- Sep. 17, 2021
- Sep. 13, 2021
- Sep. 10, 2021
Based on a speech given by Kazuo Ueda, Member of the Policy Board, at the Fall Meeting of the Japan Society of Monetary Economics, on October 25, 2003
February 27, 2004
Bank of Japan
Click on ko0402b.pdf (29KB) to download figures.
I have been asked on this occasion to discuss issues related to the risk management of financial institutions. Given my comparative advantage, I would like to take this opportunity to talk about the risk management of central banks. The subject is still very broad. I could discuss issues related to the internal management of the Bank of Japan or issues involved in risk management measures to ensure smooth functioning of the BOJ-NET, the central bank's payment and settlement system. The most important risk management issue for central banks relates to their efforts to maintain price stability, which is their primary policy objective. Of course, uncertainties about the future prevent central banks from fully achieving this policy objective. Accordingly, central banks should monitor the probability of failing to attain the policy objective and minimize it through their policy implementation. Further, even if some degree of price instability is unavoidable, central banks should, to the greatest extent possible, take every measure to minimize the negative impact on the welfare of the public. This is the most important aspect of central banks' risk management.1
Today, I would like to discuss a somewhat narrower topic related to central banks' risk management. Central banks usually take policy measures as deemed appropriate in order to attain their policy objectives. In the area of monetary policy, most such policy measures take the form of financial transactions in the money market and lending to and borrowing from counterparty financial institutions. Through these measures, central banks are exposed, to a greater or lesser extent, to market and credit risks. In the case of private financial institutions, such risks are taken on in the pursuit of profit maximization and their capital functions as a cushion against unexpected losses arising from the materialization of these risks. Central banks also take risks, but instead of maximization of profits, achieving price stability is their main objective. Central banks incur unexpected losses, even with cautious policy implementation, and the question emerges whether central banks need to hold capital against such losses. Furthermore, to what extent does capital inadequacy or, in some cases, insolvency of central banks, hinder the attainment of their primary policy objective?
The role of central bank capital has become an increasingly popular topic recently. Unlike in the past, when the topic was discussed in relation to central banks in developing countries, growing attention has recently been paid to the role of central banks' capital in developed countries. This reflects the use of, or increasing likelihood of the adoption of, risky operations by central banks in order to overcome or prevent deflation. Some of these operations run the risk of severely impairing the balance sheet of a central bank.
Among major central banks, the Bank of Japan has started to purchase stocks from private banks in late 2002, in addition to its holdings of foreign currency and large amounts of medium- to long-term Japanese government bonds (JGBs). In 2003, the Bank further broadened its purchase operations to include asset-backed securities. Such operations have met with two sharply different responses. Some believe that the Bank has taken on more risks than is appropriate for a central bank and that it is subject to the possibility of serious deterioration in its balance sheet and of loss of public confidence in the Bank. Others take the view that overcoming deflation through such unconventional market operations would lead to improvements in the economy and consequent increases in tax revenue. Increased tax revenue would make it easy for the government to recapitalize the Bank in the event that it became insolvent as a result of risky unconventional operations. They also argue that seigniorage revenue will rise as the inflation rate returns to a normal positive value due to the Bank's action, making temporary insolvency an insignificant problem.2
I must say that there is a grain of truth in both arguments. However, neither captures the full complexity of the role of capital in central banking. In what follows I would like to present my view on the topic, touching also on the criticism that the Bank has been too conservative in its policy decisions with too much concern about the health of its balance sheet.
Let us first take a look at how much capital central banks actually hold. Table 1 shows the ratio of capital to total assets for a number of central banks. There are four points to notice in the table. First, the ratio varies from country to country. Second, it is generally lower for central banks in developed countries. Third, insolvent central banks are concentrated in developing countries. And fourth, among the G3 countries, the capital ratio for the United States is low, while the ratios for Japan and the EU are somewhat higher. The significant disparity in the ratios suggests that there is no consensus among central banks about the desirable level of capital.
