## wpiea2022229-print-pdf

## Source details

**Canonical URL:** [wpiea2022229-print-pdf](https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022229-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2022/english/wpiea2022229-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2022/english/wpiea2022229-print-pdf.pdf.json)

---

### Transition 2017-21: background and progress
- Uzbekistan began transitioning monetary policy away from an exchange rate peg in September 2017, focusing monetary policy on achieving price stability.
- Central Bank of Uzbekistan (CBU) removed from list of state bodies under government control, enabling greater operational independence.
- Restrictions on access to foreign exchange abolished, resulting in a 50 percent depreciation of the official exchange rate and unifying the exchange rate.
- New central bank law enacted in October 2019 set price stability explicitly as the primary goal and started the CBU’s transition to an inflation targeting (IT) regime.
- IMF technical assistance provided in November 2017, February 2018, February 2019, September 2020, and December 2021.

### IT framework, objectives, and CBU strategy
- IT essentials and preconditions:
  - IT entails targeting inflation directly and explicitly; adopted by 41 countries, including three in the Caucasus and Central Asia.
  - General preconditions: institutional independence, stable macroeconomic fundamentals, well-developed analytical capabilities, and a healthy financial system.
  - Additional requirements: an effective monetary policy instrument, understanding of the transmission mechanism, improved economic databases, and reduced dollarization.
  - Five main IT elements (Hammond, 2012): (i) price stability primary objective; (ii) public numerical inflation targets; (iii) informative strategy using broad information including inflation forecasts; (iv) transparent communication; (v) enhanced accountability.
- CBU strategy and numerical targets:
  - December 2019: CBU announced IT implementation by 2023; formal IT strategy adopted January 2020.
  - Numerical targets: reduce the inflation rate to 10 percent by the end of 2021 and to 5 percent by the end of 2023.
  - Strategy elements: coordination CBU–Cabinet of Ministers; adjust remaining administered prices to (at least) cost recovery levels (by February 2020); reduce preferential credit; keep overall fiscal deficit below 1.5 percent of GDP.
  - Progress uneven: substantial CBU-controlled implementation; pending government actions on credit policy and utility tariffs.

### Institutional, operational, and legal reforms
- 2019 central bank law: price stability overriding objective, grants operational independence, prohibits CBU financing the government, enhances accountability.
- Executive board composition: 9 members — governor, 1st deputy governor, 5 deputy governors, and 2 independent members; law required two independent, non-executive members.
- Accountability: three-member audit committee; CBU financial statements externally audited.
- Monetary policy decision process: formalized in seven stages; executive board holds eight regular meetings a year.
- Operational changes:
  - Operating target shifted from banks’ reserves to policy rate and interbank rate.
  - CBU produces liquidity forecasts; reformed standing facilities; consolidated lending facilities; introduced overnight deposit facility.
  - New OMO framework (early 2020): FX swaps, Repo auctions, deposits, and CBU securities auctions.
  - These changes helped keep interbank rate within the policy rate corridor.

### Analytical, forecasting, and communication capacity
- Forecasting and models:
  - CBU developed short-term forecasting models and adopted an FPAS model for key macro indicators and policy simulations.
  - Limitations: large structural break in 2017; limited statistical coverage (lack of detailed unemployment statistics; limited quarterly national accounts data for early 2022).
  - Monthly inflation expectations survey of households and firms.
- CBU models (short-term forecasting suite; frequency; key variables preserved):
  - ARIMA (Monthly; Core inflation, inflation)
  - BVAR (Monthly; Inflation, credit)
  - FAVAR (Monthly; Inflation; Foreign inflation, Exchange rate, CPI components)
  - VECM (Monthly; M0, remittance, FX, interest rate, CPI, GDP)
  - DFM (Monthly; GDP; Oil price, industrial production, export, remittances)
  - ARDL (Quarterly; Core inflation; Output gap, PPI, International food prices, FX, Money supply)
  - DOLS (Quarterly; Core inflation; Output gap, PPI, International food prices, FX, Money supply)
  - ARIMA* (Daily; Cash in circulation) — ARIMA* forecasts daily impact of autonomous factors on liquidity.
- Communication enhancements: annual Monetary Policy Guidelines for next 3 years; press conferences and press releases after board meetings; quarterly review of monetary policy since 2018; monthly monetary and financial statistics since 2018; quarterly statistical bulletin; use of social media including Telegram.

### Inflation developments, exchange rate dynamics, and expectations
- Inflation path:
  - After the 50 percent depreciation in September 2017, inflation surged temporarily but trended down due to tight monetary policy.
  - Inflation remains relatively high with inertia from steady exchange rate depreciation, price deregulation and increases in administered prices, and global supply/commodity pressures.
- Inflation expectations:
  - Households’ and firms’ expectations remain above headline inflation due to short CBU track record and general expectation bias.
  - Number of households and firms expecting inflation within 9-12 percent (close to the CBU’s 10 percent 2021 target) has been steadily increasing.
- Exchange rate and intervention:
  - Exchange rate broadly moved in line with fundamentals; CBU intervened daily to avoid undue volatility because of shallow market and large government/SOE transactions.
  - CBU interventions sterilized domestic currency injected when buying gold (selling foreign currency to offset gold purchases) — interventions aimed to sterilize gold purchases and stabilize excess volatility.
  - CBU’s stock of international reserves remained relatively stable during 2017-20.
  - Early 2021: CBU implemented FX market reforms to promote market trading and increase FX flexibility.
  - Day-to-day exchange rate volatility limited compared with many IT peers, but increased more recently.

