The following publication has been lightly reedited for spelling, grammar, and style to provide better searchability and an improved reading experience. No substantive changes impacting the data, analysis, or conclusions have been made. A PDF of the originally published version is available here.
In December 2000, the Federal Reserve Bank of Chicago and the Financial Stability Forum cosponsored a symposium on “Designing an Effective Deposit Insurance Structure: An International Perspective.” The symposium was intended to generate informed feedback on its recommendations to countries introducing or modifying deposit insurance schemes. Representatives of the Financial Stability Forum discussed their analyses and recommendations with a select group of leading financial and banking economists. The purpose of this Fed Letter is to summarize the discussion at the symposium and the resulting conclusions.
The Financial Stability Forum and the Working Group on Deposit Insurance
Following a number of recent disruptions in international financial markets (the Asian crisis, Russian devaluation, and systemic concerns resulting from Long-Term Capital Management) the Group of 7 set up the Financial Stability Forum (FSF) in 1998. Comprised of international regulatory and monetary authorities, the FSF was directed to study means to manage risk in the global financial system. This was to be accomplished by developing standards and codes of good practice, improving the information exchange among financial regulatory authorities, evaluating the existing vulnerabilities within the international system, and developing priorities and programs which best address perceived weaknesses in these markets.
In an attempt to address industry fragilities, the FSF has organized working groups to address specific areas of concern. Past groups have evaluated highly leveraged institutions, the volatility of capital flows and excessive short-term external indebtedness, offshore financial centers, and international supervisory/regulatory standards. The most recently formed working group of the FSF is concerned with deposit insurance issues. In 1999 a study group was created to evaluate the contribution that deposit insurance can make to financial stability. The study group became a formal FSF working group in April 2000 and includes experts from several countries, the International Monetary Fund, and the World Bank. The objective of the working group is to provide guidance on the elements of effective deposit insurance systems for countries considering the adoption of a limited-coverage deposit insurance system or the reform of an existing system. The guidance will emphasize practical assistance and will cover such matters as public policy objectives and the basic conditions that should exist within a country wishing to put a system in place.
Recently the group has begun a series of outreach meetings in different regions of the world with parties interested in deposit insurance issues. In December 2000 one such meeting was held at the Federal Reserve Bank of Chicago that included working group members, industry experts selected to critique the working group’s efforts to date, and other interested parties mainly from the academic community. Members of the working group discussed four major areas: 1) the public policy objectives for deposit insurance, 2) moral hazard, 3) ensuring a country’s readiness for deposit insurance, and 4) the transition of a country from blanket coverage to limited protection.
J. P. Sabourin, Canada Deposit Insurance Corporation and Chairman of the FSF Working Group on Deposit Insurance, began the symposium by discussing the approach that the working group had taken to achieve its mission. He argued that the purpose of the group was not to develop a set of standards or best practices that countries would be encouraged to use in initiating or adjusting their deposit insurance scheme. Nor was the approach to be taken to encourage or discourage countries to introduce deposit insurance. Indeed, most of the group’s research suggested that a “cookie-cutter” approach would not suit all countries, but rather that the details would depend on the characteristics of the country. There was no single best approach. This was an interesting introduction to the symposium because much of the critique that followed often addressed these very issues, with working group members indicating that making such recommendations would go beyond the mandate provided to them by the FSF and discussants arguing that the mandate should be expanded.
Insurance objectives and infrastructure requirements
Greg Cowper, Canada Deposit Insurance Corporation, presented the working group’s analysis of public policy objectives for deposit insurance.1 The report was the result of a thorough review of the academic literature and a survey of numerous countries that had introduced deposit insurance schemes. The group was interested in finding differences and similarities in objectives across countries, in finding trade-offs between various objectives, and in determining how these affected the structure of deposit insurance schemes introduced. Finally, the group would analyze the findings and make some limited recommendations.
