## aieelaea

## Source details

**Canonical URL:** [aieelaea](https://www.imf.org/-/media/files/publications/dp/2026/english/aieelaea.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/dp/2026/english/aieelaea.pdf.md)
- [Structured JSON version](/-/media/files/publications/dp/2026/english/aieelaea.pdf.json)

---

### Executive Summary — Overview
- Long-term inflation expectations critically shape macroeconomic outcomes by influencing how quickly inflation responds to shocks and how costly it is for monetary policy to restore price stability.
- The post‑COVID inflation surge acted as a stress test for monetary frameworks across Latin America, many of which shifted toward inflation targeting over the past quarter century.
- This document evaluates:
  - How well long‑term inflation expectations remained anchored during the post‑COVID episode.
  - Drivers of cross‑country and temporal variation in anchoring.
  - How anchoring affects macroeconomic shock transmission.
  - Whether policy actions can strengthen or weaken anchoring.
  - Lessons from historical transitions to new monetary regimes in the region.

### Three key messages
- Long‑term inflation expectations remained more resilient than short‑term expectations during the recent inflation surge.
  - Near‑term expectations rose in line with contemporaneous inflation, while long‑term beliefs generally stayed much closer to monetary policy objectives, even when fiscal policy was expansionary.
  - This resilience reflects a multidecade improvement in anchoring that began after very high inflation in the 1970s, 1980s, and 1990s.
- Strong anchoring yields material macroeconomic benefits.
  - With well‑anchored expectations, adverse external shocks (for example, declines in the terms of trade) produce much smaller and shorter‑lived increases in inflation; economies with weak anchoring see more pronounced price effects.
  - For monetary policy shocks, evidence is nuanced: strongly anchored expectations are associated with smaller real activity responses to interest rate surprises, consistent with a less sensitive expectations channel.
- Credibility responds asymmetrically to policy behavior.
  - Unexpectedly tight policy produces only small and delayed gains in anchoring.
  - Unexpectedly loose policy leads to sizable and rapid deterioration in anchoring.
  - Case studies show coherent institutional support, consistent communication, and well‑sequenced operational reforms are essential for successful transitions to modern monetary frameworks; absent these, transitions can fail even under adverse external conditions.

### Headline statistic
- 4.5 percent at the end of 2023. YoY = year over year.

### Main findings and empirical patterns
- During the post‑COVID surge, longer‑term inflation expectations in Latin America generally remained remarkably well anchored and compared favorably to other EMDEs, where de‑anchoring was more significant.
- The distribution of long‑term expectations in Latin America is asymmetric and skewed toward the upside; by contrast, expectations in AEs are weakly skewed to the downside.
- Improvements in anchoring over the past 20 years were driven by improvements in both the upper and lower tails of the distribution—a pattern shared with other EMDEs.
- The distribution of beliefs is tight but generally off target:
  - Beliefs about future inflation in Latin America are generally less dispersed than in other EMDEs and similarly dispersed to those in AEs.
  - Average beliefs are further from target than in AEs, consistent with capable policymakers operating in imperfectly credible policy frameworks.
- Better anchoring mitigates the inflationary cost of terms‑of‑trade shocks and, more ambiguously, dampens real‑activity responses to monetary policy shocks.
- Weak anchoring reflects common beliefs across forecasters rather than idiosyncratic disagreement, suggesting underlying institutional constraints (rather than individual incompetence) may limit central banks’ ability to hit targets.
- The persistent slight positive skew in long‑term expectations could reflect lingering fiscal concerns or psychological scars from past high‑inflation episodes.

### Definition and measurement of anchoring
- Anchoring: how tightly private sector long‑term inflation forecasts are pinned to the central bank’s target over time.
- Main measurement: Bems and others (2021) index, summarizing:
  - how close long‑term expectations are to the monetary authority’s target;
  - how stable long‑term expectations are over time;
  - the extent of agreement across professional forecasters.
- Data source: Consensus Economics’ professional forecasters’ CPI inflation forecasts at three‑ and five‑year‑ahead horizons.
- Construction details:
  - For each country and survey date observe mean long‑term forecast and cross‑sectional dispersion.
  - Index constructed by standardizing and averaging three metrics: (1) root mean square deviation of long‑term forecasts from the inflation target, (2) time‑series standard deviation of forecasts, and (3) cross‑sectional dispersion among forecasters.
  - Higher values indicate stronger anchoring.
  - Coverage runs from the 2000s until 2025 for Argentina, Chile, Brazil, Colombia, Mexico, and Peru.
- Alternative measures noted but not used: model‑based forward‑looking weights, market breakeven volatility and level, and shock‑response measures; Consensus forecasts correlate well with alternative measures at many horizons.

### Quantitative indicators and series coverage (figures and axes)
- Anchoring index panel ranges and ticks cited: 0.15, 0.1, 0.05, 0, −0.05, −0.1 for long sample.
- Three‑year‑horizon anchoring index axis labels cited: 0.13, 0.125, 0.12, 0.115, 0.11, 0.105 across 2017–25.
- Figure coverage years cited explicitly: 2000–25 and 2017–25 in anchoring index figures; distributions and decomposition panels include 2005–24 or 2005–25 ranges as shown.

### Cross‑country patterns and notable exceptions
- LA5 = Latin America 5 (Brazil, Chile, Colombia, Mexico, and Peru).
- Country highlights:
  - Chile and Peru: professional forecasters’ long‑term expectations barely moved in response to higher contemporaneous inflation; medians stayed around the target or, for Peru, the target range.
  - Mexico and Brazil: modest upward shifts in longer‑term forecasts, mostly contained.
  - Argentina: after abandoning the currency board in 2001, lack of movement toward sustained IT and flexible exchange rate, fiscal dominance, erosion of monetary policy independence, return of price and wage indexation produced significantly less anchored expectations than the rest of the region; some measures improved after recent reforms.

### Distributional analysis and decompositions
- Distribution reconstruction: three‑year‑ahead expectations reconstructed using beta distributions matched to mean, standard deviation, minimum, maximum from Consensus Economics.
- Decompositions:
  - Anchoring Up versus Down: in LAC, de‑anchoring is asymmetric and more strongly driven by expectations of above‑target inflation than below‑target inflation.
  - Common versus Idiosyncratic components:
    - Decomposition into common component (root mean square error relative to target) and idiosyncratic component (standard deviation).
    - In LAC, the common component contributes more to de‑anchoring than the idiosyncratic one; forecasters generally agree on departures from target.
    - Forecaster disagreement in LAC is small and similar to AEs.
    - Persistent higher common‑component deviations suggest institutional credibility gaps relative to AEs.

