## cr18278

## Source details

**Canonical URL:** [cr18278](https://www.imf.org/-/media/files/publications/cr/2018/cr18278.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2018/cr18278.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2018/cr18278.pdf.json)

---

### Advanced Practices in Risk Management

- Fiscal risk diagnosis and recent history
  - Kazakhstan faces fiscal risk from volatility of oil prices, financial sector exposure, state-owned enterprise (SOE) risks, and other macroeconomic and specific shocks.
  - Oil price shocks of 2008-09 and 2014-15 led to:
    - large declines in revenues and exports;
    - costly bailouts of highly dollarized banks in 2009 and 2016;
    - a large transfer to KazMunaiGas in 2015 to ensure timely debt service payments.
  - The oil price shocks in 2008-09 and 2014-15 led to a decline of nominal GDP by nearly two standard deviations relative to the average; best practice identifies a macro shock when nominal GDP falls by one standard deviation (IMF 2016).

- Rationale for comprehensive fiscal risk management
  - Analyzing and managing fiscal risks:
    - helps credibility and sustainability of public finances;
    - supports macro stability and strengthens central forecasts;
    - informs implications for government liquidity, financing needs, and solvency.
  - Identified current practice gaps:
    - risk management practices are often incomplete, fragmented, and lacking quantitative analysis;
    - no systematic analysis of forecast errors for macroeconomic assumptions or fiscal forecasts;
    - line ministries provide statements of risks but typically lack quantified risk information;
    - MoF central reporting is not analyzed from a risk perspective or summarized into a statement of fiscal risks.

- Four-step framework recommended for fiscal risk management
  - Step 1: Identifying and Quantifying Risks — identify, assess, and disclose risks; quantified assessments and disclosure create a feedback loop that strengthens identification and assessment.
  - Step 2: Risk Mitigation — mitigation depends on risk nature (endogenous vs. exogenous, continuous vs. discrete).
  - Step 3: Provisioning — three types: expensing costs up-front in the budget; creating budget contingencies; or establishing buffer funds.
  - Step 4: Accommodation — remote, discrete fiscal risks are generally accommodated by maintaining sufficient fiscal space.

- Institutional arrangements and controls
  - Key elements:
    - A formal risk management policy defining conditions under which fiscal risks may be assumed.
    - Clear responsibilities: individual ministries identify, estimate, analyze, and monitor risks within their functions; some oversight (SOEs, PPPs) centralized at the MoF.
    - A central risk oversight body: monitors aggregate fiscal risk exposures, establishes risk-warning indicators, undertakes ‘war-gaming’, and assesses adequacy of mitigation practices.
    - Central controls over major risks: authority to approve contracts exposing government to fiscal risk vested in a central authorizing entity (usually the MoF or a cabinet committee); decisions assessed as part of budget process.

- Advanced practices expectations (summary)
  - Macroeconomic Risks: budget documentation includes sensitivity analysis, alternative scenarios, and probabilistic forecasts of fiscal outcomes.
  - Specific Fiscal Risks: main specific risks disclosed in a summary report with magnitude estimates and likelihood where practicable.
  - Long-Term Fiscal Sustainability Analysis: multiple scenarios published over at least the next 30 years using a range of assumptions.
  - Budgetary Contingencies: budget includes an allocation for contingencies with transparent access criteria and in-year reporting.
  - Asset and liability Management: liabilities and significant asset acquisitions/disposals authorized by law; balance-sheet risks managed according to published strategy.
  - Guarantees: all government guarantees and gross exposure published at least annually; maximum value authorized by law.
  - Public-Private Partnerships: annual publication of total rights, obligations, exposures, expected receipts/payments; legal limit on accumulated obligations.
  - Financial Sector Exposure: quantify and disclose explicit support to the financial sector at least annually; regular assessments of financial sector stability.
  - Natural Resources: annual estimates of volume and value of major natural resource assets under different price and extraction scenarios.
  - Environmental Risks: identify and quantify main fiscal risks from natural disasters and manage them according to a published strategy.
  - Sub-national Governments: financial condition/performance published quarterly; limit on liabilities or borrowing.
  - Public Corporations: disclose all direct/indirect support and publish an annual report on overall financial performance, including quasi-fiscal activities.

