## smpfea

## Source details

**Canonical URL:** [smpfea](https://www.imf.org/-/media/files/publications/dp/2023/english/smpfea.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/dp/2023/english/smpfea.pdf.md)
- [Structured JSON version](/-/media/files/publications/dp/2023/english/smpfea.pdf.json)

---

### Executive Summary — key context and recent inflation dynamics
- Amid persistent post-COVID-19 inflation and spillovers from Russia’s war in Ukraine, CCA countries faced strong price pressures; inflation is estimated to have peaked in early 2023 but still exceeds central bank targets.
- Core inflation remains high because of second-round effects, surges in global energy and food prices, and domestic demand pressures.
- Annual CPI inflation levels "ranging between 8.3 percent in Armenia to 20.7 percent in Kazakhstan as of January 2023" remain relatively high compared to pre-pandemic levels and country-specific targets.
- Uncertainty and downside risks: regional tensions, financial turmoil related to international monetary policy normalization, and a growth slowdown in key trading partners.

### Progress in monetary frameworks and institutional issues
- Legal frameworks strengthened since the 1990s; broad de jure central bank independence established.
- Countries transitioning to inflation-targeting regimes: Armenia, Georgia, Kazakhstan, the Kyrgyz Republic, and Uzbekistan.
- Countries relying on the exchange rate as an operational target: Azerbaijan, Tajikistan, and Turkmenistan.
- Most CCA central banks place price stability as the primary objective and have numerical inflation targets, but anchoring inflation expectations remains challenging.
- De facto independence can be compromised where governments influence board members or governance; de jure flexible exchange rates often differ from de facto arrangements.

### Four key challenges hampering monetary policy design and implementation
- Structural weaknesses impair transmission:
  - Weak bank competition, absence of a well-established benchmark yield curve, high dollarization, small and illiquid capital markets, and limited cross-border capital mobility.
  - Result: transmission from the interest rate to inflation is relatively weak in outer years; transmission from the exchange rate to inflation remains relatively strong.
  - Policy implication: foster financial market development, deepen financial inclusion, expand role of financial markets (including developing a benchmark yield curve), and reduce dollarization.
- Exchange rate management and FX interventions (FXIs):
  - In some de jure flexible regimes, central banks frequently intervene to smooth exchange rate fluctuations.
  - Empirical analysis for Armenia, Georgia, and the Kyrgyz Republic indicates central banks “lean against the wind” by selling (purchasing) foreign currency in periods of exchange rate depreciation (appreciation).
  - The “leaning against the wind” is asymmetric—leaning more against depreciation—but does not deplete FX reserves below prudent levels.
  - More symmetric FXIs would help bolster FX reserves, allow exchange rates to move in both directions, and help deepen FX markets.
- High financial dollarization:
  - Dollarization remains high despite a downward trend in recent years attributed to prudent macroeconomic policies, macroprudential measures, and the move to floating exchange rates.
  - Survey responses: most CCA countries have an official de-dollarization policy.
  - All countries apply higher reserve requirement ratios and additional capital requirements for FX-denominated liabilities relative to domestic currency liabilities.
  - Further scope exists to reduce dollarization through stronger macroeconomic policy frameworks, macroprudential measures, central bank credibility, and development of financial markets and institutions.
- Nascent central bank communication and credibility:
  - Central banks increasingly use social media in addition to announcements and press releases; press releases are more detailed, predictable, and transparent.
  - Credibility challenges persist; sharing more and better information would help anchor inflation expectations.

### Empirical and analytical contributions
- Draws on original surveys of country authorities and IMF country teams and on the methodology of Unsal, Papageorgiou, and Garbers (2022).
- Presents novel empirical work on:
  - Monetary policy transmission in the CCA (Annex 1).
  - The link between foreign exchange interventions and exchange rate dynamics (Annex 2).
  - Drivers of financial de-dollarization.
  - Effects of central bank communication in the CCA.

### Policy recommendations and next steps (tailored)
- Strengthen legal and governance frameworks:
  - Expand operational independence where needed; ensure majority of board directors are nonexecutive; set stricter rules for subsidized lending and other quasi-fiscal mandates.
- Improve operational frameworks:
  - Improve liquidity forecasting; coordinate cash-flow projections with government; remove interest rate caps for policy instruments; introduce intraday and overdraft credit facilities.
- Enhance transparency, communication, and financial literacy:
  - Disclose more and better information on policy measures, including FX operations; improve financial literacy; adopt forward-looking communication explaining monetary and FX decisions relative to projected inflation and medium-term targets.
- Address structural impediments complementary to monetary policy:
  - Coordinate central bank efforts with government policies to address institutional and structural deficiencies; strengthen governance and credibility of public institutions; promote economic diversification.
- Implementation caveats:
  - Calibrate tools depending on shocks, country characteristics, and initial conditions; coordinated policy action essential as central banks alone cannot address broader institutional deficiencies.

### 1. Energy Exporters — frameworks, evidence, and implications
- Operational frameworks after independence:
  - Monetary-targeting: operational target = growth rate of reserve money; intermediate target = growth rate of broad money; instrument = open market operations. Drawbacks: limited by financial market development, fiscal dominance, unstable money multipliers and velocity.
  - Managed exchange rate: operational target = exchange rate vs. US dollar or basket; instrument = FXIs; sometimes a band used. Requires ample FX reserves, prudent fiscal policy, subdued capital flow shocks, and aligned business cycles; risks include inability to fully shield external shocks and step devaluations.
- Evidence on volatility of monetary aggregates (standard deviation of growth rates):
  - Money multiplier:
    - Armenia = 10.1
    - Kazakhstan = 15.7
    - Azerbaijan = 16.8
    - Kyrgyz Republic = 6.8
    - Georgia = 10.6
    - Tajikistan = 12.4
  - Money velocity:
    - Armenia = 7.9
    - Kyrgyz Republic = 11.7
    - Georgia = 11.1
    - Tajikistan = 15.4
    - Azerbaijan = 7.9
    - Uzbekistan = 8.6
    - Kazakhstan = 12.0
- Transition to inflation targeting: countries have started transition but effectiveness questioned amid supply-side disruptions; IMF (2015) best-practice principles reiterated verbatim.
- FXI transparency: only Armenia, Georgia, and the Kyrgyz Republic publish daily FXI data; others publish only monthly aggregates.
- Macroprudential tools: capital and liquidity buffers, differentiated reserve requirements, loan-to-value and debt-to-service caps, limits on open FX positions; recommendation to expand tools to nonbank financial institutions.
- Structural environment: larger agriculture share, higher food/imported goods in consumption, low diversification, high foreign-currency debt, weak productivity, high dollarization, procyclical current accounts—making economies vulnerable to real shocks.
- Inflation dynamics:
  - Spikes during crises, particularly in oil-importing economies; spikes coincide with large exchange rate depreciations and high pass-through via imported consumption goods and supply shocks.
  - Central banks typically accommodate temporary first-round effects; second-round effects may require tighter policy.
- Financial sector indicators: nonperforming loans contained in most countries but with spikes; conservative capital adequacy ratios; credit-to-GDP increased but weak financial development impairs transmission.
- Mobile money: deepening financial inclusion, reducing cash demand, shifting monetary aggregate composition and increasing velocity.

