Journal of Financial Services Research 22:1/2 5±27, 2002

# 2002 Kluwer Academic Publishers. Manufactured in The Netherlands.

E-Finance: An Introduction
FRANKLIN ALLEN
The Wharton School, University of Pennsylvania

JAMES MCANDREWS*
Federal Reserve Bank of New York

PHILIP STRAHAN
Boston College

Abstract

E-®nance is de®ned as ``The provision of ®nancial services and markets using electronic communication and
computation''. In this paper we outline research issues related to e-®nance that we believe set the stage for further
work in this ®eld. Three areas are focused on. These are the use of electronic payments systems, the operations of
®nancial services ®rms and the operation of ®nancial markets. A number of research issues are raised. For
example, is the widespread use of paper-based checks ef®cient? Will the ®nancial services industry be
fundamentally changed by the advent of the Internet? Why have there been such large differences in changes to
market microstructure across different ®nancial markets?

Key words: E-®nance, internet, research.

1. Introduction

How important are electronic communication technologies and the Internet for ®nance? Is
the Internet the most signi®cant development for many decades or is it just one among
many? One view is that it will fundamentally transform the ®nancial services industry and
®nancial markets. Another is that the net represents the latest in a long line of electronic
technologies that have reshaped the ®nancial industry. These issues were addressed at a
conference held at the Federal Reserve Bank of New York on February 23 and 24, 2001.
This volume contains papers presented there. The purpose of this article is to review these,
frame the issues related to the use of the Internet and other types of electronic
communication technologies in ®nance, and stimulate future research. We attempt to draw
out the many interesting questions to be answered below.
What exactly is e-®nance? The de®nition that is used as the basis of this article is

*Corresponding author. 33 Liberty Street, New York, NY10045, U.S.A. E-mail: [email protected].
The views expressed in this paper are those of the author and do not necessarily re¯ect the views of the Federal
Reserve Bank of New York or of the Federal Reserve System.
6 ALLEN ET AL.

The provision of ®nancial services and markets using electronic communication and
computation.

Claessens et al. (this volume) provide an overview of the state of e-®nance around the
world. The developments can be divided into two broad areas. The ®rst is the impact on
banking and ®nancial services. They argue that the advent of the Internet and other
electronic communication means has fundamentally altered many aspects of the banking
industry. Many of the services traditionally provided by banks are being provided by other
entities. The second broad area is the transformation of ®nancial markets. These no longer
need to be associated with a physical place. As a result, trading systems for equities, bonds,
and foreign exchange are becoming global. All these changes have important signi®cance
for public policy towards the ®nancial services industry and ®nancial markets. They
consider the implications for safety and soundness regulation, competition policy,
consumer and investor protection, and global public policy.
Three important trends in the ®nancial services industry have been accelerated by the
emergence of the Internet. These are improved price transparency, differential pricing, and
transformation of distribution channels. Clemons et al. (this volume) consider the impact
of these trends on the ®nancial services industry. Improved price transparency potentially
increases competition and reduces pro®t margins. Evidence from other industries that
make extensive use of the Internet, however, suggests that there are limits to this process.
It appears that transaction costs of search remain suf®ciently high that differential pricing
is possible and this will become increasingly important in ®nancial services. As increased
use of the Internet leads to the unbundling of services and promotes disintermediation,
there will be a transformation of distribution channels and an important restructuring of the
industry.
Section 2 of our overview considers broadly the impact of e-®nance technologies on
®nancial services ®rms. Online banking began in the mid-1990s and is steadily becoming
more important. In contrast to banks, however, insurance companies have used the Internet
a relatively small amount. Electronic brokerage services have been an important
development in recent years. Firms such as Schwab and E* Trade have developed
signi®cant discount brokerage businesses over the net. While the Internet is new and
exciting, we discuss how other e-®nance technologies, many of which have been around
longer than the Internet, have also reshaped the ®nancial services industry, particularly the
banking sector. For example, starting in the 1970s, automated teller machines (ATMs)
began to alter the ways that consumers interact with banks. In the 1980s, greater use of
electronic computation and data analysis changed the way credit decisions are made; these
innovations have both reduced the cost of lending on average and enhanced the liquidity
and marketability of loans.
The use of electronic communication in ®nance, in fact, goes back much further than the
1970s. As long ago as 1918, the Fedwire payment system allowed electronic settlement of
payments between banks over the telegraph. This use of electronic communications in
payments systems has steadily increased over time. Now virtually all large payments
between banks and corporations are done electronically. In some countries, such as those
in Scandinavia, electronic payments systems are becoming increasingly widely used at the
consumer level. In the United States, however, the paper-based check clearing system still
E-FINANCE: AN INTRODUCTION 7

predominates. In section 3, we discuss in some detail how e-®nance is changing the

payments systems. In our view, insuf®cient attention has been paid to these developments.
Our purpose in this section is to highlight the many interesting theoretical and policy
questions surrounding the evolution of payments systems and, we hope, stimulate further
research.
We consider the role of electronic communications in the securities markets in section 4.
Nasdaq was an early example of a stock market that was based on electronic
communication means. Many others around the world have followed suit and now the
New York Stock Exchange remains one of the few stock markets that retains traditional
means of trading. The foreign exchange market was a telephone-based dealer market for
many years. Recently it has increasingly migrated to the Internet and this trend is expected
to continue. Finally, the bond market has also traditionally been a dealer market. Unlike
the stock market or the market for foreign exchange, however, the bond market has not
been transformed by the advent of electronic trading systems.
Concluding remarks are contained in section 5.

2. E-Finance and the ®nancial services industry

This section begins by describing how different kinds of ®nancial services ®rms,
depository institutions, insurance companies, and securities companies, have deployed
e-®nance technologies. Then, we discuss how these new technologies have reshaped the
role and structure of the ®nancial services sector. We attempt throughout the section to
draw out the research questions and policy issues raised by the advent of new e-®nance
technologies.

2.1. Adoption of e-®nance by ®nancial services ®rms

2.1.1. Financial intermediaries. By the end of the 1990s, e-®nance technologies had
arguably affected all aspects of the business of banking and ®nancial intermediation,
with the possible exception of lending to large businesses. Depository institutions have
used electronic information technologies, for example, to make credit decisions to con-
sumers since the 1980s. Having computation abilities that allow the use of large
databases has made this possible. For consumers, the application and approval process
for both mortgages and credit cards has become suf®ciently automated so that it can be
done without any personal contact with the lender. The ability to make these credit
decisions in this way depends crucially on standardized information provided to lenders
by a small number of credit bureaus that keep track of individuals' credit histories.
With respect to credit cards, their use as a medium to make payments, along with debit
cards, has grown dramatically, fueled by rapid communications technologies that allow
vendors to validate a person's credit worthiness in seconds.
In recent years, information technologies have also been used systematically for lending
to small businesses. Since the early 1990s, banks and ®nance companies have been using
credit scoring models to lend to small businesses on a wide scale. These models use
8 ALLEN ET AL.

