IEO Evaluation of Exchange Rate Policy

Chapter 5: Findings and Recommendations

International Monetary Fund. Independent Evaluation Office
Published Date:
August 2007
  • ShareShare
Show Summary Details

Main Finding

54. In the period reviewed (1999–2005), the IMF was simply not as effective as it needs to be to fulfill its responsibilities for exchange rate surveillance. This judgment is not meant to detract from the dedicated and hard efforts of staff, nor to fail to recognize the inherent complexity and lack of professional consensus on many of the issues discussed in this report. However, the effectiveness of IMF surveillance in fostering international cooperation depends, ultimately, on the IMF’s adeptness in focusing on the key analytical issues of the day (which have shifted radically over time) and in engaging in effective dialogue with its members, individually and collectively. While by no means evident in all countries, this evaluation observed serious shortcomings in both respects that resulted in an “effectiveness gap” in the IMF’s main line of business. The reduced traction with advanced economies is in danger of being extended to large emerging market economies, and beyond. Such an evolution is corrosive, breeds cynicism among the staff as well as the members, and builds on perceptions of a lack of evenhandedness. Unless the shortcomings are successfully addressed in the period ahead, and as the number of countries looking else-where for policy advice and support continues to grow, there could be serious implications for the ability of the IMF to discharge its responsibilities in the future.

Rules of the Game and Guidance to Staff


55. The rules of the game for exchange rate surveillance are unclear, both for the IMF and member countries. The confusion may reflect to some degree the complex nature of the consensus reached in the 1977 Decision, and the failure subsequently to translate and adapt that understanding into more specific guidance on key points. Yet, the heart of the matter lies with the failure of the IMF to have the appropriate degree of engagement with all of its members. For the staff to do a better job fulfilling its responsibilities, it needs to be both more responsive to members’ concerns and more forthright, and it requires the more active support of management and the Executive Board.

56. Operational guidance for staff is insufficiently clear (or, in some cases, absent). For example, the requirements to assess exchange rate regimes and levels are not very specific. The evaluation identified two key priorities:

  • (1) The stability of the system. The IMF is charged with responsibility for oversight of the international monetary system, but the last Executive Board review of this topic was in 1999. No recent Board review has therefore assessed whether the stability of the international monetary system is best preserved by the choices of exchange regimes (and exchange rate levels) now made by the membership. Consequently, there is no updated framework that would guide policy advice in individual country contexts. An updated review could have considered, for example, the extent to which large reserve accumulations, among a host of very large shifts in public and private asset positions, affect the workings and stability of the system.

  • (2) The use and limits of intervention in intermediate regimes.

    • Use of intervention. Emerging market and developing countries have been wrestling with multiple challenges, including how to maintain monetary—or inflation—control in circumstances of large inflows of capital (or aid and natural resource revenues). Allowing a nominal appreciation may facilitate monetary control, but could adversely affect export performance and growth. Insufficient attention has been paid to this trade-off, for example by investigating over what time period intervention—together with other policies, including fiscal measures or changes in capital controls—might modify the assumption that increases in the real exchange rate cannot be resisted.

    • Limits to reserves accumulation (or, more generally, to the accumulation of public sector net foreign assets). Immediately after the financial crises of the 1990s, guidance was appropriately developed on rebuilding reserves to sensible minima, but insufficient Executive Board guidance was developed on what might constitute reasonable upper bounds, and why.

57. Management assigned insufficient focus and attention on conducting effective dialogue with authorities. While staff’s discussions with the authorities were generally seen as two-way and were found useful in most cases, a clear message also emerged that authorities in many countries were seeking greater value added.

  • (1) The effectiveness of the dialogue was hampered in some cases because staff teams did not bring with them sufficient expertise and experience. Financial market and foreign exchange market expertise needed to be complemented by cross-country experience, attuned to country-specific circumstances. Moreover, management did not make sufficiently clear that, in all cases, staff’s general advice (e.g., on regime change) should be based on their judgments on the readiness of appropriate implementation capacity, with technical advice to be provided on such aspects, as necessary.

  • (2) The IMF has not always been well positioned to deliver messages that would add value to the appropriate decision makers. When exchange rate policy is a live issue, it often requires the attention of ministers and government leaders. To be effective in providing advice, the IMF needs to be expert at communicating messages at this highest political level, as well as at the more technical level at which discussions normally take place. Communicating at this level requires skill and the involvement of senior management, a good understanding of decision-making processes, and of where messages need to be given to have impact. Advice has to be presented both orally and in written form in ways that are persuasive to hard-pressed ministers (which means being very brief and very clear).


58. Clarify the rules of the game for the IMF and its member countries. As discussions proceed on surveillance policy initiatives, a revalidation of the fundamental purpose of surveillance would be an important goal. Central to this is the requirement on countries, and the IMF, to consider the consequence for others of an individual economy’s policies, including exchange rate policies and other measures that affect exchange rates. Since relevance and effectiveness cannot be legislated, however, the key lies in ensuring the trust and willingness of countries to cooperate within whatever legal framework is in place.

59. Practical policy guidance should be developed on key analytical issues. This would be based on the latest research and cross-country experience and would help to ensure an evenhanded approach across the membership. Two priorities would be:

  • (1) On the stability of the system. An Executive Board policy review of the stability of the system of exchange regimes and exchange rates should be conducted periodically, taking into account the array of chosen regimes, global liquidity conditions, and other issues. The conclusions would provide an updated framework for guidance in individual country cases.

  • (2) On the use and limits of intervention. As input to developing guidance to staff, given the many circumstances of countries and the different roles assigned to the exchange rate, authorities could be asked during Article IV consultations to describe the range of reserves holdings/public net foreign positions they expect to hold over the period ahead, and the reasons for establishing such a range. Discussion could then take place both on the range presented and on the arguments to justify it, which would provide a benchmark for subsequent discussions. Guidance would reflect various considerations, including precautionary motives for reserves, intertemporal savings of natural resource incomes, and potential problems for monetary management and competitiveness, as well as the implications for adjustment in the world economy.

60. Management should give much greater attention to ensuring effective dialogue with authorities. This task should be assigned as much weight as developing the right advice.

  • (1) Management should develop a strategic approach to identify opportunities to improve the effectiveness of the dialogue, involving senior management and with support, when necessary, from Executive Directors. This would also involve ensuring the staff team has the right kind of expertise; planning whom to engage in discussions and when; calibrating the format of the message to particular needs. In the performance appraisal process, the success in ensuring effective dialogue would be defined and rewarded.

  • (2) Management and the Executive Board need to adjust the incentives to raise controversial issues. They need to send staff a clear signal that they will be supported when they take time to understand the authorities’ views, when they have difficult messages to deliver, both to the authorities and back to the Board, and when there are difficulties with the provision of information by the authorities.

Problems in Implementing Existing Policy Guidance


61. Clear descriptions of exchange rate regimes have remained elusive. The inconsistencies among de facto regime classifications, as well as between de jure and de facto classifications, were allowed to drag on for the whole period covered by this evaluation, contributing to a lack of clarity in analysis. There has been a failure to build consensus at the Executive Board to resolve this.

62. Staff and management too often failed to provide analytical backing for their recommendations for regime shift; and on other occasions they missed opportunities to give a clear view on the regime choice made by the authorities. The lack of analysis supporting regime advice gave support to the notion that the IMF’s advice, at times, was based on fashion rather than tailored to the country-specific circumstances. Too often, also, staff assessments of existing regimes stopped at the backward-looking statement that “the regime in place had served the country well,” with insufficient analysis to back this up.

63. On exchange rate levels, while analysis has improved, there were too many cases where staff’s assessments were not stated clearly. Quantitative analysis of exchange rate levels increased but was still far from universal, and the choice of methodology some-times appeared arbitrary. For developing countries, greater attention to cost and other data would have strengthened the analysis of exchange rate levels.

64. Data provision for the purpose of exchange rate surveillance was a serious problem. Staff appear not to have flagged to the Executive Board the full extent to which the data shortcomings hampered the conduct of exchange rate surveillance, including when authorities were unwilling to provide data, and in cases when Executive Board discussion may have been materially affected. In not pursuing data issues more forcefully, including those related to intervention, staff gave high weight to maintaining smooth relations with the authorities and/or perceived a lack of support by management and the Executive Board for a stronger stance.