### How anchoring affects shock transmission — empirical evidence
- Definitions:
  - Strong anchoring: observations above the 75th percentile of the anchoring index.
  - Weak anchoring: observations below the 25th percentile of the anchoring index.

A. Terms‑of‑Trade Shocks
- Method: local projections estimating cumulative impulse response of CPI to a decrease in the Gruss and Kebhaj (2019) terms‑of‑trade index over a 12‑month horizon; specification includes interactions with anchoring index and flexible exchange rate dummy; 12 lags of inflation and 2 lags of the terms‑of‑trade shock; country and time fixed effects.
- Panel includes: Argentina, Brazil, Bulgaria, Chile, Colombia, Croatia, Hungary, India, Indonesia, Malaysia, Mexico, Peru, Philippines, Poland, Romania, Russia, Thailand, Türkiye, and Ukraine.
- Period: January 1998 to June 2025.
- Key empirical result:
  - A 1 percent decrease in a country’s terms of trade leads to a gradual rise in the CPI, reaching an increase of about 0.2 percent after one year on average.
- Heterogeneity:
  - Inflation increases when terms of trade worsen are concentrated in countries with weakly anchored expectations (below the 25th percentile).
  - Strongly anchored countries (above the 75th percentile) show muted inflation responses.
  - The difference between strongly and weakly anchored responses is statistically significant.

B. Monetary Policy Shocks
- Method: monetary policy shocks from Checo, Grigoli, and Sandri (2024) (constructed from forecast errors and purged of macro information); local projections with country and time fixed effects, flexible exchange rate dummy, interactions with anchoring indicator, 12 lags of inflation, and 2 lags of MP shocks; h = 1,...,12.
- LA5 main sample: Brazil, Chile, Colombia, Mexico, and Peru; monthly data from 2002 to 2022.
- Shocks standardized to one standard deviation.
- Findings for LA5:
  - Tighter monetary policy reduces real activity and increases bond yields.
  - Magnitudes of these responses are smaller when inflation expectations are strongly anchored; differences across anchoring levels are statistically significant.
  - Consumer price responses in the LA5 to a monetary policy shock are statistically insignificant in general; small sample yields a price puzzle (prices rise after monetary tightening).
- Expanded EMDE evidence:
  - Monetary tightening induces a statistically significant drop in inflation over three years when inflation expectations are well anchored, and a more muted price response when expectations are weakly anchored.
  - Implication: better anchoring improves the sacrifice ratio for monetary policy.

C. Asymmetry and expectations channel (Annex 2)
- Tightening of monetary policy reduces economic activity regardless of anchoring level.
- Negative monetary policy shock (loosening) has an insignificant effect on economic activity when inflation is strongly anchored but a weakly stimulative effect when inflation is weakly anchored.
- When long‑term expectations are weakly anchored, monetary loosening has a persistently more expansionary effect than when expectations are well anchored.
- Response of 12‑month‑ahead CPI expectations:
  - If expectations are strongly anchored, tightening raises inflation expectations over a one‑year horizon (information channel).
  - If expectations are weakly anchored, tightening causes expectations to fall (expectations channel).
  - Differences across anchoring levels are statistically significant.

### Policy conduct, reputation channel, and anchoring dynamics
- Question: Can routine policy actions improve long‑term anchoring?
- Method: local projections with change in anchoring index as dependent variable and aggregated annual monetary policy shocks (from Checo, Grigoli, and Sandri (2024)) as exogenous; annual data for 2000–22 for LA5 countries.
- Findings:
  - Positive (more hawkish) policy surprises lead to improvements in anchoring; effects are small, delayed, and weakly significant.
  - Negative (more dovish) policy surprises lead to much larger deteriorations in anchoring.
  - Quantification: the anchoring variable increased for the LA5 countries after the global financial crisis of 2007–09 by 0.01 per year on average. A negative monetary policy surprise of one standard deviation produces approximately the same deterioration in anchoring after one year — i.e., it wipes out around one year’s worth of gains in anchoring, on average.
  - Interpretation: credibility is hard to gain but easy to lose; negative surprises have disproportionate adverse effects on anchoring.

### Monetary regime transitions — narrative case evidence (selected)
- Conceptual framework for transitions classified by:
  1. Monetary policy regime (nominal anchor).
  2. Pace and nature of the transition (gradual/rapid; orderly/disorderly).
  3. Underlying institutional capacity and initial conditions.

Chile (1989–99): Gradual, orderly transition to IT
- BCCh gained legal independence in late 1980s; gradual disinflation strategy.
- 1995–1999: reoriented operations to interbank interest rate; 1999 abandoned exchange rate band and formally adopted full IT with point target of 3 percent and a symmetric tolerance band.
- Outcomes:
  - Inflation fell from above 20 percent in the late 1980s to low single digits by the early 2000s.
  - Long‑term inflation expectations in Chile barely budged from the 3 percent target during the post‑COVID period.
- Lessons:
  - Gradual, transparent transitions supported by operational modernization and fiscal surpluses anchored expectations and allowed exchange rate flexibility to act as shock absorber.

Brazil (1999): Rapid shift to IT
- Crawling band unraveled January 1999; float and formal IT in June 1999.
- Initial central targets: 8 percent (1999), 6 percent (2000), and 4 percent (2001) with ±2 percentage point tolerance bands.
- Institutional features: quarterly inflation reports, COPOM minutes, open‑letter accountability mechanisms.
- Outcomes: rapid adoption succeeded with statutory transparency and clear operational tools; policy rule reacted strongly to forecast inflation.

Argentina (2016–18 and 2023–25): hybrid and transitionary approaches
- 2016–18 IT attempt failed amid fiscal deficits, fiscal dominance concerns, dollarization, and elevated inflation.
- December 2023 stabilization: sharp devaluation and monthly devaluation schedule of 2 percent per month as part of broader plan; combined exchange rate adjustment, tight monetary policy with control over monetary aggregates, and deep fiscal consolidation.
- Monthly inflation fell from over 20 percent in late 2023 to below 3 percent by late 2024 and early 2025.
- By April 2025 policy moved toward greater FX flexibility with a widening band and a monetary aggregate targeting framework.
- Lessons: pragmatic hybrid regimes can be effective as bridges from very high inflation, but sustainability challenges (e.g., real appreciation undermining competitiveness) can emerge.

Other rapid and sustained transitions
- Uruguay (2002), Jamaica (IT elements introduced 2017; formal IT 2020), Nicaragua (durable crawling peg with gradual slowdown in crawl rates from 5 percent to 0 percent during 2019–2024).

