- Omotunde Johnson
- Published Date:
- April 2002
Professor Hamada has written a paper in which he has tried to cover a varied set of issues of great importance when discussing the welfare cost of systemic risk, financial instability, and financial crises in general. In the process, he demonstrates the immense difficulties in trying to be concrete and precise and the dilemma faced by public policy. I shall comment directly on his paper but focus more on the principles that have been guiding the public policy response in this area.
Nature and Sources of the Risks
Systemic risk is the risk that one or more important financial or nonfinancial firms or organizations will fail due to serious credit or liquidity problems and, in the process, will trigger a chain of events causing serious disruption of production, income loss, and debt settlement problems in the economy, ensuing ultimately in suboptimal intertemporal consumption patterns. As Furfine (1999) notes, it is useful to distinguish between the risk that some financial shock causes a set of markets or organizations simultaneously to fail to function efficiently from the risk that failure of one or a small number of firms or organizations will be transmitted to others due to explicit financial linkages. Those concerned with banking crises tend to focus on the first type of systemic risk and those focusing on payment systems tend to be concerned mainly with the second type of systemic risk.
Systemic risks can arise from macroeconomic factors (demand or supply shocks, poor fiscal or monetary management), from poor governance in financial firms (excessive risk taking, miscalculations, criminal activity), from technical failures (including those resulting from design flaws of major systems), and from human error. Some of the macroeconomic factors could be propelled by natural forces (like weather) and some by coordination failures in market processes. As Professor Hamada rightly points out, there is reason to worry that, increasingly, systemic crises due to deliberate acts (security breaches, vandalism, and electronic tampering) could become quite important.
Deliberate acts, especially via the Internet, have made difficult the probability calculations as well as the determination of the equilibrium point at which the marginal benefit of an additional dollar spent on, say, security measures will be no greater than the marginal social cost if the feared event occurs. These acts, again especially those that are Internet related, also raise important questions about the public good nature of the preventive and resolution policies and measures to address systemic risks—an issue quite well known to those working in the area of payment systems. Hence, these acts raise important questions about the distribution among members of the society of the preventive and resolution costs incurred.
As regards financial instability and the related financial crises, as Professor Hamada again notes, the reference, typically, is to events such as currency crises, monetary instability, and sharp exchange rate fluctuations. Financial instability can arise from fluctuations in real demand and supply in the domestic or world economy, as well as from unstable domestic monetary and fiscal policies. The consequences are unpredictable movements in commodity prices, interest rates, and exchange rates.
Financial crises, Professor Hamada states, are sudden and precipitous movements of financial variables—especially those that are unexpected and unidirectional. In a sense, therefore, a financial crisis can be an instance of financial instability, perhaps aggravated by a coincidence of other unfortunate developments, such as banking system crises. A financial crisis can also be a systemic crisis. Hence, it is possible to think of financial crises as being the more inclusive terminology.
Nature and Measurement of Welfare Costs
There are many ways one can approach the discussion of welfare costs. The costs ultimately flow from the fact that financial crises are unpredictable. From that perspective, the assessment could begin with a calculation of the welfare costs of living in an uncertain world. One can then decide to take an accounting approach. For instance, one can begin with the cost of calculation (including forecasting and risk measurement and management), insurance, and the resources devoted to protective measures (security, backups, etc.). Then one can add a pecuniary component for the psychological cost of coping with uncertainty.
Even with a fair amount spent on calculation, the welfare costs of living in a world in which crises are unpredictable in their timing, severity, and duration must take into account the cost of behavior (panics and runs) during crises. As Professor Hamada puts it, anxiety can be contagious and a chain of mistrust and fear can engender a chain of failures and bankruptcies. Professor Hamada invokes the concept of a “contagion multiplier,” which I would regard as a factor one could use—as an additional safety valve—to augment the amounts spent on calculation, insurance, and protective measures with a view to reducing the probability of financial crises and the severity and duration of crises when they occur.
If society is going to organize itself to reduce the probability of financial crises and the cost when crises occur, then it is important to understand the nature of the costs of the crises when they occur as well as the relationship between expenditures on calculation and protection, on the one hand, and the probability of crises, on the other. We have now entered the difficult world of efficient policymaking, including institutional organization, for financial crisis prevention and resolution. As an example, Professor Hamada draws attention to something well-known to those interested in payment systems: financial instability caused by rapid structural changes and other shocks to the financial system make difficult the task of monetary policy. Developing the appropriate capacity to deal with this kind of world is part of the welfare cost of financial instability.
