Standards

Standards and codes for sound financial management are the aspect of the architectural agenda to which the official community has devoted the most effort and to which it most regularly points as evidence of its commitment to reform. Standards have been developed for everything from macroeconomic policy and data dissemination to institutional and market infrastructure. The IMF compiles a list of initiatives taking place under this heading, has developed a Special Data Dissemination Standard that covers key economic data, and has promulgated codes for transparency in monetary and fiscal policy. It has prepared and published several rounds of Reports on the Observance of Standards and Codes (ROSCs) and cooperated with the World Bank in developing a joint Financial Sector Assessment Program (FSAP) to identify the vulnerabilities of national financial systems. The Financial Stability Forum, created in 1999 as a venue for discussing common financial problems and sharing information on regulatory best practice, has convened a Follow-Up Group on incentives for fostering the implementation of standards. In all, official compendia list upwards of 64 standard- and code-related initiatives.⁴

International standards for sound practice at the national level are an obvious vehicle for addressing challenges for collective stability in a world of sovereign states. It can be argued that they are the only viable approach to providing the collective good of stable financial markets in the absence of a super-national body with the power to override national monetary and financial decisions.⁵ To an extent they have always been the focus of efforts to enhance financial stability. In the late 19^th century, such standards were limited to the monetary arena.⁶ That we refer to the dominant monetary arrangement of that period as the gold standard is no coincidence, in other words. The gold standard was the internationally accepted standard for sound financial management: adhering to it was necessary and sufficient for a country to participate actively and constructively on international capital markets (Bordo and Rockoff, 1996).

Standard setting limited to the regulation of monetary affairs sufficed so long as there was little pressure to use monetary policy to achieve goals other than currency stability and no fully articulated theory of stabilization policy. It was viable as long as financial integration was limited mainly to purchases of government and railway bonds. The early decades of the 20_th century saw a challenge to this limited agenda. Fiscal and monetary instruments had been directed toward other objectives during World War I, leading to recognition of the need to extend standards into additional areas. With the failure of efforts at the Brussels and Genoa Conferences of 1920 and 1922 to develop standards for fiscal policy, central bank independence, and international cooperation in the conduct of monetary policy, countries retreated into financial autarky, soon followed by the outbreak of the Great Depression. International standards for financial practice were rendered redundant by tight regulation of financial transactions, international financial transactions in particular. Thus, the standards to which countries were held by the Bretton Woods Agreement were essentially limited to exchange rates (IMF members were obliged to declare par values and to consult regarding their adjustment) and current account convertibility. These obligations were fundamentally oriented toward stabilizing and sustaining international trade rather than international finance.

The recovery of market-based financial transactions was responsible for the subsequent expansion of these functions. The growing cross-border repercussions of national economic policies, transmitted by unprecedentedly large capital flows, led to the Second Amendment of the IMF Articles of Agreement empowering the IMF to conduct “firm surveillance” of its members’ macroeconomic policies. The Herstatt crisis led to the Principles for the Supervision of Banks’ Foreign Establishments (the Basle Concordat agreed to in 1983), while the debt crisis, which created worries that international banks were dangerously undercapitalized, led to the Basle capital adequacy standards for internationally-active banks (the Capital Accord published in 1988). The subsequent growth of international capital markets then elicited widespread recognition of the need for more far-reaching changes in national financial practice. Ultimately, this effort foundered on the attempt to amend the Articles of Agreement to oblige IMF members to render their currencies convertible on capital account, an incompletely thought through attempt to apply the “gold standard solution” of unilateral liberalization to a world where the precondition for successful financial liberalization were more complex.

But there was no denying that international financial markets had become more liquid and deeply integrated.⁷ The range of claims traded internationally was greater than ever before; no longer were these dominated by government and railway bonds. The number of banks, corporations and governments able to fund themselves offshore was larger than before World War I (the prior period of financial globalization) or in the immediately succeeding years. Short-term capital movements were larger. More credit was available to highly-leveraged institutions. As a result of these changes, there were more points along the chain of financial transactions where things could go wrong, with negative effects not just for the country in which the problem originated but for its neighbors and potentially the world as a whole.

This recognition led in the first half of the 1990s to a broadening of IMF conditionality to encompass institutional reforms related primarily to the operation of financial markets.⁸ It inspired Morris Goldstein’s proposal for an international banking standard and to the creation of the Fund’s Special Data Dissemination Standard in 1996.⁹ But it was the Asian crisis that sealed the case for international standards. As interpreted by G-7 governments and the multilaterals, that crisis reflected weaknesses in prudential supervision and regulation, auditing and accounting, bankruptcy and insolvency procedures, corporate governance, and financial transparency. Preventing the recurrence of such crises required steps to upgrade national practice in these areas. The absence of other mechanisms and experience with the Basle Accord pointed to standards as the logical means of achieving this end. And applying peer pressure, conditioning multilateral assistance on progress in these areas, and encouraging market participants to focus on countries’ adherence to these standards were obvious ways of encouraging conformance in the absence of a global regulator with enforcement powers.

This historical digression is meant to suggest that there is a logic for the current approach to strengthening the financial architecture. The globalization of finance renders countries vulnerable to the destabilizing repercussions of financial problems abroad. Because integration is even more extensive than a century ago, so too is the scope for destabilizing spillovers. Increasingly, financial regulation has the character of a global public good. But in the absence of a global financial regulator with enforcement powers, the only mechanism for supplying this public good is national initiatives to upgrade supervision, regulation, and practice. Standards are a focal point for the peer pressure, conditionality, and market discipline that provide the incentives to carry out this task.

To say that this process has a logic is different from saying that it will necessarily have the desired effect. For one thing, there are worries that the standard-setting process has already lost its focus. The international policy community has promulgated or recognized upwards of 60 standards, raising the question of whether governments will take any of them seriously. Morris Goldstein (2001) has reasonably asked whether officials have created a bureaucratic Frankenstein.

