Statistical Implications of Inflation Targeting
Chapter

16 Inflation Targeting: Statistical Issues Suggested by the Experience of Six Emerging Market Countries

Author(s):
Carol Carson, Claudia Dziobek, and Charles Enoch
Published Date:
September 2002
Share
  • ShareShare
Show Summary Details
Author(s)
Edgar Ayales,, Randall C. Mgerris, and Alfredo Torrez 

This chapter reviews some of the macroeconomic statistics that are used by six emerging market countries that have adopted inflation-targeting regimes—that is, policy frameworks built around the pursuit of an explicit inflation target as the overriding objective of monetary policy—and suggests statistical issues for discussion.

The next section provides, for each country, a brief description of the information requirements for inflation targeting. Basic information requirements specific to inflation-targeting regimes are (1) an explicit inflation target, (2) inflation forecasts, and (3) data for interest rates and possibly other variables that are used in the day-to-day operations of monetary policy, as well as other macroeconomic data that are not necessarily specific to inflation-targeting regimes.

Subsequently, the chapter describes the IMF’s treatment of inflation-targeting regimes in its operations, including IMF-supported programs, and describes some data implications. This section considers adjustments that have been made to the IMF’s analytical framework, which has traditionally focused on monetary or exchange targets. It also describes recent updates in the IMF’s International Financial Statistics (IFS)—the IMF’s most comprehensive published set of data—to include information on the monetary policy frameworks of member countries and, in particular, to separately identify those countries that have inflation-targeting frameworks. It then discusses possible adjustments to the IMF’s presentation of country data, which currently focuses to some extent on the IMF’s traditional financial programming variables.

The final section presents some discussion issues that explore whether, and to what extent, the compilation, reporting, and publication by the IMF of data pertaining to inflation-targeting countries should be expanded or refined.

Country Experiences

Brazil

In June 1999 Brazil formally instituted inflation targeting (Decree 3088). Under the decree, the National Monetary Council sets the inflation target, which is specified as a range, or band, around a midpoint inflation rate. For 2001, the midpoint was 4 percent and the target range was 2 to 6 percent inflation. The current inflation target has a 3.5 percent midpoint and an upper bound of 6 percent by the end of 2002. The inflation targets are based on the broad consumer price index (CPI). If the target is not met, the governor of the central bank must send an open letter to the minister of finance explaining the reasons for the deviations from the target band.

The CPI is compiled by the Department of Price Indices of the Brazilian Institute of Geography and Statistics (IBGE).1 The CPI is compiled on a monthly basis, and the data together with the IBGE’s comments are released at a press conference within 9 to 15 days after the end of the reference month. Changes in methodology for the CPI are announced at least one month in advance. The data are presented in Indicators, an IBGE monthly publication.

For monetary policy purposes, the Central Bank of Brazil (CBB) analyzes (1) the inflation forecasts derived from two econometric forecasting models; (2) core inflation, based on a subset of the CPI components; (3) monetary aggregates, Ml and M2; and (4) other economic indicators, including producer prices, imports and exports, capital flows, salaries and wages, and capacity utilization. In addition, the CBB’s daily survey of market expectations of future inflation provides information about the public’s perceptions about future price developments.

The inflation-forecasting models are (1) a small-scale model that includes output growth, unemployment, exchange rates, and interest rates, but does not include monetary aggregates; and (2) a vector autoregressive (VAR) model that includes a monetary aggregate among the variables. The small-scale model does not contain a monetary aggregate, but the CBB continues to monitor movements in the monetary base and the money and credit aggregates. External shocks (for example, oil price increases) and the resulting effects on exchange rates are taken into consideration in fine-tuning the inflation forecasts.

Monetary policy decisions are delegated to the CBB’s Monetary Policy Committee, which meets every five weeks.2 The day-to-day conduct of monetary policy centers on the use of open market operations for offsetting movements in the overnight interbank rate.

The CBB publishes a quarterly report on the underlying causes of inflation, as well as a monthly statistical bulletin in which data pertaining to the CPI are presented. The CBB also reports to the Senate on recent movements in the monetary base, Ml, and M2.