Once a crisis does occur, one must add to the above-mentioned costs of living in a world in which financial crises are unpredictable in their timing, severity, and duration the disruptions in production and intertemporal consumption patterns that are engendered by an actual crisis. Professor Hamada suggests looking at this welfare cost in terms of consumer surplus using the Harberger triangle, an approach with much merit but also limited by its static nature and the difficulty of quantification in pecuniary terms. It is therefore not surprising that when he tries to quantify the costs of systemic crises and financial crises, Professor Hamada looks mainly at the income (GDP) losses and does not tie those in with the consumer surplus approach. But I do feel also that the truly general approach would entail a sort of enhanced Harberger approach.
Thus, for the case of systemic risk, when Professor Hamada asks, what is the probability of a crisis and if one occurs what will be the loss to the economy, he illustrates the magnitudes under highly simplified assumptions. And, of course, the welfare loss from a systemic crisis depends on the probability of the crisis occurring, the duration of the crisis, the recovery rate of defective claims and other damages, and the availability of facilities to dampen the disruption (such as lender of last resort facilities). Similarly, in addressing financial instability and financial crises in general, Professor Hamada takes into account mainly the income loss from the financial crises.
The welfare costs of financial instability poses some paradox. The fluctuations give rise to the same costs of living in an uncertain world as discussed before. But, as Professor Hamada indicates, price fluctuations can have beneficial welfare effects, due to the convexity of expenditure and profit functions with respect to prices. Hence, ultimately the welfare costs of price and financial fluctuations come from the fact that the variations are not easily predicted. Professor Hamada notes that the changes in financial variables during a financial crisis can be so abrupt and drastic that they can force changes in the monetary system itself. When the crises are worldwide, they can force changes in the international monetary system. Again, contagion effects that have a psychological content (“contagion of anxieties”) are present: the public loses trust in monetary policy, in the international monetary regime, and in the payment system itself.
An interesting and valuable point made by Professor Hamada in this regard is that the welfare cost of financial crises will depend on the monetary regime. He proceeds to give a brief and general discussion of how to look at the welfare costs of financial crises under different monetary regimes and “situations”—seven in all—that are currently highly topical, including currency board and dollarized systems. His underlying message is that there are opportunity costs of adopting a particular monetary regime. The clear implication is that countries should bear this in mind, and in assessing the welfare costs of financial crises they should try to isolate those costs that are traceable to the monetary regime they have chosen. Unfortunately, it is difficult for a country to rank (international) monetary regimes on the basis of the country’s potential economic performance under the different regimes, including the propensity for financial crises or the severity of financial crises. This is no surprise given the difficulty of controlling for some of most important causal factors involved. In any event, Professor Hamada does not go deeply into this issue and nor will I.
No doubt the welfare costs of financial crises would depend on other country-specific factors such as the degree of integration in financial markets, of interlinkages in the payment system, and of concentration in the financial sector. In addition, the interaction of these factors with the degree of liberalization in financial markets, including international capital mobility, in determining the probability and severity of financial crises for any given set of preventive measures is also a constant issue for public policy. More generally, once again, the ultimate challenge is determining the socially optimal mix of preventive, insurance, and resolution policies.
From a public policy perspective, why the interest in the welfare costs of systemic risk, financial instability, and financial crises in general? The obvious answer, is, first, that we want to know if the risks are worth worrying about. If so, second, we would like to investigate how to design socially efficient policies and measures to reduce the incidence (that is, the probability) and the severity of the crises. Third, when crises do occur, we would like to have in place policies and measures that mitigate the wealth effects and the distributional effects that have adverse social, political, and economic consequences. Optimal policymaking must address all three problems simultaneously. For instance, the resources that that are socially efficient to expend in designing and implementing policies to prevent crises from occurring would be affected by the cost of policies that can mitigate the effects of crises when they do occur and the probability of occurrence of the crises if no preventive policies are put in place.
From a fundamental public policy perspective, there is a clear consensus that we do need to worry about the risks, especially systemic risks. In response, the challenge is to design policies that are welfare enhancing. For it is very easy for the policies to be welfare reducing for a number of reasons. In particular, (1) they could be ineffective in achieving their objectives and hence the resources devoted to their design and implementation would be a deadweight loss; (2) they could inhibit private initiative and dampen incentives to create wealth; and/or (3) they could be far more costly to design and implement than the effects of the crises they prevent. So the question becomes, how can society ensure that it design socially efficient policies to address the problems that can result from inadequate and inappropriate attention to systemic risks, financial instability, and financial crises in general?
I would posit that at the analytical core of the appropriate policy response is a tendency toward putting in place measures to (1) ensure that there is a clear understanding of the risks, especially by those who are the sources of, or have some control over, the risks; (2) assign clearly the responsibility for managing the risks; and (3) provide adequate and appropriate incentives to those responsible to manage the risks in a socially efficient way. In a sense, we don’t know the optimal solution ex ante, so we create an environment that will tend to a solution that is close to the optimal; one can call this “an enabling environment.”