### Cross‑case policy implications
- Appropriate monetary regime is context dependent; in very high inflation episodes, monetary targets, exchange rate pegs, or hybrids may be preferable to immediate IT.
- IT can anchor expectations even when adopted rapidly, conditional on transparency and accountability.
- Broad institutional support—especially fiscal discipline and political backing—is essential for success; IT regimes are unlikely to succeed when fiscal or other policies undermine the central bank’s objective.
- Anchoring is slowly built but quickly lost; avoid surprising markets with extra‑loose policy because de‑anchoring is rapid and sizable; gains from unexpected tightness are modest and slow.

### Data, sample coverage, and methodological notes (Annex highlights)
- Primary data: Consensus Economics Forecast inflation expectations; CPI from Haver Analytics; terms‑of‑trade shocks from Gruss and Kebhaj (2019); exchange rate classification from Ilzetzki, Reinhart, and Rogoff.
- Anchoring index components: (1) deviation from target, (2) variability over time, (3) cross‑forecaster dispersion.
- Distribution reconstruction: beta distributions fitted to mean, standard deviation, minimum, maximum of quarterly forecasts; x‑axis expressed as percentage point deviations from target; annual averages of quarterly distributions used where shown.
- Figure and sample coverage examples (selected):
  - Argentina: Anchoring Index Start 1/1994 End 10/2025 nobs 382.
  - Brazil: CPI Start 1/1994 End 10/2025 nobs 382; Anchoring Index Start 1/1994 End 10/2025 nobs 382.
  - Chile: CPI Start 1/1994 End 10/2025 nobs 382; Anchoring Index Start 1/1994 End 10/2025 nobs 382.
  - Mexico: Anchoring Index Start 1/1994 End 10/2025 nobs 382.
  - Peru: Anchoring Index Start 1/1997 End 10/2025 nobs 346.
  - Panel estimation period for terms‑of‑trade analysis: January 1998 to June 2025.
  - Monthly LA5 monetary policy shock sample: 2002 to 2022.
- Asymmetric shock responses and robustness analyses described in Annex 2 and Annex 3 (cross‑sectional decomposition).

*IMF DEPARTMENTAL PAPERS — Anchoring Inflation Expectations (excerpts provided).*

### Executive Summary.................................................................................................1

### Executive Summary

### Overview
- Long-term inflation expectations critically shape macroeconomic outcomes by influencing how quickly inflation responds to shocks and how costly it is for monetary policy to restore price stability.
- The post‑COVID inflation surge acted as a stress test for monetary frameworks across Latin America, many of which shifted toward inflation targeting over the past quarter century.
- This paper evaluates:
  - How well long‑term inflation expectations remained anchored during the post‑COVID episode.
  - Drivers of cross‑country and temporal variation in anchoring.
  - How anchoring affects macroeconomic shock transmission.
  - Whether policy actions can strengthen or weaken anchoring.
  - Lessons from historical transitions to new monetary regimes in the region.

### Three key messages
- Long‑term inflation expectations remained more resilient than short‑term expectations during the recent inflation surge.
  - Near‑term expectations rose in line with contemporaneous inflation, while long‑term beliefs generally stayed much closer to monetary policy objectives, even when fiscal policy was expansionary.
  - This resilience reflects a multidecade improvement in anchoring that began after very high inflation in the 1970s, 1980s, and 1990s.
- Strong anchoring yields material macroeconomic benefits.
  - With well‑anchored expectations, adverse external shocks (for example, declines in the terms of trade) produce much smaller and shorter‑lived increases in inflation; economies with weak anchoring see more pronounced price effects.
  - For monetary policy shocks, evidence is nuanced: strongly anchored expectations are associated with smaller real activity responses to interest rate surprises, consistent with a less sensitive expectations channel.
- Credibility responds asymmetrically to policy behavior.
  - Unexpectedly tight policy produces only small and delayed gains in anchoring.
  - Unexpectedly loose policy leads to sizable and rapid deterioration in anchoring.
  - Case studies show coherent institutional support, consistent communication, and well‑sequenced operational reforms are essential for successful transitions to modern monetary frameworks; absent these, transitions can fail even under adverse external conditions.

### Main findings and empirical patterns
- During the post‑COVID surge, longer‑term inflation expectations in Latin America generally remained remarkably well anchored and compared favorably to other EMDEs, where de‑anchoring was more significant.
- The distribution of long‑term expectations in Latin America is asymmetric and skewed toward the upside; by contrast, expectations in AEs are weakly skewed to the downside.
- Improvements in anchoring over the past 20 years were driven by improvements in both the upper and lower tails of the distribution—a pattern shared with other EMDEs.
- The distribution of beliefs is tight but generally off target:
  - Beliefs about future inflation in Latin America are generally less dispersed than in other EMDEs and similarly dispersed to those in AEs.
  - Average beliefs are further from target than in AEs, consistent with capable policymakers operating in imperfectly credible policy frameworks.
- Better anchoring mitigates the inflationary cost of terms‑of‑trade shocks and, more ambiguously, dampens real‑activity responses to monetary policy shocks.
- Weak anchoring reflects common beliefs across forecasters rather than idiosyncratic disagreement, suggesting underlying institutional constraints (rather than individual incompetence) may limit central banks’ ability to hit targets.
- The persistent slight positive skew in long‑term expectations could reflect lingering fiscal concerns or psychological scars from past high‑inflation episodes; precise causes remain for future work.
- Historical evidence indicates that formal moves toward more advanced monetary frameworks (for example, inflation targeting) generally coincide with better‑anchored expectations, but institutional support and political conditions matter crucially.

### Context and institutional evolution
- Monetary frameworks in Latin America evolved substantially after the extreme inflation episodes of the 1980s and 1990s.
  - By the early 2000s, the Latin America 5 (Brazil, Chile, Colombia, Mexico, and Peru; LA5) converged on some version of inflation targeting combined with a flexible exchange rate and more independent central banks.
  - Inflation targets typically converged toward 2 to 3 percent with a tolerance band.
  - Frameworks became more transparent with inflation reports, published targets and ranges, clearer reaction functions, and greater willingness to let exchange rates respond to shocks.

### Policy implications and priorities
- Preserve and strengthen institutional credibility, as credibility is the principal constraint on achieving fully anchored expectations.
- Ensure policy coherence across institutions to prevent common‑belief de‑anchoring driven by fiscal or political constraints.
- Maintain frameworks that keep inflation expectations aligned with long‑term objectives in the face of future shocks.
- Recognize the asymmetry in credibility dynamics: avoid surprising markets with extra‑loose policy, because de‑anchoring is rapid and sizable; gains from unexpected tightness are modest and slow.

