II. Building Blocks for Increasing Monetary Policy Effectiveness

During the last 20 years, LA5 central banks undertook a comprehensive macroeconomic and financial sector reform. Significant institutional, policy, and operational changes took place (see Appendix 1). Four of the five countries approved new central bank legislation.³ All five central banks adopted IT and enhanced the use of market-based instruments to conduct monetary policy. These reforms improved the capacity of central banks to conduct effective monetary policy, which helped reduce inflation to historical low rates.⁴ Over time, the LA5 central banks built credibility and gained a reputation of being committed to price stability, which was critical for weathering the recent crisis without experiencing major macroeconomic disarray.

A. Enhancing Central Bank Autonomy and Accountability

New legislations strengthened central banks’ autonomy.⁵ New laws gave most central banks the primary mandate to focus on preserving price stability and provided central banks with political independence to take monetary policy decisions as needed to achieve their mandate. A milestone in the reform adopted in most countries was to delink monetary policy decisions from the political cycle, by appointing central bank board members for a longer period than—or in overlap with—the presidential term. In addition, some laws limited and others banned the central bank from lending to the government. The laws also granted central banks the freedom to choose the monetary operations they judged suitable to reach their stability goals. To conduct monetary policy, most LA5 central banks were assigned goal and instrument autonomy (Table 1), a practice not common in advanced IT countries, where most central banks enjoy only instrument autonomy.⁶ This autonomy was compounded with the governments’ obligation to preserve central banks’ capital integrity, except in Mexico.

Table 1.

Key Institutional Arrangements for IT in LA5 Countries as of end–2009

^a/Central banks have target autonomy when they set the inflation target without necessarily agreeing with the government. Central banks have—only—instrument autonomy when they are free to choose the policy means to achieve the target.

^b/Refers to the Minister of Finance.

Table 1.

Key Institutional Arrangements for IT in LA5 Countries as of end–2009

Countries	Adoption date	Type of central bank autonomy a/	Decision- making body	Composition of the CB Board		Frequency of policy meetings
Brazil	Jun. 1999	Instrument	CB Board	8 internal		Every 6 weeks
Chile	Sep. 1999	Target + instrument	CB Board	5 internal		Monthly
Colombia	Sep. 1999	Target + instrument	CB Board	6 internal	MoF b/	Monthly
Mexico	Jan. 2001	Target + instrument	CB Board	5 internal		Monthly
Peru	Jan. 2002	Target + instrument	CB Board	1 internal	6 external	Monthly

^a/Central banks have target autonomy when they set the inflation target without necessarily agreeing with the government. Central banks have—only—instrument autonomy when they are free to choose the policy means to achieve the target.

^b/Refers to the Minister of Finance.

View Table

Table 1.

Key Institutional Arrangements for IT in LA5 Countries as of end–2009

Countries	Adoption date	Type of central bank autonomy a/	Decision- making body	Composition of the CB Board		Frequency of policy meetings
Brazil	Jun. 1999	Instrument	CB Board	8 internal		Every 6 weeks
Chile	Sep. 1999	Target + instrument	CB Board	5 internal		Monthly
Colombia	Sep. 1999	Target + instrument	CB Board	6 internal	MoF b/	Monthly
Mexico	Jan. 2001	Target + instrument	CB Board	5 internal		Monthly
Peru	Jan. 2002	Target + instrument	CB Board	1 internal	6 external	Monthly

^a/Central banks have target autonomy when they set the inflation target without necessarily agreeing with the government. Central banks have—only—instrument autonomy when they are free to choose the policy means to achieve the target.

^b/Refers to the Minister of Finance.

The laws put the central bank boards in charge of making the policy decisions. In contrast to the LA5, in other IT central banks, a monetary policy committee performs this task (e.g., Korea, New Zealand, Thailand, South Africa, and the United Kingdom).⁷ The governing arrangement of the LA5 central bank boards differs slightly from those in many other IT central banks. Except for Peru, the boards are comprised of full-time members, as opposed to some other IT central banks in advanced countries and emerging markets, where the board also includes external members (for instance, the United Kingdom, Australia, Hungary, Poland, and the Philippines). In addition, in Colombia, the Minister of Finance chairs the board and in Chile the minister can participate in board sessions without casting votes.⁸

The new laws also strengthened central bank accountability to the markets and the society at large. Central banks became responsible for achieving inflation targets and for disclosing and explaining their policies and goals (see Appendix 1). Most central bank laws require central bank governors to explain monetary policy and inflation developments and outlooks to the legislative body. Before the legal reform, central banks only needed to prepare and publish an annual report. In practice, these reports did not help decision making of market participants or guide inflation expectations.

The autonomy and accountability of LA5 central banks is high compared to that of central banks in other developing and emerging market countries. Following the legal reforms, the autonomy and accountability of LA5 central banks increased substantially, and using a metric similar to that used in the literature, arguably tripled.⁹ The index we have used weighs legal provisions that strengthen autonomy and accountability and, hence, expands the Cukierman and others index that only focuses on the autonomy of central banks (Appendix 2).¹⁰ Today, the LA5 central banks are among the most autonomous central banks in Latin America and are, at least, similar compared to other important emerging market central banks that also sharply changed their central bank legislation (Table 2). On average, they rank better than the other 10 countries in our sample under the vast majority of criteria included in the autonomy index (Table 3).¹¹

Table 2.

Central Bank Autonomy in LA5 and in Other Countries as of 2009

Source: IMF’s central bank legislation data base and authors’ calculations.

^(*)It incorporates the provisions from the presidential decree that established the inflation targeting regime in 1999.

Table 2.

Central Bank Autonomy in LA5 and in Other Countries as of 2009

LA5 countries	CWNE index	Other LA countries	CWNE index	Other selected IT countries	CWNE index
Brazil (*)	0.81	Argentina	0.77	Czech	0.86
Chile	0.95	Costa Rica	0.63	Indonesia	0.91
Colombia	0.82	Dominican	0.82	Poland	0.85
		Rep.
Peru	0.86	Guatemala	0.74	Thailand	0.75
Mexico	0.85	Uruguay	0.74	Turkey	0.81

Source: IMF’s central bank legislation data base and authors’ calculations.