Let us take a closer look at individual cases in order to discuss some possible causes of, and the impact on the conduct of monetary policy of large operating losses or insolvency of a central bank.3 Stella (1997, 2000) and others have examined these questions and found that many of the cases of insolvency have occurred in Latin America. Some central banks in Asia, Africa, and Eastern Europe have also become insolvent. In developed countries, the Bundesbank experienced insolvency in 1977-1979.
The causes of insolvency varied, but most cases followed one of three patterns. In the first pattern, the central bank had assumed some of the costs arising from efforts to contain a financial crisis. In the second, an appreciation of the currency resulted in revaluation losses on foreign reserves, or a depreciation of the currency caused an increased burden on the central bank's foreign currency-denominated debt, which was accumulated in its cooperation with the government's trade policy that involved foreign currency transactions. And in the third, the government failed to make interest payments on government issued securities held by the central bank.
An important question here is whether insolvency has interfered with any of these central banks' conduct of monetary policy. In many cases, they faced difficulties in attaining their policy objective of maintaining price stability, and as a consequence, an extremely high rate of inflation had emerged. Two major causes of the difficulties may be noted. First, attempts to overcome insolvency by higher seigniorage revenue naturally led to high rates of inflation. And second, explicit reliance on the fiscal authority for injection of capital to overcome insolvency generated interventions by the fiscal authority in the conduct of monetary policy, creating distortions in policy decisions.
One example of insolvency was the case of the central bank of Venezuela. It was insolvent from the 1980s through the 1990s due to the following factors: payment of import subsidies in the form of an overvalued exchange rate; the central bank's involvement in an export tax scheme; and the sharing of the cost of policy measures in the containment of the financial crisis in 1994-95. After becoming insolvent, the central bank tightened its monetary policy in order to halt the surge in inflation accelerated by expansionary fiscal policy. However, due to profit-squeezing resulting from the central bank's issuance of its debt with high interest rates at the time, it was forced to abandon the tightening policy, thus leaving inflation unattended.4
Another example is the experience of the central bank of Jamaica from the 1980s through the mid-1990s. As a result of its support for the government's industrial policies in the early 1980s, the central bank held on its balance sheet a significant amount of foreign currency-denominated debt. Following the depreciation of the currency in the mid-1980s, it started to post large losses due to the burden of its interest payments on this debt. The government compensated the bank for the losses by contributing government bonds, but as the bonds were non-interest bearing, they did not help to improve profits, and the central bank became insolvent. Meanwhile, the bank fought the accelerating inflation by withdrawing liquidity through a large-scale issuance of certificated deposits, but this further expanded its losses as interest rates were rising, and as a consequence, sufficiently tight monetary policy to halt inflation could not be implemented.
The Philippine Central Bank, the former central bank of the Philippines, also faced insolvency and high inflation due to a similar mechanism. Measures to counter inflation were not taken because the central bank was obliged to support government policies.5
There are cases, however, where insolvency of a central bank did not lead to high inflation. The central bank of Chile was insolvent from 1997 through 2000, but the inflation rate in 2000 was below 4 percent, and thus price stability was maintained. The difference between the case of Chile and the last three cases is that the extremely tight Chilean fiscal policy up to around 1997 combined with relatively accommodative monetary policy resulted in stable macroeconomic policy on the whole.6, 7
Deterioration in the balance sheet of a central bank poses risks not only to price stability, but also to its function as the fiscal agent for the government and its responsibility to maintain an orderly settlement system. Stella (2002) notes that when significant deterioration in a central bank's balance sheet is observed, financial disintermediation away from the formal payment system occurs and hence the efficiency of settlement often declines.
Summing up the experiences of insolvent central banks, my conclusion is that the maintenance of a sound balance sheet is, in general, neither a necessary nor a sufficient condition for fulfilling a central bank's responsibility, but there have been cases where an unhealthy balance sheet became a major obstacle to price stability.
Let us now move on to a more theoretical analysis of the role of capital for central banks.
