Speech

Inflation Targeting
Introductory Remarks by Horst Köhler
Managing Director
International Monetary Fund
Seminar on the Statistical Implications of Inflation Targeting
International Monetary Fund, Washington, D.C.
February 28, 2002

Ladies and gentlemen, it is a pleasure to welcome you to this Seminar on the Statistical Implications of Inflation Targeting. Your presence here today confirms to me how important this topic is for our members.

Exchange rate policy has always been central to the work of the IMF. Ever since the breakdown of the Bretton Woods system, there has been a widespread desire to avoid excessive volatility in the exchange rates of the world's major currencies. And the IMF's bilateral and multilateral surveillance have become important ways for the international community to signal its views and seek better coordination of underlying economic policies. Our Research Department's analysis points to a persistent overvaluation of the US dollar and the undervaluation of the euro. But in a world of highly integrated global capital markets, I do not see any realistic alternative to floating exchange rates among the major currencies. So our members need to deal with currency misalignments by concentrating on the fundamentals-especially, in the cases of the EU and Japan, by accelerating the pace of key structural reforms.

Looking beyond the major currencies, an important conclusion of the IMF's research and reviews of country experience is that no single exchange rate regime is appropriate for all members in all circumstances. Nevertheless, there is now an increasing tendency among our members to choose "corner solutions." A country that is willing to abandon all monetary policy discretion may be able to adopt a hard peg, such as a currency board arrangement. Because this deprives the country of instruments to deal with external shocks, living safely under a hard peg obliges a country to have not only a disciplined fiscal policy, but also particularly sound financial and corporate sectors and considerable wage and price flexibility. As shown tragically in Argentina, when these conditions do not hold, it can be very difficult for a country to arrange a timely exit strategy.

We advise emerging market countries to be very cautious about adopting pegged or heavily managed exchange rate systems. When a country is open to international capital flows, the reaction to any hint of unsustainable macroeconomic policies can be swift and severe. And a country considering an exchange rate peg needs to be fully aware of the associated costs, including the possibility that extraordinarily high interest rates might be required at times of severe financial market pressure. Especially, its domestic financial institutions and businesses must be well prepared to live with such policy adjustments. Where there is doubt that these requirements will be met, a flexible exchange rate regime is a better choice.

For other developing countries, there is in principle a wider range of choice in exchange rate regimes, provided these are backed by the appropriate macroeconomic and structural policies. And we recognize that very few countries, advanced or developing, are indifferent to the behavior of their exchange rates. But on balance, we see floating exchange rates as the safest solution for a wide range of countries. Unfortunately, this has the disadvantage of leaving the public and markets without a clear anchor, which can make a country vulnerable to accelerating inflation in response to domestic or external shocks. And that, ultimately, is why we are here today.

Inflation targeting has been adopted by several countries with floating exchange rate systems, as a way to anchor inflation expectations. Many other countries are actively considering this possibility, and some are committed to do so under Fund-supported programs. After a decade of experience, we have seen inflation targeting regimes in action through all phases of the business cycle, and we are beginning to have a basis for assessing their performance and refining our views on how best to manage them.

As the IMF stresses to countries that are considering the adoption of inflation targeting, the keys to success are transparency and credibility. Once a political decision has been taken to make the inflation target the primary objective of monetary policy, it is crucial for the monetary authorities to keep the public regularly informed about their actions to meet that objective and the basis for the judgments that they make. Perhaps even more that other monetary regimes, inflation targeting obliges the central bank to safeguard its credibility in pursuing the inflation goal. For this reason, inflation targeters are almost invariably countries in which the central bank has a high degree of operational independence. But it is also important to avoid a deflationary bias, which would impose unnecessary costs on society and risk undermining the political basis for the inflation targeting regime and the independence of the central bank.

While there are now many analyses of the experience under inflation targeting, relatively little has been written about the implications for economic statistics. This is an important omission. In preparing for inflation targeting, most countries have adopted an existing, well-known price index as the target-generally the national consumer price index-and they have used a selection of available data series in forecasting inflation and assessing the effects of monetary policy. Moving forward, it will be important to explore whether this is the most appropriate index to target, whether some components of the CPI should be systematically excluded or added to the index, and what new statistics might be needed for related analysis and forecasting.

It is also possible that the institutional arrangements for producing economic statistics may need to be adapted in a country with an inflation targeting regime. Just as with a central bank, it is important for the national statistical agency to be operationally independent, while receiving adequate resources from the government to do its job. At the same time, there should be good coordination with the central bank. In many countries the central bank sits at one end of town, and the statistics agency at another, or-as in my own country-even in another city. The mixture of senior central bank officials and those from statistics agencies at this seminar provides a particularly valuable opportunity to exchange ideas on these issues.

For countries with inflation targeting regimes, forward-looking indicators such as stock indices, real estate prices, and derivatives yields provide crucial information for assessing the appropriateness of monetary policy. This type of data is also important for the work of the IMF in promoting stability of domestic and international capital markets, and we often face gaps in information and questions over the suitability over of what we do have. While such information is generally produced by markets or private firms, there might be a role for the public sector-either in collecting these statistics, setting standards or guidelines, or at least encouraging full disclosure of the methodology that was used to assemble the data.

Finally, I would like to comment on why the IMF, in particular, has chosen to host a seminar on the statistical implications of inflation targeting. This topic brings together many of the key themes of our ongoing work to improve the functioning of the international monetary system-in particular, strengthening IMF surveillance, enhancing transparency, and using internationally-recognized standards and codes as new rules of the game for global governance. Our Special Data Dissemination Standard was the first of the IMF's core standards to be introduced, under the leadership of Carol Carson. In parallel with the SDDS, during the past year we have begun using the IMF's new Data Quality Assessment Framework, as a systemic approach to assessing country practices and capabilities in the statistical area. While the Data Quality Assessment Framework was not designed with inflation targeting countries specifically in mind, it is clear that data quality in the sense of the framework is precisely what underpins a well-functioning inflation targeting regime.

The SDDS was designed for countries seeking access to international capital markets, regardless of their exchange rate regimes. But another one of our objectives at this conference is to consider whether the SDDS might need to be enhanced to take account of special data needs of countries with inflation targeting regimes. Whatever the outcome, it is noteworthy, and certainly welcome, that virtually all countries that have adopted an inflation targeting regime are also among the 50 subscribers to the SDDS.

Ladies and gentlemen, you can trust that the IMF will be listening closely to the presentations from country representatives at this seminar. We will also be offering several of our own papers, sharing our recent work, and welcoming feedback. And, I would add, if any of the countries represented here today would like to request our help in developing the statistical underpinnings for an inflation targeting regime, we would be pleased to be your partner. Thank you for coming, and I wish you a fruitful discussion.