## _cr14243

## Source details

**Canonical URL:** [_cr14243](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2014/_cr14243.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2014/_cr14243.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2014/_cr14243.pdf.json)

---

### Income and income inequality: findings and analysis
- Kazakhstan experienced robust GDP growth since 2000—averaging 8 percent—which contributed to a sharp rise in per capita income.
- Growth incidence curve (2004–09): real GDP growth rates for groups in the middle and lower parts of the income distribution exceeded rates for the upper part, reducing the relative income gap.
- Poverty trends (2001–2012):
  - National poverty declined from 47 percent in 2001 to 4 percent in 2012.
  - Rural poverty declined from 59 percent in 2001 to 6 percent in 2012.
  - Urban poverty declined from 36 percent in 2001 to 2 percent in 2012.
- Rural poverty gap is wider in Kazakhstan relative to some neighbors.
- Regional disparities (2011):
  - Share of people below subsistence minimum ranges from 1.7 percent in Astana to over 10 percent in south Kazakhstan.
  - High poverty observed in both non-oil and oil-rich regions; oil sector functions as an enclave—capital intensive with limited job creation and weak spillovers.
- Policy measures taken:
  - In 2013 the government increased the tax burden on real estate and properties and raised tax rates on luxury goods by amending the tax code (changes effective January 2014).
  - Medium-term intent to strengthen progressivity of the income tax.
  - Expenditure measures include cash transfers to migrants to settle in target areas (particularly north Kazakhstan) and increased social orientation of the national budget focused on improving access to education and health for low-income families.
  - Specific plans: eliminate shortage of space in schools by 2017, provide free preschool education by 2020, and introduce a compulsory health insurance system in the medium term.
- Policy guidance and structural recommendation:
  - Fiscal redistribution should be consistent with fiscal sustainability.
  - Better targeting of transfers reduces fiscal cost and the tax burden required to finance them.
  - Promoting economic diversification and structural transformation (Kazakhstan 2050 Vision) requires sizeable investments in physical, human, and institutional capital, prioritizing infrastructure investment and enhancing investment efficiency through structural reforms and selective financial support.

### Employment: findings, risks, and projections
- Unemployment and labor force (2013):
  - Labor force in 2013: 9.1 millions.
  - Employed in 2013: 8.6 millions.
  - Unemployed in 2013: 0.5 millions.
  - Unemployment rate in 2013: 5.2 percent.
- Unemployment trends:
  - Unemployment rate declined rapidly since 2000; 5.2 percent in 2013, less than half of its early-2000s level.
  - Youth unemployment fell substantially owing to targeted interventions, including expanded vocational and training programs.
- Female labor force participation:
  - FLFP reported at 67 percent in 2012, higher than emerging market peers.
- Quality of employment:
  - About 30 percent of employed people engage in vulnerable employment (unpaid family workers and own-account workers).
  - If measured vulnerable employment rate were reduced to the emerging market average of 20 percent, the measured unemployment rate would be above 10 percent.
  - Labor statistics quality concerns due to ambiguous treatment of self-employment; substantial portion of rural labor force recorded as self-employed.
- Sectoral and structural issues:
  - Weak job creation since the global crisis; manufacturing employment anemic despite industrialization program; employment in agriculture declining.
  - Agriculture accounted for more than a quarter of total employment in 2011 while contributing about 6 percent of GDP.
  - Long-term employment elasticity in Kazakhstan: 0.23 (Text Table 1).
  - Contributing factors: capital-intensive oil sector with limited spillovers; underdeveloped private sector constrained by state control and limited access to commercial credit; oversized public sector (about 22 percent of employed people work in public entities); skills mismatch—only 41 percent of Kazakhstani firms provide formal training (comparators: Russia 52 percent, Poland 61 percent, Malaysia 50 percent).
- Medium-term projections and requirements:
  - Projected labor force in 2019: 9.9 millions.
  - Projected new entrants to the labor force up to 2019: 0.8 millions.
  - Total unemployed and new entrants in 2019: 1.2 millions.
  - Employment elasticity: 0.23.
  - Change in employment required to absorb entrants: 9.0 percent.
  - Annual growth in employment required to absorb entrants: 1.4 percent.
  - Required real GDP growth to absorb entrants for 2014: –196.4%
  - Average real GDP growth, 2003–13: 7.1%.
  - Average real GDP growth projected, 2014–19: 5.2%.
  - IMF staff projection: unemployment rate envisaged to rise to 6.3 percent in 2019 from 5.2 percent in 2013, under current policies; unemployment could rise above 7 percent linked to capital-intensive oil projects.
- Policy responses and programs:
  - Government “Road Map” (Regional Employment and Retraining Strategy, with ILO) adopted to contain sudden rises in unemployment via public works.
  - Road Map components: additional financing for housing and utilities projects; construction and maintenance of local roads; maintenance of social infrastructure (schools and hospitals); creation of social jobs (including 50 percent co-financing of salaries for selected target groups); youth internships and vocational training and retraining.

