Alfred has every right to grumble. He spent "five long years" huddled in public disclosure rooms at the Department of Labor poring over paper-based retirement plans and auditors' reports before launching BrightScope. Today, his San Diego-based company sells Web-based software that provides corporate plan sponsors, asset managers and financial advisers with in-depth retirement plan ratings and investment analytics. Many of BrightScope's ratings are published online for free, while advisers and large enterprises fork over anywhere from $5,000 to $200,000 per year for highly sophisticated and customized prospecting tools.
A team of nearly 15 data analysts processes more than 60,000 retirement plans per year while integrating data captured from SEC filings, financial industry regulatory authorities, census data and public websites. Once collected, the data is then cleansed and linked together using proprietary mapping algorithms.
"There's a big difference between free open data and actionable intelligence," says Alfred. "Companies don't want to purchase data that requires data analysts and infrastructure and integration. So there's a ton of work on both the manual side and the engineering side to get the data into a format that can be reliably used by our clients." That's a complex process that, he says, could be simplified with a bit of help from the government.
The good news is that the government is taking steps to improve data quality. Calcbench has benefited from that effort. The New York-based startup has created a sophisticated engine that turns complex financial data such as earnings statements, cash flow statements and companies' balance sheets into a more readable format. To ensure accuracy, Calcbench uses proprietary artificial intelligence tools that sift through the data to detect errors such as misdated year-end reports. Finance professionals such as investors, auditors and industry researchers use the resulting repackaged and cleansed data to compare financial ratios, examine entire industries and review competitors' disclosure data.
"A lot of what we're charging for isn't the data," says Alex Rapp, Calcbench's CTO and co-founder. "It's how we structure it, how we store it for you and how we solve a tremendous number of business problems by making data comparable across different companies and industries. That's something the data doesn't do itself."
Nor is it a service that would have been possible without the U.S. government's insistence, starting in 2009, that the financial information in corporate SEC filings must be in XBRL (Extensible Business Reporting Language) -- a freely available and global standard format for exchanging business information. Whether you're downloading quarterly filings directly into spreadsheets or analyzing year-end statements using off-the-shelf software, XBRL converts fragmented financial data into a machine-readable format. And that makes life easier for Calcbench's techies.
But as government agencies inch toward cleaner, more structured data sets, a handful of startups are offering powerful new tools for dealing with older data. "These intermediaries are going to make open data more accessible to other types of companies because the reality is, open data is still pretty messy at the local, state and federal level," says Gurin.