Recognizing this problem, the Financial Stability Forum has designated 12 standards as deserving priority in implementation.¹⁰ These are concerned with macroeconomic policy and transparency (with corresponding standards for monetary and fiscal policy transparency, data dissemination, and data compilation), institutional and market infrastructure (with standards for insolvency, corporate governance, accounting, auditing, payments and settlement, market integrity, and market functioning), and financial regulation and supervision (with standards for banking supervision, securities regulation, insurance regulation, and financial conglomerate supervision).¹¹ Although this is a step in the right direction, the process would be more effective with even more attention to priorities and less effort to be comprehensive.

Then there is the question of whether these standards are suited to the circumstances of the countries to which they are applicable. Although Hong Kong and Singapore are represented in the Financial Stability Forum, lowand middle-income emerging markets are not. Private-sector standard-setting bodies like the International Accounting Standards Committee have subcommittees responsible for addressing the problems of emerging markets, but standard setting is still dominated by the G-7 countries, their regulators, and their market participants. To the extent that these individuals lack familiarity with the Asian model of bank-centered finance, industrial policy insulated from rent seeking, and reciprocal control, they may underestimate its viability. And standards requiring arm’s-length dealing between banks and corporations, Western-style prudential regulation and supervision, and resource allocation guided by asset prices may be corrosive of a bank- and government-led development model with a track record of success.

My own view is that the Asian crisis and historical experience generally have shown that financial integration implies the need for a degree of institutional harmonization. Asian countries cannot sell corporate securities to international investors without requiring their issuers to employ international standards for auditing, accounting, and financial reporting; otherwise, even a rumor about the condition of the issuer, however ill founded, may provoke panic and destabilize the market. They cannot allow banks to fund themselves offshore without removing implicit guarantees and applying internationally-recognized regulatory standards; otherwise, foreign lenders will have no reason to discipline borrowers, who will respond by levering up their bets. Governments cannot borrow abroad unless they are prepared to provide information; otherwise revelations about the public finances may lead panicked investors to herd out of the markets all at once. The choice, then, is whether to participate actively on international financial markets, in which case international standards along the lines of those developed by market-led economies like the United Kingdom and the United States are the only game in town, or to limit such participation, in which case a variety of other institutional arrangements are possible. My own reading is that Asian policy makers see their economies as exiting the phase of government-led extensive development for a subsequent phase of more heavily market-led development in which financial market integration will play a prominent role, creating irresistible pressure for institutional convergence around international standards of the sort currently being promulgated.

Even if Asian countries opt for financial integration and embrace internationally-recognized rules and procedures, standards are unlikely to contribute much in the absence of appropriate institutions. Pastor (2000) makes this point for legal rules. The law is a system of interdependent concepts; few legal rules and concepts can be understood and applied without reference to other legal rules and concepts. A rule will not deliver efficient and equitable outcomes in the absence of a body of precedent, reflecting accumulated knowledge of its functioning in the context of a living legal system. Hence, standardized rules are unlikely to be effective where a complementary body of law does not exist. IOSCO’s standards for securities regulation will not have their desired effect in the absence of an adequate body of commercial law (governing, inter alia, the duties of company directors and officers, takeover bids and other transactions intended to change control, and private rights of contract and property). Similarly, an international standard for the equitable and efficient determination and resolution of insolvency will not be workable in a country that lacks an independent judiciary.¹²

The objections in the preceding paragraph are really of two types. One is that standards will not work when the institutions required for their application are absent. Legal standards will not work when an independent judiciary and rule of law are absent. Financial standards will not work in the absence of sanctions against corruption and malfeasance—including the dissemination of blatantly false and misleading balance-sheet information. But this is not just a critique of standards: it is a challenge to viability of a market economy, which cannot function in such an environment. International standards may be just a small part of the solution to these problems, but they are part of the solution just the same. This objection is not a fundamental critique of the standard-centered agenda, only a reminder that it is just a small part of the larger task.

The other objection is that rigid and detailed international standards discourage experimentation and the adaptation of best practice to local conditions. Superior arrangements compatible with the local legal culture and economic context tend to be discovered through a trial-and-error process informed by local knowledge. Attempting to micro-manage institutional arrangements at the international level stifles local innovation.¹³ The counter is that the standards promulgated by the international community are general: they set broad benchmarks for minimally acceptable practice but allow governments to meet those desiderata in different ways. This is a caution that standards that are too specific can be counterproductive. It suggests that standards that take the form of broad principles (such as the Basle Committee’s Core Principles for Effective Banking Supervision) offer a better model than those setting out detailed methodologies (such as the IMF’s Special Data Dissemination Standard, or SDSS).

Then there is the issue of resource cost. Studies of the cost of implementing WTO commitments—only a small subset of the relevant standards—have found that these can be very substantial for developing countries (see, for example, Finger and Schuler, 1999). Collecting and processing data and strengthening regulatory and supervisory standards require technicians and data processors, computer programmers and bank inspectors. However commendable the effort, these investments may come at the expense of programs supporting socially vulnerable groups and risk eliciting a political backlash in cash-strapped economies (Soludo and Rao, 1999; Park and Wang, 2001).¹⁴ If the G-7 and the multilaterals are serious about investments in standards, then they should consider earmarking grants for this purpose.

Finally, what incentives are there for countries to take ownership of the relevant standards—to implement them, as distinct from simply voicing the intent? Those incentives are of four types: surveillance, conditionality, regulation, and market discipline. I leave the discussion of IMF conditionality to a subsequent section. Suffice it to say that the notion that the IMF should be actively involved in this process is controversial. This would encourage the further expansion of IMF surveillance and conditionality and entangle the Fund even more deeply in the economic and social affairs of its members, something to which its shareholders and Managing Director have both declared their opposition. There is more sympathy for the notion that ROSCs, FSAP reviews, and the Fund’s quarterly reports on progress of countries subscribing to the SDDS provide an objective basis for the financial surveillance conducted by Article IV missions, in the context of which the relevant warnings can be issued and congratulations extended.¹⁵ The Financial Stability Forum’s willingness to name names in its review of offshore financial centers and the readiness of the Financial Action Task Force to do the same for money laundering suggest a new resolve to provide candid assessments.¹⁶ Still, it is hard to imagine that standards-centered surveillance not backed by standards-related conditionality would do much to alter the behavior of governments.