### Labor market outlook and structural reform priorities
- Growth and employment requirements:
  - To maintain unemployment at the current level (around 5 percent), medium-term GDP growth should be boosted to 6½ percent, 1¼ percentage points higher than envisaged in staff’s macroeconomic projection.
  - Increasing employment-output elasticity to the middle-income country group average (0.25) could support keeping unemployment at the current level without boosting medium-term growth.
- Structural reform priorities:
  - Improve the business climate to energize the private sector and boost employment; support SMEs and enterprise development.
  - Address weaknesses in education to reduce skill mismatches; align secondary education with market needs, increase private sector involvement, and balance higher education and vocational training.
  - Strengthen the rule of law and lower the role of the state in the economy.
- Authorities’ actions and international cooperation:
  - Revamping a college internship program and a job placement program to bolster youth employment.
  - Plan to accelerate structural reforms in cooperation with development partners (ADB, EBRD, World Bank) focusing on financial sector, SMEs, skills, investment climate, regional development, and institutional reforms.
- Inclusivity and fiscal policy:
  - Growth has been broadly inclusive; income inequality and unemployment compare favorably to peers, but rural poverty rates remain higher than some poorer regional peers.
  - Fiscal policy can help reduce income inequality through better targeting of transfers to reduce fiscal cost and tax burdens.

### Monetary policy, VAR and Taylor-rule findings
- VAR specification (2003: M1–2014: M2): endogenous variables—refinance rate, deposit rate, lending rate, and inflation; exogenous variables—international food and energy prices.
- VAR empirical findings:
  - Inflation does not respond to changes in any of the interest rates.
  - The refinance rate responds to shocks to inflation.
  - "the NBK policy (refinance) rate does not affect money market interest rates."
  - "changes in the deposit, lending, or refinance rates do not induce any change in the inflation rate."
  - "the refinance rate responds to shocks in inflation, rather than the other way round."
  - Conclusion: current monetary policy instruments used by the NBK are unable to signal the stance of monetary policy and are not effective in ensuring price stability.
  - Results robust to generalized IRFs, different lag specifications, and inflation Granger causes the refinance rate.
- Taylor-rule specification and estimation (2003: Q1–2013: Q3):
  - Baseline: it = rn + πt + 0.5 (πt - π*t) + 0.5 ȳt.
  - Backward-looking and forward-looking specifications estimated; forward-looking estimated using two-step Heckman (1976) procedure to correct for endogeneity.
- Taylor-rule estimation results (Text Table 1 summary):
  - Backward-looking specification indicates a statistically significant response of the refinance rate to past inflation and past output gap; strong interest rate smoothing (high ρ).
  - Forward-looking specification: current interest rates do not respond to future changes in inflation or output gap (variables statistically insignificant after correction).
  - No response to changes in the exchange rate (d not statistically significant).
  - Consistency: both VAR and Taylor-rule analyses indicate lack of causality from the policy rate to inflation.
- Assessment:
  - Monetary policy in Kazakhstan over the sample period is estimated to be backward-looking.
  - The NBK’s official refinancing rate does not fully signal the stance of monetary policy; weak transmission from refinance rate to money market rates and weak influence on inflation.
  - Monetary policy needs to become more forward-looking to contain inflationary pressures and anchor expectations.

### Dollarization: determinants and policy measures
- Facts and implications:
  - Financial dollarization: ratios of foreign currency deposits and loans to total remain high despite a gradual fall prior to recent devaluation.
  - Dollarization limits monetary policy autonomy, weakens transmission mechanisms, increases need for reserve cushions, deepens exchange rate pass-through, and raises credit/solvency/liquidity risks.
- Determinants (model, 2000: Q1–2014: Q1):
  - ∆depdollarizationt = β0 + β1 interest-differentialt-1 + β2 exchanget-1 + β3 exchange-volatilityt-1 + β4 inflationt-1 + β5 inflation-volatilityt-1 + β6 financial-developmentt-1 + β7 asymmetry-dummyt-1 + εt
  - Dependent variable: percentage change in deposit dollarization; independent variables lagged one period.
- Empirical findings:
  - "Inflation volatility and asymmetry of exchange rate policy toward depreciation drive deposit dollarization in Kazakhstan."
  - Inflation volatility matters more than the level of inflation.
  - The asymmetric exchange rate policy—allowing depreciations but resisting appreciations—has been a major incentive for residents to hold foreign currency deposits.
- Macro-level de-dollarization recommendations:
  - Adopt an inflation targeting regime with flexible exchange rates and absence of fiscal dominance.
  - Credible disinflation policies reduce dollarization over time.
  - Policy examples: widen exchange rate bands (Poland late 1990’s); increase FX loan interest rates (Croatia); raise domestic deposit interest rates above foreign-currency rates (Turkey, Egypt, Hungary, Poland early 1990’s); deepen local-currency securities markets with credible indexation (Chile, Mexico 1980’s; Bolivia, Israel, Turkey early 2000’s); remove administrative controls on interest rates (Egypt, Lithuania, Poland early 1990’s); apply unbiased taxation on foreign-currency vs local-currency financial income.

