Digital Financial Reporting

XBRL US has announced that the 2009 version of the XBRL taxonomy forUS GAAP is available for public review.  Information about the taxonomy, the review process, and downloads for the taxonomy can be found here.

The finalized version of this taxonomy is slated to be available by February 2009.

I was helping my daughter do a project for school and came across a really good example of what is wrong with business reporting today.  Now, this "report" example is not from a business.  It is from the U.S. Central Intelligence Agency (CIA).  But the same shortcomings of this example exist for business reports.

The project my daughter has to do is a report on a country for a geography class.  The teacher of the class provided this resource to use on the project, the World Factbook.  My daughter picked the country of Bulgaria for her report.  Information about Bulgaria in the CIA's World Factbook can be found by clicking here.

This is a phenomenal set of information!  I won't spend time describing the details of the information, you can go look at it yourself.  Let me just say that it includes both numeric and textual information about a country and data exists for every country in the world.

Now, this information has a taxonomy.  This describes the "fields" of information, providing a description (or in XBRL terms "documentation" for the concept).  Each country has a "report", again here is the report for Bulgaria.  In addition it has comparisons, such as this one for GDP for each country.  All these listings are provided here and called rank order pages.  You can even download "the publication" and use it locally.

And that is sort of the problem.  You can "download the publication" or "view the publication", at least from the perspectives for the specific reports provided by the creators of the information.

But what if you want something else?  Suppose you wanted to compare, say, GDP with the number of cell phone users in a country?  That would mean rekeying information.

What if this information was made available in XML, say even in XBRL?  If the information were available in XBRL one could literally write queries and populate, say, an Excel spreadsheet with data from the World Factbook and do whatever comparison one might desire, rather than being limited by the existing reports due to the effort it would take to rekey the information to get what you desired.

Governments generate vasts amount of information.  It has to be the case that the data which makes its way to the CIA's World Factbook is housed in some sort of database.  That database has some application which generates the HTML files which make up the fact book.

The CIA's wonderful World Factbook is just one publication of the U.S. government.  There are hundreds of agencies and departments which generate other information using taxpayer dollars.  It would be a trivial thing to take this data and generate it as XBRL (or some other XML format).  There would be benefits of the government standardized on XBRL, but I won't go down that path here.

All this is similar to situations which exist in pretty much every business around the world.  Users within a company gets what they are given and they typically have to rekey information to get something different.  Business Intelligence (BI) applications solve this problem to a degree.  BI applications store all this information in multidimensional formats, providing for flexible access to information.

Now if the U.S. government standardized on, say Pentaho because it is open source, as its BI front end which every U.S. taxpayer; then all this data could be made available in the Pentaho GUI.  However, it would likely be the case that IBM, Oracle, Microsoft, and other BI software vendors would prefer that their tool be used.  So, that probably won't happen.

But what if the data was made available using XBRL?  I would speculate that all the BI tools would then rush to support XBRL, allowing for importing XBRL into each BI application.  Taxpayers using the information could pick the BI application they prefer.  Or, user could simply use Excel or some other tool to grab all this data made available.

Some would say that the XBRL data would not be helpful as the XBRL is too hard to use.  Well, it may be harder to use the information if it is ONLY provided in an XBRL format.  But it is impossible to reuse the information using automated processes in existing formats (like the World Factbook's HTML format) and certainly time consuming to rekey data to get the view you desire.

Anyway, I hope the point that I am trying to make comes through.  To summarize it again, current business reporting practices makes it extremely difficult to use information outside the formats provided to business users.  XBRL, particularly when used with the multidimensional model, enables more flexible access to information.

Perhaps with Barack Obama's understanding of technology and interest in using technology, his administration might perhaps make use of XBRL for sharing information such as the CIA's World Factbook or other U.S. Federal Government publications, enabling easier reuse of that information.