Skip to Content

Reform of the International Financial System


Speech by Dr. Eisuke Sakakibara At the ManilaFramework Meeting in Melbourne on March 26th, 1999


Ladies and gentlemen. It is my great pleasure to have an opportunity to addressthis group. As we all know, this group was created with the aim of forestallingfinancial crisis in Asia and the Pacific, and, in the event of a crisis, ofminimizing its impact on the economic welfare of the peoples of the region.Since the turbulence in the financial markets is not yet over, even though ithas subsided considerably, it is all the more important for this group toemphasize the need for continued vigilance.

Today, I should like to talk a little about the lessons we needto learn from the financial crises of the latter part of this decade, especiallyabout the question of how to cope with the free capital movements thatcharacterize today’s world economy. Its immediate implication is whether thereis room for improvement in IMF programs, but the matter does not stop there: infact it warrants a thorough review of the structure and function of theinternational financial architecture as a whole. In my remarks today, therefore,I would like to suggest some measures that I think would be effective instrengthening the architecture, including measures to improve IMF programs, sothat abrupt shifts in capital flows will not lead to yet another serious crisis.

1. The Nature of the Crisis

Ladies and gentlemen. Since the crisis erupted in Asia in 1997,I have persistently insisted that it was not the “Asian” crisis but acrisis of global capitalism. And I think it is fair to say now that manyhave accepted this proposition, and the crises of 1997 and 98 should be analyzedas a continuation of the “global” crisis that started in Mexico andArgentina in 1994-95. Unlike the Mexican crisis of 1982, when external factorssuch as a steep rise in the U.S. interest rate and the sudden appreciation ofthe U.S. dollar played a major part in triggering the crisis, there were noapparent external causes of the 1994-95 crisis. International conditions,including the U.S. market, were stable and economic reforms in both Mexico andArgentina were well received by the international community. Some economists,notably Rudiger Dornbusch, argued that overvalued currencies were the directcause, as in the case of the Asian crisis of 1997. Indeed, throughout the crisisfrom 1994 to 1998, overvaluation of real effective exchange rates was one of thefactors triggering the panic. Also, the short-term debts of Mexico and Argentinain 1994 exceeded the level of foreign reserves. In particular, the short-termofficial debt of Mexico denominated in U.S. dollars (tesobonos) of around $28billion scheduled to be paid within several months in 1995 far exceeded thelevel of foreign reserves which at that time was only $6 billion. A similarsituation existed between private short-term debts and the level of foreignreserves in Thailand, Indonesia, and South Korea in mid-1997. In Asian countriesit was private, short-term debts, not official debts such as tesobonos which hadaccumulated.

Despite some growing signs of vulnerability, these crisesfrom Mexico to South Korea were not predicted by market participants andanalysts until certain events, political uncertainty, or bankruptcies of bigcorporations triggered panic. Risk premiums in loans remained low, and ratingagencies such as Standard & Poor’s and Moodies kept their relatively highrating of sovereign bonds up until the onset of the crises. Many analysts andfinanciers, particularly at the outset of the crises, argued that the lack ofproper disclosure or high level of transparency hampered the appropriateassessment of risks. However, objective evidence and data seem to indicate thatthe pertinent information such as the real effective exchange rates, short-termforeign debts in the private sector, current account balances, and balancesheets of banking sectors was largely available. The problem was that thisinformation was not appropriately incorporated into the risk assessment of themarkets. Particularly when considering factors in the behavior of nonbankfinancial institutions such as hedge funds and pension funds, one is inclined tobelieve that riding with the herd mentality has been more prevalent thanrational and detailed calculation of emerging market risks. Moreover, so-calledrational calculations a la LTCM turned out to be misleading in that their modelsassumed some stable equilibrium.

Thus, looking more objectively at the details of these crises,one is led to believe that they are testaments to the inherent instability ofliberalized international capital markets where sudden reversals of marketconfidence cause periodic panics of different magnitude and duration. Also, itis interesting to note that both the Mexican and South Korean crises occurredimmediately after these countries joined the OECD and began to conform to thecode of capital liberalization of the organization. Indeed, after thesubstantial liberalization of the capital account of five Asian countries, -South Korea, Indonesia, Malaysia, Thailand, and Philippines- around 1993 andthereafter, approximately $220 billion dollars in private capital flowed intothe region during the three-year period from 1994 to 96. The reversal of flowsin 1997 due to the sudden shift in confidence amounted to roughly $100 billion.No country or region can tolerate this kind of sudden shift in market sentimentfrom euphoria to panic that causes a huge reversal of private capital flows.