Chile

Inflation targeting in Chile is sometimes dated as early as 1990, when the Central Bank of Chile (CBC) announced an annual inflation target for 1991. Full-fledged inflation targeting began in September 1999, at the same time a floating exchange rate system replaced the crawling peg system that had been in effect since 1985.3 Starting in January 2001, the CBC enhanced the inflation-targeting framework by moving from a year-end inflation target to an inflation target range of 2 to 4 percent over a one- to two-year targeting horizon.

The Instituto Nacional de Estadísticas (INE) compiles the monthly CPI, using a Laspeyres formula (1998 = 100) covering the Greater Santiago area.4 The INE disseminates the CPI through a news release within five days after the end of the reference month. No government officials outside the INE have access to the data before the release to the public. The data are usually released without accompanying commentary.

The inflation target is specified in terms of “headline” inflation, which is based on the full CPI. Monitoring and policy decisions are based in part on core inflation, which is calculated from a CPI that excludes fuel and certain food items with volatile prices. Core inflation, compiled by the CBC, is viewed as the most reliable indicator of inflationary pressures in the economy.

The CBC analyzes and forecasts data for both headline and core inflation. A VAR model is used to forecast inflation over various time horizons. The variables in the VAR model include the CPI, interest rates, wages, GDP, money supply, and exchange rates.

The CBC monitors the external current account deficit, nominal exchange rate, GDP growth, unemployment rate, wage growth, fiscal deficit, and market interest rates. The CBC also monitors current and expected movements in the monetary aggregates—albeit not as intermediate targets—as well as monthly data on the sources and uses of net international reserves.

The main policy instrument for achieving the target is open market operations for influencing the interbank interest rate. The CBC board determines the overnight interbank rate that is consistent with achievement of the inflation target.

The CBC submits semiannual reports, in January and May, to a Senate commission, and an annual report to the full Senate in September. To increase transparency, the CBC publishes, three times a year, a Monetary Policy Report, which contains a detailed description of the inflation target and its policy actions. The Monetary Policy Report communicates the central bank’s policies to the financial markets and other sectors of the economy. The CBC also publishes a calendar of its board meetings, six months in advance, and a summary of the board decisions on monetary policy, within 90 days after each meeting.

Colombia

In September 2000, the central bank of Colombia, Banco de la República (BDR), announced that it had formally adopted an inflation-targeting framework for monetary policy. The inflation target is based on the annual rate of growth of the national CPI. The target, initially set for a two-year period, was 8 percent for 2001 and 6 percent for 2002. Monetary policy had previously been guided by an intermediate target for the growth of base money and a trading band for the exchange rate.

The national CPI, a monthly fixed-weight Laspeyres index (December 1998 = 100), is compiled by the Departamento Administrativo Nacional de Estadísticas (DANE).5 The data are disseminated in the form of a press release, no later than five working days after the end of the reference period. Methodological changes are announced at the time the data are disseminated to the public.

The BDR uses two types of inflation forecasting models: (1) a time-series model that is based entirely on historical trends and (2) a VAR model that is specified to include consumer prices, output gap, exchange rates, and monetary aggregates. The inflation forecasts are finetuned using data on wages, employment, capacity utilization, and the fiscal deficit. Data from a quarterly survey of market participants’ expectations about future inflation, as well as information about the international economic outlook, are also used.

The BDR prepares quarterly inflation reports that contain data on current and forecasted inflation and a review of its fiscal and monetary policies.

Czech Republic

The Czech Republic adopted inflation targeting in December 1997, along with a floating exchange rate regime. In the initial period of inflation targeting, the Czech National Bank (CNB) used “net” inflation6 in specifying its main inflation target. The CNB’s medium-term inflation target for the end of 2000 was announced along with the adoption of inflation targeting in December 1997. In April 2001, the CNB decided to switch to targeting headline inflation, based on the full CPI, and to express the inflation target as a range. The announced target ranges were 3 to 5 percent by January 2002 and 2 to 4 percent by December 2005. The Czech Statistical Office compiles the monthly data for net and headline inflation.