### 3. Challenges in Monetary Policy — transmission, FXIs, dollarization
- Design and implementation:
  - Five CCA countries announced transition to inflation targeting: Armenia, Georgia, Kazakhstan, the Kyrgyz Republic, and Uzbekistan.
  - FPAS rollout supported by IMF capacity development; structural factors complicate forecasting and policy effectiveness.
- A. Weak monetary policy transmission — channels and empirical findings:
  - Channels: interest rate, bank lending, balance sheet, asset price, exchange rate, expectations.
  - Structural impediments: excess liquidity parked at central bank, weak bank competition, absence of benchmark yield curve, illiquid stock/derivatives markets, imperfect capital mobility, fear of floating, importance of supply-side shocks.
  - Empirical findings (panel VAR and IRFs):
    - A 1 percent shock to the real policy rate results in a 0.5 percent decline in the rate of inflation in the first year.
    - The impact of a 1 percent policy rate shock dissipates and becomes insignificant in outer years.
    - The policy rate shock does not have a significant impact on output at a 95 percent confidence level (negative and significant at 90 percent).
    - A 1 percent exchange rate depreciation results in a 0.3 percent increase in the rate of inflation in the first year.
    - Exchange rate pass-through to inflation is 30 percent.
    - Money supply does not have a significant impact on inflation and output growth; standard deviation of money velocity growth ranges between 6.9 percent in Azerbaijan and 15.6 percent in Tajikistan (2005–20); money multiplier growth ranges between 8.9 percent in Armenia and 17 percent in Azerbaijan (2005–20).
    - When exchange rate is set exogenous in panel VAR, response of inflation to interest rate shock is 0.22 percent (less than half of baseline), consistent with exchange rate appreciation amplifying interest rate hikes.
  - Overall: transmission to output muted; interest-rate-to-inflation effect short-lived; exchange rate channel plays greater role.
- B. Exchange rate management and FXIs — empirical micro-evidence:
  - FXIs common across CCA, including countries with inflation-targeting-like frameworks; exchange rate vs. Russian ruble relevant in some intervention decisions.
  - FXIs respond to exchange rate movements and volatility; central banks “lean against the wind”:
    - FX sales (purchases) took place for 23 (30) percent of weeks in sample for Armenia.
    - FX sales (purchases) took place for 17 (19) percent of weeks for Georgia.
    - FX sales (purchases) took place for 29 (10) percent of weeks for the Kyrgyz Republic.
    - FX sales more likely when exchange rate depreciates against its 12-week moving average; FX purchases more likely when exchange rate appreciates.
    - Leaning is asymmetric toward leaning against depreciation but has not led to deterioration of reserve adequacy metrics.
- C. High dollarization:
  - Origins: instability, hyperinflation, and large depreciations in the 1990s.
  - Recent trends:
    - A decade ago, share of FX deposits in total deposits averaged 60 percent across the CCA (constant exchange rates); ranged from 34 percent in Armenia to about 80 percent in Georgia.
    - Deposit dollarization fell to 38 percent in 2022; Georgia highest at 57 percent in 2022.
    - Credit dollarization declined from 62 percent in 2010 to 28 percent in 2022 (CCA average).
    - Kazakhstan’s credit dollarization fell from 74 percent in 2010 to single digits in 2022.
    - Gap between deposit and credit dollarization narrowed due to regulatory limits on net open FX positions.
  - De-dollarization drivers: prudent macro policies, macroprudential measures, floating exchange rates; policy scope remains (credible frameworks, market development, regulatory measures).
  - Directionality: deposit dollarization is driving credit dollarization.

### 1. Dollarization of Deposits — regulation and empirical drivers
- Measurement: deposit and credit dollarization computed at constant exchange rates (method described in source).
- Reserve requirement differentials cited as policy tools; examples:
  - Azerbaijan: 0.5 percent (reserve requirement differential)
  - Georgia: close to 20 percent (reserve requirement differential)
- Short-term drivers (empirical VAR results):
  1. Increasing spread between reserve requirement ratios on foreign currency and local currency has contributed to deposit and credit de-dollarization in several countries.
  2. Prudential measures (restrictions on FX lending to unhedged borrowers, higher provisions on FX loans, differentiated capital risk weights, tighter capital requirements against FX positions) contributed to credit de-dollarization in several countries.
  3. Extending yield curve and domestic debt market development not associated with sufficient de-dollarization, possibly due to low market development.
  4. Effects of exchange rate volatility and inflation on dollarization are mixed.
- Financial stability context: negative net international investment positions above 70 percent of GDP in several CCA countries exacerbate contractionary impact of FX depreciation on balance sheets.

### Communication, transparency, and press releases ("Length")
- Reading easiness measured by Flesch index; analysis performed on English press releases.
- Overall sentiment of press releases in the region is negative; more negative tone increases interest rate pass-through via signaling.
- Central bankers’ views (Joint Vienna Institute survey, June 2021):
  - Monetary policy reports and press releases seen as most relevant communication tools; social media ranked lower.
  - All surveyed officials reported central bank communications impact interest rates, inflation expectations, and exchange rate.
  - Four of six central banks consider impact on inflation expectations as high; five of six consider impact on retail interest rates as low.
- Progress on IMF (2015) best-practice principles: improvements noted but gaps remain (board selection/dismissal, operational independence in practice, weak liquidity forecasting, limited transparency on FXIs).
- National Bank of Kazakhstan numerical inflation target example:
  - 2021–22: 4–6 percent
  - 2023–24: 4–5 percent
  - From 2025 onwards: 3–4 percent

### Policy recommendations (implementation considerations emphasized)
- Continue modernization guided by IMF best-practice principles; adapt outside inflation-targeting regimes where appropriate.
- Strengthen legal frameworks to close governance gaps and ensure operational independence in practice.
- Improve transmission by:
  - Reducing cross-purpose instruments (simultaneous FXIs and subsidized lending).
  - Enhancing liquidity forecasting and government transaction coordination.
  - Deepening financial markets, improving inclusion, developing a benchmark yield curve, and reducing dollarization.
- Advance de-dollarization through credible macro frameworks, absence of fiscal dominance, deep domestic markets, and regulatory measures where credit dollarization exceeds deposit dollarization; recognize de-dollarization is gradual and requires prolonged low and stable inflation.
- Gradual move to greater exchange rate flexibility to cushion shocks and preserve competitiveness; FXIs can smooth transitions but require clarity and transparency.
- Use the Integrated Policy Framework (IPF) to select policy mixes that jointly consider monetary policy, FXIs, and macroprudential regulation; weigh short-term benefits of FXIs and macroprudential measures against potential adverse side effects.