information about borrower quality, such as the credit history of the proprietor, to estimate
the likelihood that a particular small business loan will default; loan applications with a
suf®ciently low default likelihood (high ``score'') are granted. Survey evidence suggests
that large banks have been the ®rst to use credit scoring for their small business loans, and
that these models are generally used only for very small, small business loans such as those
under $100,000 (Mester, 1997). Smaller banks have also gotten access to credit scoring
technologies, however, through the efforts of third parties such as Fair Isaac. These
companies aggregate information on the performance of loans to many small businesses
(and other borrowers) to produce generic credit scores that can be used by any lender, large
or small. In this way, these third parties allow smaller ®nancial institutions to take
advantage of a technology that comes with very substantial scale economies.
The widespread use of these information technologies have made it increasingly easy
for banks to remain at ``arm's length'' from their borrowers. Over the past two decades,
for example, the probability that a bank communicates with its small business borrower in
person rather than with the phone or mail has declined from 59% in 1973 to 36% in 1993.
Moreover, the average geographical distance from a bank to its typical small business
borrower has increased from 16 miles during the 1973±1979 period to 68 miles during the
1990±1993 period (Petersen and Rajan, 2001).
Surveys by the Federal banking regulators suggest that depository institutions in very
recent years have also invested heavily in e-®nance technologies (e.g., website on the
Internet) as a means to distribute their products to retail customer. As described by Furst et
al. (this volume), by 1999 banks with transactional websitesÐwebsites capable of
allowing customers to conduct businessÐheld more than 90% of banking assets. Banks
have used the Internet as an alternative means to distribute their traditional credit and
payments products to retail customers. For example, Furst et al. ®nd that most Internet
banks offer balance inquiry and funds transfer, bill payment and credit applications online.
In contrast, fewer banks with transactional website offer brokerage, cash management,
®duciary, or insurance services online, although this may re¯ect the fact that fewer banks
provide these services.
The rapid increase in the use of the Internet represents a continuation of banks' efforts to
replace their costly branch network with alternative distribution channels such as the
telephones, the mail and ATMs (Radecki et al., 1997). Although ATMs have existed since
the middle of the 1970s, as discussed in section 2 their numbers have grown dramatically
in recent times. Having said that, the number of of®ces at banks and savings institutions
continues to grow, albeit slowly. Over the second half of the 1990s, for example, the total
number of of®ces at depositories rose from about 80,000 in 1995 to a little more than
85,000 in 2000.1
Furst et al. also found that large banks adopted these technologies very aggressively
during the late 1990s, and that these technologies have been associated with relatively high
pro®ts and ef®ciency. An open question, however, is whether adoption of the Internet (and
other e-®nance technologies) has itself enhanced pro®ts, or whether the more ef®cient and
higher pro®ts banks were simply ®rst to adopt the technology. They report that de novo

1 Figures are based on statistics reported on the FDIC's website, http://www2.fdic.gov/hsob.

E-FINANCE: AN INTRODUCTION 9

banks that attempt to use websites as their primary means for interacting with customers
have not been successful to date, suggesting that e-®nance technologies cannot supplant
traditional distribution channels entirely. This suggests a slightly broader question: are
banks using the Internet to enhance their existing delivery channels (``bricks and clicks'')
or to supplant them (``bricks or clicks'')? In other words, are branches and websites
substitutes or complements?

2.1.2. Insurance. Like the securities underwriting business, the insurance industry
also has not seemed to adopt e-®nance technologies in large numbers yet. For example,
in 1999 12% of U.S. insurers used the Internet to sell products amounting to 1% of
the total sales (Banks, 2001). Insurance companies tend to buy and hold securities
(mainly debt securities) issued or originated by someone else, thus limiting the poten-
tial usefulness of e-®nance on the asset side of the business. On the liability side, slow
adoption of e-®nance technologies may re¯ect the relative infrequency with which an
insurer and retail customer interact. In contrast to bank depositors, policyholders gener-
ally interact with their insurer only at the time of sale and when ®ling a claim. Also in
contrast to banking services, insurance policies tend to be quite heterogeneous and
consumers tend not to be well informed about the products, making automation of the
sales process dif®cult. Differences in regulation may also play a role. The issue of
why the insurance industry has been relatively slow to use the Internet needs further
study.

2.1.3. Securities ®rms. E-®nance has also affected the securities industry, particularly
for the broker±dealers in the secondary securities markets. During the 1990s, the dis-
count brokerage business rapidly gained market share. Discounters typically rely on
fees from retail customers with few ancillary services, thereby allowing individuals to
trade securities at very low cost. From 1990±1999, employment at discount brokers
rose from 2.4±5.3% of total employment in the securities industry. Moreover, the dis-
counters enjoyed a higher pro®t rate ( pre-tax return on equity) than national full-line
securities ®rms or large investment banks in every year from 1990±1999 (Securities
Industry Fact Book, 2000).
Will e-®nance technologies reshape the primary securities markets to the same extent
that it has affected the brokerage business? The traditional approach to bringing a new
security to the primary market involves ``book building,'' a process by which investment
banks assess the demand for a security issuance from a small number of well-connected
institutional investors before ®nalizing the terms of the offering. Wilhelm (1999) argues
that the relationship-based approach to book building has allowed information to be
collected ef®ciently in an environment of high search and information costs. By lowering
these search and information costs, however, e-®nance technologies may reduce the
comparative advantage of the relationship-based approach relative to a more arm's length
process, such as an online auction.
Some online IPOs have succeeded over the past ®ve years, beginning with the Spring
Street Brewery in 1995 (Banks, 2001). More than 3500 people bought shares in the $1.6
million offering over the Internet. The shares were subsequently traded on an Internet
bulletin board called Wit-trade. The SEC investigated the primary issue market and the
10 ALLEN ET AL.

secondary trading and issued a no-action letter. In 1999, W.R. Hambrecht created an online
auction (OpenIPO) available to any individual or institutional investor that has brought
several companies public. OpenIPO operates a Dutch auction in which participants pay the
minimum price necessary to sell all of the shares offered. According to the Hambrecht
website, six companies have used the OpenIPO format successfully since 1999.2 Whether
the online auction model of securities underwriting represents a real threat to traditional
securities underwriting methods remains to be seen. Many studies have found that
securities are underpriced when ®rst sold to the public, suggesting that issuers sacri®ce
substantial amounts of capital by using the traditional underwriting approach (Loughran et
al., 1995). Thus, an important question for future empirical research is whether securities
sold in an open auction yield less underpricing than securities sold using the traditional,
relationship-based technology.

2.2. The effects of e-®nance on the ®nancial services sector

2.2.1. Disintermediation. The advent of e-®nance technologies furthers the long-

standing evolution of the ®nancial services sector from one dominated by ®nancial
intermediaries to one dominated by capital markets and ®nancial institutions that hold
marketable securities as assets. Traditional ®nancial intermediaries transform illiquid,
hence non-marketable, assets (e.g., bank loans) into liquid liabilities (e.g., demand
deposits). This role has become less important over time as the liquidity of ®nancial
assets originated by intermediaries has increased. Liquidity depends crucially on the
ability of buyers and sellers to agree on the value of a ®nancial assets. Liquidity is
reduced by asymmetric informationÐsellers knowing more than buyersÐbecause
buyers assume, rationally, that informed sellers want to keep high quality assets and
sell low quality ones. E-®nance technologies reduce asymmetric information because
they lower the costs of communication, computation and data processing, thus
allowing buyers and sellers of ®nancial assets to have more equal access to informa-
tion.
Securitization of small business loans offers a concrete example of how e-®nance
technologies enhance the liquidity of a formerly illiquid class of ®nancial assets. Small
business lending had traditionally been based on information developed from a long-term
relationship between lender and borrower. As a result, outsider investors had no reliable
way to evaluate these loans and they remained illiquid. Over the past decade, however,
banks and ®nance companies have begun to replace relationship-based lending with an
automated credit scoring system that allows outside ratings agencies to evaluate the credit
risks inherent in a pool of these loans. According to a recent Federal Reserve Board Report
on Congress (2000), more than 300 large small business lenders use a common credit
scoring system developed by Fair Isaac. Because pools of small business loans can now be

2 They are Briazz, Inc., Peet's Coffee, Nogatech, Andover.net, Salon.com, and Ravenswood. See http://
www.wrhambrecht.com/offerings/completed.html.
E-FINANCE: AN INTRODUCTION 11