### Preconditions and quantitative indicators for IT (selected values preserved)
- Overall assessment: progress toward IT pre-conditions but gaps remain.
- Selected quantitative indicator highlights (post-adoption values preserved exactly as in source Table 1):
  - Technical infrastructure (post-adoption): ICs 0.98, EMs 0.97, UZB (2021) 0.83.
  - Data availability (post-adoption): ICs 0.94, EMs 0.92, UZB 0.50.
  - Models capable of conditional forecasts (post-adoption): ICs 1.00, EMs 1.00, UZB 1.00.
  - Financial system health (post-adoption): ICs 0.60, EMs 0.48, UZB 0.40.
  - Private bond market capitalization to GDP (post-adoption): ICs 0.31, EMs 0.07, UZB 0.00.
  - Institutional independence (post-adoption): ICs 0.78, EMs 0.72, UZB 0.65.
  - Central bank legal mandate (post-adoption): ICs 0.44, EMs 0.62, UZB 0.50.
  - Central bank independence (post-adoption): ICs 0.72, EMs 0.64, UZB 0.92.
  - Exchange rate pass-through (0=high, 1=low) (post-adoption): ICs 0.50, EMs 0.44, UZB 0.00.
  - Extent of dollarization (post-adoption): ICs 1.00, EMs 0.75, UZB 0.50.
  - Trade openness (0= fully open, 1= fully autarkic) (post-adoption): ICs 0.16, EMs 0.21, UZB 0.81.

### Institutional independence: Modified Cukierman Index (selected values)
- De-jure central bank independence index increased from 0.38 to 0.78 (0–1 scale; 1=maximum).
- Selected subcomponents (UZB 2016 -> UZB 2020; EU peers 2020 for comparison):
  - Overall Central Bank Independence: 0.38 -> 0.78 (EU peers 0.89).
  - Central Bank Board: 0.44 -> 0.44 (EU peers 0.79).
  - Central Bank objectives (Fundamental objective): 0.00 -> 1.00 (EU peers 1.00).
  - Central Bank lending (aggregate): 0.25 -> 0.82 (EU peers 0.86).
  - Accountability: 0.75 -> 0.75 (EU peers 1.00).
  - Central Bank transparency (subcomponent): 0.00 -> 0.00 (EU peers 1.00).
- Note: As a de-jure index, it may overestimate actual independence where rule of law is weak.

### Monetary policy transmission: constraints, empirical assessment, and elasticities
- Main transmission constraints:
  - Limited development of financial markets.
  - High degree of dollarization.
  - Government credit policies and large state-owned banks/SOEs footprint.
- Dollarization levels (2021):
  - Deposits dollarization: 41 percent.
  - Credit dollarization: 50 percent.
- State-owned banks:
  - 11 state-owned banks out of 35 banks, accounting roughly for 82 percent of total banking sector’s assets.
- Empirical elasticities (Percentage Points) — OLS regressions (sample periods: 2010M1-2017M8; 2017M9-2022M7; 2010M1-2022M7):
  - Lending Rates:
    - UZB: 0.30*** ; 0.59** ; 0.37***
    - Selected IT peers¹: 0.68**
  - Deposit Rates:
    - UZB: -0.05 ; 0.35 ; 0.04
    - Selected IT peers¹: 0.77***
  - Note: Stars denote significance levels: ***, **, * represent 1%, 5% and 10% respectively.
  - ¹ includes Poland, Hungary, and Romania.
- Interpretation:
  - Transmission from the policy rate to lending rates has strengthened; pre-2017 transmission may reflect administrative pricing by state-owned banks.
  - Transmission to deposit rates increased but statistically insignificant.
  - Overall transmission remains constrained relative to IT peers.

### VAR analysis: policy rate, NEER, and core inflation
- VAR setup:
  - Monthly data from 2016:M12 to 2020:M10.
  - Endogenous Y_t: policy rate, nominal effective exchange rate (NEER), core inflation rate.
  - Exogenous X_t: global energy prices and food prices.
  - GIRF by Pesaran and Shin (1998) used.
  - Variables (except policy rate) in logs; first differences used due to integration order.
  - Core inflation defined as headline inflation excluding regulated prices and fruits and vegetables.
- Results:
  - Policy rate appears to reduce core inflation, but the response is statistically insignificant.
  - Inflation appears to respond to exchange rate changes, but also statistically insignificant.
  - Combined implication: constraints on effectiveness of monetary policy transmission in Uzbekistan; similar results for headline inflation.

### Econometric and data limitations
- Reliable interbank market data available only since April 2018; government securities market shallow with low trading volumes.
- Structural changes (2017 liberalization) create breaks limiting econometric inference.
- Consequence: econometric analysis can only address overall transmission from policy rates to short-term retail interest rates.

### Policy recommendations: increasing credibility and coordination
- Central bank governance and transparency:
  - Improve de-facto independence and predictability.
  - Address high rotation and limited independent executive board members.
  - Recommendations:
    - Set terms (in years) for all executive board members and clarify selection/dismissal criteria.
    - Ensure a majority of independent board members or a large proportion with oversight roles; independent members should lead external audit committees.
    - Adopt IFRS and publish audited financial statements.
- Coordination with government:
  - Enhance coordination on preferential lending programs, utilities price regulation, and public sector wage policies.
  - Government should provide predictability and close coordination to support IT.

### Policy recommendations: improving transmission, FX and de-dollarization
- Financial market development:
  - Deepen interbank market: establish money market working group; develop benchmark rate and standard master repo agreement; develop clearing and settlement for interbank repo.
  - CBU published methodology for the benchmark interbank interest rate in July 2022.
  - Working group established in 2020 with EBRD support (3 EBRD experts, CBU departments, commercial banks’ representatives); meets at least once a quarter.
- FX market reforms:
  - Increase transparency on FX intervention rules (principles, objectives, operational guidelines, tactics).
  - Support transition to price discovery via FX multiple price auctions.
  - Adopt a version of the Global FX code or introduce formal market arrangements.
  - Increase moderate FX volatility so agents perceive fluctuations as reversible.
- De-dollarization measures:
  - Fiscal, prudential, and market-development measures to incentivize local-currency saving:
    - Higher reserve requirements on foreign currency deposits.
    - Raising insurance premiums on dollar deposits.
    - Holding reserve requirements for foreign currency deposits in local currency.
    - Requiring tighter provisioning on foreign currency loans.
    - Developing instruments and markets to hedge currency risks.
  - International examples provided (country measures preserved in source).

### Structural reforms to strengthen transmission
- Reduce state role in banking and economy:
  - Impose hard budget constraints on state-owned banks.
  - Improve SOE and state bank governance; appoint independent and qualified board members.
  - Reduce government funding and explicit assistance/guarantees to SOEs.
  - Consider eventual privatization of state-owned banks.
  - Strengthen bank supervision.
  - Promote market-determined energy pricing to avoid distortions.
- Rationale: ensuring firms and SOEs face hard budget constraints increases monetary policy efficacy.

### Transition considerations and concluding assessment
- Constraints during transition: large state role in economy and banking sector; administrative prices; government lending at preferential rates; fiscal dominance.
- Central bank must design policy acknowledging these limitations and coordinate with government to reduce state role.
- Progress achieved: substantial improvements in CBU institutional and governance set-up, forecasting capacities, monetary operations, policy formulation, and communication.
- Remaining priorities:
  - Enhance CBU governance.
  - Develop domestic capital and FX markets.
  - Reduce dollarization.
  - Reduce the role of the government in the economy and banking sector.
- Final assessment: large government role in bank ownership and broader economy, combined with government lending policies, has constrained monetary policy transmission.

*Source: wpiea2022229-print-pdf - Conclusion and selected chapters*

### Conclusion .............................................................................................................

### Conclusion

### Transition 2017-21: Background and progress
- Uzbekistan began transitioning monetary policy away from an exchange rate peg in September 2017, focusing monetary policy on achieving price stability.
- The Central Bank of Uzbekistan (CBU) was removed from the list of state bodies under government control, paving the way for greater operational independence.
- Restrictions on access to foreign exchange were abolished, resulting in a 50 percent depreciation of the official exchange rate and unifying the exchange rate.
- A new central bank law enacted in October 2019 set price stability explicitly as the primary goal of monetary policy and started the CBU’s transition to an inflation targeting (IT) regime.
- The IMF provided technical assistance in November 2017, February 2018, February 2019, September 2020, and December 2021.

### IT framework, objectives, and preconditions
- IT entails targeting inflation directly and explicitly; IT has been adopted by 41 countries, including three in the Caucasus and Central Asia.
- General preconditions for IT include: institutional independence, stable macroeconomic fundamentals, well-developed analytical capabilities, and a healthy financial system.
- Additional emphasized requirements: an effective and clear monetary policy instrument, understanding of the transmission mechanism, building economic databases, and reducing dollarization in the financial sector.
- Five main elements of an IT framework (Hammond, 2012) highlighted: (i) price stability as the primary objective; (ii) public announcement of numerical inflation targets; (iii) an informative monetary policy strategy using broad information including inflation forecasts; (iv) transparent communication of plans, objectives, and rationale; (v) enhanced accountability.

### CBU strategy, targets, and constraints
- In December 2019 the CBU announced IT implementation by 2023; a formal IT strategy was adopted in January 2020.
- Numerical targets in the strategy: reduce the inflation rate to 10 percent by the end of 2021 and further to 5 percent by the end of 2023.
- Strategy elements: (i) close coordination between the CBU and the Cabinet of Ministers to remove obstacles to effective IT transition; (ii) plan for adjusting remaining administered prices to (at least) cost recovery levels (by February 2020); (iii) reducing provision of credit on preferential terms; (iv) keeping the overall fiscal deficit to below 1.5 percent of GDP.
- Progress has been uneven: substantial implementation within CBU control, while areas requiring government action (credit policy, utility tariffs) remain pending.