### Annex 1 — empirical analysis of monetary policy transmission (methodology and key statistics)
- Baseline: panel VAR on annual data for seven CCA countries (Armenia, Azerbaijan, Georgia, Kazakhstan, Kyrgyz Republic, Tajikistan, Uzbekistan) for 2000–20.
- Endogenous variables: real GDP growth (y), CPI inflation (p), changes in real central bank policy rate (r), real base money growth (m), nominal local currency per US dollar exchange rate growth (e). Two lags; country and time fixed effects; Cholesky ordering Y = [y, p, r, m, e].
- Granger causality p-values:
  - Effect on output growth (y):
    - Policy rate (r): 0.306
    - Money supply (m): 0.374
    - Exchange rate (e): 0.151
    - Jointly (p, r, m, e): 0.005***
  - Effect on inflation (p):
    - Policy rate (r): 0.000***
    - Money supply (m): 0.771
    - Exchange rate (e): 0.017**
    - Jointly (y, r, m, e): 0.000***
- Variance decomposition (share of forecast error variation explained by impulse variables, percent):
  - Output growth (y):
    - 1 year: y 0.974, p 0.013, r 0.000, m 0.004, e 0.008
    - 2 years: y 0.963, p 0.012, r 0.002, m 0.006, e 0.017
    - 3 years: y 0.932, p 0.041, r 0.002, m 0.007, e 0.018
    - 4 years: y 0.921, p 0.052, r 0.003, m 0.006, e 0.018
    - 5 years: y 0.917, p 0.054, r 0.003, m 0.007, e 0.019
  - Inflation (p):
    - 1 year: y 0.002, p 0.893, r 0.031, m 0.002, e 0.072
    - 2 years: y 0.002, p 0.832, r 0.036, m 0.017, e 0.114
    - 3 years: y 0.003, p 0.817, r 0.044, m 0.017, e 0.119
    - 4 years: y 0.003, p 0.817, r 0.045, m 0.016, e 0.118
    - 5 years: y 0.004, p 0.813, r 0.048, m 0.017, e 0.118
- Notable quantitative conclusions:
  - Share of inflation variation explained by the exchange rate reaches 12 percent in the fifth year (0.118), more than two times higher than variation explained by the policy rate (5 percent; 0.048) in the same horizon.
  - Only 2 percent of inflation variation is explained by monetary base (reported as 0.017 or 0.002 depending on horizon; narrative notes "Only 2 percent").
  - For output growth, policy variables explain a small fraction of variation.
- Impulse responses:
  - Monetary policy shocks mostly affect nominal variables (inflation) rather than real variables (real GDP growth).
  - A 1 percent shock to the real policy rate reduces inflation by 0.5 percent in the first year.
  - A 1 percent exchange rate depreciation increases inflation by 0.3 percent in the first year.
  - Money supply shocks do not have a significant impact on inflation.
  - Country-level monthly VARs (excluding output) support dominance of exchange rate channel; interest rate and money shocks largely insignificant at country level.

### Annex Table 2.1 and ordered logit evidence on FX interventions
- Ordered logit model (dependent variable: FX purchases = −1, no interventions = 0, FX sales = 1) — maximum likelihood estimation; robust standard errors reported.
- Coefficients for deviation of exchange rate from its 12-week moving average (percent):
  - Armenia: 0.399*** (0.063)
  - Georgia: 0.162*** (0.032)
  - Kyrgyz Republic: 0.280*** (0.058)
- Coefficients for 12-week standard deviation of exchange rate changes (percent):
  - Armenia: −0.386** (0 .16 4)
  - Georgia: 0 .13 8 (0.145)
  - Kyrgyz Republic: −0.282** (0 .137)
- Model sample and fit:
  - Observations: Armenia 74, Georgia 3579, Kyrgyz Republic 534
  - Log-likelihood: Armenia −728 .9, Georgia −485.5, Kyrgyz Republic −444.3
  - Pseudo Rsq: Armenia 0.0587, Georgia 0.0274, Kyrgyz Republic 0.0439
  - Significance notation: *** p < 0.01, ** p < 0.05, * p < 0 .1.
- Event study ("leaning against the wind") — estimated magnitudes:
  - FX sales occur in weeks when exchange rate depreciates relative to normal periods by about:
    - Armenia: 0.4 percent
    - Georgia: 1 percent
    - Kyrgyz Republic: 0.2 percent
  - Following FX sales, relative depreciation comes down and is not significantly different from zero by the third week in all countries.
  - FX purchases occur in weeks when exchange rate appreciates relative to normal periods by about:
    - Armenia: 0.3 percent
    - Georgia: 0.45 percent
    - Kyrgyz Republic: 0.4 percent
  - Following FX purchases, relative appreciation comes down and is not significantly different from zero by the third week in all countries.
- Interpretation: estimation results support evidence of asymmetric “leaning against the wind” interventions aimed at smoothing exchange rate fluctuations rather than maintaining an overvalued exchange rate.

*Source: smpfea - Executive Summary; 1. Energy Exporters; 3. Challenges in Monetary Policy; Annex 1; Annex Table 2.1 (IMF Departmental Paper).*

### Executive Summary

### smpfea - Executive Summary

### Key context and recent inflation dynamics
- Amid persistent post-COVID-19 inflation and spillovers from Russia’s war in Ukraine, CCA countries faced strong price pressures in recent years; inflation is estimated to have peaked in early 2023 but still exceeds central bank targets.
- Core inflation remains stubbornly high due to second-round effects, surges in global energy and food prices, and domestic demand pressures.
- Annual CPI inflation levels "ranging between 8.3 percent in Armenia to 20.7 percent in Kazakhstan as of January 2023" remain relatively high compared to pre-pandemic levels and country-specific inflation targets.
- Uncertainty and downside risks include regional tensions, financial turmoil related to international monetary policy normalization, and a growth slowdown in key trading partners.

### Progress in monetary frameworks
- Since independence in the 1990s, considerable progress has been made: legal frameworks strengthened and broad de jure central bank independence established.
- Countries transitioning to inflation-targeting regimes: Armenia, Georgia, Kazakhstan, the Kyrgyz Republic, and Uzbekistan.
- Countries relying on the exchange rate as an operational target: Azerbaijan, Tajikistan, and Turkmenistan.
- Most CCA central banks place price stability as the primary objective and have numerical inflation targets, but anchoring inflation expectations remains challenging.

### Four key challenges hampering monetary policy design and implementation
- Structural weaknesses impair monetary policy transmission:
  - Weak bank competition, absence of a well-established benchmark yield curve, high dollarization, small and illiquid capital markets, and limited cross-border capital mobility.
  - Result: transmission from the interest rate to inflation is relatively weak in outer years; transmission from the exchange rate to inflation remains relatively strong.
  - Policy implication: strengthening transmission requires fostering financial market development, deepening financial inclusion, expanding the role of financial markets (including developing a benchmark yield curve), and reducing dollarization.
- Exchange rate management and FX interventions (FXIs):
  - In some de jure flexible regimes, central banks frequently intervene to smooth exchange rate fluctuations.
  - Empirical analysis for Armenia, Georgia, and the Kyrgyz Republic indicates central banks “lean against the wind” by selling (purchasing) foreign currency in periods of exchange rate depreciation (appreciation).
  - The “leaning against the wind” is asymmetric—leaning more against depreciation—but does not deplete FX reserves below prudent levels.
  - More symmetric FXIs would help bolster FX reserves, allow exchange rates to move in both directions, and help deepen FX markets.
- High financial dollarization:
  - Dollarization remains high despite a downward trend in recent years attributed to prudent macroeconomic policies, macroprudential measures, and the move to floating exchange rates.
  - Survey responses: most CCA countries have an official de-dollarization policy.
  - All countries apply higher reserve requirement ratios and additional capital requirements for FX-denominated liabilities relative to domestic currency liabilities.
  - Further scope exists to reduce dollarization through stronger macroeconomic policy frameworks, macroprudential measures, central bank credibility, and development of financial markets and institutions.
- Nascent central bank communication and credibility:
  - Central banks increasingly use social media in addition to announcements and press releases.
  - Press releases discuss topics consistent with central bank objectives and have become more detailed, predictable, and transparent.
  - Central banks still face credibility challenges; sharing more and better information would help anchor inflation expectations.

### Empirical and analytical contributions of the paper
- Draws on original surveys of country authorities and IMF country teams, and on the methodology of Unsal, Papageorgiou, and Garbers (2022).
- Presents novel empirical work on:
  - Monetary policy transmission in the CCA (Annex 1).
  - The link between foreign exchange interventions and exchange rate dynamics (Annex 2).
  - The drivers of financial de-dollarization.
  - The effects of central bank communication in the CCA.