standardized and rated, securitization of small business loans have been able to occur,
growing tenfold between 1995 and 1999, from $241 million to $2.3 billion.3
While the growth rates are impressive, securitization of small business loans has
remained a small part of this market. In lending to consumers, however, securitization has
become widespread. By the end of the 1990s, over half of residential mortgages were
being securitized, about 45% of credit card loans were securitized, and about 10% of other
consumer loans were securitized (Mishkin and Strahan, 1999). Thus, an interesting
question for research is whether securitization will continue to spread. Will securitization
of small business loans become as pervasive as in mortgage and consumer lending? What
about lending to large businesses? And, how important are systematic data processing
technologies in reducing the adverse selection problems that make assets illiquid and hard
to securitize?
As a result of the increasing liquidity of ®nancial assets, depository institutions have lost
market share to mutual funds and pension funds. Between 1980 and 1998, for example, the
share of all ®nancial institution assets held by depositories fell from 58% to 31%, while the
share held by mutual funds and pension funds rose from 21±49% (Mishkin and Strahan,
1999).4 Moreover, the decline in the market share of intermediaries arguably understates
the decline of traditional ®nancial intermediation because an increasing share of the assets
that remain on their balance sheets could be sold or securitized. This change re¯ects
disintermediation; assets have migrated from intermediaries that provide asset
transformation services, to ®nancial institutions that provide very little or no asset
transformation. (Mutual funds and pension funds hold mainly marketable securities as
assetsÐcommercial paper, corporate debt, mortage-backed securities, government
securities, etc.) As this discussion suggests, traditional intermediation has been declining
for many years. An interesting research question looking ahead is whether e-®nance
technologies will push this process along, or whether banks and other intermediaries can
use these technologies to encourage re-intermediation, as argued by Domowitz (this
volume).
How will disintermediation affect monetary policy? By weakening the link between the
liabilities and assets in the banking system, disintermediation creates a challenge for
central bankers. The central bank has direct in¯uence only over the liabilities of banks and
other depository institutions, but innovations such as securitization allow banks to
continue to originate loans even if funding becomes scarce. Monetary policy is thought to
operate both through its effects on interest rates and through its effects on lending by
banks.5 As bank lending becomes less important and less tied to their liabilities, however,
this second ``credit channel'' of monetary policy may lose its potency (Lown and Morgan,
2001). Estrella (2001) in fact shows that securitization in mortgage markets has reduced

3 These ®gures include conventional small business loans and the unguaranteed portion of loans backed by
the the Small Business Administration. See Federal Reserve Board Report on Congress (2000).
4 The market share of insurance companies and other ®nancial institutions has remained relatively constant
over this period.
5 There is a large literature on the channels of monetary policy that we will not cite here. See, for example,
Bernanke and Blinder (1992) and Gertler and Gilchrist (1994).
12 ALLEN ET AL.

the effects of a monetary policy shock and, at the same time, altered the channels through
which those shocks operate. At a minimum, disintermediation complicates the life of the
central banker by increasing the already substantial degree of uncertainty inherent in the
job.
Another important question is whether disintermediation will change the way the
®nancial system allows agents to share risks. Allen and Gale (1997) argue that bank-based
®nancial systems eliminate risk through inter-temporal smoothing. Banks are able to build
up reserves in good times and run them down in bad times. Disintermediation prevents this
from happening because assets will be marked to the market and the current owners will
obtain the full increase in value. The extent to which e-®nance leads to this type of
disintermediation remains an empirical issue.

2.2.2. Consolidation. Consolidation in the banking sector has gone hand-in-hand with
disintermediation, perhaps re¯ecting the joint effects of e-®nance. At the same time that
it lowers transactions and information costs, e-®nance technologies raise scale econo-
mies. On the liability side, for example, electronic payments technologies require large
®xed investments and often require networks that exhibit increasing returns. On the
asset side, for example, credit scoring models rely on statistical analysis of default risks
which perform better with larger databases, thus giving large lenders a signi®cant
advantage over their smaller competitors. Studies using the experience of the 1980s
typically found very limited economies of scale in the banking industry, but more
recent studies using data from the 1990s suggest scale economies up to $10±25 billion
in assets.6 Although data limitations make it dif®cult to study effects of e-®nance tech-
nologies directly, recent evidence suggests that electronic payments processing at the
Federal Reserve Banks exhibit signi®cant scale economies (Hancock et al., 1999).
In part as a result of these increased economies of scale, banking has consolidated very
dramatically over the last decade. The largest ten banking organizations in the United
States accounted for 27% of all operating income in 1990, compared to their share of 45%
in 1999.7 This trend toward increased consolidation, however, has been conspicuous by its
absence in the rest of the ®nancial services industry. In securities, for example, the share of
revenue from the top 10 ®rms accounted for 57% of industry revenues in 1999, down from
64% in 1990 (Securities Industry Association, 2000). In life insurance, the share of assets
held by the largest eight ®rms fell from 42% in 1988 to 35% in 1996, while in property and
casualty insurance the share of assets held by the largest eight ®rms rose only slightly,
from 33% in 1988 to 36% in 1996 (Berger et al., 1999).
Given these facts, an important question is: why is banking the only segment in
®nancial services to consolidate? Two key differences between banking and other
®nancial institutions come to mind. First, banking was inef®ciently balkanized during
the 1970s and 1980s because of regulations restricting both geographic and product
markets. Thus, deregulation generated much of the consolidation of the past two decades

6 For an overview of this large literature, see Berger et al. (1999).

7 Operating income equals net interest income plus non-interest income. Figures are based on the authors'
calculations from data in the 1990 and 1999 fourth quarter Reports of Income and Condition.
E-FINANCE: AN INTRODUCTION 13

(Jayaratne and Strahan, 1998). Second, e-®nance technologies have had perhaps there
greatest effect on the banking industry (and on other ®nancial intermediaries), where the
asset transformation role has been dramatically affected by reductions in transactions
and information costs. While the role of the ®nancial intermediary itself is threatened by
e-®nance technologies, the same cannot be said for either securities ®rms, or other
®nancial institutions that rely little on asset transformation (e.g., mutual funds, pension
funds, or insurance companies). Because e-®nance technologies bring large scale
economies, it makes sense that we have seen much more consolidation in banking than
in other parts of the ®nancial services industry. Moreover, in the securities industry e-
®nance may explain some of the declines in consolidation as discount brokers
experienced dramatic growth, thus taking market share away from the large, full-service
investment banks.
From a public policy perspective, the consolidation in U.S. banking raises few antitrust
concerns, at least over the short run. At the same time that e-®nance technologies
increase scale economies, they also lower barriers to entry into new markets, both
geographic and product. Petersen and Rajan (2001) ®nd that small businesses tend to
borrow from more distant banks now than in the past, and that this increase in distance
occurred because banks now use communications and information technologies more
intensively when making credit decisions. While research of banking markets from the
1980s suggested that prices and pro®ts were higher in more concentrated local markets,
evidence from the 1990s is less supportive of the idea that local market concentration
raises the price of banking services (Hannan, 1997; Radecki, 1998; Strahan and Hannan,
2000).
Like the United States, banking markets in Europe have been opened to greater
competition with deregulation and the possibility of cross-border banking. Also like the
United States, consolidation has occurred in European banking, but most of this
consolidation has been between banks operating in the same country. Danthine et al.
(1999) report the slow growth of cross-border banking in Europe during the 1990s. In
1992, for example, 4.7% of deposits held by French households and non-bank businesses
were with foreign banks; this percentage increased to just 5.8% by 1998. The ®gures did
increase in France and in other European countries, but they increased quite slowly. It is an
important research issue why European deregulation has not yet created a single banking
market. The relatively slow process of ®nancial integration in Europe represents a
signi®cant puzzle, particularly because the deregulation occurred at a time when e-®nance
technologies have made it easier for customers to bank anywhere. Monetary union, to be
completed in 2002, would seem to remove the last remaining explicit impediment to
creating a uni®ed banking and ®nancial system in Europe.

2.2.3. Access to credit. How will e-®nance technologies affect access to credit, parti-
cularly for borrowers that rely on a relationship with their lender? Small businesses
tend to concentrate their borrowing at a single bank with which they have a long-term
relationship, and the cost of credit seems to be lower when banks forge a relationship
with them (Petersen and Rajan, 1994; Berger and Udell, 1995). As banks invest more
of their capital in automating the lending process, less may be available to invest in
these relationships. This requires further research.
14 ALLEN ET AL.