The CNB uses open market operations to steer interest rates in the economy, manage liquidity in the money markets, and signal the CNB’s monetary policy stance. The main instrument for open market operations is 14-day repurchase agreements; shorter-term repurchase agreements are used on occasion. Foreign exchange and direct securities transactions, which are viewed as fine-tuning instruments, are used on an ad hoc basis to smooth the interest-rate effects of unexpected fluctuations in money-market liquidity.

The CNB forecasts both net and headline inflation. The headline inflation forecast, announced at the end of January 2002, has an 18-month horizon (February 2002 through July 2003). This headline inflation forecast (stated as a range of 2.3 to 3.6 percent per annum) has been disseminated to the public.

Since the inception of inflation targeting, the CNB has introduced monthly and quarterly surveys of inflation expectations. The respondents to the monthly survey are 12 firms in the financial corporations sector. The quarterly survey covers 160 nonfinancial corporations and 600 households, as well as the panel from the monthly survey.

The CNB’s quarterly Inflation Report contains information on actual and forecasted inflation. Commentary on Monetary and Economic Development provides an evaluation of monthly and annual CPI developments. The CNB is planning to include 18-month forecasts of headline inflation in its Inflation Report, signaling its commitment to controlling inflation over the medium term. The Inflation Report also contains an evaluation of monetary developments.

Mexico

Mexico formally adopted inflation targeting in January 2001, well after the adoption of a floating exchange in late 1994. The target is 5 percent, and the time horizon for achieving the target is three years.

The Bank of Mexico (BOM) compiles the national CPI and a core inflation index (CII)7 that excludes the most price-volatile goods and services covered in the national CPI. Both indices are published biweekly, within 10 days after the end of the reference period.8 Government officials do not have access to the data before it is released to the public. The price data are released without accompanying commentary. However, the BOM publishes a monthly report and a quarterly report that provide comprehensive analysis of the CPI, the CII, and the inflation measures that are based on these indices. The BOM does not publish its own inflation forecasts.

The BOM has no official model for forecasting inflation, but it is studying the use of formal inflation modeling that would be similar to the models used in other inflation-targeting countries. The BOM obtains data on private analysts’ inflation expectations. In formulating monetary policy, the BOM compares the level and trend of inflation expectations with the official inflation target, while monitoring aggregate demand, wage settlements, and exchange rates.

The day-to-day conduct of monetary policy involves the use of discount-window administration and, in particular, setting of the required levels of settlement balances that banks are required to hold at the BOM.9 The authorities have found that the exchange rate tends to be less volatile when both the level of settlement balances and short-term interest rates are used to gauge the tightness of monetary policy. The BOM plans to place more emphasis on movements in short-term interest rates when the inflation rate is more in line with inflation in industrial countries.

The BOM monitors the monetary base, Ml, M2, net international reserves (NIR), and interest rates, but it does not place much weight on the monetary aggregates or interest rates in the formulation of monetary policy. Whenever it influences the settlement balances in the banking system, the BOM issues a press release to communicate its views to the markets and thereby improve its accountability and the transparency of its actions.

Thailand

Thailand instituted an inflation-targeting regime in May 2000, nearly three years after it adopted a floating exchange rate in July 1997. Before inflation targeting, monetary policy focused on the setting of monetary base targets by the Bank of Thailand (BOT). The inflation target is a range, which is currently set at 0 to 3.5 percent inflation over a two-year horizon. The inflation target is specified in terms of the quarterly average rate of core inflation.

The price statistics are compiled by the Trade and Economic Indices Bureau of the Department of Internal Trade, Ministry of Commerce. Headline inflation data are derived from the full CPI, and core inflation data are derived by excluding raw food and energy (electricity and gasoline) from the core price index, because of the short-term volatility of their prices.

The BOT uses a medium-scale macroeconomic model (core inflation model) for economic forecasting and for simulating the effects of exogenous shocks and policy changes. The core inflation version of this econometric model contains (1) separate equations for the headline (full) and core CPI, (2) equations for five other price indices, (3) an equation for average earnings, and (4) equations for five price deflators.10 An equation for core inflation expectations (as a function of current and lagged core inflation) is also included in the model.

