With the success of ATMs, banks had the incentive to develop new products and new delivery channels, such as home banking via phone and Personal computer ATM networks allowed banks to reach new customers outside the markets served by their branches and created the opportunity for greater price competition.

This past October, FleetBoston, newly renamed after the merger of Fleet Financial and BankBoston, took a jump into cyberspace. The firm announced plans to spend $100 million over 18 months to launch a high-powered new Internet site that will offer not only traditional banking services, but also stock trading, mutual funds, credit cards, mortgages, investment advice, financial news, and bill payment, all in one Web site. Blaise Heltai, managing director of Fleet's Internet strategy, told The Boston Globe, "We eventually want you to buy every financial services product or service you could ever need through us.” This most certainly is not your father's bank.

With this ambitious venture into cyberspace, Fleet joins a number of other banks in what is only the latest move in two decades of banking industry upheaval brought about by enormous advances in information technology. These advances have affected nearly all aspects of the business of banking.

In 1980, the banking industry consisted of a large number of relatively small firms operating in geographically distinct local markets. Products and services - primarily taking deposits and making loans - were delivered via the branch and the calling officer, which emphasized face-to-face contact with customers. These customers were, for the most part, relatively unsophisticated and trusted their bankers to act in their best interest. Twenty years later, with dramatic advancements in IT, banking customers have become increasingly savvy, making use of multiple distribution channels and demanding an ever-increasing variety of complex products. And competition has emerged from nontraditional quarters to take advantage of new technology and challenge old certainties.

What banks deliver and how they deliver it have changed dramatically. And these changes are likely not over yet. While no one can foresee the future, a look back over the last 20 years at the impact of information technology on banking may provide clues about what lies ahead as banks navigate into the twenty-first century.

THE IMPACT ON PRODUCTIVITY AND PROFITS

The past two decades have witnessed enormous reductions in the cost of information technology. Between 1986 and 1995, the computing power of the average PC increased elevenfold while the price declined. At the same time, a revolution in telecommunications reduced the cost of transmitting data by 90 percent since 1980. Such cost reductions have made it ever less expensive to acquire, store, transmit, and transform data into information. They have also created enormous changes in data-intensive industries such as financial services - which is, after all, fundamentally about processing information.

For commercial banking firms, these advances in IT have resulted in dramatic productivity gains. One early example was the introduction of the automatic teller machine (ATM), which first appeared in the United States in 1968. Most certainly, the introduction of ATMs made the distribution of some banking services more "efficient.” Before ATMs, withdrawing funds, account inquiries, and transferring funds between accounts all required face-to-face interaction between the customer and a bank teller. The bank's costs for these transactions included wages of tellers and back-office personnel, the cost of maintaining the premises, and other related expenses. ATMs automated this process and, to the extent that they were simply substituting a machine for a bank teller, costs per transaction fell significantly. The Wall Street Journal reported that a typical transaction by a teller costs between 90 cents and $2 per transaction, whereas the same transaction processed via an ATM costs only 40 cents.

Perhaps surprising, however, is that such productivity increases do not necessarily translate into overall cost reductions for banks. Why? Because advances in IT can do more than simply automate a banking activity. They also have "indirect” effects, both on consumer preferences and on the structure and competition of the banking industry. These indirect effects can alter banks' costs and revenues in a number of complicated and contradictory ways, with the end result on profits uncertain.

In the case of ATMs, as customers became comfortable with the new technology, they began demanding greater convenience and higher-quality products. But the costs of providing these new services were not necessarily below the costs of traditional bank accounts. Customers began making more frequent withdrawals which, in turn, forced banks to process an increasing number of transactions - potentially at significant cost. Soon customers decided that access to a single ATM at the bank branch was not enough; they wanted broader ATM accessibility. Banks responded either by investing in expensive ATM networks or by allowing their customers to have access to accounts via networks built by others.

Customers also began to demand more elaborate services from ATMs. The original machine was a simple cash dispenser; today banks can install sophisticated ATMs that scan checks, give out cash to the penny, let customers apply for loans, and allow for face-to-face discussion with a service representative via video. Thus, what started as a way to automate the services of a bank teller eventually developed into a new and improved delivery system for bank products. Yet, providing this system was costly, requiring a sizable investment in information technology and continued maintenance of sophisticated high-speed computer systems.

The impact of IT on revenues is similarly complicated. With better-quality products and services, banks should be able to charge more, all else equal. In the case of ATMs, the improved features and increased usage meant that banks might expect to receive increased fee revenue for processing customer transactions. But the proliferation of ATM networks also allowed banks to reach customers outside the geographic markets served by their branches. This created the opportunity for greater price competition, as consumers could choose the lowest-cost provider rather than a neighborhood bank. Online banking may have a similar effect on revenues. As people become comfortable shopping and applying for products such as mortgages and credit cards online, these products may turn into commodities, and reduce the profit margins that banks previously enjoyed. In the end, the impact on revenues depends on whether the higher prices associated with new and better products outweigh the lower prices that come with increased competition.