Conditionality applies only to countries with programs. This raises the question of how to apply official incentives to the others. The policy community has backed away from direct regulatory incentives.¹⁷ Whereas the early proposal for revising the Basle capital adequacy standards proposed heavier charges on loans to countries failing to comply with the relevant standards, the revised proposal omits this recommendation.¹⁸ The recommendation of the Financial Stability Forum that implementation should be fostered by “encouraging” regulated institutions to take information on standards into account, that regulators should urge caution in dealing with counterparties located in jurisdictions which have gaps in their observance of standards, and that they should require issuers to disclose information on the observance of standards in prospectuses for international sovereign bond issues, is weak soup in comparison.

Given all this, if there is to be pressure to comply, it will have to come from the markets. A review by the Financial Stability Forum (2000) suggests that institutional investors and fund managers have rather limited familiarity with these standards and codes (Clark, 2000). That Standard and Poor’s is now producing corporate governance ratings for individual firms based on the OECD principles is a promising development, as is the fact that two commercial agencies have offered to rate countries’ compliance with a variety of other standards. But whether the market has the requisite attention span remains to be seen.

Exchange Rates

The crises of the 1990s, including Asia’s in 1997-98, made it impossible to deny that “intermediate” exchange rate arrangements—those lying between very hard pegs and relatively free floats—are fragile and crisis prone. These crises provided compelling evidence of the costs when pegged rates collapse. The decline in the value of the currency, when it comes, is “larger, more rapid and more unanticipated than when a depreciation occurs under a floating exchange-rate regime” (Mishkin, 2001). The distress suffered by banks and firms with unhedged foreign currency exposures is greater. There is at least impressionistic evidence that the associated output losses are larger.

These effects can be understood as consequences of the build-up of unhedged foreign exposures among banks and corporations led to believe that they are insured against currency fluctuations by the authorities’ commitment to peg the rate, and of the shock to confidence that results when the anchor for monetary policy, namely the one and same exchange rate peg, is withdrawn abruptly. A more prudent arrangement, the conclusion follows, would be to move to a more flexible rate voluntarily rather than under duress. Currency fluctuation would then remind borrowers of the need to limit their borrowing and hedge their exposures, and the central bank and government would have to develop independent sources of credibility. Alternatively, the country could adopt a hard peg to a foreign currency or make that foreign currency the exclusive circulating medium and legal tender, eliminating currency risk and allowing policy credibility to be imported quickly rather than laboriously grown at home.

This wisdom became increasingly conventional following the Asian crisis. In an April 1998 speech on international financial reform, then-U.S. Treasury Secretary Robert Rubin declared that the IMF should no longer provide financial assistance to support overvalued currency pegs. His successor, Lawrence Summers (2000, p. 8), stated that “the choice of appropriate exchange rate regime for economies with access to international capital markets, increasingly means a move away from the middle ground of pegged but adjustable fixed exchange rates towards the two corner solutions of either flexible exchange rates or a fixed exchange rate supported, if necessary, by a commitment to give up altogether an independent monetary policy.” Stanley Fischer, drawing on IMF staff’s assessment of the de facto exchange rate regime in IMF member countries, showed that the proportion with intermediate arrangements (neither hard pegs nor floats) was significantly lower in 1999 than 1991 (34 percent versus 62 percent for all countries, 42 percent versus 64 percent for emerging markets).¹⁹ Fischer’s testimony to the Meltzer Commission forecast that more countries would move to the corners.²⁰ IMF staff’s most recent paper for the Executive Board on exchange rate arrangements (Mussa et al., 2000) similarly concludes “for developing countries with important linkages to modern global capital markets… the requirements for sustaining pegged exchange rates have become significantly more demanding. For many emerging market economies, therefore, regimes that allow substantial exchange rate flexibility are probably desirable. Some emerging market countries, of course, may go in the other direction—toward hard currency pegs (such as currency boards), supported by the requisite policy discipline and institutional structures.”

Thus, an international system of hard pegs and relatively free floats is seen as the monetary component of the new international financial architecture. For members of the “missing middle school”, this consensus, together with the observed movement away from intermediate arrangements, is an area of real progress in rationalizing the international financial system (see for example Fischer, 2001).

Less heartening is that many countries—most recently Turkey—have moved away from intermediate arrangements not in a measured way but as the result of a crisis. Also disturbing is the suspicion that a not insignificant number of emerging markets that claim to operate floating rates intervene heavily to damp currency fluctuations, in practice operating what amounts to soft (“noncredible”) pegs (Calvo and Reinhart, 2000). Insofar as the authorities have changed the name but not the reality of the exchange rate regime, firms and banks, still confident that they are protected from large currency fluctuations by the authorities’ implicit commitment to stabilize the rate, may fail to hedge their exposures, while the authorities, for their part, will fail to invest in the development of a new monetary anchor and a source of credibility independent of the exchange rate. The crisis problem may have been disguised, in other words, but it has not been eliminated.

It is not hard to understand why countries are reluctant to abandon intermediate exchange rate arrangements. Abandoning the national currency for the dollar (or the euro) is a symbolic sacrifice, as acknowledged even by those who believe that dollarization (used henceforth as a generic term) has more benefits than costs. Abandoning a peg for greater flexibility will be a shock to the balance sheets of both the public and private sectors. It will undermine confidence if the exchange rate had previously served as the anchor for monetary policy and if the commitment to peg it had been an important source of policy credibility for the government and the central bank. Knowing that exiting from the peg will be a shock to confidence and it may precipitate a recession, governments are understandably inclined to put off the decision.²¹

It follows that if countries are not going to exit voluntarily from soft pegs, the IMF needs to counter this status-quo bias. The Council of Foreign Relations Task Force on Strengthening the International Financial Architecture (Council on Foreign Relations, 1999) argued that the Fund should provide the appropriate incentives by committing not to provide large-scale financial assistance to governments intent on defending overvalued currency pegs. But the qualification that the Fund should not support “overvalued” currency pegs provides a convenient pretext for exceptions, and the pressure to make exceptions on political grounds will remain great. In other words, the CFR did not explain how such a commitment could be made credible. The Meltzer Commission, while urging the more active use of Article IV consultations to remind countries of the risks associated with pegged rates, did not propose that their abandonment should be a precondition for IMF assistance.