2. Were the IMF programsappropriate?

Much criticism is being raised with regard to the IMFprograms for the crisis-hit Asian countries. Although I do not necessarilyagree with all the criticism, I believe that it is crucial that not only the IMFbut also its shareholders are frank about admitting their shortcomings, if any,and that they learn lessons from them, because in the end it is the shareholdersthat make IMF decisions and are responsible for their outcome. As one of themajor shareholders, Japan has a keen sense of responsibility. And because theIMF must continue to be the central institution in the international financialsystem, I strongly believe that its role needs to be improved. From thisviewpoint, it seems to me that the preliminary assessment of the programs inAsia carried out by the IMF staff (‘IMF-Supported Programs in Indonesia,Korea, and Thailand: a Preliminary Assessment’), which was published earlythis year, was a very constructive step forward, but that there still is a needfor further review. Although I do not intend to go into detail, I would like tolist some points that I think are important.

First of all, it should be noted that the growth outlookfor program countries was too optimistic when the programs were at the designstage. Since the economic outlook was a basis for all subsequent advice onpolicy direction, the failure to grasp the severity of the economic situationwas critical.

Secondly, in the area of fiscal policy, the IMF appliedits traditional approach -- which it has applied to Latin America and otherregions, namely, tightening of fiscal policy aimed at reducing current accountdeficits -- to the Asian crisis countries, which had not had large currentaccount or budget deficits before the crisis. Moreover, this policy stance forAsia was based on an optimistic economic outlook for the economy. In my opinion,this contributed to a further contraction of domestic demand. Indeed, the IMFaccepted an easing of the fiscal stance at a later stage, which would seem toindicate that the fiscal stance adopted at the outset may have beeninappropriate.

The IMF staff appears to suggest that fiscal tightening wasassigned a modest role in the programs. If this is true regarding the programfor Indonesia, I do not understand why there was such harsh criticism by theIMF towards the Indonesian authorities in January, 1998 when they announced abalanced budget. This criticism had a severe adverse impact on marketconfidence, and in the event, the IMF accepted an increase in budgetexpenditures at a later stage, which makes me wonder whether the earliercriticism was warranted.

Thirdly, regarding monetary policy, the IMF required theprogram countries to raise interest rates in order to restore market confidenceand to ease pressure on their currencies in the exchange market. This advice wasrepeated for Russia and Brazil. However, I think that the relationship betweentight monetary policy and developments in the exchange rate following theoutbreak of a crisis is quite unstable, at best.

In the Asian programs, interest rates were raised to haltinflation. However, in many cases what occurred was one-off general priceincrease due to currency devaluation and not inflation, for example, caused byexcessive demand. Attempts to contain this one-off price increase through verytight monetary policy have, in many cases, led to an unnecessary overkill ofeconomic activity.

Another purpose of the interest rate increase was to stemcapital outflows by making the return on domestic assets more attractive.But, again, if confidence has fallen too far, monetary tightening will merelyconvince investors that there will be a further economic downturn, and will notgive them any incentive to stay. In hindsight, I wonder whether the IMF couldnot have proposed another way to stem the torrent of capital outflow.


Incidentally, again in Indonesia’s case, therewas an argument that large-scale liquidity support to the financial sector,which was out of line with the program assumptions, was the cause of thecontinuous depreciation of the rupiah. To start with, however, liquidityexpansion became necessary because the authorities needed to respond promptly tothe financial panic caused by the measures in the IMF program, namely, the hastyclosing of 16 banks and the requirement to quickly improve the capital adequacyratio. Without such liquidity expansion, the Indonesian economy would have beenmore seriously wounded. Moreover, this argument is hard to sustain, since itcannot explain why downward pressure on the Thai baht and the Korean wonpersisted, when the authorities of these countries achieved the programs’monetary targets roughly within the programme assumptions.


Lastly, as I pointed out earlier, there were indeed structuralproblems in the crisis-hit Asian countries, particularly in the financialsectors. The IMF was right in raising concerns and trying to address the issues.The question is: did the IMF programs include only those structural policymeasures that were indispensable to overcoming the immediate crisis, and didnot the programs require too rapid reform?