The BOT also makes use of data from a survey of market participants’ expectations (that is, forecasts) for headline inflation. The respondents to the survey, conducted by Reuters (Thailand), are “research houses” that include the research units in Thailand offices of major investment banking firms such as Merrill Lynch Phatra, Salomon Smith Barney, and Goldman Sachs and other research organizations such as the Thai Farmers Research Institute. The respondents provide GDP growth forecasts, as well as data on their expectations for CPI growth.

The monetary policy decisions are made by the Monetary Policy Committee, established in July 2001 and composed of eight senior officials of the BOT and two executives who serve as committee advisors from outside the BOT.

Under inflation targeting, the BOT continues to monitor developments in the monetary and financial markets. In particular, monitoring changes in the 14-day security repurchase rate and forecasting future levels of banking system liquidity, as measured by the monetary base, are still essential elements of the day-to-day conduct of monetary policy. More weight is now attached to movements in the individual currency and bank-reserve components of the monetary base. Government revenue, expenditures, and financing are taken into account in forecasting movements in the components of the monetary base.

The BOT recognizes that other monetary and financial data play an important role in policy formulation, given the diffuse nature of the transmission of monetary policy actions to the financial markets and the real sector. These data include (1) repurchase agreement, swap, foreign exchange, and securities transactions of money-market institutions, which have potential interest rate consequences; and (2) deposit and loan rates, credit aggregates, nonperforming loans, financial sector profitability, and Basel capital ratios, which are used in gauging the liquidity and solvency of the financial sector.

The BOT presents inflation analyses in its monthly, quarterly, and annual publications on economic and monetary conditions. The BOT’s quarterly Inflation Report contains detailed treatment of price developments. It includes BOT’s forecasts for headline and core inflation, as well as the results of the Reuters (Thailand) survey of the inflation expectations of research houses.

Issues for the IMF

Surveillance and IMF-Supported Programs

Inflation targeting has led the IMF to adapt the analytical framework used in its operations. The IMF’s financial programming framework has been built around targets for international reserves and for money and credit aggregates.11 The IMF adapted this framework for countries with inflation-targeting regimes, particularly given that the inflation target replaces exchange rate (and reserve) or monetary aggregate targets. As the following examples illustrate, adjustments are reflected in the countries surveyed in this paper, although it appears that the authorities and the IMF continue to monitor the traditional variables.

Following Brazil’s mid-1999 adoption of inflation targeting, the IMF-supported program at first continued to specify a net domestic asset (NDA) ceiling and an NIR floor. The ceiling was phased out by June 2000, but the NIR floor was retained as a benchmark for policy purposes. The IMF-supported program currently translates a year-end inflation target into a continuous targeting of inflation within a band, starting in 2001.

Similarly, the IMF-supported program in Colombia shifted from a focus on NDA and NIR targets to an inflation target, starting in late 2000 (IMF, 2001c,d). New procedures were developed to monitor monetary developments with emphasis on a quarterly path of inflation. In monitoring inflation, the IMF staff considers the central bank’s inflation forecast.

In the IMF’s regular macroeconomic consultations (Article IV) with Chile (IMF, 2001b), the Czech Republic (IMF, 2001e), Mexico (IMF, 2001h,i), and Thailand (IMF, 2001g), staff analyses also have been adjusted to reflect those countries’ inflation-targeting frameworks. Modification of these consultations has implications for the country data work and statistical publications of the IMF. Some of these issues are discussed below.

Data Published in International Financial Statistics

The growing prominence of inflation-targeting regimes is reflected in the new presentation of information on the exchange rate arrangements of the member countries, as introduced in the April 1999 issue of International Financial Statistics (IFS). Previously, the Exchange Rate Arrangements table in IFS grouped the member countries by exchange rate regime only.12 The table now contains a new classification of exchange rate regimes,13 showing a disaggregation of -member countries by monetary policy framework and identifying those countries that have inflation-targeting frameworks.14 Exchange rate regimes and anchors of monetary policy (as of December 31, 2001) from the June 2002 issue of IFS are reproduced in Appendix 16.1.