65. Discussion of policy spillovers, including the regional or systemic impact of large countries’ policies (as well as the effects of intervention activities on those countries in whose currencies such interventions take place) remained infrequent. Multilateral and financial surveillance had not been well integrated with bilateral surveillance during the evaluation period. Analysis of spillovers remained spotty for most countries, and attempts to assess the effects of intervention activities on other members in the context of Article IV consultations remained limited.


66. Management and the Executive Board should resolve inconsistencies and ambiguity over the issue of regime classification. Whatever solution is found would benefit from being approved by the Executive Board and would involve removing the stigma of particular labels. For Article IV staff reports for countries with intermediate regimes (all but independently floating rates and hard pegs), the priority should be to have an unambiguous description of the authorities’ regime, including how it works in practice. The description could be agreed to by the authorities and staff, or differences of view should be described clearly to the Board. Subsequent Article IV consultations could revalidate the existing description, or revise it.

67. IMF advice on exchange rate regimes should be backed up more explicitly by analytic work. Analysis Fund-wide could be improved by strengthening the framework for considering regime choice, building on work already done in some departments. For regimes in place, in Article IV staff reports it could be helpful to describe concisely the policy assumptions underlying a forward-looking staff assessment that the chosen regime will remain appropriate. Any differences of view on the assumptions would be reflected in the report. The assumptions laid out in one Article IV consultation would then provide markers for discussion at the next. When little had changed, the discussions on this issue would be appropriately short.

68. To improve assessments of the exchange rate level, the IMF should be at the forefront of developing the needed analytical framework, while more successfully translating existing methodologies into advice that is relevant to discussion of individual country cases. The genuine difficulty in doing this is no excuse for not making more progress. While improvements in methodology are often considered for the advanced and emerging market economies, scope exists for improving data and analysis for developing countries. (For example, thought could be given to working with other agencies to assemble cost data.)

69. Management and the Executive Board should consider further what lies behind the apparently serious problems of data provision for surveillance, and how incentive structures can be improved. A full analysis lies beyond the scope of this evaluation.

70. Incentives should be given to develop and implement guidance for the integration of spillovers into bilateral and regional surveillance. In addition to interdepartmental work to improve existing methodologies, a panel of senior officials in member countries could be asked to give advice on policy feedbacks—the “what if” question—that they would find useful to explore. In many cases, greater financial market expertise may be required to inform staff advice and contribute to discussion with authorities.

Management of Work on Exchange Rates


71. The work on exchange rates has not been as well organized and managed as it should have been. An enormous amount of activity on exchange rate issues was not well integrated. Despite some progress made (including the CGER), research—from inside and outside the IMF—and multilateral studies were not consistently distilled and absorbed into frontline operational work. Both country authorities and staff would have welcomed more practical help on analysis, cross-country comparisons, and financial market and foreign exchange market experience.

  • (1) Responsibility for exchange rate issues is scattered throughout the IMF. Area departments lead the discussion with authorities, and develop their own analysis; INS (training), MCM (classification issues and exchange market expertise, and the Global Financial Stability Report), PDR (policy development and review), RES (WEO and research), and STA (data issues) are all involved. The structure diffuses responsibility and accountability for prioritizing, pursuing, and disseminating work on exchange rate matters.

  • (2) Lack of understanding of financial markets has been identified as a factor that in the past limited the value of IMF advice. As suggested in the IEO evaluation of Multilateral Surveillance (IEO, 2006c), part of the problem may be that knowledge that resides in ICM (and now MCM) may not yet be well integrated into the work of the rest of the IMF. Another problem may be the scarcity of practical experience among IMF staff. Country officials interviewed for this evaluation attached particular weight to advice, and wanted more of it, from those who have practical experience in handling foreign exchange market and financial market issues.


72. Management should address how to bring better focus to the analytical work on exchange rates

  • (1) Management should clarify responsibility and accountability for exchange rate policy issues and actively use a forum like the Surveillance Committee to ensure proper focus on key issues, and to discuss a variety of different views and perspectives. The integration of financial sector work would be an important element. A key role of the structure should be to prioritize exchange rate policy issues and initiatives from across the IMF, including a multiyear agenda for policy, research, and statistical work.