^(*)It incorporates the provisions from the presidential decree that established the inflation targeting regime in 1999.

View Table

Table 2.

Central Bank Autonomy in LA5 and in Other Countries as of 2009

LA5 countries	CWNE index	Other LA countries	CWNE index	Other selected IT countries	CWNE index
Brazil (*)	0.81	Argentina	0.77	Czech	0.86
Chile	0.95	Costa Rica	0.63	Indonesia	0.91
Colombia	0.82	Dominican	0.82	Poland	0.85
		Rep.
Peru	0.86	Guatemala	0.74	Thailand	0.75
Mexico	0.85	Uruguay	0.74	Turkey	0.81

Source: IMF’s central bank legislation data base and authors’ calculations.

^(*)It incorporates the provisions from the presidential decree that established the inflation targeting regime in 1999.

Table 3.

Criteria behind the CWNE Index in the LA5 and Other Groups of Countries

Source: IMF’s central bank legislation data base and authors’ calculations.

^(*)In the case of Brazil, it incorporates the provisions from the presidential decree that established the inflation targeting regime in 1999.

Table 3.

Criteria behind the CWNE Index in the LA5 and Other Groups of Countries

CWNE index and its components	Average LA5 Countries (*)	Average other LA countries	Average other selected IT
Central Bank Board	0.73	0.49	0.55
Mandate	1.00	0.70	0.96
Policy formulation	0.92	0.97	0.87
Lending to government	0.93	0.83	0.83
Financial autonomy	0.76	0.68	0.51
Accountability	0.77	0.84	0.63
Total CWNE index	0.86	0.75	0.74

Source: IMF’s central bank legislation data base and authors’ calculations.

^(*)In the case of Brazil, it incorporates the provisions from the presidential decree that established the inflation targeting regime in 1999.

View Table

Table 3.

Criteria behind the CWNE Index in the LA5 and Other Groups of Countries

CWNE index and its components	Average LA5 Countries (*)	Average other LA countries	Average other selected IT
Central Bank Board	0.73	0.49	0.55
Mandate	1.00	0.70	0.96
Policy formulation	0.92	0.97	0.87
Lending to government	0.93	0.83	0.83
Financial autonomy	0.76	0.68	0.51
Accountability	0.77	0.84	0.63
Total CWNE index	0.86	0.75	0.74

Source: IMF’s central bank legislation data base and authors’ calculations.

^(*)In the case of Brazil, it incorporates the provisions from the presidential decree that established the inflation targeting regime in 1999.

B. Adopting an Appropriate Monetary Policy Framework

With a clear mandate on price stability, each of the LA5 central banks adopted an IT regime as their new monetary policy framework. IT provided an anchor for inflationary expectations and allowed for increased monetary policy flexibility, including on the exchange rate front. The central banks refined their IT frameworks gradually over a decade. To measure monetary policy success, they all chose inflation target ranges or a point inflation target with some tolerance bands.¹² The targets are set indefinitely, except in Brazil and Colombia, where the central banks review and may change the inflation target annually. They all used the consumer price index as the target measure. The horizon for bringing inflation back on target is one year, except in Chile, where it is up to two years (Table 4).¹³

Table 4.

Key Parameters about IT Regimes in LA5 Countries as of end-2009

Source: Central banks’ websites.

^a/In Brazil, the National Monetary Council (see Appendix 1) sets the inflation target for the end of each year (t) by end-June of two years in advance (t-2).

^b/The Bank of the Republic of Colombia has a long-run inflation target and reviews annually the inflation target for the following year.

Table 4.

Key Parameters about IT Regimes in LA5 Countries as of end-2009

Countries	Last change of inflation target	Target for inflation rate	Target measure	Frequency of revisions of inflation target	Horizon for bringing inflation back on target
Brazil	Jun. 2003	4.5% ± 2%	CPI	Every year a/	1 year
Chile	Sep. 1999	2% - 4%	CPI	No revisions	2 years
Colombia	Oct. 2009	2% - 4%	CPI	Every year b/	1 year
Mexico	Jan. 2003	3% ± 1%	CPI	No revisions	1 year
Peru	Jan. 2002	2% ± 1%	CPI	No revisions	1 year

Source: Central banks’ websites.

^a/In Brazil, the National Monetary Council (see Appendix 1) sets the inflation target for the end of each year (t) by end-June of two years in advance (t-2).

^b/The Bank of the Republic of Colombia has a long-run inflation target and reviews annually the inflation target for the following year.

View Table

Table 4.

Key Parameters about IT Regimes in LA5 Countries as of end-2009

Countries	Last change of inflation target	Target for inflation rate	Target measure	Frequency of revisions of inflation target	Horizon for bringing inflation back on target
Brazil	Jun. 2003	4.5% ± 2%	CPI	Every year a/	1 year
Chile	Sep. 1999	2% - 4%	CPI	No revisions	2 years
Colombia	Oct. 2009	2% - 4%	CPI	Every year b/	1 year
Mexico	Jan. 2003	3% ± 1%	CPI	No revisions	1 year
Peru	Jan. 2002	2% ± 1%	CPI	No revisions	1 year

Source: Central banks’ websites.

^a/In Brazil, the National Monetary Council (see Appendix 1) sets the inflation target for the end of each year (t) by end-June of two years in advance (t-2).

^b/The Bank of the Republic of Colombia has a long-run inflation target and reviews annually the inflation target for the following year.

The new policy framework also required some reorganization of relevant central bank departments. For instance, the Central Bank of Brazil created a new Research Department in 1999 to centralize macroeconomic modeling and inflation forecasting.¹⁴ Also, the Central Bank of Peru replaced in 2005 the existing sector-based organization (monetary, fiscal, external, real, and global) that responded to the logic of financial programming, with a process-based institutional arrangement comprising processes for macroeconomic statistics, the international economy, monetary policy, macroeconomic analysis, and research.

