- Omotunde Johnson
- Published Date:
- April 2002
There is a wealth of interesting material in Mr. Mahoney’s paper. He has been very open and frank in letting us see how Moody’s assesses financial stress. I agree with much of what Mr. Mahoney said, and those who are being assessed should pay close attention. Even if they disagree, the fact that a major rating agency follows a particular approach, of itself, has implications for the markets and financial flows. While I can go along with much of the first part of the paper, however, I do have some doubts about Mr. Mahoney’s statement that these factors provide a necessary but insufficient indicator of vulnerability and, therefore, about some of the policy implications toward the end of the paper to which I will return later.
Mr. Mahoney suggests four factors that have predictive power in identifying vulnerability to stress: systemic solvency, accounting transparency, quality of supervision, and funding stability. Systemic insolvency is said to arise when a large proportion of a system’s banks is insolvent. In this connection, a useful distinction is made between “accounting insolvency” and “economic insolvency.” Moody’s focuses on the latter. Economic insolvency is defined as occurring when an institution’s unrealized credit losses exceed its capital, reserves, and realistic, near-term earnings prospects. This must involve Moody’s adjusting a bank’s reported figures for unrealized credit losses and/or near-term earnings prospects. The adjustments must be enormous, and pervasive, since Mr. Mahoney claims that no bank in the world shows in its annual accounts that it is insolvent, while with the suggested adjustments apparently one-quarter of the banks in the world are so.
This brings us quickly to the second factor: accounting transparency. Mr. Mahoney is certainly right in arguing that accounting transparency is key in making a good assessment of a banking system’s financial condition. The general lack of any prior awareness of imminent banking system problems in Asia has been widely attributed to problems of accounting transparency. At least as important, once the crisis emerged, the lack of accounting transparency undoubtedly led to serious delays in several of the countries before the magnitude of the crisis could be properly assessed and the appropriate set of remedial measures implemented.
One of the most interesting parts of Mr. Mahoney’s paper is his listing of some of the “accounting games” commonly used by banks: minimizing reported delinquencies; overvaluing collateral; establishing and financing unconsolidated subsidiaries; transferring problem loans to subsidiaries beyond the jurisdiction of the regulator; moving bad assets from branch to branch; and misclassification of problem loans. My first observation on this section is that this listing provides a good justification for the banks to be as transparent as possible in their reporting. Mr. Mahoney and his colleagues are aware of the various ways in which a bank’s true position can be disguised, and if a bank follows such practices, they will make adjustments to try to take account of them. To follow on from this, I suspect that when Moody’s does not know how much adjustment to make, it is unlikely to be timid—in other words, there may be an overadjustment, particularly where there is little “hard” information to go on. Indeed, commentators in several countries have argued that during the banking crisis the outside raters, auditors, and consultants may have been overzealous in their adjustments. But that would be an appropriate risk-averse response in a situation of uncertainty, and the way to address it would be to reduce the uncertainty.
My next observation relates to Mr. Mahoney’s throwaway phrase that the banks’ accounting games are often abetted by their regulators. In some cases, of course, this may well be true. Bankers may be powerful people who have friends who can put pressure on regulators, and in many countries regulators may be very vulnerable to such pressures. In addition, there may be a deep, common interest in some countries between owners and regulators in maintaining the position that the banks in the country are sound. But the message to the regulators from Mr. Mahoney’s listing is similar to the message to the banks: the accounting games are well known; private sector rating agencies, other countries’ regulators, and other financial sector observers know the right questions to ask to see if these games are being played; and they will make their own adjustments if they do not receive open and credible answers. With global understanding of transparency having passed a certain critical threshold, there is not likely to be much mileage any more—if indeed there ever was—in pursuing these games. Once again, outsiders are unlikely to be timid in making adjustments to reported figures if they feel that regulators are hiding anything. Indeed, the major consequence of any abetting of questionable practices by regulators would be to undermine the credibility of those regulators, with far more dire consequences for financial flows to a country than if the regulators had been open about their problems up front.
This leads directly to Mr. Mahoney’s identification of “quality of supervision” as the third factor for assessing financial vulnerability. There perhaps may be some redundancy in having it as a separate factor. A bank’s regulators should be in a much better position than Moody’s to make, and enforce, any of the suggested adjustments to a bank’s reported accounts that Mr. Mahoney identified at the outset. So—assuming that Moody’s has got its adjustments right—if the adjustments are large in a particular case, this implies that the regulators are doing it wrong. Similarly, if the adjustments are large, and right, this implies that the supervisors are abetting the bank’s practices willfully or through neglect—in either case this is an indictment of supervision.
