Fact or Fiction? Let’s Set the Record StraightPosted Wednesday, June 27
Al Berkeley, Chairman, Princeton Capital Management
On May 23, 2018, the House Financial Services Committee passed H.R. 5054, in a 32-23 vote. This legislation, which aims to reduce the burden on small filers, is an exemption from XBRL filing for companies with revenue less than $250 million. While the intent is good, the repercussions of the passage of an exemption like this, are anything but.
The rationale for the exemption is based on two misperceptions: first, that the cost of XBRL preparation is so high that small companies are shying away from going public because of it; and second, that investors aren’t using data in XBRL format anyway.
But a lot has changed since an early version of this bill was passed back in 2013. Arguments made to support this bill contain inaccuracies, often based on outdated studies and data.
We want to set the record straight. Read on.
XBRL preparation for small companies costs $50,000 per year.
In 2014, the AICPA and XBRL US conducted a survey that found the average annual cost of XBRL preparation for small filers was $10,000; the median was $8,000. In 2018, this survey was conducted again, and results show that the price of XBRL preparation has actually declined 45% to average less than $5,500 per year; with a median of $2,500.
Regulators don’t use XBRL data.
The Securities and Exchange Commission (SEC) does use XBRL-formatted data for their own analysis. While we have heard this anecdotally for some time, a recent rule proposal for Auditor Independence With Respect to Certain Loans or Debtor-Creditor Relationships, states in a footnote, “This estimate is based on staff analysis of XBRL data submitted with EDGAR filings of Forms 10-K, 20-F and 40-F and amendments filed during the calendar year of January 1, 2017 to December 31, 2017.” This is just one example demonstrating that XBRL data is used by regulators in analysis required to prepare rule proposals.
The SEC also recently had XBRL data posted to Google BigQuery, which now carries full SEC public company datasets.
And most importantly, the SEC’s 2018-2022 Strategic Plan states that the Commission will:
“Enhance our analytics of market and industry data to prevent, detect, and prosecute improper behavior. Data analytics are essential to rooting out wrongdoing in our markets. The SEC will continue to invest in the data and tools needed for our enforcement and examination programs to uncover and prosecute violations of the federal securities laws.”
Only 10% of investors use XBRL-formatted data. Other investors rely on data from traditional HTML filings.
Investors typically obtain most or all of their financial data from third party commercial providers such as Thomson Reuters, S&P Market Intelligence, CalcBench, idaciti, and Intrinio, among other large and small data providers. All of these organizations use XBRL data in their database offerings to investors. Securities analysts and portfolio managers typically know what data provider they use but they are unlikely to know that the provider is extracting financials through the company’s XBRL submission.
Why do data providers use the XBRL submission rather than the HTML? Because it’s substantially easier, more timely, and less expensive, to process.
The CFA Institute, which represents investment professionals (over 120,000 members in 140 countries), published a blog containing data from a study that found that 90% of investor respondents get some or all of their data from third-party providers. Only a small percentage of investors manually extract all the data they use from the company’s source documents.
Investors don’t want XBRL data.
Companies are required to post their XBRL-formatted data on their web site; and most would tell you that no one is downloading that data. That’s because investors don’t get their data off individual corporate web sites. They get it, as noted above, primarily from third party data providers.
What investors want is the benefit of machine-readable data. Companies that report without machine-readable data, will simply be ignored in the investors screening process.
And studies show that investors want more, not less, automated data. In a 2016 study on Data and Technology: Transforming the Information Landscape, the CFA Institute noted that financial information contained in the earnings release and the MD&A would also be highly valuable in structured format:
“… data from earnings releases remain unstructured, and XBRL versions are voluntary. We believe that requiring companies to tag their earnings releases … will be beneficial for investors…Some very rich data exist in the management’s discussion and analysis (MD&A) section of filings. Unfortunately, the MD&A section falls outside the scope of the XBRL mandate. Requiring this section and other numeric data to be tagged would open up a trove of valuable data for all investors.”
Investors don’t look at financial fundamentals for small companies, like those in the biotech industry.
While alternative data like clinical trials and regulatory issues are certainly important, financial information like cash burn rate, options issuance, stock-based compensation, and operating income, are critical to understanding the health and future performance of a company, regardless of size. If supporters of the bill truly believe that financial fundamental data is not important for small companies, then why are small companies required to report it at all?
XBRL is an outdated technology, more state-of-the-art technologies should be used.
XBRL was initially conceived of in 1999, and implemented by the SEC for corporate reporting in 2009. But a good, functioning standard evolves. That’s why XBRL International, the global nonprofit consortium, has an active, industry-driven, standards committee that is tasked with continuous improvement. The XBRL 2.1 Specification, in use today, was released in 2013; it has evolved to meet changing technologies and industry requirements. In 2009, XBRL was based only on XML; today there are versions of XBRL for JSON and CSV, again, to meet the changing needs of the market.
But regardless of the ongoing development of the spec, working with a mature, tested and proven specification, is not a bad thing. Good standards are made to last. HTML was first proposed by Tim Berners-Lee in 1989, but HTML continues to be widely used 29 years later, its value as a standard has not changed.
Company management will make the best decision about what their investors need.
HR 5054 provides an option for small companies to file in XBRL even though the legislation would eliminate their requirement to do so. Many believe that company management is better equipped to know what their investors need than the investors themselves; therefore, if filing in XBRL is important, company management will make the decision to file in XBRL.
Despite this option, if the bill were to pass, most small companies would opt out of XBRL. Corporations focus on meeting compliance requirements; if not required, the assumption would be that computer-readable data is not needed. Investors, not management, should determine if computer-readable data is needed by investors. Earlier in this blog, I noted that investors do want structured, computer-readable data. In fact, they want more of it, not less. Investors and company management are not always on the same page.
XBRL preparation requires companies to hire XBRL experts to prepare their filings.
In 2009, when companies first began to file in XBRL, there was not a lot of expertise at public companies, and some may have hired consultants for help in getting on track initially. Today, there is a mature, competitive marketplace of tools and services that make XBRL preparation dramatically easier than it was nine years ago.
More and more companies are managing the process internally with the same staff that handle traditional financial accounting activities. Many companies have opted for disclosure management systems which handle XBRL, document, and press release preparation – all the components needed for compliance. These tools make the entire disclosure process more efficient than it was in the days before XBRL.
Small businesses are not going public because of the burden of XBRL.
The cost of going public ranges from $10.1 million for companies with revenue less than $100 million, to $14 million for companies with revenue between $100 and $250 million, according to a study conducted in 2017 by PwC. Based on those figures, average annual XBRL costs as a percent of total IPO costs are around 0.039% to 0.054%. And it should be noted that XBRL-formatting is not even required in the first year of an IPO.
It’s important to get the facts right about XBRL. Don’t throw out a tool for investors that is improving the efficiency of information in the capital markets.