Currently Being Moderated

“We can rebuild him,” begins the old television series The Six Million Dollar Man. “We have the technology.”

 

Five Vs of Data 04-09-12-AWith those words, U.S. intelligence chief Oscar Goldman launched Steve Austin’s career (the fictitious astronaut’s, not the stone-cold wrestling personality’s) as a red-jumpsuit-clad cyborg action hero. The bionic body parts that saved Austin’s life after the violent end of a test flight also made him “better ... stronger ... faster” than other Cold War operatives -- and 1970s pop icons.

 

Electronic traders don’t have bionic body parts, such as Austin’s night vision-capable 20.2:1 zoom eye, but they do have visualization tools to better handle unstructured data. Visualization tools can process data volumes that would be impossible for lone humans to process, but the technology cannot see everything by itself.

 

“Visualization offers an invaluable machine-human interface,” said Neil McGovern, SAP senior marketing director and my fellow Trading & Risk Technology blogger. It allows market participants to quickly make sense of both historical and real-time data.

 

And making sense of data in real time is increasingly crucial.

 

Picture This

 

A clear majority of electronic trading participants want access to continuous risk information, according to a poll taken last week during a Sybase webcast. More than two-thirds of respondents indicated that they want to see their positions and exposures to P&L, sensitivities, etc. as soon as possible.

 

McGovern hosted Thursday’s webcast, “Using Visualization to meet Big Data challenges in Capital Markets Applications,” in which he discussed his “Five Vs” of data. The fifth “V,” visualization, was the six-million-dollar center around which the webcast turned.

 

Five Vs of Data 04-09-12-BThe aim of visualization is to stop looking at the numbers, and keep your eye on the trends, according to Peter Simpson, senior vice president of research and development with Stockholm-based data visualization group Panopticon Software. Seeing hundreds of columns graphically (see graph) is much more effective than looking at five columns in a tabular view.

 

Panopticon EX and Sybase ESP provide the tools necessary for the bionic trader to see that big picture, make sense of it, and trade as appropriate.

 

The combined technology queries a database to analyze the data graphically, noted Jeff Wootton, Sybase’s product management team leader for CEP and RAP, as well as McGovern’s and my fellow Trading & Risk Technology blogger. And ESP users don’t have to store data, and then query it; they can perform continuous queries on data streams.

 

Fantastic Four

 

The four other “Vs” were: Value, generating value by mining unstructured data for information is significantly more difficult than doing so with structured (e.g. market) data due to both the complexity of understanding unstructured data and its higher noise ration. Unstructured data encompasses a wide variety of information, including news, weather and social media feeds, as a London-based hedge fund proved when it beat the market by analyzing mood swings on Twitter.

 

Five Vs of Data 04-09-12-CThe velocity of trades skyrocketed after 2007 (see graph). And faster speeds will be more critical than ever this year because the volume of data will continue to grow. Humans will create 1.5 zettabytes of digital information in 2012, as McGovern noted. That is enough data to fully load 50 billion 32 GB iPads, which would take up the entire volume of Taipei 101 more than 22 times over.

 

Finding value amongst the mottled volume of Big Data in real time can be a daunting task. But half of the respondents during last week’s Sybase webcast found Big Data valuable for risk management. Another 36 percent indicated it was valuable for trading.

 

That is because, to quote Austin’s boss, “we have the technology.” Working in concert with these visualization tools will make the right human traders “better ... stronger ... faster” than ever before.

Comments

Actions

Filter Blog

By author: By date:
By tag: