Information about Asia and the Pacific Asia y el Pacífico
IEO Evaluation of Exchange Rate Policy

Chapter 3: What Has Been the Quality of IMF Analysis and Advice?

International Monetary Fund. Independent Evaluation Office
Published Date:
August 2007
  • ShareShare
Information about Asia and the Pacific Asia y el Pacífico
Show Summary Details

18. There is a disconnect between the perceptions of staff and of the authorities regarding the quality of IMF exchange rate analysis and assessment—particu-larly in respect of advice to major advanced and large emerging market economies. While two-thirds of all staff respondents considered that the IMF’s exchange rate policy advice had improved or significantly improved over the period covered by the evaluation (since about 1999), country authorities overall were less convinced. Just over half of authorities’ survey respondents saw improvement. Responses grouped by type of country revealed a more differentiated, and—taken with the observations of impact noted above—more worrisome pattern (see Figure 3.1). In particular:

  • Within the group of large emerging market economies, appreciation for the quality of the IMF’s advice was significantly weaker: about 70 percent of respondents from the authorities reported that the quality of policy advice was unchanged or worse than a few years ago. By contrast, 70 percent of responses from staff working on these economies considered IMF advice to have improved or significantly improved. Notably, these economies had received much attention over the period (involving crisis prevention and resolution efforts in some). In part, the authorities’ responses may reflect discontent with the implementation of IMF surveillance in general, and the very challenging policy environment that they face, for which there were no easy answers. Certainly the quality of the IMF’s advice cannot be judged simply by how popular it is. Nonetheless, in the IEO’s opinion, the authorities’ views were validated by examples they provided of how the advice could have been improved.
  • A mixed message could be taken from the advanced economies, the majority of whose respondents reported no change (in the case of major advanced economies) or improved quality (in the case of other advanced economies), but also for the most part said that IMF views had little or no bearing on their decisions.
  • The most appreciative of the IMF’s efforts were the authorities in 60 percent of other emerging market and developing economies, whose perceptions of improved quality, as well as of impact, matched those of the staff. However, only limited comfort should be drawn from this result. The authorities in these countries, many of whom saw IMF advice as instrumental in the context of IMF-supported programs, also indicated several areas in which the quality of advice could be improved significantly. Moreover, with the prospect of fewer IMF-supported programs, greater analytical capacity in many countries and further European integration, the challenges for the IMF to remain relevant in these economies will increase too.

Figure 3.1.Perceived Change in the Overall Quality of IMF Staff’s Exchange Rate Analysis and Assessment

Responses on the change in overall quality of IMF staff’s analysis and assessment over the evaluation period

(In percent)

All in all, the results were indicative of a gap between the existing quality of advice and that which would be found useful by many authorities, especially in advanced and emerging market economies. Interviews with country authorities gave credence to this finding. While some officials stressed that the quality of analysis was excellent, and clearly valued, others (and not just those who may have disagreed with the advice given) were quite blunt in saying that it fell short of what would have been appropriate and helpful.

19. What could explain these different perceptions? The evaluation found several aspects in which the quality of exchange rate advice had improved, but also examples of why it had failed to persuade. It focused on eight elements of quality, including aspects of advice and its analytical basis for which some guidance had been given to staff.1

  • Coverage of exchange rate issues, including linkages with other policy areas (see the section “Coverage of Exchange Rate Issues”).
  • A description of the exchange arrangement has been a standard requirement throughout, but, in addition, from 2004, staff have been required to “accurately identify and describe the de facto exchange rate regime” in place (see the section “Regime Identification”).
  • A description of intervention policies and practices is necessary to describe and assess exchange rate policy (see the section “Intervention and Related Policies”).
  • The requirement for staff to “give a candid assessment of the impact of deficiencies in the timeliness and/or quality of data provided to the IMF on the staff’s ability to conduct effective surveillance,” from the 1997 BSR was taken up in subsequent Board meetings, including on the IMF’s reserves template, and in a 2005 guidance note on data provision for surveillance purposes (see the section “Data Requirements”).
  • An assessment of the exchange rate level is required, according to the Board guidance, from the 2000 BSR (see the section “Analysis of the Level of Exchange Rates”).2
  • An assessment of the exchange rate regime is to be made in all cases, guidance also dating from 2000 (see the section “Advice on Exchange Rate Regimes”).
  • The integration of multilateral and regional perspectives (see the section “Multilateral and Regional Perspectives”).
  • The consistency and evenhandedness with which advice was provided (see the section “Consistency and Evenhandedness of Advice”).

Coverage of Exchange Rate Issues

20. Strikingly, in a number of cases, substantive discussions with the IMF did not cover important exchange rate topics that were live issues for the country at the time. Some authorities perceived that in discussions with the IMF certain exchange rate issues received less focus than in internal debates (Figure 3.2). The failure to cover topics comprehensively was reflected in gaps or limited discussion in staff reports submitted to the Executive Board (though, on occasion, the lack of a substantive exchange of views with the authorities would be difficult to discern from reading the staff report and should have been flagged more clearly). Examples arose in a wide variety of circumstances and, for instance, in at least 5 of the 30 economies whose experience was reviewed in-depth (China, Korea, Mexico, Saudi Arabia, and the United Kingdom), there had been no meaningful two-way discussions on certain exchange rate issues for at least part of the period under review, or their treatment in staff reports was pro forma (lacking detail or much analytical content). 3 In some cases, IMF staff did not deal in a substantive way with possibly contentious issues (such as assessments of the appropriateness of a country’s exchange rate level, regime choice, or limits to accumulating international reserves). Staff interviewed by IEO explained that the reasons for not being more forthright on such issues included not only judgments on the relative importance of issues, but also the desire to preserve the IMF’s relationship with the country in question, and insufficient support from management or the Executive Board—an observation that is supported by the staff survey. In other cases, policy constraints and market or political sensitivities meant that the authorities were either hesitant or unwilling to discuss certain issues.

Figure 3.2.Survey of Authorities: Relative Emphasis Given by the Authorities and Staff, by Policy Issue

Authorities’ responses about the relative emphasis given to exchange rate issues in internal discussions compared to those with IMF staff

(In percent)

21. That said, in several countries, there was much more to IMF advice than met the eye in staff reports. In those cases, the exchange rate discussions were much more intense than suggested by Article IV staff reports. For example, detailed discussions on regime choice took place, with little or no documentation in staff reports or related selected issues papers. Staff activities in these cases ranged from informal workshops to confidential staff notes and meetings, extending over several years in some cases, with the authorities and staff exploring a variety of alternative policy options in the process. The staff received praise for this work, but it could only have been accomplished on the understanding that it not be divulged to the Executive Board. While it is reassuring that this work was carried out in some countries and was highly appreciated, the lack of reporting to the Executive Board of substantive issues in the context of Article IV consultations, which is not a new issue, does raise issues of accountability as well as the appropriate bounds of confidentiality.

22. Although exchange rate issues cannot sensibly be considered in isolation from domestic policy settings, evidence was mixed on how well the discussion of exchange rate issues was integrated with that of other relevant policy areas. In the surveys, both the authorities and staff agreed that coverage of linkages in discussions was good overall. However, a sizable minority of the authorities’ responses (35 percent) suggested room for improvement, an assessment in line with other sources of evidence. In the desk reviews:

  • Integration with monetary and fiscal policies was found to be good, with structural policies also well integrated for most countries. Discussions in staff reports were characterized by a focus on the consistency of these policies with the exchange rate regime and the external environment.4
  • Coverage and integration of financial sector and financial stability issues has improved over time—and significantly so in the context of countries’ Financial Sector Assessment Program (FSAP) exercises and related technical assistance mis-sions.5 The indepth review of 30 economies found that FSAPs benefited the integration of financial sector and stability issues into staff analysis. At the same time, about 40 percent of the surveyed staff saw analysis of financial stability issues as well as better analytical tools (e.g., balance sheet analysis) as areas where improvements could still be made.6
  • Global and regional spillovers were an area that, despite recent improvements, remained infrequently covered. Guidance from the 1997 BSR called for staff to incorporate spillover effects by focusing “on the international as well as the domestic implications” of the policies of regionally or systemically important member countries.7 Progress was made particularly in the treatment of international ramifications of U.S. policies, reflected in numerous staff papers using a variety of multicountry simulation models. For other country cases, however, progress was much less pronounced—an issue that is taken up in more detail in the section “Multilateral and Regional Perspectives.”

Regime Identification

23. A review of the most recent staff reports across the IMF membership suggested that classifications by staff of de facto exchange rate regimes were not always obvious or unambiguous; moreover, tensions between de jure and de facto classifications have remained unresolved throughout the entire period covered by this evaluation. In 12 cases (6 percent of the IMF membership), there appeared to be inconsistencies between MCM’s classification of the de facto regime and the descriptions provided in either the body of the Article IV staff report or its annex on IMF relations.8 In at least 3 of the cases from the in-depth review of 30 economies, the staff’s classification of the de facto regime conveyed, at least temporarily, a misleading impression of the regime in place, reflecting in part the continuing tensions between authorities’ de jure classification and the de facto classification based on staff judgment, for which consensus did not exist.9

Intervention and Related Policies

24. Analysis of intervention and related policies has been lacking in various ways. Coverage of intervention policies in staff reports was found to be incomplete, in at least 5 cases (with floating or managed floating regimes) from the in-depth desk review of 30 economies (euro area, Japan, Korea, Norway, and Singapore) reflecting, to different degrees, a more general lack of attention. This included insufficient analysis of past intervention episodes (including their effectiveness) for otherwise floating exchange rate regimes; missing assessments of whether intervention activities had been in line with the authorities’ stated intentions; and incomplete analysis of the effects of changes (beyond the narrow definition of reserves) in the net foreign assets of government agencies or government-sponsored enterprises (see Box 3.1), whether for balance of payments or other purposes. Aspects of intervention policies that received almost no staff attention were intervention tactics, that is, the specifics of how intervention is implemented and the extent to which such practice is consistent with the stated intervention goals, and the “exit problems” involved in withdrawing from large-scale intervention activity.10 Staff in general did not consider the effects of intervention activities (including those conducted in the context of fixed exchange rate regimes) on the countries whose currencies were used for intervention—or on the currencies of third countries.11

Box 3.1.IMF Surveillance of Intervention Policies

Foreign exchange market intervention is an important topic for exchange rate surveillance, with the 1977 Surveillance Decision—in outlining “Principles for the Guidance of Members’ Exchange Rate Policies”—placing a strong emphasis on members’ activities in this area. Yet, the evaluation found that intervention policies are insufficiently covered in the IMF’s surveillance of members’ exchange rate policies. In practice, Article IV staff reports and internal documents rarely describe the nature of intervention activities in any detail and few of them analyze such issues as the effectiveness of such activities, optimal levels of reserves, or intervention tactics and implementa-tion. This is despite the existing academic literature on some of these issues, which could have provided guidance for staff in conducting such analysis.1 An exception is the analysis of the quasifiscal costs of countries’ intervention activities, which are a more or less standard feature of staff assessments in countries with pronounced foreign exchange interventions.

A key aspect of staff’s treatment of intervention policies is a narrow focus on the use and accumulation of international reserves, which tends to disregard economically very similar activities outside the traditional boundaries of intervention policies, such as those associated with government-controlled investment funds and their investment policies. Being fiscally induced, such activities differ from “traditional” sterilized or nonsterilized interventions. However, to the extent that these activities are targeted at—or are otherwise expected to affect—the real exchange rate, an assessment of that country’s intervention policy against its stated rationale should be complemented by taking explicitly into account the impact of those government-controlled funds on capital flows.2

This, in turn, will require relatively detailed information on countries’ public sector net foreign assets—data that are not currently available to staff for all countries.

In some cases, the staff may have to judge whether a particular intervention policy is appropriate or not. While the 1977 Surveillance Decision suggests certain developments that “might indicate the need for discussion with a member,” 3 the evaluation finds that these “pointers” seldom guide the staff’s internal assessments of intervention policies, while being largely absent from any material submitted for discussion by the Board. But quite apart from any guidelines that would help define the legal issues, what is lacking is practical guidance on what would and would not constitute sensible and appropriate activity, in different circumstances and with different purposes, that could form the basis of a cooperative discussion.

The staff, in coming to an informed position, should have an estimate of the equilibrium real exchange rate in order to assess if intervention broadly defined (i.e., practiced through reserve movements, fiscal, or other means and motivated for balance of payments, fiscal, or other reasons) is keeping the exchange rate low or high, and forming policy advice on the basis of that assess-ment. Although there is no universally accepted methodology for assessing the exchange rate level, the staff could choose the concept of equilibrium exchange rate that, in their judgment, best suits the task at hand,4 and then supplement this analysis with an assessment of other quantitative and qualitative indicators. However, equilibrium exchange rate models are infrequently used in staff analysis in any case, and not typically in connection with discussion of intervention policies (see the section “Analysis of the Level of Exchange Rates”).

1 See, for example, Boyer (1978) on “optimal intervention,” Williamson (1973) and Ben-Bassat and Gottlieb (1992) on “opti-mal reserves,” and Edison (1993) on the “effectiveness of inter-ventions.” Recent examples of IMF work in the area are IMF (2004b) and Ishii and others (2006).2 The same applies to official borrowing or lending and capital controls, as highlighted in the Surveillance Decision. The use of this broader concept is not intended to question the traditional distinctions based on the motivation of policies. For example, in this context, the Executive Board noted that “[m]onetary or interest rate policies adopted for demand management purposes or other policies adopted for purposes other than balance of payments purposes would not be regarded as action to influence the exchange rate.” See IMF (1974) and Crockett and Goldstein (1987).3 See “Principles of IMF Surveillance over Exchange Rate Policies,” paragraph 2.4 See Background Paper 3 for more information on these models and their key assumptions.