LA5 central banks improved in communicating clearly the rationale of their policy strategy and individual policy decisions. Table 5 suggests that transparency in LA5 central banks has improved remarkably after each country adopted inflation targeting.¹⁵ After introducing IT, the LA5 central banks emphasized a forward-looking approach, explaining to market participants where monetary policy is heading. Since the beginning of the current decade, LA5 central banks prepare and publicly disclose inflation or monetary policy reports three or four times a year and, more recently, they also elaborate and disseminate financial stability reports. They decide the level of the policy rate during preannounced meetings, disclose the rationale behind their rate and other policy decisions, and some publish the minutes of their board meetings (Brazil, Chile, and Colombia). In general, LA5 central banks also constantly communicate with the market and issue press releases as needed. The Central Bank of Brazil created the Investors Relations Group in 1999 with the aim of enhancing two-way communication with the private sector, in particular with investors, to support the effectiveness of IT. Central banks also make available data on market views about key macroeconomic variables, in particular, about expectations on inflation and economic activity, as well as central bank forecasts. Similarly, LA5 central banks have produced some basic material to educate economic agents and made it available at their websites.¹⁶

Table 5.

LA5 Central Banks’ Transparency ^*/

(before and after adopting inflation targeting)

Source: Survey conducted to LA5 central banks.

^*/Questions asked in the survey and the coding of the answers correspond to the categories and the criteria in the transparency index by Crowe and Meade (2008).

^**/Based on central banks’ legislation (not asked in the survey).

Table 5.

LA5 Central Banks’ Transparency ^*/

(before and after adopting inflation targeting)

	Brazil		Chile		Colombia		Mexico		Peru
	After	Before	After	Before	After	Before	After	Before	After	Before
1. Statutory objective **/	0.5	0.5	1.0	1.0	1.0	1.0	1.0	0.5	1.0	1.0
2. Explicit numerical target for inflation	1.0	0.0	1.0	1.0	1.0	1.0	1.0	1.0	1.0	0.0
3. Central bank publishes surveys markets use to estimate inflation expectations	1.0	0.0	1.0	0.0	1.0	0.0	1.0	0.0	1.0	0.0
4. The central bank publishes forward-looking analyses such as forecasts	1.0	0.0	1.0	0.0	1.0	0.0	1.0	0.0	1.0	0.0
5. The central bank publishes minutes of policy meetings	1.0	0.0	1.0	0.0	1.0	0.0	0.0	0.0	0.0	0.0
6. The central bank publishes voting patterns of policy meetings	1.0	0.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
7. The central bank publishes explanations the same day of policy changes	1.0	0.0	1.0	0.0	1.0	0.0	1.0	1.0	1.0	0.0
8. The central bank publishes explanations the same day that it decides no policy change	1.0	0.0	1.0	0.0	1.0	0.0	1.0	0.0	1.0	0.0
9. The central bank publishes discussions of risks to outlook or forecasts	1.0	0.0	0.5	0.0	1.0	0.0	1.0	0.0	1.0	0.0
10. The central bank publishes discussions of shocks or forecast errors after the facts	1.0	0.0	1.0	0.0	1.0	0.0	1.0	0.0	1.0	0.0
Totals	9.5	0.5	9.5	2.0	9.0	2.0	8.0	2.5	8.0	1.0
Differences (after – before)	9		7.5		7		5.5		7

Source: Survey conducted to LA5 central banks.

^*/Questions asked in the survey and the coding of the answers correspond to the categories and the criteria in the transparency index by Crowe and Meade (2008).

^**/Based on central banks’ legislation (not asked in the survey).

View Table

Table 5.

LA5 Central Banks’ Transparency ^*/

(before and after adopting inflation targeting)

	Brazil		Chile		Colombia		Mexico		Peru
	After	Before	After	Before	After	Before	After	Before	After	Before
1. Statutory objective **/	0.5	0.5	1.0	1.0	1.0	1.0	1.0	0.5	1.0	1.0
2. Explicit numerical target for inflation	1.0	0.0	1.0	1.0	1.0	1.0	1.0	1.0	1.0	0.0
3. Central bank publishes surveys markets use to estimate inflation expectations	1.0	0.0	1.0	0.0	1.0	0.0	1.0	0.0	1.0	0.0
4. The central bank publishes forward-looking analyses such as forecasts	1.0	0.0	1.0	0.0	1.0	0.0	1.0	0.0	1.0	0.0
5. The central bank publishes minutes of policy meetings	1.0	0.0	1.0	0.0	1.0	0.0	0.0	0.0	0.0	0.0
6. The central bank publishes voting patterns of policy meetings	1.0	0.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
7. The central bank publishes explanations the same day of policy changes	1.0	0.0	1.0	0.0	1.0	0.0	1.0	1.0	1.0	0.0
8. The central bank publishes explanations the same day that it decides no policy change	1.0	0.0	1.0	0.0	1.0	0.0	1.0	0.0	1.0	0.0
9. The central bank publishes discussions of risks to outlook or forecasts	1.0	0.0	0.5	0.0	1.0	0.0	1.0	0.0	1.0	0.0
10. The central bank publishes discussions of shocks or forecast errors after the facts	1.0	0.0	1.0	0.0	1.0	0.0	1.0	0.0	1.0	0.0
Totals	9.5	0.5	9.5	2.0	9.0	2.0	8.0	2.5	8.0	1.0
Differences (after – before)	9		7.5		7		5.5		7

Source: Survey conducted to LA5 central banks.

^*/Questions asked in the survey and the coding of the answers correspond to the categories and the criteria in the transparency index by Crowe and Meade (2008).

^**/Based on central banks’ legislation (not asked in the survey).