The specification for the inflation-targeting countries is given as:

… public announcement of medium-term numerical targets for inflation with a commitment by the monetary authority to achieve these targets. Additional key features include an increased communication with the public and markets about plans and objectives of monetary policymakers and increased accountability of the central bank for obtaining its inflation objectives. Monetary policy decisions are guided by the deviation of forecasts of future inflation from the announced inflation target, with the inflation forecasting acting (implicitly or explicitly) as the intermediate target of monetary policy.15

The Exchange Rate Arrangements table in the April 1999 issue of IFS showed eight countries in the category of inflation-targeting frameworks—of those countries described above, Chile and the Czech Republic, as well as Australia, Canada, New Zealand, Poland, Sweden, and the United Kingdom. In the June 2002 issue of IFS, the countries identified as having an inflation-targeting framework as of December 31, 2001 included Brazil, Chile, Colombia, the Czech Republic, Mexico, and Thailand, as well as 10 other countries.

Issues for Discussion

Review of the experiences of Brazil, Chile, Colombia, the Czech Republic, Mexico, and Thailand reveals several broad similarities among their inflation-targeting frameworks, as well as country-specific differences. Some similarities and differences, along with data issues prompted by their implementation of inflation-targeting frameworks, are summarized below.

Inflation Targets

All six countries announce inflation targets, but the target specifications differ with regard to (1) the target horizons (one to two years, three years, or longer), and (2) the specification of a target that is a single inflation rate, or one that is a range, or band, around a single inflation rate.

Issue: Would descriptions of the inflation targets be useful information in IFS?

Inflation Measures

Some countries’ inflation targets are based on growth rates for the full CPI (headline inflation), whereas other countries’ targets are based on more narrowly defined measures of consumer prices (core inflation).16

Issues: Should the focus be only on the price statistics on which the country’s inflation targeting is based? Should both full-CPI and core-inflation measures be published in IFS?

Inflation Forecasting

The inflation forecasting models used by the six countries differ with respect to econometric technique and the variables included.

Issues: Looking ahead, what additional data might be included in the specification of the forecasting models or for the fine-tuning of the inflation forecasts?

Surveys of the Public’s Expectations of Inflation

Some of the countries’ inflation-targeting frameworks make use of surveys of market expectations about future inflation. In addition to their usefulness in inflation forecasting and monitoring, such data provide a potentially important indicator of the credibility of an inflation-targeting regime. At the outset of inflation targeting, significant differences may exist between the authorities’ inflation forecasts and the public’s perceptions about expected inflation. As official inflation targets are met over time, the gain in credibility should be reflected in a convergence of official inflation forecasts and the forecasts embodied in the surveys of market expectation, particularly if the public has timely access to the official forecasts and views them as reliable.

Issues: Would further work on surveys of market expectations be useful? Would it be useful to study the practices (sample size and stratification, periodicity, questionnaire design, and so forth) used in conducting such surveys?

Interest Rates

Inflation targeting involves a shift in emphasis from quantity data (for example, money and credit aggregates) to price data, as well as modifications in the analytical use of both quantity and price data. In addition to the direct use of data on the prices of commodities and non-financial services (for example, CPI data), inflation targeting may spawn new uses for data on the prices of credit—that is, interest rates. Under inflation targeting, an interbank interest rate (overnight or otherwise short-term) retains its traditional role as an indicator for calibrating the monetary policy actions. Interest rates also play a role in inflation forecasting—as explanatory variables in econometric forecasting models and possibly in the direct analysis of expected inflation as embedded in the term structure of interest rates.

Issues: Would additional data series on interest rates (including, possibly, forward rates) be potentially useful for inflation forecasting? Which interest rate series should be reported to the IMF and which series should be published in IFS17or other IMF publications? Should the published data include yield series for short-, medium-, and long-term government securities18to provide the data user with data for at least three points along the yield curve? Should data on forward rates be reported to the IMF?