The findings of the literature review and survey of individual countries were not particularly surprising. The typical public policy objectives of deposit insurance were categorized into three groups: to contribute to financial system stability, protect less financially sophisticated depositors, and enhance the ability of regulators to achieve other related policy objectives. Other objectives included developing formal mechanisms for resolving failed banks, contributing to an orderly payments mechanism, and avoiding or quickly resolving financial crises. Alternative policy objectives included increased competition resulting from decreasing barriers in the deposit-taking industry, increased concentration of the cost of bank failure toward the banking industry, enhanced economic growth, the ability to help move countries away from a 100% explicit deposit guarantee toward a regime with limited coverage, and the ability to enact additional bank legislation and regulation. While system stability and protection of the less-sophisticated investor are standard textbook objectives of deposit insurance, the objectives of the last general category were sometimes less traditional.
Although there were similarities in the survey responses, there did appear to be differences in objectives across countries. Often these differences resulted from country-specific circumstances concerning the financial infrastructure; i.e., the set of legal, regulatory, and accounting institutions that support financial markets and financial intermediation.2 The more developed the financial markets, the more common it was for the private sector to have a greater role in the administration of deposit insurance and greater was the reliance on market discipline. Based on their analysis, the working group recommended that countries should carefully delineate the objectives of deposit insurance, tie them to the abilities of the countries, and ensure that the structure of the insurance scheme is consistent with the stated objectives.
Jose Carlos Jaime, Seguro de Depositos Sociedad Anonima, Argentina, presented the working group’s analysis of how the deposit insurance scheme should fit within other components of the country’s safety net. As stressed earlier by Sabourin, Jaime reiterated that the group’s analysis started from the premise that a country had decided to establish an explicit deposit insurance system. It was not the group’s objective to argue the advantages and disadvantages of doing so. Given that this decision had been made, how then should the deposit insurance program operate relative to the lender of last resort function and the activities of the bank supervisor?
Obviously, the performance of any deposit insurance scheme will be affected by the operation of the lender of last resort and the effectiveness of supervision. While the presence of an insurance system is not adequate to prevent the spread of systemic risk, it can decrease the spillover effects and their associated costs. But the insurance system must work in harmony with other aspects of the safety net. If the supervisor is slow to respond to troubled banks, the cost of intervention for the deposit insurance fund is larger. While the lender of last resort can resolve liquidity problems, extending its use to insolvent institutions can also result in additional costs for the insurance fund. Thus, there are certain requirements of each element of the safety net if the three elements are to function most effectively.
The lender of last resort attempts to avoid having the problems of individual institutions spill over to others. However, to accomplish this without generating future monetary problems, the central bank must have an adequate stock of market securities to sterilize any liquidity injections. The more difficult it is for a central bank to sterilize liquidity injections, the more negative the potential effect of a liberal rediscounting policy.
Bank supervision is intended to avoid errant behavior of individual banks. This is accomplished by monitoring bank behavior and imposing constraints. While some constraints are imposed after a bank has engaged in errant behavior, the realization by banks that constraints will be forthcoming often serves as a source of discipline on bank actions before it engages in the behavior.
Finally, how can deposit insurance best interact with these other two safety net functions? It was argued that it was important to realize what exactly was being protected by deposit insurance—was it the banks, bankers, depositors? While the trend has definitely moved away from protecting the banker, and was legislatively mandated in the U.S., there has been a movement to protect banks for fear that the failure of individual banks may result in systemic problems. This has not enhanced the effectiveness of deposit insurance. If a supervisor realizes that failures will always be covered, then there may be a tendency to delay corrective action resulting in unnecessarily high resolution costs.
The discussants for the first two presentations were Carl Tannenbaum, ABN AMRO; Randall Kroszner, University of Chicago; Gordon Roberts, York University; and Robert Bliss, Federal Reserve Bank of Chicago.3 Tannenbaum argued the potential benefits from deposit insurance were significant. By decreasing volatility and increasing consumer confidence, the potential for economic growth would improve. He questioned, however, whether all the desired objectives could be achieved without having 100% coverage. While the smaller, less-sophisticated investor could be protected by partial coverage, the lack of perfect information in markets would occasionally lead to market disruptions. While 100% coverage could address these concerns, it would also create moral hazard problems that would offset some of the potential benefits.