25. Staff have generally supported a country’s accumulation of reserves for precautionary purposes, but not for purposes of maintaining competitiveness (see Box 3.2). About half of the sample of 30 economies covered in the in-depth desk review accumulated significant reserves in the evaluation period, especially in more recent years. Their motives included: (1) self-insurance against disorderly market conditions and volatility; (2) intergenerational and Dutch disease considerations (in countries with large natural resources or aid flows); and (3) concerns about competitiveness and export/industrial performance. IMF staff have generally endorsed the accumulation of reserves on precautionary grounds and in countries with large natural resource endowments. But they have advised against accumulating reserves aimed at containing the appreciation of the exchange rate in the event of strained competitiveness (including in low-income countries, when international reserves had reached a more prudent level). Because explicit analysis of an adequate level of precautionary reserves (linked to the exchange rate regime, nature of shocks, and country conditions; see Table 3.1) is often absent, assessments of the appropriateness of such policy measures have remained highly judgmental.12

Box 3.2.Views on Reserves Accumulation, 2000–05: Selected Cases

Many in the IEO’s sample of 30 economies accumulated significant reserves during the evaluation period. This box summarizes the experiences of seven such countries and the corresponding views of IMF staff and the Executive Board. In general, the IMF supported the accumulation of reserves for precautionary motives and in resource-rich cases, but did not on competitiveness grounds.

Continued support of reserves accumulation

Brazil. The authorities explained the accumulation of reserves, pursued in the context of a flexible exchange rate regime, in terms of reducing external vulnerabilities. Staff generally supported the authorities’ argument, but cautioned against the perception that they were defending a particular exchange rate level. The Board supported the vulnerability-reducing motive of reserves accumulation.

Norway. Norway maintains an inflation targeting framework and its central bank does not intervene in foreign exchange markets. Its natural resource wealth is managed through a petroleum fund that was set up to mitigate Dutch disease effects and for intergenerational equity, with oil and gas resources expected to be depleted over the medium term. Staff and the Board consistently supported the use of the government fund to sterilize the macroeconomic impact of oil revenues and called for a consistent rules-based fiscal policy to minimize appreciation pressure.

Shift to limiting reserves accumulation

Korea. Initially, the rationale for intervention was to rebuild reserves after the currency crisis and to moderate appreciation, but from about 2001 it also began to include the need to limit volatility. As early as 2000, IMF staff saw Korea’s level of reserves as adequate and argued that intervention should only be undertaken in disorderly market conditions. From 2003, staff increasingly argued for greater exchange rate flexibility, with broad Executive Board support.

Peru. The authorities’ rationale for reserves accumulation shifted from vulnerability concerns associated with economic shocks and high dollarization to competitiveness and price stability concerns. IMF staff were supportive of higher reserves until about 2004, when they began to argue for limiting the reserves buildup and for allowing greater flexibility of the exchange rate (because of inconsistency with the stated exchange regime, sterilization costs, and their view that competitiveness was not a concern). The Executive Board broadly supported the precautionary buildup of reserves and argued also for greater exchange flexibility, with more divided views on the latter in recent years.

Russia. Staff initially supported the authorities’ aim to limit the pace of real appreciation, arguing that a fairly stable exchange rate was a reasonable compromise between nominal appreciation pressures in the face of capital inflows, concerns that too rapid an appreciation would jeopardize output recovery, and uncertainty surrounding the recovery in money demand. This view included the assessment that continued intervention would seem appropriate. Among staff, however, doubts were expressed about the consistency of such advice with the objective of reducing inflation. Eventually, changing views on the persistence of strong terms of trade gains led to repeated advice in 2002–03 to scale back intervention and avoid further delays in allowing increased exchange rate flexibility. While the Board’s views developed broadly along with those of the staff, some Directors remained supportive of the authorities’ preference for targeting both inflation and the exchange rate in the face of real appreciation pressures.

South Africa. The authorities initially built up reserves in order to unwind the net open foreign position from past interventions. Once the net open foreign position was closed, staff supported the authorities’ early stance to increase reserves, especially against the background of gradual capital account liberalization; by 2005, however, staff began to argue that reserves were adequate. The Board broadly supported the evolving views of the staff.

Tanzania. The authorities’ stance on reserves accumulation stemmed from competitiveness concerns linked to external resource flows. They continued to build reserves by using only a portion of aid receipts to limit the monetary impact of increased government spending. From 2002, the staff did not see a problem with the level of the exchange rate and called for a greater absorption of foreign assistance. The Board, while supporting the buildup of reserves early on to create a buffer, endorsed the staff position and suggested structural reforms to ease competitiveness concerns.

Table 3.1.Coverage of Optimal Reserves Levels in Article IV Issues Papers, 2001–051
EconomyIssues PaperYearName of AnalysisConcenrned with High/Low Reserves
AngolaInternational Reserve Adequacy in Angola2003ExplicitLow
BulgariaExternal Sustainability and Vulnerability2004ImplicitLow
Central African Economic and Moetary CommunityReserve Adequacy in a Currency Union2005ExplicitHigh/Low
ChileChile’s Holdings of Foreign Reserves2004ExplicitHigh
HaitiReserve Adequacy in Haiti2005ExplicitLow
KazakhstanAn Assessment of External Vulnerability2001ImplicitLow
KoreaForeign Exchange Crises, Money Demand, and International Reserves2001ImplicitLow
LibyaOil Fund for Saving and Stabilization—Reform Options for Libya2003ImplicitHigh
LithuaniaCurrent Account Sustainability2005ImplicitLow
MalaysiaMalaysia’s Resilience to Unanticipated Shocks: Initial Results2002ImplicitLow
MauritaniaManaging Oil Wealth2005ImplicitHigh
MexicoReserve Adequacy in Mexico2003ExplicitHigh
NamibiaInternational Reserves and Investment Decisions by Institutional Investors2004ImplicitLow
NorwayThe Norwegian Government Petroleum Fund and the Dutch Disease2005ImplicitHigh
Slovak RepublicSlovakia’s Current Account Deficit: Why So Large and Is It Sustainable?2002ImplicitLow
South AfricaThe Case for Building International Reserves2004ExplicitHigh/Low
TunisiaAssessing Reserves Adequacy2004ExplicitHigh/Low
UkraineExternal Risks and Opportunities2005ImplicitLow
West African Economic and Monetary UnionThe Adequacy, Sources and Costs of International Reserves in the WAEMU2005ExplicitHigh/Low

Based on a desk review of exchange-rate-related issues papers for the entire IMF membership.

Based on a desk review of exchange-rate-related issues papers for the entire IMF membership.

26. In general, IMF staff have been reluctant to consider the use of intervention, including in the context of money or inflation targets, beyond the building of precautionary reserves.13 During the evaluation period, many economies experienced large inflows of aid, private capital, or natural resource revenues, which have put upward pressure on their real exchange rates. Staff’s reluctance to support the idea of intervention in these cases appears to be based on the assumption that the path of real appreciation would be identical, whether induced by a nominal exchange rate appreciation, or by an intervention-spurred increase in money and prices. However, this assumption is open to challenge. With underdeveloped capital markets, or underemployed resources, there are plausible reasons why this assumption may not hold, especially in the short run. Authorities’ concerns about the potential harm to the export sector from excessive nominal exchange rate appreciation may be warranted and should therefore be discussed on their merits.14 Further analysis should be worthwhile to draw out the different implications for policy advice, depending on the source of the inflow—whether aid, private capital, or income from natural resource exports.

