Supervision and Regulation of Network Banks
Aleksander BERENTSEN <firstname.lastname@example.org>
Public computer networks, in particular the Internet, have the potential to transform the financial services sector by providing a fast, cheap way to sell financial services. Low setup costs and the transnationality of the Internet could remove significant barriers to entry in the financial services industry. Cross-border provision of services and the high mobility of network banks could challenge the ability of national and international authorities to establish and enforce banking regulations. This paper considers the supervision and regulation of banks providing financial services on public computer networks for the mass retail market, i.e., deposit taking and lending of money on retail and small- and medium-size company markets.
The Internet's explosive growth has initiated considerable activity in the financial services industry. For this industry, the Internet and, in particular, the World Wide Web, serve as new vehicles for transmitting financial information, comparable to the invention of the telegraph 150 years ago and its use for transmitting financial information. Although computer networks only transport financial information, many predict radical changes including the "dissolution of geographic markets into virtual financial systems" and the "loss of national independence."
The facts are that (a) financial services are information commodities and (b) public computer networks offer a fast, cheap way to trade information. Public computer networks can radically improve efficiency in the financial services industry. Increased efficiency relies on three characteristics of these networks:
The gains in efficiency will affect banking competition and banking regulation. While the small setup costs and the cross-border provision of financial services will spur competition, it is the Internet's borderless nature that could pose a major challenge for the regulation and supervision of financial intermediaries.
The mass retail market, i.e., deposits taking and the lending of money on retail and small- and medium-size company markets, could be affected most. The retail sector, so far, has been less affected by telecommunication technology than have other financial sectors, and the mass retail market has remained largely a national business. In contrast, large corporate banking is already international, and the cross-border provision of financial services in these markets is the rule rather than the exception. The prospect of an international mass retail market for financial services is a likely candidate for public concern and intervention because regulatory efforts are often directed toward protecting small depositors.
In this paper, the first section on network (Internet) banks distinguishes network retail banking and "traditional" retail banking; the second section on banking regulation identifies the need for public intervention in the banking industry; and the third section examines supervision and regulation of network banks.
Exponential growth of the Internet started when the World Wide Web was invented in the early 1990s.  The emergence of financial service providers on the World Wide Web is more recent. Security First Network Bank in Atlanta, Georgia, was the first bank to use the Web as its main channel for offering traditional financial services, such as transaction accounts. The bank opened its business on October 18, 1995.
Despite its youth, the Web has initiated many activities in the financial services industry; in December 1996, more than 1,490 banking institutions were providing financial information on it.  There is also a growing demand for online financial services. For example, Wells Fargo in San Francisco increased its online banking base from 20,000 to 270,000 users in 18 months. 
Because banking on the Internet is so new, no systematic research on Internet-based financial services and their economic implications is available. Existing literature consists entirely of anecdotal evidence from individual companies, usually in trade journal articles with no empirical basis (Kalakota 1996). Moreover, there exists no research on the prudential regulation of network banks.
Analysis of network banks and their regulation must start with identifying the characteristics distinguishing network banking from traditional banking. Our main proposition is that in the relation between network banks and customers, the physical location of the network bank or the customer plays no role. We term this fact the irrelevance of physical location. In all other respects network banking works like traditional retail banking. In particular, they could provide the same services: facilitating transactions, portfolio management, risk transformation, and monitoring.
The irrelevance of physical location is based on the fact that marginal costs of providing financial services on a public computer network are not related to a customer's location; more importantly, a customer's marginal costs for obtaining financial services are independent of the bank's location. In contrast, the "traditional" retail bank's location is an important (cost) factor for bank customers and banks. In "traditional banking" a customer's marginal costs for obtaining financial services increase with distance to his bank. Consequently, traditional banks compete for location and (over) invest in branches and ATM-systems to collect deposits (Neven, 1990).
The irrelevance of the physical location has three implications.