The reality is that the IMF is an institution of many members. Reflecting their diverse structures, histories, and circumstances, there is no consensus among them about the appropriate exchange rate arrangement, forcing the Fund to qualify its recommendations. Nor is it clear that the IMF is any better placed than governments to trade off the political and economic costs of recession now for the benefits of a more robust exchange-rate regime later. The Fund’s own status-quo bias is clear in the recent cases of Argentina and Turkey. The institution was a strong supporter of Argentine convertibility until the final months of 2001, reflecting fears that abandoning that regime would be a sharp shock to confidence. In the case of Turkey, the 2000 IMF program sought to sustain the lira’s crawling peg regime for 18 months rather than moving immediately to a more flexible rate.

This last case is revealing of the underlying dilemmas. Diagnoses of Turkey’s problems were informed by the Asian crisis. The banks had received implicit guarantees as the price of being used as instruments of the government’s industrial and agricultural policies. The combination of a pegged exchange rate and an open capital account encouraged them to leverage their bets. A more flexible exchange rate could solve this problem by encouraging the banks to more prudently manage their exposures, but a sudden change in the rate would disturb balance sheets and confidence. Hence, the IMF program sought to move gradually in the direction of greater flexibility over a period of 18 months, at the end of which the banks would have limited their exposures sufficiently that neither they nor the Turkish economy would be destabilized.

The Turkish crisis showed that this program was shaped by wishful thinking. The preconditions for holding the exchange rate stable are formidable when political commitment is tenuous and banks are weak. The knowledge that the authorities plan to move to greater flexibility in the future creates the perception of a one-way bet. And hopes that the banks will hedge their exposures today in response to the knowledge that the exchange rate will be allowed to float more freely tomorrow will be dashed if those banks are public or well connected politically and if tomorrow is not scheduled to come for 18 months. In the end, Turkey is just another example of the difficulties of operating an intermediate arrangement for even a limited period of time.

Post-Asian crisis statements by the U.S. Treasury recommended that emerging markets move toward more flexible rates. This reflected worries that widespread dollarization might create political and diplomatic complications if the Federal Reserve did not adjust U.S. monetary policy to accommodate the needs of the newly dollarized economies, together with skepticism of the workability of monetary union. Europeans have always been more sympathetic to pegs, reflecting lessons drawn from the currency turbulence of the 1930s, and more optimistic about the prospects for monetary union. And now that a European Monetary Union exists, joining it provides an obvious solution for Eastern European countries seeking to eliminate the exchange rate problem.

The IMF, for its part, has overcome its initial skepticism regarding currency boards, dollarization, and monetary unification, having been influenced by the successes of these arrangements in the 1990s. IMF economists on the way to the staff cafeteria traverse a corridor lined with cases displaying the currency notes of the institution’s members; throughout the 1990s they regularly referred to this fact, proclaiming that the rule of “one country, one currency” was one of the most robust regularities in monetary economics. Europe’s success in launching the euro would appear to have rendered this conviction less firm.²² And, while not exactly an enthusiast of dollarization, the Fund did not resist Ecuador’s decision to adopt the U.S. currency.

But if only the corners are viable, who should move to which one? In Europe, where there is a commitment to political as well as economic integration, monetary union is the best option. Each country enjoys currency stability vis-à-vis its principal trade and financial partners but flexibility useful for facilitating adjustment against the rest of the world. The commitment to political integration allows the creation of institutions of shared governance, in turn enabling each member to have a voice in the common monetary policy. This is a luxury not enjoyed by an El Salvador that unilaterally adopts the dollar or a Hong Kong that pegs to it via a currency board. Absent a comparable commitment to political integration in Latin America and Asia, there are reasons to question whether there exist realistic prospects for monetary union on those continents in coming decades.²³

A hard peg is the obvious solution for small countries with ties to larger neighbors and/or weak institutions. Small countries that trade heavily with a single larger partner and rely on it for external finance satisfy the classic optimum currency area criteria for pegging. It makes eminent sense from this point of view for El Salvador to adopt the U.S. dollar and for Estonia to peg to the euro.²⁴ Similar arguments can be made for countries with underdeveloped financial markets that leave firms and banks unable to hedge against currency fluctuations, and with weak institutions, chronic budget deficits, and bankingsector problems that prevent the authorities from credibly committing to low inflation. Advocates of dollarization will argue that we have just described the universe of emerging markets. In other words, since all emerging markets are characterized by these conditions, the world is destined to move toward three currency blocs centered on the dollar, the euro, and the yen.²⁵

Even leaving politics aside, there are reasons to doubt this forecast. The collapse of Argentina’s currency board is indicative of some of the problems with wider dollarization. Brazil, not the United States, is Argentina’s most important trading partner. Argentina attracts foreign investment (or hopes to!) from Spanish banks and European bond markets. Hence, its single-currency peg created serious difficulties for the economy when the euro or the real depreciated against the dollar. Asia faces a similar dilemma, as evidenced by the near-fatal consequences in 1997 of the combination of dollar pegs and dollar-yen fluctuations.²⁶

Thus, there remains a case for independent floating by countries with the capacity to operate such a system, assuming—contrary to the assertions of the advocates of dollarization—that these countries exist. Such countries will presumably have diversified trade and financial linkages. They will have independent central banks, well regulated financial systems, efficient fiscal institutions, and stable political institutions. These preconditions should enable monetary policy makers to acquire credibility and encourage them to follow sound and stable policies without orienting monetary policy around a particular value for the exchange rate. As their commitment to do so gains credibility, the volatility of the exchange rate (which reflects uncertainty about future policy) will diminish accordingly.

This requires the authorities to articulate and implement an alternative monetary policy operating strategy.²⁷ The leading candidate is inflation targeting. Inflation targeting entails an institutionalized commitment to price stability as the primary goal of monetary policy, mechanisms rendering the central bank accountable for attaining its monetary policy goals, the public announcement of targets for inflation, and a policy of communicating to the public the rationale for the decisions taken by the central bank.²⁸ Central bank independence is needed to give the monetary authorities the leeway necessary to commit to price stability. And a stable fiscal policy and banking system are needed to avoid problems of fiscal dominance that would otherwise prevent the central bank from subordinating other goals to the objective of price stability or cause its independence to be undermined.