The fourth identified factor is sources of funding, which is taken as a proxy for vulnerability to liquidity pressures. Funding from institutional investors, the interbank market, foreign bank and nonbank depositors, or from commercial paper holders is defined as “hot” or “noncore.” I am slightly surprised at this terminology and the lumping together of some of these groups. Both wholesale and retail depositors comprise many disparate elements, and it seems a little strange in the present environment to argue that a bank that has well-informed depositors is going to be more vulnerable than one where depositors are not well-informed. Indeed, one can argue the reverse—it is the existence of well informed depositors that puts the necessary discipline on a bank to act prudently. This, indeed, underlies the model of supervision of the Reserve Bank of New Zealand and the recent moves toward transparency more generally.
That being said, there clearly is something in what Moody’s is trying to pick up: liquidity problems are undoubtedly a major initiator of overt banking crises. I would suggest two alternative measures to the suggested breakdown between wholesale and retail funding. One is to go back to the key factors specified earlier and examine the regulations on liquidity in a country and how well one believes that the supervisors are enforcing them. Secondly, closer to the fourth factor listed in Mr. Mahoney’s paper, and leaving aside data problems, 1 is to look at changes over time in sources and maturity of banks’ funding. While a prudent and well-managed bank might survive, and indeed thrive, on wholesale or short-term funding over the long term, sudden recourse to such funding could well be a signal of emerging problems.
The suggestion to look at the rate of change rather than some preset benchmark may well be applicable more generally across a range of indicators, including those in Mr. Mahoney’s first key indicator of vulnerability. While a range of factors has been identified that, in principle, may predict banking vulnerability, much less work has been done to determine the benchmark levels that would lead to an “amber light” or “red light” as to emerging dangers. Indeed, it may well be that, across countries and across different types of bank within a country, there is no single set of useful benchmarks. An assessment of rates of change over time in key indices for a particular bank, or a particular banking system, could therefore be particularly useful.
Having gone through the four broad categories of factors for predicting systemic risk, Mr. Mahoney suggests that these produce necessary but not sufficient conditions for an actual crisis. For the crisis to actually emerge, it is argued that one also needs the “Injudicious Introduction of Market Discipline” and/or “External Payments Crises.” The implication is that, as long as market discipline is kept at bay and the country avoids an external payments crisis, a system of insolvent banks can keep going indefinitely without falling into overt crisis. China is given as an example. The policy prescription of this analysis is that countries are urged to be very cautious in liberalization.
There must, however, be serious questions about this analysis. In particular, if a system keeps going despite a large number of insolvent banks, this must be because the public expects that the government will support the banks, or the banks’ customers, if necessary. However, while a government may wish to support such banks or their customers, its resources are by no means infinite. Attempts by a government to limit its commitment to a banking system—for instance, by introducing a limited deposit guarantee—which Mr. Mahoney may consider to be injudicious introduction of market discipline, may reflect a necessary response to recognition of the fiscal implications of unlimited support. And, even if the government does not recognize the limits to its support, the public may well do so. Limits to public support are likely to be a function of the government’s overall indebtedness; as the stock of debt rises, there is a corresponding decline in the government’s ability to spend money, including on supporting the banking sector, and hence in the credibility of any strategy that it has for handling the banks. Moreover, one can argue that the sort of conditions that lead to an increase in banking system insolvency are in many countries the same as those that lead to an increase in fiscal indebtedness—for instance, as in Russia, the unwillingness of powerful groups to pay their financial obligations (whether as bank borrowers or taxpayers), or pressure for large expenditures (whether through bank or budget financing) for various groups. Pervasive bank insolvency, if unaddressed, is likely to be synonymous with increasing bank insolvency and, indeed, precrisis insolvency. It may well also be connected with overt, or disguised, fiscal problems and currency problems, which ultimately can all lead to a mutually reinforcing full economic crisis.
It is difficult to argue that—for instance among the transition economies—those that liberalized furthest and fastest experienced the biggest banking crises. Among the countries of central and Eastern Europe, for example, one of the biggest banking crises was that of Bulgaria. In the mid-1990s, Bulgaria was among the slowest in introducing market discipline into the economy. Nevertheless, by mid-1995 the banking system had become insolvent in aggregate, despite earlier recapitalizations of the state banks; meanwhile, the government itself had built up an unsustainable stock of debt. With the public’s awareness of these problems, there were runs on (or rather long and slow-moving lines outside) many of the banks. The government’s commitment to a blanket guarantee for depositors in early 1996 could not substantially improve the situation since it was known that the government had insufficient resources to make the required payments and, in the event, delayed payments so that their real value became eroded by the emerging hyperinflation.
