Big data is certainly a big topic in financial services. On the first day of the Sibos conference in Boston, the importance of data usage and analytics was on display. Data analytics was the first session of the conference's technology track.
Participants in the panel session "Big data is still hungry: crunching more data, faster and cheaper" confirmed that big data was a top priority for banks, but they also indicated that spending is proceeding cautiously. After years of limited success with data initiatives, banks are unwilling to throw money at data management and analytics projects, said Lee Fulmer, managing director of global cash management at JPMorgan Chase and former CTO for its corporate and investment bank. Banking technology "has been tarred by data warehousing projects that we did 10 years ago," Fulmer told the audience. "Collectively, we spent hundreds of millions of dollars as an industry."
However, big data is not seen as a pure technology project at many banks. "We are not launching big data [projects] as technology initiatives. We are looking to deliver value to the business," and JPMorgan is applying big data to particular initiatives that are identified by the business.
Paul Ventisei, global head of architecture at HSBC, said during the session that big data technology has a low barrier to entry, and banks can experiment with different models before launching larger projects. "You have to prove value to be able to invest in new technology. Luckily, there is a lot of open source, so you can start ideas relatively small, and you can scale them to petabytes of data."
[Do you aspire to the C-suite or some other spot in upper IT management? Then bulk up your credentials around today's most pressing IT movement, digital business, at the InformationWeek IT Leadership Summit.]
In order to get backing from the business, HSBC has developed small projects that can be demoed in the company's innovation lab, Ventisei said. "We wet their appetite, so to speak, and that generates interest."
A quick electronic audience poll during the session found that banks are also at varying stages of maturity when it comes to how they implement big data projects. For instance, just 30% of the audience indicated that their bank had been working on big data for 18 months to three years, while 19% said their banks' initiatives started in the past 6-18 months.
As with all new technology, there are obstacles to successful implementation of big data. One of the most complex issues facing banks are the ever-changing data privacy laws, Ventisei said. When different countries look at privacy in different ways, "how does a global bank stay in compliance?"
Lee said looking three or four years into the future to predict public attitudes toward data sharing and privacy laws is almost impossible. To prepare, firms systems need to be flexible. "What we have now may not be legal in the future. If we don't think it through, it could impact product sets, customer servicing, and more. We have to think about our data models now, so we can segregate data in the future" if needed.
According to Fulmer, financial firms also need to come to grips with the seemingly continuous push by business leaders and technology providers to move everything to the cloud. "Data is a competitive asset. Why would I outsource it to someone else? Although there is the regulatory [and privacy] aspect of using the cloud, I'm more concerned about the competitive aspect. Anyone who considers outsourcing data better get their head examined."
However, banks will need to figure out the regulatory and competitive balance for using hosted services. Ventisei mentioned the efficiencies of scale that large hosting providers such as Microsoft, Google, or Amazon can provide. Banks currently have large data centers but do not approach the scale of the mega-providers. The gap between the cloud providers and banks will continue to grow as well, he said.
"When we look at the future of data center, we [UBS] are going to be sub-scale soon," he said. "We are OK for now, but the day will come. To not use the scaled data centers in the future is unlikely."Greg MacSweeney is editorial director of InformationWeek Financial Services, whose brands include Wall Street & Technology, Bank Systems & Technology, Advanced Trading, and Insurance & Technology. View Full Bio