Kroszner argued that while he agreed with many of the general ideas in the papers and presentations, he thought the discussion was too general and ignored many of the trade-offs involved with various elements of the safety net. He stressed the need for effective corporate governance mechanisms to ensure that bank management has the correct incentives to prudently manage risk. Failure to introduce such systems can lead to inappropriate cross-subsidizations and poor risk-management procedures. Unless sufficient precautions are taken, one can end up doing more harm than good by introducing a poorly structured insurance scheme.
Roberts and Bliss stressed the problems involved with introducing effective insurance schemes. Countries attempting to limit an existing scheme may find it more difficult to reconstruct the existing program than to introduce a new scheme where one did not previously exist. Thus, optimal changes may differ across countries. To the extent possible, they recommended the incorporation of market discipline when introducing or modifying existing schemes through mechanisms such as coinsurance and limited coverage.
Issues in implementing deposit insurance systems
The second session of the symposium emphasized the problems involved with introducing deposit insurance arrangements, particularly moral hazard issues and the problems resulting from decreasing coverage in an attempt to rely on alternative disciplining mechanisms. George Hanc, Federal Deposit Insurance Corporation, discussed the moral hazard problems that are inherent to any deposit insurance program. He argued that corporate governance, market discipline, and regulatory discipline need to be incorporated simultaneously. To do so, however, required an infrastructure that was sufficient to allow these mechanisms to be effective. For example, market discipline requires that market participants have sufficient access to meaningful information about the bank. Thus, accounting standards and regulatory rules and guidelines (and perhaps mandatory disclosure) need to be established and enforced. Hanc stressed that a comprehensive regulatory package included a strong and knowledgeable supervisory group, prompt supervisory responses, meaningful capital requirements, and effective risk-management systems. He concluded moral hazard was inherent to any deposit insurance scheme. With care, one could limit its impact by utilizing some combination of market and supervisory discipline along with an effective corporate governance scheme. The proper mix between these three was probably country-specific.
Carlos Isoard, Board of Governors, Instituto para la proteccion al Ahorro Bancario, discussed the problems involved with moving from blanket coverage (100%) to limited coverage, an objective which was thought to be desirable, but difficult to achieve. Isoard reiterated the message of other working group members indicating that many transition issues were country-specific and did not lend themselves to general guidelines. He argued that it was important for the parties implementing the transition to have credibility and to provide a clear set of deliverables concerning the new program’s objectives, coverage, funding source, and institutional arrangements. The actual specifics could vary significantly across countries depending on the situation in each country and would be dependent on things such as the economic and political framework, the existing regulatory and accounting environment, and institutional arrangements. For example, the decision as to whether the new insurance system should be managed by an umbrella supervisory agent or an independent institution with no other regulatory responsibilities would vary according to the current situation in each country. Similar differences could exist with respect to the decision to have ex ante or ex post funding, the optimal size of the fund, its investment policy, whether the funding was on or off budget, etc. Isoard concluded by summarizing the approach taken by Mexico to transition to a limited coverage environment. The coverage reduction period began in 1999 and is scheduled to end in 2005.
The discussants for this session were Robert Eisenbeis, Federal Reserve Bank of Atlanta; George Kaufman, Loyola University Chicago; and George Pennacchi, University of Illinois at Urbana-Champaign. Eisenbeis stated that the working group had provided a comprehensive laundry list of items to be considered in implementing or moving to a new deposit insurance scheme, but there was little direction provided as to how to put it all together. He argued that by its very nature, deposit insurance introduces trade-offs. Hard decisions have to be made. How do you design a system to share the risk among the bank, its depositors, and the taxpayer? He thought that many of the group’s stated objectives were probably well beyond the scope of deposit insurance. For example, while deposit insurance can protect the unsophisticated depositor (a relatively minor problem in most countries) it cannot prevent macroeconomic instability. That is most often caused by poor macroeconomic policy. Fundamentally, what is it that deposit insurance should be trying to accomplish? Eisenbeis argued that the main objective should be to make bank failures independent events. While insurance programs may aid that process, other policies such as prompt corrective action probably play a more important role. He argued that what was most needed was the development of a financial infrastructure to enable private markets to discipline market participants, to have swift supervisory response when problems are encountered, and have established guidelines as to how losses are to be distributed when they occur.
