Semantics Help Wall Street Cope With Data Overload
Modern technology has made it possible for financial firms to get more data, faster and with more detail than ever before. Transparency requirements and globalization of the markets only increase the information overload. And the new SEC mandate for XBRL (eXtensible Business Reporting Language) financial statements will expose even more structured information for smart applications to analyze.
Further, not only is the sheer volume of information vastly increased over even the past 10 years, but the shape of the information and options available to access it are in rapid evolution. The Web, for example, comprises more than 30 billion documents. And in just the last two to three years, a huge leap in the amount of consumer-generated content has created a new alternative to traditional market surveys.
Fortunately, semantic technologies -- software that encodes meaning to raw data -- may come to the rescue. Filters, aggregators, intelligent triggers, smart data summaries, and even automated merger predictors and latent trend analytics now are available in the information solutions marketplace.
Opening the Data Flood Gates
Prior to the SEC 2002 Regulation Fair Disclosure ruling, which aims to prevent selective disclosure of information by public companies to market professionals and shareholders, it may have been possible for an analyst to bypass the noise of the information flood via direct communication with managers in order to piece together a coherent picture of a company's position. Post-Reg FD, however, analysts must differentiate themselves not by the relationships they cultivate, but by their skill in slicing and dicing the abundance of market data. Semantic technology just may supply an edge.
Large and small providers of investment research are responding to the changing landscape. Dow Jones and Thomson Reuters, for example, both have made heavy investments in semantics, though in different ways.
Dow Jones has developed a semantic tool, known as Synaptica, that maps information into structures called ontologies, which basically are smart graphs. Ontologies can create models that reflect the fact that companies have subsidiaries and branches in diverse geographical locations. While this may seem to be pretty basic functionality, it actually empowers search to become more precise and comprehensive, according to Christine Conners, Dow Jones' global director for semantic technology solutions.
In searches for "Bank of America," for example, ontologies enable Dow Jones to aggregate the world of data that could affect the target company through people, locations and relationships -- including business partners and acquisitions, Conners explains.
Semantics further allows specific tags or elements of news to be used to drive algorithmic trading, Conners adds. Thus, she says, customers of the Dow Jones ALGO news product were able to make use of the "Bank of America in Advanced Talks to Buy Countrywide" news item to perform immediate trades -- providing a competitive advantage, especially for the 2 minutes when DJ had the scoop on that bit of news.
According to Conners, the firm's value proposition emphasizes speed of information availability. "Optimization of our [internal] user interface in concert with semantic technologies is what really gives us our competitive advantage," she says, noting that Dow Jones processes 180,000 articles from 17,000 sources daily.
Thomson Reuters made a major investment in semantic technology for the long term with its acquisition of ClearForest last year. ClearForest tagging and analytics technology promises to deliver "complex associations, relationships and concepts" out of free-form text, the company says on its Web site. After all, according to Tom Tague, VP and Calais/ClearForest evangelist at Thomson Reuters, in the end, text is what moves the markets.
Tweet This