Wall Street can be so hard to please. It seems like only yesterday that bank stock analysts had discovered the importance of information technology and began to view systems as a key factor in whether a bank could go it alone or needed to put itself up for sale. Now, a number of top analysts have begun wondering aloud if banks are spending too much on technology, or at least too much on unproven systems.

"When Wachovia says that they are spending $30 million to $40 million to build a customer information file, is that a good investment decision?" asked Thomas K. Brown, an analyst for Donaldson, Lufkin & Jenrette Inc., at a recent American Banker roundtable. "Could it be $10 million? Should it be $60 million?".

Analysts' rumblings are becoming more audible not because they don't believe technology is critical. Rather, it's because they don't have the detailed information they need to weigh big-ticket technology purchases, they rarely are technology experts, and the bottom-line benefits of systems aren't easy to quantify, at least in the short term.

And, as Robert Albertson of Goldman, Sachs & Co. noted, banks tend to sound alike when they talk about their technological prowess. "Everybody is saying kind of the same thing," said Mr. Albertson, "and then you may find vast differences when you actually say, 'Show me what you've got.' " So how much should banks be spending on systems? Is the industrywide increase in spending - which has exceeded the rate of inflation for four years - the right amount? How can the benefits of pricey new systems be evaluated?

Bank consultants say there are no easy answers to these questions. Some are troubled that technology expenses are rising. Others say Wall Street should be more concerned about banks that aren't spending enough on technology. Most agree the rate of spending isn't going to decline anytime soon.

James McCormick, president of First Manhattan Consulting Group in New York, said it's helpful to look at three kinds of spending - on core processing systems, so-called data warehouses, and home banking initiatives via PC.

Banks that already have companywide systems for core applications are ahead of their counterparts when it comes to integrating even large acquisitions, Mr. McCormick said. Banks that don't have common systems will need to spend more over the next three years, he said. "Some banks are much more effective here," said Mr. McCormick. "It's not shades of gray, but almost black and white."

He added that several banks widely admired a decade ago for their proprietary systems now have a lot of catching up to do because they were slow to deploy newer applications. "It's interesting how the tides shift," he said.

And while the development of electronic channels such as PC banking have received generous media attention, Mr. McCormick said banks generally aren't spending as much as the hype would suggest. Said Mr. McCormick, "You really don't need to spend all that much in that area."

It's the third category - data warehousing systems - that has both the most promise and the greatest risk of profligate spending. Banks are now turning to such systems to identify customers who are likely prospects to buy profitable products. The systems can be costly but the returns in boosting revenue and retaining customers can be significant. But, Mr. McCormick said, some big banks are looking at these systems without making a prudent business case.

"We've received some RFPs (requests for proposals) which are driven by the IT side of the house that are undisciplined by business realities," he said. "Of course, that is goaded on by the technology vendors who would love you to put every piece of data into a data warehouse because they are going to make $10 million, $20 million, and $30 million over three years."

That kind of undisciplined approach, he said, is a formula for failure. Banks that will be successful in this arena must first think through exactly what business decisions they will need to make and what data and analysis are necessary to make them. Then they must build the capability to capture that data. "Way down the list," he said, "is the amount of effort necessary to build the warehouse once those other two things have been accomplished."

Bill Arnold, a principal in Towers Perrin's Atlanta office, agreed. "The bigger issue to me is not technology, it's people," he said. "Do they have the right competencies and capabilities in their people - both in the line and in executive management - to understand what's required and how they are going to use it? I have found few - not none but few - traditional bankers that really understand retailing and how to use this information."

Those views probably won't do much to quell analysts' concerns about excessive technology spending. "The banking industry has had a 10-year, stellar record of cutting costs, and we're happy about that," said Goldman Sachs' Mr. Albertson. "Now that that's been in place for so long, we don't want to see it upset by a sinkhole of spending in technology that may not be as good as people think."

A big question for analysts, consultants, and bankers alike is how effective data warehouses would be in generating revenue. And that can be tough to measure.

Mr. Arnold says one gauge of success is winning a bigger market share. But he concedes that measure can be misleading. The benefit to a bank that captured a larger portion of profitable customers in a region at the expense of unprofitable consumers wouldn't be reflected in a larger share of the market but in strong revenue growth.

Another test is looking at improvements in efficiency - particularly those that can be traced to the revenue, and not the cost, side of the equation.

First Manhattan's Mr. McCormick said the best way to judge success is evaluating the internal rate of return of a product marketing and sales strategy. He said he would want to know the effectiveness of sales efforts by phone, mail, and in the branch, exactly how much it costs the bank to originate an account, and expected annual revenue over five years.

"You want to be finding over and over and over again, whether that's originating new small-business or home equity loans, or whether that's selling checking accounts," he said.

Other lessons in prudent technology spending can be found by looking at banks that have a long track record of earnings growth and improved efficiency.

Fifth Third Bancorp, for example, recently spent nearly $5 million on three new processors for its transaction processing subsidiary.

George Landry, executive vice president at the unit, Midwest Payment Systems Inc., said the bank continued to operate its mainframe processing system efficiently while it waited for International Business Machines Corp. to develop a hardware platform that offered the benefits of client/server computing with the reliability of their existing systems. "It's almost like we were patient and didn't get on the first ship that sailed. And now we are really on the best ship," said Mr. Landry, who noted that the bank had never before bought a brand-new system from IBM. "Now we've kind of leapfrogged the whole industry." The new parallel enterprise server from IBM is expected to reduce energy expenses by 97%, floor space requirements by 94%, and maintenance costs by 65%.

Mr. Landry said the company waited because the interim alternatives did not offer the software capability the bank wanted. "The most expensive thing you have to watch out for is software and software development," he said.

The executive said the cautious approach reflects how Fifth Third makes technology decisions. "We generally don't react to some of the media trends."

In any event, consultants say banks will need to spend more in technology to stay competitive with both bank and nonbank competitors. The idea that investments may rise doesn't worry most. "We're about to see a new round of technology spending emerge, if it's not already here ... on creating customer marketplace information data bases and the real-time capability to profile customers and look at their needs, and life events, and purchase behaviors," said Mr. Arnold. "If I were an analyst I would be more concerned that a bank is not spending money on this kind of stuff than if it is."

Not everyone agrees. "I think they are spending too much," said Charles Callen, senior vice president and managing director of CSS Index in New York. "I think they've got to redirect themselves to some of the front-end technologies and find ways to ... improve the working relationship and the teaming with business users. Many banks have made great strides on that, and I think there is still a lot more distance to go there." The bottom line is that analysts remain perplexed by bank technology spending.

"We don't have a whole lot of information that is specific enough," said Mr. Albertson. "We are probably not as educated as we should be."

Subscribe Now

Access to authoritative analysis and perspective and our data-driven report series.

14-Day Free Trial

No credit card required. Complete access to articles, breaking news and industry data.