ESSAYS ON ISSUES THE FEDERAL RESERVE BANK MARCH 2001OF CHICAGO NUMBER 163

Chicago Fed Letter

The Internets place in thebanking industryHow the Internet will affect bankingis one of the most intriguing questionsin the ongoing evolution of the U.S.banking industry. Internet bankinggives customers the ability to accessvirtually any type of banking service(the main exception for now beingcash) in any place and at any time. Ifcustomers adopt this new way of bank-ing in large numbers, banks may beable to shed much of their investmentin expensive brick and mortar branch-es. But Internet banking remains awork in progress, and for many U.S.banks the initial Internet experiencehas been disappointing.

In this Chicago Fed Letter, I argue thatthe Internet is chiefly a new deliverychannelnot a new productandbased on this argument, I propose asimple framework for analyzing thestrategic interactions between physicalbranches, automated teller machines(ATMs), the Internet, and otherbank delivery channels. For mostbanks, the future of the Internet liesin how well it can be integrated withmore traditional delivery channels.But in the end, profitability will de-pend primarily on the quality of theproducts and services banks deliverto their customers, and not necessarilyon how those products and servicesare delivered.

Changing delivery channels

The way that U.S. commercial banksdeliver products and services to theircustomers has changed substantiallyover the past decade (figure 1). Bankmergersmany of which combinedtwo banks from different geographicmarketsreduced the total numberof banks by about one-third duringthe 1990s. But despite having fewer

banks, the U.S. nowhas more bankingpoints of sale than adecade ago. The num-ber of branch locationshas increased fromabout 60,000 to about70,000, and the num-ber of ATMs has sky-rocketed. The typicalbank now has a greatergeographic reach, andcovers those marketswith a denser networkof branches and ATMs.

More recently, bankshave augmented theirdistribution networkswith transactional web-sites, which allow customers to openaccounts, apply for loans, check bal-ances, transfer funds, and make andreceive payments over the Internet.The number of banks with transac-tional websites is increasing rapidlyfrom near zero just a few years ago,to 1,100 at year-end 1999, to an esti-mated 2,000 plus by the end of 2001.And the recent introduction of wire-less Internet banking promises tofurther increase the convenience ofWeb-based banking.1

The Internet is also transformingtraditional bank distribution channels.For example, the recent increase inATMs includes the introduction ofautomated banking machines (ABMs).Often deployed at banking kiosks,ABMs combine at a single locationan ATM for getting cash and deposit-ing checks, an Internet connectionto the banks website, and often a tele-phone for accessing customer service.Similarly, the increase in bank branch-es over the past decade includes theintroduction of mini-branches, inwhich Internet kiosks are placed side-by-side with teller windows.

New product or new package?

When a retailer like Eddie Bauer sellsa pair of jeans, the point of sale mightbe a physical store, a telephone order,or an Internet purchase. Regardless,the customers choice of a deliverychannel does not affect the natureof the product. This analysis can beapplied to most banking services, re-gardless of whether the point of saleis a physical branch, an ATM or ABM,or the Web. With a few exceptions, atransactional Internet website is nota new financial productrather, it isa new delivery channel for existingfinancial products.

In some ways, the introduction of theInternet banking channel parallelsthe introduction of ATMs several de-cades ago. ATMs did not introduceany new financial services, but theyoffered customers more convenientaccess to a limited array of existingfinancial services, primarily the safe-keeping of deposits, liquidity services,and information on account balances.Like ATMs, Internet banking (support-ed by other developments like creditscoring technology, check imaging,

number

1. Bank delivery channels, 199199

Banks(right scale)

1991 93 95 97 990

50,000

100,000

150,000

200,000

250,000

0

2,500

5,000

7,500

10,000

12,500number

Transactional websites(right scale)

ATMs(left scale)

Branches(left scale)

Sources: Data on ATMs from American Bankers Association,Bank Network News, Ernst & Young, and Dove Associates;banks and branches from FDIC; transactional websites from OCC.

and check truncation) has increasedthe convenience of accessing an evenwider array of existing banking services.