For instance, there were more than fifty items of structuralreform required in the program for Indonesia, including reforms of the fooddistribution system. I doubt whether all of these reforms were absolutelynecessary to resolve the crisis. The reform measures were also too ambitious,as evidenced by the closure of 16 banks in Indonesia in a matter of a few dayswithout adequate protection for depositors. Structural reform is not somethingthat can be implemented by a single decision by the decision-maker, like achange in interest rates. It has to be endorsed through the political processand introduced in a way that does not cause undue friction in society, becauseeach country has its own traditions, history, and culture, which are reflectedin the economic structure. To think that it is possible to carry out structuralreform overnight is to ignore this basic fact.

Of course, some reforms become possible only at a time ofcrisis: but this does not mean that everything can be put into one bag. If thetargets the IMF sets for structural reforms are too many and too ambitious, theyare likely to be missed amid the political confusion after the crisis, resultingin a weakening, rather than a strengthening, of market confidence. I stronglybelieve that, in light of the fact that the 1997-98 crisis was caused by rapidand massive capital outflows, the rapid drain on capital should have beenstopped first, and then structural reform should have been pursued from along-term perspective.

3. Lessons Learned

Ladies and Gentleman. In my opinion, the lessons we need tolearn from our common experience with the 1997-98 crisis are as follows:

First, we must recognize the fact that a crisis may take placeas a result of a shift in investor confidence, almost regardless of thefundamental strength of a given economy. This is because private capital tendsto move in one direction (the ‘herding’ problem), and, as a result, hugeinflows or outflows en masse can make the economy overheat or collapse.The international financial system, whose basic structure was established in thedays of fixed exchange rates and limited private capital movements, needs to beadjusted to this reality.

Second, more specifically, an appropriate exchange rateregime must be adopted by each emerging economy. There have been many caseswhere a country tries to defend a rigid exchange rate system at all costs untilfinally it has to abandon the pre-set target and opt for a flexible regime, andin the process economic activity is stifled by too tight fiscal and monetarypolicies.

Third, the IMF and the World Bank need to be strengthened. Inparticular, as successive G7 communiques have pointed out, IMF programs needto be strengthened to adapt to the new reality of free capital movements,and the IMF’s procedures should be improved so that input from shareholderswill be better reflected.

Fourth, ways to ensure closer involvement of private-sectorcreditors in crisis prevention and crisis resolution need to be foundquickly.

These are, I think, the core of the wide-ranging issuesconcerning the so-called international financial architecture although there aremany other issues, including a need for more transparency, standards for goodpractices, and enhanced supervision. I would now like to turn to specificproposals which Japan has been emphasizing.

4. Towards a New InternationalFinancial Architecture

a) Safeguards against InternationalCapital Movements

Ladies and Gentleman. First of all, we need enhanced safeguardsagainst international capital movements. In order to strengthen the monitoringof capital movements as an important part of IMF surveillance, the IMF andnational authorities should collect more detailed data on inflows and outflowsof capital. When liberalizing capital accounts, care must be taken that theliberalization is conducted in appropriate sequence. Emerging economiesshould also strengthen prudential regulations in order to preventexcessive exposure to foreign creditors and check surges in capital inflows.

Monitoring should also be strengthened on the side of creditors.In this connection, measures are needed to better deal with the activities of highly-leveragedinstitutions (‘HLIs’) such as hedge funds, including those establishedin off-shore locations. This is important in two senses: first, HLIs use a lotof mathematically calculated probability formulae in deciding their investmentstrategies, but there is always a probability that strategies based on theseformulae will go badly wrong, as in the case of Long Term Capital Managementlast autumn. Because HLIs are interwoven deeply in the trading successions, ifone HLI has trouble, it quickly spreads to the counterparty and then throughoutthe whole system. In such a case, the unwinding of positions will put furtherdownward pressure on the market, or when the liquidity is thin the market itselfmay collapse, and a vicious spiral may ensue. This is why the Basle Committee onBanking Supervision stated in its report early this year that better riskmanagement on the part of financial institutions is crucial. For this purpose,all countries concerned should take measures, including strengthening prudentialregulations, if necessary. Second, HLIs often have such great resources andleverages at hand that they can dominate the market in a small economy, whichmay enable them to manipulate market developments. Making the activities of HLIsmore transparent is one way to address this problem. In addition, ways must beexamined to minimize the possibility of market manipulation by large-scaleinvestors, for the sake of the healthy functioning of the markets in emergingeconomies.

