In the wake of the financial crisis in 2008, the US Government mandated machine-readable financial reports via XBRL.
That was a wonderfully progressive move that could dramatically change the efficiency and reliability of the capital markets by bringing financial reporting into the 21st century. (If you are new to XBRL, this post and it’s links will serve as a good starting point).
Now that great initiative is in danger of being overturned. This would be a terribly regressive step. An efficient, fair capital market is key to a prosperous free enterprise society; so this matters to us all.
Which is why we need to #SaveXBRL.
This post has the story and this is the bit that matters:
“Smaller Firms Want Out
An even larger debate surrounding XBRL is taking place in Congress. Some corporations have protested to lawmakers that using the XBRL tagging system is too costly and that it puts undue burdens on smaller firms.The legislation that addressed the issue, the Small Company Disclosure Simplification Act, was first introduced by representatives Robert Hurt (R-VA) and Terri Sewell (D-AL), and passed the House Financial Services Committee on March 14 by a vote of 51 to 5. The proposal is now part of a bill called H.R. 5405, which passed the House on September 16, requiring that the SEC exempt public companies with annual revenues under $250 million from filing their financial statements in the XBRL format for up to five years. The bill awaits a vote in the Senate.”
To understand why this is the wrong way to solve a legitimate problem, travel with a financial data item through the financial reporting process:
- Step # 1. Start as an electronic bit in an accounting/ERP system. The data is now perfectly machine-readable and gets aggregated and processed in the most efficient way.
- Step # 2. The data is converted into into human-readable form for the SEC. This post describes this very well:
“I will never forget talking to the CFO of one of three largest corporations in world, and he told me that the only time their numbers are on actual paper is when they send their reports to the SEC. That’s because in the corporate world, everything is electronic and digital,”
- Step # 3. Somebody extracts the data from a PDF or HTML file and turns it back into a machine-readable bit in XBRL format. That “somebody” is probably working for an outsourcing firm that is being paid by the company doing the reporting, because they have to comply with that SEC mandate.
You can see why this song is playing to packed houses:
“release those poor small companies – the life blood of our economy – from yet another regulatory burden”.
In other words, stop demanding Step # 3. Kill off the XBRL Mandate. Return us all safely to the 20th Century. Forget that the Internet happened.
The solution is not to eliminate Step # 3. The solution is to eliminate Step # 2.
Imagine the poor overloaded folks at the SEC surrounded by piles of paper. They are dedicated, smart and hard working. They will therefore have evolved a system that sort of works – poring over individual company filings and marking something odd about a data item in a footnote with a yellow pen and then digging though a pile of documents to look on page 256 of another report (having cleverly marked the page) to correlate something odd on that other company’s filing…
Imagine if all the data was in XBRL electronic format and they could let an algorithm do the grunt work, so that they could do the higher-level work needed to catch the bad guys and maybe avoid a repeat of the financial system’s “cardiac arrest moment” in September 2008.
The algos could process thousands of companies to look for that anomaly, that weird thing that says, “something looks fishy”. The data surfaced by the algos still requires the higher-level cognitive and pattern matching skills of humans. This is about empowering the SEC staffers to be more efficient. I imagine that they would vote for this change.
The work done by SEC staffers is impossible without better systems. The devil is in the details, or to put that in financial reporting language:
“the devil is in the footnotes”
The footnotes are where a company buries that embarrassing fact that they want investors and regulators to gloss over. Scam artists use that footnote technique; that is an edge case. More normally it is used by companies to simply “accentuate the positive” while abiding by the letter of the reporting law – a kind of obfuscation through obscurity.
We want the SEC to be efficient and to catch the bad guys. Imagine a tech-empowered SEC staffer being able to check a data point across thousands of reporting firms through a single algorithm that can be continuously improved.
Saving XBRL is not just about the efficiency of the SEC and catching bad guys. The bigger issue is about making the capital markets an efficient place for small companies to raise money and for retail investors to make money. This is where the story about saving small companies from the regulatory burden is so misleading.
If you eliminated Step # 2, there is no burden, even for small companies. It is straight through processing. Eliminating Step # 2 might entail a tweak to the SEC Mandate to mandate not just human-readable conversion of XBRL to HTML but the transmission of the raw XBRL data feed. Once that mandate was in place, CFOs would call their Accounting/ERP software reps and demand XBRL built into their core systems.
So much for the burden of XBRL. Why do small public companies need XBRL?
Small public companies need XBRL for discoverability.
It is the same reason that anybody running an online site/blog wants to be discovered by Google. You want your site to be discovered.
The CEO/CFO of a small publicly listed company wants to be discovered by investors. Obscurity is their biggest enemy.
These emerging growth companies (less than $250m in annual revenues) are too small to motivate analysts to pore over their individual filings. However if all their data is machine-readable, their company can be discovered by algos looking for bargains. Those algos could be used (and written) by individual retail investors who would get a chance to discover bargains (and shortable overpriced junk) that the big firms are not tracking. Small companies raising money efficiently and individual investors making money in the stockmarket sounds like a worthwhile objective.
I was motivated to get interested in the arcane subject of XBRL thanks to a superb article by Daniel Roth in Wired from the dark days of February 2009 called:
Overturning the XBRL Mandate killed would be a huge blow to small public companies and retail investors. The reporting burden can be fixed technically. I think saving XBRL is at least as important as Net Neutrality and when the digital people arise, the politicians and regulators are forced to listen. That is why I think it is both important and feasible to #SaveXBRL