Appendix 16.1:Exchange Rate Regimes and Anchors of Monetary Policy as of December 31, 2001
Monetary Policy Framework
Exchange Rate

Regime

(number of

countries)
Exchange rate anchor01,02Monetary

aggregate

target
Inflation-

targeting

framework
IMF-supported or

other monetary

program
Other
Exchange arrangements with no separate legal tender (40)Another currency as legal tenderCFA Franc ZoneEuro Area06,07
ECCU03WAEMU04CAEMC05Austria
EcuadortAntigua & BarbudaBenintCameroon†Belgium
El Salvador08DominicaBurkina Faso†C. Afr. Rep.†Finland
KiribatiGrenadaCôte d’lvoire†Chad†France
Marshall Isl.St. Kitts & NevisGuinea Bissau†Congo, Rep.ofGermany
MicronesiaSt. LuciaMali†Eq. GuineaGreece
PalauSt. Vincent & theNiger†Gabon†Ireland
PanamaGrenadinesSenegal†Italy
San MarinoTogo†Luxembourg
Netherlands
Portugal
Spain
Currency board arrangements (8)Argentina†
Bosnia and Herzegovina†
Brunei Darussalam
Bulgaria†
China: Hong Kong SAR
Djibouti†
Estonia†
Lithuania†
Other conventional fixed-peg arrangements (including de facto peg arrangements under managed floating) (40)Against a single currency (30)Against a composite (10)
ArubaBotswana09China, People’s
Bahamas, The01FijiRep. of*10
BahrainKuwait
BangladeshLatvia†
BarbadosLibyan A.J.
BelizeMalta
BhutanMorocco
Cape VerdeSamoa
China, People’s Rep. of*10Seychelles
Comoros11Vanuatu
Iran, I.R. of09,10
Jordan†10
Lebanon10
Lesotho†
Macedonia. FYR†10
Malaysia
Maldives10
Namibia
Nepal
Netherlands Antilles
Oman
Qatar10,12
Saudi Arabia10,12
Sudan10
Suriname09,10
Swaziland
Syrian Arab Republic09
Turkmenistan10
United Arab Emirates10,12
Zimbabwe10
Pegged exchange rates within horizontal bands (5)13Within a cooperative arrangement ERM II (1)14
Other band arrangements (4)Hungary*
DenmarkCyprus
Egypt09
Hungary*
Tonga
Crawling pegs (4)Bolivia†
Costa Rica10
Nicaragua†
Solomon Islands10
Exchange rates within crawling bands (6)15BelarusRomania†10Israel*
Honduras†Uruguay†
Israel*Venezuela, Rep. Bolivariana de
Managed floating with no pre-announced path for exchange rate (42)Ghana†Thailand†AzerbaijanAlgeria06
Guinea†Cambodia09Angola06
Guyana†CroatiaBurundi06
Indonesia†EthiopiaDomin. Rep06,09
Jamaica†IraqEritrea06
MauritiusKazakhstanGuatemala*
Sâo Tomé & Príncipe†KenyaIndia06
Kyrgyz RepublicMyanmar06,09,10
SloveniaLao, PDR09Paraguay06
Sri Lanka†MauritaniaSingapore06
TunisiaNigeriaSlovak Rep.06
PakistanUzbekistan06,09
Russian Federation
Rwanda
Trinidad & Tobago
Ukraine
Vietnam
Yugoslavia
Zambia
Independently floating (41)Gambia, The†AustraliaAlbaniaAfghanistan09,10
Malawi†Brazil17ArmeniaHaiti06
Mongolia†CanadaCongo, Dem. Rep.Japan06
Perut18Chile09GeorgiaLiberia06
Philippines†Colombia†MadagascarPapua New
Sierra Leone†Czech Rep.MoldovaGuinea06
Turkey†IcelandMozambiqueSomalia09,16
Yemen†KoreaTajikstanSwitzerland06
MexicoTanzaniaUnited States06
New ZealandUganda
Norway
Poland
South Africa
Sweden
United
Kingdom
Source: IMF, International Financial Statistics (IFS), June 2002, pp. 2–3.Note: The term “country,” as used in this publication, does not in all cases refer to a territorial entity that is a state as understood by international law and practice; the term also covers some territorial entities that are not states but for which statistical data are maintained and provided internationally on a separate and independent basis. For classification of exchange rate regime, see IFS.

A country with a * indicates that the country has more than one nominal anchor that may guide monetary policy. It should be noted, however, that it would not be possible, for practical purposes, to infer from this table which nominal anchor plays the principal role in conducting monetary policy.

A country with † indicates that the country has an IMF-supported or other monetary program.

ECCU: East Caribbean Currency Union. These countries have a currency board arrangement.

WAEMU: West African Economic and Monetary Union.