First, it enables cross-border trade of financial services in
the retail market. Without legal restrictions, agents can shop
for financial services anywhere on this planet as long as the
network bank has access to a local ATM network to obtain cash.
However, the emergence of electronic money, respectively digital
cash, could even remove this requirement. White (1996) has stressed
What strikes me as the most exciting potential development to come from the new payment technologies is that, as they lower the cost of wiring money from $20 to 2 cents or less per transaction, they give ordinary small savers affordable access to offshore banking. With direct deposit of paychecks and with analog currency available at ATMs whenever we want it, many of us no longer need to visit our bank in person. Why not keep your accounts with a reputable bank (perhaps a branch of a major Swiss bank) in the Bahamas or Cayman Islands?
Second, network banks can move their physical location without changing their relation to their customers. In particular, they can move their business across nations "overnight." Thus, network banks can react faster than traditional banks to changing economic conditions and regulatory requirements. For example, they can easily shift their production to low factor cost countries and nations with inexpensive regulatory regimes.
The potential of a bank to move its entire legal location overnight is novel. There exists no experience with this type of behavior, and it is possible that one will never observe a (honest) bank to move its business overnight. Nevertheless, this potential will have legal and economic implications because the credible threat of dislocation will change the relation between banks and their regulators.
Third, it changes the behavior of bank customers by cutting the previously close and long-standing ties between a bank and a customer.  If banking customers can open a new banking relationship at very low costs they are likely to switch banking relationships more often. Particularly, bank customers are likely to switch immediately if their is any indication that the bank could become insolvent. This will increase the chances of bank runs. It also makes it more difficult for banks to forecast their operations, and it increases the volatility in this industry.
Any attempt to understand the prudential regulation of banks requires examining the nature of financial intermediation, the potential for market failures, and the attempt to correct these failures through public intervention. The theory of financial intermediation stresses four functions of financial intermediaries: 
According to the first function, financial intermediaries facilitate transactions by organizing the payment system. They offer check clearing and payment products such as credit cards, debit cards, and travelers checks. The second function, portfolio management, refers to the management of diversified portfolios consisting of financial liabilities issued by firms and governments and sold to the public. The third function, risk transformation, refers to the transformation of risky assets issued by firms into fixed interest deposits demanded by households and to the transformation of illiquid assets into liquid liabilities to provide liquidity insurance to households. The fourth function, monitoring, refers to monitoring activities of financial intermediaries, which reduces problems of moral hazard and asymmetric information in relation between firms and financiers (Hellwig, 1991).
Banking regulation theory is closely linked to these functions. The issue is twofold:
It is commonly accepted that any banking regulation must rely on some form of market failure. If banks were only to offer transaction and portfolio management services, regulations would be unnecessary. Fama (1980) demonstrates that these two functions do not cause market failures. The potential for market failures is associated with the third and fourth functions. Particularly, it is the financing of illiquid assets with short-term deposits and the potential of bank runs that create a need for public intervention and the establishment of a safety net to guarantee the stability of the financial system (Baltensperger and Dermine, 1990).
Diamond and Dybvig's (1983) were the first to model bank runs. Their model suggests an equilibrium in which all depositors try to close their accounts, forcing the bank to sell illiquid assets, resulting in the failure of the otherwise solvent bank. The interesting aspect of their model is that there is no underlying real reason for the bank run: the illiquid asset is a completely safe investment. It is the expectation about the behavior of other depositors that drives the behavior of any individual. In the bank run equilibrium, these expectations are fulfilled.
A related type of market failure stresses the "contagious" nature of bank runs (Baltensperger and Dermine, 1987). A bank failure can trigger a run on other, solvent banks when bank customers of the solvent bank assume that the value of banks' assets are highly correlated with each other. In most countries there exists a public deposit insurance or a lender-of-last-resort agency to prevent bank runs. Deposit insurance makes deposits risk-free, thereby eliminating the incentive for early withdrawals.