With the success of ATMs, banks had an incentive to develop new delivery channels, such as home banking via telephones and PCs. Debit cards, electronic check clearing, cash management, derivative securities, risk management, stored-value cards, and electronic forms of currency are also examples of products that are new, or newly reinvented, because of IT. Federal Reserve Economists Franklin Edwards and Frederic Mishkin find that the share of commercial bank income accounted for by activities other than interest on loans almost doubled between 1980 and 1994.

How has this played out to date? Allen Berger and Loretta Mester, also of the Federal Reserve, find that the banking industry's cost per unit of output has risen in recent years, but so has its profits. One way to interpret this result is that the quality of banking products has improved, but in a way that is hard to measure accurately. These new and better products may have raised costs, yet generated revenues greater than the cost increases.

CHOOSING A STRATEGIC DIRECTION

Banks that do not make investments that take advantage of new technology may find that they are losing customers to the better-quality or lower-cost products of firms that do. But using IT as a strategic weapon can be quite tricky, entailing high costs and an uncertain payoff.

Picking the right time is key. At first, customer resistance and the high price of new technology make investments risky. With ATMs, for example, there were early concerns about security and accuracy; and even today, many people are still reluctant to use ATMs to deposit checks. Competing technologies may exist, and an early commitment to the wrong one can doom a product or business as competitors and customers move on. As the price of IT drops and consumers become comfortable with the new technology, firms can invest with less concern about customer acceptance, although they may face an entrenched competitor who got there early.

With ATMs, banks divided themselves up and pursued several different strategies, according to Ralph Kimball, an economist at the Boston Fed, and William Gregor, senior VP of Gemini Consulting, in Cambridge, Massachusetts. Traditionalists offered only a few ATMs, primarily as an accommodation for customers needing to cash checks outside normal banking hours. Others positioned multiple machines at their branches as substitutes for tellers, encouraging customers to use the new, more cost-effective delivery channel. These banks did not attempt to reach past their branches by establishing an outside network. Still others, such as New England's BayBanks (now part of FleetBoston), invested early and aggressively in an extensive off-site network of ATMs as a way to extend their distribution and focus on customers with less need for face-to-face contact. It backed up this network with a 24-hour telephone center and a glossy catalog that detailed its products.

Each of these strategies had risks. The traditionalists risked losing customers willing to substitute technology for face-to-face contact. Innovators willing to invest heavily early on, like BayBanks, chose a high-risk and, as it turned out, high-return strategy that hinged upon customer willingness to adopt the new technology. Followers - those waiting to see whether consumers accepted the technology - risked being too late to grab the market share needed to survive.

Now, some of these same issues have resurfaced in online banking. With the cost of an online transaction at about 5 cents, just a fraction of an ATM or teller transaction, banks have been eager to convert customers. They started trying in the 1980s, when several banks invested millions to develop home banking products. But, unfortunately for the banks, consumers were not quite ready for the change and many of these products failed. Is the timing better today? According to Forrester Research Inc., of Cambridge, Massachusetts, the number of households that bank online will increase by over 350 percent in the next three years, from 3.7 million households to 13.7 million. The success of online brokerage companies, such as Charles Schwab and E*Trade, also suggests that more customers may now be willing to move their financial transactions online.

As with ATMs, banks are adopting several different online strategies. Some have no Web banking to speak of and continue to rely on branches and ATMs. Others are encouraging existing customers to switch to the Web for its cost advantages, with sites where customers can get balances and transfer funds between accounts. Still others, such as FleetBoston, Citigroup, and Wells Fargo , Co., are undertaking a bolder strategy. They have chosen to offer one-stop financial services on the Web, including real-time balances, corporate research, portfolio calculators, and stock trading. As with ATMs, these banks hope not only to lower costs, but also to increase revenues by providing a broader range of products to new and existing customers. In entering early, they may also be able to "lock in” customers who like the convenience of a one-stop site and later find unwinding the interlocking account and payment arrangements necessary to switch banks too much bother.

And a novel strategy is being pursued by a handful of Internet-only banks - banks with headquarters but with no bricks-and-mortar branches. They are hoping that, at least for some customers, banking will come to resemble the credit card industry, where all transactions are handled via the phone, mail, or electronically. But, although they keep costs low, Internet-only banks also face disadvantages, particularly in dispensing cash and accepting deposits, which still require an ATM.

Which strategies will pay off? BayBanks succeeded in luring new customers by placing its ATMs everywhere, then encouraging cardholders to use them with aggressive marketing and high-energy television and print ads. The investment paid off in increased market share, and its network became one of the most utilized in the country. By 1990, over 90 percent of BayBanks' customers carried cards, as compared to 65 percent nationally.