The multi-dimensional nature of this definition explains why there is no consensus about which emerging markets practice inflation targeting. Brazil, Chile, the Czech Republic, Israel, South Africa, Poland, Colombia, Thailand, Mexico, the Philippines, and South Korea are all cited in this connection. While the length of this list would appear to pose a challenge to those who argue that floating backed by inflation targeting is not viable in emerging markets, there is the question of whether these countries are actually practicing inflation targeting or are really just covertly pegging their currencies. That some, like Israel, also maintain bands for their exchange rates, while others, including South Korea, intervene to limit or offset exchange rate fluctuations, has been pointed to as evidence that they are really covert peggers. And since a soft commitment to peg the exchange rate is the worst of all arrangements, it follows according to this view that these countries would be better off dollarizing.

But the fact that their central banks alter monetary policy when the exchange rate moves may not in fact mean that they are not inflation targeting. Insofar as depreciation is a leading indicator of inflation, the standard inflation targeting framework suggests tightening monetary policy when the exchange rate weakens. This is the prescribed response when there is a change in the direction or availability of capital flows due to, say, a rise in foreign interest rates, or a deterioration in foreign investor sentiment toward the country. A higher foreign interest rate implies less capital inflow for a given domestic interest rate and therefore a weaker currency. As the exchange rate weakens, higher import prices are passed through into inflation. The appropriate response is thus to raise interest rates. This is not because the central bank cares about the exchange rate in and of itself, but because it cares about inflation.

If the disturbance is to the foreign component of aggregate demand (to the terms of trade or export demand), again the exchange rate will weaken, since export revenues will have declined while nothing else affecting the foreign exchange market will have changed in the first instance. In addition, demand will decline, since foreigners are demanding fewer of the country’s exports. Now there are two offsetting effects on inflation: while higher import prices will be passed through into inflation, weaker aggregate demand will be deflationary. If the second effect dominates, then inflation will decline with the growth of the gap between potential and current output, and the appropriate response for an inflation-targeting central bank will be to cut interest rates regardless of the weight it attaches to output variability. The more policy relevant case is probably the one in which this shock, by depreciating the exchange rate, is inflationary on balance. If the central bank attaches a high weight to output variability, it still may want to cut interest rates. If, on the other hand, it attaches a high weight to deviations of inflation from target, it may instead raise interest rates to limit currency depreciation and inflation in the short run, while still allowing the exchange rate to adjust eventually to its new long-run equilibrium level. In other words, it will acknowledge that the weakness of demand requires a weaker exchange rate, but it will still smooth the downward adjustment of the currency by leaning against the wind in order to prevent a sharp spike in inflation.

This framework suggests a test of whether countries like South Korea have really begun to practice inflation targeting. Find instances where the shock is to commodity rather than capital markets, since these are the cases where the weakness of output creates a case for interest rate cuts and allowing the exchange rate to adjust. Consider not just the tendency for the central bank to lean against the wind in the short run but also its willingness to allow the exchange rate to adjust subsequently. 2001 is a useful data point insofar as the shock to the Korean economy is mainly from commodity markets (reflecting the U.S. recession, continuing Japanese difficulties, and the high-tech slump). An inflation targeting central bank will tighten as inflation heats up, but it will also allow the exchange rate to adjust if the slump persists. My reading of Korean experience is that the central bank, while concerned with inflation, nonetheless allowed the won to depreciate in 2000-01 with the slowdown in the global electronics industry and therefore in domestic growth. This does not look like fear of floating to me.

A final characteristic of emerging markets that may affect their ability to target inflation is liability dollarization. In many emerging markets, the obligations of banks, corporations, and governments are denominated in foreign currency, while the bulk of their revenues are domestic-currency denominated.²⁹ When the exchange rate depreciates, their balance sheets suffer, and this “financial accelerator” significantly depresses output and employment.

The simplest way of thinking about liability dollarization is as reducing the response of output to currency depreciation.³⁰ While depreciation renders domestic goods more competitive, as before, now it also weakens the balance sheets of banks, firms, households, and governments, depressing consumption and investment. Consider the response to a negative shock to capital markets. Weaker consumption and investment due to adverse balance-sheet effects now imply less inflation in the intermediate run. An inflation-targeting central bank will therefore feel less compelled to raise interest rates in order to push up the exchange rate and damp down the increase in import prices. If the shock to the exchange rate emanates instead from commodity markets, higher import prices will still be passed through into inflation, but now aggregate demand will be even weaker than before because of the adverse balance-sheet effects. Since output is lower and inflation is no higher than in the absence of liability dollarization, again there will be less pressure to hike interest rates in order to stabilize the currency and damp down inflation, and more incentive to cut interest rates to stimulate production. Surprisingly, this suggests, regardless of the source of shocks, that reluctance to let the exchange rate adjust will be less in the presence of liability dollarization.

While this may seem counterintuitive, it is just a specific illustration of the general point that when the central bank worries more about variables other than inflation, either because those variables have a heavier weight in its objective function or because the parameters of the model cause those other variables to be displaced further from their equilibrium levels (where the latter is the case presently under discussion), it will move more gradually to eliminate discrepancies between actual and target inflation. Because the exchange rate must move more to increase output and employment, and because measures which would limit its fluctuation and thereby reduce imported inflation tend to destabilize the economy, the now weaker tendency for depreciation to stimulate activity means that the central bank will do even less to limit depreciation.³¹

Clearly, those who argue that liability dollarization creates fear of floating have something else in mind, presumably that the balance-sheet effects of currency depreciation are so strong that a cut in interest rates, which weakens the exchange rate, depresses output on balance. In this case, a negative shock to capital markets still fuels inflation through higher import prices, encouraging the authorities to raise rates. But now, in addition, it lowers output through the adverse balance-sheet effect. The appropriate response, which damps down inflation and stabilizes output by limiting balance-sheet damage, is to raise interest rates and push the exchange rate back up toward its preshock level.“Fear-of-floating” type behavior results. If the disturbance is instead to commodity markets, the weaker exchange rate again means more imported inflation and lower levels of output.³² Again, interest rate hikes are the appropriate response to both problems, since a higher interest rate that strengthens the exchange rate not only damps down inflation but also strengthens balance sheets. Again, the central bank will not hesitate to raise interest rates. Again, its response will resemble fear of floating.