Clear understanding is facilitated by systematic analysis of the risks and how they arise and then communicating this information to the relevant parties in a language that those parties can understand. Alternatively, it could be a requirement for obtaining the right to engage in some risky activity that a “person” demonstrate a clear understanding of the risks involved. For instance, one of the Core Principles for Systemically Important Payment Systems in the Report of the Task Force on Payment System Principles and Practices of the Committee on Payment and Settlement Systems says that a system’s rules and procedures should enable participants to have a clear understanding of the system’s impact on each of the financial risks they incur through participation in the system (Committee on Payment and Settlement Systems, 2001). In the banking area, examination of the internal models and process used in the management of a bank’s risk should provide a reasonably informative picture of the bank’s understanding of the risks it faces.
In deciding how responsibility for managing risks should be assigned, so as to facilitate social efficiency and wealth creation, an approach that has been found useful is to attempt to assign the responsibility for risk management according to comparative advantage in managing such risks, as this greatly reduces the social cost of risk management. It is sometimes argued that many legal systems that have emerged through a long evolutionary process, based on norms and customs of society, have tended to obey this principle. From this perspective, laws are frequently mere formalizations of actual and desired norms, with the advantage that having laws also helps to resolve the problem of who will bear the cost of enforcement.
Since Coase (1960), many economists have tended to argue that, in the absence of transaction costs, and with freedom of contract, it really does not matter how rights and obligations (and, by the fact itself, risks) are initially assigned. Secondary reallocations will take care of apparent inefficiencies in the initial allocation. The initial assignments could, of course, engender important wealth effects. Also, with sufficiently high transaction costs, secondary reallocations could be frustrated. Nevertheless, it would appear that, for a payment system, for example, freedom of the system participants as a group to set the rules, and freedom of individual participants to negotiate among themselves the assignment of responsibilities for managing the risks that arise in their relations with each other, facilitate social efficiency.
Little wonder, then, that in assigning responsibility for management of risks, permitting freedom of contract within “appropriate” limits1 is often fostered. This would seem particularly pertinent for today’s financial world, as it would give flexibility in rule making in a rapidly changing financial and technological environment. In other words, the argument is that, given the basic legal and general institutional framework, participants in a financial system, individually and as a group through their governing bodies, should sometimes be given broad freedom to formulate mutually advantageous rules that are enforceable in all relevant jurisdictions. This would permit the design of rules that take into account special circumstances and the modification of those rules on a timely basis, in response to changing circumstances or new opportunities.
Freedom of contract can greatly help provide adequate and appropriate incentives to those responsible for managing the risks in a socially efficient way. But easily the most important principle in creating incentives for efficient risk management would seem to be fostering internalization of costs and benefits. This principle, which is particularly useful in property rights formulation,2 is also important in determining liability rules and assigning risk management obligations. Complete internalization is fostered by a cost-reward structure in which the values created by any particular activity are made to accrue to those who bore the cost of undertaking the activity. Each individual’s wealth is made to depend upon the net value created by the resources under that individual’s control. Improving the institutional framework by more clearly defining property rights, opening markets, and allowing greater competition, are some of the ways of fostering internalization. Complete internalization means no divergence between private and social marginal costs and benefits and thereby promoting social efficiency in the allocation and use of resources.
Internalization provides private agents with the incentives to find efficient methods to manage risks assigned to them; and market mechanisms also evolve in ways that enable agents to pool and shift risks in well-known ways. Examples include insurance (risk pooling); developing internal risk management processes and models (to design and implement policies like diversification, credit ratings, early warning signals); and developing price forecasting capability. Moreover, specialized agents in bearing risks lead to arrangements such as forward contracts to mitigate the psychological costs of financial instability.
It is not always possible or efficient to attain complete internalization, due to the high cost of internalization or unquantifiable elements of costs and benefits (such as some of the value placed on “convenience” or “safety”). A practical approach to facilitating social optimality could then be simply to ensure that institutional and organizational arrangements (rules and governance arrangements) are put in place that tend to promote social optimality and then to conclude that any residual divergences observed between private and social (marginal) costs and benefits, at the equilibrium point of private choice, reflect the impact of unquantifiable factors and of internalization costs.3
Developing standards, via some process of consensus, is one of the ways countries foster adequate and appropriate incentives for socially efficient management of risks. A slight problem is that standards in general limit freedom of contract; hence the question of where or how to draw the line and set limits to freedom of contract becomes important. Ultimately, a balance must be struck between the positive effect on incentives of freedom and flexibility to alter policies, practices, and procedures, and the negative effects on third parties of not constraining the contracting parties from exercising this freedom in ways that have adverse effects on third parties without appropriate compensation.
We would argue that when the case for freedom of contract is placed next to the case for limiting this freedom, the conclusion as to how to draw the line often, but not always, appears straightforward. In the case of payment systems, for example, it would appear that the participants in a system should be allowed to determine the rules, and hence the policies, practices, and procedures governing the system, and some oversight agency representing the public interest should assess the rules to ensure that they conform to, say, the core principles for systemically important payment systems.