Beyond the direct effect on relationship customers, the competition and consolidation
fostered by e-®nance technologies can also conceivably harm the relationship-based
customer. Some studies, for example, suggest that increased competition can hurt small
and young ®rms by reducing the incentive for banks to forge long-term relationships with
them (Petersen and Rajan, 1995; Boot and Thakor, 2000; Marquez, forthcoming). In
competitive markets, for example, it may be dif®cult for borrowers to commit to
maintaining a long-term relationship. Lenders may, therefore, be less willing to offer
credit on good terms early in the life of a ®rm because they have little con®dence that they
can recover their investment if the ®rm prospers.
Will consolidation also potentially harm relationship borrowers? Small banks have
traditionally been important lenders to small ®rms, and some authors have argued that this
role re¯ects their comparative advantage in relationship lending. Information stemming
from a relationship is generated and controlled by a human being (the loan of®ce) rather
than a machine (the computer), and small banks may be better able to control the agency
problem that is generated by the private information held by the loan of®cer than a larger,
more bureaucratic, bank.
While there may be theoretical reasons why increased competition and consolidation
could reduce credit availability to relationship borrowers, one can argue the other side as
well. The ®rst-order effect of banking competition is to spur innovation and force prices
closer to marginal costs. Moreover, increased bank size that comes with consolidation also
ought to lower lending costs overall, both because size reduces the need to hold costly
capital, and because banks' incentives to monitor their borrowers effectively are enhanced
when their probability of failure is reduced through diversi®cation (Diamond, 1984). The
empirical evidence needed to sort this issue out remains inconclusive and further work is
needed.8

3. E-Finance technologies in payment services

Electronic communications technologies have been used in the banking sector for many
years, particularly in interbank payment systems. One of the early applications of
electronic communication networks to ®nance was the Fedwire payment system. By 1918
this linked the accounts of banks held at the twelve Federal Reserve Banks across the
United States using leased telegraph wires and inaugurated electronic settlement of
accounts. This facility, combined with the ability to settle in central bank balances,
eliminated the ¯uctuating exchange rates that had previously been common between bank
balances due from banks located in different regions of the country.9 This early application

8 See, for example, Keeton (1996, 1997), Peek and Rosengren (1996, 1998), Strahan and Weston (1998),
Craig and Santos (1997), Kolari and Zardkoohi (1997a,b), Zardkoohi and Kolari (1997), Walraven
(1997), Berger et al. (1998), Sapienza (1998), Berger et al. (1999), Cole and Walraven (1999), Jayaratne
and Wolken (1999), Bonacorrsi di Patti and Dell'ariccia (2000), Black and Strahan (forthcoming), and
Cetorelli and Gambera (forthcoming).
9 See Gilbert (1997) and Garbade and Silber (1979) for a review of the early history of the Federal
Reserve's activity in payments and some of its effects.
E-FINANCE: AN INTRODUCTION 15

of electronic communication in ®nance displays one of the heralded features of Internet

communications: the importance of distance was reduced as telegraphic instructions to
adjust central bank balances replaced the need to physically ship coin and currency. In this
case, the advent of the specialized intermediary, the central bank and the electronic
communication method served to substitute for existing currency exchanges.
Electronic payment systems have evolved over the decades. Interbank payment systems
in the industrialized countries typically utilize dedicated telephone networks and
mainframe computer systems to manage their payments, which are characterized by
high volumes and values of payments. Many other payment systems, some intended for
lower-value payments, have also adopted electronic infrastructure. In the United States the
automated clearing house (ACH) system was designed and built in the 1970s and is widely
used for the payment of wages and other recurring payments.10 Many European Giro
systems have adopted electronic formats to reduce paper processing, as did the credit card
associations, both in the 1970s. The electronic communication and computation
technologies associated with these payment systems are clearly complementary to
banks' activities.
Another type of electronic technology banks have invested in is the ATM and the
network facilities that allow depositors remote access to their bank accounts at any time of
day. While the number of bank of®ces has continued to rise gradually over the past 20
years, the number of ATMs has exploded, from 18,500 in 1980 to 324,000 by 2000
(Thompson Financial, 2001). ATMs add convenience for the depositor, but they do not
seem to substitute for the branch. Rather it seems they are complements.
Although electronic payments systems have long played the dominant role in the
United States for interbank transactions, the same cannot be said for customer
transactions. The check is still the single most widely used non-cash payment
instrument in the United States. However, its dominance, which dates from the 19th
century, is eroding. While checks are estimated to have made up 78% of non-cash
transactions in 1994, their share fell to 70.7% in 1998 (BIS, 2000). Their share of
payment volume has been falling as the fast growth in credit and debit card payments
outpaces the relatively stagnant growth in check payments. Credit and debit card
payments increased their share over that time span from 18.7±24.5% of noncash
payments in the United States. Debit card payments alone increased in number by a
factor of 4.2 in that time period.
Although there has been an expansion in the market share of card payments in the
United States, checks continue to be the primary means of bill payment. As checks appear
to be more costly than some electronic alternatives, an important research issue is why
checks continue to predominate in this use.11
Two alternative hypotheses have been advanced to explain the seemingly slow adoption
by consumers of electronic means of bill payment. One hypothesis suggests that the rate of
adoption of electronic alternatives to the check is inef®ciently slow because of a

10 The ACH was built partly in response to the ``paper crisis'' of the late 1960s, which also spurred
adoption of electronic methods for settling securities transactions.
11 See Wells (1996) for a comparison of the private and social costs of checks.
16 ALLEN ET AL.

coordination failure.12 The other hypothesis suggests that the rate of adoption is ef®cient,
and that the continued prevalence of the check for bill payments re¯ects both the implicit
advantages of the check enjoyed by people who write checks and the costs of adopting the
current generation of alternative electronic means of payment. We will explore these
hypotheses in greater detail.
Payment systems are examples of network goods: they consist of originating and
receiving participants and intermediaries using complementary components of a
technology that, when used jointly, produces a transfer of funds. Costs are borne by all
the various participants in a payment system. The intermediaries recover their costs by
charging their customers, the senders and receivers of payments, either explicit fees or
implicit fees through earnings on deposits held by the intermediary. It is quite common in
the United States for banks to offer a package of banking services to their depositors that
includes a zero marginal cost for writing many checks and some minimum balance the
depositor must keep in his or her account.
When an electronic alternative means of payment becomes available, it too imposes
costs on each participant. Even if the total cost of the electronic method is less than that of
a check payment, it may not be adopted in the short-run at least, because some
participant's costs rise relative to the costs of using a check. Consider a consumer, for
example, when he or she is considering whether to adopt an electronic bill-payment
service. The consumer may well experience an increase in costs to adopt the electronic
alternative. First, the consumer has already learned how to make payments by check, while
the electronic services are unfamiliar. Second, electronic services typically charge explicit
fees for making electronic payments. Third, the consumer may not be able to pay all their
bills via the electronic service, and so may still choose to maintain the zero-marginal-cost
checking service from their bank. Fourth, the consumer will still need to open their bills,
maintain the balance in their checking account, etc. For some consumers, the
inconvenience of actually writing a check and the cost of buying postage may not
impose higher costs relative to the explicit costs and alternative format for the electronic
alternative. Finally, the check may offer conveniences relative to some electronic
alternatives, including the ability to stop payment, automatic provision of a receipt and
proof of payment, control of the timing of payment, etc.
Each of the other participants faces a similar set of comparisons in making their choice
of whether to adopt the alternative system. Absent an ability to make side-payments
among the parties, the lack of adoption by one class of participants can result in a
potentially lower cost alternative system going unused. Some payment instruments,
including general-purpose credit cards and point-of-sale debit cards do transfer side-
payments, known as interchange fees, from one class of intermediary to another. Other
payment instruments, including the check, the automated clearing house, most giro
systems and wholesale systems do not require participants to pay interchange fees; instead
those payment instruments are ``par'' payment instruments, which transfer the full value
of the payment between participants. In addition to this difference among alternative

