Big Data: Everyone’s talking about it, but what is it, and more importantly, can it make money? Basically computation of such large, numerous and sophisticated datasets that the process requires new paradigms of storage and analysis, Big Data presents big challenges to be able to address it correctly, but creates big opportunities for those who can do so.
In the past, a firm’s value was in the experience of traders, and their ability to assimilate information and spot, recognize and react to events. This task became onerous as data volumes increased, prompting firms to translate that experience into algorithms with greater capacity, which can react faster and more often.
But even the best algorithms risk being overwhelmed by the amount of data out there—especially as firms seek to incorporate new inputs, unstructured data, information on supply and demand, among numerous other content types.
For example, this week, we report on how StarCompliance—a provider of software that monitors the personal trading activity of employees of financial firms to ensure they are complying with the firm’s code of ethics—is using content from news aggregator Acquire Media to reconcile trading activity against news activity to spot potential insider trading by individuals or the firm, using a weighted trailing average news impact score calculated by Acquire on pre-filtered content, so as not to overwhelm compliance officers.
Of course, there are some projects so mind-bogglingly complex that they make this important service look like child’s play. And key to making these understandable (i.e. valuable) at the consumer level—whether that consumer is a human monitoring compliance or an algorithm poised to fire off trades—is reducing complexity early in the process. For example, in this week’s Open Platform, Jeff Wootton of SAP recommends using event stream processing technology to perform “smart capture”—which performs some analytical processes within the data stream itself to reduce the amount of work that must be performed to gain valuable insight once the data reaches its destination.
In a real-time environment, this could be used to create new derived inputs to support trading, but risk is an obvious sweet spot for the application of Big Data. With so many initiatives underway, the best way for firms to address these and ensure compliance is to capture as much data as possible to cover all their bases.
As Waters reporter Steve Dew-Jones reported from last week’s Big Data Online Summit, hosted by WatersTechnology, the “holy grail” of Big Data is “continuous risk analysis,” said independent information architect Tom Dalglish. “Pre-trade risk analysis is where regulation is driving us. If that’s taking 30 minutes, it’s too long; we need it to be in seconds,” Dalglish said.
And Big Data is key to achieving this across datasets as diverse as market data analytics, intraday risk assessment, and real-time data on liquidity and funding, but breaking down traditional business silos to leverage the skills and knowledge trapped in those silos across the enterprise, reported Tim Bourgaize Murray, citing remarks at the Summit by SAP’s Stuart Grant.
But before this can begin, firms must address the people problem of Big Data: finding the right engineers to build a Big Data architecture that will deliver the desired results, and to develop the analytics that will reap those results, reported Anthony Malakian from the summit, citing Howard Halberstein, lead solutions architect for Unix at Deutsche Bank, who said assembling and storing the data is less problematic than knowing how to use it. “When you have that many variables… to correlate… where do you start?” Halberstein said. “I’ve seen things get mired when they get to the analytics piece [because] they can’t get any usable information from this large pool of data.”
In short, while Big Data will be a requirement going forward, the most advantage will be gained by those who can overcome these challenges and be first to use it to their advantage across their lines of business.
Jesse Lund talks about real uses for DLT in the capital markets, lessons learned while rolling out IBM's blockchain platform, and what’s ahead for 2018, and into 2019.Subscribe to Weekly Wrap emails