*Source: Executive Summary, Anchoring Inflation Expectations (IMF Departmental Paper).*

### 4.5 percent at the end of 2023. YoY = year over year.

### aieelaea - 4.5 percent at the end of 2023. YoY = year over year.

### Key headline
- 4.5 percent at the end of 2023. YoY = year over year.

### Main findings on inflation expectations in Latin America (LA5)
- The post-COVID inflation episode was characterized by generally well-anchored public expectations of inflation in Latin America, with some cross-country differentiation.
- Chile and Peru: professional forecasters’ long-term expectations barely moved in response to higher contemporaneous inflation; medians stayed around the target or, for Peru, the target range.
- Mexico and Brazil: modest upward shifts in longer-term forecasts, mostly contained.
- As supply and commodity shocks faded and domestic demand cooled, headline CPI in the LA5 started to fall; in many cases, by 2024–25, headline CPI was back within or near target bands.
- Well-anchored expectations reduced policy trade-offs during the surge; poorly anchored expectations increase the cost of stabilization (larger monetary tightening and output loss) and raise interest-rate premia for governments and firms.

### Definition and measurement of anchoring
- Anchoring defined as how tightly private sector long-term inflation forecasts are pinned to the central bank’s target over time.
- Main measurement used: Bems and others (2021) index, which summarizes three aspects:
  - how close long-term expectations are to the monetary authority’s target;
  - how stable long-term expectations are over time;
  - the extent of agreement across professional forecasters.
- Data source for index: Consensus Economics’ professional forecasters’ CPI inflation forecasts at three- and five-year-ahead horizons.
- Construction details:
  - For each country and survey date observe mean long-term forecast and cross-sectional dispersion.
  - Index constructed by standardizing and averaging three metrics: (1) root mean square deviation of long-term forecasts from the inflation target, (2) time-series standard deviation of forecasts, and (3) cross-sectional dispersion among forecasters.
  - Higher values indicate stronger anchoring.
  - Coverage runs from the 2000s until 2025 for Argentina, Chile, Brazil, Colombia, Mexico, and Peru.
- Alternative measures noted but not used here: model-based forward-looking weights, market breakeven volatility and level, and shock-response measures; Consensus forecasts correlate well with alternative measures at many horizons.

### Quantitative indicators from figures and series coverage
- Anchoring index panels and scales referenced:
  - Panel ranges and ticks include values: 0.15, 0.1, 0.05, 0, −0.05, −0.1 for long sample.
  - Three-year-horizon anchoring index axis labels include: 0.13, 0.125, 0.12, 0.115, 0.11, 0.105 across 2017–25.
- Geographic grouping and label conventions:
  - LA5 = Latin America 5 (Brazil, Chile, Colombia, Mexico, and Peru).
  - Data labels use ISO country codes where applicable.
- Figure coverage years explicitly cited: 2000–25 and 2017–25 in anchoring index figures; distributions and decomposition panels include 2005–24 or 2005–25 ranges as shown.

### Cross-country patterns and notable exceptions
- Improvement in anchoring observed in major Latin American economies, especially during the early 2000s.
- Argentina as counterexample:
  - Convertibility in the 1990s delivered dramatic disinflation from triple digits to low single digits.
  - After abandoning the currency board in 2001, lack of movement toward sustained IT and flexible exchange rate, fiscal dominance, erosion of monetary policy independence, return of price and wage indexation produced significantly less anchored expectations than the rest of the region.
  - Some measures of anchoring have improved after reforms in recent years.

### International context
- EMDEs in central and eastern Europe: rapid move from very high and volatile inflation in the 1990s to relatively well-anchored expectations in the 2000s.
- EMDEs in Asia: more gradual path with credibility building around exchange rate management before transitioning to explicit inflation targets.
- Latin America ultimately embraced similar institutional architectures that deliver price stability.
- Post-COVID: many EMDEs, especially in Latin America, tightened policy quickly as headline inflation rose above target; AEs reacted more slowly but from a position of long-standing well-anchored expectations.

### Distributional analysis of long-term beliefs
- Distributions of three-year-ahead expectations reconstructed using beta distributions matched to summary statistics (mean, standard deviation, minimum, maximum) from Consensus Economics.
- Cross-country and time variation:
  - Chile: distributions consistently narrow and centered on official target.
  - Other LA5 countries: more persistent deviations in center and spread; notable weakening in anchoring after the pandemic in some cases (e.g., Brazil), with partial reversals in recent years.
- Emphasis: anchoring depends on average level, dispersion, and tail behavior of beliefs.

### Decompositions of anchoring and their implications
- Anchoring Up versus Down (tail decomposition):
  - Recompute Bems index separately for beliefs above target and below target using fitted distributions.
  - Key conclusions:
    - In LAC, de-anchoring is asymmetric and more strongly driven by expectations of above-target inflation than below-target inflation.
    - For other EMDEs, asymmetry is even larger between 2008 and 2019.
    - In AEs, de-anchoring is more symmetric, though below-target de-anchoring increased after the post-2012 low-inflation period.
    - In LAC and EMDEs broadly, anchoring improved over time due to both above-target expectations falling and below-target expectations increasing; improvements began with better-anchored beliefs from below and later in the upper tail.
- Common versus Idiosyncratic components:
  - Decompose departure from target into:
    - common component: root mean square error relative to target (captures average departure across forecasters);
    - idiosyncratic component: standard deviation of the distribution (captures disagreement across forecasters).
  - Findings:
    - In LAC, the common component contributes more to de-anchoring than the idiosyncratic one, implying forecasters generally agree on how much future inflation will depart from target.
    - Forecaster disagreement in LAC is remarkably small and similar to AEs.
    - Both components have historically been better in LAC than in other EMDEs, though the gap in the common component has narrowed recently.
    - Persistent higher common-component deviations suggest institutional credibility gaps relative to AEs; idiosyncratic components remaining low are consistent with high-quality individual policymakers with strong personal credibility.
  - Interpretation:
    - Common-component de-anchoring signals generalized lack of faith in monetary policy and the inflation target (institutional credibility).
    - Idiosyncratic de-anchoring signals concerns about individual policymakers’ ability or willingness to respond to shocks (policymaker capability or heterogeneity).