C. Improving the Operational Framework

LA5 central banks revamped their operational frameworks. They have converged in choosing a short-term interest rate (the “policy rate”) as their operational target. Central banks in Brazil, Chile, Mexico, and Peru chose a target for a market-based overnight interest rate as their policy rate. In turn, the Bank of the Republic of Colombia chose a central bank rate, namely the floor rate for the daily overnight liquidity injection auctions. The transition was swift in Brazil, Chile, and Colombia, but took longer in Mexico and Peru. Bank of Mexico adopted a policy rate as the operational target only in January 2008, after having followed some steps to replace its previous quantity-based operational target, the so-called “corto.” The Central Bank of Peru, whose financial system is highly dollarized, adopted a policy rate as an operational target in late 2003, after going through a gradual transition from quantities to prices (Box 1).¹⁷

Box 1.

The Transition from a Quantity to a Price Operational Target in Mexico and Peru

Mexico: A gradual strategy guaranteed a smooth transition from a quantity to a price operational target in Mexico. From September 1995 onward, Banxico had been conducting monetary operations by defining a target for bank reserves (the “corto”). Since March 2004, Banxico monetary policy announcements have signaled the intended changes in “monetary conditions,” which in practice have steered interbank overnight rates. Since then, the interbank market has in practice followed central bank signals for short-term interest rates, although the formal announcements were still in terms of the size of the “corto”. Subsequently, Banxico has refrained from changing the “corto” even if it would sometimes alter its stance for the monetary conditions—the last move in the “corto” took place in March 2005. However, until January 2008, the Mexican central bank did not announce a target for the interbank overnight interest rate. Instead, it referred to tightening or loosening monetary conditions by a certain number of basis points. As it announced the formal move towards adopting the overnight interbank interest rate as its operational target, Banxico explained that the “corto” had successfully helped bring inflation down to 4 percent in 2003 from 52 percent in 1995. However, the central bank needed to be more specific about the desired level for interest rates in an environment with more stability in financial markets and lower inflation.^{/1 /2}

Peru: Peru also adopted a gradual strategy to move from a monetary to an interest rate operational target. Until 2000, monetary policy controlled money base growth. In the beginning of 2001, the BCRP changed its operational target to the aggregate balance of commercial banks’ current accounts with the central bank. It also set a calendar of monetary policy meetings in which it announced the targeted range for the current account balance. In January 2002, Peru adopted inflation targeting as its monetary policy regime. Also in the beginning of 2002, the BCRP complemented the current account target by setting up a corridor for overnight interbank rates between the central bank’s overnight deposit rate and rediscount rate. This corridor became the monetary policy operational variable in the beginning of 2003. In September 2003, the BCRP announced that it would target, through its open market operations, an interbank overnight interest rate close to the center of the corridor. The changes in the operational target reduced volatility of the interbank interest rate, and improved the transmission to longer term interest rates.

View Full Size

• Download Figure
• Download figure as PowerPoint slide

¹See Banxico’s document titled “Instrumentación de la Política Monetaria a través de un Objetivo Operacional de Tasa de Interés (Anexo 3 del Informe sobre Inflación Julio-Septiembre 2007).” ²The idea behind the adoption of a volume target, when the “corto” was introduced, was that inflation and interest rates were so high and volatile that a policy interest rate would hardly be relevant.

To drive short-term interest rates to the policy rate, LA5 central banks conduct open market operations. They borrow or lend bank reserves (which is usually known as liquidity management) to keep their demand and supply in equilibrium and interbank market rates close to the policy rate. The Bank of Mexico is a net lender and has a net asset position with banks, while the central banks of Brazil, Chile and Peru usually—although not always—are net borrowers and have a net liability. The Bank of the Republic of Colombia has had a net asset position most of the time in the last five years, which has occasionally turned to a net liability. All LA5 central banks use repos for lending domestic currency. The central banks of Chile and Peru issue their own securities to withdraw (borrow) liquidity. The Brazilian central bank resorts to reverse repos of government securities and the Bank of the Republic of Colombia issues term deposits. The Bank of Mexico also injects liquidity through collateralized credits (guaranteed by the compulsory deposits at the central bank).

Central banks rely on liquidity forecasting to guide their open market operations. They have developed efficient means to anticipate factors that could lead to liquidity imbalances, which could in turn cause market interest rates to deviate from policy rates. Such factors include government cash flows and central bank foreign exchange operations, among others. Coordination with the government bodies in control of government cash flows and public debt has been crucial in improving liquidity forecasting.

LA5 central banks have also set up an interest rate corridor around the policy rate, which helped keep market interest rates close to target. The ceiling and the floor of the corridor are standing facilities offered for liquidity provision and for liquidity absorption, respectively. For different countries, however, the widths of such corridors differ. Chile has the narrowest corridor while Mexico has the widest. When defining the corridor width, central banks weigh two opposing objectives. On the one hand, a wider corridor relies more on liquidity forecasting and open market operations to stabilize short-term rates. On the other hand, a narrower corridor provides a stricter automatic control over short-term market rates, but if too narrow, may inhibit interbank trading.

A. Lower Macro-financial Vulnerabilities

The global crisis found the LA5 countries with relatively low macroeconomic vulnerabilities. By the early 2000s, the LA5 countries had already achieved low and stable inflation and had kept fiscal and external imbalances in check (see Appendix 3). Taking advantage of a commodity price boom and easy global financial conditions during 2003-07, the LA5 took the opportunity to reduce macroeconomic vulnerabilities. Their current account balances improved and they were able to build fiscal and external buffers that helped mitigate the negative effects of exogenous shocks. They also further improved public debt profiles and built up international reserves.

Foreign reserve coverage increased in most of LA5 over the decade before the global financial crisis. As a percent of GDP, international reserves increased substantially in Brazil and Peru, slightly in Mexico, and remained stable in Colombia. Although reserve coverage almost halved in Chile, it still stood at high levels. The increase in Chilean sovereign wealth funds provided extra insurance against a sudden stop or other sudden need of external liquidity.¹⁸ Methods to build up reserves varied, with some countries using methods that limit the impact on exchange rate flexibility, including through auctions for options (Mexico and Colombia) and pre-announcing the intended reserve buildup (Chile).¹⁹

The capital flow composition improved and external financing needs decreased. Foreign direct investment dominated capital flows, and debt flows declined sharply, reducing the share of external debt and annual external financing needs. In particular, median external financing needs halved as a percent of GDP in the decade before the global financial crisis—to 6.8 percent of GDP in 2007 from about 15 percent of GDP in 1996.