### Policy recommendations and next steps (tailored to country circumstances)
- Strengthen legal and governance frameworks:
  - Expand operational independence of central banks where needed despite existing de jure mandates.
  - Ensure the majority of board directors are nonexecutive.
  - Set stricter rules for subsidized lending and other quasi-fiscal mandates.
- Improve operational frameworks:
  - Improve liquidity forecasting.
  - Coordinate better cash-flow projections with the government.
  - Remove interest rate caps for policy instruments.
  - Introduce intraday and overdraft credit facilities.
- Enhance transparency, communication, and financial literacy:
  - Disclose more and better information on policy measures, including FX operations.
  - Improve financial literacy among the public.
  - Step up communication efforts and adopt a forward-looking strategy in which central banks explain monetary and FX policy decisions based on projected changes in inflation and their consistency with the medium-term target.
- Address structural impediments complementary to monetary policy:
  - Coordinate central bank efforts with other government policies to address institutional and structural deficiencies.
  - Strengthen governance and credibility of public institutions and promote greater economic diversification to improve monetary transmission and achieve price stability.
  - Sequence and prioritize structural reforms to reflect their positive externalities for monetary policy effectiveness.

### Implementation caveats and policy trade-offs
- Central banks must carefully assess the use of available tools and calibrate policies depending on the nature of shocks, country characteristics, and initial conditions.
- Central banks alone cannot address deficiencies associated with the broader institutional environment; coordinated policy action is essential.

*Source: smpfea - Executive Summary (IMF Departmental Paper)*

### 1. Energy Exporters

### 1. Energy Exporters

### Monetary and exchange rate frameworks: characteristics and weaknesses
- After independence in the early 1990s, CCA countries adopted exchange rate and monetary targeting frameworks to stabilize high inflation.
- Monetary-targeting framework: operational target = growth rate of reserve money; intermediate target = growth rate of broad money; instrument = open market operations. Drawbacks: limited by financial market development, fiscal dominance, unstable money multipliers and velocity as financial markets develop.
- Managed exchange rate framework: operational target = exchange rate vis-à-vis the US dollar or a currency basket; instrument = FX interventions (FXIs); sometimes a band is used. Requires ample FX reserves, prudent fiscal policy, subdued shocks to capital flows, and aligned business cycles. Risks: inability to fully shield external shocks, deviations of the peg from equilibrium, and step devaluations.
- Both frameworks have drawbacks in the CCA context: monetary targeting is prone to structural shifts in money demand; exchange rate pegs are prone to step devaluations. Example: money multipliers and velocity have fluctuated widely over the last two decades.

### Evidence on volatility of monetary aggregates (exact values from source)
- Money multiplier (standard deviation of growth rate):
  - Armenia = 10.1
  - Kazakhstan = 15.7
  - Azerbaijan = 16.8
  - Kyrgyz Republic = 6.8
  - Georgia = 10.6
  - Tajikistan = 12.4
- Money velocity (standard deviation of growth rate):
  - Armenia = 7.9
  - Kyrgyz Republic = 11.7
  - Georgia = 11.1
  - Tajikistan = 15.4
  - Azerbaijan = 7.9
  - Uzbekistan = 8.6
  - Kazakhstan = 12.0

### Transition to inflation targeting: rationale and best-practice principles
- Most CCA countries have started transitioning to inflation targeting; effectiveness questioned after COVID-19 supply-side disruptions.
- IMF (2015) best-practice principles for effective inflation targeting (preserved phrasing):
  - First, central banks have a clear legal mandate and operational independence to pursue it.
  - Second, price stability is the primary objective of monetary policy (at least in the medium term).
  - Third, central banks have a numerical medium-term inflation objective to operationalize the price stability mandate and guide policy actions.
  - Fourth, central banks also consider the implications of policy decisions on output and financial stability.
  - Fifth, central banks have an effective operational framework, generally centered on the controlling short-term interest rates.
  - Sixth, central banks adopt a forward-looking strategy that maps the price stability objective into policy decisions.
  - Finally, clear communication is a central element of the monetary policy framework to explain policy decisions and outcomes and provide guidance about the future.
- The COVID-19 pandemic and supply chain disruptions highlighted limitations of inflation targeting in addressing supply-side inflationary pressures, though the IMF (2015) principles remain universal for modernizing monetary policy frameworks.

### Central bank independence, mandates, and transparency
- Many CCA central banks have strengthened legal frameworks and established de jure independence through reforms and new central bank chapters; de jure independence has improved over time.
- De facto independence can be compromised where governments influence board members or governance.
- Most CCA central banks view price stability as the primary objective and set numerical inflation targets. Some central banks have additional legal mandates (for example, Armenia, Georgia, and Kazakhstan have legal mandates to maintain financial stability).
- Exchange rate management remains central in monetary policy frameworks, including in countries transitioning to inflation targeting. De jure flexible exchange rate arrangements often differ from de facto arrangements showing more rigidity.
- De facto, none of the CCA countries was classified as freely floating and not more than three CCA countries were classified as floating from 2007 onward.
- FXI transparency: only three CCA countries (Armenia, Georgia, and the Kyrgyz Republic) publish daily FXI data; others publish only monthly aggregates.

### Macroprudential policy and financial stability tools
- CCA central banks use macroprudential measures to address systemic risks influenced by commodity prices and remittance flows.
- Macroprudential tools include capital and liquidity buffers, differentiated reserve requirements, loan-to-value and debt-to-service caps, and limits on open FX positions.
- Some central banks are legally mandated to maintain financial stability in addition to price stability.
- The IMF Integrated Policy Framework (IPF) suggests combining monetary and macroprudential instruments, among other tools, depending on country characteristics and initial conditions.
- Recommendation in source: expand macroprudential instruments to nonbank financial institutions to address data and supervisory gaps and minimize leakages.

### Macroeconomic and structural environment shaping monetary policy
- CCA structural features relevant for monetary policy:
  - Larger share of agriculture in value-added.
  - Higher share of food and imported goods in consumption baskets.
  - Low economic diversification; high dependence on commodity exports or remittances.
  - High share of foreign-currency denominated debt and weak labor and capital productivity.
  - High dollarization, limited access to international markets, procyclical current accounts.
- CCA economies are more vulnerable to real shocks—supply-side, policy, and external shocks—as inferred from:
  - Negative correlation between inflation and output.
  - Large standard deviations in government spending.
  - Volatile terms of trade and real effective exchange rates.
- Limited ability to smooth shocks due to underdeveloped financial markets and limited access to international capital markets.

### Inflation dynamics, exchange rate pass-through, and credibility
- Inflation tends to spike during crises, particularly in oil-importing economies; spikes coincide with large exchange rate depreciations and high pass-through to CPI via imported consumption goods and supply shocks (for example, food price shocks).
- Central banks typically accommodate temporary first-round effects; second-round effects may require tighter monetary policy.
- The extent of second-round effects depends on how well inflation expectations are anchored, underpinned by credibility of the monetary policy framework and ability to mitigate destabilizing exchange rate swings.
- The announced transition to inflation targeting by five CCA central banks has not led to a pronounced reduction in inflation rates (Figure 7), possibly due to shocks after announcements or incomplete transitions.

### Financial sector indicators and monetary transmission
- Nonperforming loans (percent of total gross loans): the share has been reasonably contained in most CCA countries over the last decade, but some countries experienced spikes.
- Capital adequacy ratios (capital in percent of risk-weighted assets): CCA countries maintain conservative ratios.
- Financial deepening has improved but markets remain relatively shallow; the credit-to-GDP ratio has increased but weak financial development impairs monetary transmission.
- Other constraints: governance-related structural factors, high informality, poor accounting practices, lack of well-established yield curves on government securities, and limited access to international capital markets.