### Data and methodological notes
- Primary data: Consensus Economics Forecast inflation expectations data (professional forecasters).
- Index methodology: follow Bems and others (2021) with standardized and averaged metrics (root mean square deviation from target, time-series standard deviation, cross-sectional dispersion).
- Distributional reconstruction: beta distributions fitted to mean, standard deviation, minimum, and maximum of quarterly forecasts; x-axis expressed as percentage point deviations from target; annual averages of quarterly distributions used where shown.
- Sample coverage specifics: LA5 included Brazil, Chile, Colombia, Mexico, and Peru; country coverage includes Argentina, Chile, Brazil, Colombia, Mexico, and Peru for the long series.

*IMF DEPARTMENTAL PAPERS  •  Anchoring Inflation Expectations*

### 1. Common (Root Mean Square Error)

### aieelaea - 1. Common (Root Mean Square Error)

### How Anchoring Affects Shock Transmission — Overview
- Focus: empirical evidence on how the propagation of macroeconomic shocks depends on the degree of anchoring of inflation expectations.
- Two shock types analyzed: a terms-of-trade shock (exogenous) and a monetary policy shock (policy choice).
- Strong versus weak anchoring defined as observations above the 75th percentile or below the 25th percentile of the anchoring index, respectively.

### A. Terms-of-Trade Shocks — Estimation and Findings
- Methodology:
  - Estimate cumulative impulse response of CPI to a decrease in the Gruss and Kebhaj (2019) terms-of-trade index over a 12-month horizon using local projections.
  - Specification includes interactions between anchoring index, flexible exchange rate dummy, and terms-of-trade shock; 12 lags of inflation and 2 lags of the terms-of-trade shock; country and time fixed effects.
  - Panel of EMDEs includes: Argentina, Brazil, Bulgaria, Chile, Colombia, Croatia, Hungary, India, Indonesia, Malaysia, Mexico, Peru, Philippines, Poland, Romania, Russia, Thailand, Türkiye, and Ukraine.
  - Period: January 1998 to June 2025.
- Key empirical result:
  - A 1 percent decrease in a country’s terms of trade leads to a gradual rise in the CPI, reaching an increase of about 0.2 percent after one year on average.
- Heterogeneity by anchoring:
  - Inflation increases when terms of trade worsen are concentrated in countries with weakly anchored expectations (below the 25th percentile).
  - Strongly anchored countries (above the 75th percentile) show muted inflation responses.
  - The difference between responses for strongly and weakly anchored countries is statistically significant and economically meaningful.
- Figure annotations and inference:
  - Results show cumulative inflation response to a 1 percentage point month-on-month deterioration in commodity terms of trade.
  - Confidence interval indicated at the 90 percent level in the figure note.

### B. Monetary Policy Shocks — Estimation and Findings
- Context and theoretical ambiguity:
  - Better anchoring could (a) strengthen policy by lowering output cost to control inflation if agents expect return to target; or (b) weaken transmission by making expectations less responsive and dampening the expectations channel.
- Methodology:
  - Use monetary policy shocks identified by Checo, Grigoli, and Sandri (2024) (constructed from forecast errors and purged of macro information).
  - Local projections similar to the terms-of-trade specification with country and time fixed effects, flexible exchange rate dummy, interactions with anchoring indicator, 12 lags of inflation, and 2 lags of MP shocks; h = 1,...,12.
- Sample for main exercise:
  - LA5 countries: Brazil, Chile, Colombia, Mexico, and Peru.
  - Monthly data from 2002 to 2022 for LA5 countries.
  - Monetary policy shock is standardized to one standard deviation.
- Empirical findings for LA5:
  - Tighter monetary policy reduces real activity and increases bond yields.
  - Magnitudes of these responses are smaller when inflation expectations are strongly anchored.
  - Differences across anchoring levels are statistically significant and consistent with the importance of the expectations channel.
- Consumer price response:
  - In the LA5, consumer price responses to a monetary policy shock are statistically insignificant in general, and separating by anchoring level yields a price puzzle (prices rise after monetary tightening) in this small sample.
  - Extending the exercise to a larger sample of EMDEs with additional controls: monetary tightening induces a statistically significant drop in inflation over three years when inflation expectations are well anchored, and a more muted price response when expectations are weakly anchored.
  - Implication: better anchoring improves the sacrifice ratio for monetary policy (less output cost per unit of disinflation), per the EMDE evidence.
- Additional notes:
  - The expectations channel and asymmetrical responses to positive versus negative monetary policy shocks are explored further in Annex 2 (summary reference only in source).
  - Short-term inflation expectations decrease following tighter monetary policy when long-term anchoring is weaker, consistent with a stronger expectations channel in weakly anchored settings.

### C. Policy Conduct and Anchoring — Reputation Channel Evidence
- Research question: Can routine policy actions (e.g., “acting tough” with extra-tight policy) improve long-term anchoring of expectations?
- Methodology:
  - Local projections with change in anchoring index as dependent variable and aggregated annual monetary policy shocks (from Checo, Grigoli, and Sandri (2024)) as the exogenous variable.
  - Annual estimation because the anchoring index is slow moving and evolves yearly; monthly shocks aggregated to annual frequency.
  - Country and time fixed effects included; shocks split by sign to distinguish tighter versus looser policy surprises.
  - Annual data for 2000–22 for LA5 countries.
- Empirical findings:
  - Positive (more hawkish) policy surprises lead to improvements in anchoring; effects are small, delayed, and weakly significant.
  - Negative (more dovish) policy surprises lead to much larger deteriorations in anchoring.
  - Quantification: the anchoring variable increased for the LA5 countries after the global financial crisis of 2007–09 by 0.01 per year on average. A negative monetary policy surprise of one standard deviation produces approximately the same deterioration in anchoring after one year — i.e., it wipes out around one year’s worth of gains in anchoring, on average.
  - Interpretation: credibility is hard to gain but easy to lose; negative surprises have disproportionate adverse effects on anchoring.
- Figure annotation:
  - Dark bands at the 68 percent confidence interval and lighter bands at the 90 percent confidence interval in the figure note.
  - Monetary policy shock standardized to one standard deviation.