This formulation has some peculiar implications. For one, a negative commodity market shock that reduces export demand and depresses output must be offset in the new equilibrium by an appreciated exchange rate, not a depreciated one. This is a world where overvaluation is good for output because its favorable financial effects dominate its adverse competitiveness effects, even in the long run, which hardly seems realistic.

A possible reconciliation is that when the exchange rate depreciates by a large amount, the adverse balance-sheet effects dominate, but when it depreciates by a small amount, the favorable competitiveness effects dominate. Large depreciations cause severe financial distress because they confront banks and firms with asset prices for which they are unprepared while doing little to enhance competitiveness because of the speed with which they are passed through into inflation. For small depreciations, the balance of effects is the opposite. Small depreciations are more likely therefore to satisfy the conditions for an expansionary devaluation.³³

If the exchange rate falls sufficiently to enter the first range, then an inflation-targeting central bank will raise interest rates sharply with the goal of pushing up the currency and minimizing the financial damage to banks, firms, and households. But if the depreciation is modest, so too will be the rise in interest rates; the central bank will allow the currency to fall to a new lower level so long as the competitiveness effects continue to dominate the balancesheet effects.³⁴

Thus, whether emerging markets can implement an inflation targeting regime that allows the exchange rate to fluctuate more freely depends on the precise extent and effects of their liability dollarization. If even a small depreciation of the exchange rate threatens to destabilize balance sheets and the macroeconomy (that is, the country immediately enters the zone where depreciation and lower interest rates are contractionary and destabilizing financially), then the central bank will not be willing to let the exchange rate move. Although the preceding propositions for how the central bank should respond flow directly from the standard inflation-targeting framework, inflation targeting and a hard peg are basically indistinguishable under these conditions. If the perceived advantage of inflation targeting is that it permits greater exchange rate flexibility, then the advantages of inflation targeting are correspondingly less in highly dollarized economies. Inflation targeting then has no obvious advantages over a hard peg, which has the merits of simplicity, transparency, and credibility.

For countries where the adverse balance sheet effects dominate only when exchange rate movements exceed a critical threshold, inflation targeting will be viable so long as shocks and exchange rate fluctuations are small, while the desire to intervene and stabilize the exchange rate will dominate when they grow large. The additional exchange rate flexibility promised by inflation targeting will be feasible, but the central bank’s appetite for indulging in it will have limits. When those limits are reached, intervention to stabilize the exchange rate will become its overriding objective.³⁵

This discussion suggests that exchange rate arrangements consistent with the new architecture will vary not by region but with the characteristics of individual countries. Small countries that depend heavily on trade and financial transactions with a single larger partner will prefer to peg rigidly or perhaps even to adopt the latter’s currency. Larger countries where the extent of liability dollarization is limited, fiscal institutions are reasonably strong, and the political commitment to price stability is firm will prefer to practice inflation targeting and allow the exchange rate to vary. There are examples of both kinds of arrangements in Latin America (contrast El Salvador and Brazil) and Asia (contrast Hong Kong and Korea). The present perspective does not suggest that exchange rate arrangements will be uniform within regions or vary in some systemic way across them. It does not suggest a distinctively Asian solution to the exchange rate problem.

Regional Responses

The previous point notwithstanding, calls are frequently heard for a regional response. The Asian crisis elicited proposals for a new institution to free Asian countries from dependence on the IMF. IMF financial assistance, in the Asian view, was too little, too late, and the Fund’s conditions were not well suited to the Asian model. The IMF’s insistence on high interest rates as a means of restoring confidence proved counterproductive, given the high debt-equity ratios of corporations in the region. Requiring governments to close problem banks in the midst of the crisis produced a credit crunch, given the dependence of Asian economies on bank finance. The Fund’s call for tax increases and public spending cuts was inappropriate for countries that did not enter the crisis with fiscal imbalances and only served to aggravate the recession. IMF surveillance failed to place sufficient weight on the dangers of premature and poorly sequenced capital account liberalization and on the special risks posed by large players in small financial markets. And the Fund’s infatuation with flexible exchange rates was incompatible with Asian countries’ preference for trade-friendly pegs. In response, the concept of an Asia Fund was tabled by the Japanese government soon after the outbreak of the crisis, leading to the negotiations that ultimately culminated in the so-called Chiang Mai Agreement on an Asean+3 network of currency swaps.

As the preceding paragraph makes clear, there are several rationales for institution building at the regional level. The fact that crises have a regional component (Glick and Rose, 1999) sharpens the incentive for neighboring countries to engage in mutual surveillance and to extend assistance to one another in the face of potentially contagious threats to stability. Because intraregional trade is of growing importance (a quarter of Asean’s trade is within the grouping and another 17 percent is with Japan) and many Asian countries sell into the same markets outside the region (whether measured by the geographical direction of trade or the similarity of its commodity composition), unilateral exchange rate depreciation can seriously erode the competitiveness of a country’s neighbors.³⁶ The popular perception is that many of these countries have similar economic structures and, by implication, vulnerabilities; in conjunction with the fact that they have common creditors, this implies that difficulties in one can undermine investor sentiment and the availability of finance to the others.

The subtext here is that the conditions attached to foreign assistance would be better attuned to the realities of the Asian model were they formulated by Asian countries for Asian countries. Asian policy makers, the argument goes, better understand the distinctive characteristics of the Asian model, allowing them to reach appropriate conclusions about the structure of assistance and conditionality. Obviously, this is a disputable point: it is difficult to think of three economies whose structures differ more than South Korea, Hong Kong, and China.³⁷

It is hard to quibble with the argument for a market in ideas. The IMF has no monopoly on macroeconomic and financial wisdom. Regional funds are presented as a mechanism for intensifying this competition and allowing good ideas to drive out bad. But would they? Here it is important to distinguish between technical assistance and financial assistance. There is no reason to discourage countries from taking technical assistance from the provider with the best track record.³⁸ But development assistance is not the core competency of the IMF; rather, it is domain of the World Bank, the regional development banks (including the Asian Development Bank), national agencies such as USAID, and nongovernmental organizations.³⁹ There is already a market for ideas, in other words. But in the case of financial assistance in these of crisis, it is not clear that competition among funding agencies will result in the good ideas crowding out the bad. Governments will have an incentive to shop around for the most generous assistance and the least onerous terms. Only if one is convinced that IMF conditionality is uniformly too strict and its terms too paltry will competition guarantee better outcomes. Those who see the IMF as too willing to lend and the resulting moral hazard as corrosive of market discipline will not be convinced.⁴⁰