### Micro-prudential measures to reduce dollarization and sequencing to IT
- Micro-prudential measures discussed and suggested:
  - Minimum capital requirements for foreign currency-induced credit risk.
  - Request credit bureaus to provide currency-specific debt information.
  - Impose higher reserve requirements on foreign currency deposits (examples: Armenia, Belarus, Bolivia, Croatia, Peru, Romania, Serbia, Turkey in the 2000’s).
  - Remunerate reserve requirement on local currency deposits at a higher rate than for foreign currency deposit reserve requirement.
  - Hold reserve requirements for foreign currency deposits in local currency (examples: Croatia, Haiti, Serbia in the 2000’s).
  - Tighter provisioning requirements on foreign currency loans (examples: Albania, Croatia, Mozambique mid 2000’s).
  - Raise insurance premiums on dollar deposits (Peru).
  - Develop markets for instruments to hedge currency risks (Peru, Israel).
  - Require banks to hold higher liquid-asset ratios on short-term liabilities, higher for foreign currency than domestic currency.
- Conclusion and recommended near-term steps for transition to inflation targeting (IT):
  - Introduce a clear policy rate instrument supported by open market operations to anchor key interbank rates around the NBK’s policy rate.
  - Gradual widening of the exchange rate band.
  - Introduce micro-prudential measures, including capital and reserve requirements, to increase confidence in the domestic currency.
  - Ensure multiple objectives of financial or exchange rate stability do not conflict with the ultimate goal of price stability.
  - Enhance open communication of NBK’s policy intentions and operations to help anchor expectations.

### External sector and trade balance findings
- Reserves and reserve adequacy:
  - Official reserves improved slightly in 2014: Q1 after declines in 2012–13.
  - Total foreign exchange reserves including national oil fund (NFRK) assets are well above the IMF’s Assessing Reserve Adequacy (ARA) metric.
  - ARA suggested reserve adequacy range is 100–150 percent of the metric; combined international reserves and NFRK assets are together well above the ARA metric.
- External debt projections and shocks:
  - External debt to GDP ratio projected to decline from 73 percent in 2014 to 65 percent in 2019.
  - A one-time real depreciation of 30 percent in 2015 would raise the external debt to GDP ratio to 97 percent in 2019 versus 65 percent under the baseline.
  - A combined 0.25 standard deviation shock to the interest rate, growth rate and non-interest current account balance would raise the external debt to GDP ratio to 75 percent in 2019.
  - Decline in external debt driven by noninterest external current account surpluses and nondebt creating capital inflows and automatic debt dynamics; capital inflows expected to remain stable.
- Exchange rate devaluation and trade balance (2003: Q1–2013: Q3):
  - Short-run effects: mixed across trading partners; short-run improvements with Austria, France, Romania, and Russia (these represent about 35 percent of total external trade); J-curve effects for Canada and Switzerland (about 3 percent of total trade).
  - Long-run effects: little overall long-run effect; statistically significant long-run relationship between exchange rate and trade balance found in only a couple of countries (France and Romania, about 10 percent of total trade).
  - Half-life estimates: trade balance adjusts within two to three quarters on average; fastest for France and Romania (within one quarter), slowest for Canada (more than six quarters).
  - About 40 percent of external trade (China, Germany, Italy, Japan, Netherlands, Turkey, United States) is unaffected by currency devaluation.
  - Policy implication: with oil exports around 80 percent of total exports, devaluation of the tenge is unlikely to have much effect on the export side; short-run improvement may occur mostly via reduced imports; broader need for a more diversified and competitive export structure.

*Prepared by SeokHyun Yoon. Source: IMF staff estimates and projections as presented in the chapter.*