12 See Humphrey and Berger (1990), Humphrey et al. (1998), Humphrey et al. (2000), Mester (2000),
Wells (1996), and McAndrews and Roberds (1999) who review these possibilities.
E-FINANCE: AN INTRODUCTION 17

systems, some payment systems such as Visa and Mastercard are ``branded,'' operate with
a strategy to provide a uniform level of service across end-users, and advertise their
services. While these two dichotomous characterizations of payment systems largely
coincide with whether the system is operated by private sector or the public sector, some
systems operated by the private sector, such as the ACH in the United States (whose rules
are determined by the National Automated Clearing House Association, a private sector
group), explicitly rule out interchange fees, and conduct little brand advertising.
Two important questions arise in considering this brief review of the hypothesis that
adoption of electronic alternatives to the check might be inef®ciently slow in the United
States. Why do banks offer zero-marginal-cost check services? Standard competitive
theory suggests that by offering a menu of alternative account relationships, bundling
minimum balances and alternative cost check services, banks are engaging in price
discrimination. It also suggests that check writing is a normal good, that is, that those
customers who can place a larger balance on deposit are more likely to write many checks.
Such a pricing structure is not necessarily inef®cient; Wilson (1998) points out situations in
which nonlinear pricing schedules can be ef®ciency enhancing. Generally, these situations
include economies of scale in production and demand heterogeneity, or economies of
scope in the production of related products. However, if a potentially less costly alternative
means of making payments is introduced, banks may have few unilateral incentives to alter
their price structure. In such a case, a coordinated movement of prices might overcome the
reluctance of banks to unilaterally change their prices. Humphrey et al. (1998) examine
such a coordinated move to explicit pricing for checks by banks in Norway. They ®nd that
there is signi®cant demand elasticity for checks, indicating that checks and other payment
services are highly substitutable. More empirical research is needed to further test this
hypothesis. The policy question of how and whether to craft incentives to move more
quickly away from the check means of payment is affected by the results of such research.
The second question that occurs is what are the effects of having some payment
instruments settle at par, and others settle with a payment of an interchange fee between
the intermediaries involved. Interchange fees open up the possibility of a redistribution of
the costs of adopting a particular payment instrument among the participants. Such an
option can potentially overcome the reluctance on the part of some of the participants to
adopt a payment instrument that otherwise imposes costs on them alone. In particular, as
Baxter (1983) points out, the payer and the payee both experience bene®ts from making
and receiving a payment, while the two (or more) intermediaries involved experience
costs. While the sum of the bene®ts may exceed the sum of the costs, it may not be the case
that the bene®ts accruing to the payer, for example, exceed the costs to the payer's bank.
Interchange fees can redistribute the bene®ts in such a way as to allow each intermediary
to cover its costs without the need to charge a price that exceeds its customer's willingness
to pay. The presence of interchange fees in credit and debit cards may have allowed those
means of payment to overcome some of the inherent challenges to the adoption of a new
payment instrument.13 However, interchange fees have been the focus of antitrust disputes

13 Baxter (1983), Carlton and Frankel (1995), Chakravorti and To (1999), Schmalensee (2000), and Rochet
and Tirole (2000) examine the economics of interchange fees.
18 ALLEN ET AL.

as well. The role of interchange fees, therefore, is an important issue for research in
electronic payments, and may affect the policy question mentioned above. How is one to
craft incentives to move more quickly away from check payments?
The other hypothesis explaining the continued use of checks in bill payments also relies
on the network aspect of a payment instrument. The hypothesis is that there has been
insuf®cient development of complementary systems to support electronic alternatives to
check payment. To take one example of this lack of development, consider the bills
companies send to their customers requesting payment. Most bills are in paper form and
delivered by mail. They often are delivered with a return envelope. The check is a
complementary form of payment for a paper bill, while an electronic means of payment
may not be as well integrated with paper processing of account numbers and mail delivery.
The hypothesis that electronic alternatives to the check are not well developed can be
elaborated in a number of ways, all suggesting that complementary technologies are not
yet in place to endow electronic means of payment equal net bene®ts to the check.
The relatively rapid adoption of the use of debit cards and, more recently, the growth of
personal online payment systems, raise questions about the importance of having
complementary technologies, legal structure and business agreements available to foster
the growth of payment networks. These examples of successfully introduced payment
instruments also raise the question of how differentiated is the demand for payment
instruments. Research into these questions is likely to be useful and likely to shed light on
whether incumbent payment networks, such as those offered by the credit card
associations, are especially well positioned to take advantage of future developments
and market demands.
Debit card usage has grown dramatically in the 1990s in the wake of two
complementary developments. First was the growth and widespread use of ATMs, cards
to access ATMs, and the supporting networks in the 1980s. The presence of ATM networks
allowed those organizations to take advantage of economies of scope in the use of their
network communication facilities to provide point-of-sale debit card services. The
widespread use of the ATM cards allowed consumers to adopt that means of payment
without having to acquire yet another card for the purpose of debiting of one's account at
the point of sale. Banks had to place devices at the merchant's locations, but this was an
easier task than doing that and inducing consumers to carry the cards. The second
development was the decision of the credit card networks to offer debit cards on a
widespread scale. The credit card networks had widespread merchant acceptance networks
and network communications facilities, and could leverage those in providing debit card
services.
In addition to these supply-side considerations, the demand for speed and convenience
at the point of sale was increasingly satis®ed by card methods of payment as electronic
authorization methods improved. The speed of card payment authorization has increased
over time so that such payments became faster to complete than check payments. This
technological development, combined with the many consumer rights enjoyed with card
payments (due both to business arrangements and to government regulation) have made
card payments increasingly advantageous to consumers.
An interesting use of the card payment system is personal online payments. These were
introduced in response to the success of online auctions. The online auction service, eBay,
E-FINANCE: AN INTRODUCTION 19

auctions several hundred thousand items per day. Because many of the sellers are
individuals, credit cards often are not accepted for payment for a purchase. For such
purchases, the main form of payment available to the parties, prior to the introduction of
personal online payment systems, was the check. Checks typically do not provide any sort
of instantaneous authorization or guarantee that the check will be honored by the bank on
which it was drawn. Instead, the check must be ``cleared'' to determine that the funds will
be ®nally transferred to the seller. In addition, of course, the buyer must deliver the check
to the seller, typically sending the check through the mail. These factors imply that in a
typical online auction transaction, paid for by check, the seller will not be fully informed
about the payment by buyer for many days after the auction occurs.
The delay and lack of integration with the online auction occasioned by a check
payment, in contrast to a credit card, led to the innovation of personal online payments.
One way to describe the personal online payment providers is to say that they provide
intermediation of credit card receipts: they accept a credit card payment from a buyer in an
auction, and deliver the payment to the seller by adding credit to the seller's credit card
line of credit. This is accomplished in various ingenious ways, with communication of the
payment advice made by e-mail. There has been considerable entry into this ®eld, and
many alternate providers offer services. A basic transaction occurs when the buyer funds
an account at the provider (usually using a credit card) and noti®es the seller via e-mail
that he has initiated a payment. The seller clicks on a link in the e-mail to contact the
provider to receive payment. The seller receives payment after establishing an account
into which the funds are transferred. The provider pays out these funds to the seller by
crediting the seller's credit card (or by sending the funds to the seller's bank account using
the automated clearing house or by check).
This innovation suggests many useful questions about developments in ®nancial e-
commerce. First, it is noteworthy that personal online payment systems developed in
response to a speci®c demand. There have been many attempts to create ``e-money'' with
certain features in the last decade. Most of these attempts were in response to a perceived
future demand, often a demand that did not materialize, or for which credit cards were
suf®cient. The concrete demand by buyers and sellers in online auctions was a signi®cant
market for a speci®c solution to making payments online. It led to a unique solution to this
need, which was unforeseen just a few years earlier. Are other distinct demands likely to
surface as e-commerce continues to grow? Second, the demand was related to personal
transactions. Individual consumers do not typically have access either to more
sophisticated and costly communications networks, such as private interdealer value-
added telecommunications networks (VANs), nor to the acceptance networks of the credit
card systems. It is in this realm of personal transactions, therefore, in which few close
substitutes existed, that the innovation took root quickly. One question that arises is
whether banks will allow individuals to more easily accept credit card payment directly.
This question leads to many other questions regarding the technological feasibility of
allowing persons to accept credit card payments, and the business case for doing so.
Finally, e-mail was used as a building block for communication, and the clearing and
settlement networks of the credit card associations were relied upon for that part of
payment services. These ``off-the-shelf'' technologies, one a communications and one a
®nancial technology, provided many advantages for this form of payment. By building in
20 ALLEN ET AL.

these strong technological complementarities, the providers of these payment systems

could piggy-back on the widespread use of e-mail and credit cards. The use of these
complementary and widespread technologies therefore allowed many consumers to easily
adopt personal online payments. A question here is how likely are consumers to adopt
relatively customized technologies, such as a particular type of e-money stored-value card,
for example? Many of the research questions here are best tackled with market research
approaches or with an experimental approach.
The adaptation of credit card payments and the use of them rather than checks in online
payment systems illustrates an important research issue. This is the extent to which credit
and debit cards will take over from checks and eventually become the primary payments
system, at least for personal payments. One could imagine, for example, if credit cards
paid interest on positive balances they would become an attractive substitute for checking
and savings accounts. This will be particularly true if the time for undertaking transactions
using credit cards is further reduced. Other interesting policy issues ¯ow from this
possibility for substitution of credit card accounts for checking accounts, in particular,
would positive balances on a credit card account be considered a bank deposit?