An additional basis for market failures is asymmetric information
between banks and their depositors. Banks are better informed
about the quality of their loans and the security of their assets
than depositors are. Depositors can become better informed by
monitoring banks; monitoring bank solvency, however, is expensive
and requires skills small depositors may not have. In addition,
information about bank solvency has the characteristic of a public
good. This view is emphasized in Dewatripont and Tirole (1992,
One neglected, although certainly quantitatively important, feature of banks [ is ] the dispersed nature of the debt-holders or depositors. Small depositors typically have no time or expertise to perform the monitoring and control that the optimal governance structure [. . .] requires. And even if they did, they would be tempted to free ride on each other's monitoring and exercise of control.
According to Dewatripont and Tirole (1992), when a bank is in trouble, bank managers and equity holders have an incentive to gamble for resurrection. As a consequence, debt holders of banks, i.e., depositors, must take control when bank performance is bad because their incentives are to limit risk taking. A large number of small free-riding depositors, however, cannot perform this task, which suggests a role for public intervention. A public agency would have to regulate banks ex-ante by imposing capital requirements and limiting the growth of deposits. In addition, a public agency would have to intervene ex post acting on behalf of small depositors in bad times.
In summary, modern theory of financial intermediation suggests two reasons for public intervention:
A further reason for government intervention in the banking industry, which is not directly connected to the theory of financial intermediation, is the interest of the public to fight money laundering and to guarantee the stability of the payment and the financial system.  This requires a certain control of national and international money flows, which is more difficult the faster and more unpredictable transactions are. Particularly, the emergence of digital money flows where the transacting parties remain anonymous would increase these problems to a considerable extent.
A final note is required, here. Any banking regulation induces unwanted side effects. For example, deposit insurance that prevents bank runs lowers the incentives of banks to invest prudentially. The U.S. deposit insurance system, for example, is blamed for hundreds of billions of dollars of taxpayers' losses by creating moral hazards. Deposit insurance lowers the reputation-building incentives of financial institutions. As a consequence, financial institutions are likely to take more risks and to monitor less if they have deposit insurance than if they do not. Present regulatory theory does not offer clear-cut recommendations, owing to the complexity of the welfare analysis (Vives, 1991). It is therefore important to balance carefully the benefits of any regulation versus the potential for harmful side effects.
This section considers whether network banks provide additional reasons for public intervention and whether they require different treatments. In the first section of this paper, we noted that in the relationship between the network bank and the customer the physical location of either party is irrelevant. In the second section, we identified two reasons for public intervention: bank runs and the asymmetric information between banks and their depositors. Here, we identify three questions with regard to the supervision and regulation of network banks. The questions are not outlined in detail. They should be read with the previous discussion in mind.
Could the provision of financial services via public computer networks increase the need for banking regulation and supervision?
An increase in public intervention would be necessary if network banks were more vulnerable to bank failures, bank runs, and systemic risks. Consider the potential for bank runs first. Network bank are more vulnerable to bank runs for the following reasons:
The second reason for market failures is asymmetric information between the bank and a large number of free-riding depositors and the incentives of bank management and equity holders to gamble for resurrection when the bank is in trouble.
In summary, network banks are more vulnerable to bank runs because news and rumors spread quickly over public computer networks. The main source of instability could be security problems. Network banks have also more options to avoid outside interference in order to gamble for resurrection because they can credibly threaten to move their business. As a consequence, network banks should be regulated and supervised at least to the same degree as ordinary banks. In the initial stage, before network banking matures, the potential for bank runs suggests that governmental authorities should monitor network banks as closely as possible and provide information to the public. This would be also to the advantage of network banks because it would give the public more confidence to use these banks, which in turn would help network banking to take off.
