Like all buzzwords, Big Data now has an almost Rashomon effect on people especially as financial firms are trying to make sense of the various definitions and strategies that are emerging around it.
The standard definition of Big Data is that it involves large sets of key information—a typical dataset exceeds 30 petabytes—that can overwhelm traditional IT infrastructures. For securities firms, this is frequently a front-office problem, particularly for algorithmic and high frequency trading, real-time risk analysis, and pricing systems that require huge amounts of market data. But there are many signs that managing Big Data issues is quickly turning into middle- and back-office opportunities for vendors and their clients.
For instance, Digital Reasoning, an unstructured data analytics provider to U.S. intelligence agencies is moving into financial services and this morning announced that Cristóbal Conde, the former president and CEO of SunGard has joined its board of directors. In his new role there, Conde will help the company apply its technology, which “reads, resolves and reasons across hundreds of millions of documents,” to the many unstructured data problems facing financial services firms, according to company officials.
Reconciliation, enterprise-wide risk management, security master databases and the myriad strains of unstructured data that support structured products immediately come to mind as potential candidates for Big Data management analysis.
Yet developing ways to analyze and manage Big Data will be daunting tasks. Even so, some brave senior-level IT managers at large investment banks and buy-side firms are attempting to do so. In fact, they are using the fuzziness of Big Data to their advantage when it comes to getting the funding they need to revamp middle- and back-office operations. I am working on the November print edition of FTF News and have spoken to several key players in the industry about what Big Data means for operations. What they have discovered is illuminating.
First, almost all of them report that there is a need to manage Big Data for real-time, risk management purposes, regulatory reporting, and the move to execute and clear over-the-counter (OTC) trades—just to name three situations. Big Data is serving as a vehicle for major, over-arching data backbone projects that have budgets in the hundreds of millions and expiration dates a decade away. The business drivers are the usual suspects—cost savings, new regulations and the competitive advantages that come from operational efficiencies.
This scenario harkens back to the now quaint Y2K projects in the last years of the 20th Century when anything and everything was overhauled to avoid fiascos. But there is a big difference with these current projects—they have to show an astronomical ROI—one source says they have to have cost savings that approach $1billion.
One important way this can be achieved, it is argued, is by eliminating the duplication and redundancies of market data products and services within the enterprise. These firms hope a streamlined, silo-busting data backbone across the enterprise will dramatically shrink their extremely high data bills.
Yet my sources tell me that these behemoth projects almost always stumble and fall under the weight of their own grandiosity; sometimes the sponsors are not there when the project ends due to political battles or a failure to meet key project goals in time. However, given that Big Projects impress the C-level executives, they are getting funding.
I am told that the best approach for IT managers charged with making the Big Data Dreams real is to complete less sexy but more deliverable projects such as a security master overhaul while working under the Big Data umbrella. Breaking it down as much as possible appears to be the prudent course at least for getting the green light—while the funding lasts. Huge projects have a way of getting stalled or rendered irrelevant by major shifts in the markets.
Among the many clouds (not the computing kind) on the horizon the most disturbing is the European sovereign debt crisis, which has been looming for months. Events over the coming days, weeks and months will clarify where Europe is headed. If the debt crisis causes an economic earthquake for Europe as some fear, then there will be severe aftershocks here.
If those aftershocks hit, I predict that most Big Data projects will quickly wind up on the shelf.
Need a Reprint?
Leave a Reply