If indeed it were liberalization that caused the banking crises, one would expect that the parts of the banking sector most exposed to market forces would be the most insolvent and most likely to precipitate the crisis. This frequently has not been the case. Although the public’s belief that the government would stand behind the state banks has meant that liquidity pressures have in some cases emerged first in the private banking sector, once the entire sector is reviewed, problems are often seen to be at least as deep-seated among the state banks. In Bulgaria, for example, the two well-known “ailing banks” in the mid-1990s were both state banks. Some of the private banks also became insolvent, but most of the insolvency remained in the state bank sector. In Indonesia, the authorities performed a triage on the private banks, largely on the basis of a bank’s capital asset ratio (CAR); the “A” banks were those deemed strong enough to survive without government support; those “B” banks—the banks with less good but still acceptable CARs—those that could provide viable business plans and satisfy a number of other tests were deemed eligible for joint recapitalization with the government; and the “C” banks, those which had unacceptably bad CARs, were closed. In 1999, 5 percent of the banking sector (38 banks) were “C” category or “B fail” private banks and were closed; however, 15 percent of the sector were either “A” category or “B pass” private banks. By contrast, of the state banks, comprising around 50 percent of the banking sector, on the basis of the private sector tests, all would have been clearly in the “C” category and would have been closed had the government not deemed them all too big to fail. Because the government waited so long to address banking sector weakness—problems in the Indonesian banking sector were apparent from the start of the 1990s and the World Bank pressed the authorities to take early action to restructure the state banks—the costs of bank restructuring increased manyfold, and the banking crisis was accompanied by a full-blown economic crisis, the costs of which will be borne for many years to come.
More generally, in a world where countries have to be able to demonstrate macroeconomic flexibility—for instance, in response to external shocks—having a weak and unliberalized banking system is not going to be helpful. For instance, economic conditions may call for the depreciation of the exchange rate. The existence of a weak and unliberalized banking system may well deter the authorities from making such a move out of concern for the effect of depreciation on the banking system. Indeed, prolonged adherence to an unsustainable exchange rate has been a feature of many of the recent banking and economic crises.
I, therefore, draw a different lesson about how the authorities should protect themselves against banking crises. I do not think that the answer is to liberalize slowly—for many countries that is not an option anyway, since old means of control are no longer available or effective. Liberalization and globalization are with us anyway, unless a country takes a very different route and cuts itself off from the major trends in the world economy. However, these trends make it all the more urgent that there are accelerated advances in improving accounting standards, quality of supervision, and—I would add—very importantly, legal reform. These infrastructural developments are no longer technical afterthoughts; they are now central for economic policymaking.
In this regard, the recent emphasis on introducing standards, and encouraging countries to comply with these standards as part of the new international financial architecture, provide a critical way forward for many countries. This will be discussed in more detail later in the seminar. For instance, on the data provision side, the Special Data Dissemination Standard (SDDS) has been developed by the IMF in conjunction with an extensive outreach program among the membership. While subscription to the SDDS is voluntary among members, the standard is uniform across subscribers. The Basel Core Principles and the Codes on Monetary and Financial Policies are further examples that have been recently developed. As these become better known in the wider world, and more experience is gained with them, it is likely that there will be increasing reliance on these standards in assessing a country’s vulnerability. One would expect that a paper on the private sector’s perspective in assessing financial sector vulnerabilities would in a few years’ time include the role of these standards as a central element. This would save Moody’s a great deal of work, and would provide assessed countries confidence that they understood fully the basis on which they were being assessed and that assessment was being conducted uniformly.
Returning finally to the theme of the session—Assessing the Ability of a Financial System to Withstand Financial Stress—I would suggest that there is an additional element beyond those identified by Mr. Mahoney in his paper. I would mention also the possible role of macroprudential indicators. Mr. Mahoney states that there is a risk of systemic vulnerability if there are many banks in a system that demonstrate individual vulnerability, but “many banks” is not defined. Attempts to develop a consistent framework for this issue have led recently to a number of institutions, including the IMF, starting to develop a set of macroprudential indicators. So far, in most cases, this has involved drawing up long lists of indicators of vulnerability. The next stage is to narrow this range and preferably to devise a theoretical basis for determining this narrowing. One can argue that the operationality of this work is roughly where we were on monetary indicators in the 1970s—most people agreed that money was important, but there were many ways of measuring it, and little agreement on what one should be focusing on most intensely. The IMF is very much involved in trying to take forward work in this area. Hopefully, in a few years, a paper on private sector assessments of financial sector vulnerability would also look at these macroprudential indicators.
In conclusion, I would reiterate that I found Mr. Mahoney’s paper interesting and useful. This is a rapidly evolving field, and the contribution of private sector agencies such as Moody’s will be integral as we all go forward to improving our understanding of how financial sector vulnerability can be predicted and addressed.
One issue on which it would be interesting to elaborate is the quality of the data to which Moody’s and other outside agencies have access.