Exchange Rate and International Reserves, 1998–2005

Sources: IMF, World Economic Outlook and Information Notice System.

1Program definition of net international reserves.

2Gross international reserves.

Data Requirements

27. Serious data problems appeared to have hampered effective surveillance. Staff reported that data problems impaired their ability to conduct exchange rate analysis and provide advice in 37 percent of countries. In 23 percent of countries, authorities were unwilling to provide relevant data. In part, lack of reliable intervention data, as well as related information (e.g., on intervention tactics and the investment policies of government-controlled asset management vehicles), seems to have limited the staff’s ability to properly assess intervention activities. Data have also remained incomplete on international reserves and authorities’ intervention and reserves management activities (see Box 3.3). In some cases, the full scale of countries’ reserves holdings, and broader concepts of govern-ment-controlled net foreign assets, raised difficult issues. Also, several of the big reserves holders do not disclose the currency composition of their reserves—for lack of participation in both the Composition of Foreign Exchange Reserves (COFER) database and the Special Data Dissemination Standard (SDDS).15

Box 3.3.Data Issues in IMF Exchange Rate Surveillance

Data shortcomings seem to have impaired the surveillance of a significant proportion of IMF members in recent years. The IEO found that staff reported material problems with data availability and quality in 90 of 191 economies in the two most recent Article IV consultations through 2005. Likewise, of the 115 countries for which country-specific information was identified by the IEO survey of IMF staff, 42 appeared to have had problems with availability or quality of data that—in the staff’s view—had impaired their ability to conduct exchange rate analysis.1 More than 40 percent of staff surveyed by the IEO also identified the availability of data as an area where significant improvement could be made that would raise the overall quality of exchange-rate-related analysis.

The causes of data shortcomings differ across countries. In some cases, such as those that have undergone transition or civil unrest, authorities themselves have not had the data. In other cases, authorities collect but seem to be unwilling to share important pieces of relevant information, such as records of intervention activity, material components of foreign exchange reserves, or unconventional intervention measures that are likely to affect exchange rates. Such data are not always essential for high-quality surveillance, but the IEO found that in about a quarter of cases IMF staff appeared to be conducting discussions from a position of informational disadvantage.2 Under such circumstances, it is difficult to see how staff advice could be effective.

Data problems do not always relate to reserves. In the case of Greece, for example, knowledge of the extent of shortcomings in fiscal data, which were not apparent to the IMF at the time, would have affected surveillance discussions in the run up to the country’s adoption of the euro in 2001. Staff responses to the IEO survey suggest that the authorities were unwilling to share critical information in several of those 30 economies that the IEO had selected for in-depth study. The desk review came across one case where underreporting of transactions had significantly affected that country’s international reserves and was not fully apparent from staff reports submitted for discussion at the Executive Board. In two other cases, reserves-related data issues were reported to the Board. In all three cases, the respective problems have subsequently been addressed.

1 Specifically, IMF staff working on 42 countries agreed or strongly agreed with the statement that “for [country name, as specified], the availability and quality of data has impaired staff’s ability (italics in original) to conduct exchange rate analysis and provide related advice.”2 In 26 out of 115 country cases, IMF staff agreed or strongly agreed with the statement that the authorities were “unwilling (italics in original) to share some critical data/information needed for exchange rate analysis and related advice.” In a similar number of cases, staff judged the authorities as technically not capable of furnishing critical data.

28. At the same time, staff appear to have been hesitant to pursue such data issues more forcefully. Because by definition official intervention always involves a coun-terparty, often a correspondent bank handling the actual trades, some information is bound to exist in the market that can be—and, on occasion, has been—sought out by staff. More generally, however, although staff are required to certify that data are adequate for effective surveillance, the evidence from the staff survey raises questions as to why the certification is granted so often. For staff to take a stronger stand when authorities are unwilling to share the critical information needed for surveillance, however, requires support by senior management and the Board, which—according to the staff survey and interviews—was perceived as lacking.

Analysis of the Level of Exchange Rates

29. While efforts have been made to enhance the analytical basis of staff assessments of exchange rate levels, “forthright assessments” have not been provided in all cases.16 While some description of exchange rate levels is contained in virtually every Article IV report, the in-depth review of IMF documents for the sample of 30 economies finds 5 cases with little or no analysis of exchange rate levels over part of the 1999–2005 period. China and Saudi Arabia are very different examples of countries for which the IEO found that a “forthright” assessment had not been made—in part because of analytical difficulties, but seemingly also because staff did not discuss with authorities and report what were potentially contentious levels-related issues.17

30. The use of sophisticated methodologies in the IMF’s analysis of exchange rate levels has increased, but is still limited and documentation could have been significantly better. For example, in 2005 there were only 25 cases for which one or more such techniques were used (see Table 3.2).18 In general, staff could have explained better how they reached their assessments of levels. At times, the choice of methodology appeared arbitrary, casting doubts on the results and their usefulness.19 In selecting methodologies, more attention should have been given to the particular strengths and weaknesses of individual approaches, and to how these relate to the circumstances of the economy in question and existing measures of competitiveness. So, while the staff’s work on exchange rate levels may have become more sophisticated, its impact on the quality of the resulting assessments and advice is difficult to establish. Also, given the large “error margins” inherent in all methodologies of equilibrium exchange rate determination, staff have generally been hesitant to attach much emphasis to model-based exchange rate assessments. The IMF’s Consultative Group on Exchange Rate Issues (CGER) approach, for example, is met by staff with a degree of skepticism, with only 40 percent of the surveyed staff who had worked on relevant countries finding it very useful.20 However, in its review of country cases, the IEO saw scope for more developing countries to have benefited from greater quantitative analysis of exchange rate levels and competitiveness. A prerequisite would be to have more and better data, including on costs, so that the link between assessments of competitiveness and of exchange rate levels, in periods of structural change, can be improved.

Table 3.2.Staff Use of Analytical Methods for Exchange Rate Level Assessments, 2000–051
Instances of Staff Use
YearNumber of CountriesPPP/adjusted PPPMacroeconomic balance/CGERFEER/BEEROther

See Background Document 3 for a description of the various methodologies and more detail on staff use.

See Background Document 3 for a description of the various methodologies and more detail on staff use.

Advice on Exchange Rate Regimes

31. Assessments of countries’ exchange rate regimes are a standard feature of Article IV reports, usually taking the form of a statement noting that the regime in place has served the country well. When advice was given over the evaluation period, it tended to be in the direction of more flexible exchange rates. In particular, based on the last two staff reports through 2005, IMF staff were found to have advised countries to adjust their exchange rate regimes (including monetary frame-works) in 63 cases. In 51 of these, they advised in favor of more exchange rate flexibility, which was linked to a proposed switch to inflation targeting in 8 cases. For the 30 economies reviewed over the 1999–2005 period, explicit regime advice was given in 12 cases, mostly in the direction of enhanced flexibility.21 Although such advice may not be unreasonable, particularly in a medium- or long-term context, greater flexibility may not always be desirable, and a particular view should not be taken for granted. What is striking is:

  • the frequent lack of formal, country-specific analysis backing such advice, which is likely to have limited the Executive Board’s ability to judge the merit of staff’s advice on a case-by-case basis; and
  • the lack of a Board-endorsed view since 1999 that this is indeed the right strategic direction for the IMF to be taking.22 The lively debates over exchange rate advice have taken place in other fora, in informal discussion, and in individual country cases.