  • (2) The structure of staff teams could be reconsidered. Better integration of financial market and foreign exchange market expertise at headquarters would be a start. But it is unlikely that this could bridge entirely the “expertise and experience gap” that was identified in this evaluation as a factor in some cases. Perhaps, on limited occasions, consultants or senior officials from a pool of foreign exchange market practitioners could join Article IV mission teams (in addition to TA missions, as now) to provide relevant expertise and cross country experience that would directly add value to the discussions with the authorities.

Confidentiality and Executive Board Oversight


73. There have been some limited cases where keeping the Board fully informed of the engagement of staff and management on an exchange rate policy issue would have been incompatible with being an effective interlocutor. In some instances, country authorities are simply not willing to discuss issues candidly with the IMF, in either bilateral or multilateral settings, if they believe the content of such discussions would be revealed to the Executive Board (and hence, potentially, to officials in all member countries), let alone markets. Yet it is clearly in the interests of the IMF (and the broader international community) that staff and management be engaged. This poses a real dilemma for accountability. While such instances may be relatively few, it is important that the Executive Board, management, and staff agree on new procedures to respect the very real confidentiality concerns that exist, while ensuring that steps are in place to provide adequate accountability. Simply pretending that no issue exists is not a responsible response.


74. An understanding is needed on what are the expectations for inclusion in the Article IV staff report, what may be mentioned orally at Board meetings, and what may be understood to have been discussed between staff and the authorities on the clear understanding that it would not be revealed to the Executive Board. The aim would be to clarify the accountability of the Board and management for what happens, while defusing what is obviously a major tension and a barrier to effective dialogue.

  • (1) Confidential policy discussions about possible policy actions in the case of contingencies should be a regular feature of the dialogue with member countries. Such discussions are likely to become more important since the speed required to respond to capital market events requires any preparations to be accomplished in advance. It should be understood that, for at least a subset of countries, staff would be expected to conduct “what if” scenario exercises looking at contingent plans for domestic policy shifts (including exit strategies), as well as for exogenous developments, and policy shifts and different exchange rate paths in other countries. While the staff report for a country might not discuss such scenarios, the Board would need to be assured that such exercises had been discussed.

  • (2) How can the Board exercise its accountability and oversight functions in this area?

    • The IEO evaluation of the IMF’s engagement with Argentina (IEO, 2004) made some suggestions that might be relevant, including:

      “Establish guidelines whereby the Board could explicitly authorize management to withhold certain issues from discussion in a full Board meeting, with a presumption that, once the sensitivity is no longer present, management’s decision is ex post subjected to Board scrutiny.”

    • A further option, which could avoid information being conveyed in any way to the Executive Board, would be to charge an independent party with the task of periodically reviewing all IMF activities on exchange rates not reported to the Board, and to provide the Board with a regular report certifying that necessary work had been done (for example on contingencies); assessing the effectiveness of such activities (without revealing countries or details); and giving a ruling on whether the information not shared with the Board was withheld for good reason.

Facilitating Multilateral Policy Coordination


75. Over the evaluation period, the scope for countries to act in concert to deal with “global imbalances” was not fully explored and alternative analysis of these imbalances, and related adjustment scenarios, could have received more attention. The following lessons can be drawn from earlier episodes of exchange rate policy coordination that may still have relevance. Success is made more likely by:

  • (1) Advance planning of various scenarios, and constantly validating conclusions against new information.

  • (2) Explicitly recognizing policy interdependencies and countries’ appropriate reactions to policy decisions taken by others.

  • (3) Supplementing regular staff discussions with policy dialogue between management and the highest political levels, and building up ways to communicate collectively with relevant groups of countries.


76. Opportunities for potential multilateral concerted action deserve to be a key strategic management focus. This work should, for the most part, be based on rigorous and compelling analysis of scenarios and involve a strategic plan to build consensus amongst key players. To highlight and learn more about policy interdependencies, this could involve alternative sets of scenario-based policy recommendations at the individual country level that are explicitly conditional on policy actions taken in other countries.


    Other Resources Citing This Publication