4. The impact on ®nancial markets

In this section we consider the impact of electronic communication and computation on

stock markets, bond markets, and foreign exchange markets. In recent years a very large
literature has developed on market microstructure (see, for example, O'Hara (1995) for an
overview). Most of this literature is concerned with understanding the operation of stock
markets. Traditionally stock markets were at physical locations and operated with face-to-
face communication. The development of the over the counter market for stocks into the
Nasdaq trading system was an early example of e-®nance in the context of markets.
Subsequently most stock exchanges in the world including the London Stock Exchange,
The Tokyo Stock Exchange and the Frankfurt Stock Exchange have moved to electronic
trading. The New York Stock Exchange, which is the largest by market capitalization in
the world, still uses physical trading. However, even they have introduced the Network
NYSE platform that allows retail and institutional investors to engage in electronic
trading.
The foreign exchange (FX) and bond markets provide an interesting contrast to stock
markets. These have traditionally been dealer markets that operate over the telephone.
There has not been a physical location and trading is done directly by pairs of dealers or
with the help of brokers that intermediate between them. In recent years the foreign
exchange market has started to rapidly move to electronic trading. In contrast the bond
market has been slow to change and is still largely a telephone market. The similarities and
differences between these markets raise a number of important issues. The ®rst is why they
traditionally have such a similar market microstructure. The second is why there has been
such a difference in the speed with which they have moved to electronic trading.
E-FINANCE: AN INTRODUCTION 21

4.1. Stock markets

Many stock markets around the world have adopted electronic trading methods. In the
United States, the Nasdaq market was created in 1971 to allow dealers to make over-the-
counter trades on an electronic system of linked screens. It has grown rapidly and has
become one of the main equity markets in the United States. Regulatory pressures in the
mid-1990s led to the entry of many electronic communications networks (ECNs) in the
trading of Nasdaq-traded stocks. These electronic systems allow a wider set of participants
to view limit orders (orders to buy or sell speci®c amounts of stock at various prices), as
well as allowing for the possibility of executing trades electronically. In more recent
developments, exchange based markets (including the New York Stock Exchange) have
implemented various automated order execution systems, either to trade small orders,
supplementing their ¯oor-based trading systems, or as the primary means of trading. There
are many important questions that have been raised and addressed regarding the growing
electroni®cation of stock trading. The effects of electronic trading on the cost, speed, and
overall liquidity of trading are continuing important research questions, as are the effects
on the economies of scale and market structure of exchanges.
As described by Weston (this volume), the recent growth of ECNs has enhanced
competition in trading for the stocks traded on ECNs and Nasdaq. ECNs allow traders to
transact directly with each other at a small fee in an electronic marketplace, thereby
eliminating the need to compensate the dealer via the bid-asked spread. The ECNs allow
traders to view the bids and offers in their limit-order book. In some cases these electronic
limit-order books are available to the public on the Internet.14 Traders can ``hit'' these bids
and offers if they offer attractive prices, and ``crossing'' trades are automatically matched,
thereby short-circuiting the traditional role of the securities dealer. ECNs can also allow
traders to route orders to the dealer that offers the best price for the order. Weston shows
that bid-asked spreads have declined with the growth of trading on ECNs, and that the
spreads narrowed the most in the stocks that are most frequently traded on the ECNs,
accounting for other factors. It is an important research issue to further investigate the
source of these lower spreads. It is not clear whether ECNs brought these lower spreads
from greater transparency of pricing, from greater market interconnection, or from lower
resource costs or faster speed of routing orders to best-priced dealers. These issues are
important ones in better understanding market microstructure in an interconnected dealer
market.
Practitioners are not all of one mind on the ef®cacy of electronic order execution for all
market environments. In particular, some argue that adoption of electronic systems for all
¯oor-based exchanges is not as likely to lead to improvements in market ef®ciency as in
dealer markets.15 Domowitz (this volume) provides cross-country evidence on this issue.
He ®nds that the adoption of automated execution technology can reduce the total costs of
order execution signi®cantly, even in comparison to traditional, ¯oor-based exchanges. He

14 See McAndrews and Stefanadis (2000) for a descriptive review of ECNs.

15 These views can be found in testimony reported in the United States Senate Committee on Banking,
Housing, and Urban Affairs (2000).
22 ALLEN ET AL.

hypothesizes that these reduced costs can result from lower development and operating
costs, as well as important liquidity advantages that arise from the greater transparency of
the trading system, when compared with a specialist or non-automated system. These
hypotheses frame research agenda that should be investigated. In other work, notably
Domowitz and Steil (1999), Domowitz has investigated the effects of automated execution
technology on market merger activity, cross-border trading, changes in methods of
exchange governance and competition. These are topical issues, especially in Europe,
where signi®cant mergers of exchanges have occurred, and where several exchanges have
many exchanges have changed governance structures in the direction of becoming for-
pro®t stock-based corporations.16
So®anos (2000) points out that much trading, as in ``upstairs market'' trading, avoids
exchanges altogether. This trading of large blocks of stock often occurs without posting
limit orders with specialists on exchange ¯oors or in automated systems. Instead, the
orders stay ``hidden'' until the trader sees an opportunity to move a large block, and only
then announces the willingness to trade. So®anos expresses the view that until automated
systems can accommodate these traders, the adoption of such a system might result in even
more off-exchange trading ( possibly at higher cost), which a researcher might not observe
when calculating the costs of trading pre- and post-adoption of an automated system. Until
automated systems can replicate all the functions of a ¯oor-based system and its
institutions, such as specialists, So®anos does not believe that automated systems are
capable of replacing the full functions of a ¯oor-based trading system. Again, these views
yield important questions for researchers in market microstructure.
Madhavan (2000) argues that the Internet has reversed a centuries-old trend towards
market consolidation. The reason is that recent trends in terms of improved access to
information and reductions in trading costs have greatly increased intraday volatility. This
has a number of undesirable effects including increasing institutional trading costs. The
Internet not only allows easy access to information but also acts as a coordination device.
This creates a challenge to regulators since the possibilities for manipulation are
signi®cantly increased. However, in the longer run the network externalities associated
with the Internet are likely deepen markets and improve price ef®ciency. The public policy
issues Madhavan's analysis raises are an important topic for research.

4.2. Foreign exchange markets

Given the enormous volumes of foreign exchange that are traded on a daily basis ($1.5
trillion a day according to Banks, 2001) relatively little academic work has been concerned
with foreign exchange markets. Lyons (2001) contains an excellent account of the work
that has been done. He and others have developed a market microstructure approach to the
foreign exchange markets.
The foreign exchange market has traditionally been a multiple dealer market. Lyons
(2001) points out that it is characterized by a number of factors. It has an enormous trading

16 See McAndrews and Stefanadis (2001) for a review of some of these developments in Europe.
E-FINANCE: AN INTRODUCTION 23

volume and trades between dealers account for most of this volume. Another important
factor is that trade transparency is low. A large proportion of the trading, roughly two thirds,
is interdealer trading. For many years it was a telephone-based market. Two major systems
(Reuters and EBS) were developed for providing quotes. Initially, trades were still done
over the telephone. However, these systems have developed into full trading platforms.
Dealers are able to observe the best bid and offer in the market. Allen et al. (2001) point out
that the market has rapidly performed the transition from a telephone market to an
electronic market. The Bank for International Settlements (BIS) performs a triennial survey
of trading between dealers. In 1995 their survey showed that 20±30% was conducted
electronically, by 1998 this had risen to 50% and by 2001 it was expected to be over 90%.
Although the interdealer market for FX has largely become electronic the market
between large corporations and dealers has been less affected and communication over the
telephones remains important. Several Internet platforms such as FXall and Atriax aim to
capture this market by offering executable quotes.
This outline of the operation of the FX market raises a number of interesting questions.
The ®rst is why the foreign exchange market operated as a decentralized telephone market.
The second is why the interdealer market moved so quickly to an electronic market and
why the same has not yet happened to the corporate market.