### 1. Income and Income Inequality __________________________________________________________ 3

### 1. Income and Income Inequality

### A. Income Inequality: findings and analysis
- Kazakhstan experienced robust GDP growth since 2000—averaging 8 percent—which contributed to a sharp rise in per capita income (Figure 1).
- Growth incidence curve (2004–09) indicates a decrease in inequality: real GDP growth rates for groups in the middle and lower parts of the income distribution exceeded rates for the upper part of the distribution, reducing the relative income gap (Figure 2).
- Poverty trends (2001–2012):
  - National poverty declined from 47 percent in 2001 to 4 percent in 2012.
  - Rural poverty declined from 59 percent in 2001 to 6 percent in 2012.
  - Urban poverty declined from 36 percent in 2001 to 2 percent in 2012 (Figure 3).
- Rural poverty gap (defined as rural poverty rate, in percentage of the rural population, less national poverty rate, in percentage of the nation's population) is wider in Kazakhstan relative to some neighbors (Figure 4).
- Regional disparities in income and subsistence-level poverty (2011):
  - Share of people below subsistence minimum ranges from 1.7 percent in Astana to over 10 percent in south Kazakhstan (Figure 5).
  - High poverty observed in both non-oil and oil-rich regions; oil sector functions as an enclave—capital intensive with limited job creation and weak spillovers.
- Policy measures taken:
  - In 2013 the government increased the tax burden on real estate and properties and raised tax rates on luxury goods by amending the tax code (changes effective January 2014).
  - Medium-term intent to strengthen progressivity of the income tax.
  - Expenditure measures include cash transfers to migrants to settle in target areas (particularly north Kazakhstan) and increased social orientation of the national budget focused on improving access to education and health for low-income families.
  - Specific plans: eliminate shortage of space in schools by 2017, provide free preschool education by 2020, and introduce a compulsory health insurance system in the medium term.
- Policy trade-offs and guidance:
  - Fiscal redistribution should be consistent with fiscal sustainability to support growth and the capacity to finance higher redistribution over the longer term.
  - Better targeting of transfers reduces fiscal cost and the tax burden required to finance them.
- Structural recommendation:
  - Promoting economic diversification and structural transformation (as envisaged in the Kazakhstan 2050 Vision) is essential—requires sizeable investments in physical, human, and institutional capital, prioritizing infrastructure investment and enhancing investment efficiency through structural reforms and selective financial support.

### B. Employment: findings, risks, and projections
- Unemployment trends:
  - Unemployment rate declined rapidly since 2000; unemployment rate was 5.2 percent in 2013, less than half of its early-2000s level (Figure 6).
  - Youth unemployment fell substantially owing to targeted interventions, including expanded vocational and training programs.
- Female labor force participation:
  - Female labor force participation (FLFP) reported at 67 percent in 2012, higher than emerging market country peers (Figure 7). High FLFP contributed to output growth and reduced income inequality.
- Quality of employment concerns:
  - High share of self-employment and vulnerable employment: among employed people, about 30 percent engage in vulnerable employment (unpaid family workers and own-account workers) (Figure 9).
  - If Kazakhstan’s measured vulnerable employment rate were reduced to the emerging market average of 20 percent, the measured unemployment rate would be above 10 percent (Figure 10).
  - Labor statistics quality concerns due to ambiguous treatment of self-employment; substantial portion of rural labor force recorded as self-employed.
- Job creation and sectoral composition:
  - Relationship between job creation and growth has been weak since the global crisis (Figure 11).
  - Job creation in manufacturing is anemic despite accelerated industrialization program; employment in agriculture has been continuously falling due to unproductive farm structure (Figure 12).
  - Agriculture accounted for more than a quarter of total employment in 2011 while contributing about 6 percent of GDP.
- Structural factors limiting employment elasticity:
  - Long-term employment elasticity in Kazakhstan is low relative to comparators (Text Table 1: Kazakhstan 0.23; regional and income-group comparisons shown).
  - Contributing factors:
    - Capital-intensive oil sector with limited employment spillovers.
    - Underdeveloped private sector constrained by state control, business climate distortions, limited access to commercial credit, and lack of long-term investment.
    - Oversized public sector: about 22 percent of employed people work in public entities; public sector provides higher compensation and benefits, crowding out private employment.
    - Skills mismatch: excess demand for workers with higher and vocational education and excess supply of workers with general secondary education and below; only 41 percent of Kazakhstani firms provide formal training (comparators: Russia 52 percent, Poland 61 percent, Malaysia 50 percent).
- Medium-term unemployment outlook and projections:
  - Demographics and labor supply:
    - Labor force in 2013: 9.1 millions.
    - Unemployment rate in 2013: 5.2 percent.
    - Unemployed in 2013: 0.5 millions.
    - Employed in 2013: 8.6 millions.
    - Projected labor force in 2019: 9.9 millions.
    - Projected new entrants to the labor force up to 2019: 0.8 millions.
    - Total unemployed and new entrants in 2019: 1.2 millions.
  - Employment elasticity and required changes:
    - Employment elasticity: 0.23.
    - Change in employment required to absorb entrants: 9.0 percent.
    - Annual growth in employment required to absorb entrants: 1.4 percent.
    - Required real GDP growth to absorb entrants for 2014: –196.4% (text table entry preserved as presented).
    - Average real GDP growth, 2003–13: 7.1%.
    - Average real GDP growth projected, 2014–19: 5.2%.
  - IMF staff projections indicate unemployment rate envisaged to rise to 6.3 percent in 2019 from 5.2 percent in 2013, under current policies (Figure 13).
  - Scenario caveat: medium-term growth prospects are closely linked to capital-intensive oil projects (e.g., Kashagan); unemployment could rise above 7 percent.
- Policy responses and programs:
  - Government “Road Map” (Regional Employment and Retraining Strategy, in cooperation with ILO) adopted as part of anti-crisis program to contain sudden rises in unemployment through creation of public works.
  - Road Map objectives:
    - Contain increase in unemployment via short-term employment and job creation in public works and social programs.
    - Rehabilitate social infrastructure and facilities to support sustainable development.
  - Program components include:
    - Additional financing for housing and utilities projects.
    - Construction and maintenance of local roads.
    - Maintenance of social infrastructure (schools and hospitals), including outside main cities.
    - Creation of social jobs (including 50 percent co-financing of salaries for selected target groups).
    - Youth internships and vocational training and retraining.

