Greater transparency, lower filing costs, reduced risks and better data quality – these long-promised benefits of XBRL are finally becoming a reality. As the XBRL landscape continues its swift evolution, we discussed the latest developments with five experts – as well as with a representative of the SEC.
Vintage question > A lot of consumers of XBRL data lament the low quality of the data. In what ways do you think XBRL implementation and data gathering can be improved?
Campbell Pryde > This is the main task we have on our plate right now at XBRL US – improving the quality of the filings. We’re finally reaching the point at which there are enough people taking in this XBRL data and making it available that people are coming to us with much more realistic questions about how the data can be improved, because they’re noticing enough of the errors. We have an initiative underway to address all the errors, an effort in which Vintage is involved – the Data Quality Committee.
We’re trying to set very prescriptive guidance to help filers get their data into a form that can be used by the capital markets. We’re doing that in two ways: by issuing guidance and also by publishing rules that companies can run over their filings before they file with the SEC. An important component of that, obviously, is SEC buy-in. We’re spending a lot of time with the SEC – doing regular updates with them to make sure that they’re supporting the initiative and monitoring the results.
Pranav Ghai > What I would say is that there are small errors, and then there are really big errors. People complain about the errors in XBRL, but there are an equal number, if not more, in data collected by hand. And XBRL is transparent enough that the errors usually pop out at you.
The second thing is that Calcbench does a lot of data cleansing for our clients. Our systems catch a lot of the low-hanging fruit. Once you have a certain number of data points to look at, you can discern patterns in them. For instance, you know that shares outstanding is typically in the millions or more. If you see something with shares outstanding in the trillions or the thousands, you sniff it out and you say, “Hey, something is off here.” You can catch some of that stuff rather easily. It’s true that there are other types of errors that are harder to catch, and we’re working on those. But my guess is that even the best data vendors haven’t corrected them yet.
In addition, XBRL US has created the Data Quality Committee, as Campbell mentioned, and the idea is that this consortium is putting together a set of rules that will be guiding principles for how companies should file their XBRL. The idea is simply that it will increase adoption by decreasing the error rate. That is one very tangible step the community is taking. These principles are then being put into practice by the filers, meaning the corporates themselves, or by the companies that handle their filings, such as Vintage.
Jaret Klekota > Two studies have been done on XBRL in recent years, one by Columbia University’s Center for Excellence in Accounting and Security Analysis and another by the CFA Institute. And both studies came to the conclusion that the reliability of the data needs to be improved in order for it to be better utilized. They concluded that one possible way of improving it is to require some type of assurance or audit of the XBRL. In the CFA Institute survey, 50% of respondents said they thought that XBRL should be incorporated into the standard financial statement audit and 19% said they thought a separate audit should be done around XBRL. So there are different groups that have been advocating for ways to help improve quality. I also think education is key for companies — understanding what all the rules and risks are. A lot of companies need to step back and ask themselves if they have a formal enough process in place for filing. Do they have formal controls and policies in place that make them comfortable with the quality of their XBRL?
Vintage question > XBRL data can obviously be a useful tool for investors, but can it be beneficial for companies as well? In what ways could it be used?
Ilya Vadeiko > I would not recommend that each individual business maintain their own database of XBRL data of other peer companies, because it requires considerable experience to achieve good comparability and usability. However, if they use XBRL data for reporting and make it available to aggregators such as FinDynamics, they could subscribe and obtain the high value data of their peers at a low cost. The benefits are significant if XBRL is used wisely.
It ensures consistency within one report and across many periods if used correctly, which is important for reliable analysis. In the past, if an analyst wanted to go beyond the three primary financial statements into more interesting disclosure areas, the amount of effort and time required to collect the information from unstructured HTML reports grew exponentially.
This is why a handful of traditional data aggregators that provide some normalized disclosure data from HTML reports charge enormous fees, because that data used to be manually collected and still had a fair number of errors. With XBRL, it’s becoming a lot easier and more reliable and verifiable.
Campbell Pryde > We’re actively trying to give public companies access to the data – we have an API that people can use to pull the data into a spreadsheet. We have a number of public companies that use that data to look at their own filings, and then do comparative analysis as well. They can do their own modeling for their own corporate purposes.
Pranav Ghai > There’s a really nice anecdote I can share with you from one of our customers about this.
The CFO of a company went to a presentation that a group of consultants were giving, and their whole approach was something like, “You must do corporate benchmarking to figure out where you stack up against your peers. But we can do it for you. It will cost you half a million dollars a year.” This CFO came back from the meeting and said to one of his employees, “Look, they want to charge us half a million bucks a year, but I know we can do the work. And we can do it better because we understand our market. Go find some tools that’ll help.” That employee did his homework and he came to us. And we were able to give him the data using XBRL. That was it. It was literally that easy.