CAEMC: Central African Economic and Monetary Community.

The country has no explicitly stated nominal anchor, but rather monitors various indicators in conducting monetary policy.

Until they are withdrawn in February 2002, national currencies will retain their status as legal tender within their home territories.

For El Salvador, the printing of new colones, the domestic currency, is prohibited, but the existing stock of colones will continue to circulate, along with the U.S. dollar, as legal tender until all notes physically wear out.

Member maintained exchange regimes involving more than one market. The regime shown is that maintained in the major market.

The country has a de facto regime that differs from its de jure regime.

Comoros has the same arrangement with the French Treasury as do the CFA Franc Zone countries.

Exchange rates are determined on the basis of a fixed relationship to the SDR, within margins of up to ±7.25%. However, because of the maintenance of a relatively stable relationship with the U.S. dollar, these margins are not always observed.

The band width for these countries is: Cyprus (±2.25%), Denmark (±2.25%), Egypt (±3%), Hungary (±15%), and Tonga (±5%).

ERM II: European Exchange Rate Mechanism Phase II.

The band for these countries is: Belarus (±5%), Honduras (±7%), Israel (±22%), Romania (unannounced), Uruguay (±3%), and Venezuela (±7.5%).

There is no relevant information available for the country.

Brazil maintains an IMF-supported program.

Peru’s exchange rate regime has been reclassified, retroactively, because Peru has been maintaining an independently floating exchange rate.

Source: IMF, International Financial Statistics (IFS), June 2002, pp. 2–3.Note: The term “country,” as used in this publication, does not in all cases refer to a territorial entity that is a state as understood by international law and practice; the term also covers some territorial entities that are not states but for which statistical data are maintained and provided internationally on a separate and independent basis. For classification of exchange rate regime, see IFS.

A country with a * indicates that the country has more than one nominal anchor that may guide monetary policy. It should be noted, however, that it would not be possible, for practical purposes, to infer from this table which nominal anchor plays the principal role in conducting monetary policy.

A country with † indicates that the country has an IMF-supported or other monetary program.

ECCU: East Caribbean Currency Union. These countries have a currency board arrangement.

WAEMU: West African Economic and Monetary Union.

CAEMC: Central African Economic and Monetary Community.

The country has no explicitly stated nominal anchor, but rather monitors various indicators in conducting monetary policy.

Until they are withdrawn in February 2002, national currencies will retain their status as legal tender within their home territories.

For El Salvador, the printing of new colones, the domestic currency, is prohibited, but the existing stock of colones will continue to circulate, along with the U.S. dollar, as legal tender until all notes physically wear out.

Member maintained exchange regimes involving more than one market. The regime shown is that maintained in the major market.

The country has a de facto regime that differs from its de jure regime.

Comoros has the same arrangement with the French Treasury as do the CFA Franc Zone countries.

Exchange rates are determined on the basis of a fixed relationship to the SDR, within margins of up to ±7.25%. However, because of the maintenance of a relatively stable relationship with the U.S. dollar, these margins are not always observed.

The band width for these countries is: Cyprus (±2.25%), Denmark (±2.25%), Egypt (±3%), Hungary (±15%), and Tonga (±5%).

ERM II: European Exchange Rate Mechanism Phase II.

The band for these countries is: Belarus (±5%), Honduras (±7%), Israel (±22%), Romania (unannounced), Uruguay (±3%), and Venezuela (±7.5%).

There is no relevant information available for the country.

Brazil maintains an IMF-supported program.

Peru’s exchange rate regime has been reclassified, retroactively, because Peru has been maintaining an independently floating exchange rate.

This chapter benefited from comments of desk economists in the Asia and Pacific, European I, and Western Hemisphere Departments of the IMF and authorities in the respective countries.

The CPI measures consumer prices of goods and services purchased by households in the metropolitan areas of 11 cities. The areas covered by the index represent 40 percent of the urban population and 30 percent of the total population of the country. The expenditure weights in the CPIs compiled by the IBGE are based on the Household Budget Survey conducted in 1995–96.

The minutes of the Monetary Policy Committee’s meetings are published one week after the meeting.