An increase of public intervention would be also necessary if network banks would facilitate money laundering and other criminal activities. As suggested by an anonymous referee, there is a widespread fear that the increased speed and volatility of national and international monetary transaction could be used by criminals to break audit trails, etc., especially if the criminals own the bank. According to a Reuters report (PARIS, 2.12.98), "The future of criminal money laundering appears to be in cyberspace, officials say, as electronic money cards and the Internet provide an alternative to the methods now in use. Gold, casinos and real estate are now favored ways to cleanse big sums of dirty money right now, replacing banking channels that get clogged by tight regulation, the world's top authority on money-laundering said Thursday. The 26-nation Financial Action Task Force (FATF) [. . .] warned that in the future, modern technologies such as the Internet and electronic banking cards might prove ideal camouflage for criminals looking to turn ill-gotten gains into respectable funds. To keep pace with the fast-moving money launders, FAFT said it would carefully monitor the Internet and so-called electronic purse systems, whereby cash is passed from person to person via electronic chips, leaving no audit trail in its wake."
How does the international dimension of network banking potentially affect supervision and regulation?
Existing regulations for retail banking, deposit insurance, and lender-of-last-resort facilities are designed to deal with domestic or branches of foreign retail banks with domestic depositors. In the first section of this paper we concluded that a customer's marginal costs of buying financial services are independent of the network bank's physical location. Thus, in the absence of legal restrictions, network banks could attract a large number of foreign customers, giving rise to the following issues for supervision and regulation:
According to Baltensperger and Dermine (1990, p. 18) international
banking raises two specific issues:
The first concerns supervision and regulation. Does domestic regulation apply to other banks operating in the country ("national treatment principle") or does it apply to the foreign component of domestic banks ("home country principle")? The second issue concerns the extent of responsibility of the domestic lender of the last resort and of the domestic insurance system. Do they cover branches or subsidiaries of domestic banks operating abroad? Do they cover branches and subsidiaries of foreign banks operating domestically?
In the international banking context, one specific issue is to determine which country is responsible for regulating and supervising foreign branches of banks. The United States, for example, applies the "national treatment" principle to foreign banking organizations operating in its jurisdiction. In contrast, the European Commission has opted for the "home country principle."
Network banking and international banking differ in one important respect: network banks do not establish branches. Consequently, regulators of network banks face a different problem. They would either have to supervise and regulate banks in their jurisdiction that mainly attract funds and lend to foreign residents. Or, they would have to supervise and regulate network banks established in a foreign country with a large number of domestic customers. Thus the question is, Who is going to regulate a network bank with mostly foreign customers? The difference between international banking and network banking is depicted in Graph 1 and Graph 2 below.
In the context of network banking, the problem is to determine which country -- in terms of efficiency -- is responsible for supervising and regulating a network bank that attracts mostly small depositors and borrowers from abroad. A related question is, which country should provide the safety net? For example, will it be necessary and, if so, feasible to insure a large number of citizens who have uninsured deposits abroad?
When a domestic network bank has mostly foreign depositors, the incentives of domestic regulators to supervise the bank or to step in and provide lender-of-last-resort assistance when the bank is in trouble could be affected. Consider, for example, a Swiss network bank that attracts a large number of small German depositors and borrowers. When the bank fails, the Germans would be hurt while Switzerland's reputation as a safe haven would suffer. If the bank were sufficiently large, both countries would have an incentive to step in as a lender of last resort. However, the German Bundesbank would prefer that the Swiss National Bank do the dirty work, and the Swiss National Bank would like to see the German Bundesbank step in. In contrast, consider the failure of a Swiss bank subsidiary operating in Germany. The failure of this bank would hurt not only German customers but also Germany's reputation as a safe haven. It would also, but to a lesser extent, damage Switzerland's reputation for secure investments.
While this is an extreme example, it shows that the incentives of regulators and lender-of-last-resort agencies could respond differently when a bank's customer base consists mainly of foreign residents. It also shows that supervision and regulation of network banks is inherently an international policy issue. To create a stable financial environment, international coordination is essential.