4.3. Bond markets

Similarly to the FX market, compared to stock markets relatively little academic work has
been done on the operation of bond markets. The structure of bond markets has
traditionally been very similar to that of foreign exchange markets. The secondary trading
in government, municipal and corporate bond markets is done over the telephone in
multiple dealer markets.
To illustrate the operation of bond markets we will focus on the markets for government
securities. These are large in terms of volume. Fabozzi (2001) reports that the volume of
U.S. Treasury securities that is traded daily was around $200 billion per day in the ®rst half
of 1999. The primary market involves auctions that are open to all but where a special role
is played by primary government securities dealers. They are expected to participate
meaningfully in the Treasury auctions and interact directly with the Federal Reserve Bank
of New York in open market operations. They also supply market information to the Fed.
The principal market makers in the secondary market are the primary dealers. Interdealer
brokers provide dealers with electronic screens that post bid and offer prices. Trades are
typically executed over the telephone.
In contrast to the FX market the move towards electronic trading has been relatively
slow. In the U.S. Treasury market Allen et al. (2001) report that 40% of securities were
traded electronically in 2000. In contrast for corporate bonds only 10% were traded
electronically. This is not because of a lack of bond trading platforms. Banks (2001) points
out that in 1997 there were 11 online platforms. By 1999 this had increased to 40 and by
2000 to more than 80.
This sketch of the bond markets and the comparison to experience in FX markets
suggests a number of important research issues. Why are bonds, particularly government
24 ALLEN ET AL.

bonds traded in dealer markets and not on exchanges? Are the factors that lead to this
structure the same as for FX or different? Why has electroni®cation been so slow
compared to FX and other ®nancial markets? Will the delay in electroni®cation be
temporary or more long lasting?

5. Concluding remarks

E-®nance is not new. For example, the Fedwire used electronic communications system as
early as 1918. The Nasdaq market involved the electronic trading of stocks as early as
1971. The difference today is that electronic communication and computation is now used
much more widely than before. A large number of people have access to the Internet and
this has vastly changed the opportunities for the use of electronic payments systems, the
operations of ®nancial services ®rms and ®nancial markets. We have argued that this
change raises a number of important research issues. For example, is the widespread use of
paper-based checks ef®cient? Will the ®nancial services industry be fundamentally
changed by the advent of the Internet? Why have there been such large differences in
changes to market microstructure across different ®nancial markets? We look forward to
these and other questions being answered as the emerging ®eld of e-®nance develops.

References

Allen, F., and D. Gale. ``Financial Markets, Intermediaries, and Intertemporal Smoothing.'' Journal of Political
Economy 105 (1997), 523±546.
Allen, H., J. Hawkins, and S. Sato. ``Electronic Trading and its Implications for Financial Systems.'' Working
paper, Bank of England, 2001.
Bank for International Settlements. Statistics on Payment Systems in the Group of Ten Countries. Basel:
Switzerland, 2000.
Banks, Erik. E-Finance: The Electronic Revolution. London: John Wiley & Sons, 2001.
Baxter, William F. ``Bank Interchange of Transactional Paper: Legal and Economic Perspectives.'' Journal of
Law & Economics 24 (1983), October, 541±588.
Berger, Allen N., Rebecca S. Demsetz, and Philip E. Strahan. ``The Consolidation of the Financial Services
Industry: Cause, Consequences and Implications for the Future.'' Journal of Banking and Finance 23 (1999),
135±194.
Berger, Allen N., and Gregory F. Udell. ``Some Evidence on the Empirical Signi®cance of Credit Rationing,''
Journal of Political Economy (1992), 1047±1081.
Berger, Allen N., and Gregory F. Udell. ``Relationship Lending and Lines of Credit in Small Firm Finance,''
Journal of Business 68 (1995), 351±382.
Berger, Allen N., and Gregory F. Udell. ``The Economics of Small Business Finance: The Roles of Private
Equity and Debt Markets in the Financial Growth Cycle.'' Journal of Banking and Finance 22 (1998), 613±
673.
Berger, Allen N., Anthony A. Saunders, Joseph K. Scalise, and Gregory F. Udell. ``The Effects of Bank Mergers
and Acquisitions on Small Business Lending.'' Journal of Financial Economics 50 (1998).
Bernanke, Ben, and Alan Blinder. ``The Federal Funds Rate and the Channels of Monetary Policy.'' American
Economic Review 82, no. 4 (1992), 901±921.
Bonacorrsi di Patti, Emilia, and Giovanni Dell'ariccia. ``Bank Competition and Firm Creation.'' Mimeo,
January 30, 2000.
E-FINANCE: AN INTRODUCTION 25

Boot, Arnoud, and Anjan V. Thakor. ``Can Relationship Banking Survive Competition?'' Journal of Finance 55,
no. 2 (2000), 679±714.
Black, Sandra E., and Philip E. Strahan. ``Entrepreneurship and Bank Credit Availability.'' Journal of Finance
(forthcoming).
Carlton, Dennis W., and Alan S. Frankel. ``The Antitrust Economics of Credit Card Networks.'' Antitrust Law
Journal 63, no. 2, Winter (1995), 643±668.
Chakravorti, Sujit, and Ted To. ``A Theory of Merchant Credit Card Acceptance.'' Federal Reserve Bank of
Chicago Working Paper Series, WP-99-16, November, 1999.
Cetorelli, Nicola, and Michele Gambera. ``Bank Structure, Financial Dependence and Growth: Intenational
Evidence from Industrial Data.'' Journal of Finance (forthcoming).
Cole, Rebel A. ``The Importance of Relationships to the Availability of Credit.'' Journal of Banking and Finance
22 (1998), 959±977.
Cole, Rebel A., Lawrence G. Goldberg, and Lawrence J. White. ``Cookie-Cutter versus Character: The Micro
Structure of Small Business Lending by Large and Small Banks.'' Proceedings of Conference on Business
Access to Capital and Credit, March 8±9, 1999.
Craig, Ben R., and Joao A.C. dos Santos. ``Banking Consolidation: Impact on Small Business Lending.''
Working paper, Federal Reserve Bank of Cleveland, 1997.
Cyrnak, A., Hannan, Timothy H. ``Bank Lending to Small Businesses and the Use of CRA Loan Data to
Measure Market Structure.'' Federal Reserve Board working paper, 1998.
Danthine, Jean-Pierre, Francesco Giavazzi, Xavier Vives, and Ernst-Ludwig von Thadden. The Future of
European Banking, CEPR monograph on European Integration 9, 1999.
Diamond, Douglas. ``Financial Intermediation and Delegated Monitoring.'' Review of Economic Studies 51
(1984), 393±414.
Domowitz, Ian, and Benn Steil. ``Automation, Trading Costs, and the Structure of the Securities Trading
Industry.'' Brookings-Wharton Papers on Financial Services 2 (1999), 33±92.
Estrella, Arturo. ``Securitization and the Ef®cacy of Monetary Policy.'' Presented at the Federal Reserve Bank of
New York Conference on Financial Innovation and Monetary Transmission, July 2001.
Evans, David S., and Richard Schmalensee. Paying with Plastic: The Digital Revolution and Borrowing. MIT
Press, 1999.
Fabozzi, F. J. The Handbook of Fixed Income Securities. New York: McGraw-Hill, 2001.
Federal Reserve Board of Governors. 2000. Report to Congress on Markets for Small Business and Commercial
Mortgage Related Securities Markets, available online at http://www.federalreserve.gov/boarddocs/
RptCongress/markets2000.pdf.
General Accounting Of®ce, Payments, Clearance, and Settlement: A Guide to the System, Risks and Issues,
June, 1997.
Gertler, Mark, and Simon Gilchrist. ``Monetary Policy, Business Cycles and the Behavior of Small
Manufacturing Firms.'' Quarterly Journal of Economics (1994), 309±340.
Gilbert, R. Alton. ``The Advent of the Federal Reserve and the Ef®ciency of the Payments System: The
Collection of Checks, 195-1930.'' Explorations in Economic History 37 (April, 2000), 121±148.
Hancock, Diana, David B. Humphrey, and James A. Wilcox. ``Cost Reductions in Electronic Payments: The
Roles of Consolidation, Economies of Scale and Technical Change.'' Journal of Banking and Finance 23
(1999), 391±421.
Hannan, Timothy H. ``Market Share Inequality, the Number of Competitors, and the HHI: An Examination of
Bank Pricing.'' Review of Industrial Organization 12 (1997), 23±35.
Huck, Paul, Sherrie L. W. Rhine, Philip Bond, and Robert P. Townsend. ``A Comparison of Small Business
Finance in Two Chicago Minority Neighborhoods.'' Proceedings of Conference on Business Access to
Capital and Credit, March 8±9, 1999.
Humphrey, David B., Setsuya Sato, Masayoshi Tsurumi, and Jukka M. Vesala. ``The Evolution of Payments in
Europe, Japan, and the US: Lessons for Emerging Market Economies.'' World Bank Policy Research
Working Paper, 1676, October, 1996.
Humphrey, David B., L. Pulley, and Jukka M. Vesala. ``Cash, Paper and Electronic Payments: A Cross-Country
Analysis.'' Journal of Money, Credit and Banking 28, no. 4 (1996), 914±939.
Humphrey, David B., L. Pulley, and Jukka M. Vesala. ``The Check's in the Mail: why the US Lags Behind in the
26 ALLEN ET AL.