### Financial inclusion, mobile money, and implications for monetary aggregates
- Increasing use of mobile money services is deepening financial inclusion and reducing demand for cash.
- Mobile money services expand access, especially in rural areas, and do not require bank accounts.
- Switching from cash to mobile money can lead to a structural shift in monetary aggregate composition, weigh on seigniorage revenues, and increase velocity of money.

### Policy implications and recommendations (explicit themes from source)
- Strengthen legal and operational independence of central banks and clarify mandates consistent with IMF best-practice principles.
- Maintain price stability as primary objective while considering output and financial stability implications.
- Adopt numerical medium-term inflation objectives to operationalize price stability mandates.
- Build effective operational frameworks centered on short-term interest rate control and forward-looking strategies that map objectives into policy decisions.
- Improve transparency and communication to guide expectations and explain policy actions and outcomes.
- Strengthen macroprudential frameworks and ensure consistency between macroprudential and monetary policies; expand tools to cover nonbank financial institutions.
- Enhance FXI transparency and clarify FX intervention strategies to improve public understanding of FX policy.
- Deepen domestic financial markets and improve access to international capital markets to strengthen shock smoothing and monetary policy transmission.
- Monitor and address financial stability risks related to international monetary policy normalization and adjust macroprudential measures as needed.

*Italic: Sources: IMF, International Financial Statistics; IMF staff calculations; IMF (2015); IMF (2020); Poghosyan (2022); Khandelwal and others (2022); Atamanchuk and Tokuoka (2023); Shirono and others (2021); Sahay and others (2020); Armas and Singh (2022).*

### 1. Energy Exporters

### 1. Energy Exporters

### Monetary policy credibility and communication
- Absence of well-established inflation expectation surveys makes it difficult for central banks to assess policy effectiveness and communicate the appropriate monetary policy stance.
- Building credibility in this environment is challenging for an inflation-targeting central bank.

### Fiscal considerations affecting monetary policy
- Some central banks are pressured to provide subsidized lending or conduct other quasi-fiscal activities.
- The high share of FX-denominated government debt leaves the fiscal accounts vulnerable to exchange rate depreciations, imposing constraints on executing monetary policy and the FXI strategy.
- Higher market rates resulting from tighter monetary policy increase the public debt service burden, putting pressure on central banks to undertake less tightening than is appropriate.
- Lack of consistency between monetary and fiscal policies complicates the task of achieving price stability.
- Until authorities address these constraints, they will continue to hamper monetary policy effectiveness.

### Capital flows, capital account openness, and the role of FXIs
- Capital flows and capital account openness have substantial benefits, but they can also increase vulnerabilities to external shocks and complicate the conduct of monetary policy.
- Capital inflows can boost development and competition in CCA financial markets.
- Tightening international financial conditions can lead to capital outflows and put pressure on the local currency, prompting central banks to raise policy rates irrespective of inflation rates and risking procyclical tightening.
- The IPF provides an analytical framework to consider the use of FXIs and capital flow management measures to help monetary authorities achieve their objectives in the presence of frictions, such as shallow markets and currency mismatches, provided they do not substitute for warranted macroeconomic policy adjustments or support a misaligned exchange rate to gain unfair competitive advantage (IMF 2020).
- Constant use of FXIs and macroprudential instruments may perpetuate the same frictions that justify their deployment (for instance, policies that aim to stabilize the exchange rate can encourage the accumulation of unhedged FX liabilities and discourage FX market development).
- Policymakers should weigh the short-term benefits of FXIs and macroprudential instruments against their potential costs and side effects.

### Dollarization and currency mismatches
- High dollarization creates trade-offs between price and financial stability, weakening monetary transmission.
- CCA countries have a structurally high reliance on foreign currency borrowing by the private and public sectors.
- Some CCA countries issue Eurobonds and can access international capital markets, but others rely on concessional financing from international development institutions.
- This leaves them vulnerable to destabilizing swings in the exchange rate.
- Amid large currency mismatches, a significant share of the domestic financial system becomes less sensitive to domestic monetary policy, weakening monetary transmission.
- Capital inflows can exacerbate currency mismatches, leaving recipient countries susceptible to local currency depreciations and imposing additional constraints on monetary policy (Poirson and others 2020, Khandelwal and others 2022).
- Financial stability considerations constrain central banks from letting the exchange rate fully cushion external shocks.

### Additional challenges specific to the CCA’s energy exporters
- These countries have ample FX reserves and tend to rely on managed exchange rate regimes, making them vulnerable to large changes in the equilibrium exchange rate following terms of trade shocks.
- Restoring equilibrium requires large movements in the nominal exchange rate, creating difficult trade-offs between price and exchange rate stability.
- Governments are direct and indirect recipients of energy export revenues; terms of trade shocks also affect fiscal positions.
- Ensuring consistency between fiscal and monetary policy actions requires increased efforts in the face of these shocks.

*Source: smpfea - 1. Energy Exporters*

### 3. Challenges in Monetary Policy

### 3. Challenges in Monetary Policy

### Design and Implementation
- Five CCA countries—Armenia, Georgia, Kazakhstan, the Kyrgyz Republic, and Uzbekistan—have announced their transition to inflation targeting.
- Progress has been mixed; announcement of inflation targets has helped facilitate modernization efforts.
- IMF capacity development has supported rolling out the Forecasting and Policy Analysis System (FPAS) to develop a toolkit for forward-looking monetary policy formulation (Mæhle and others 2021).
- Structural factors—frequent missed targets, noisy data, volatile interest and exchange rates, and poor communication—complicate the interpretation of central bank intentions and reduce the effectiveness of standard forecasting models and monetary policy actions.

### A. Weak Monetary Policy Transmission
- Monetary policy transmission channels described:
  - Interest rate channel.
  - Bank lending channel.
  - Balance sheet channel.
  - Asset price channel.
  - Exchange rate channel.
  - Expectations channel (monetary policy credibility is key).
- Structural features in CCA that impair transmission:
  - Structural excess liquidity: mutual distrust among banks, regulatory and institutional shortcomings, remittance inflows, and irregular cash flows from budgetary operations lead banks to park excess liquidity at the central bank at submarket rates; mopping up large volumes of excess liquidity is costly and limits the impact of policy rate changes on lending volumes.
  - Weak bank competition: small number of banks, asset concentration, government-owned banks in some countries, shallow nonbank financial sector, large interest margins, and subsidized lending by some central banks weaken pass-through of policy rates to market rates.
  - Absence of a well-established benchmark yield curve: illiquid secondary government securities markets, shallow securities markets, and securities held until maturity hamper transmission from short-term policy rates to medium- and long-term rates.
  - Small and illiquid stock and derivatives markets: limited ability to mark physical capital to market and to hedge risks, weakening the asset price channel.
  - Imperfect capital mobility: not all CCA countries issue Eurobonds, limited private sector access to international capital markets, weak integration makes interest parity less likely to hold; central banks resort to FXIs to operationalize exchange rate channel.
  - Fear of floating: high exchange rate pass-through and dollarization, high share of foreign-denominated public debt, and political economy pressures lead central banks to restrain exchange rate flexibility and intervene in FX markets.
  - Importance of supply-side shocks: high share of imported goods and foods in the CPI basket increases headline inflation volatility and reduces short-run effectiveness of monetary policy; underscores importance of targeting medium-term inflation.
- Empirical findings (authors’ analysis):
  - A 1 percent shock to the real policy rate results in a 0.5 percent decline in the rate of inflation in the first year.
  - The impact of a 1 percent policy rate shock dissipates and becomes insignificant in the outer years.
  - The policy rate shock does not have a significant impact on output at a 95 percent confidence level (footnote: the impact is negative and significant at 90 percent confidence level).
  - A 1 percent exchange rate depreciation results in a 0.3 percent increase in the rate of inflation in the first year.
  - The relatively high exchange rate pass-through to inflation is 30 percent.
  - Money supply does not have a significant impact on inflation and output growth, potentially due to highly volatile money velocity and multiplier:
    - The standard deviation of the money velocity growth ranges between 6.9 percent in Azerbaijan and 15.6 percent in Tajikistan (2005–20).
    - The money multiplier growth ranges between 8.9 percent in Armenia and 17 percent in Azerbaijan (2005–20).
  - When the exchange rate is set as an exogenous variable in a panel VAR (i.e., switching off endogenous exchange rate amplification), the response of inflation to an interest rate shock is 0.22 percent, less than half of the response in the baseline model—consistent with exchange rate appreciation amplifying the impact of an interest rate hike on inflation.
  - The response of output growth to policy rate and money supply shocks remains insignificant when the exchange rate is set as exogenous.
- Overall assessment:
  - Compared to advanced economies, transmission from policy instruments to output is relatively muted in the CCA.
  - Transmission from the interest rate to inflation is short-lived.
  - The exchange rate channel plays a greater role than other channels due to direct impact of exchange rate fluctuations on inflation.
  - Some channels, such as the asset price channel, are not working well because of shallow financial markets.