See Schaechter, Stone, and Zelmer (2000) for a description of Chilean inflation targeting and inflation experience in the decade leading up to September 1999.

The full CPI is based on the prices of more than 1,300 goods and services in 368 categories within five major groups: food, housing, clothing, transportation and communications, and other. The methodology used for computing the CPI is described in detail in the INE’s Indice de Precios al Consumidor, IPC, base: December 1998 = 100, Aspectos Metodológìcos. A summary of the methodology is available on the Chile page of the IMF’s Dissemination Standards Bulletin Board (DSBB).

The methodology used to compile the CPI is described in the DANE’S statistical bulletin and is summarized in the IMF’s DSBB.

“Net” inflation in the Czech Republic differs somewhat from the construct for core inflation as found in some of the other five countries. Both core inflation and net inflation are based on a subset of the items in the full CPI. Construction of core inflation calls for the exclusion of items that display substantial short-term price variability. The Czech Republic’s net inflation measure excludes items that have regulated prices and adjusts for the price effects arising from adjustments in indirect taxes and the abolition of subsidies.

Mexico is the only one of the six countries that relies on its central bank to compile the CPI. Compilation of the CPI by a central bank or other policymaking institution could raise concerns about conflict of interest—real or perceived. In Mexico, such concerns are ameliorated by having the data certified by the Organization International de Estandares, a nonresident organization that provides a semiannual independent certification of the quality of the CPI data for Mexico.

The methodology for computing the national CPI, which comprises four subindices (agricultural products, education, administered goods and services, and core components), is posted on the IMF’s DSBB.

To signal its monetary policy stance, the BOM announces a daily target for net settlement balances.

Price equations are specified for the non-oil import, energy, retail petroleum, raw food, and farm price indices. Equations are specified for the public investment, government consumption, export, import, and GDP price deflators. Thus, the model includes equations for raw food and energy prices, even though these items are excluded from the measure of core inflation.

This is true for IMF-supported programs as well as for the IMF’s regular macroeconomic consultations with member countries (conducted in accordance with Article IV of the IMF’s Articles of Agreement). Blejer and others (2001) discuss the IMF-supported programs in inflation-targeting countries.

The classifications were (1) currency-pegged, either to a single currency (disaggregated by individual currency) or to a composite (that is, a basket of currencies), (2) flexibility limited to a single currency or group of currencies, and (3) more flexible arrangements (disaggregated by “other flexible arrangements” and “independently floating”).

The classifications are (1) exchange arrangements with no separate legal tender (countries that are “dollarized” or belong to a currency union), (2) currency board arrangements, (3) other conventional fixed peg arrangements, (4) pegged exchange rates within horizontal bands, (5) crawling pegs, (6) exchange rates within crawling bands, (7) managed floating with no pre-announced path for the exchange rate, and (8) independently floating.

The classifications for other types of monetary policy frameworks are (1) exchange rate anchor, (2) monetary aggregate targeting, (3) inflation-targeting framework, (4) IMF-supported or other monetary program, and (5) other.

Conceivably, the inflation target could be specified in terms of a broader measure of inflation, more akin to an implicit price deflator for GDP, but this approach would be practical only if data for the price series could be produced on a timely basis.

The interest rate sections of the current IFS pages are not uniform with respect to the interest rates presented. Deposit and lending rate series are shown for each country, the discount rate (central bank rate) is shown for each country except Mexico, and a money market (interbank) rate is shown for each country except the Czech Republic. Savings rates are shown for Brazil, Chile, and Mexico; foreign-currency deposit and lending rates are shown for Chile; a refinancing rate is shown for the Czech Republic; and an average cost of funds is shown for Mexico. The Brazil page includes two treasury bill rates—for domestic- and for foreign-currency-denominated treasury bills—but no series for a government bond yield. The Thailand page contains a government bond yield, but no treasury bill rate. The Chile, Colombia, and Czech pages do not contain rates (yields) on government securities. Of the six country pages, only the Mexico page contains both a treasury bill rate and a government bond yield. The sampling methods for deposit and loan rate quotations and the computational features for interest rate series differ across countries.

The government securities markets would need to have sufficient depth, breadth, and resiliency for providing reliable yield quotations.

    Other Resources Citing This Publication