Adoption of Cost Saving Electronic Payments.'' Journal of Financial Services Research 17, no. 1 (2000),
17±39.
Jayaratne, Jith, and Philip E. Strahan. ``Entry Restrictions, Industry Evolution and Dynamic Ef®ciency:
Evidence from Commercial Banking.'' Journal of Law and Economics 41 (1998), 239±274.
Jayaratne, Jith, and John D. Wolken. ``How Important are Small Banks to Small Business Lending? New
Evidence from a Survey of Small Businesses.'' Journal of Banking and Finance 23 (1999), 427±
458.
Keeton, William R. ``Do Bank Mergers Reduce Lending to Businesses and Farmers? New Evidence from Tenth
District States.'' Federal Reserve Bank of Kansas City, Economic Review 81, no. 3 (1996), 63±75.
Keeton, William R. ``The Effects of Mergers on Farm and Business Lending at Small Banks: New Evidence
from Tenth District States.'' Working paper, Federal Reserve Bank of Kansas City, 1997.
Kolari, J., and A. Zardkoohi. ``The Impact of Structural Change in the Banking Industry on Small Business
Lending.'' Report to the Small Business Administration, 1997a.
Kolari, J., and A. Zardkoohi, A. ``Bank Acquisitions and Small Business Lending.'' Working paper, Texas A&M
University, 1997b.
Loughran, T., J. Ritter, and K. Rydqvist. ``Initial Public Offerings: International Insights.'' Paci®c-Basin
Finance Journal 2 (1994), 165±199.
Lown, Cara, and Donald P. Morgan. ``Credit Effects in Monetary Policy, Then and Now.'' Presented at the
Federal Reserve Bank of New York Conference on Financial Innovation and Monetary Transmission, March
30, 2001.
Lyons, R. K. The Microstructure Approach to Exchange Rates. Cambridge, A: MIT Press (forthcoming).
Madhavan, Ananth. ``In Search of Liquidity in the Internet Era.'' paper presented at the Ninth Annual Financial
Markets Conference of the Federal Reserve Bank of Atlanta 2000.
Marquez, Robert. ``Competition, Adverse Selection and Information Dispersion in the Banking Industry.''
Review of Financial Studies (forthcoming).
Mester, Loretta J. ``The Changing Nature of the Payments System: Should New Players Mean New Rules?''
Business Review, Federal Reserve Bank of Philadelphia, March/April: 3±26, 2000.
McAndrews, James, and William Roberds. ``A General Equilibrium Analysis of Check Float.'' Journal of
Financial Intermediation 8 October (1999), 353±377.
McAndrews, James, and William Roberds. ``The Economics of Check Float.'' Economic Review, Federal
Reserve Bank of Atlanta, Fourth Quarter, 1999, 1±11.
McAndrews, James, and Chris Stefanadis. ``The Emergence of Electronic Communications Networks in the US
Equity Markets.'' Current Issues in Economics and Finance, Federal Reserve Bank of New York 6, no. 12
(2000).
McAndrews, James, and Chris Stefanadis. ``The Development of Pan-European Stock Exchanges.'' Mimeo,
Federal Reserve Bank of New York, September 2001.
Mishkin, Frederic S., and Philip E. Strahan. ``What Will Technology do to the Financial Structure?'' In: The
Effect of Technology on the Financial Sector, Brookings-Wharton Papers on Financial Services, Robert
Litan and Anthony Santomero, eds, 1999, 249±287.
O'Hara, M. Market Microstructure Theory. Cambridge, MA: Blackwell, 1995.
Peek, Joe, and Eric S. Rosengren. ``Small Business Credit Availability: How Important is the Size of the
Lender?'' In: A. Saunders, and I. Walter, eds, Financial System Design: The Case for Universal Banking.
Burr Ridge, IL: Irwin Publishing, 1996, 628±655.
Peek, Joe, and Eric S. Rosengren. ``Bank Consolidation and Small Business Lending: It's not Just Bank Size that
Matters.'' Journal of Banking and Finance 22 (1998).
Petersen, Mitchell A., and Raghuram G. Rajan. ``The Bene®ts of Firm-Creditor Relationships: Evidence from
Small Business Data.'' Journal of Finance 49 (1994), 3±37.
Petersen, Mitchell A., and Raghuram G. Rajan. ``The Effect of Credit Market Competition on Lending
Relationships.'' Quarterly Journal of Economics 110 (1995), 407±443.
Petersen, Mitchell A., and Raghuram G. Rajan. ``Does Distance Still Matter? The Information Revolution in
Small Business Lending.'' Mimeo, March, 2001.
Radecki, Lawrence J. ``The Expanding Geographic Reach of Retail Banking Markets.'' Federal Reserve Bank of
New York Economic Policy Review 4 (1998), 15±34.
E-FINANCE: AN INTRODUCTION 27

Radecki, Lawrence J., John Wenninger, and Daniel K. Orlow. ``Industry Structure: Electronic Delivery's
Potential Effects on Retail Banking.'' Journal of Retail Banking Services 19, no. 4 (1997), 57±63.
Rochet, Jean-Charles, and Jean Tirole. ``Cooperation among Competitors: The Economics of Payment Card
Associations.'' Mimeo, Institut D'Economie Industrielle, University of Toulouse, 2000.
Sapienza, Paola. ``The Effects of Banking Mergers on Loan Contracts.'' Northwestern University working
paper, 1998.
Schmalensee, R. ``Payment Systems and Interchange Fees.'' Unpublished manuscript, Massachusetts Institute of
Technology, December 2000.
Securities Industry Association. Securities Industry Fact Book, available online at http://www.sia.com/
publications/html/fact_book.html, 2000.
So®anos, George. ``Comments.'' Financial Market 2000 Conference on E-Finance, Federal Reserve Bank of
Atlanta, October 2000.
Strahan, Philip E., and James P. Weston. ``Small Business Lending and the Changing Structure of the Banking
Industry.'' Journal of Banking and Finance 22 (1998), 821±845.
Thompson Financial. ``EFT Data Book.'' ATM & Debit News, 1, no. 44 (September 13, 2001).
United States Senate Committee on Banking, Housing and Urban Affairs. The Financial Marketplace of the
Future: Hearings. 106th Congress, 2nd session, February 29, 2000.
Wells, Kirsten E. ``Are Checks Overused?'' Quarterly Review, Federal Reserve Bank of Minneapolis, 20 (1996),
Fall: 2±12.
Wilhelm, William J. ``Internet Investment Banking: The Effect of Information Technology on Relationship
Banking.'' Journal of Applied Corporate Finance Spring (1999).
Zardkoohi, A., and J. Kolari. ``The Effect of Structural Changes in the US Banking Industry on Small Business
Lending.'' Working paper, Texas A&M University, 1997.