### B. Exchange Rate Management and FXIs
- All CCA countries intervene in the FX market and manage exchange rate movements to varying degrees and at different times; some have de jure floating regimes while others heavily manage the exchange rate.
- FXIs remain an important monetary policy instrument across CCA, including in countries that adopted inflation-targeting-like frameworks.
- The exchange rate of local currencies vis-à-vis the Russian ruble has recently become important in intervention decisions in some countries.
- CCA countries have accumulated sizeable precautionary FX reserves; large reserves insure against capital flow movements and external shocks but are costly and may encourage currency mismatches.
- Challenges adopting a de facto fully floating exchange rate regime:
  - Policy rate transmission to aggregate demand is weak; central banks use a mix of the policy rate, FXIs, and macroprudential instruments.
  - Under inflation targeting the policy rate is main instrument and central banks are not expected to react to exchange rate movements unless they jeopardize price stability; in practice, weak interest rate transmission leads central banks to smooth exchange rate fluctuations via interventions.
  - IPF and related literature suggest that a combination of instruments (policy rate, FXIs, macroprudential) can be more effective than a single instrument response to certain external shocks in non-reserve issuing emerging economies.
- Empirical analysis of FXIs in Armenia, Georgia, and the Kyrgyz Republic (de jure floating regimes) — summarized conclusions:
  - FXIs are common even in countries that announced a transition to inflation targeting and adopted a floating exchange rate regime.
    - FX sales (purchases) took place for 23 (30) percent of weeks in the sample for Armenia.
    - FX sales (purchases) took place for 17 (19) percent of weeks for Georgia.
    - FX sales (purchases) took place for 29 (10) percent of weeks for the Kyrgyz Republic.
  - FXIs respond to exchange rate movements and volatility:
    - FX sales (purchases) are more (less) likely when the exchange rate depreciates against its 12-week moving average.
    - FX sales (purchases) are less (more) likely in periods of high exchange rate volatility.
  - Central banks “lean against the wind”:
    - Sell foreign currency in periods of exchange rate depreciation and purchase in periods of appreciation.
    - FX sales occur in weeks when the exchange rate depreciates relative to normal periods and relative depreciation comes down following FX sales.
    - FX purchases occur in weeks when the exchange rate appreciates relative to normal periods and relative appreciation comes down following FX purchases.
  - This “leaning against the wind” is asymmetric toward leaning against depreciation, but this has not led to deterioration of reserve adequacy metrics in these countries—suggesting interventions aim to smooth fluctuations rather than maintain an overvalued exchange rate.
  - The level of exchange rate management is relatively modest and consistent with the announced floating regimes, despite some recent spikes in management activity.

### C. High Dollarization
- Dollarization challenges:
  - High dollarization affects macroeconomic stability, monetary policy transmission, and financial sector development.
  - Dollarization stems from a history of instability, hyperinflation, and large exchange rate depreciations in the 1990s, leading households to prefer holding dollars as a store of value.
  - Highly dollarized balance sheets cause agents to prefer hard currency to preserve real consumption amid macroeconomic uncertainty (Ize and Levy Yeyati 2005).
  - Armenia, Georgia, the Kyrgyz Republic, and Tajikistan have negative net international investment positions above 70 percent of GDP, exacerbating contractionary impacts of currency depreciation on balance sheets.
- Recent trends in dollarization:
  - Financial dollarization has declined in recent years due to prudent macroeconomic policies, macroprudential measures, and moves to floating exchange rates.
  - A decade ago, the share of FX deposits in total deposits averaged 60 percent across the CCA (measured at constant exchange rates); deposit dollarization rates varied from 34 percent in Armenia to about 80 percent in Georgia.
  - Deposit dollarization in the CCA fell to 38 percent in 2022; Georgia remained the highest in the region at 57 percent in 2022.
  - Credit de-dollarization has been more pronounced:
    - The share of FX loans in total loans declined from 62 percent in 2010 to 28 percent in 2022 (CCA average).
    - Kazakhstan’s credit dollarization fell from 74 percent of all credit in 2010 to single digits in 2022.
  - The gap between deposit and credit dollarization has narrowed because of regulatory limits on banks’ net open foreign exchange positions.
  - Over the last decade, CCA dollarization rates declined significantly and converged to those in other EMDEs; other EMDEs saw declines in loan dollarization and slight increases in deposit dollarization rates (regional comparisons noted).

*Source: IMF Departmental Papers — Strengthening Monetary Policy Frameworks in the CCA (excerpt: "3. Challenges in Monetary Policy")*

### 1. Dollarization of Deposits

### 1. Dollarization of Deposits

### Overview and recent trends
- Deposit and credit dollarization are computed at constant exchange rates (foreign exchange deposit/credit (the end of month stock) divided by the same day exchange rate and then multiplied by the exchange rate as of May 31, 2021). Deposit/credit dollarization is a ratio of foreign exchange-denominated deposit/credit counted in local currency to total deposit/credit, multiplied by 100.
- Reported regional changes: deposit dollarization increased by 12 percentage points in emerging Europe, 13 percentage points in LAC, and 1 percentage point in emerging Asia.
- After introducing limits on open FX positions, the gap between loan and deposit dollarization has narrowed considerably since 2010.
- In Azerbaijan, Georgia, and Kazakhstan, credit dollarization has dropped significantly more than deposit dollarization, mainly due to regulatory restrictions on FX lending.

### Policy measures and regulatory approaches in the CCA
- Most CCA countries have an official de-dollarization policy in place.
- All CCA countries apply higher reserve requirement ratios and additional capital requirements for FX-denominated liabilities versus domestic currency liabilities.
- Reserve requirement differential examples:
  - Azerbaijan: 0.5 percent
  - Georgia: close to 20 percent
- Macroprudential and administrative measures in use:
  - Bans on FX lending to specific segments (for example, mortgages) in some countries.
  - Georgia introduced a floor on FX loans.
  - Measures such as differentiated loan-to-value and payment-to-income ratios are less common and more recent.
- All CCA countries issue long-term government securities in local currency, with Uzbekistan being the latest to do so; however, local currency debt accounts for a small share of total public debt in the CCA.
- There is scope for further regulatory measures to mitigate risks from FX lending, particularly where credit dollarization exceeds deposit dollarization, although some countries (for example, Georgia) already apply nearly all available regulatory tools.

