We take these ments seriously.However,my observation is that the submissions we receive from filers are improving “with age and experience”.We have noted a gradual rction in the number of custom tags selected as extensions so would note that there is greater standardisation across tags.Yet,our monitoring tells us that some chronic issues remain and filers should expect that we will ntinue to municate about these issues,providing guidance and FAQs on our web am pleased to note that there is definite improvement as filers gain experience with the taxonomy.

Another positive observation i would make is that third party tools are improving.There has been some but still not enough emphasis placed on software ntrols embedded in the third party software tools,thus preventing some of these chronic issues from finding their way to the SEC –catching them at the point of preparation.We ntinue to see and applaud third party service providers investing in their products,but in the long run,full integration of XBRL tagging into the preparation process is the best way to rce st and risk.

INLINE XBRL

We have a healthy interest in making the submission of XBRL data easier and are nsidering inline XBRL.Inline XBRL eliminates the need to create a separate XBRL attachment and allows the filer to embed the XBRL tags in a single document.We believe that the use of iXBRL creates a good opportunity to improve both efficiency and quality by allowing the submission of a single financial statement document.We understand the value of improving the experience of the filer and nsumer and see inline XBRL as a possible tool in this discussion with potential benefits to all involved.

nsidering the current environment,i would note that our experience tells us that greater efficiency and improved data quality are realised gradually over filing periods and external survey data supports this ment.For first time XBRL filers,particularly those submitting footnotes using detailed tags,the learning curve is steep.Filers have told us emphatically that the initial foray into detailed tagging has been the most difficult phase and but that they are able to significantly rce the number of resources devoted to creating the XBRL in subsequent filings.

We also begin to see best practices emerging in the filing/preparer munity and more and more XBRL experts who are well versed in what it takes to perform the work.We have also heard the ncerns raised in the press about the quality and reliability of the XBRL data.We believe all stakeholders would agree that nsumers of the data legitimately demand aess to high quality and reliable data.ncerns about quality must be addressed if the data is to be used.At the SEC,we will ntinue to speak about our observations and listen to input about improvements.

NSUMPTION

From an investor and analyst standpoint we hear anecdotally that there is great interest in our XBRL data.ncerns about quality remain,but the detail and broad pany verage,particularly of smaller panies is useful.We know that analysts and data aggregators get their data from a variety of sources and that structured data ntent is often richer,more reliable and less time nsuming to process than anizing less structured data.

ntinuing the discussion about using the XBRL data,i will mention a key initiative we have undertaken at the SEC in my Division,the Division of enomic and risk Analysis.Our Division Director,Dr.Craig lewis has developed a statistical model we call the Acunting Quality model which is being tested.The model identifies anomalies in financial statement filings by focusing on discretionary aruals and predicts which panies may have engaged in potential earnings manipulation.The Acunting Quality model will use XBRL data in its calculations.We believe the XBRL database has a lot to offer AQM,due to its rich detail and pany verage.SEC examiners uld use AQM to direct their attention to areas in the financial statement that deserve further attention,perhaps to improve disclosures or to investigate questionable acunting practices.AQM is an excellent example of how the SEC finds value in XBRL data.

OTHER REGUlATORS

We talk periodically to other regulators around the globe and share experiences.We are very excited about the us ernment’s open data initiative which promotes transparency for data that the ernment llects and makes publicly available.the initiative calls for llected data to be structured,aessible and reusable.There is a tremendous amount of data available from many,many sources.Many of the anisations we oversee are international and llect and disseminate data in a variety of structures and anisations that effectively anise and integrate the data they must llect and nsume will be most suessful in using the data to their advantage.We believe that XBRL data can play an important role in data integration.

My background is in operations for the financial services industry.We were responsible for preparing information for others to nsume.When i began my career,my pany had a great deal of ntrol over what information was made public.Now all of that has changed –there are a many opportunities for those interested in gathering information about a pany to gain aess to information from a variety of sources.At the SEC,we want to stay nnected to the changes in data,aess and technology around the world.We also want to use data in fulfilling our mission.Timely,aurate and easily aessible XBRL data fits very nicely within this strategy.

Technology changes nstantly and i would not prescribe one solution to satisfy all needs –XBRL has many advantages and strengths in the way it standardises financial data elements and allows machines to read,pare and ntrast data.on principal,auracy,timeliness and standardisation are key elements of a usable data base and there are a number of ways to solve for this.We are trying to ensure that the best and most easily nsumed data can be used by the people who want it,whether it is an institutional or individual investor.

We expect to be broad minded in how we view the future and future solutions.However,i would suggest we address the issues just mentioned since for all its strengths;there is plenty of room for improvement in the XBRL experience.Better software tools are beingdeveloped and this momentum must ntinue.iXBrl uld simplify the preparation.Input from industry groups and other stakeholders including analysts and software providers is improving the taxonomy but perhaps more can be done to simplify.I wouldn’t say that XBRL is always the right answer;nor would i say that a regulatory mandate is always a necessary ingredient for solving data transparency issues.Industry and the munity can go a long way toward solving the issues.Gathering,structuring,anising and disseminating data is an area where we can work together with all stakeholders to find a solution.

A final area of interest is the academic area.We have a number of academics in our midst in the Division of enomic and risk Analysis and they are interested in XBRL data.From an academic and an analyst standpoint,the lack of a time series is something that is a bit of a hindrance.I expect as time goes on and we add years of data to the database that will change and there will soon be a healthy interest in the XBRL data on the part of the academics for research.This has benefits for everyone,as the more eyes that are reviewing and menting on the data the better the quality.

原文链接：