More Data, Better Results

 
 
By Dennis McCafferty  |  Posted 2010-08-20
 
 
 

A U.S.-based bank seeks a better way to predict customer behavior. An international bank looks to an IT solution to help manage its mergers and acquisitions. A credit-card company must reduce its exposure to risk, while a Wall Street options exchange simply has a need for speed.

Such developments are key indicators of what finance-sector companies are seeking in IT solutions. Battered by what’s been described as the harshest economic downturn since the Great Depression, these enterprises must make better use of data to slash operating costs and increase revenue. And they’re willing to invest to do so.

Despite the uncertainty of a double-dip recession, enterprise-IT spending in the financial-services sector worldwide will increase this year to $554.6 billion, from $531.4 billion in 2009, according to industry researcher Gartner, based in Stamford, Conn. In 2008, spending hit $565 billion, so there’s clearly room to grow when the economy improves. 

Gartner expects software as a service (SaaS), cost-optimization and risk management/compliance solutions to emerge as key drivers in the demand for tech products. “We’re seeing a moderate recovery in 2010,” says Susan Cournoyer, managing vice president of technology and service-provider research at Gartner, “but we expect a strong recovery beyond that.”

Julien Courbe, a principal at the financial services advisory practice at New York-based industry consultancy PricewaterhouseCoopers, agrees that IT spending in the financial services sector is cautiously rising, and adds that there are many promising high-growth niches. Retail banks and insurers, for example, are projecting a 50 percent to 100 percent increase in online users during the next two years, and demand will be high for solutions to serve these customers.

“There is a great deal of focus on the development of the online-revenue channel,” Courbe says. “But banks, insurers and asset managers continue to struggle with profitability, and are also looking for ways to intelligently reduce costs and, at the same time, develop new revenue streams. Customer relationship management solutions can help meet those goals. CRM is seen as a way to help customers better protect investments and returns while still gaining a greater ‘share of wallet’ among existing customers.”

Baseline recently spoke with four enterprises in the banking/finance industry that are benefiting from IT solutions. In all four cases, the companies are looking for a better way to use or manage data. Yet each case presents its own distinct story about how IT can provide these services.

Tapping Into CRM  

In the banking industry, “upselling” and “cross-selling” programs are where the money is made. A person who walks into a branch to deposit a week’s paycheck is a good customer, but a walk-in who deposits a week’s paycheck into a certificate of deposit (CD) is even more valued.

To gain better predictive insight into its customers, Memphis-based First Tennessee Bank sought an advanced CRM tool to save on marketing costs and time by distinguishing the customer niches that are likely upsells. For years, First Tennessee had relatively little data about this.

“We’d know how much business customers were doing with us,” says Tanner Mueller, who manages the marketing database for First Tennessee, which oversees more than $20 billion in assets. “And we’d know, for example, whether they were customers who only had a checking account. That made them a good candidate for a savings account. We’d know if they had a CD about to mature, because that would make them a good candidate for a new CD. But it really wasn’t a lot of demographic data to work with.”

That’s when First Tennessee deployed IBM’s SPSS predictive-analytics solution, which enables the bank to consider more than 125 sources of information. This includes external data, such as fluctuations in the Federal Reserve interest rate, along with a wealth of internal data about customers, including income, occupation and even the publications they’re likely to read. Using this broader breakout, the bank’s ROI on a number of key cross-sell campaigns increased by 600 percent.

“To go from using just two or three demographic points—and maybe five or six banking transactional ones—to [125 sources of information] is of great value to us,” says Mueller. “We now rank customers based on a score system and push that data to those on staff who can make the best use of it. It’s great to go to our product managers with this kind of predictive information, which has a lot more meat than what we were providing before.”

The State Bank of India has also signed on with IBM’s CRM packages in order to better share critical data throughout its organization—including foreign offices, associates and subsidiaries—and enable more informed business decisions. “This solution will allow our bank staff to concentrate on valuable tasks such as upselling and cross-selling, performing more sophisticated analysis and delivering a higher level of customer service,” says Krishna Kumar, deputy managing director of information technology.

Bringing Banks Together With Virtualization

Mergers and acquisitions are business-as-usual for the banking/finance industry, and Brussels-based Dexia Bank Belgium needed help to make a successful transition after buying two other significant banks. All three of the banking organizations had the basics when it came to IT infrastructure: local servers and file services and access-management solutions. There was also a mix of needed applications that couldn’t be easily migrated into the new, unified structure. When Dexia attempted to bring all the technology together, the outcome was predictable.

“The engineers would say the infrastructure was outdated,” says Thierry Abeels, a member of the executive board of Dexia’s technology services. “The users would say that accessing services was too complex. The managers would say that fixing it was too expensive. 