### Short-term drivers of financial de-dollarization (empirical findings)
- A standard recursive VAR was used to examine roles of macro-variables, prudential measures, and development of financial capital markets on credit and deposit de-dollarization in the CCA, estimating impulse response functions and forecast error variance decompositions.
- Key empirical findings:
  1. Increasing the spread between reserve requirement ratios on foreign currency and the local currency has contributed to deposit and credit de-dollarization in several countries.
  2. Various prudential measures—discouraging bank lending in foreign currency to unhedged borrowers, raising provisions on foreign currency-denominated loans, introducing differentiated capital risk weights on foreign currency loans, tightening capital requirements against foreign exchange positions—have contributed to credit de-dollarization in several countries.
  3. Countries’ concerted efforts to extend the yield curve and develop the domestic debt market have not been associated with sufficient de-dollarization, possibly due to these markets’ low level of development.
  4. The effects of exchange rate volatility and inflation on financial dollarization are mixed.
- Directionality: deposit dollarization is driving credit dollarization, not the other way around.

### Financial stability context and market development
- CCA countries have made significant progress developing capital markets over the last decade, but domestic bond market development remains incomplete (need for systematic approach toward money markets, primary and secondary markets, investor base, and market infrastructure).
- Several CCA countries have a negative net international investment position above 70 percent of GDP, which exacerbates the contractionary impact of FX depreciation on balance sheets.

*Sources: National authorities; and IMF staff calculations.*

### 1. Length

### 1. Length

### Press releases: focus, tone, and readability
- Reading easiness measured by Flesch index; Flesch index defined as: "206.835 – 1.015*average words per sentence – 84.6*average syllables per word."
- Analysis performed using English versions of press releases.
- Overall sentiment of press releases in the region is negative; a more negative tone increases interest rate pass-through due to the signaling channel of communication.
- Figures and word clouds show country-specific focus: Armenia, Georgia, Kazakhstan, Azerbaijan, Uzbekistan, and the Kyrgyz Republic emphasize inflation or growth themes in press releases.

### Country groupings and definitions referenced
- AEs: Australia, Canada, euro area, Iceland, Israel, New Zealand, Norway, and Sweden.
- EMDEs include Albania, Brazil, Chile, Colombia, Czech Republic, Hungary, Indonesia, Mexico, Moldova, Mongolia, Peru, Philippines, Poland, Romania, Russian Federation, South Africa, Serbia, Thailand, Türkiye, and Ukraine.
- Country abbreviations use International Organization for Standardization country codes.

### Transparency and similarity metrics
- Central bank transparency is reported on an index with range 0–15 (0 = minimum and 15 = maximum).
- Overall transparency in the region has increased in sync with global trends; upgrades driven by greater disclosure of policy decisions and their motivation.
- Figures include measures of average similarity and average sentiment of press releases across time and country groups.

### Views from CCA central bankers (survey)
- Survey conducted as part of the Joint Vienna Institute initiative (June 2021).
- Central banks view monetary policy reports and press releases as the most relevant communication tools; social media postings rank significantly lower.
- All surveyed officials reported that central bank communications have an impact (either low or high) on interest rates, inflation expectations, and exchange rate; none reported "no impact."
- Distribution of views:
  - Four of six central banks consider the impact of communications on inflation expectations as high.
  - Five of six central banks consider the impact of communication on retail interest rates as low.

### Progress and gaps on best-practice principles for modernizing monetary policy frameworks
- Principles taken from IMF (2015); traffic-light assessment used to reflect progress by country teams.
- Key principles and assessments:
  - Clear legal mandate and operational independence: All CCA central banks have a legal mandate to pursue monetary policy objectives and many have monetary policy committees; however, some laws leave gaps in selection and dismissal of board members and their qualifications. De facto independence can be compromised by government influence.
  - Price stability as primary objective: Most central banks pursue price stability as main objective. Turkmenistan is the only country with a fixed exchange rate regime. In Azerbaijan, de jure free float is de facto stabilized relative to the US dollar.
  - Numerical medium-term inflation objectives: Except for Turkmenistan, all CCA central banks have a numerical inflation objective. Some target a point, others a range. Example—National Bank of Kazakhstan targets:
    - 2021–22: 4–6 percent
    - 2023–24: 4–5 percent
    - From 2025 onwards: 3–4 percent
  - Consideration of output and financial stability: Some central banks are mandated to ensure financial stability and contribute to financial development; this necessitates macro- and micro-prudential policies and addressing supervisory and data gaps, especially in non-bank financial sector.
  - Effective operational framework centered on controlling short-term interest rates: Further work needed; transmission from policy rates to money market, deposit, and lending rates is relatively weak in parts of the region. Causes include use of multiple instruments (reserve requirement ratios, FXIs, subsidized lending), structural excess liquidity, and weak liquidity forecasting.
  - Forward-looking strategy and FPAS (forecasting and policy analysis system): Armenia, Georgia, and Kazakhstan have made good progress with FPAS; others remain in initial stages due to institutional bottlenecks and capacity constraints.
  - Clear communication: Progress made (Armenia introduced regular press conferences; Kyrgyz Republic improved monetary reports; most CCA countries have well-organized websites) but gaps remain. Need more transparency on monetary operations and FXIs; timely publication of such information available only from a few central banks.

### Policy recommendations and implementation considerations
- Continue modernization guided by best-practice principles; adaptations possible outside inflation-targeting regimes.
- Strengthen legal frameworks to close gaps in board selection/dismissal and ensure operational independence in practice.
- Improve monetary policy transmission by:
  - Reducing use of instruments that work at cross-purposes (e.g., simultaneous FXIs and subsidized lending).
  - Enhancing liquidity forecasting, including government transaction forecasts.
  - Deepening financial markets, improving financial inclusion, developing a benchmark yield curve, and reducing dollarization.
- Advance de-dollarization through:
  - Credible monetary and exchange rate frameworks.
  - Absence of fiscal dominance.
  - Deep domestic financial markets offering long-term instruments.
  - Regulatory measures to mitigate FX lending risks where credit dollarization exceeds deposit dollarization.
  - Acknowledge that de-dollarization is a gradual process requiring low and stable inflation for an extended period.
- Gradual transition to greater exchange rate flexibility to cushion external shocks, facilitate relative price adjustment, preserve export competitiveness, limit current account imbalances, and protect FX reserves; FXIs can be used sparingly to smooth transitions but require clarity, transparency, and communication.
- Use the Integrated Policy Framework (IPF) to select policy mixes that jointly consider monetary policy, FXIs, and macroprudential regulation in presence of disruptive capital flows and market frictions; balance short-term benefits of FXIs and macroprudential measures against potential adverse side effects (unhedged FX liabilities, erosion of market discipline).
- Improve communications to enhance credibility:
  - Streamline press releases to be explicit on monetary policy stance.
  - Explicitly cite numerical monetary targets to help anchor inflation expectations.
  - Explain views on current and future economic conditions, rationale for policy measures, and outcomes of monetary policy actions.
  - Publish more transparent and timely data on central bank operations and FXIs.

*Source: IMF staff estimates and IMF Departmental Papers — Strengthening Monetary Policy Frameworks in the CCA.*

### Annex 1. Empirical Analysis of the Monetary

### Annex 1. Empirical Analysis of the Monetary Policy Transmission Mechanism in the CCA

### Methodology
- Baseline: panel VAR on annual data for seven CCA countries (Armenia, Azerbaijan, Georgia, Kazakhstan, Kyrgyz Republic, Tajikistan, and Uzbekistan) for the period 2000–20.
- Endogenous variables: real GDP growth (y), CPI inflation (p), changes in real central bank policy rate (r), real base money growth (m), nominal local currency per US dollar exchange rate growth (e).
- Lags: two lags in the VAR.
- Controls: country and time fixed effects.
- Identification: Cholesky decomposition with recursive ordering Y = [y, p, r, m, e].
  - Implication: y and p are non-policy variables observed contemporaneously by policymakers; r, m, and e are policy variables affecting non-policy variables with a lag.
- Stability: eigenvalue stability conditions satisfied (modulus of eigenvalues is less than 1).
- All variables in differences are I(0).