- Existing mitigation tools in Kazakhstan
  - Current tools include:
    - ceilings for borrowing and limits on transfers from the NFRK;
    - a budget reserve to finance expenditures not planned when budgets were prepared;
    - annual limits on state guarantees set in the budget.
  - Information is produced (scenario-based forecasts by MNE, medium-term ceilings, centralized reporting by MoF) but not systematically analyzed and disclosed as a fiscal risk statement.

- Recommendations and technical assistance
  - Produce a single fiscal risk statement:
    - potentially via a newly-established “fiscal risks unit” housed at the MoF or the MNE, in collaboration with other ministries, the NBK, and SOEs;
    - include quantitative macro-fiscal sensitivity and scenario analysis, ideally with balance-sheet analysis;
    - make the statement part of the budget document submitted to parliament and publish it, taking into consideration sensitive information and moral hazard concerns.
  - IMF technical support areas:
    - Fiscal Risk Management and Stress Testing: identify sources of risk (macro shocks, guarantees, PPPs, financial sector) and quantify them; examine monitoring, management, mitigation, and reporting arrangements.
    - Fiscal Transparency Evaluation: assessment against the Fiscal Transparency Code; analysis of fiscal vulnerability based on transparency indicators; action plan for reform priorities.
    - Fiscal Oversight of SOEs: focus on institutional responsibilities, monitoring SOE performance and fiscal impact, identification and management of quasi-fiscal activities, and sector-wide reporting.

### Key Elements of the New Regime of Natural Resource Taxation

- Goals and overview
  - Revisions aim to encourage exploration and incentivize investment.
  - Petroleum sector changes: introduction of an optional alternative tax on profits, elimination of the commercial discovery bonus, and more favorable treatment of exploration expenses.
  - Mining sector changes: removal of the excess profits tax (EPT) and increased reliance on production-based instruments; a comprehensive review of mining taxation is recommended.

- Petroleum sector: alternative tax vs. standard regime
  - Main features of the alternative tax:
    - optional “alternative” tax on profits with a sliding rate scale varying with price;
    - below US$50/bbl, the alternative tax rate is zero;
    - replaces sector-specific production and profit-based instruments under the standard regime.
  - Comparative simulation (illustrative stylized project assumptions and outcomes):
    - Project description data shown in figure: Size: 1,000 MMBbl; Costs: $30.0/Bbl; Oil price: $60.0/Bbl (real); IRR pre tax: 20%; Discount Rate: 10.0%.
    - Assumed stylized oil field: 1 billion barrels, unit cost approximately US$30/bbl.
    - At an assumed US$60/bbl price level, and pre-tax return of 18.3 percent:
      - Standard regime yields a discounted average effective tax rate (AETR) of over 100 percent.
      - Alternative regime yields an AETR around 70 percent.
      - Using a “breakeven price” approach (to meet a hurdle post-tax real rate of 12.5 percent):
        - Standard regime breakeven price is over US$120/bbl.
        - Alternative regime breakeven price is US$56.5/bbl.
    - Trade-off: reliance on profit-based instruments under the alternative regime shifts revenue risk to the government and is more progressive (government share of total benefits rises with profitability but bears more downside for less profitable projects).

- Mining sector: removal of EPT and implications
  - The removal of the excess profits tax (EPT) followed observation that very few companies (aside from coal producers) paid the EPT, due largely to tax planning and avoidance.
  - Concerns:
    - companies with refining operations or non-resource activities create possibilities for transfer pricing, especially if ringfencing rules are not clear.
    - recommendation for careful review of ringfencing and transfer pricing rules to support profit-based instruments.
  - Compensatory changes: to offset loss of EPT from coal producers, the export rent tax was increased from 2.1 to [text truncated in source].

- Transparency and regulatory environment
  - Super-giant fields (Karachaganak, Tengiz, and Kashagan) are governed by project-specific contracts with confidential fiscal terms, complicating assessment of expected revenues and fiscal risks.
  - Recommendations:
    - further improvements in the investment and regulatory environment to secure new investment offshore;
    - in line with international norms, authorities and oil companies are encouraged to disclose all contractually agreed fiscal terms, particularly for the “super-giant” fields, to allow clear public understanding of fiscal benefits.

- Preparation attribution
  - Prepared by Alpa Shah.

### Interest-rate rules: role, estimation, and simulation results

- Role and cautions
  - Interest-rate rules provide guidance and a communication tool for monetary policy.
  - Should not be followed mechanically; have limitations where central banks lack full independence.
  - Modern central banks present outcomes from several rules to inform decision-making.