I pointed out that fiscal authorities have an incentive to interfere in the conduct of monetary policy when supporting a financially distressed central bank. However, it may be easily seen that, as a first approximation, the budgetary cost for the fiscal authority of recapitalizing an insolvent central bank is zero. Recapitalization to resolve the insolvency needs to be covered by, for example, increased issuance of government bonds. However, the central bank can purchase the additionally issued government bonds using recapitalization funds. Furthermore, flows of interest payments by the government on the bonds held by the central bank would be transferred to the government. Therefore, the government does not have to use any "real" money in this operation.
Perhaps reflecting such thinking, Meltzer (1999) has argued that "I see no reason to believe there would be any doubt about the government's obligation to stand behind the BOJ. No central bank has ever faced default, and no responsible government would permit that to happen. It is unclear to me what could be meant by failure of a central bank."
Such arguments seem to be, however, based on a rather naive view of the relationship between the central bank and the government, and the procedures by which the government formulates budgets. First of all, even if the government's recapitalization of a central bank would contribute to the central bank's conduct of monetary policy to maintain price stability, an important question is whether the government would have the incentive to do so. If that were not a significant problem, central bank independence from the government would not be an issue in the first place. The issue of central bank independence has become a subject of discussion in light of the risk that the government, if in charge of monetary policy, is prone to choose a socially undesirable rate of inflation. 8
In reality, fiscal authorities often face constraints that are not captured merely by the consolidated balance sheet of the central bank and the government. Specifically, the single year budgeting principle and the diversity and complexity of interests within the government give rise to huge inter- and intra-ministry negotiation costs when reshuffling is required between different categories of expenditure and revenue. For example, the fiscal authority usually prefers a stable stream of profit transfers from the central bank every fiscal year, since compensating for any shortfall with other revenue items would inevitably entail adjustment costs.
Thus, even if the government does not incur any financial loss in recapitalizing an insolvent central bank, it may not find it easy to take such action. More seriously, the government may take advantage of the opportunity of capital injection to the central bank to influence monetary policy if it wants a different target inflation rate from the central bank.
The theoretical view that regards temporary insolvency of a central bank as permissible because of its earning potential based on seigniorage cannot be accepted at face value, either. During the insolvency, it is quite likely that the fiscal authorities, which are often responsible for regulating a central bank, would interfere in various aspects of the central bank's policy actions. A more crucial point for an insolvent central bank is that the amount of seigniorage it can generate with a reasonable inflation rate is limited.9 If, therefore, the central bank intended to overcome insolvency in a short period of time solely by earning seigniorage, it would have to aim for a high inflation rate, sacrificing its goal of price stability.
Furthermore, we also need to take into account the possibility that a central bank, concerned about the various possible outcomes described above, might hesitate to adopt a monetary tightening policy even before becoming insolvent since some of such policy actions themselves increase the probability of insolvency. This more realistic view about the relationship between central banks and governments is helpful in understanding the often heard view that deterioration in the balance sheet of a central bank would undermine investors' confidence in the central bank and this will, in turn, result in a fall in the value of the currency.
I must add, however, that the above discussion and examples in the last section illustrate the fact that the extent to which a deterioration in the financial strength of a central bank would impede the central bank's ability to maintain price stability is a function of, among other things, the relationship between the central bank and the government. This implies that there is no single answer to the question of the desirable level of a central bank's holdings of capital.10 From a practical perspective, if the central bank and the government agree on the adequate level of capital or a formula to calculate it, this greatly reduces the adjustment costs between the two entities.