*Prepared by SeokHyun Yoon. Source: IMF staff estimates and projections as presented in the chapter.*

### 15.       A combination of reforms and higher growth is essential to absorb new entrants to

### _cr14243 - 15.       A combination of reforms and higher growth is essential to absorb new entrants to

### Labor market outlook and growth requirements
- To maintain unemployment at the current level (around 5 percent), medium-term GDP growth should be boosted to 6½ percent, 1¼ percentage points higher than envisaged in the staff’s macroeconomic projection.
- Staff simulations indicate that increasing the employment-output (employment elasticities) responsiveness could substitute for higher growth:
  - Raising the employment-output elasticity to the middle-income country group average (i.e., 0.25) would support keeping unemployment at the current level without boosting medium-term economic growth.

### Structural reform priorities to foster job creation
- Higher economic growth alone is unlikely to substantially reduce overall or youth unemployment without structural reforms improving labor market responsiveness.
- Priority reform areas and expected effects:
  - Improving the business climate to energize the private sector and boost employment over the medium term; supporting enterprise development by improving infrastructure and promoting small and medium-sized enterprises (SMEs).
  - Addressing weaknesses in the education environment to reduce skill mismatches; aligning secondary education with market needs, increasing private sector involvement, and striking the right balance between higher education and vocational education training.
  - Strengthening the rule of law and lowering the role of the state in the economy to promote a more vigorous private sector.

### Authorities’ actions and international cooperation
- Authorities are revamping a college internship program and a job placement program to bolster youth employment and reduce labor market mismatches.
- The authorities plan to accelerate structural reforms in close cooperation with development partners (the Asian Development Bank, the European Bank for Reconstruction and Development, and the World Bank), focusing on: the financial sector, SMEs, skills, the investment climate, regional development, and institutional reforms.

### Inclusivity, fiscal policy, and overarching reform agenda
- Kazakhstan’s economic growth has been broadly inclusive; income inequality and unemployment compare favorably to peers, but rural poverty rates remain higher than some poorer regional peers.
- Recent efforts to reduce income inequality via faster employment growth have been relatively weak.
- Fiscal policy can help reduce income inequality through better targeting of transfers to reduce fiscal cost and tax burdens while achieving distributional objectives more efficiently.
- An ambitious structural reform agenda is paramount to becoming a dynamic emerging market economy and ensuring sustainable and inclusive growth; key priorities include strengthening human capital and institutions and lowering the role of the state in a more diversified economy.

### External sector and reserve adequacy (key findings)
- Official reserves improved slightly in 2014: Q1 after declines in 2012–13; total foreign exchange reserves including national oil fund (NFRK) assets are well above the IMF’s Assessing Reserve Adequacy (ARA) metric.
- The ARA suggested reserve adequacy range is 100–150 percent of the metric; combined international reserves and NFRK assets are together well above the ARA metric.
- External debt dynamics and projections:
  - External debt to GDP ratio projected to decline from 73 percent in 2014 to 65 percent in 2019.
  - A one-time real depreciation of 30 percent in 2015 would raise the external debt to GDP ratio to 97 percent in 2019 versus 65 percent under the baseline.
  - A combined 0.25 standard deviation shock to the interest rate, growth rate and non-interest current account balance would raise the external debt to GDP ratio to 75 percent in 2019.
  - The decline in external debt is mainly driven by noninterest external current account surpluses and, to a lesser extent, by nondebt creating capital inflows and automatic debt dynamics; capital inflows are expected to remain stable.

*Source: IMF staff report excerpt (Republic of Kazakhstan).*

### 5.      This section estimates a multivariate vector autoregressive (VAR) model to examine

### 5.      This section estimates a multivariate vector autoregressive (VAR) model to examine

### VAR specification and estimation period
- Model: multivariate VAR of order p (Equation (1)).
- Endogenous variables Yt: the refinance rate, deposit rate, lending rate, and inflation.
- Exogenous variables Xt: international food and energy prices.
- Sample period: 2003: M1–2014: M2.
- Errors: εt is a vector of iid error terms.
- Lag order determined by standard lag selection criteria.