However, some of the financial servic-es that banks offer over the Internetare new. For example, some banks areusing the Internet to offer accountaggregation, which organizes in oneplace all the data from a customersmultiple relationships with banks, in-surance companies, and brokeragefirms. (Prior to financial deregulation,customers tended to have relationshipswith fewer financial institutions, so ac-count aggregation was less necessary.And prior to the Internet, the logisticsof collecting data and mailing it tocustomers made this a less cost-effec-tive service.) Another example is thebusiness-to-business marketplace, wherebanks use the Internet to bring togeth-er prospective buyers and sellers ofstandardized business inputs (e.g.,chemicals or paper products). If thesemarkets are constructed efficiently,buyers and sellers benefit from betterprices and more timely delivery, andbanks can benefit by providing financ-ing for the deals that result.

Choosing a distribution strategy

Not all banking products, and not allbanking customers, adapt well to theInternet channel. Transferring funds,paying bills, and applying for a creditcard do not require personal con-tact or a large physical space, andare therefore well suited for Internetdelivery. But applying for a businessloan, closing on a home mortgage,and estate planning are complextransactions, which typically requirea secure physical space and/or person-to-person communication. And gettingcash is impossible over the Internet,requiring either branches or ATMs.Because of such limitations, mostbanks that offer Internet delivery donot rely on it entirely.

The mix of delivery channels a bankchooses has consequences for its ex-penses, the convenience of its custom-ers, and the quality of the productsand services it delivers. Figure 2 cate-gorizes bank delivery channels accord-ing to the distance that customerstypically must travel to use them

(vertical axis) and theamount of in-personservice that customersreceive (horizontal axis).

As a banks mix of de-livery channels shiftsvertically from the topof the figure toward thebottom, there are ben-efits for both the cus-tomer and the bank:convenience increasesbecause customersdont have to travel asfar to perform transac-tions, and bank expens-es tend to fall becauseless physical overheadis necessary to facilitate the transac-tion. According to some recent esti-mates, branch banking costs about$1.07 per transaction, telephonebanking costs about $0.55 per transac-tion, ATM banking costs about $0.27per transaction, and Internet bankingcosts about $0.01 per transaction.2But there is a tradeoff: Shifting to amore convenient, lower cost mix ofdelivery channels also tends to reduceperson-to-person contact with the cus-tomer. As a banks mix of deliverychannels shifts horizontally from rightto left in figure 2, some customerswill experience a reduction in (eitheractual or perceived) service quality.

Note that the data displayed in figure 1indicate that the mix of bank deliverychannels has been shifting from thetop right corner of figure 2 towardthe bottom left corner of figure 2.Does the resulting increase in cus-tomer convenience offset the declinein service quality? This shift may ormay not be a profitable move for anygiven bank, depending on the natureof the financial services it sells, thepreferences of its customers, and theamount of cost savings from the newdistribution strategy.

One potentially successful distribu-tion strategy is to occupy the entirespace in figure 2. A click and mortarbank augments its existing brick andmortar branches, ATM locations, andother delivery channels with a trans-actional Internet website. This ap-proach arguably avoids the tradeoff

between customer convenience andin-person quality by allowing custom-ers to choose the mix of delivery chan-nels that works best for them. Theclick and mortar strategy has beenadopted by all of the largest U.S.banks. An increasing number of fullservice community banks are also im-plementing this strategy, chiefly as adefensive move aimed at retaininghigh-value customers who want to usethe Internet for some of their bank-ing transactions.

Another potentially successful strategyis to occupy only the bottom left cor-ner of figure 2. An Internet-only or pureplay Internet bank operates no brickand mortar branches. With the excep-tion of arrangements for customers toget cash and deposit checks at ATMmachines, banks using this distributionstrategy deliver all of their productsand services over the Internet. Thevery nature of this delivery channelprecludes person-to-person customerservice, and although this can limitthe ability of a pure play Internet bankto charge premium prices, reducedspending on physical overhead maypotentially offset these revenue limita-tions. Internet-only banking is often re-garded as a niche strategy that focusesonly on the most Internet-savvy bank-ing customers and/or delivers only alimited array of financial services.