In other words, when it comes to the policies, practices, and procedures being implemented by a payment system, there should be neither a regime of only self-regulation nor one of strict regulatory requirements (interpreted as rules and regulations imposed as minimum requirements by some public sector agency that also supervises and inspects the systems to ensure strict compliance). The participants would agree on general guidelines on risk management. Some of these guidelines would involve centralized management and others would involve decentralized management. Where decentralized management is involved, participants would be allowed to work out their own solutions and the governing body of the payment system would then assess the solutions to ensure that they conform to the guidelines and standards set by the organization. Thus, for example, participants would be allowed to determine their bilateral exposure limits through bilateral negotiations. I believe that, in this particular area, this is in fact the typical approach.
Even abstracting from the risks due to intentional acts, it is doubtful that the efficient social policy is to so protect a system as to reduce the probability of crisis to zero. Recognizing this reality, public policy tends to have measures in place to address crises when they do arise, thereby mitigating the adverse welfare effects. Liquidity support (typically from the central bank) and fiscal support from the government summarize the most important of these policy responses. The basic analytical issues involved in considering fiscal support will be addressed in a later session of this seminar, even if indirectly. Liquidity support often is seriously considered when the system faces some liquidity shock.
Liquidity shocks of a systemic nature (including those that give rise to unexpected credit squeeze) could be payments related. Hence, the monetary authorities have taken a keen interest in the measures that payment system operators have, or can, put in place to reduce the likelihood that the payment system would be the origin of liquidity shocks in the financial system—thereby forcing unplanned and inflationary injection of base money or reduction of interest rates by the monetary authorities. In addition, of course, the monetary authorities are equally concerned about preventing unwarranted deflation resulting from liquidity shocks while avoiding the moral hazard problem of being seen as a guarantor always ready to provide enough liquidity to prevent gridlock or a breakdown of the payment system or, indeed, bank failures for other reasons. In this respect, concerns about inflation matter, but because of the monopoly of the central bank in the production of the legal tender—the medium of exchange and the ultimate means for settling debts—consideration of the adverse welfare effects of liquidity shocks4 may be unavoidable.
In reality, it is probably the case that most liquidity shocks do not originate in the payment system, even though they end up putting pressure on the payment system in view of their impact on the ability of individual financial institutions to settle their obligations after clearing. This, among other things, motivates country authorities to introduce measures to improve the soundness of financial systems. Hopefully, then, as bank licensing becomes more stringent, banking supervision and governance of banks improve, and as bank branching spreads, liquidity shocks emanating from within the financial system but outside the payment system will decline in significance. Liquidity shocks emanating from outside the domestic financial system—abroad or due to other domestic factors—will continue to occur, of course.
In short, the approach of policymakers has been to introduce policies to improve the soundness of the financial system while, at the same time, instituting measures to reduce settlement risks in the payment system by ensuring payment finality5 without special intervention by the central bank. Central banks may have become seen increasingly as guarantors of safety of payments settlement, because of legal obligations of the central bank to promote monetary and financial stability and (related to this) because of a growing consensus that final settlement of clearing systems should take place normally on the books of the central bank. Here, the dominant consideration is risk—in this case systemic risk. In reaction, central banks have had to work toward ensuring the existence of measures to contain settlement risks in payment systems and thereby reduce the chances that they would be called upon to “bail out” a payment system to prevent a systemic crisis.
Coase, Ronald,1960, “The Problem of Social Cost,” Journal of Law and Economics, Vol. 3 (October), pp. 1–44.
Committee on Payment and Settlement Systems, 2001, Core Principles for Systemically Important Payment Systems, Report of the Task Force on Payment Systems Principles and Practices (Basel: Bank for International Settlements).
Furfine, Craig H.,1999, “Interbank Exposures: Quantifying the Risk of Contagion,” BIS Working Paper No. 70 (Basel: Bank for International Settlements).
Johnson, Omotunde E.G.,1998, “The Payment System and Monetary Policy,” IMF Paper on Policy Analysis and Assessment 98/4 (Washington: International Monetary Fund).
Johnson, Omotunde E.G., and others,1998, Payment Systems, Monetary Policy, and the Role of the Central Bank (Washington: International Monetary Fund).
The appropriate limits to freedom of contract is one of the many contentious areas in law and economics. For an introduction see, for example,Trebilcock (1994).
Property rights are rights over assets: rights to use an asset (user rights); rights to reap income from the asset and to contract over the nature and amount of this income; and rights to transfer to another party the ownership or user rights to the asset.
See Johnson and others (1998), pp. 67-68.
These adverse welfare effects, again, include disruption of production and the inducement of suboptimal intertemporal consumption patterns.
Payment is final when it becomes irrevocable and unconditional.