The Internet, particularly the World Wide Web, is changing our lives. Public computer networks are part of today's technology that drives the information revolution -- an analogy to the Industrial Revolution of two centuries ago. This analogy stresses the common feeling that some significant and exciting developments are underway. The financial sector could be radically changed by information technologies, and public computer networks could play a major role in this turmoil. The financial services sector could be affected comparably more than any other sector in the economy because financial services are information commodities which can be easily traded on public computer networks.
Two results emerge from this analysis.
A major conclusion of the banking literature is that the need for public intervention comes from the potential instability of banking markets (Baltensperger and Dermine, 1990). Lack of enforcement and lack of adequate regulations could adversely affect the financial system's stability. Despite its importance, there exists no systematic research on the economic implications of financial intermediation on public computer networks. This paper is a first incomplete step in analyzing the economic implications of network banks. Further analysis will be valuable in identifying emerging problems, and it will be a base for future policy recommendations. In particular, the emergence of network banks requires revision of current national and international banking regulations. It is important that new regulations be based on reliable analysis to avoid either overreaction that would stifle these innovations or neglect the potential for instability in the world of banking.
* I would like to thank an anonymous referee, Daniel Rubinfeld, Heidi Seney, and Kevin Siegel for helpful suggestions and the Swiss National Science Foundation for financial support. I would particularly like to thank Edward J. Valauskas who published a slightly different version of this paper on the Internet at http://www.firstmonday.dk.
4. An anonymous referee has stressed this point.
5. See, for example, Baltensperger and Dermine (1987; 1990), Dewatripont and Tirole (1994), and Vives (1991). For a recent survey of modern banking theory literature, consider Bhattacharaya and Thakor (1993).
6. An anonymous referee has suggested adding this reason for government intervention.
E. Baltensperger and J. Dermine, 1987. "Banking Deregulation in Europe," Economic Policy, Volume 2, number 4, pp. 63-109.
E. Baltensperger and J. Dermine, 1990. "European Banking, Prudential and Regulatory Issues," In: J. Dermine, (ed.), European Banking in the 1990s. Oxford: Basil Blackwell.
S. Bhattacharaya and A. Thakor, 1993. "Contemporary Banking Theory," Journal of Financial Intermediation, Volume 3, pp. 2-50.
M. Dewatripont and J. Tirole, 1993. "Efficient Governance Structure: Implications for Banking Regulation," In: C. Mayer and X. Vives, (eds.), Capital Markets and Financial Intermediation. New York: Cambridge University Press.
M. Dewatripont and J. Tirole, 1994. The Prudential Regulation of Banks. Cambridge, Mass.: MIT Press.
D. Diamond and P. Dybvig, 1983. "Bank Runs, Deposit Insurance, and Liquidity," Journal of Political Economy, Volume 91, pp. 401-19.
E. Fama, 1980. "Banking in the Theory of Finance," Journal of Monetary Economics, Volume 6, pp. 39-57.
M. Hellwig, 1991. "Banking, Financial Intermediation and Corporate Finance," In: A. Giovannini and C. Mayer, (eds.), European Financial Integration. New York: Cambridge University Press.
R. Kalakota, 1996: The Impact of Cybercommunications on Traditional Financial Services, http://www.ctr.columbia.edu/citi/cybcompap/ravi.htm
D. Neven, 1990. "Structural Adjustment in European Retail Banking: Some Views from Industrial Organization," In: J. Dermine, (ed.), European Banking in the 1990s Oxford: Basil Blackwell.
X. Vives, 1991. "Banking Competition and European Integration," In: Giovannini and C. Mayer, (eds.), European Financial Integration. New York: Cambridge University Press.
L. H. White, 1996. "The Technology Revolution and Monetary Evolution," In: The Future of Money in the Information Age. Cato Institute's 14th Annual Monetary Conference, http://www.cato.org/moneyconf/14mc-7.html