There has been quite a bit of discussion about the idea that XBRL filings should be comparable, and if they are not, that somehow is a surrogate indicator of lower quality XBRL. Yet this flies in the face of one of the key promises of XBRL - "Tell your story, your way".
Companies are different, and after years of attempting to create one-size fits all reporting. The SEC tried, and IFRS continues to think they have a one-size fits most (except SMEs). It remains clear that it is what is different about companies that enables them to be successful.
Three Motivations to Create Extensions
For years the example used was that of a major computer manufacturer and service provider, which included a negative expense line for 'IP expense'. The expense was negative because the company was bringing in over a $1 billion in revenue from patent licenses. The problem was that the data aggregators consistently aggregated (well, it is in the name) would combine all of their expenses into one 'other expenses' line, thus distorting the company's position, and message. They were (are) proud of their portfolio of patents, and reflect that in their business reporting.
We also see the example of the giant Zombie banks. Looking at their XBRL we see extension rates well above 50%. They are telling their story, their way - by intentionally making it difficult for simple Zombie to Zombie comparisons to be run. Difficult in fact, for anyone to perform automated analysis, including regulators.
There are also companies that have limited resources to spend on their external reporting, and XBRL has added to their burden. Sometimes creating a new extension is simply faster and easier than digging through 16,000+ elements, reading detailed definitions, and wondering why their exact concept is missing. Equally, as the US GAAP taxonomy evolves year on year, how many companies are reviewing their extensions, confirming that an extension created in a prior year is still required.
Three examples, three motivations, one outcome: more extensions.
1. Transparency if wonderful. We are different, and we want the investing community to know that we are different. We have unique line items and footnoted facts because we want to demonstrate why we are the better investment.
2. We'll happily pay for opacity. We are different, and exploitation of our differences enables us to be successful. Enabling easy comparisons between us and our 'peers' actually will reduce our ability to exploit our unique advantages - whatever they are. Transparency helps regulators and competitors, not us.
3. We are too busy and with no benefit from investing limited resources in XBRL, we'll get this done a quickly and cheaply as possible. If we can produce XBRL that passes the SEC's validation checks, then that is good enough for us.
One example uses XBRL to improve the quality of available information and increase transparency. The other harnesses the power of XBRL to protect their opacity. "Our 'black box' is what keeps us profitable, reduces competitors ability to match us, and keeps the regulators in the dark (without appearing to want to keep regulators in the dark)". The third simply does not have the resources to waste on XBRL, there's real work that needs doing.
They are not going away
There is simply too large a need for extensions, and too many different motivations. There are also over a million extensions already created. These are not going away. Some, possibly most, are either duplicates or are so similar as the make if difficult to differentiate. Yet these are not going away. If anything, expect the total number of extensions to continue to rise.
After all, even if the SEC, the FASB, the IASB, or any group, attempts to analyse extensions to identify a reduced set of new taxonomy elements, the three motivations outlined above will act as a 'headwind' to companies migrating off their extensions.
So while we should see fewer 'errors', we will not see significant drops in extensions that are there specifically to influence comparability or reduce the 'auto consumption' of financial and business information. Controlling the message is what business reporting is all about, not providing transparent reporting.
Only once they have driven down the number of errors will the SEC have the energy or resources to drive down the number of extensions - and for each one, the SEC will need to demonstrate that the filer did not, in the filers' view, have an adequate justification for the extensions created and used.