One of the top challenges accountants face in their efforts to use big data is that much of the data is unstructured, such as the plain text found in the Management’s Discussion and Analysis (MD&A) sections of a company’s 10-Q and 10-K reports.
Meanwhile, other data that accountants are increasingly using – or data that they will be using more commonly – is also unstructured, including huge volumes of text that resides across the Internet. These include online reports, regulatory information, generally accepted accounting principles information, social media sites, online forums, and other online data sources.
Accountants and other professionals often struggle to gather and examine vast streams of unstructured data – including trying to pinpoint and act on key words in financial statements.
But predictive analytics tools let accountants analyze both structured and unstructured data sets across financial databases, Excel spreadsheets, web-based financial statements, and other systems where critical information resides. Predictive analytics and data discovery tools can help accountants identify and manage potential risks before it’s too late to act on them.
For instance, the manipulation of event timing and irregular access to critical earnings and performance data can generate the risks in stock option backdating cases, according to an article in the Journal of Accountancy. Email data – as opposed to data contained in general ledgers – generally reflect the timing and intent of the information and activities that lead up to transactions in these cases.
Accountants and other examiners can use predictive analytics and data discovery tools to evaluate email traffic surrounding each stock grant to determine the intended data of the grant along with the date that it was actually communicated. Meanwhile, analysis of the electronic files used to support a stock grant memorandum may reveal that a memo was created, finalized, and printed well after the memo date and grant date that have been entered.