32. Over the evaluation period, advice in favor of flexibility was not always backed up by formal analysis.23 While it is difficult in practice to separate cleanly the logic and timing of advice on regimes from advice pertaining to misalignments, recent IMF advice on exchange rate policy has mostly been couched in terms of calls for greater exchange rate flexibility. Formal analyses of exchange rate levels were used for only 25 of the 63 economies to which recent regime advice was given, and regime suitability was analyzed in only 10 cases (Table 3.3). This pattern is consistent with the observation, from the IEO’s sample of 30 economies, that analysis of regime choice was often of a largely conceptual nature, appeared to be influenced by concerns about exchange rate levels, and tended to lag—rather than lead—the IMF’s general direction of advice. 24 In addition, in a few of the sample economies, IMF staff and management also pressed the authorities to move quickly, usually against their preference for a more gradual approach, and failed to fully appreciate country-specific factors, especially—but not always—in a program context.25 This finding was supported by survey evidence, particularly among the large emerging market economies and in interviews. The IMF was acting against the background of the lack of clearcut guidance from the academic literature on regime choice, which has tended to discuss regime decisions in the context of a limited number of economic characteristics, but without developing operational tools to aid practical choice.26 Whatever the reasons, the IMF may have overemphasized the benefits of a rapid move to more flexible exchange rate regimes, while insufficiently appreciating country-specific obstacles to implementation and other reasons why country authorities may prefer to remain—or remain longer—with a more managed exchange rate system.

Table 3.3.Exchange Rate Regime Advice and Its Analytical Basis1
Nature of Staff AdviceNumber of Cases2Regime Sustainability or Suitability Analyzed3Formal Exchange Rate Level Analysis4Currency Deemed Over- (Under-) Valued by Staff
More flexibility5172011 (10)
Less flexibility1100 (0)
Implementation; management of existing regime244112 (6)
Number of economies with advice563102511 (11)
No specific advice given12814388 (4)
Total number of economies191246319 (15)

As recorded in the IEO desk review of the two most recent staff reports through 2005.

Advice has been given to 63 economies overall (in 13 cases, advice on flexibility and management of the regime was given simultaneously), of which 2 were advanced economies, 10 were large emerging market economies, and 51 were other emerging market and developing economies.

Based on tools such as optimum currency area criteria and analysis of economic shocks.

Analysis of exchange rate level explicitly involved tools other than interpretation of REER charts.

Net of double counting.

As recorded in the IEO desk review of the two most recent staff reports through 2005.

Advice has been given to 63 economies overall (in 13 cases, advice on flexibility and management of the regime was given simultaneously), of which 2 were advanced economies, 10 were large emerging market economies, and 51 were other emerging market and developing economies.

Based on tools such as optimum currency area criteria and analysis of economic shocks.

Analysis of exchange rate level explicitly involved tools other than interpretation of REER charts.

Net of double counting.

33. A key problem was the lack of appreciation on the part of staff and management for the challenges posed by implementation. In the words of one senior official—expressing a sentiment shared by others—“the more complex or country-specific the [implementation] issue, the less useful the IMF’s advice.’’ Some 40 percent of the authorities’ survey responses indicated that attention to implementation issues could be improved. This view was particularly pronounced within the group of large emerging market economies and among those authorities that had received advice on their exchange rate or monetary policy regimes during the last two consultation cycles. (Among the latter, 60 percent would have liked to see broad advice being developed into concrete advice on issues of implementation.)27 Interviews suggested that technical obstacles to the implementation of more flexible exchange rates tended to be underestimated by staff, who were perceived by the authorities as having insufficient technical expertise or practical experi-ence. Technical assistance (TA), to the extent it was provided, was in general valued by both staff and the authorities as being an important component of IMF advice.28

34. In individual cases, demands for urgent action and insufficient attention to detail led to unintended outcomes that reflected badly on the advice given. In the case of Egypt, in January 2003, both staff and the authorities had not given sufficient attention to ascertain that the microeconomic preconditions for successful floating (e.g., a functioning foreign exchange interbank market and a monetary policy framework) had been put in place. The IMF had provided technical assistance in advance, but there was no careful checking of whether the recommendations had been implemented. In the event, the flotation attempt was aborted in the wake of unfavorable market dynamics. These were triggered by pentup demand for foreign exchange and open positions on bank balance sheets (which had been identified in the earlier technical assistance, in broad terms, as potential sources of risks), as well as a lack of sustained support from other policies. In the case of Ukraine, in 2005, an undue sense of urgency conveyed by the IMF’s policy advice complicated communications with the authorities to an extent that policy actions were taken without proper preparation.29 Reasons for staff to advocate a “rush to more flexibility,” either at the level of implementation (as in Egypt) or more generally (as in Ukraine), could have included substantive concerns, for example, that the economy was open to even greater risks of substantial shock if adjustments were delayed and that no other policy option were available to cope with these risks. However, in the cases mentioned above, such risks did not appear great; or at least, evidence was not presented to support going ahead without having the “first best” groundwork in place. It appears, therefore, that, in these cases, one of the abiding lessons from the Asian crises was being applied too readily.

Multilateral and Regional Perspectives

35. Despite increased attention to global imbalances and capital flows in recent years, multilateral considerations did not feature prominently in most bilateral discussions. Multilateral aspects were explicitly referred to in one-sixth of the recent staff reports covered by the desk review of the full IMF membership, while regional aspects were referred to in about a third (Table 3.4). 30 In the desk reviews, however, depth has often been found lacking, with references to multilateral developments not fully integrated into the staff report or supported by formal analysis. It is also the case that multilateral considerations have been heavily dominated by concerns about global current account imbalances, while regional considerations tended to focus on economic conditions in major regional trading partners. Scant attention, however, was being given to other issues, such as financial market and balance sheet spillovers, as reflected by the relatively limited analysis of exchange-rate-related issues in the IMF’s multilateral surveillance that have been relevant for bilateral surveillance (see Table 3.5),31 and the failure in bilateral surveillance to pick up financial market issues that may have affected several countries at once.

Table 3.4.Coverage of Multilateral and Regional Issues, by Country Group1
IssueMajor AdvancedOther AdvancedLarge EmergingOther Emerging/Developing
Resolution of global imbalances7730
Global capital markets2165
Spillovers from trading partners/competitors12622
Other regional spillovers12416
Memorandum item:
Number of economies in group82120142

Based on the last two Article IV reports through 2005.

Based on the last two Article IV reports through 2005.