But, as with ATMs, banks' online strategies are not risk-free. Much will depend on customer acceptance. Will people want Internet banking? Will they prefer the convenience of one-stops or will they want to divide their business among specialty providers? So far, "most institutions have experienced the Internet as a money-losing proposition,” Chuck Farkas, managing director of Bain and Co.'s global financial services practice in Boston, told The Boston Globe. As for the future, only time will tell.

THE CHANGING BOUNDARIES OF THE FIRM AND THE INDUSTRY

The changes brought about by IT - new products, more sophisticated customers, changing cost structures, and enhanced competitive pressures - have all combined to transform the structure of the banking industry. And with further development of new technology, the industry will likely continue to evolve.

Advances in IT have surely changed the optimal size of a bank. Some technologically intensive products, such as processing payments, are more efficiently produced on a large scale; and the banking industry's recent wave of mergers and acquisitions suggests that bankers, at least, believe the "efficient” size of a bank has increased. The number of U.S. banking organizations fell from about 18,000 in 1985 to just under 10,400 in 1998, while the percentage of banking assets held by the largest 10 banks rose from about 25 percent to 35 percent over the same period. Although much of this consolidation was due to the elimination of restrictions on interstate banking and branching, much was also attributable to advances in IT.

However, the advantages of scale have not been felt equally across all banking products. Loans to businesses, for example, which tend to be specialized and handled on an individual basis, have shown less dramatic efficiency gains. Moreover, small competitors can often get the advantages associated with economies of scale by outsourcing some of their activities to specialists. A 1995 Brookings study suggests that, although consolidation will continue, 3,000 to 4,000 banking organizations should still be around a decade from now. According to this report, a few banks will be very large, but most will be relatively small.

Advances in IT have also resulted in new database technology and data-mining techniques that may expand the range of services that banks offer their customers. This technology allows firms to use customer information gathered in one part of their company, say banking, to increase sales in the others, say insurance or brokerage services, and is one of the factors driving recent industry consolidation. Large financial supermarkets, such as Citigroup, formed from the merger of Citibank, Travelers Insurance, and Salomon Smith Barney, are hoping to take advantage of the "cross-selling” opportunities created by IT. Federal legislators, concerned about privacy issues, are finalizing legislation that will put limits on the use of such data-mining techniques, especially on the sale of customer information to outside firms. But the days of a bank offering only traditional deposit accounts and making standard loans are likely over.

Advances in IT have opened up market niches for competitors from unexpected places. Many firms, not just banks, can now use statistical models to evaluate risk efficiently, originate loans, transform them into marketable securities, and sell them to obtain funding to make more loans. Countrywide Mortgage caused a radical change in the residential mortgage business when it equipped its salesforce with laptops and sent them into the field to take applications. This not only reduced the inconvenience for clients, but also reduced its own need for bricks-and-mortar facilities and for the data entry needed to process applications. Through its use of IT, Countrywide dramatically increased its market share, growing from a small start-up in 1969 to the largest U.S. independent residential mortgage lender and servicer in 1998.

Finally, information technology will likely continue to transform some banks into new types of financial institutions whose business bears little resemblance to that of a traditional bank. For example, State Street Bank in Boston is no longer a bank in the conventional sense; last October, it sold off its last bit of business taking deposits and making loans. State Street now relies on a very profitable IT-driven business, focusing on complex accounting and record-keeping activities for institutional investors, such as mutual funds. Other banks have taken on firms such as Amazon.com, using technological expertise to help smaller businesses build and manage online stores. For a fee, these banks will track inventory, generate shipping information, authorize customer payments, and even build the Web site. Although closely related to financial services, these activities are hardly traditional bank lines of business.

LOOKING FORWARD

As we look ahead to the next generation of information technology, it is doubtful that the current landscape - which types of firms offer which products, the degree to which physical proximity to the customer matters, the best size for firms - will remain unchanged. More likely, the types of products delivered and how they are delivered will continue to evolve. In a sea of change, it is also reasonable to expect some real surprises.

Today, with an estimated 190,000 ATMs already installed in the United States, the new frontier is on the Web. If online banking succeeds, it will almost certainly change the types of products that banks offer. But it is also likely to further break down the geographic advantage of local firms and intensify price competition. The overall impact on costs and revenues is hard to predict, even as the cost of an individual transaction on the Web drops.

From one point of view, IT can be said to have created havoc in the banking industry and to have placed in jeopardy huge investments in human and physical capital. Careers have been disrupted or abbreviated, and venerable institutions have been dismantled. Yet, because each problem also represents an opportunity, IT has provided benefits for those who are able to successfully adapt.

BANKS CHOOSE DIFFERENT STRATEGIES IN SITING THEIR ATMS Distribution strategies at large U.S. bank holding companies* In placing their ATMs, even the largest banks follow no single strategy. Some, such as First Union and National City, are relatively cautious, placing an average of one ATM at each branch, and relatively few outside of them. Chase Manhattan makes more effort to substitute ATMs for tellers, with an average of more than two ATMs per branch. Others, such as Bank One have built a large network of ATMs, with an average of five outside ATMs per branch.