### VAR empirical findings (impulse response functions and causality)
- Main empirical messages:
  - Inflation does not respond to changes in any of the interest rates.
  - The refinance rate responds to shocks to inflation.
- Specific findings from generalized impulse response functions (IRF):
  - First: "the NBK policy (refinance) rate does not affect money market interest rates." Shown by statistically insignificant responses of deposits or lending rates to shocks in the refinance rate.
  - Second: "changes in the deposit, lending, or refinance rates do not induce any change in the inflation rate," evident from statistically insignificant responses of inflation to generalized one standard-deviation shocks in all of these interest rates.
  - Third: "the refinance rate responds to shocks in inflation, rather than the other way round." This response is positive and statistically significant, indicating that higher inflation induces the NBK to raise its refinance interest rate; shocks to the policy rate do not affect inflation.
- Policy implication from VAR results:
  - "These combined results suggest that the current monetary policy instruments used by the NBK are unable to signal the stance of monetary policy and are not effective in ensuring price stability."
- Robustness:
  - Results hold using generalized IRFs (not sensitive to ordering).
  - Results robust to different lag specifications.
  - Inflation Granger causes the refinance rate.

### Is Monetary Policy Forward-Looking? — Background and Taylor-rule specification
- NBK objective: achieve price stability and keep annual inflation within a 6–8 percent range.
- Taylor-rule baseline (Equation (1) in that section):
  - it = rn + πt + 0.5 (πt - π*t) + 0.5 ȳt
  - where it is the short-term policy interest rate, rn is the natural rate of interest, πt is inflation, π*t is the inflation target, and ȳt is the output gap.
- Extended specifications:
  - Interest rate smoothing: it = (1-ρ) i*t + ρ it-1 + εt (Equation (2)).
  - Target rate specification: i*t = rn + πt+k + (β-1)(πt+k - π*t) + γ ȳt+k + η ∆xt (Equation (3)).
  - Output gap: ȳt = GDPt - GDPtrend_t (Hodrick-Prescott) (Equation (4)).
- Testable empirical model (Equation (6)):
  - it = a + b πt±1 + c ȳt±1 + d ∆xt + ρ it-1 + εt
  - Backward-looking specification: k = -1; forward-looking specification: k = +1.

### Taylor-rule estimation methodology
- Estimation period: 2003: Q1–2013: Q3.
- Estimation approach:
  - Backward-looking specification estimated by OLS.
  - Forward-looking specification estimated using a two-step Heckman (1976) procedure (following Kim (2006) and Kim and Nelson (2006)) to correct for endogeneity:
    - First step: regress inflation and output gap on instruments (four lags of inflation, output gap, global commodity prices, and interest rate) and obtain residuals.
    - Second step: include residuals in original Taylor-rule specification; forward-looking model then estimated by OLS.
  - ADF unit root tests used to determine levels/differences.
  - Robust standard errors reported.

### Taylor-rule estimation results (Text Table 1 as reported)
- Text Table  1. Kazakhstan
  - Backward0.0030.1190.012      -0.026       0.9170.90
    (0.004)   (0.066)*   (0.005)**   (0.018)   (0.061)***
  - Forward     0.001      0.080      -0.002     -0.034       0.961        0.86     
           (0.005)    (0.067)           (0.006)           (0.021)           (0.071)***
  - Robus t s ta ndard errors are in parentheses.
  - *Signi fi cant at 10 percent; * *Si gnif icant at 5 percent; * ** Si gnif ica nt a t 1 per cent

### Interpretation of Taylor-rule results
- Monetary policy in Kazakhstan over the sample period is estimated to be backward-looking:
  - Backward-looking specification: refinance rate shows a statistically significant response to past inflation and past output gap—NBK raises its refinance rate in the current period in response to higher inflation or overheating in the previous period.
  - Forward-looking specification: current interest rates do not respond to future changes in inflation or output gap (variables statistically insignificant after correcting for endogeneity).
- Both specifications indicate:
  - A strong interest rate smoothing effect (high ρ).
  - No response to changes in the exchange rate (d not statistically significant).
- Consistency with VAR results: both analyses indicate lack of causality from the policy rate to inflation.

### Does Dollarization Hinder the Move Toward Inflation Targeting? — Overview
- Rising dollarization ratios can complicate domestic monetary policy and delay adoption of an effective monetary policy framework.
- The report examines:
  - Extent of dollarization in Kazakhstan relative to emerging markets.
  - How a dollarized banking system complicates macroeconomic policy.
  - Drivers of dollarization in Kazakhstan and potential de-dollarization measures.

### Dollarization: facts and implications
- Financial dollarization in Kazakhstan: ratios of foreign currency deposits and loans to total remain high despite a gradual fall prior to recent devaluation.
- Dollarization complicates macro policy and increases financial risks:
  - Limits monetary policy autonomy and weakens transmission mechanisms.
  - Calls for extra reserve cushions and deepens exchange rate pass-through on inflation in managed exchange rate regimes.
  - Raises credit, solvency, and liquidity risks, including divergence between onshore and offshore interest rates on dollar deposits.