### Granger causality and variance decomposition (key statistics)
- Granger causality test results (reported p-values):
  - Effect on output growth (y):
    - Policy rate (r): 0.306
    - Money supply (m): 0.374
    - Exchange rate (e): 0.151
    - Jointly (p, r, m, e): 0.005***
  - Effect on inflation (p):
    - Policy rate (r): 0.000***
    - Money supply (m): 0.771
    - Exchange rate (e): 0.017**
    - Jointly (y, r, m, e): 0.000***
- Interpretation: policy variables have a significant joint Granger effect on inflation and output; bivariate tests indicate that only the policy rate and the exchange rate Granger-cause inflation individually, while none of the policy variables Granger-cause output individually.
- Variance decomposition (share of forecast error variation explained by impulse variables, percent):
  - Response variable: output growth (y)
    - 1 year: y 0.974, p 0.013, r 0.000, m 0.004, e 0.008
    - 2 years: y 0.963, p 0.012, r 0.002, m 0.006, e 0.017
    - 3 years: y 0.932, p 0.041, r 0.002, m 0.007, e 0.018
    - 4 years: y 0.921, p 0.052, r 0.003, m 0.006, e 0.018
    - 5 years: y 0.917, p 0.054, r 0.003, m 0.007, e 0.019
  - Response variable: inflation (p)
    - 1 year: y 0.002, p 0.893, r 0.031, m 0.002, e 0.072
    - 2 years: y 0.002, p 0.832, r 0.036, m 0.017, e 0.114
    - 3 years: y 0.003, p 0.817, r 0.044, m 0.017, e 0.119
    - 4 years: y 0.003, p 0.817, r 0.045, m 0.016, e 0.118
    - 5 years: y 0.004, p 0.813, r 0.048, m 0.017, e 0.118
- Notable numeric findings:
  - The share of inflation variation explained by the exchange rate reaches 12 percent in the fifth year (reported as 0.118 in table for 5 years), more than two times higher than the variation explained by the policy rate (5 percent; reported as 0.048 in table for 5 years).
  - Only 2 percent of inflation variation is explained by monetary base (reported as 0.017 or 0.002 depending on horizon; narrative statement: "Only 2 percent").
  - For output growth, policy variables explain a small fraction of variation, ranging from 0.3 percent (policy rate) to 2 percent (exchange rate) in narrative terms.

### Impulse response results and interpretation
- Panel VAR IRFs indicate:
  - Monetary policy shocks mostly affect nominal variables (inflation) rather than real variables (real GDP growth).
  - A 1 percent shock to the real policy rate reduces inflation by 0.5 percent in the first year.
  - A 1 percent shock to the exchange rate depreciation increases inflation by 0.3 percent in the first year.
  - The exchange rate channel amplifies the interest rate channel.
  - Money supply shocks do not have a significant impact on inflation.
- Monthly country-specific VARs (country-level):
  - Use monthly data excluding output (no monthly output series available).
  - Exogenous variables: changes in global oil prices and U.S. 3-month T Bill rates.
  - Variables measured as 12-month changes; dummy variables for structural shifts.
  - Lags selected by Schwartz-Bayes information criterion.
  - Findings: support for the exchange rate channel in most countries; interest rate and monetary shocks largely insignificant for inflation at country level.
- Interpretation: results consistent with shallow financial markets in CCA countries, strengthening the exchange rate transmission channel relative to other traditional channels.

### Principal conclusions and policy implications
- Exchange rate channel dominates other transmission channels in the CCA.
- Monetary policy tends to affect nominal variables (inflation) more than real variables (output), supporting the notion that monetary policy should focus on stabilizing nominal variables.
- Policy instruments:
  - Policy rate changes have a significant effect on inflation (per Granger tests and IRFs).
  - Exchange rate movements explain a larger share of inflation variation than the policy rate over the medium term.
  - Monetary base (real base money growth) has limited explanatory power for inflation.
- Implication for monetary frameworks: emphasis on nominal targets (for example, inflation targeting) is supported by the evidence that policy operates more strongly through nominal channels, but the prominent role of the exchange rate channel and shallow financial markets should be taken into account in framework design and communication.

*Source: IMF staff calculations and analysis in Annex 1 of the provided IMF departmental paper.*

### Annex Table 2.1. Determinants of FX Interventions: Ordered Logit Model

### Annex Table 2.1. Determinants of FX Interventions: Ordered Logit Model

### Ordered Logit Model Results
- Dependent variable: ordered categorical variable for foreign exchange purchases (−1), no interventions (0), and foreign exchange sales (1).
- Estimation method: maximum likelihood estimator. Robust standard errors in parentheses.
- Coefficients (with robust standard errors) and significance:
  - Deviation of exchange rate from its 12-week moving average (percent)
    - Armenia: 0.399*** (0.063)
    - Georgia: 0.162*** (0.032)
    - Kyrgyz Republic: 0.280*** (0.058)
  - 12-week st. dev. of exchange rate changes (percent)
    - Armenia: −0.386** (0 .16 4)
    - Georgia: 0 .13 8 (0.145)
    - Kyrgyz Republic: −0.282** (0 .137)
- Model fit and sample:
  - Observations: Armenia 74, Georgia 3579, Kyrgyz Republic 534
  - Log-likelihood: Armenia −728 .9, Georgia −485.5, Kyrgyz Republic −444.3
  - Pseudo Rsq: Armenia 0.0587, Georgia 0.0274, Kyrgyz Republic 0.0439
- Significance notation: *** p < 0.01, ** p < 0.05, * p < 0 .1.

### Estimated Probabilities of FX Interventions
- Reported are estimated probabilities of FX interventions from the ordered logit model (Annex Figure 2.5).
- FX categories referenced: FX purchases, FX sales, Neutral (no intervention).

### Event Study Findings: "Leaning against the wind"
- Summary statement: Estimation results provide evidence of “leaning against the wind.”
- FX sales:
  - Happen in weeks during which the exchange rate depreciates relative to normal periods by about:
    - Armenia: 0.4 percent
    - Georgia: 1 percent
    - Kyrgyz Republic: 0.2 percent
  - Following the sale, the relative depreciation comes down and becomes not significantly different from zero on the third week in all countries.
- FX purchases:
  - Happen in weeks during which the exchange rate appreciates relative to normal periods by about:
    - Armenia: 0.3 percent
    - Georgia: 0.45 percent
    - Kyrgyz Republic: 0.4 percent
  - Following the purchase, the relative appreciation comes down and becomes not significantly different from zero on the third week in all countries.

### Event Study Analysis Details
- Reported are results of the event study analysis on changes in the local currency vis-à-vis US dollar exchange rate around FX interventions (Annex Figure 2.6).
- Time = 0 refers to the week of FX interventions. Filled squares indicate significance at 95 percent confidence level.
- Data sources cited for event study: National Bank of the Kyrgyz Republic; and IMF staff calculations.

*Source: IMF staff calculations. Note: Pseudo Rsq = statistical measure used to assess the goodness of fit of the ordered logit model.*

---


_Source: https://www.imf.org/-/media/files/publications/dp/2023/english/smpfea.pdf_