Safeguards against capital movements should also be an importantelement in IMF programs. Measures such as controls on capital outflows ofresidents and a temporary standstill on foreign liabilities owed by public- aswell as private-sector borrowers should be legitimized as possible options,depending on the situation. In this connection, the IMF should move ahead withthe policy of lending into arrears. Needless to say, care must be taken thatsuch controls do not restrict current transactions, as opposed to capitaltransactions, or undermine investor confidence.

b) Exchange Rate Policy

Second, an appropriate exchange rate policy needs to be soughtby each country. In the case of emerging economies, it might be generallyappropriate for them to peg their currencies to a basket of currencies ofthe developed countries with which they have the closest trade and investmentinterdependence, adjusting the peg periodically to reflect developments in realeffective exchange rates and current and capital account balances among otherthings. There is no simple set formula, however, and a case-by-case perusal ofcountries’ specific situations is essential. Needless to say, the IMF needs toplay an important role in this exercise.

c) Strengthening IMF Surveillance andPrograms

I just pointed out that elements such as monitoring of capitalflows and selecting the appropriate exchange rate regime should be emphasized inIMF surveillance. In addition, a better understanding of the realeconomy of the country in question, rather than an overemphasis on monetarydevelopments, should also be pursued.

With regard to programs, together with improvement in thefields of safeguards against capital movements and exchange rate policy, policyadvice in other areas should also be strengthened. As for fiscal policy,countries which have good track records and which have not had significantbudget deficits or current account deficits in the past should not be requiredto implement too strict a fiscal policy immediately after a crisis. Theunavoidable expansion of budget deficits after a crisis should be addressed as amedium- to long-term challenge to be tackled once the situation stabilizes.

As for monetary policy, while monetary tightening hasbeen used in many cases for the purpose of defending an exchange rate, such apolicy needs to be used with caution. As I said earlier, once investorconfidence in a currency is lost, monetary tightening may have limited effectson defending the exchange rate. On the other hand, tight monetary policy and theresulting high real interest rates will have an adverse effect on a country’seconomic activity and fiscal position. Furthermore, it is important to recognizethat there are cases where the IMF should shift priority to providing liquidityto the domestic economy to mitigate the deflationary impact of the crisis. Inorder to avoid a situation where liquidity injected in connection with an IMFprogram leads to additional capital outflows and a decline in reserves, measuressuch as the introduction of controls on capital outflows of residents should beconsidered.

On structural policy, in times of crisis, the IMF shouldrefrain from requiring too broad or ambitious structural reform. Once thesituation stabilizes, necessary structural reforms should be pursued as medium-to long-term challenges, with the World Bank and other relevant institutionstaking the lead, while the IMF would continue to participate in discussions withthem.

d) Private Sector Involvement

There are already sufficient studies and proposals on the issueof private sector involvement. This is a critically important question,since bailing out private investors by means of financial assistance fromofficial sources, in particular the IMF, is not justified from the viewpoint offairness and moral hazard.

I think that in times of crisis it might be appropriate to makeIMF programs conditional on the maintenance of exposure and/or on an agreementto roll over loans from major private sector creditors. Some IMF facilities,such as the CCL, should be linked to a member’s efforts to secure privatesector involvement through, for instance, collective action clauses andcontingent financing arrangements.

5. Conclusion

Ladies and Gentleman. How to reform the international financialarchitecture is a complex and difficult challenge. There have been manysuggestions from government officials, academics, journalists, and so on, but onsome issues there is still a wide difference of opinion. But, we need to find away to forge agreement among us in order to build a foundation for the sounddevelopment of the world economy in the next century, just as at Bretton Woods,which resulted in amazing economic development and the reduction of poverty inmost of the post- World War II world.

The task before us is certainly huge, but I am convinced thatwith our collective wisdom we will be able to agree on the comprehensive reformof the international architecture. Otherwise, the 21st century couldbecome the century of major turbulences and collapse of the global capitalism.

Thank you very much.