“XBRL Mistakes Really Hurt: Why Accuracy is Crucial for Your Company’s Communications with Financial Markets” (read full)
(Dimensions is an online magazine by Merrill Corporation)
The article discusses the implications of mistakes that filers make in their XBRL reports submitted to SEC. Although oftentimes such mistakes make a big difference in the interpretation of the financial report, one should not overlook the fact that if compared with plain HTML reports, XBRL facilitates a lot the process of searching and removing such mistakes. The companies just have to realize how easy the process of cross-checking reports has become and harness some new tools such as XBRLAnalyst platform for reviewing the results of their XBRL filing process. In short, as soon as the companies begin using the results of their XBRL work for internal financial planning and analysis, the quality of XBRL reports will become satisfactory to the investor community too.
For the article in Dimensions, I had to reflect on several key questions concerning XBRL Data quality.
1. How does poor quality or errors in XBRL filings impact the users of financial information?
In fundamental research, I dealt with large data sets generated from physical processes. We were used to the fact that the data needs to be cleaned before performing any analysis. This process actually stimulates our thinking about the nature of the errors. It also helps later on in the development of techniques for identifying rare events and outliers that are usually of great interest to the financial analysts. What is important to realize is that there are systematic and so to say unpredictable errors (called stochastic noise in scientific terms). I think the number of systematic errors decreases as XBRL matures and the standard setters including FASB and SEC work hard to eliminate those. The unpredictable errors are of great interest to researchers in the accounting field, because they may result from poor software quality, lack of technical qualification of reporting personnel or even intentional distortion of financial information in the reports. All that probably existed before XBRL but we never had the right tools to investigate it.
Before the emergence of XBRL, the original data source, i.e. financial reports, would inevitably be distorted by the normalization process of data aggregators. There was no efficient way of working with the raw financial information reported to SEC. We find our role at FinDynamics in developing the right software (in scientific terminology – the right instrument) for working with the raw XBRL data in an efficient way. We deliver that data to Excel in undistorted form so the accuracy of our data is identical to that of the original source. Obviously, the data quality of the original source is an irritating issue for us and for our users. However, it is something that one can deal with, because it is still “cleaner” than what the data aggregators may ultimately offer. In the past – before XBRL, the ability to extract efficiently the financial information reported in HTML or PDF format did not exist at all. So, the data quality should be perceived and evaluated in comparison with what existed before XBRL. A group of researchers from UWCISA (University of Waterloo’s Centre for Information Integrity and Information System Assurance) with whom we collaborate, has recently shown that the quality of data provided by some leading aggregators is far from perfect .
What we think is important to highlight today is the fact how easy and cost effective it has become to identify and quantify such errors due to XBRL. In my opinion, it is a great advantage of XBRL standard rather than a drawback. It is only a matter of time when the quality of XBRL data stops being an issues or concern for filers and consumers of financial information.
2. Why the XBRL quality issue persists?
I think the companies need to integrate XBRL with their data management systems and use it for internal financial planning, analysis and reporting. It is like if you were a chef and never tried any of your own dishes. One of the strangest and common places where companies make errors is DEI (Document Entity Information), which contains some simple information such as company name or reporting period. It goes to show that it’s not the complexity of XBRL standard or technical skills required that results in such mistakes in XBRL reports. If the companies were to consume their own XBRL reports internally, they would have noticed such mistakes right away. For that, they need to consume XBRL information in every day task, for example for financial analysis in Excel. This is why we focus on enabling XBRL in Excel.