To summarize, the government and the central bank might disagree on the target inflation rate. In addition, the government has a great interest in the market operations of the central bank, as they affect the expenditures and revenues of the fiscal authorities. This applies regardless of the operations' effect on the net revenue, which is revenue minus expenditures. As a result, the tension between the government or the fiscal authorities and the central bank is likely to heighten when the balance sheet of the central bank deteriorates substantially.11
This section discusses the current financial condition of the Bank of Japan. Chart 1 shows the Bank's capital ratio calculated by the conventional formula used by the Bank, which is the ratio of capital to banknotes in circulation. The capital ratio has declined markedly in the past three years, and is currently close to the lower bound of the 8-12 percent range, which is the Bank's own standard for its financial soundness. The primary reason for the decline in the capital ratio is the expansion of the Bank's balance sheet due to its quantitative easing policy, as well as to a slower pace of increase in profits and, consequently, reserves due to the extremely low level of interest rates.12
Assuming that the level of capital will only change slowly, future developments in the Bank's capital ratio will depend on the amount of banknotes issued. The Bank's capital ratio will decline further from the current low level if the pace of increase in banknotes remains at around 5 percent.13 However, since the ratio of banknotes in circulation to GDP has already reached twice the historical average (see Chart 2), the pace of increase in banknotes might decline significantly in the future, resulting in a recovery in the capital ratio.
A more essential issue here concerns risks pertaining to possible changes in the Bank's capital. As the Bank has been conducting unconventional market operations, it is subject to the risk of incurring several types of capital loss. On its balance sheet, JGBs account for the largest share in the Bank's assets, and as of February 10, 2004, had reached 66 trillion yen. In contrast, the Bank's capital was just over 5 trillion yen and thus a 10-percent fall in the value of JGB holdings due to a rise in interest rates would be sufficient to make it insolvent.
The Bank's insolvency in this sense, however, needs to be interpreted cautiously. It should be noted that the Bank has not sold any JGBs outright to the market in the last 30 years. Furthermore, the Bank's revaluation method for its holdings of JGBs will be changed, in line with those of private firms and public-sector institutions, from the lower of cost or market method to the amortized cost method from fiscal year 2004. As long as the Bank does not recognize the unrealized losses on held-to-maturity JGBs, the key question with regard to maintaining its financial soundness in the accounting sense is whether there is a possibility that the Bank will be required to sell a large volume of its JGB holdings in the near future.14
A starting point for answering the above question is to consider what kind of market operation the Bank will need to conduct when it brings the economy to an appropriate level of inflation after overcoming the current deflation. Depending on developments in the demand for the monetary base, the Bank could be forced to carry out large-scale liquidity absorbing operations in order to keep short rates at target. Using outright JGB selling operations would lead to realization of large capital losses. Alternatively, the Bank could absorb funds by issuing short-term BOJ bills. In this case, the Bank would incur losses only when interest rates on such bills exceeded yields on JGBs held by the Bank. Compared to losses incurred as a result of outright sales of JGBs, losses due to the yield differentials would be likely to be smoothed out over a longer period. The total amount of the losses could also be smaller depending on the shape of the yield curve during such operations. Thus it seems that during the exit period from the current quantitative easing policy, there would be only a small possibility of the Bank facing insolvency in accounting terms. Even if the Bank becomes insolvent during the exit period, it is likely that the extent of shortfall in capital would be modest and insolvency would be resolved relatively quickly after the exit period using operating profits.15
To put it another way, the Bank's current financial condition does not preclude further purchases of JGBs, but at the same time, extremely large purchases of JGBs would inevitably heighten the probability of the risks discussed above materializing.
Besides large holdings of JGBs, the Bank's risk assets include foreign currencies and stocks. Its holdings of foreign currencies were 4.3 trillion yen as of October 2003 and those of stocks were almost 2 trillion yen. Given the size of these assets, it is clear that the Bank would not become insolvent solely as a result of revaluation losses on one of these two types of asset, although price fluctuations of such assets within usual ranges can have non-negligible effects on the Bank's profits.16
I have pointed out that political-economy constraints in the real world may prevent central banks from fulfilling their responsibility to maintain price stability, if their financial strength deteriorates substantially. At times, the government, from a short-term perspective, can be attracted to excessively easy monetary policy in order to stimulate the economy. Likewise, the fiscal authorities at times want to boost seigniorage through high inflation. The likelihood of the central bank giving in to such political pressures rises if its financial strength deteriorates to the point where it is obliged to seek financial support from the government.