### Institutional and operational reforms
- The 2019 central bank law: sets price stability as overriding objective, grants operational independence, makes the CBU accountable for fulfilling objectives (including banking and payment system stability), and prohibits the CBU from financing the government.
- The law required two independent, non-executive members to be appointed to the executive board. The CBU board has 9 members: the governor, the 1st deputy governor, 5 deputy governors, and 2 independent members.
- Accountability enhancements: a three-member audit committee and a requirement for CBU financial statements to be externally audited.
- Monetary policy decision-making was formalized in seven stages and the CBU executive board holds eight regular meetings a year.
- Operational improvements: banks’ reserves initially became the operating target, later replaced by the policy rate and interbank rate; CBU began producing liquidity forecasts; reformed standing facilities including consolidation of lending facilities and introduction of an overnight deposit facility; new OMO framework introduced in early 2020 for FX swaps, Repo auctions, deposits, and CBU securities auctions.
- These changes helped keep the interbank rate within the policy rate corridor determined by the CBU’s overnight facilities.

### Analytical, forecasting, and communication capacity
- The CBU developed a suite of short-term forecasting models and adopted an FPAS model to forecast key macroeconomic indicators and simulate policy scenarios.
- Limitations remain due to a large structural break in 2017 and limitations in statistical coverage (including lack of detailed unemployment statistics and limited quarterly national accounts data for early 2022).
- The CBU conducts an inflation expectations survey monthly.
- CBU models listed (short-term forecasting suite): ARIMA (Monthly; Core inflation, inflation), BVAR (Monthly; Inflation, credit), FAVAR (Monthly; Inflation; Foreign inflation, Exchange rate, CPI components), VECM (Monthly; M0, remittance, FX, interest rate, CPI, GDP), DFM (Monthly; GDP; Oil price, industrial production, export, remittances), ARDL (Quarterly; Core inflation; Output gap, PPI, International food prices, FX, Money supply), DOLS (Quarterly; Core inflation; Output gap, PPI, International food prices, FX, Money supply), ARIMA* (Daily; Cash in circulation). ARIMA* forecasts daily impact of autonomous factors on liquidity.
- Communication stepped up: annual Monetary Policy Guidelines for the next 3 years, press conferences and press releases after board meetings, quarterly review of monetary policy since 2018, monthly monetary and financial statistics since 2018, quarterly statistical bulletin, and use of social media including a Telegram account.

### Legal and banking sector reforms
- New laws on foreign exchange, payment system, and banking approved in late 2019.
- Government instructions consolidated preferential lending programs to within 3 out of 13 state-owned banks, and preferential loans could no longer be issued at interest rates lower than the CBU policy rate.
- Authorities developed a strategy to reform the banking sector and restructure state-owned banks.
- State-owned banks’ corporate governance improved with appointment of independent supervisory board members with international banking experience replacing government officials.

### Monetary policy stance and outcomes
- The real monetary policy rate (nominal rate deflated by expected inflation) was mostly negative until mid-2017; the CBU increased the nominal policy rate as an initial step toward a new regime.
- Since 2017 the real monetary policy rate has been positive most of the time and since 2019 it has fluctuated around 2-3 percent, indicating a more aggressive approach to monetary policy.

*Source: wpiea2022229-print-pdf - Conclusion*

### 15.   Inflation  has  been  gradually  declining  in  recent  years  and  the  CBU  has  been  able  to  influence

### 15.   Inflation  has  been  gradually  declining  in  recent  years  and  the  CBU  has  been  able  to  influence 

### Inflation developments and expectations
- After a 50 percent depreciation in September 2017, inflation experienced a temporary surge but has been on a downward trend due to a consistent tight monetary policy stance.
- Inflation remains relatively high and shows inertia due to:
  - steady depreciation of the exchange rate;
  - further price deregulation and increases in administered prices;
  - global trends, including supply chain challenges and commodity price increases.
- Households’ and firms’ inflation expectations remain above headline inflation, attributed to:
  - the CBU’s still relatively short track record; and
  - households’ and firms’ general expectation bias (perceiving current inflation higher than actual inflation), as documented in the literature (e.g., Van Duyne 1982, Coibion et al. 2018, and Murphy & Rohde 2018).
- The CBU appears to be making progress influencing inflation expectations:
  - the number of households and firms expecting inflation within 9-12 percent — close to the CBU’s 10 percent target for 2021 — has been steadily increasing.

### Exchange rate dynamics and CBU intervention
- The exchange rate has broadly moved in line with fundamentals.
- Due to a relatively shallow market and large government and SOE transactions, the CBU has intervened on a daily basis to avoid undue volatility.
- CBU intervention has been significant but has not altered longer-term exchange rate trends; the CBU’s stock of international reserves remained relatively stable during 2017-20.
- Intervention has largely followed the “neutrality principle,” which calls for sterilizing domestic currency injected when the CBU buys gold from domestic gold producers:
  - sterilization is done by selling foreign currency to offset gold purchases; thus interventions aim both to sterilize gold purchases and to stabilize excess volatility, although markets cannot distinguish these objectives.
- Prior to 2017, strict foreign exchange controls limited exchange rate volatility and hampered financial market development.
- Since the 2017 liberalization of the foreign exchange market, the CBU has gradually shifted to more market-based foreign exchange operations to deepen the market and facilitate trade.
- In early 2021 the CBU implemented FX market reforms to promote market trading and increase FX flexibility.
- Day-to-day exchange rate volatility has been relatively limited compared with many inflation-targeting peers, though it has increased more recently.