A final strategy is to occupy only thetop right corner of figure 2. A brickand mortar bank does not operate atransactional website, but may operate

2. Bank delivery channels: A set of choices

Unit bank

Branch bankDrive-through

ATM

TelephoneInternet

Wireless

Person-to person contact MoreLess

Clos

erFa

rther

s

s

s

s

Dis

tanc

e to

cus

tom

er

Michael H. Moskow, President; William C. Hunter,Senior Vice President and Director of Research; DouglasEvanoff, Vice President, financial studies; CharlesEvans, Vice President, macroeconomic policy research;Daniel Sullivan, Vice President, microeconomic policyresearch; William Testa, Vice President, regionalprograms and Economics Editor; Helen OD. Koshy,Editor; Kathryn Moran, Associate Editor.

Chicago Fed Letter is published monthly by theResearch Department of the Federal ReserveBank of Chicago. The views expressed are theauthors and are not necessarily those of theFederal Reserve Bank of Chicago or the FederalReserve System. Articles may be reprinted if thesource is credited and the Research Department isprovided with copies of the reprints.

Chicago Fed Letter is available without charge fromthe Public Information Center, Federal ReserveBank of Chicago, P.O. Box 834, Chicago, Illinois60690-0834, tel. 312-322-5111 or fax 312-322-5515.Chicago Fed Letter and other Bank publications areavailable on the World Wide Web at http://www.frbchi.org.

ISSN 0895-0164

a non-transactional website where cus-tomers can check account balancesand get information on products andprices. Banks that use this distributionstrategy deliver all of their productsand services through traditional fullservice branches, augmented by ATMmachines. Although this traditionalapproach is likely to remain a profit-able strategy for some communitybanks into the near future, any strat-egy that completely excludes Internetbanking options is unlikely to be prof-itable in the long run. As time passesand a greater percentage of the pop-ulation want to do at least some oftheir banking on the Web, these banksare likely to lose an increasing num-ber of their high-value loan and de-posit customers.

Is Internet banking profitable so far?

Just a few years ago, pundits werepredicting that the Internet channelwould soon eclipse brick and mortarbranches, and that Internet-only bankswould quickly capture a large shareof the banking market. More recently,these predictions have swung like apendulum. Indeed, some analysts nowargue that pure play Internet bankingis a flawed business model.3

The reality probably lies somewherebetween these two extreme positions.To date, only a handful of serious stud-ies have examined the performanceof the Internet banking channel.Not surprisingly, the assessments ofthese studies tend to be less extremethan conclusions drawn in the finan-cial press.

Two of these studiesone performedat the Office of the Comptroller ofthe Currency, the other at the FederalReserve Bank of Kansas Citycomparethe performance of click and mortarbanks with that of brick and mortarbanks.4 These two studies find a num-ber of similar results. Most important-ly, they find that profitability at theInternet banks tends to be higherthan, or is at least comparable to, prof-itability at the more traditional banks.While the direction of causation inthese studies is not completely clearfor example, it may be that well-man-aged, profitable banks are more likely

to start up transactional websitesthese studies suggest that the Internetdelivery channel can be part of aprofitable banking strategy.

A third study, performed at theFederal Reserve Bank of Chicago,compares the performance ofInternet-only banks and thrifts withthat of banks and thrifts that operatebranches (after controlling for anumber of outside factors).5 Thestudy finds relatively low profits atthe Internet-only institutions, causedin part by high labor costs, low fee-based revenues, and difficulty gener-ating deposit funding. However,rather than concluding that Internet-only banks are necessarily unprofit-able, the study stresses that it maysimply be too soon to judge this busi-ness modelboth Internet-onlybanks and their customers are stilllearning how to efficiently use thisdelivery channel, and overall demandfor Internet-only banking is likelyto grow.

