Another analysis I did was to check how the companies which make up the Dow Industrials report their organization information and basis of reporting. You can see the results of this analysis here.
To summarize, this is what I see (note that of the 30, there were 28 10-Q filings and 2 10-K filings):
- 14 companies, about half of the Dow, use the concept us-gaap:OrganizationConsolidationAndPresentationOfFinancialStatements DisclosureTextBlock
- 9 companies use the concept us-gaap:SignificantAccountingPoliciesTextBlock
- 2 companies used the concept us-gaap:BasisOfPresentationAndSignificantAccountingPoliciesTextBlock
- 1 company used the concept us-gaap:NatureOfOperations
- 1 company used the concept us-gaap:QuarterlyFinancialInformationTextBlock
- And finally 3 companies chose to add an extension concept which they created
An interesting side note is that it seems that all 3 companies who chose to create an extension concept used the same filing agent.
The information shown on the summay page of the analysis is for only one fact. If you look at the information within the fact for the concept reported, it tends to contain information about the organization, basis of reporting, nature of operations, and significant policies all packed into that one concept. I did not look at every financial report closely, you can examine that if you are interested.
This packing of information into one big box brings up a point which will likely lead to discussion. In the past when reporting information in say an HTML or document format, you could combine the information as one saw fit. With XBRL you basically need to put the individual pieces of information into some "box", people call these "tags", a term I try and avoid using because the information is not "tagged", it is reported for a fact using a specific concept. I will leave that point alone for now.
But, so here you have specific concepts in the US GAAP taxonomy. The "flow" of the financial report which a company had created in the past does not necessarily follow the "flow" offered by the US GAAP taxonomy. What I mean is that there tends to be exactly four cases:
- One-to-one correlation. There is a one-to-one correlation between what a company reports and what the US GAAP taxonomy provides.
- Company groups, US GAAP taxonomy decomposes. The company groups things together, but the US GAAP taxonomy decomposes that same set of information into individual pieces. For example, the company has one specific set of information which the US GAAP taxonomy decomposes into three specific pieces.
- US GAAP taxonomy groups, company decomposes. The US GAAP taxonomy groups something, but the company chooses to decompose that one group into individual pieces. For example, the US GAAP taxonomy has one set which the company chooses to decompose into three individual pieces.
- Mismatched sets. The final option seems to be where the US GAAP taxonomy and the company have what amounts to mismatched sets. In this case the pieces simply don't really correlate nicely.
Lots of discussion will likely occur over the coming years as those used to having full control over information presented try and put that information into what amounts to little boxes. It will likely be the case that in 80 percent of the cases there will be no issues really. But in 20 percent of the cases, there might be some issues.
What is occurring is similar to trying to convert an analog record into a digital CD. Some information is lost in the process, but that information lost is deemed no relevant. I had a friend to swears he can hear a significant difference between a record and a CD. Me, I cannot really tell the difference.
I anticipate that there will be many, many discussions during the coming years as these sorts of questions are resolved.