- NBK’s current approach
  - NBK uses a complex interest-rate rule as part of its quarterly projection model (QPM).
  - Initial QPM rule combined the world interest rate (approximated by the US Federal funds rate) and a Kazakhstan premium linked to the terms of trade and trade balance.
  - With transition to inflation targeting, a new complex rule adopted that is essentially a weighted average of:
    - a balance-of-payments-based rule; and
    - a rule depending on lagged policy rate, the equilibrium real rate, deviation of inflation from its target, and the output gap.

- Simpler rules and small-scale models
  - Simpler Taylor-type rules and small-scale two-equation models (output gap and inflation equations, with lags) can provide tractable and communicable insights.
  - Basic model estimated using quarterly data for 1Q 2003-4Q 2017:
    - Sample: 2003Q1–2017Q4
    - Included observations: 60
    - Inflation defined as the log difference over 4 quarters minus the inflation target—set at 6 percent.
    - Output gap is the log difference of actual and potential GDP extracted from an HP-filter.
    - Equilibrium real rate assumed equal to 4 percent.
  - Unit root tests were overall inconclusive, with p values around 0.06-0.07 (Phillips-Perron); the KPSS test did not reject stationarity.

- Estimated model results (selected)
  - Output gap equation (GAP):
    - GAP(-1) coefficient: 0.823801 (Std. Error 0.070914, Prob. 0.0000)
    - RR(-1) coefficient: -0.245767 (Std. Error 0.084206, Prob. 0.0051)
    - RR(-2) coefficient: 0.234006 (Std. Error 0.083058, Prob. 0.0067)
    - R-squared: 0.752174; Adjusted R-squared: 0.734151
  - Inflation equation (P):
    - GAP(-1) coefficient: 0.017046 (Std. Error 0.055494, Prob. 0.7599)
    - P(-1) coefficient: 1.261463 (Std. Error 0.055060, Prob. 0.0000)
    - P(-2) coefficient: -0.280772 (Std. Error 0.054406, Prob. 0.0000)
    - R-squared: 0.949228; Adjusted R-squared: 0.943480

- Optimal and robust rules (as presented)
  - Optimal interest-rate rule reported in source (symbolic/rule presentation preserved as in source).
  - Robust rule derived via invariant set approach reported in source (symbolic/rule presentation preserved as in source).
  - Uncertainty sets can be calibrated to confidence intervals of estimated parameters and regression variances.

- Simulation scenarios and assumptions
  - Four scenarios:
    1. No uncertainty.
    2. Parametric uncertainty only: coefficient on output gap in inflation equation = point estimate + one standard deviation; coefficient on lagged real rate = point estimate − one standard deviation.
    3. Parametric uncertainty plus a persistent negative shock to output of 0.25 percent per quarter.
    4. Scenario 3 plus a temporary positive shock to inflation of 0.5 per quarter for four quarters, starting from the second quarter.
  - Rules compared: (i) optimal rule; (ii) robust rule; (iii) classical Taylor rule.

- Simulation findings (summary)
  - Scenario 1 (no shocks/uncertainty):
    - Optimal rule performs best overall; robust and Taylor rules yield outcomes close on inflation.
    - Robust rule produces more aggressive interest-rate adjustments, lower output volatility, but larger instantaneous loss due to penalization of interest-rate changes.
  - Scenario 2 (parametric uncertainty):
    - Robust rule brings inflation back to target faster than optimal and Taylor rules.
    - Higher loss under robust rule due to more aggressive interest-rate adjustments and somewhat higher output gap volatility.
  - Scenario 3 (parametric uncertainty + persistent negative output shock):
    - Robust rule exhibits superior performance and stabilizes the system.
    - Optimal rule leads to instability as its timing and magnitude of interest-rate adjustments are ill-suited for the widening output gap.
  - Scenario 4 (adds temporary positive inflation shock to Scenario 3):
    - Robust rule achieves significantly better stabilization.
    - It prescribes a relatively steep interest-rate increase during the inflation shock and larger cuts thereafter.

