and academia, who are tasked with developing guidance and validation rules to prevent or detect inconsistencies or errors in XBRL data submitted to the SEC.

Why was the DQC created?

The DQC was formed in 2015 with a goal to help corporate issuers improve the quality and usability of their XBRL financial data filed with the SEC. We recognized that more guidance was needed for issuers to help them navigate through the US GAAP Taxonomy and consistently tag their financials.

How many rules have been developed by the DQC?
To date, we have published 14 rulesets, with over 100 individual rules, each of which covers multiple concepts in the US GAAP Taxonomy. The negative-value-input rule, for example, applies to hundreds of concepts to help issuers identify potential problems prior to filing.

Which rules detect the most errors?
The most common error we have been seeing of late is when an issuer incorrectly uses a member on an axis, followed by missing calculations on face financials. A common example of the former is when an issuer uses the Legal Entity Axis for facts that apply only to the Consolidated Entity. The third biggest category of errors continues to be negative-value errors, but we have seen a significant decline in error count since our rules began tracking that issue in filings.
What is the process for identifying new rules? Who participates in evaluating the rules and their impact?
Rules are developed through a rigorous seven-step process, where we gather input about recognized errors from filing agents, data providers, and data consumers, then consider the potential impact of the establishment of a rule. We write up clear, unambiguous error messages, test the rules against historical filings, then publish them for a public review period to get further input. Once the public exposure period closes, we incorporate market feedback and publish the freely available rules for all to use.

How does a company catch up on learning more about the DQC rules and their use to improve the quality of their XBRL in SEC filings?

Most issuers access and use the rules through their own EDGAR/XBRL tool or service provider. We certify vendors to make sure they can successfully run the rules in their applications. Issuers have the option of running the rules against their filing on the XBRL US website as well.

Which DQC rules are incorporated as checks into the revised EDGAR Filer Manual (Volume II)? Will more become part of it in time?

The 2021 release now includes six DQC rules covering situations including incorrect negative-value input, reversed calculation, required calculation parent in the cash-flow statement, element values not in alignment, and an axis that contains inappropriate members. These have been incorporated into the US GAAP Taxonomy and are referenced in the Filer Manual. The FASB will be adding more rules over time with upcoming releases.

Why did the SEC decide to rely on the DQC rules and not just create its own rules or merely reference the DQC rules as a resource? What’s the significance of this decision by the SEC?

The XBRL US DQC rule development and vetting process involves participation by many stakeholders. It is fairly fluid because we need to accept and react quickly to changes in the US GAAP Taxonomy, in how investors are accessing it and in how preparers are interpreting how to work with it. These factors make it more easily handled by industry because we can more quickly adapt to changes and reflect input in the
rules that may change over time. That said, the SEC and FASB offer a lot of guidance through the EDGAR Filer Manual, reports and guidance, and tools that they provide to issuers.

Is this an acknowledgment by the SEC of the importance of data quality and accurate XBRL tagging?
The inclusion of the DQC rules in the US GAAP Taxonomy, along with the reference to the XBRL US DQC in the EDGAR Filer Manual, is simply further confirmation that data quality is of critical importance to the Commission and that every filer should be using these freely available rules. And it clearly points out that the XBRL-formatted financials are just as important as the paper-based version of the company’s financials.

Does this indicate the potential for broadened XBRL-tagging requirements in more SEC filings and other parts of them, or potential enforcement for repeated errors?

The SEC has made a number of rule proposals that point to greater reliance on XBRL for reported disclosures. For example, variable-annuity and life-insurance companies and business-development companies will soon be required to submit their disclosures in XBRL format. The acceptance of the DQC rules is simply further evidence that the Commission recognizes how XBRL can be used to further improve the quality of information. This kind of automated validation simply would not be possible if the data were not in structured, standardized format.

Should an SEC filer first check its filings against the rules on the XBRL US website before trying to submit it with the SEC?

Absolutely. Every filer should use the rules to check its XBRL financials, but filers should first turn to their filing agent or application provider, because many of these have the rules built in for ease of use. If the filer is
working with an application that does not, then yes, use the rules at the XBRL US website.

Does Inline XBRL allow more validation and error-checking than standard XBRL?

Inline XBRL is extremely helpful for manual reconciliation and review, as it allows viewers to see how a fact was tagged if they look at it in a viewer, such as the SEC viewer. This can be very helpful for organizations where multiple departments—e.g., legal, accounting, investor relations—are involved in reviewing and approving a document before it is finalized. But automated validation and error-checking is performed the same way on an XBRL document, whether it is reported in Inline XBRL, XML, JSON, or even CSV.

How will the EDGAR system inform filers of defects in XBRL quality?