Fiscal frameworks were significantly strengthened. All five countries adopted fiscal responsibility laws. Chile and Mexico also put in place a fiscal rule and Colombia has announced its intent to introduce one. These fiscal frameworks reflected a political will that helped reduce fiscal vulnerabilities. The public debt composition improved markedly, with major drops in the share of foreign currency-denominated debt and public debt issued abroad. Over the last decade, the behavior of public debt levels as a percent of GDP varied. Although the gross public debt to GDP ratio increased in Brazil, it fell significantly in Chile and Peru and moderately in Colombia and Mexico.²⁰ In all countries, however, the public debt burden denominated in foreign currency and issued abroad fell substantially as a percent of GDP. The public external debt to GDP also fell slightly in all countries (See Appendix 3).

Prudential regulation and supervision were markedly upgraded. The improvements in financial soundness indicators noted earlier reflected not only good times from a favorable external environment but also, and more importantly, significant reforms in prudential regulation and banking supervision. These reforms imposed some prudence on financial institutions for managing the good times. Colombia and Peru also adopted measures to smooth credit growth along the cycle by adopting countercyclical loan-loss provisions. The LA5 made significant progress in the compliance with Basel core principles, although there is still room for improvement.²¹ Most countries also took steps toward a gradual adoption of the Basel II Accord.

The soundness of the banking system increased. As opposed to several advanced countries, financial institutions in the LA5 countries were not exposed to toxic assets, at least in systemic amounts. In general, capital adequacy ratios largely exceeded the 8 percent minimum of the Basel Accord, averaging about 16 percent at the time of the Lehman events. Nonperforming loans had been declining to below 5 percent of total loans in the context of rapid credit growth. Bank profitability had also been increasing during the decade of the 2000s (Figure 1). The degree of financial dollarization in Peru, the only dollarized economy in the group, also fell substantially over the period, even after adjusting for the effect of currency appreciation.²²

View Full Size

LA5 Financial Soundness Indicators

(2000–2009)

Citation: IMF Working Papers 2010, 292; 10.5089/9781455211739.001.A001

Sources: National authorities; and IMF staff calculations.1/ The official definition of soundness indicators vary by country. The solid lines indicate the median of the indicators across the countries in the group, while the shaded area contains the second to the fourth quintiles.

• Download Figure
• Download figure as PowerPoint slide

The better monetary, fiscal, and prudential frameworks were themselves a source of strength and lower macroeconomic vulnerabilities. Monetary policy frameworks had anchored expectations and granted the central bank more credibility. The more flexible exchange rate regimes made a significant difference, reducing the macroeconomic and financial effects of the real exchange rate appreciations. They helped avoid Krugman-type balance of payments crises when capital inflows suddenly reversed. Moreover, production and consumption suffered less than they would have if interest rates had needed to increase or remain high to defend the nominal exchange rate under a peg regime. In addition, the anchored inflation expectations allowed the public sector to increase the share of its debt denominated in local currency, which reduced sovereign exposures to exchange rate risk. Unlike in past episodes of financial turmoil, the stronger frameworks, lower vulnerabilities, and some good luck allowed the region to escape this time currency, banking, and/or public debt crises.²³

B. Coping with the Surge in Commodity and Food Prices

Before the recent global financial crisis exploded, the LA5 countries were facing rising inflation, as did the rest of the world. The inflationary pressures reflected a combination of rising costs and domestic demand. The cost-push associated with the surge in food and oil prices coincided with easy external financial conditions and favorable terms of trade that increased domestic demand, appreciated the domestic currency, and led to sharp increases in credit growth. Commodity and energy prices had been rising in world markets since the early 2000s, but they swelled in 2007 up to the first half of 2008. With fuel and food items accounting for a high percent of CPI baskets in LA5 countries and with rising demand, inflation started to increase, in particular in Chile and Colombia—where inflation rose to close to 10 percent and 8 percent, respectively, at the peak.

LA5 central banks reacted—with a short lag—by tightening monetary policy. The central bank of Colombia was the first to tighten and kept a policy of raising the policy rate (by 400 basis points) that had already started on 2006Q2. The Central Bank of Chile reacted swiftly since the beginning of 2007 raising the policy rate (by 325 basis points up to 2008Q3). Central banks in Peru, Brazil, and, in particular, Mexico increased their policy rate more gradually, but quickened the increases during 2008Q2 and 2008Q3 (Table 4). These policy rate hikes were transmitted to bank rates, which increased in real terms (Figures 2, 3, and 4). Initially, most central banks expected that the supply shock would be temporary and would not significantly affect core inflation. However, as the supply shocks became persistent, second-round effects started to kick in, thereby increasing core inflation (Figures 2 and 5). In some countries demand pressures also played a role as capacity constraints were binding (Brazil and Colombia) and credit was growing fast (Peru).²⁴ On average, the increase in the policy rate started earlier in the LA5 than in other countries in Latin America (Figure 6).

View Full Size

Inflation and Policy Rates in the Wake of the Supply and Financial Shocks

(Annual percentage rate)

Citation: IMF Working Papers 2010, 292; 10.5089/9781455211739.001.A001

Source: LA5 central banks’ websites.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Monetary Policy Transmission, Policy and Bank Interest Rates

(Annual percentage rate)

Citation: IMF Working Papers 2010, 292; 10.5089/9781455211739.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Monetary Policy Transmission, Policy and Real Bank Interest Rates²⁵

(Annual percentage rate)

Citation: IMF Working Papers 2010, 292; 10.5089/9781455211739.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Commodity Prices and Consumer Price Indices

(CPI sub-indices of consumer prices expressed in U.S. Dollars, SA, September 2008=100)

Citation: IMF Working Papers 2010, 292; 10.5089/9781455211739.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Inflation and Central Banks’ Interest Rates in Latin America

(Percentage rate)

Citation: IMF Working Papers 2010, 292; 10.5089/9781455211739.001.A001

Source: LA5 central banks’ websites.

