When Credit Suisse announced this past September that it was going to be "aggressively reviewing" its market data spending as part of a major cost-cutting campaign, it underscored an industrywide sea change -- market data content doesn't carry the cachet it once did. "We believe that by operating at best-in-class standards in this area, we can achieve significant cost savings," said Brady Dougan, head of Credit Suisse's investment banking division, in a company memo. As a result, the Credit Suisse staff now is required to identify any nonessential market data, and new market data requests require COO approval. Credit Suisse executives were not available to comment on the firm's market data strategy for this story.
But keeping costs under control is only half of the market-data transformation taking place, according to TowerGroup analyst Tom Price, who says aggregated market-data feeds make little sense in today's low-latency world. "There's more of an appetite for direct exchange feeds," says Price. "That's why content has been eclipsed by infrastructure itself. The spend is going to be on the infrastructure side, and spend on content is really going to slow down."
In addition, Price adds, Reg NMS in the U.S. and MiFID in Europe will create even more data, which further drives the need for robust infrastructures powered by increased bandwidth. "Market-data infrastructure is really going to be one of the main drivers in electronic success," he asserts. >>
JPMorgan Worldwide Securities Services understands the critical role a robust infrastructure plays in generating clean, accurate data. "We are leveraging our operations and pulling them together into a Center of Excellence in order to apply best practices" to data management, relates Peter Serenita, SVP and chief data officer, JPMorgan Worldwide Securities Services. "We measure the straight-through-processing rate of the information that comes from the vendors to see where the failures happened. We determine the root cause of those and work with either the vendor or internal staff to remedy the straight-through-processing break. This has been a big internal focus for us."
The Center of Excellence isn't out to generate more data; rather, it's goal is better data, according to Serenita, who says that while it is possible to pay less and still get quality data, it doesn't happen overnight and requires a lot of "heavy lifting." "We take measurements that are really trying to understand where the leverage points are," explains Serenita. "Our typical leverage points are the traditional three legs of the stool: vendor, operational and technology costs. You can't work just one leg of the stool -- you have to work all three simultaneously."
In essence, when JPMorgan receives data from outside vendors, the data undergoes a myriad of business rules, validation and internal checks. If the data should drop out of one of those rules, it hits an operational repair queue, Serenita relates, indicating that an operations person needs to take action. "It might be that the data didn't hit a tolerance check or that there were inconsistencies between two vendors," he says. "The operations staff would then need to find out what the truth is, repair it and send it on its way -- whether to the front office, operations or client reporting systems."
And while a Global Market Reference Data (GMRD) infrastructure was implemented at JPMorgan before Serenita was promoted to his current role in late summer 2006, he says the firm remains committed to improving its capabilities. The GMRD system holds more than 5 million securities in its source database and was able to help JPMorgan triple the number of securities that it processes through automated feeds without manual intervention, resulting in significant cost savings, according to Serenita, who is the first JPMorgan executive to hold the title of chief data officer.
"It's particularly important to partner with our vendors to work through some of the information we get from them and how they can expand their product set," Serenita says. The partnership creates a win-win situation because JPMorgan is able to "uni-source" its data instead of multisource, he explains, and the vendor is able to improve or extend its functionality as a result of the partnership.
Another program that extends the GMRD's capabilities is Central Price Capture (CPC). While GMRD includes all of the securities-master or securities-reference data, CPC involves gleaning data from outside market-data vendors to price those securities on a daily basis, according to Serenita. "We can then manage that information holistically and internally so that we can reduce our costs," he relates. "In the past, our operations group might work the same data multiple times. Now, we will receive the pricing data once, view and scrub it once, tolerance check it once, and then, by the time it comes out of the CPC system, it will be a scrubbed, validated price that the rest of Worldwide Securities Services can use for their valuations and client services."
Serenita says his firm also is in the midst of spending about $600 million on new technology services, a portion of which includes the continuing build-out of its market-data infrastructure, an area that is critical not only to JPMorgan's operational efficiency, but to client relations as well. "Without it we would be very fragmented," says Serenita. "Our clients would see that fragmentation," he adds.
"We want to provide our clients with very clean, consistent data," Serenita continues. "If we were to put manual procedures in place to manage [market data], it would increase the cost to everyone, and we wouldn't be generating attractive price points for our clients."
State Street Sizes Up Compliance
When it comes to enhancing and scaling data infrastructures, JPMorgan Worldwide Securities Services is not alone. John White, principal, global data management, at Boston-based State Street Global Advisors, also understands the challenge of accurate pricing for clients in the face of "more data coming through the door than ever before." And, similar to JPMorgan, White says State Street is devoting significant resources to building a sound infrastructure over the next few years to house data that will take up gigabytes of space.
"That's actually becoming quite a challenge for us," says White. "What if I bring in historical pricing -- which is end-of-day pricing -- for a security going back 10 years?" he asks. "Now all of a sudden I have to grab every tick for the past 10 years. You then understand that the breadth of data, and the type of technology and storage you need now is very different," White explains.
In addition to ramping up bandwidth and scalability, State Street also is actively examining contracts with data vendors, another layer of complexity in the data-management puzzle. Data compliance is a top priority at State Street, as vendors and regulators increasingly are conducting audits to ensure that companies actually are paying for the services they use, according to White.
As a result, State Street currently is reexamining all market-data vendor contracts to ensure it is properly using the data for which it pays, White says. If discrepancies are found, he notes, State Street will immediately set out to update or renegotiate its vendor contracts. "That's really taking on a huge new scope because we're starting to hit a maturity stage within the life cycle of investment content," says White. "Head executives of many organizations are saying, 'I'm not only looking to make sure our users are getting good, quality data. I'm also looking to make sure I'm not violating any contracts.'"
While White declines to elaborate on the specific vendors with which he works, he says the Investment Management Data Services Group at State Street sources the data and works with both technology and investment teams to create the appropriate applications and data feeds for end clients. "[State Street's] client services is working with us, trying to figure out how to get data out to its clients," he relates. "It's not just about making that feed for them; it's also being sure that when we go back to the vendors we're working with, we show them what we are trying to do and the content that is required," White adds. "Do we have to put it into a PDF format because you don't want it in an electronic format? What are the rules if we put data onto a secure intranet or extranet?" he questions.