Table 3.5.Selected Coverage of Exchange Rate Issues in the World Economic Outlook, 2000–05
How will global imbalances adjust?9/2005Appendix
Learning to float: experience of emerging market countries in the early 1990s9/2004Chapter
How did Chile, India, and Brazil learn to float?9/2004Box
Foreign exchange market development and intervention9/2004Box
The effects of a falling dollar4/2004Box
How concerned should developing countries be about G-3 exchange rate volatility?9/2003Chapter
Reserves and short-term debt9/2003Box
Are foreign exchange reserves in Asia too high?9/2003Chapter
How have external deficits adjusted in the past?9/2002Box
Market expectations of exchange rate movements9/2002Box
Weakness in Japan, global imbalances, and the outlook3/2002Appendix
How did September 11 affect exchange rate expectations?12/2001Box
What is driving the weakness of the euro and the strength of the dollar?5/2001Chapter
The weakness of the Australian and New Zealand currencies5/2001Box
Convergence and real exchange rate appreciation in EU accession countries10/2000Box
Why is the euro so undervalued?10/2000Box
The pros and cons of dollarization5/2000Box
Source: IMF, World Economic Outlook.

36. In particular, coverage often lacked depth even in cases for which regional or multilateral issues would be expected to be important. This has included limited coverage of potential regional spillovers emanating from such economies as the euro area, Brazil, or Russia.32 Moreover, bilateral and multilateral surveillance have focused unevenly on common factors underpinning exchange rate developments and associated policy responses. For example, while attention was given to self-insurance motives and the accumulation of reserves across a large number of countries from the late 1990s, discussions of the abundant global liquidity in later years (and the possibility of its being temporary in nature) were not translated—in bilateral or multilateral surveillance—into analysis of common patterns across countries of those intervention policies that seek to contain exchange rate appreciation pressures to help preserve competitiveness. Consistent with this, some 45 percent of the respondents to the IEO’s survey of country authorities found that IMF staff had rarely identified, or sufficiently integrated into their analysis, spillovers affecting their countries. The authorities were roughly split on whether or not spillovers emanating from their countries had been sufficiently integrated. In both cases, dissatisfaction was felt particularly among the large emerging market economies (Figure 3.3). 33

Figure 3.3.Survey of Authorities: Views on the Integration of Spillovers

(In percent)

An example in this context is the IMF’s treatment of global imbalances—the key multilateral surveillance issue of the last few years (Box 3.4).

Box 3.4.The Treatment of Global Imbalances in Bilateral Consultations, 2003–05 1

Having emerged as a major macroeconomic policy issue at some point in 2002, global imbalances were given prominence in late 2003, with the near-simultaneous pub-lication of the G-7 communiqué in Dubai and an analysis of reserves accumulation in the IMF’s World Economic Outlook report. Concerns at the time were—and continue to be—driven mainly by the risk of disorderly adjustment and related cross-country spillovers should exchange rate and macroeconomic policies in major surplus and deficit economies fail to facilitate a timely correction of these imbalances.

The IMF was among the early proponents of the “disorderly adjustment” view, with references to global imbalances starting to appear consistently in internal IMF documents and staff reports for a number of Asian countries in 2003 and, beginning in 2005, for other major surplus economies. In 2002, large-scale reserves accumulation had been identified as an issue with potential multilateral implications in internal IMF discussions on China. Earlier, starting in 1999, references to global imbalances and disorderly dollar adjustment had appeared in Article IV reports for the euro area and the United States, though in the context of imbalances between the G-3 economies.

Despite the importance attached to the issue by both IMF staff and management,2 and the inherent need for coordinated—possibly, IMF-facilitated—policy responses, the integration of multilateral considerations into analysis at the individual country level generally lacked depth. In particular, with policy advice being formulated largely on the basis of cyclical, country-level considerations that would provide “first best” recommendations for the country in isolation, interdependencies between individual country policies and the responses of other countries were not sufficiently integrated into staff analysis. Accordingly, staff recommendations—while consistent with global adjustment over the medium to long term—became increasingly inconsistent with staff’s own assessments of the need for and relative urgency of policy responses at the multilateral level.3

As a result, scope for active policy coordination—for example, by providing alternative sets of policy recommendations that are explicitly conditional on policy actions taken in other countries—was insufficiently exploited. Despite increasing coverage of the topic in bilateral Article IV consultations and repeated calls by the Executive Board for a “globally coordinated and calibrated policy response” the Board did not “force” a more systematic approach towards the resolution of global imbalances. Finally, following high-profile remarks by certain country officials in the course of 2005, the search for a coordinated policy response was further complicated as the IMF was increasingly seen as reacting to outside pres-sure; this—according to a number of country officials—lowered the credibility of its policy prescriptions.

1 See Background Document 5 for more detail.2 See, for example, Köhler (2003) and de Rato (2005).3 The IMF’s views on the size and urgency of any adjustments and on the corresponding risks for disorderly market conditions were not universally shared, even among IMF staff. Internal debate of competing views, however, has remained limited

The Consistency and Evenhandedness of Advice

37. The consistency—or evenhandedness—of IMF advice is another important aspect of quality: no clear-cut cases of uneven treatment were identified in the sample of 30 economies, but more could have been done to counter the perceptions of inconsistency, which remain strong. Consistency requires that advice be given across the membership in ways that adjust for different circumstances, while also allowing for the fact that a variety of measures can be used to deal with similar challenges.34 Claims of inconsistency are as difficult to dispel as they are to prove. However, providing better explanations for particular policy advice would reduce the risk of inconsistency, as well as the risk of being accused of it. But care also needs to be taken that similar types of assessments are delivered with similar degrees of analytical detail to preserve an evenhanded approach. Three types of potential inconsistencies or lack of evenhandedness were examined briefly, but they raise questions rather than provide answers.

38. One possible inconsistency arises from the lack of in-depth analysis of countries’ intervention policies, which could lead the IMF to treat reserves accumulation—or more broadly, public sector net foreign asset positions—unevenly across countries. For example, would the IMF’s view on reserves accumulation in China, or Peru, be different if long-term developments, such as aging or receipts from natural resource exports, were taken into account, with sizable parts of the country’s net foreign assets accumulated in a dedicated fund for the benefit of future generations, or with a state-owned holding company? How would these examples compare to the analyses of large, government-controlled net foreign asset positions in other countries; and how and on what basis would the IMF make judgments in such cases?

39. A second possible lack of evenhandedness can arise from an unwillingness to raise sensitive issues with advanced economies, while having less compunction in doing so with other countries. An example from the sample of 30 countries was staff and management’s reluctance to raise with the U.K. authorities the issue of exchange rate regime choice while the United Kingdom was deciding whether or not to adopt the euro.35 Given the significance of euro adoption during the evaluation period, the case of Greece is also of interest. The 1999 Article IV staff report for Greece noted that weaknesses in data “complicated the assessment of economic conditions,” but the true extent of these weaknesses and their implications were not unearthed until later years. This raises the question of whether more effective surveillance would have altered the assessment of policies in the period before the adoption of the euro in 2001. When the data problems were later revealed, no report was discussed at the Executive Board on the causes and implications. These examples have done nothing to dispel the notion that advanced countries are treated differently, though by themselves they cannot prove it either.36 It is particularly important, for both substantive and signaling reasons, that the exchange-rate-related advice to advanced economies (including those in the euro area) is evenhanded and perceived as such.