### Preconditions for transitioning to inflation targeting (IT)
- The paper follows the IMF (2006) methodology covering four key areas: (i) technical infrastructure of the central bank; (ii) financial system health; (iii) central bank institutional independence; and (iv) economic structure.
- Overall assessment: Uzbekistan has made progress toward IT pre-conditions but gaps remain.
- Key findings by area:
  - Technical infrastructure of the central bank:
    - Improvements in forecasting and modeling suggest Uzbekistan is close to levels observed in other emerging economies that implemented IT.
    - Data availability has improved significantly, though further work remains.
  - Financial system health:
    - Reported banking indicators are consistent with countries that implemented IT.
    - Underdeveloped capital market and short maturity of domestic bonds indicate further financial development is needed to match other emerging economies with IT regimes.
  - Institutional independence:
    - Despite improvements, economies with IT displayed stronger operational independence, more robust mandates, and greater governor job security.
    - Recent increases in fiscal deficits and public debt may complicate monetary policy going forward.
  - Economic structure:
    - High foreign exchange pass-through, a high degree of dollarization, and sensitivity to commodity prices imply continued broader reforms are needed to improve economic structure before or during an IT transition.

- Quantitative indicator highlights (selected, values preserved as in source Table 1):
  - Technical infrastructure (post-adoption): ICs 0.98, EMs 0.97, UZB (2021) 0.83.
  - Data availability (post-adoption): ICs 0.94, EMs 0.92, UZB 0.50.
  - Models capable of conditional forecasts (post-adoption): ICs 1.00, EMs 1.00, UZB 1.00.
  - Financial system health (post-adoption): ICs 0.60, EMs 0.48, UZB 0.40.
  - Private bond market capitalization to GDP (post-adoption): ICs 0.31, EMs 0.07, UZB 0.00.
  - Institutional independence (post-adoption): ICs 0.78, EMs 0.72, UZB 0.65.
  - Central bank legal mandate (post-adoption): ICs 0.44, EMs 0.62, UZB 0.50.
  - Central bank independence (post-adoption): ICs 0.72, EMs 0.64, UZB 0.92.
  - Exchange rate pass-through (0=high, 1=low) (post-adoption): ICs 0.50, EMs 0.44, UZB 0.00.
  - Extent of dollarization (post-adoption): ICs 1.00, EMs 0.75, UZB 0.50.
  - Trade openness (0= fully open, 1= fully autarkic) (post-adoption): ICs 0.16, EMs 0.21, UZB 0.81.

### Institutional independence: Modified Cukierman Index
- The CBU’s de-jure central bank independence index increased from 0.38 to 0.78 on a 0–1 scale (1=maximum), indicating improved legal/institutional independence.
- The index indicates remaining room for improvement, particularly in enhancing CBU board governance.
- Selected subcomponents (UZB 2016 -> UZB 2020; Selected EU-Peers 2020 shown for comparison):
  - Overall Central Bank Independence: 0.38 -> 0.78 (EU peers 0.89).
  - Central Bank Board: 0.44 -> 0.44 (EU peers 0.79).
  - Central Bank objectives (Fundamental objective): 0.00 -> 1.00 (EU peers 1.00).
  - Central Bank lending (aggregate): 0.25 -> 0.82 (EU peers 0.86).
  - Accountability: 0.75 -> 0.75 (EU peers 1.00).
  - Central Bank transparency (subcomponent): 0.00 -> 0.00 (EU peers 1.00).
- Note: As a de-jure index, it may overestimate actual independence where rule of law is weak.

### Monetary policy transmission: constraints and recent improvements
- Main constraints on transmission:
  - Limited development of financial markets;
  - High degree of dollarization;
  - Government credit policies;
  - Large footprint of state-owned banks and state-owned enterprises (SOEs).
- Interest rate transmission hampered by limited financial market development:
  - Interbank market developed rapidly since early 2020 but signs of segmentation persist (some banks consistently borrowers or lenders).
  - Shallow government-securities market reflects low demand for T-bills (due to low domestic savings) and low supply of T-bills (linked to high domestic interest rates related to high inflation).
  - Substantial interbank lending is unsecured, with risk limits restricting access for some banks.
- Dollarization:
  - Uzbekistan’s dollarization stood at 41 and 50 percent of deposits and credit, respectively, in 2021.
  - Dollarization reduces the base of credit and deposits over which monetary policy operates; higher dollarization weakens transmission.
  - Headline figures suggest dollarization increased due to exchange rate depreciation and banks’ increased foreign borrowing to fund credit.
  - After correcting for exchange rate revaluation effects, dollarization has broadly stabilized, suggesting the CBU’s relatively tight monetary policies have helped discourage dollarization.
- State-owned banks and government lending initiatives:
  - There are 11 state-owned banks out of 35 banks, accounting roughly for 82 percent of total banking sector’s assets.
  - State-owned banks are not fully operating on market-based rules, weakening pass-through from the CBU policy rate to lending and deposit rates.
  - Easy access of these banks to government funding and external funding weakens monetary policy transmission.
  - Government lending programs at preferential below-market rates also weaken transmission.
  - Limited bank competition (due to dominance of state-owned banks) further constrains transmission.
- SOEs:
  - Large SOE footprint dampens competition and price flexibility.
  - Many SOEs do not operate on full market principles; their prices adjust with significant lags and often follow a centralized, discretionary process.
  - This adds to inflation inertia and reduces monetary policy effectiveness, as SOEs often face softer budget constraints and better financing access.