### Monetary Regime and Anchoring — Narrative Evidence (Latin America)
- Approach:
  - Narrative case-study approach to study regime changes because quantitative methods struggle with rare, diverse, and endogenous regime-change events.
  - Presents three main case studies of monetary regime changes and additional LAC cases to capture motivations, sequencing, and outcomes.
- Conceptual framework for classifying transitions (three dimensions):
  1. Monetary policy regime: nominal anchor (examples: monetary aggregate targeting, IT, exchange rate targeting).
  2. Pace and nature of the transition: gradual or rapid; orderly or disorderly.
  3. Underlying institutional capacity and initial conditions: operational capacity of central bank, political and fiscal constraints, prevailing macroeconomic conditions.
- Illustrative case study — Chile 1989–99 (Gradual and Orderly Transition to IT):
  - Chile’s transition combined phased move to greater exchange rate flexibility, institutional reform, operational modernization, and credibility building.
  - Historical context: prior history of very high inflation, peaking at several hundred percent per year in the mid-1970s.
  - Key steps:
    - Late 1980s: Central Bank of Chile (BCCh) gained legal independence and price stability was established as primary mandate.
    - Chosen strategy: gradual disinflation rather than rapid stabilization.
    - Early/mid-1990s: BCCh used a crawling exchange rate band as nominal anchor while introducing explicit inflation targets (initially multiyear targets published as ranges).
    - Emphasis on transparency, forward guidance, and systematic communication to align private expectations with the disinflation path.

*Italic source attribution: IMF DEPARTMENTAL PAPERS — Anchoring Inflation Expectations (excerpts provided).*

### 1. Chile in the 1990s: A Gradual Path to IT

### 1. Chile in the 1990s: A Gradual Path to IT

### Modernization of monetary operations and institutional change
- Between 1995 and 1999, BCCh progressively reoriented its operating framework toward the interbank interest rate as the main policy instrument—moving away from quantity-based instruments.
- In 1999 Chile abandoned the exchange rate band and allowed full exchange rate flexibility, formally adopting a full-fledged IT regime with a point target of 3 percent and a symmetric tolerance band.
- The ability to let the exchange rate float—without compromising inflation control—reflected how expectations had become adequately anchored.

### Outcomes and transmission
- Inflation fell from above 20 percent in the late 1980s to low single digits by the early 2000s.
- Long-term inflation expectations in Chile barely budged from the 3 percent target during the post-COVID period, even when inflation itself reached double digits.
- Anchored expectations, a credible institutional framework, and a transparent communication strategy allowed the floating exchange rate to act as a shock absorber without undermining the nominal anchor.

### Key lessons from Chile
- A gradual, transparent, and institutionally grounded transition—supported by modernization of operational tools and consistent policy implementation—can anchor expectations and deliver lasting macroeconomic stability even after episodes of deep indexation and past instability.
- A supportive fiscal stance—characterized by sustained budget surpluses throughout the 1990s—ensured disinflation under the transition to IT was not undermined by fiscal imbalances and insulated monetary policy from financing pressures and fiscal dominance.

---

### 2. Brazil in 1999: A Rapid Shift to IT

### Context and policy shift
- The crawling band regime unraveled in January 1999 after contagion from the Russian crisis; Brazil entered the float and formalized an IT regime in June 1999.
- Initial central targets: 8 percent (1999), 6 percent (2000), and 4 percent (2001), with ±2 percentage point tolerance bands.
- The central bank centered operational control on the Selic interbank rate and adopted a forward-looking reaction to inflation expectations.

### Institutional and accountability mechanisms
- Emphasis on transparency and accountability: quarterly inflation reports, timely COPOM minutes, and an open-letter mechanism requiring the BCB Governor to explain misses, corrective measures, and the expected horizon back to target.
- Over time, tolerance-band adjustments: from ±2.5 percentage points starting in 2003 to ±1.5 percentage points starting in 2017.
- Policy framework evolution: moved from calendar-year targets to a rolling, long-term framework with a continuous 3 percent target starting in January 2025.

### Outcomes and interpretation
- Despite initial absence of deep FX markets and other preconditions, targets helped stabilize expectations; the policy rule reacted strongly to forecast inflation, leading to lower inflation and output volatility relative to pre-IT periods.
- Brazil’s experience shows rapid formal adoption of IT can succeed if coupled with robust statutory transparency, clear operational tools, and credible communication of trade-offs when shocks hit.

---

### 3. Argentina 2016–18 and 2023–25: Hybrid and Transitionary Approaches

### 2016–18 attempt and failure
- Argentina’s abortive IT attempt in 2016–18 failed amid persistent fiscal deficits, fiscal dominance concerns, high dollarization, and elevated inflation and inflation expectations.
- Debt monetization and perceived fiscal dominance conflicted with the high real interest rates needed to achieve overly optimistic disinflation targets, producing a credibility shock in late 2017 and a framework replacement in October 2018 by a monetary targeting rule.

### 2023–25 stabilization: hybrid regime and outcomes
- In December 2023 authorities enacted a sharp devaluation and announced a monthly devaluation schedule of 2 percent per month (a “crawling peg”) as part of a broader macroeconomic stabilization plan.
- The strategy combined exchange rate adjustment, tight monetary policy with strict control over monetary aggregates and liquidity, and government-backed fiscal consolidation including deep spending cuts and reduction of quasi-fiscal deficits.
- Monthly inflation, which had surged to over 20 percent in late 2023, began to fall steeply to below 3 percent by late 2024 and early 2025.
- By April 2025, the official policy stance moved toward greater FX flexibility—allowing the peso to float within a large widening band—combined with a monetary aggregate targeting framework.

### Lessons from Argentina 2023–25
- A pragmatic hybrid regime blending exchange rate adjustment, tight monetary and fiscal discipline, and aggregate-money management can be useful as a bridge for countries starting from a very high inflation environment where direct move to IT is complicated.
- Such frameworks can be challenging to sustain long term: by early 2025 sustained real appreciation began undermining competitiveness and reversing gains in current account balancing and FX reserve accumulation.

---

### 4. Other rapid and sustained transitions

### Uruguay, Jamaica, and Nicaragua
- Uruguay (2002): exit from a crawling band amid a crisis; operationally relied on the monetary base before shifting to the policy rate in 2007 and used a hybrid framework.
- Jamaica: began introducing IT elements as early as 2017 and formally adopted IT in 2020 concurrent with legislative reform strengthening central bank independence; relied on rules-based FX auctions, surrender requirements, and strong front‑loaded monetary tightening during 2021–23.
- Nicaragua’s crawling peg (begun 1993) is highly durable:
  - Rate of crawl adjusted: reduced from 5 percent to 3 percent in 2019, then to 2 percent in 2020, 1 percent in 2023, and 0 percent in 2024.
  - The de jure crawling peg is in practice equivalent to a stabilized arrangement, supported by ample reserve coverage and fiscal discipline.

---

### 5. Cross-case lessons and policy implications

### Main conclusions from narrative evidence
- The appropriate choice of monetary regime is highly context dependent; in episodes of exceptionally high inflation (high double digits or above), regimes such as monetary targets, exchange rate pegs, or hybrid arrangements may be better suited than formal IT.
- IT can anchor inflation expectations even when adopted rapidly in challenging environments; transparency and accountability are crucial for building confidence.
- Broad institutional support—particularly fiscal discipline and political backing—is essential for success; IT regimes are unlikely to succeed when fiscal or other policies undermine the central bank’s objective.