Among the most vocal proponents of a regional response are those who advocate a cooperative exchange-rate stabilization arrangement for Asia. The most prominent proposal is for a common basket peg with weights on both the U.S. and Japanese currencies.⁴¹ The basket structure of the peg would absorb many of the effects of yen-dollar fluctuations, insulating Asia’s economy from an important source of shocks from outside the region, while common weights in the different baskets would rule out intra-regional exchange rate fluctuations due to fluctuations in non-Asian currencies.⁴² The second element in particular presupposes cooperation at the regional level.

But dollar-yen fluctuations are not the only source of shocks to Asian financial markets. Intra-regional exchange-rate fluctuations are not the only transmission belt. In a world of high capital mobility, simply declaring common basket pegs is not enough to ensure their viability. It does not insulate the economies in question from all disturbances, nor does it ensure the robustness of the resulting exchange rate system. Sustainability may require very extensive financial assistance to protect the participating currencies against speculative pressures. This, in turn, presupposes institution building to lend credibility to that commitment and to render it incentive compatible. And to the extent that currency pegs can be laid low by financial problems internal to the region (banking-sector problems, for example), robustness will require measure strengthening prudential supervision and regulation.⁴³ This presupposes institutions of mutual surveillance to render this commitment credible and incentive compatible. The countries extending the swaps will be prepared to do so only if they believe that those receiving them will not follow reckless policies leading to reckless demands for support—and only if they believe that they will be paid back.

Thus, financial support on the scale needed to sustain a regional exchange rate stabilization agreement is likely to be forthcoming only if mechanisms are developed compelling the participants to strengthen their financial policies and systems, together with institutions to monitor their compliance and sanctions for failure to comply. By implication, the hope that a regional arrangement of this sort would not have to be burdened by the oversight and conditions that accompany IMF programs is likely to be disappointed.

The Asean+3 system of swaps and credits illustrates several of these dilemmas. Swap agreements already existed between some Asean members at the time of their 1997 crisis, but these were ineffective in containing the crisis. One diagnosis of this failure was that their scale was inadequate; hence, the new arrangement, encompassing not just Asean but also China, South Korea, and Japan, marshals more than $800 billion of reserves. This is an impressive number, but it pales in comparison with the financial resources of investors worldwide. Under what conditions, then, can such a system deliver the promised exchange rate stability? Only when its members pursue financial reforms that limit the likelihood that their reserves will have to be utilized.⁴⁴ In particular, only when they have strong financial systems will they be able to raise interest rates to defend their currencies (without ratcheting up the pressure on already fragile financial institutions to intolerable levels) rather than immediately appealing for foreign support. It is no coincidence, then, that the Chiang Mai Initiative committed its signatories to cooperation on prudential supervision and mutual surveillance of their banking systems.⁴⁵ Whether this commitment will be pursued with vigor is not yet clear. But if the signatories are serious about extending significant financial assistance to one another through their network of swap lines, then there will be pressure to move in this direction. It is also revealing that the Asean+3 countries linked the expansion of their swap arrangements to the development of an Early Warning System designed to identify the need for policy adjustments that might avert or minimize the need to draw. With the expansion of the swap network to include China, South Korea, and Japan, there is now predictable discussion of the need to expand the Early Warning System. One wonders whether the members would be willing to extend financial support for the exchange rates of countries that did not heed their early warnings, and if so what conditions would be attached to such assistance.

It is not clear, in other words, that this arrangement will ultimately oblige its participants to meet fewer conditions than its IMF analog. One waits to see what conditions the strong-currency countries attach when they are asked to extend significant credits.⁴⁶

This brings us finally to IMF conditionality. The expansion in the number and range of conditions attached to IMF loans became a bone of contention following Asia’s crisis. The feeling was that the Fund’ structural conditions were invasive, that they were formulated without reference to the government’s capacity to muster support for reform, and, to the extent that they pointed out structural weaknesses of the program countries, that they did more to destroy than to strengthen confidence. The 1997 Indonesian program is seen as typifying the problem: its loan conditions dealt with reforestation, the national car program, local content programs for motor vehicles, the compulsory 2 percent after-tax charitable contribution, restrictive market agreements for cement and paper, the forced planting of sugar cane, the introduction of a micro-credit scheme for small businesses, and the elimination of the Clove Marketing Board.⁴⁷ An Asian Fund would be less invasive, the argument runs, given the tradition in the region of non-interference in other countries’ economies.

Although the number of structural conditions attached to IMF programs peaked in 1997, the increase in such measures was not specific to Asia’s crisis. The number of structural conditions per program more than doubled between the late 1980s and early 1990s and doubled again between the first and second half of the most recent decade (IMF, 2001a, p. 3, Figure 1). Explanations for the trend include the priority that came to be attached to restoring growth in highly-indebted countries in the 1980s (complaints that the Fund placed too much weight on stabilization and therefore was too tolerant of recession led to this emphasis on growth, which in turn required structural reform) and the emphasis on structural transformation and institution building in formerly centrally planned economies in the 1990s.⁴⁸ While these experiences help to explain why the Fund became accustomed to giving growth-related advice and applying structural conditions, they do not justify its preoccupation with micro- and sectoral reforms in East Asia, a region with an admirable record of growth and no history of deep structural problems. They do not explain why the conditions attached to the Fund’s 1997 and 1998 programs with Indonesia, South Korea, and Thailand were so numerous and detailed.