Kaufman suggested that the working group be bolder and strengthen their recommendations instead of continually emphasizing country-specific solutions. While one size may not fit all, total flexibility may not result in the adoption of efficient programs. He also suggested that the group better incorporate the findings of the existing economic literature into their recommendations. Pennacchi argued that while the group had laid out rather sensible guidelines, he thought they could go further by analyzing how the private sector would address issues such as moral hazard and risk management. When possible, market solutions should be used to limit regulatory intrusion and encourage proper risk management. He argued that this solution might restrict the activities of banks. Financial innovation has allowed banks to take on additional risk using financial instruments that are hard for supervisors to monitor. Private markets would address this problem by increasing reliance on collateral and requiring that deposits be backed with relatively transparent assets. Supervisors should take similar actions.
Summary
There was general agreement by the discussants that deposit insurance reform is a difficult issue, and the working group should be praised for their willingness to take on the tough issues. They also stressed the importance of the topic and the need for a thorough understanding of the trade-offs involved with the objectives. Typically, one wants the forces of the marketplace to influence firm behavior, but regulators are most concerned about discipline resulting in bank deposit runs. While solvent banks with access to a liquidity source need not be concerned with these runs, regulators have shown they are concerned. This suggests that one needs to stress these standard trade-offs and evaluate evidence of the existence of such trade-offs. Obviously, deposit insurance, cannot solve all the problems—particularly those of conflicting objectives.
There were strong recommendations by the discussants to better ground the proposed guidelines within the existing research that has been done in this area. What theoretical and empirical evidence exists on the topics covered? Are the guidelines consistent with that work? Is the literature adequate? The working group should initiate and/or encourage others to conduct research as needed in the relevant areas. That work could parallel the current development of guidelines.
It was also argued that if the recommendations/guidelines are path-dependent, then the authors should say as much. The proposals are frequently not starting with a clean slate and trying to implement the most efficient insurance scheme. Rather, it was argued, the authors should explicitly recognize this and say, for example, that they are starting at, perhaps, step 4 of a 10-step process. If the process had started at step 1, it may have evolved differently and may have resulted in a more optimal insurance program. Then, however, there is a need to justify why step 4 was the appropriate starting point. Once these clarifications are made, the debate about the details can take place on more common ground. This may resolve many of the problems with different views concerning appropriate directions the working group should be taking.
There were also issues raised about drawing conclusions about the behavior of market participants if deposit insurance is introduced, based on behavior in the current environment. Behavior after a regulatory change may be very different from that prior to the change. The quality of the credit analysis, the level of monitoring taking place, the amount of disclosure, and the quality of data provided to investors would all change as market discipline is imposed and regulatory reform is introduced.
Finally, there were suggestions that although country-specific characteristics need to be taken into account in introducing a deposit insurance scheme, efforts should be made to find commonalties across countries. What common characteristics need to be introduced? What, if any, are the infrastructure requirements? The important items discussed appear to be related to the use of corporate governance, other aspects of market discipline, means to protect the deposit insurance fund, risk-based, forward-looking insurance premiums, clear coverage limitations and advanced decisions concerning closure policies. These were some of the major issues raised by the discussants. Accordingly, they should be some of the important issues evaluated going forward.
Notes
1 The papers discussed in this article and related material could be found on the Canada Deposit Insurance Corporation website at the time of the original publication of this letter.
2 See also Michael H. Moskow, 2001, “Financial infrastructure in emerging economies,” keynote address, Symposium on Intermediation: Banking in Emerging Markets, University of Michigan, Ann Arbor, MI, June 15.
3 Gordon Roberts was the scheduled discussant but was unable to attend because of inclement weather. Robert Bliss presented Roberts’ written comments and included additional comments of his own.