### Empirical and policy insights
- Better anchoring of inflation expectations helps mitigate the impact of external shocks: the pass-through of terms-of-trade shocks in well-anchored countries is statistically indistinguishable from zero.
- Effects of anchoring on monetary transmission are mixed: when anchoring is weaker, interest rates remain higher for longer and output declines more when policy tightens; stronger anchoring can moderate transmission but provides greater macroeconomic stability.
- Anchoring is slowly built but quickly lost; anchoring responds asymmetrically to unexpected monetary tightening and loosening, underscoring the need to safeguard central bank credibility.
- The success of IT as a long-term nominal anchor is confirmed, but wider institutional support beyond monetary policy—especially fiscal discipline and political backing—is critical for transition and sustained success.

*IMF Departmental Papers — Anchoring Inflation Expectations (chapter excerpt: "Chile in the 1990s: A Gradual Path to IT")*

### Annex 1. Extra Charts and Tables

### Annex 1. Extra Charts and Tables

### Measurement and Composition of the Anchoring Index
- Anchoring index constructed from three metrics: (1) deviation from target, (2) variability over time, and (3) cross-forecaster dispersion.
- Data labels in figures use International Organization for Standardization (ISO) country codes.
- Source for figures: Consensus Economics Forecast inflation expectations data.

### Inflation Anchoring by Country Group
- Country groups and inclusions:
  - LAC includes Brazil, Chile, Colombia, Mexico, and Peru.
  - Advanced economies include Australia, Austria, Belgium, Canada, Czech  Republic,  Denmark,  Estonia,  Finland,  France,  Germany,  Greece,  Hong  Kong  Special  Administrative  Region  (SAR),  Ireland,  Italy,  Japan,  Korea,  Latvia,  Lithuania, The Netherlands, New Zealand, Norway, Portugal, Singapore, Slovak Republic, Slovenia, Spain, Sweden, Switzerland, Taiwan Province of China, the United Kingdom, and the United States.
  - Europe EMDEs include Bulgaria, Croatia, Hungary, Poland, Romania, Russia, Türkiye, and Ukraine.
  - Asia EMDEs include China, India, Indonesia, Malaysia, Philippines, and Thailand.
- Metrics constructed following Bems and others (2021) methodology using Consensus Economics Forecast inflation expectations data.
- The anchoring index is a standardized variable. LAC = Latin America and the Caribbean; EMDEs = emerging market and developing economies.

### Country-level Measurement Composition (Example: Argentina)
- Anchoring index and its three component measurements displayed for Argentina (figure).
- Note: Anchoring index based on three metrics: (1) deviation from target, (2) variability over time, and (3) cross-forecaster dispersion.

### Data Availability (Annex Table 1.1)
- Source: Haver Analytics (CPI); Consensus Economics and staff calculations (Anchoring Index); Gruss and Kebhaj (2019) (Terms-of-Trade shocks); Ilzetzki, Reinhart, and Rogoff exchange rate classification (Exchange rate regime).
- Note: CPI = consumer price index. nobs = number of observations.

- Argentina: CPI Start 1/1997 End 10/2025 nobs 346; Anchoring Index Start 1/1994 End 10/2025 nobs 382; Flex Start 1/1994 End 11/2015 nobs 251; Shock Start 2/1994 End 12/2021 nobs 335
- Brazil: CPI Start 1/1994 End 10/2025 nobs 382; Anchoring Index Start 1/1994 End 10/2025 nobs 382; Flex Start 7/1994 End 10/2025 nobs 369; Shock Start 2/1994 End 12/2021 nobs 335
- Bulgaria: CPI Start 1/1994 End 10/2025 nobs 382; Anchoring Index Start 1/2007 End 10/2025 nobs 226; Flex Start 1/1997 End 10/2025 nobs 346; Shock Start 2/1994 End 12/2021 nobs 335
- Chile: CPI Start 1/1994 End 10/2025 nobs 382; Anchoring Index Start 1/1994 End 10/2025 nobs 382; Flex Start 1/1994 End 10/2025 nobs 382; Shock Start 2/1994 End 12/2021 nobs 335
- Colombia: CPI Start 1/1994 End 10/2025 nobs 382; Anchoring Index Start 1/1997 End 10/2025 nobs 346; Flex Start 1/1994 End 10/2025 nobs 382; Shock Start 2/1994 End 12/2021 nobs 335
- Croatia: CPI Start 1/1994 End 10/2025 nobs 382; Anchoring Index Start 1/2007 End 10/2025 nobs 226; Flex Start 10/1994 End 10/2025 nobs 373; Shock Start 2/1994 End 12/2021 nobs 335
- Hungary: CPI Start 1/1994 End 10/2025 nobs 382; Anchoring Index Start 1/1998 End 10/2025 nobs 334; Flex Start 1/1994 End 10/2025 nobs 382; Shock Start 2/1994 End 12/2021 nobs 335
- India: CPI Start 1/1994 End 10/2025 nobs 382; Anchoring Index Start 1/1995 End 10/2025 nobs 370; Flex Start 1/1994 End 10/2025 nobs 382; Shock Start 2/1994 End 12/2021 nobs 335
- Indonesia: CPI Start 1/1994 End 10/2025 nobs 382; Anchoring Index Start 1/1995 End 10/2025 nobs 370; Flex Start 1/1994 End 10/2025 nobs 361; Shock Start 2/1994 End 12/2021 nobs 335
- Malaysia: CPI Start 1/1994 End 3/2018 nobs 291; Anchoring Index Start 1/1995 End 10/2025 nobs 370; Flex Start 1/1994 End 10/2025 nobs 382; Shock Start 2/1994 End 12/2021 nobs 335
- Mexico: CPI Start 1/1994 End 10/2025 nobs 382; Anchoring Index Start 1/1994 End 10/2025 nobs 382; Flex Start 1/1994 End 10/2025 nobs 367; Shock Start 2/1994 End 12/2021 nobs 335
- Peru: CPI Start 1/1994 End 10/2025 nobs 382; Anchoring Index Start 1/1997 End 10/2025 nobs 346; Flex Start 1/1994 End 10/2025 nobs 382; Shock Start 2/1994 End 12/2021 nobs 335
- Philippines: CPI Start 1/1994 End 10/2025 nobs 382; Anchoring Index Start 1/2009 End 10/2025 nobs 202; Flex Start 1/1994 End 10/2025 nobs 376; Shock Start 2/1994 End 12/2021 nobs 335
- Poland: CPI Start 1/1994 End 10/2025 nobs 382; Anchoring Index Start 1/1998 End 10/2025 nobs 334; Flex Start 6/1995 End 10/2025 nobs 365; Shock Start 2/1994 End 12/2021 nobs 335
- Romania: CPI Start 1/1994 End 10/2025 nobs 382; Anchoring Index Start 1/1998 End 10/2025 nobs 334; Flex Start 4/2001 End 10/2025 nobs 295; Shock Start 2/1994 End 12/2021 nobs 335
- Russia: CPI Start 1/1994 End 10/2025 nobs 382; Anchoring Index Start 1/1998 End 10/2025 nobs 334; Flex Start 12/1999 End 10/2025 nobs 295; Shock Start 2/1994 End 12/2021 nobs 335
- Thailand: CPI Start 1/1994 End 10/2025 nobs 382; Anchoring Index Start 1/1995 End 10/2025 nobs 370; Flex Start 1/1994 End 10/2025 nobs 376; Shock Start 2/1994 End 12/2021 nobs 335
- Türkiye: CPI Start 1/1994 End 10/2025 nobs 382; Anchoring Index Start 1/1998 End 10/2025 nobs 334; Flex Start 2/1998 End 7/2018 nobs 220; Shock Start 2/1994 End 12/2021 nobs 335
- Ukraine: CPI Start 1/1994 End 12/2019 nobs 312; Anchoring Index Start 1/1998 End 12/2019 nobs 264; Flex Start 9/1998 End 1/2014 nobs 185; Shock Start 2/1994 End 12/2019 nobs 311