### Quantitative assessment of transmission
- The paper documents features of transmission from policy rate to retail interest rates, inflation, and the exchange rate using:
  - OLS estimation for transmission from monetary policy rates to retail rates.
  - Vector autoregressive (VAR) model to assess transmission from the policy rate to the inflation rate and exchange rate.
- These approaches have caveats but provide useful insights into transmission effectiveness.

*Source: IMF staff working paper text (chapter 15) from wpiea2022229-print-pdf.*

### 26.   Structural and data limitation constrain the econometric analysis of the transmission from monetary

### wpiea2022229-print-pdf - 26.   Structural and data limitation constrain the econometric analysis of the transmission from monetary policy rates to retail interest rates

### Limits of econometric analysis and data
- The transmission from policy rates to retail rates consists of three stages:
  - i) transmission from the central bank policy rate to the interbank market interest rate;
  - ii) transmission from the interbank market to the yield curve of safe assets;
  - iii) transmission from the yield curve to retail interest rates.
- Data and market underdevelopment constrain analysis for stages i) and ii):
  - Reliable interbank market data available only since April 2018.
  - Government securities market is shallow, with very low trading volumes.
- Consequence: econometric analysis can only address the overall transmission from policy rates to short-term retail interest rates.

### Empirical approach and elasticities of retail rates to policy rate
- Methodology: approach similar to Angeloni and Ehrmann (2003) using simple OLS regressions due to limited observations.
  - Regression specification: ∆Interest rate "x_t" = α + β∆Monetary Policy Rate_t + ε_t.
- Comparison periods for Uzbekistan:
  - pre-September 2017 and post-September 2017 (post-liberalization of the foreign exchange market and adoption of a new monetary policy approach).
- Key estimated elasticities (Percentage Points):
  - Sample periods listed: 2010M1-2017M8; 2017M9-2022M7; 2010M1-2022M7
  - Lending Rates:
    - UZB: 0.30*** ; 0.59** ; 0.37***
    - Selected IT peers¹: 0.68**
  - Deposit Rates:
    - UZB: -0.05 ; 0.35 ; 0.04
    - Selected IT peers¹: 0.77***
  - Note: Stars represent significant levels: ***, **, * represent 1%, 5% and 10% respectively.
  - 1/ includes Poland, Hungary, and Romania.
- Interpretative points from estimates:
  - The transmission from the policy rate to lending rates has strengthened.
    - Caution: significant pre-2017 transmission may reflect state-owned banks using the policy rate administratively to price loans.
  - Transmission from the policy rate to deposit rates has increased but is not statistically significant.
  - Overall, the transmission mechanism in Uzbekistan appears constrained compared with other inflation-targeting (IT) countries.

### VAR analysis of policy, exchange rate, and inflation
- Model specification:
  - A multivariate VAR of order n estimated on monthly data from 2016:M12 to 2020:M10.
  - Endogenous vector Y_t embeds the policy rate, the nominal effective exchange rate (NEER), and the core inflation rate.
  - Exogenous vector X_t includes global energy prices and food prices.
  - Generalized Impulse Response Functions (GIRF) by Pesaran and Shin (1998) used to avoid orthogonalization and ordering dependence.
  - All variables except the policy rate are expressed in logarithm form.
  - Augmented Dickey-Fuller tests show all variables are non-stationary and integrated of order one; first differences used.
- Definition note:
  - Core inflation is defined as headline inflation excluding regulated prices as well as fruits and vegetables prices.
- Empirical IRF results and implications:
  - The policy rate appears to reduce the core inflation rate, but the response is statistically insignificant.
  - Inflation appears to respond to changes in the exchange rate, but this is also statistically insignificant.
  - Combined implication: constraints exist on the effectiveness of monetary policy transmission in Uzbekistan.
  - Similar conclusions hold for headline inflation.

### Challenges and policy recommendations — Increasing credibility
- Central bank governance and transparency:
  - Improve de-facto independence and predictability of policy to bolster credibility.
  - Address high rotation and limited number of independent executive board members.
  - Recommendations:
    - Set terms (in years) for all executive board members and clarify criteria for selection and dismissal.
    - Ensure a majority of independent board members or a large proportion of the board with oversight roles; independent members should lead external audit committees in best-practice countries.
    - Adopt IFRS and publish audited financial statements so central bank policies are reflected transparently on its balance sheet.