The SEC recently announced that the EDGAR system will inform filers of certain quality issues using warning
messages driven by errors identified through the rules of the XBRL US Data Quality Committee.

Do you have data on the most common XBRL errors? Will the SEC also collect data on this that it will share with XBRL US?

We maintain an archive of errors generated so that we can track how use of the rules has affected the market, and so we can determine if we need to revise any rules already available. They can be found on our website through charts showing aggregated real-time filing errors.
How has the use of DQC rules impacted the perception and use of XBRL by data users?

The charts of aggregated real-time filing errors show a definite decline in the number of errors in public company filings, which provides further evidence that most corporate issuers are using the rules.

Do you envision that the DQC rules related to IFRS filings will be incorporated in future SEC warnings?

Both the FASB and the IASB have observers who sit in on our DQC meetings and work with us closely as we develop and refine the rules. We will continue to develop rules for IFRS filers just as we do for US GAAP filers. In Europe, a similar process is being discussed as the European Securities and Markets Authority recognizes that DQC-type rules will be helpful for their implementation as well.

FASB: J. Louis Matherne and David Shaw
Which DQC rules are incorporated as checks into the revised system? Will more become part of it in time?
The SEC supports new data-quality-enhancing checks included by the FASB in the DQC Rules Taxonomy (DQCRT). The DQCRT was developed by the FASB through public review in collaboration with market participants and the XBRL US Data Quality Committee. The FASB expects to add more DQC rules to the DQCRT in future updates.

How are these rules selected for inclusion in the FASB DQCRT?

The FASB acceptance process for including DQC rules in the DQCRT involves reviewing the DQC process for designing, exposing, testing, and addressing feedback for the rule; identifying the risks of including the rule in the DQCRT; and independently testing the rule. Our baseline expectation is that the rule has been available for use after formal approval by the DQC for at least one annual cycle to confirm that it is working as expected. Once the FASB determines these criteria have been met—and in consultation with the SEC, XBRL US staff, and the DQC—the rule becomes a candidate for inclusion in the next DQCRT release, at which time the SEC will consider it for inclusion in EDGAR.

Why is it useful for the SEC to rely on the DQC rules and not just create its own rules, or merely reference them as a resource?

While the SEC has a number of validation checks that are aimed at improving data quality, such as including one on negative values, the DQC rules are the result of a market effort to address data quality by market participants with first-hand experience in dealing with data-quality issues and developing solutions that can be implemented in XBRL-tagging software. This is the SEC being responsive to and engaging in an effective manner with the community it regulates. Additionally, including these market-developed rules in EDGAR gives them more prominence with all market participants, increasing the likelihood of data-quality improvements across all registrants.

What is the significance of that decision by the SEC?

Even though the rules are warnings from the EDGAR system, they are being made visible by the regulatory authority and are not dependent on the tools being used.

How does the FASB get involved in the SEC decision related to the EDGAR Filing Manual?

In general, the FASB, in consultation with its Taxonomy Advisory Group (TAG) and other constituents, provides input to the SEC on ways to improve the EFM rules, with corresponding positive effects on data quality based on market observations and taxonomy-design requirements. The DQCRT is simply one highly visible example of this involvement.

Why is the FASB part of this process? Why does the SEC not take the rules directly from XBRL US?

The FASB is a part of the DQC process by participating in the DQC’s rules group and proposing rules to include based on taxonomy modeling, and as an active observer to the DQC meetings. Including the DQCRT with the GAAP Taxonomy facilitates this process, as it assists in ensuring the rules properly validate against the GAAP Taxonomy.

Why did the FASB in the US GAAP 2021 taxonomy include new data-quality-enhancing checks?

The DQCRT was added to the FASB Taxonomies with the specific goal of enhancing data quality by increasing exposure to the XBRL US DQC rules. The first version of the DQCRT in 2020 included three DQC rules.

The 2021 DQCRT includes three more. Future DQCRT updates will include additional rules. Our approach is to methodically build out the DQCRT in a measured way to make sure it works as intended and give us
the ability to recalibrate, if necessary.

Which DQC rules are you considering for inclusion in the next taxonomy? How will you select those rules?

We are currently soliciting input for which rules to add in the 2022 DQCRT. We are looking to add another two to three rules that meet our criteria for inclusion in the DQCRT. Our objective is to add rules that have higher lists of exceptions so that we achieve broader coverage in addressing more pervasive data-quality issues.

Note: The views expressed here are those of the respondents and not necessarily those of their respective organizations.

To read the full article in Dimensions Vol. 2021, No. 2, click here.

Reach out to jump start a partnership that will bring speed, security, accuracy and efficiency to all of your complex content and communication requirements.