- Quantitative variability measures (Table 2 excerpts)
  - Scenario 1:
    - Inflation: Optimal rule 0.0561; Robust rule 0.0523; Taylor rule 0.0541
    - Output gap: Optimal rule 0.0039; Robust rule 0.0014; Taylor rule 0.0026
  - Scenario 2:
    - Inflation: Optimal rule 0.0675; Robust rule 0.0225; Taylor rule 0.0424
    - Output gap: Optimal rule 0.0015; Robust rule 0.0043; Taylor rule 0.0016
  - Scenario 3:
    - Inflation: Optimal rule 5.3905; Robust rule 0.9902; Taylor rule 2.5171
    - Output gap: Optimal rule 1.5054; Robust rule 0.1899; Taylor rule 0.5942
  - Scenario 4:
    - Inflation: Optimal rule 2.4897; Robust rule 1.0038; Taylor rule 1.5543
    - Output gap: Optimal rule 1.5740; Robust rule 0.8622; Taylor rule 1.0158
  - Note: Variability measured as the sum of squared deviations of inflation from target and of output gap from zero, divided by the length of the time period.

- Policy recommendation and toolkit implications
  - NBK could integrate interest-rate projections from additional rules:
    - Continue using the QPM rule.
    - Add optimal and robust rules as useful benchmarks.
    - Retain the original Taylor rule in the toolkit.
    - Consider other rules and variants (e.g., variations of the Taylor rule, difference rules) as further cross-checks.

- Source: IMF staff calculations and analysis as presented in the supplied content.

### Reforming reserve requirements — context, design issues, and recommendations

- Operational features of current RR system
  - NBK operates a complex system of reserve requirements (RRs) with different rates based on residency, maturity, and currency of denomination of liabilities.
  - Short-term liabilities are defined as liabilities up to one year.
  - Maintenance and calculation periods set at 28 days and fulfilment of RRs is defined in terms of averages; the average amount of the eligible reserve assets during the maintenance period should be not less than the average RR for the calculation period.
  - Currently, all cash holdings in domestic currency are recognized as eligible assets.
  - RR are not remunerated.

- Reported table of current RR rates (percent of liabilities) (Source: National Bank of Kazakhstan)
  - Categories and reported rates:
    - 2 4 0 2 2 6 0 2

- Objectives and role of reserve requirements
  - Principal functions considered: prudential, monetary control, and liquidity management; liquidity management is likely the most relevant at present.
  - NBK does not target reserve money; prudential function is covered by liquidity requirements, deposit insurance, and the NBK’s credit facility.
  - RRs can be useful for absorbing surplus liquidity and help maintain a more stable and predictable monetary base.
  - In absence of RRs, volatility of reserve money would likely track volatility of payments, as banks minimize holdings of NBK reserves that do not bear interest.

- Key issues for reform design
  - Base:
    - Currently banks are not required to hold reserves against long-term liabilities of residents in domestic currency or FX.
    - Recommendation: extend RR to all liabilities (exclude liabilities to other financial institutions subject to the same RR regime).
  - Rates:
    - Current structure employs residency, maturity, and currency, resulting in eight categories with four different rates.
    - Recommendation: simplify by eliminating residency differentiation as a minimum and retain differentiation by currency (tenge vs FX) to help reduce dollarization.
    - Suggested two-rate structure: one for tenge and one for FX.
    - If reserves are not remunerated, suggested illustrative rates: 4 percent for tenge and 6 percent for FX.
    - If higher rates are chosen, the difference between the new rates and 4 and 6 percent, respectively, should be remunerated.
  - Remuneration:
    - Unremunerated RR act as a tax on financial intermediation and widen interest spreads.
    - Recommendation: move to higher, remunerated RR combined with gradual reduction of RR meetable with cash in vaults.
    - Remuneration rate should reflect opportunity cost; appropriate rate would be that at which the NBK carries out sterilization operations.
    - Offering interest equal to the deposit facility rate would help drain liquidity at little additional cost for the NBK.
    - For FX, remuneration could be set below LIBOR to discourage dollarization.
  - Eligible reserve assets (vault cash and currency denomination):
    - Current regulation allows RR to be fulfilled with cash in vault.
    - Arguments for inclusion: cash is a central bank liability; inclusion supports banks in rural areas that operate with more cash.
    - Arguments against: definitional and measurement difficulties, potential misreporting for banks with large branch networks.
    - Recommendation: consider phasing out vault cash as eligible asset if bank-by-bank analysis shows moderate impact; phase-out could be staged.
    - Recommendation: in stable environment, prefer reserve assets denominated in domestic currency as automatic stabilizer; in periods of large ER fluctuations this may complicate bank liquidity management.
    - Given banks already hold large FX deposits at the NBK, allowing use of part of these deposits to meet FX RR could relax the current regime—advise doing this only with elimination or significant reduction of cash-in-vault eligibility.
  - Other operational issues:
    - Averaging: NBK applies averaging, but concerns exist banks may use intraday balances to avoid immobilizing liquidity; eliminating averaging could increase short-term interest rate volatility.
    - Carry-overs: some central banks allow small carry-overs between maintenance periods to smooth overnight rate volatility; consideration recommended.
    - Penalties: most central banks impose penalties for non-compliance set higher than the credit-facility rate; Ordinance 38 regulating RRs in Kazakhstan does not seem to include such provisions—gap should be filled unless regulated elsewhere.