- Coordination with government:
  - Enhance coordination on policies affecting inflation and transmission, including:
    - Preferential lending programs,
    - Utilities price regulation,
    - Public sector wage policies.
  - Government should provide predictability and close coordination to support inflation targeting.

### Improving monetary policy transmission — Financial market and FX reforms
- Financial market development:
  - Further deepen the interbank market; suggested actions include:
    - Establishing a money market working group to develop market conventions and standards for interbank transactions.
    - Developing a money market benchmark rate and a standard master repo agreement.
    - Developing clearing and settlement infrastructure for interbank repo transactions.
  - Noted action: CBU published a methodology for the benchmark interbank interest rate in July 2022.
  - Working group established in 2020 with EBRD support; includes 3 EBRD experts, key CBU departments, and commercial banks’ representatives; meets at least once a quarter.

- Foreign exchange market and FX volatility:
  - Reduce the role of FX in shaping inflation expectations by increasing moderate FX volatility so agents perceive FX fluctuations as reversible.
  - Implementation elements:
    - More transparency about FX intervention rules: principles, objectives, operational guidelines, and strategic tactics.
    - Support transition to price discovery via FX multiple price auctions.
    - Adopt a version of the Global FX code or introduce formal market arrangements.

- De-dollarization measures:
  - Fiscal, prudential, and market-development measures recommended to reduce dollarization and incentivize saving in local currency.
  - Examples of measures drawn from international experience:
    - Imposing higher reserve requirements on foreign currency deposits.
    - Raising insurance premiums on dollar deposits.
    - Holding reserve requirements for foreign currency deposits in local currency.
    - Requiring tighter provisioning on foreign currency loans.
    - Developing instruments and markets to hedge currency risks.
  - Table of country experiences (excerpted measures and examples) includes:
    - Exchange rate flexibility (Laos, P.D.R (1995); Turkey (2001)).
    - Interest rate increases on local currency deposits (Egypt (1991-1999); Estonia (1992-1994); Hungary (1995-1996)).
    - Reserve requirements more favorable for local currency liabilities (Belarus (since 2001); Croatia (2000); Pakistan (1998-2003)).
    - Loan-to-value (LTV) differentiation by currency (Hungary; Georgia (2019): LTV 85 percent local currency, 70 percent foreign currency loans).
    - Tighter provisioning on foreign currency loans (Albania (2007): additional provision of 5 percent on unhedged substandard and doubtful loans).

### Structural reforms to strengthen transmission
- Reduce the role of state-owned banks and the state in the economy:
  - Impose hard budget constraints on state-owned banks.
  - Improve SOE and state bank governance; appoint independent and qualified board members.
  - Reduce government funding and explicit financial assistance and guarantees to SOEs.
  - Consider eventual privatization of state-owned banks.
  - Strengthen bank supervision.
  - Promote market-determined energy pricing to avoid distortions and support efficient resource allocation.
- Rationale: ensuring firms and SOEs face hard budget constraints will make monetary policy changes more effective by affecting their financial decisions.

### Transition considerations for inflation targeting
- Recognize constraints during transition due to:
  - Large role of the state in the economy and banking sector,
  - Administrative prices,
  - Government lending programs at preferential interest rates,
  - Fiscal dominance.
- Policy implication: the central bank must consider these limitations in policy design and coordinate with the government to reduce the state's role.

### Conclusion — progress and remaining priorities
- Uzbekistan has made impressive progress toward meeting conditions for successful inflation targeting:
  - Improvements in the central bank’s institutional and governance set-up, forecasting capacities, monetary operations, policy formulation, and communication.
  - Further steps needed to enhance CBU governance.
- Key priorities to enhance monetary policy transmission:
  - Develop domestic capital and foreign exchange markets.
  - Reduce dollarization.
  - Reduce the role of the government in the economy and the banking sector.
- Final assessment: large government role in bank ownership and broader economy, combined with government lending policies, has constrained monetary policy transmission.

### Annex I (selected): Inflation targeting conditions — technical and financial infrastructure coding
- Data availability coding (scores):
  - 1.00 if all data were available, reliable, and of good quality.
  - 0.75 if all data available but one not reliable or of good quality.
  - 0.50 if all data available but a few not reliable or of good quality.
  - 0.25 if all data available but most highly unreliable or low frequency.
  - 0.00 if any data were missing.
- Systematic forecast process:
  - 1.00 if a periodic, systematic forecast process was already in place.
  - 0.00 if no such process was in place.
- Models capable of conditional forecasts:
  - 1.00 if yes.
  - 0.00 if no.
- Financial system health example coding:
  - Bank regulatory capital to risk-weighted assets:
    - 1.00 if regulatory capital in excess of 10 percent of risk-weighted assets.
    - 0.00 otherwise.
  - Stock market and private bond market capitalization variables scaled to UK = 1; stock market turnover ratio scaled to UK = 1.
  - Currency mismatch: self-assessment by domestically owned banks.