In the case of Japan at present, the financial condition of the Bank has not significantly deteriorated, nor does the Bank face an imminent risk of high inflation or the necessity of conducting market operations to curb it. The current policy objective of the Bank is to overcome deflation, an aim it shares with the government.
The absence of a risk of high inflation, however, does not imply that the Bank need not be concerned about its financial soundness. As I mentioned above, the Bank has implemented various types of market operations in the past several years in order to overcome deflation and to reinvigorate the distressed intermediary function of financial institutions. Some of the assets it has acquired through such operations, stocks for example, will suffer a sharp price decline if deflation is not overcome. The value of other types of assets, JGBs for example, will fall when deflation is conquered. Although risk diversification effects can be expected from holding both types of assets, accumulating either type above a reasonable level could do significant damage to the financial strength of the Bank under certain economic conditions. As discussed in the previous section, the likelihood of the Bank finding itself in such a difficult situation does not seem so great at the moment, but the probability would inevitably increase if the current deflation were to persist and the Bank conducted even more aggressive market operations. Given the discussions in earlier sections, in such a situation the Bank might find it necessary to have an advance agreement with the government on a risk-sharing scheme with regard to such aggressive market operations.
Bernanke (2003) has proposed that in the event of further aggressive JGB purchasing operations by the Bank, the government could grant the Bank the right to convert its JGBs to floating rate bonds. For the Bank, this measure would certainly be equivalent to obtaining an advance commitment from the government of financial support or recapitalization if its financial condition deteriorated in the future. But the question is, would the government have an incentive to give the Bank such an option? The answer is a partial yes, if the government shared with the Bank the policy objective of overcoming deflation and if aggressive market operations were considered effective as a means to that end. There would, however, be serious obstacles to be cleared before such an agreement could be reached between the Bank and the government.
Suppose such an agreement had been made and interest rates rose, then conversion of existing JGBs to floating rate JGBs would initially result in an increase in expenditures. If they were offset by the corresponding increased transfer of profits from the Bank, the effect of the agreement on the overall budget would be neutral as discussed above. However, an issue would remain for the government sector that the advance commitment to recapitalize the central bank would reduce the opportunities for the fiscal authorities or the government to interfere with monetary policy. In addition, there would be a less important question regarding whether the budgetary implications of the combination of an increase in interest payment expenses and an increase in transfer of profits would really be neutral for the fiscal authorities. If the government were not concerned about these points and were willing to commit itself to recapitalization, the issue of central bank independence would not have become a major issue in the first place.17
Another obvious problem with the scheme is that, although no additional interest payment burden would be imposed on the government if the Bank held floating interest rate JGBs until maturity, this would not be the case if the Bank sold them to the market. As we saw above, such a possibility cannot be ruled out. The possibility of such JGB selling operations makes it less likely that the fiscal authorities would approve the conversion option.
Since the enforcement of the new Bank of Japan Law in 1998, the degree of the Bank's financial integration with the government has been reduced (see footnote 17), and it has gained greater independence in its conduct of policy. In an attempt to overcome the intensifying deflationary trend of the economy since 1998, the Bank has conducted various types of market operations with higher risks and as a result, it has come to hold assets it previously did not. The Bank, however, is also responsible for avoiding runaway inflation when deflation ends. The Bank needs to implement its monetary policy carefully so that its financial condition does not impede the fulfillment of price stability on either front.18
My discussion has centered on the effects of various operational tools of the Bank on its financial health. Needless to say, however, a more important principle for central banks in selecting operational tools is their effectiveness in achieving price stability, and the extent to which distortionary effects of the tools on the resource-allocating function of the market can be minimized.19 I would like to emphasize that my discussion does not imply that the Bank has conducted monetary policy giving highest priority to maintaining its financial soundness.
One final remark is in order. As I mentioned earlier, many of the issues I have discussed would be of small relevance if each economic agent and policymaker were making rational decisions free of any cost. However, issues related to central bank independence and the significance of its capital are important because such decision-making costs are non-negligible in the real world. I hope economists will begin to take more interest in these issues.