40. Finally, consistency checks can also be applied to advice given to economies in similar circumstances. One view expressed in Europe, for example, was that, in the context of the launch of the European Economic and Monetary Union, the IMF had shown even less readiness to involve itself in the cases of Italy or Germany than in the case of the United Kingdom. The IEO desk reviews contrasted the IMF’s advice to some African countries that have been subject to a surge in foreign exchange receipts from aid and commodity price increases.37 At the same time, judgments about inconsistencies are very difficult to make in that different approaches can be adopted to tackle similar issues. For example, empirical research 38 finds that both Hong Kong SAR and Singapore suffered similar shocks and real developments, at least at business cycle frequencies, but had different nominal outcomes in that inflation rates were higher and more variable in Hong Kong SAR than in Singapore. While inflation rates are thought to reflect differences in the monetary (and exchange rate) regimes in place, the evidence on the performance of the two economies gives no reason to believe that current regime choices—and related IMF advice, as the IMF has consistently endorsed existing regimes in both places—have been inappropriate. Another example might be the set of varied exchange arrangements for countries in Eastern Europe in the transition to eventual euro adoption. The best way to dispel notions of possible inconsistency is to explain closely the framework under which advice is given.


41. What reasons can be given for the evidence of shortcomings in quality with respect to coverage, analysis and advice, and multilateral work (described above)? In part, the existing incentive structure was not ideally geared to producing the best result. Incentives were to manage various established processes for completing consultations, quickly and with little risk, and these may or may not be consistent, either for staff, the Board, or authorities, with fulfilling exchange rate surveillance responsibilities in a best practice way. Adverse incentives ranged from concerns that taking much space (especially with strict limits on Article IV staff report length) to justify and discuss a well-established regime was unwarranted; to concerns not to ruffle feathers, and possibly markets, when there was a genuine issue.

  • Instead of being accountable for gaining traction in their discussions with member countries, by providing relevant and forthright assessments, staff saw the maintenance of a smooth relationship with the authorities as a powerful driver. Unless staff feel they will be fully backed up by senior staff and management, and the Board, when taking a respectful but firm stand as needed in discussions, it is not surprising that opportunities for good surveillance are sometimes missed.
  • High priority is put on establishing a policy line and then sticking to it. (Examples in the period under review are the trend toward advocating more flexible exchange regimes, and the policies to deal with the perceived problem of global imbalances.) While admirable to a point, this tendency went too far because the IMF did not at the same time also encourage and reward internal questioning and challenging of that line to ensure that it stays ahead of the curve.
  • Insufficient reward was given to integrating the best elements of analysis and expertise from both inside and outside the institution. The “silo” problem has been described in other contexts, but on exchange rate advice the contrast is notable between the access to the latest thinking and ongoing research efforts, and to practical experience and expertise, and the apparent difficulty in ensuring that these elements are appropriately and quickly integrated Fund-wide into frontline advice. Of course, when there is little academic consensus on many points, the problem of distilling and establishing operational guidance is more challenging, but management oversight and the right internal structure are therefore all the more critical.

See Chapter 2 and Background Documents 1 and 2 for more detail. As noted, the specific guidance to staff on how to go about exchange-rate-related surveillance is surprisingly limited—at least relative to the centrality of exchange rate policy issues to the IMF’s responsibilities.


The 2002 Operational Guidance note (IMF, 2002) specified that “all Article IV consultation discussions and reports should include…a forthright assessment of the exchange rate level.”


In the case of China (in 2001–02), substantive engagement with the authorities was lacking on the specifics of exchange rate regime options identified by IMF staff. In the case of Korea (in 2003–04), Article IV discussions did not satisfactorily cover the authorities’ intervention policy and its consistency with the announced exchange rate policy. In the case of Mexico (in 2002–04), the staff did not assess the exchange rate level despite the fact that competitiveness was a live issue; in contrast, staff in the same period pursued with the authorities other contentious exchange-rate-related issues. In Saudi Arabia (2003–05), pronounced shifts in the terms of trade did nottrigger assessments of exchange rate levels. In the United Kingdom (2000–03), there was no substantive discussion on the issue of euro adoption, including on the merits and implementation of the so-called “five tests,” until after the authorities had made their decision.


Among the 191 economies examined, there were only 25 cases in which the two most recent staff reports were judged not to have explicitly linked exchange rate discussions to these other policy areas. In all of these 25, the exchange rate was either not regarded as a live policy issue or exchange rate issues were treated in selected issues papers, with part of the discussion of linkages covered there. See Background Documents 4 and 5 for more detail.


This finding is consistent with the IEO’s recent report on the topic; see IEO (2006b). Given the importance of FSAPs for the integration of financial sector and stability issues, the improvement in quality may be partially driven by the fact that, among the 30 economies reviewed in-depth, 24 (80 percent) had completed an FSAP.


In the broader area of country vulnerabilities, the degree of integration of the IMF’s internal “vulnerabilities exercise” into staff’s surveillance activities was also found to be good overall.


The Crow Report and the IEO evaluation of multilateral surveillance reiterated the need. See Crow, Arriazu, and Thygesen (1999) and IEO (2006c).


See Background Document 4 for more detail.


Besides the de facto classification used by MFD/MCM, there are several different classification schemes proposed by researchers. Correlations across different schemes are virtually as low as the correlation for any one of the de facto classification schemes with the de jure classification. This casts doubt on the idea that there exists an unambiguous de facto classification (see Bénassy-Quéré, Coeuré, and Mignon, 2006; and Frankel, 2004).


Examples include the practice of covert interventions, an arrangement that is typically seen as limiting the effectiveness of intervention through the signaling channel.


During the Executive Board meetings in 2005, comments by Executive Directors on the lack of such assessments remained unanswered by staff and management; and in its desk reviews, the IEO identified only two possible examples of staff analysis of intervention-related spillovers. Implications of Asian intervention policies were analyzed in a 2004 selected issues paper for the euro area, which looked at different scenarios for global rebalancing on the basis of a three-country version of the IMF’s GEM DSGE model, arguing that asset-market-based adjustments could have adverse effects on the euro area if these were to rely largely on the euro. In addition, possible regional spillover effects from yen depreciation in the context of antideflationary policies had been analyzed on the basis of simulation models in 1999/2000. This analysis, however, was not updated in the context of the interventions in 2003–04.


More recent papers have given some emphasis to the upper band of reserves accumulation. See IMF (2004b), whereas earlier studies focused on precautionary motives (IMF, 2001).


See Ho and McCauley (2003) for an analysis of the use of intervention in the context of money or inflation targets.


Absence of such information on currency composition of reserves will complicate any analysis of reserves-management-related risks of disorderly exchange rate adjustment. This is one of the key multilateral risk factors highlighted in recent Article IV reports for the United States and a number of other major economies (see Box 3.4). COFER is a database maintained by the IMF’s Statistics Department that keeps end-of-period quarterly data on the currency composition of official foreign exchange reserves, defined as monetary authorities’ claims on nonresidents in the form of foreign banknotes, bank deposits, treasury bills, short- and long-term government secu-rities, and other claims usable in the event of balance of payments needs. COFER data are currently reported on a voluntary basis by 119 countries; individual country data are strictly confidential and disseminated, including within the IMF, only in aggregated form for three country groupings. The SDDS was established to guide IMF members in the provision of their economic and financial data to the public, including data on foreign exchange positions. Subscription is voluntary, but carries a commitment by a subscribing member to observe the standard and to provide certain information to the IMF about its practices in disseminating economic and financial data; to date, there have been 64 subscriptions to the standard.