### Asymmetric Responses to Monetary Policy Shocks (Annex 2)
- Research question: whether economic response to monetary policy shocks is asymmetric across positive (tightening) and negative (loosening) shocks and whether this depends on the degree of inflation anchoring.
- Sample: Latin America sample for comparability with main paper; monthly data from 2002 to 2022 for LA5 countries in many analyses.
- Key findings:
  - A tightening of monetary policy reduces economic activity regardless of anchoring level.
  - A negative monetary policy shock (loosening) has an insignificant effect on economic activity when inflation is strongly anchored but a weakly stimulative effect when inflation is weakly anchored.
  - The difference in response from a negative monetary policy shock on economic activity between strongly and weakly anchored cases is significantly different from zero.
  - Negative (loosening) monetary shocks have smaller output responses than positive ones.
  - When long-term expectations are weakly anchored, monetary loosening has a persistently more expansionary effect than when expectations are well anchored.
  - Tighter monetary policy has similar effects whether expectations are strongly or weakly anchored.
- Possible mechanism:
  - Information channel of monetary policy (Jarociński and Karadi 2020) could explain asymmetry: credible central banks (strongly anchored) may signal a larger slowdown when loosening, prompting downward revision of private output expectations; less credible banks (weakly anchored) may have a weaker information channel and standard interest-rate transmission may dominate.
  - If the information channel were the principal driver, symmetric effects would be expected for tightening as well; no evidence found for that symmetry, warranting further research.

### Response of Economic Activity to Positive and Negative MP Shocks (Figure evidence)
- Figures report impulse responses with Ninety percent confidence interval.
- MP shock standardized to one standard deviation.
- Strongly (weakly) anchored corresponds to the 75th (25th) percentile of the anchoring index.
- LA5 = Latin America 5 (Brazil, Chile, Colombia, Mexico, and Peru).

### Response of Inflation Expectations to Monetary Policy Shocks
- Dependent variable: 12-month-ahead year-over-year consumer price inflation expectations from central bank surveys (Chile uses 11-month-ahead).
- Sample: LA5 countries; monthly data from 2002 to 2022.
- Key findings:
  - If inflation expectations are strongly anchored, a tightening of monetary policy raises inflation expectations over a one-year horizon — evidence of analysts learning from central bank actions (information channel).
  - When inflation expectations are weakly anchored, a similar tightening causes inflation expectations to fall — consistent with the expectations channel of monetary policy.
  - The difference in response across anchoring levels is statistically significant.
- Figures: Ninety percent confidence interval; monetary policy shock standardized to one standard deviation.

### Response of Prices to Monetary Policy Shocks
- Focus: impact of monetary policy shocks on domestic consumer prices conditional on anchoring.
- Expanded sample for price response: Brazil, Chile, Colombia, Hungary, India, Indonesia, Malaysia, Mexico, Peru, Philippines, Poland, Romania, Russia, Thailand, and Türkiye (monthly data from 2002 to 2022).
- Empirical specification: dependent variable is the log of the consumer price index; ΔC_i,t is a vector containing consumer price inflation, the short-term bond yield, the domestic exchange rate relative to the US dollar, industrial production, and the unemployment rate. Six months of lagged controls included. Equation specified for h = 1,...,6 with interaction terms involving anchor_i,t-1, flex_i,t, MP_i,t and their interactions.
- Key finding:
  - Inflation falls more following a contractionary monetary policy shock if inflation expectations are well anchored.
  - The decrease in inflation is statistically significant between years 2 and 3 for the well-anchored case.
  - Although confidence intervals for weakly and strongly anchored series overlap, the difference between them is statistically significant — evidence of potential gains from anchoring inflation expectations.
- Figures: Ninety percent confidence interval; monetary policy shock standardized to one standard deviation. Strongly (weakly) anchored corresponds to the 75th (25th) percentile of the anchoring index.

### Cross-Sectional Decomposition (Annex 3)
- Decomposition used for cross-sectional inflation expectations:
  - 1/T ∑_{t=1}^T 1/N ∑_{i=1}^N (Π_it − Π_t^*)^2 = 1/T ∑_{t=1}^T 1/N ∑_{i=1}^N (Π_it − Π_t)^2 + 1/T ∑_{t=1}^T (Π_t − Π_t^*)^2
  - Left-hand side: mean squared deviation of inflation expectations relative to the target averaged over period T — summary measure of inflation anchoring.
  - Right-hand side: first term captures dispersion of individual forecasts; second term reflects deviation of inflation forecasts from the target.
- Notation: Π_it denotes inflation expectation of individual i at time t; Π_t (bar) is average inflation expectation; Π_t^* is the inflation target.

*Source: IMF Departmental Papers — Annex 1. Extra Charts and Tables.*

---


_Source: https://www.imf.org/-/media/files/publications/dp/2026/english/aieelaea.pdf_