In a sense, the fact that most Asian countries had not entered their crisis with chronic macroeconomic imbalances was the very observation that motivated the emphasis on structural reforms. If macroeconomic imbalances were not the problem, then macroeconomic performance criteria were not the solution. If it was structural weaknesses in Asian financial markets that caused the crisis to take such a toll, then it was structural reform of financial markets that was called for. The argument for structural conditions on loans extended in response to the crisis is then the same as the argument for international standards to encourage the upgrading of financial markets and institutions before the fact.⁴⁹ That the Asian crisis, like many other recent crises, was a capital-account crisis rather than a current-account crisis inevitably blurred the line between measures that were needed to address the underlying instabilities and measures that were not. Solving a current-account crisis simply requires measures that will reduce absorption, while solving a capital-account crisis requires restoring confidence, which is a much more nebulous matter.

Still, if it is crises in financial markets that the IMF is in the business of averting, and flaws in the structure and regulation of financial markets that are the problem, then it is hard to justify conditions related to government policies remote from financial markets, such as reforestation or clove marketing. In principle, there is a line between structural measures that promise to significantly strengthen financial institutions and markets and others that are only peripherally related to macroeconomics and finance. The IMF, not unaware of the backlash that some of its more invasive conditions provoked, has attempted to take this distinction on board.⁵⁰ To raise the consciousness of staff and management, it has circulated an Interim Guidance Note on Streamlining Structural Conditionality that emphasizes “the need to limit conditionality to those measure that are critical for achievement of the program’s macroeconomic objectives.”⁵¹

But determining precisely which measures are critical for macroeconomic balance and financial stability is easier said than done.⁵² Some would go as far as to argue that reform of the Indonesian timber monopoly was essential to signal a new resolve to root out cronyism and corruption, and that the confidence of financial markets could not be regained in the absence of that resolve. Even the Indonesia program, in this view, was not overly ambitious or detailed. But even those who criticize these conditions as going too far would acknowledge the difficulty of devising general rules for what kinds of reforms are and are not needed for the restoration of investor confidence.⁵³ We know to be suspicious of a consensus—in this case in favor of simplified conditionality—when we see everyone from the Meltzer Commission to the Council on Foreign Relations to the Group of Twenty Four to the IMF chiming in. We suspect that they really have different things in mind, at least when it comes to moving from principle to practice.

Moving from process-based to results-based criteria is an obvious way of streamlining conditionality. The IMF’s continued willingness to lend could be based on the authorities’ observed success in recapitalizing the banking system, for example, rather than on the specific steps taken with regard to problem banks. This would avoid micro-management of reforms and give the national authorities more leeway to adopt measures appropriate to local circumstances. But insofar as results take time to materialize, it would be inconsistent with the G-7 preference for shorter-duration loans to discourage chronic borrowing and for front-loading IMF disbursements to strengthen confidence.

The proof of the pudding is in the eating: one may ask, for example, whether the Fund’s recent programs for Argentina and Turkey are significantly leaner than their predecessors. This is hard for those on the outside to judge: we have access only to the Letter of Intent (LOI), which describes not only policy measures on which the Fund’s financing is conditional but also other aspects of the authorities’ policy program. For what it is worth, Argentina’s December 2000 LOI, together with the attached Memorandum of Economic Policies, describes measures designed to update and strengthen antitrust legislation, a new regulatory framework to support development of the ports system, the consolidation of three nutritional programs, and a new regulatory framework for the telecommunications sector.⁵⁴ Turkey’s LOI of January 2001 includes provisions related to reform of the tobacco sector and the rolling out of tax identification numbers.⁵⁵

Thus, efforts to refocus and streamline IMF conditionality are one place where Asia’s message has been heard.⁵⁶ But only time will tell whether the Fund can successfully implement the principle. Could a regional fund do better? There is no obvious reason for thinking so.

Summary Assessment

I now return to the questions posed at the outset: how meaningful are the changes to the international financial architecture currently underway—are they merely cosmetic or do they constitute a significant attempt to modify the structure and management of the international financial system? Different readers will have different views of whether the glass is half empty or half full. Mine is that the effort to extend international standards to areas beyond the strictly monetary is, or at least promises to become, a very significant development affecting the structure and stability of the international financial system. Upgrading macroeconomic policy and transparency, financial market infrastructure, and financial regulation and supervision is essential for stability in a financially integrated world. And international standards, with pressure to comply to be applied by multilateral surveillance, IMF conditionality, regulation, and market discipline, are the only available means to this end in a world of sovereign states. Significant obstacles remain before this initiative can be operationalized: the design of the relevant standards is contested, there is resistance to their use in IMF conditionality, and the markets have not yet displayed a readiness to rely on them when making lending decisions. Still, I can imagine looking back years from now and seeing the standards-centered initiatives of the late 20^th and early 21^st century as a significant turning point.

The question of whether the glass is half empty or half full also applies to changes to the exchange rate system. My view is that the observed tendency for countries to evacuate the middle ground between very hard pegs and relatively free floats is important progress. It is a step in the right direction. Less reassuring is continued disagreement about the appropriate exchange rate arrangement for countries in particular circumstances, the reluctance of emerging markets to allow their currencies to fluctuate, and the impulse to restore soft pegs and bands on a regional basis. Hopefully it will not take another round of crises to demonstrate which of these tendencies are positive and which are not.

Reforms of IMF lending and conditionality are clearly in the manner of fine tuning rather than radical changes. Modest increases in interest rates and reductions in term are unlikely to do much to modify the propensity for governments to borrow from the Fund. A higher threshold of relevance to program objectives, more reliance on institutions like the World Bank in areas that are outside the core competency of the Fund, and a modest reduction in the number of structural benchmarks that a country must meet hardly represent a radical recasting of IMF conditionality. They fall short of the recommendations of the Meltzer Commission, and for good reason. My view is that this is unavoidable—that conditionality must be broader and therefore fuzzier in a financially integrated world where crises are driven by capital-account problems.

Do these initiatives address Asia’s concerns about the volatility of financial markets, the structure of conditionality, and the operation of the international monetary system? The answer is “incompletely, at best.” A more satisfactory outcome requires that Asian countries represent their views more effectively and that the multilaterals reform their voting procedures and deliberations to better enable those views to be heard. But it also requires understanding that the volatility of financial markets, the fragility of pegged exchange rates, and the breadth of IMF conditionality are unavoidable in a world of financial integration. No changes in the international financial architecture that fall short of turning back the clock on globalization can change these uncomfortable facts.