While there is no clear-cut definition of what “forthright” assessments are supposed to entail, the IEO’s review made the judgment that the following reasons would constitute failure to make such assessment: (1) absence of any analysis in situations where external developments strongly suggest that equilibrium exchange rate levels may have changed; (2) failure to bring all relevant information to bear in coming to a conclusion.


In the case of Saudi Arabia, exchange rate levels were consistently not analyzed in Article IV reports, despite pronounced, possibly long-lasting, terms of trade changes and repeated calls—in the internal review process—for more analysis. In China, by contrast, exchange rate levels were analyzed using a variety of methodologies: however, some traditional indicators of exchange rate misalignment were not brought to bear on the issue until 2005, clouding the overall assessment of renminbi levels.


Important analytical contributions were made, for example, in the case of the United Kingdom (2001), the West African Economic and Monetary Union (WAEMU) (2004–05), as well as in a selected issues paper on China (2003, later published in the IMF Occasional Paper series).


See Background Documents 3 and 5 for detail.


Isard and others (2001) listed a number of caveats on the reliability of estimates.


In the survey of authorities, a majority of the respondents saw the IMF favoring particular regimes over others, with opinions roughly split on whether the IMF’s approach had paid sufficient attention to intermediate regimes. See Background Document 6.


See Background Document 2 for more information.


Some analysis may have been provided over earlier consultation cycles. A review of selected issues papers (2001–05) for the 63 countries that have received advice on their exchange rate regimes finds 17 cases (27 percent) for which no paper on exchange rate issues was available. Another 28 countries (44 percent) had only one such issues paper over the period. Most of these papers were conceptual in nature or focused on only a subset of the issues at hand.


This applies—to different degrees—to the cases of Malaysia, Morocco, and Ukraine.


Specifically, in the case of Ukraine in 2004, attempts were made to make enhanced exchange rate flexibility a prior action for the completion of a program review—later toned down to a “demonstrable shift” in exchange rate flexibility as an important element in completing the review. This was despite a lack of compelling analytical work in support of an urgent regime adjustment and despite disagreement by the authorities. A similar attempt at leveraging the program context was made in the case of Egypt in 2002, with regard to possible access to IMF resources under the Compensatory Financing Facility (CFF), a purchase that never materialized. While staff had made a case for urgent action, the authorities’ state of readiness and management’s use of pressure in the CFF context appeared questionable. The desire to use apparent windows of opportunity in less than perfect conditions has to be set against the risks to credibility if the strategy does not work.


The IMF, for its part, has been late to develop such approaches. However, though very different in terms of methodologies, recent work by Husain (2006)—first applied in the context of Morocco, and later used for countries such as the former Yugoslev Republic of Macedonia, Serbia, and Ukraine—and Schadler and others (2005)—on the adoption of the euro in Central Europe—can be considered valuable contributions.


Within the same group, staff’s follow-up activities were judged—according to 45 percent of the respondents—to have either no or only marginal influence on implementation, which contrasts with staff’s much more favorable assessments and suggests that provision of practical assistance was falling short in the view of the authorities.


In interviews, the IEO was alerted to a case in which the findings of an exchange-rate-related TA mission may have been turned into program conditionality with undue urgency. While a more detailed assessment was deemed beyond the scope of this evaluation, the IEO screened IMF documents for the last two Article IV cycles (up to 2005) for country cases with program conditionality on exchange rate issues, identifying 10 such cases. By tracing the history of these program conditionalities, however, the IEO’s opinion was that most of these had clearly not been driven by the results of preceding TA missions and/or did not have an unrealistic timetable.


The April 2005 decision for an ad hoc step realignment of the exchange rate, which was ill-received by unprepared markets, was taken in the immediate aftermath of a meeting between IMF management and a senior country official, at which the IMF had reemphasized the need for more exchange rate flexibility.


Mention of multilateral policy issues was limited relative to the overall size of the IMF membership, but corresponded closely with the weight of such factors as country size and systemic importance. The 31 staff reports found to contain some discussion of multilateral issues, such as global imbalances and international capital market spillovers, include those for the euro area, most individual G-7 countries, and a number of non-G-7 economies and international trading centers. Most of the remaining economies experienced sizable multi-year current account imbalances over the period 1999–2005.


This evidence backs up the findings in an earlier IEO study of the uneven integration of bilateral and multilateral surveillance (see IEO, 2006c). See also Background Document 6, paragraph 20 and paragraph 35 (Figure A6.25).


See Background Document 4. In this context, the newly instituted Regional Economic Outlooks through 2005 did not seem to have been used to provide such analysis (and in the case of Russia, would need to cross over the IMF’s departmental boundaries).


Across all respondents, 39 percent agreed that the quality of their interaction with staff would have benefited from a better integration of multilateral perspectives. While only 19 percent of staff agreed with this assessment of past discussions, about 40 percent of the same respondents said that the analytical framework for the discussion of multilateral issues was lacking and that the IMF’s multilateral surveillance tools had not provided them with relevant inputs. Looking forward, 42 percent of staff saw scope for improvement of IMF analysis of policy spillovers, while 47 percent said the same about integration of bilateral and multilateral surveillance.


Consistency is not to be confused with a “one size fits all” approach, which would give insufficient attention to country circumstances.


In this case, the U.K. authorities had not been keen for the IMF to give its view and—prior to 2003—staff and management obliged, thus missing the opportunity for any influence or discussion. Preliminary work by staff in the context of the 1999 Article IV consultation had set the stage for possible subsequent development of these issues, including on the appropriate exchange rate level at which to join the euro area. In 2003, staff pressed for, and management finally agreed to, a thorough analysis of how the “five tests” had been applied by the authorities, but only after they had announced that the economic case for adopting the euro had not been made.


Evidence from the staff survey is consistent with such a ten-dency. For example, a somewhat higher percentage of respondents working on advanced economies, than for the overall sample, agreed with the statement that the need to preserve close relationships with the authorities tended to dilute coverage in staff reports.


For Sub-Saharan African countries, such as Rwanda and Tanzania, that have their own currencies, staff have generally advised the authorities to let any real appreciation from aid and commodity exports receipts take place through nominal exchange rate changes. Liquidity expansions associated with increased government spending were to be sterilized through foreign exchange sales and any adverse effects on competitiveness were to be cushioned through structural reforms (although analysis did not consider the time lags involved). For countries in monetary unions, in contrast, the advice has been for a combination of strong fiscal positions, competitiveness-enhancing structural reforms, and cautious monetary policies that would keep a lid on domestic liquidity, and, by limiting the absorption of foreign inflows, could limit or delay adjustment in the real exchange rate. It is unclear whether in these cases the contrasting advice would have implications for the ability to absorb scaled-up aid or higher commodity receipts without adversely affecting the export sectors. One factor to bear in mind in such analysis is that the CFA franc, though in a fixed peg arrangement, has appreciated with the euro. For the WAEMU, the CFA franc had appreciated in real effective terms by about 20 percent from 2000 to 2005.

    Other Resources Citing This Publication