• Download Figure
• Download figure as PowerPoint slide

Nominal exchange rates appreciated significantly in most LA5 countries as capital flowed in. Between January 2007 and mid-2008, most domestic currencies appreciated—by more than 30 percent—, in particular in Brazil and Colombia. The appreciation helped to moderate imported inflation from the upsurge in worldwide food and energy prices. The policy interest rate hikes helped further contain the inflationary pressures, but at the same time encouraged short-term capital inflows and nominal appreciation. All central banks, except Mexico, intervened in the foreign exchange market. Some did it as part of a strategy of building up international reserves that would serve as a buffer to cope with a likely weakening of the external environment. Others to reduce volatility in the foreign exchange market. Some bought foreign exchange for both purposes.

To help contain capital inflows, some LA5 countries introduced capital controls and adjusted reserve requirements. The central bank of Peru raised reserve requirements on both domestic-and foreign-currency liabilities, and differentiated, with a much higher reserve requirement coefficient, the deposits of nonresidents denominated in domestic currency. In Colombia, besides raising reserve requirements in domestic currency, the central bank imposed unremunerated reserve requirements (URR) on external borrowing and portfolio inflows. The increases in reserve requirements in domestic currency helped tighten monetary policy and slow down credit growth (Box 2). In Brazil, a presidential decree levied a 1.5 percent tax on capital inflows channeled to investment in fixed-income markets.

LA5 central banks reiterated their commitment to price stability and eventually achieved better results than other countries in Latin America. Central banks explained to the markets that raising interest rates was needed because core inflation had started to surge and inflation expectations to diverge from the inflation target tolerance band (in Chile, Colombia, and Peru). Compared to other countries in Latin America and the Caribbean, including the group of officially dollarized countries and others with a currency board in place, the LA5 were able, on average, to better contain inflation pressures (Figure 7).

View Full Size

Inflation in Latin America and the Caribbean

(Classified by exchange rate regimes. Annual percentage rate)

Citation: IMF Working Papers 2010, 292; 10.5089/9781455211739.001.A001

Source: Annual Report on Exchange Arrangements and Exchange Restrictions, IMF.

• Download Figure
• Download figure as PowerPoint slide

C. Dealing with the Global Financial Crisis Following the Lehman Collapse

The Lehman Brothers’ collapse was a turning point toward worldwide disinflation and recession. This financial turmoil had many elements of earlier global financial crises that in the past fueled domestic currency and even banking crises.²⁶ The old sudden stop and reversal of capital inflows associated with global deleveraging reappeared, leading all domestic currencies to come under pressure, reversing the previous exchange rate appreciation dilemma. The worldwide recession also led exports to tumble throughout the region. For instance, exports declined in the last quarter of 2008—with respect to the previous quarter—by about 20 percent in Brazil and Mexico.

Unconventional Measures to Ease Liquidity

LA5 central banks responded to the global financial crisis first by providing liquidity in domestic and foreign currency. In contrast to central banks in advanced economies, the LA5 did not reduce policy rates up front and rarely resorted to credit easing or quantitative easing.²⁷ Instead, they adopted liquidity easing measures in both domestic and foreign currencies to alleviate the stress associated to the sudden tightening of external financial conditions, compounded by capital outflows, and reduce the risk of a liquidity crisis.²⁸

To provide domestic currency liquidity, most LA5 countries loosened access to central bank facilities for liquidity provision. Central banks broadened collateral (Chile and Mexico) and expanded maturities (Chile and Peru) for repos. Peru also set up foreign exchange swaps to inject liquidity in domestic currency. The Bank of Mexico purchased securities (issued by the Institute for the Protection of Banking Savings), mainly to address liquidity problems in mutual funds.

The central banks in Brazil, Colombia, and Peru also cut reserve requirements. Brazil had ample room to ease reserve requirements given the prevailing high rates—compared to international standards—and used it also to achieve other specific objectives.²⁹ Colombia and Peru mostly reversed the increase in reserve requirements adopted in early- and mid-2008 after experiencing heavy capital inflows.

Public banks in Brazil, Chile, and Mexico arguably played a countercyclical role. They either increased the growth rate of its credit to the private sector or decreased it at a markedly lower rate than in private banks, especially foreign owned ones. In Brazil, this was explicitly part of the countercyclical policy response of the government to the crisis (Barbosa, 2010). In Chile and Mexico, the markedly contrarian behavior of credit to firms and households from these institutions suggests a similar approach.

Central banks were able to provide FX liquidity for a variety of reasons. First, they were able to build up an international reserve buffer during the period of favorable external financial conditions and terms of trade. Second, they followed good policies that gave them access to the IMF’s flexible credit line (Colombia and Mexico), which strongly signaled to the markets the IMF’s confidence in their fundamentals and support for their policies. Finally, they had access to a U.S. 30 billion currency swap with the U.S. Federal Reserve Bank (Brazil and Mexico).

To provide FX liquidity, LA5 central banks created mechanisms to lend foreign exchange and reversed earlier measures to contract liquidity. These mechanisms included foreign exchange swaps (Brazil and Chile). The Central Bank of Brazil also auctioned collateralized loans to banks to help finance exports and roll over external debts. Peru sold certificates indexed to the U.S. dollar but settled in domestic currency. The extensive use of alternatives to direct dollar spot sales helped preserve international reserves. To help foreign exchange liquidity easing, Brazil, Colombia and Peru also relaxed the measures affecting investments by nonresidents that had been imposed in 2007 and in the first half of 2008. In addition, Peru cut reserve requirements for dollar deposits, and Chile allowed the use of pesos to fulfill reserve requirements on dollar deposits. Some LA5 central banks also adopted new mechanisms for providing hedge against the depreciation of domestic currencies, including through derivatives (Appendix 4).