The Internets (eventual) placein banking

Although Internet-only banks mayeventually become profitable, evi-dence is mounting that banks usingthis business model are unlikely tocapture a dominant share of thefull-service banking market. A grow-ing number of Internet-only banksare specializing in niche productmarkets or customer groups. For ex-ample, iVantage Bancorp focuses oncollege students and their parents;UmbrellaBank.com attempts to buildlong-term relationships with tradi-tionally unbanked consumers;AeroBank.com concentrates on sell-ing loan and cash management ser-vices to small business owners; BMWBank cultivates an upscale customerbase at BMW auto dealerships; andState Farm Bank markets onlinebanking services through State Farminsurance agents. Meanwhile, a num-ber of large banking companies thatlaunched high-profile Internet-onlyventuresincluding Wingspan (BankOne), mbanx (Bank of Montreal),and Citi f/i (Citigroup)have beenintegrating these ventures back into

the main bank, giving their Internetcustomers full access to their branchdistribution networks. Similarly, RoyalBank of Canada purchased a Chicago-based mortgage company with 150branch offices so that customers of itsU.S. Internet-only bank, Security FirstNetwork Bank, could access bankingservices at brick and mortar locations.

At the other extreme, it seems evenless likely that traditional brick andmortar banks will retain a large marketshare in the long run without offeringtheir customers an Internet bankingoption. Today, it is difficult to imaginea successful bank that operates with-out ATMs. In the near future, it maybe just as difficult to imagine a suc-cessful bank that operates without atransactional website.

These developments suggest that themajority of Internet banking custom-ers will be served by click and mortarbanks, not by Internet-only banks. Fig-ure 3 maps out a hypothetical futuredistribution of bank delivery channels:a handful of pure play Internet banksat one extreme, virtually no brick andmortar banks at the other extreme, anda continuum of click and mortar banksin the middle. The percentage of busi-ness that any given click and mortarbank delivers over the Internet chan-nel is likely to be determined by the

mix of products it offers and the pref-erences of the customers it serves.

Conclusion

For most banks, and for most of theircustomers, banking over the Internetis still a relatively new phenomenon.Because the pace of technologicalchange is so fast, it can be difficult

to evaluate the strategic importanceand the financial impact of Internetbanking. This Fed Letter argues thatthe Internet, much like the ATM thatcame before it, is fundamentally a newdistribution channel over which bankscan deliver traditional banking prod-ucts and services. Banks that success-fully integrate this new channel withtheir preexisting branch and ATMnetworks, choosing the mix of chan-nels that best complements their prod-uct mixes and customer bases, willgain a strategic advantage. But thebusiness of banking remains the pro-vision of credit, safekeeping, transac-tions, insurance, and investmentservicesbanks that are unable to pro-vide these services efficiently in an in-creasingly competitive environmentwill not flourish, regardless of thedelivery channels they use.

Robert DeYoungSenior economist and economic advisor

nu

mbe

r of b

anks

3. Hypothetical distribution of banks

percent of business over Internet

1Currently, wireless devices are used most oftenfor a limited array of brokerage (e.g., monitoringfinancial markets, executing trades) and banking(e.g., transferring funds, checking accountbalances) transactions.

2See Luxman Nathan, 1999, Community banksare going online, Communities and Banking, Fed-eral Reserve Bank of Boston, Fall, No. 27, pp.28. Also see The Economist Newspaper Limited,2000, Branching out, The Economist: A Survey ofOnline Finance, May 20, pp. 1923.

3For example, see Dow Jones & Company, 2001,Online banks fail to realize cyber-goals, WallStreet Journal, January 10, p. C18.

4The first of these studies looks at national banks.See Karen Furst, William W. Lang, and Daniel E.Nolle, 2000, Who offers Internet banking,Quarterly Journal, Office of the Comptroller ofthe Currency, Vol. 19, No. 1, June, pp. 121.The second looks at banks in the Tenth FederalReserve District. See Richard J. Sullivan, 2000,How has the adoption of Internet banking affect-ed performance and risk in banks?, FinancialIndustry Perspectives, Federal Reserve Bank ofKansas City, December, pp. 116.

5Robert DeYoung, 2001, The financial perfor-mance of pure play Internet banks, EconomicPerspectives, Federal Reserve Bank of Chicago,Vol. 25, No. 1, First Quarter, pp. 6076.

Public Information Center

P.O. Box 834

Chicago, Illinois 60690-0834

(312) 322-5111

Return service requested

FEDERAL RESERVE BANK OF CHICAGO

Chicago Fed LetterPRESORTED

FIRST CLASS MAIL

US POSTAGE PAID

CHICAGO, IL

PERMIT 1942

0 20 40 60 80 100