### Determinants of dollarization in Kazakhstan (model and sample)
- Model for ∆depdollarizationt (Equation (1) in that section):
  - ∆depdollarizationt = β0 + β1 interest-differentialt-1 + β2 exchanget-1 + β3 exchange-volatilityt-1 + β4 inflationt-1 + β5 inflation-volatilityt-1 + β6 financial-developmentt-1 + β7 asymmetry-dummyt-1 + εt
  - Dependent variable: percentage change in deposit dollarization.
  - Independent variables (lagged one period): differential between domestic-currency and dollar deposit interest rates; exchange rate; exchange rate volatility; inflation; inflation volatility; credit-to-private-sector-to-GDP (financial development); dummy for exchange-rate asymmetry (1 for depreciation, 0 for appreciation).
- Sample period: 2000: Q1–2014: Q1.
- Estimation: OLS with robust standard errors; variables included in levels or first-differences depending on ADF tests.

### Determinants: empirical findings
- "Inflation volatility and asymmetry of exchange rate policy toward depreciation drive deposit dollarization in Kazakhstan."
  - Inflation volatility matters more than the level of inflation in explaining higher deposit dollarization.
  - The asymmetric exchange rate policy—allowing depreciations but resisting appreciations—has been a major incentive for residents to hold foreign currency deposits.
- The study uses estimated coefficients to decompose contributions of these variables to changes in deposit dollarization over the sample.

### De-dollarization policies (lessons and suggested measures)
- Successful de-dollarization requires a combination of macroeconomic stabilization and complementary microeconomic measures; gradual market-based policies are preferable to forced measures.
- Macro-level policy recommendations:
  - Adopt an inflation targeting regime with flexible exchange rates and absence of fiscal dominance to support market-driven de-dollarization.
  - Credible and successful disinflation policies reduce dollarization over time.
  - Policy examples cited from country experiences:
    - Widen exchange rate bands (example: Poland in late 1990’s), increase foreign exchange loan interest rates (example: Croatia), raise domestic deposit interest rates above foreign-currency rates (examples: Turkey, Egypt, Hungary, and Poland in the early 1990’s).
    - Deepen domestic financial markets by introducing local-currency-denominated securities with credible indexation systems (examples: Chile and Mexico in the 1980’s; Bolivia, Israel, and Turkey in the early 2000’s).
    - Remove administrative controls on interest rates (examples: Egypt, Lithuania, and Poland in the early 1990’s).
    - Apply unbiased taxation on income earned from foreign currency deposits, bonds or other financial transactions versus local currency taxes.

*Source: _cr14243 - 5.*

### 24.      At the micro level, supportive prudential regulations to make the local currency more

### _cr14243 - 24.      At the micro level, supportive prudential regulations to make the local currency more

### Micro-prudential measures to reduce dollarization and increase use of local currency
- Comprehensive prudential measures discussed by Cayazzo and others (2006) and Kokenyne and others (2010) include:
  - Minimum capital requirements for foreign currency-induced credit risk.
  - Requesting credit bureaus to provide currency-specific information on all debt.
- Suggested policies at the micro level:
  - Imposing higher reserve requirements on foreign currency deposits as in Armenia, Belarus, Bolivia, Croatia, Peru, Romania, Serbia, and Turkey in the 2000’s (see García-Escribano and Sosa (2011) and Kokenyne and others (2010)).
  - Remunerating the reserve requirement on local currency deposits at a higher rate than for the foreign currency deposit reserve requirement. See Kokenyne and others (2010) on the cases of Croatia, Israel, Nicaragua, and Romania in the 2000’s.
  - Holding reserve requirements for foreign currency deposits in local currency. Examples include Croatia, Haiti, and Serbia in the 2000’s.
  - Tighter provisioning requirements on foreign currency loans as in Albania, Croatia, and Mozambique in the mid 2000’s. Banks may be required to carry routine evaluations of currency risks, or set up reserves as a percentage of foreign currency credit that has not been evaluated (see García-Escribano and Sosa (2011) on the experience of Latin American countries).
  - Raising insurance premiums on dollar deposits, see IMF (2007) and García-Escribano (2010) on the Peruvian experience.
  - Developing markets for instruments to hedge currency risks as in Peru and Israel.
  - Requiring banks to hold liquid assets of certain percentages on their short-term liabilities, with higher requirements for foreign currency than for domestic currency liabilities. See Kokenyne and others (2010) for examples from Angola, Croatia, Cyprus, Egypt, Lebanon, and Turkey in the 1990’s and 2000’s, and Rennhack and Nozaki (2006) and García-Escribano and Sosa (2011) on the experience of Latin American countries in the early 1990’s.

