Sorry wrong number – Study finds financial results of 209 listed Indian companies don't add up

One thing investors should be able to count on is that audited financial statements reported to public regulatory repositories at least add up, but a recent study shows that is not the case. In a recent study undertaken by IRIS Business Services it was discovered that 209 listed Indian companies had discrepancies in their reports and that the numbers did not add up. IRIS discovered this by converting over 1400 listed Indian companies' primary financial statement into XBRL. It does not really matter whether the quantum of discrepancies are material. The very fact that there are discrepancies is a serious enough matter warranting further examination to prevent such occurrences in the future.

The errors were discovered in the course of creating India's first-ever corporate fundamentals database in XBRL. The project used the taxonomy approved recently by the Institute of Chartered Accountants as a first step toward full-fledged adoption of XBRL in India. It might be unfair to simply blame the auditors, as it is the responsibility of company managements to get it right, but the fact that this information has gone undiscovered for years should raise the questions, how accurate is the information reported by companies to the marketplace. Although IRIS Business Services study declined to name the errant companies, they did share the key findings of the study.

Investors who risk their hard-earned cash in equities need access to timely, relevant and accurate financial and business information. For the capital markets to operate most efficiently, information about public companies must be understandable, accessible, accurate and, most importantly, trusted by market participants. In the current state of information access, there are multiple problems in making this level of clarity, accuracy and public trust a reality.

One of the biggest roadblocks is that this information is provided in many different proprietary data formats, making it difficult to access, integrate and analyze this information in a timely, complete and accurate manner. The Internet and electronic communication has ensured that information is more freely available than ever before and that the time it takes to deliver that information has sharply decreased. The key question now is: How reusable and accurate is the information? Even when you know exactly what you are looking for and roughly where to find it, extracting information from financial and business reports today generally involves a frustrating experience and a time-consuming and largely manual effort. The biggest problem is that the format and media on which financial and business reporting data are authored varies widely between paper, html, pdf and other human readable forms or proprietary electronic formats tied to a specific software application. Each publishing format has its limitations and they can all only be interpreted by manual human processing and that is why accounting discrepancies like the ones discovered in IRIS' study go unnoticed.

This study comes at the time when the Indian capital marketplace is already being questioned about its regulatory oversight. Just recently a leading Indian outsourcing company Satyam Computer Services publicly announced that it had significantly inflated its earnings and assets for years, consequently throwing the whole local capital market into turmoil. Many observers are now asking whether similar problems might lie buried elsewhere.

The challenges associated with accurate, timely financial reporting are apparent from the number of other accounting scandals uncovered during the past few years. These scandals have been widely reported and the estimated cost has grown into the billions of dollars. ¹ Restatements of financial results by public companies soared in 2005 as auditors were forced to drill deeper into corporate accounts, in part because of a sharper focus on requirements laid out by the US Sarbanes–Oxley Act. ² This brings the attention to that a non-technological solution can go only so far. This is why the underlying design of the XBRL technology specification might be the magic wand to bring accurate and timely results because it integrates technologies with the basic accounting framework. Using underlying linkbases, XBRL allows every item defined in a financial statement to carry multiple attributes, including the interrelationship with other elements. This simple, yet powerful concept of linkbases allows every accounting element to be validated before being submitted to a regulator.

Of the various linkbases that are part of the XBRL framework, the calculation linkbase automatically calculates if the numbers are added up correctly, and thereby prevent even inadvertent errors while preparing and submitting financial statements. The underlying idea of a calculation linkbase is simple. The calculation linkbase defines basic calculation validation rules (addition/subtraction), which must apply for all instances of the taxonomy. For example two elements (concepts) A, B can be summed up to a third element (concept) C, such that C=A+B. Perhaps, this function of the XBRL specification alone is a reason for XBRL as a standard to succeed, but whether XBRL will actually prevent fraud is not a moot question. There are now six different kinds of linkbases in the specification. Each has a special purpose, but goes beyond the discussion of this study and will be explored later in the next phase of the study at IRIS Business Services.

The academic community is vital to the success of XBRL and much research has yet to be done before the pros and cons of XBRL are identified. But one thing for sure is that XBRL will unlock the information from proprietary data format and make this important information available in a timely reusable information standard, and hopefully bring trust back to capital markets.