These liquidity easing measures expanded central banks’ balance sheets. Although central banks’ balance sheets grew in LA5 countries due to these measures, the increases were not as large as in industrial countries (except Japan), in particular, the United States and the United Kingdom (Figure 8).³⁰

View Full Size

Central Bank Assets in LA5 and Selected Industrial Countries

(Index numbers. January 2006 = 100)

Citation: IMF Working Papers 2010, 292; 10.5089/9781455211739.001.A001

Source: International Financial Statistics for LA5 countries and central bank websites for industrial countries.

• Download Figure
• Download figure as PowerPoint slide

Exchange Rate Response

Foreign exchange markets were the first to dry up as the financial turmoil acquired a global dimension in September 2008. This triggered immediate responses in all LA5 countries, aiming to safeguard the economy from the consequences of excessively sharp swings in the exchange rate and in the cost of external financing.

Domestic currencies were allowed to depreciate as capital inflows started to revert. Brazil and Mexico endured the largest currency depreciation—about 30 percent between August 2008 and February 2009. The Central Bank of Peru only allowed its currency to depreciate by 11 percent. Despite large depreciations, inflation not only remained in check, but also kept falling in all LA5 countries. This partly reflected the worldwide recession and associated decline in inflation around the globe, but also a drastic decline—associated with increased monetary policy credibility—in the pass-through from currency depreciations to inflation.

Yet, some central banks sold significant amounts of foreign exchange (Figure 9). The central bank of Peru sold close to 20 percent of its international reserves over a couple pf months, to smooth out the depreciation trend in an environment of still high financial dollarization. Brazil and Mexico sold almost 10 percent of the international reserves on their books over a couple of months. In these two countries, concerns about large corporations having short dollar positions added to the motivation for avoiding a sudden and severe devaluation of the domestic currency. The Bank of Mexico used three main channels for its sales of foreign exchange: extraordinary auctions in October 2008 (roughly 35 percent of total sales), daily auctions with and without a minimum price (60 percent), and direct sales (5 percent). The Central Bank of Brazil employed non-deliverable currency swaps to supplement direct intervention in the spot foreign exchange market.³¹ Chile and Colombia followed a different path since they placed the burden of the adjustment on the exchange rate and, therefore, sold few international reserves (although Chile made use of its large fiscal funds abroad). Chile stopped buying dollars, which had amounted to US$50 million, daily, since April 2008. Colombia also stopped daily foreign exchange interventions to build international reserves in October 2008, and at the same time reactivated a mechanism of auctioning volatility-triggered options that led to sales of foreign currency as the peso depreciated.³² Despite the large exchange rate movements, inflation remained remarkably stable, suggesting that the monetary frameworks had provided a nominal anchor and reduced the exchange rate pass-through (Figure 10).

View Full Size

LA5 external adjustment after Lehman’s collapse

(International reserves (dots) and exchange rates (lines). (Index numbers))

Citation: IMF Working Papers 2010, 292; 10.5089/9781455211739.001.A001

Source: International reserves (IFS, row 1L), exchange rate index (Insdata, edna).

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Exchange Rate Pass-through, Consumer Price Inflation and Currency Depreciation

(Percent change, 12 months)

Citation: IMF Working Papers 2010, 292; 10.5089/9781455211739.001.A001

• Download Figure
• Download figure as PowerPoint slide

Changes to the policy rate

By early 2009, LA5 central banks started to cut policy rates. The pace of monetary policy loosening varied in each country, depending on inflation projections and each country’s position in the business cycle (Figure 2 and Table 6). The Central Bank of Chile was the most aggressive, slashing its policy rate by 600 basis points during January-March and 775 basis points up to July 2009. In addition, it set up a term liquidity facility to reinforce its commitment to keep the policy rate low for long.³³ At the other end, the Bank of Mexico followed a more conservative approach, because inflation and inflation expectations remained high. The Central Bank of Peru reacted with a short lag reducing its policy rate only 50 basis points in the first quarter, given a slower decline in inflation expectations.³⁴

Table 6.

Policy Rate Cuts by LA5 Central Banks in 2009

(Basis points)

Source: LA5 central banks’ websites

Table 6.

Policy Rate Cuts by LA5 Central Banks in 2009

(Basis points)

	End–2008	2009–Q1	2009–Q2	2009–Q3	2009–Q4	End–2009
Brazil	13.75%	250	200	50	0	8.75%
Chile	8.25%	600	150	25	0	0.50%
Colombia	9.50%	250	250	50	50	3.50%
Mexico	8.25%	150	200	25	0	4.50%
Peru	6.50%	50	300	175	0	1.25%

Source: LA5 central banks’ websites

View Table

Table 6.

Policy Rate Cuts by LA5 Central Banks in 2009

(Basis points)

	End–2008	2009–Q1	2009–Q2	2009–Q3	2009–Q4	End–2009
Brazil	13.75%	250	200	50	0	8.75%
Chile	8.25%	600	150	25	0	0.50%
Colombia	9.50%	250	250	50	50	3.50%
Mexico	8.25%	150	200	25	0	4.50%
Peru	6.50%	50	300	175	0	1.25%

Source: LA5 central banks’ websites

Because policy rates were high before the Lehman event, most LA5 central banks could cut them significantly. In contrast, central banks in the industrial world—where policy rates were already close to the zero bound—did not have this alternative. Most LA5 central banks did not get close to the zero bound. The Central Bank of Brazil stopped reducing the policy rate by July 2009 at 8.75 percent—4 percent in real terms when deflated by the contemporaneous inflation. The Central Bank of Chile was the only one in the LA5 to cut the policy rate to close to zero (0.5 percent), in July 2009.

Money growth, which is endogenous under inflation targeting, initially fell significantly around the Lehman events (Figure 11). The slowdown in economic activity reduced money demand, which contributed to the deceleration in money growth that, in many countries, had started earlier. As uncertainty subsided and economic activity started to recover, the low policy interest rates led to a recovery in money growth in the region, but the pace was not even across countries. As the recovery took hold, credit growth has started to pick up again (Figure 12).

View Full Size

Consumer Prices and Money Growth

(Percent change, 12 months)

Citation: IMF Working Papers 2010, 292; 10.5089/9781455211739.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Monetary Developments and Private Credit Growth

(Percent change, 12 months)

Citation: IMF Working Papers 2010, 292; 10.5089/9781455211739.001.A001

• Download Figure
• Download figure as PowerPoint slide

Box 3.

