While revisiting the U.S. House of Representatives Committee on Oversight and Government Reform report entitled The S.E.C. Designed for Failure (from May 2010), I was reminded of when I sold my first accounting software back in 1982 – almost 30 years ago. Apart from the FUD surrounding what was then considered new-fangled ‘electronic’ accounting, my main competition in those days was document-centric ledger systems.
Perhaps you remember these relics of Dickensian times (e.g. Kalamazoo’s One Write in the USA and the Simplex VAT book in UK)? These paper ledgers would have been familiar even to Pacioli and one can almost imagine Bob Cratchit copperplating entries into such a system in Scrooge’s office. But God help him if Scrooge asked for an aged debtors/receivables list. Bob’s hands would be blistered from the effort of searching through all those ledgers to compile such a basic report. He’d certainly miss his Christmas dinner.
Today, the advantages of storing accounting transactions as data in a database has all but made document ledgers obsolete. Yet the document mentality persists and continues to hold back progress when it comes to the regulatory oversight of financial markets and other risk-intensive domains.
As Darrell Issa (CA-49) points out in the report, we are still far from realizing the improved transparency and other benefits of a truly data-centric model while this document-centric mentality persists. It’s as if every corporate accounting department was expected to maintain the old paper ledgers alongside their new ERP databases. Just to be sure.
The 10-Q/K submission document, whether in paper or PDF or even HTML format, was only ever intended to be a presentation of data: static rather than interactive in purpose. Yet the reality is that most people who care about this data don’t actually read 10-Q/K documents as static narratives but treat each document as a discrete data repository to interact with – in order to search and extract data from either to find specific facts about a company or to collate data from multiple company documents for further analysis.
Today, in the world of XBRL filing, the role of the traditional 10-Q/K is merely to provide a familiar framework to tag. Otherwise it’s redundant. If you don’t believe me consider the process: We take data from our internal systems, compile it into a document in order to tag it, so the S.E.C. can convert the tagged content back into data again. To me, step 2 of this process looks redundant.
By initiating the move to a data-centric model based on XBRL, the S.E.C. has made great progress towards the improved transparency and operational efficiencies that Issa is looking for. And it has made the crucial initial step to transition from ‘designed for failure’ to ‘destined for success’.
But the document tail is still wagging the data dog.
What’s needed to break the pernicious document culture is a new kind of submission standard for filing regulatory data. One that does not rely on legacy document formats that perpetuate outdated document-centric practices but recognizes that in a data-centric world, documents are assembled from data and not the other way around.