*Source: IMF Staff estimates.*

### 1.00 if the degree of mismatch was considered as “none” or “low.”

### wpiea2022229-print-pdf - 1.00 if the degree of mismatch was considered as “none” or “low.”

### Variable coding for mismatch and bond maturity
- Degree of mismatch coding:
  - 1.00 if the degree of mismatch was considered as “none” or “low.”
  - 0.50 if the degree of mismatch was considered as “some” or “moderate” mismatch.
  - 0.00 if the degree of mismatch was considered as “high.” 
- Maturity of bonds:
  - Considers the maximum maturity of actively traded bonds (in years) and divided by 30.
  - A country with actively traded 30-year bonds were assigned a value of 1 for this variable.

### Institutional Independence — coded indicators
- Fiscal obligation:
  - Considers if the central banks has an obligation, either implicit or explicit, to finance government budget deficits.
  - Answers were coded as:
    - 1.00 if no such obligation.
    - 0.00 Otherwise.
- Operational independence:
  - Considers whether the central bank had full “instrument independence,” giving it sole responsibility for setting the monetary policy instrument.
  - Answers were coded as:
    - 1.00 if yes.
    - 0.00 Otherwise.
- Central bank legal mandate:
  - Considers if there is an inflation-focused legal mandate.
  - Answers were coded as:
    - 1.00 if inflation is the only formal objective.
    - 0.50 if other objectives are specified, but inflation takes precedence.
    - 0.00 if other objectives are specified on an equal footing with inflation.
- Governor’s job security:
  - Average of Arnone and others (2005) questions on i) who appoints the central bank governor and term of the governor in office.
- Fiscal balance in percent of GDP:
  - Considering primary fiscal balance to GDP a variable was created indicating a lack of pressure to finance fiscal deficits.
  - This ratio was converted to a score ranging from 0 to 1 using a logistic transformation.
  - This was scaled in a way that a budget that was in balance or in surplus was assigned a value of 1, and a budget deficit in excess of 3 percent of GDP was assigned a value of 0.
- Public debt in percent of GDP:
  - Using the ratio of public debt to GDP, a variable was created equal to 1 minus the ratio of debt to GDP.
  - Thus, a country with no public debt would receive a value of 1, and one with a ratio of debt to GDP equal to or greater than 100 would receive a value of 0.
- Central bank independence:
  - “overall” measure (the average of political and economic) of central bank independence following Arnone and others (2005).
  - A value of 1 indicates complete independence while values closer to 0 indicate a diminishing degree of independence.

### Economic structure — coded indicators
- Exchange rate pass-through:
  - Considers the degree of exchange rate pass-through.
  - Answers were coded as:
    - 1.00 if low or no pass-through.
    - 0.50 if moderate pass-through.
    - 0.00 if high pass-through.
- Sensitivity to commodity prices:
  - Considers the degree of sensitivity of inflation to commodity price fluctuations.
  - Answers were coded as:
    - 1.00 if not sensitive.
    - 0.50 if sensitive.
    - 0.00 if very sensitive.
- Extent of dollarization:
  - Consider the extent of deposits dollarization. A variable was coded as:
    - 1.00 if the share of dollarized deposits was 50 percent or higher.
    - 0.50 if the share of dollarized deposits was 10-50 percent.
    - 0.00 if the share of dollarized deposits was less than 10 percent.
- Trade openness:
  - Considers the ratio of exports plus imports to GDP.
  - This ratio was then scaled to that of Singapore (the economy with the largest trade share relative to GDP) and subtracted from 1, resulting in an index that would equal 1 in the hypothetical case of a completely autarkic economy, and equal 0 for an economy with a degree of trade openness comparable to that of Singapore.

### Selected references cited in the section
- Angeloni, I. and Ehrmann, M., 2003, “Monetary transmission in the euro area: early evidence,” Economic Policy, Vol. 18, No. 37, pp. 469–501.
- Al-Mashat, R., Bulif, A., Dencer, N. N., Hledik, T., Holub, T., Kostanyan, A., Laxton, D., Nurbekyan, A., Portililo, R., and Wang, H., 2018, “An index for transparency for inflation targeting central banks: Application to the Czech National Bank,” IMF Working Paper No. WP/18/210.
- Batini, N., and Laxton, D. 2005, “Under What Conditions Can Inflation Targeting be Adopted? The Experience of Emerging Markets,” forthcoming in Monetary Policy Under Inflation Targeting ed. by Schmidt-Hebel and Mishkin.
- Freedman, C. and Otker-Robe, I., 2009, “Country Experiences with the Introduction and implementation of Inflation Targeting,” IMF Working Paper WP/09/161.
- García-Escribano, M., 2010, “Peru: Drivers of De-dollarization,” IMF Working Paper WP/10/169.
- International Monetary Fund, 2020, “Annual Report on Exchange Arrangements and Exchange Restrictions 2019”.
- Laurens, M. B., Eckhold, K., King, D., Mæhle, M. N. Ø., Naseer, A., & Durré, A., 2015. The journey to inflation targeting: Easier said than done the case for transitional arrangements along the road. Working Paper WP/15/136.
- Roger, S. 2009, “Inflation Targeting at 20: Achievements and Challenges,” IMF Working Paper WP/09/236.

*IMF Working Paper excerpt (as provided).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022229-print-pdf.pdf_