“All the traditional IT solutions we had were very limited in functionality, with no mobility and high costs for maintenance. To integrate these different banks into a single one, we needed a new, modern, single front-end IT solution.”
That solution would have to be put in play within less than 10 months of the mergers for more than 6,500 people located in more than 1,100 offices—with no disruption to branch activities. And it needed to be mobile, so the
mortgage and stock-exchange experts could both work from the office and take projects home. 

Dexia chose a client-virtualization solution that resulted in a centralized architecture based on 300 Hewlett-Packard BladeSystem servers in the bank’s data center and 6,000 HP thin clients in its branches. The solution also incorporates Citrix MetaFrame Presentation Server software 3.0, which provides local and remote users with secure access to virtualized client/server applications from any location, device or connection.

The new infrastructure enabled the bank to eliminate the 1,100 legacy servers in its branches. In addition, what Dexia calls its “Branch of Tomorrow” includes a mobile office component, with HP notebooks and DeskJet printers, that enables staff to conduct sales calls from any location, 24/7, within legal restrictions. Employees’ mobile computing devices are network-based thin clients with card-reading, scanning and other capacities—all fully integrated regardless of which legacy system the employee had been working with. 

“We have no servers and no PCs within the branches anymore,” Abeels says. “We’ve moved from PCs with high failure rates to thin-client solutions where that doesn’t happen. We’ve also greatly reduced global electricity costs.” 

Of the 5 million credit card applications processed each year by Sioux Falls, S.D.-based Premier Bankcard, 2.5 million are submitted by people who fall below the FICO median credit score of 660, which deems them  “subprime” customers. Since 25 percent to 35 percent of these customers will default on
the first payment, Premier is improving risk management so it can provide credit to customers who are more likely to pay.

“As we come up with pricing models, we’re factoring in the cost of risk loss,” says Rex Pruitt, who manages profitability/risk systems for Premier. “The ‘good risk’ has to be good enough to pay for the bad. Every good card we approve helps our bottom line.”

To reduce risk, Premier depended on credit scores from FICO and other providers, but the credit score alone doesn’t tell enough of the story behind each applicant to make a good business decision. So the company turned to Enterprise Miner, a data-crunching tool from SAS that allows Premier to feed thousands of variables into the database it uses to guide its risk management practices.

“We can feed the FICO score, but we can do so much more than that,” Pruitt says. “For example, we can input whether an applicant has had a collection notice due for more than $100 within the last six months, or whether the applicant has ever filed for bankruptcy, and, if so, how long ago. We can take into the account the frequency rate of payment delinquencies. A FICO score may be low because of something that happened a long time ago, as opposed to within the last six months.”

Even those who fall short of a good credit score will often get a card, but the data mining solution allows Premier to identify those who will need greater monitoring for fraud-control mitigation. The reduction of fraud cases has resulted in $9 million in annual savings. 

In addition, Premier is saving $21.3 million a year in the avoidance of what it calls “soft fraud” cases: those in which customers secure a card without any intention of ever paying off any charges. Finally, Premier can now process and analyze information 75 percent faster than before.

Evaluating Options

Speed is also critical in the trading industry, where those who produce the quickest, most accurate numbers end up on the winning side. So technology leaders in this field are always looking for an added edge when it comes to
IT solutions.

Take the New York-based Inter-national Securities Exchange. As one
of eight options exchanges in the country, ISE provides two-sided quotes for more than 275,000 options series that trade on the exchange, and that can generate as many as a billion quotes on a peak-performance day—the equivalent of more than 25 gigabytes of data. 

ISE sends this data to the Options Price Reporting Authority (OPRA), which aggregates quote and trade information for the entire industry. “The U.S. options market is driven by a huge amount of data,” says Jeff Soule, head of market data for ISE. 

“Intra-day peak message rates can reach more than 1 million messages per second. In this case, a ‘message’ is a quote or a trade. When the market is moving quickly, the data is moving quickly right along with it. Speed and capacity are both critical to compete in this industry.”

In 2009, ISE decided to enhance its front-end trading application, PrecISE Trade, with the introduction of options analytics. It ultimately turned to Hanweck Associates to launch its VoleraFEED options analytics solution, which can perform tens of millions of options valuations per second. Price feeds, dividend maps, yield curves and rebate curves are all part of the data that the analytics engine can crunch.

Most significantly, VoleraFEED provides ISE’s members with the tools to determine an option’s volatility, or the anticipated percentage of movement (up or down) on the option’s price over a given period of time. That’s a key determining factor in pricing the option. 

Whether financial institutions want to predict customer behavior, manage mergers and acquisitions, reduce exposure to risk or perform millions of transactions per second, technology tools can enable and support those objectives.