### Conclusion and policy recommendations on monetary policy regime and sequencing
- Key assessment:
  - The chapter finds ample room for improvements before the NBK is ready to adopt IT.
  - Current issues: ineffectiveness of the policy rate, dependence on the exchange rate as a dominant monetary policy instrument, and extensive dollarization undermine the framework’s ability to ensure price stability and counteract domestic and external shocks.
  - The NBK’s official refinancing rate does not fully signal the stance of monetary policy, with weak transmission from the refinance rate to money market interest rates and weak influence on inflation.
  - Monetary policy needs to become more forward-looking to contain inflationary pressures and anchor expectations.
  - Macroeconomic stabilization policies, a deeper domestic financial market, and micro-prudential measures are needed to arrest dollarization and increase confidence in the local currency.
- Recommended near-term steps for a smooth transition to inflation targeting (IT):
  - Introducing a clear policy rate instrument supported by open market operations to help ensure that key interbank rates are anchored around the NBK’s policy rate. If current exchange rate management policies continue, large foreign exchange interventions will be required, complicating operations when combined with large open market operations (see Epstein and Portillo (2014)).
  - A gradual widening of the exchange rate band.
  - Introducing micro prudential measures, including certain capital and reserve requirements, to increase confidence in the domestic currency.
  - Ensuring that the multiple objectives of financial or exchange rate stability do not conflict or override the ultimate goal of price stability.
  - Enhancing open communication of the NBK’s policy intentions and operations to help anchor expectations and ensure a smooth transition to a new policy interest rate.

### Exchange rate devaluation and the trade balance: empirical findings (2003: Q1–2013: Q3)
- Scope and methodology:
  - Examines short-run versus long-run effects of devaluation on the trade balance in Kazakhstan over the 2003: Q1–2013: Q3 period using the traditional J-curve literature.
  - Analysis at the bilateral level between Kazakhstan and 13 top trading partners covering more than 80 percent of Kazakhstan’s overall trade.
  - Three main destinations cover almost two-thirds of Kazakhstan’s total external trade: China, Russia and the Euro area countries.
  - Uses an Autoregressive Distributed Lag (ARDL) co-integration approach (Pesaran and others (2001)) to distinguish short-run and long-run effects.
- Short-run empirical results:
  - Mixed effects of currency devaluation on the trade balance with different trading partners.
  - Short-run improvements in the trade balance observed with Austria, France, Romania, and Russia. Total combined trade with these countries represents about 35 percent of Kazakhstan’s total external trade. Based on regression results and lag selection criteria, the short-run improvement may last from one quarter up to a year. The improvement likely originates from a reduction in imports rather than an expansion in exports, as Kazakhstan’s exports to these countries are dominated by oil products denominated in dollars.
  - A J-curve (negative short-run) effect observed for Canada and Switzerland. These two countries together represent about 3 percent of Kazakhstan’s total external trade.
- Long-run empirical results:
  - Currency devaluation appears to have little long-run effects overall.
  - Evidence of co-integration between the trade balance and its determinants in almost half of Kazakhstan’s trading partners, but a statistically significant long-run relationship between exchange rate and trade balance is found in only a couple of countries.
  - Half-life estimates suggest the trade balance would adjust to exchange rate devaluation within two to three quarters on average, with the response fastest for France and Romania (within one quarter) and slowest for Canada (more than six quarters).
  - Real exchange rate devaluation, holding all else constant, has a positive long-run impact on the trade balance with France and Romania only. These two countries represent about 10 percent of Kazakhstan’s total trade. In most other models, exchange rate coefficients had the expected positive sign but were statistically insignificant.
- Trade partners unaffected by devaluation:
  - About 40 percent of Kazakhstan’s external trade is unaffected by currency devaluation. China, Germany, Italy, Japan, Netherlands, Turkey, and the United States show neither short- nor long-run devaluation effects on the trade balance, controlling for domestic and foreign incomes. These countries combined represent around 40 percent of Kazakhstan’s total external trade.
- Robustness and diagnostics:
  - Estimated models are mostly free from serial correlation or misspecification as indicated by the Lagrange Multiplier (LM) test and Ramsey’s RESET test.
  - CUSUM and CUSUMSQ tests of Brown and others (1975) show parameter stability (example shown for the model with France).
- Policy implication from trade analysis:
  - Given that Kazakhstan’s oil exports constitute around 80 percent of its total exports, the recent devaluation of the tenge is unlikely to have much effect on the export side of the trade balance going forward.
  - Some short-run improvement in the trade balance may occur mostly via reduced imports.
  - The exchange rate devaluation, as a tool to gain external competitiveness, may have limited success; what is needed is a more diversified and competitive export structure.

*Source: _cr14243 - 24.      At the micro level, supportive prudential regulations to make the local currency more (PDF chapter/section).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2014/_cr14243.pdf_