Panel Vector Autoregression

For each country i, the panel estimates

$\begin{array}{cc}\begin{array}{ll}{\mathit{x}}_{it}=\mathit{\alpha }+\mathit{\beta }{\mathit{x}}_{it-1}+{\mathit{u}}_t,& (1)\\ {\mathit{x}}_{it}=\left[\begin{array}{c}\mathit{GROWTH}\_U{S}_t\\ \mathit{RR}\_U{S}_t\\ D{(LZ)}_{it}\\ Y_{it}\\ \mathit{PIE}4-\mathit{PIETA}{R}_{it}\\ R{S}_{it}\end{array}\right],\mathit{\alpha }=\left[\begin{array}{c}{\alpha }^1\\ {\alpha }^2\\ {\alpha }^3\\ {\alpha }^4\\ {\alpha }^5\\ {\alpha }^6\end{array}\right],\mathit{\beta }=\left[\begin{array}{ccc}{\beta }^{16}& \dots & {\beta }^{16}\\ \cdot & & \cdot \\ \cdot & & \cdot \\ \cdot & & \cdot \\ {\beta }^{61}& \dots & {\beta }^{66}\end{array}\right],{\mathit{u}}_t=\left[\begin{array}{c}{u}_t^1\\ u_t^2\\ u_t^3\\ u_t^4\\ u_t^5\\ u_t^6\end{array}\right],\end{array}& \end{array}$

where, for every period t, GROWTH_US_t is the quarterly U.S. GDP growth seasonally adjusted, RR_US_t is the U.S. Fed Funds rate, D(LZ)_it is the changes in the exchange rate against the U.S. dollar in country the i, Y_it is the output gap computed for country i, computed as a latent variable in a country-version of the Global Projection Model (see Carabenciov and others (2008), and Canales-Kriljenko and others (2008), PIE4 – PIETAR4_it is the moving average (4) quarterly inflation in country i less the inflation target range, and RS_it is the monetary policy rate in country i.

Letting i equal to 1… 5, the equation system (1) is stacked for the 5 LA5 countries. The model is estimated using quarterly data for the period of 2002-09, which covers the recent international crises but excludes episodes of individual country crises that occurred prior to 2002. We use the inverse of the Cholesky factor of the residual covariance matrix with a small sample degrees of freedom correction to orthogonalize the impulse responses. In the Cholesky ordering we follow the same sequencing as presented in x_it with $y_{it}^{US}$ as the most exogenous variable and i_it as the most endogenous.

Empirical analysis shows that LA5 policy rates have systematically reacted to changes in the economic environment. This systematic response to key macroeconomic variables during 2002–09 is summarized by a simple panel vector autoregression (Box 3). The impulse responses suggest what one would expect from a well-functioning flexible monetary policy. On average, policy rates have increased (fallen) when inflation exceeded (was less than) its target and GDP was above (below) its potential level (Figure 13). Simple scatter plots between policy interest rates and inflation and GDP growth reinforce the message (Figures 14 and 15). The impulse responses also suggest that at least some policy interest rates have responded to innovations in the exchange rate, which could reflect concerns about the pass-through onto inflation or economic disruptions associated with balance sheet exposures.³⁵ The analysis also revealed a systematic response to growth in the U.S., which probably reflects the influence of the U.S. on LA5 economic activity, especially in Mexico. These impulse responses show that the policy rate responses have been systematic, but they do not identify a policy reaction function, which is beyond the scope of this paper.³⁶

View Full Size

LA5—Policy Reaction Functions^1/

(2002:Q1–2009:Q4)

Citation: IMF Working Papers 2010, 292; 10.5089/9781455211739.001.A001

Source: Fund staff estimations.^1/ Panel response of policy rate (RS) to Cholesky, one standard deviation +/- 2 standard errors of the following variables: US Federal Funds Rate (RR_US), US GDP growth (GROWTH_US), changes in nominal exchange rate (D(LZ)), deviations of inflation from target (PIE4-PIETAR), output gap (Y), and lags policy rate.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

LA5—Policy Rates and Inflation, 2005–2009

Citation: IMF Working Papers 2010, 292; 10.5089/9781455211739.001.A001

Source: National authorities and Fund staff calculations.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

LA5—Policy Rates and GDP Growth

(2005–2009)

Citation: IMF Working Papers 2010, 292; 10.5089/9781455211739.001.A001

Source: National authorities and Fund staff calculations.

• Download Figure
• Download figure as PowerPoint slide

D. Relative Inflation Performance of Flexible Frameworks

The LA5 central banks have successfully passed the tests posed by high commodity and fuel prices, capital outflows, and a global recession. They have allowed the exchange rate to move, while keeping inflation expectations well anchored around their target levels. In particular, 12-month inflation expectations have evolved within the inflation targeting band during most of the period of analysis in all LA5 central banks, even when contemporaneous inflation missed the target band (Figure 16). For instance, during periods of worldwide inflation (2007-2008), inflation expectations were lower than current inflation. This suggests that central banks’ credibility led markets to distinguish between cyclical pattern of short-term inflation and long-run trends, keeping inflation expectations anchored in the long run. Inflation expectations remained well anchored despite sharp movements in the nominal exchange rate, suggesting that the pass-through has fallen significantly compared to previous crises.

View Full Size

LA5 Inflation Rates and Inflation Expectations

(Annual percentage rate)

Citation: IMF Working Papers 2010, 292; 10.5089/9781455211739.001.A001

Source: LA5 central bank websites.

• Download Figure
• Download figure as PowerPoint slide

Although most Latin American countries were able to keep inflation relatively low, inflation volatility was significantly lower in the LA5 countries (Figure 7). Countries with less flexible exchange rate regimes depended more on external developments, which were unfavorable during this global financial crisis. They received more capital flows and faced a more volatile domestic demand (Canales-Kriljenko, forthcoming, and IMF, 2010).⁴¹