- Summary of recommended reform directions
  - Simplify the RR system by eliminating residency differentiation.
  - Maintain differentiation by currency denomination; set higher RR rates for FX liabilities to facilitate dedollarization.
  - In an environment of structural surplus liquidity, raising RR rates from current levels would support NBK sterilization efforts.
  - Consider remunerating RR above the current maximum levels of 4 percent for domestic and 6 percent for FX liabilities; remuneration rate could be set equal to the deposit facility rate for domestic liabilities and below LIBOR for FX liabilities.
  - Exclude cash in vaults from eligible reserve assets, or at least significantly limit its use; if excluded, allow banks to meet FX RR with balances on their dollar accounts at the NBK.
  - Maintain averaging and consider small carry-over provisions between consecutive maintenance periods.
  - Clearly specify penalties for non-compliance with RR.

- Prepared by Ivan Luis de Oliveira Lima, Vassili Prokopenko and Rossen Rozenov.

*Prepared from IMF staff analysis contained in the supplied content unit cr18278.*

### 1. Advanced Practices in Risk Management _____________________________________________ 6

### 1. Advanced Practices in Risk Management

### Fiscal risk: diagnosis and recent history
- Kazakhstan faces fiscal risk from diverse sources, including volatility of oil prices, financial sector exposure, state-owned enterprise (SOE) risks, and other macroeconomic and specific shocks.
- Historically, oil price shocks of 2008-09 and 2014-15 led to:
  - large declines in revenues and exports;
  - costly bailouts of highly dollarized banks in 2009 and 2016;
  - a large transfer to KazMunaiGas in 2015 to ensure timely debt service payments.
- The oil price shocks in 2008-09 and 2014-15 led to a decline of nominal GDP by nearly two standard deviations relative to the average; best practice identifies a macro shock when nominal GDP falls by one standard deviation (IMF 2016).

### Rationale for comprehensive fiscal risk management
- Analyzing and managing fiscal risks:
  - helps credibility and sustainability of public finances;
  - supports macro stability and strengthens central forecasts;
  - informs implications for government liquidity, financing needs, and solvency.
- Current practice gaps:
  - risk management practices are often incomplete, fragmented, and lacking quantitative analysis;
  - no systematic analysis of forecast errors for macroeconomic assumptions or fiscal forecasts;
  - line ministries provide statements of risks but typically lack quantified risk information;
  - MoF central reporting is not analyzed from a risk perspective or summarized into a statement of fiscal risks.

### Four-step framework recommended for fiscal risk management
- Step 1: Identifying and Quantifying Risks.
  - Advanced practice includes identifying, assessing, and disclosing risks; quantified assessments and disclosure create a feedback loop that strengthens identification and assessment.
- Step 2: Risk Mitigation.
  - Mitigation depends on risk nature (endogenous vs. exogenous, continuous vs. discrete).
- Step 3: Provisioning.
  - Three types of provisioning: expensing costs up-front in the budget; creating budget contingencies; or establishing buffer funds.
- Step 4: Accommodation.
  - Remote, discrete fiscal risks are generally accommodated by maintaining sufficient fiscal space.

### Institutional arrangements and controls
- Key institutional elements:
  - A formal risk management policy defining conditions under which fiscal risks may be assumed.
  - Clear responsibilities: individual ministries identify, estimate, analyze, and monitor risks within their functions; some oversight (SOEs, PPPs) centralized at the MoF.
  - A central risk oversight body: monitors aggregate fiscal risk exposures, establishes risk-warning indicators, undertakes ‘war-gaming’, and assesses adequacy of mitigation practices.
  - Central controls over major risks: authority to approve contracts exposing government to fiscal risk vested in a central authorizing entity (usually the MoF or a cabinet committee); decisions assessed as part of budget process.

### Table of advanced practices (summary of categories and expectations)
- Macroeconomic Risks: budget documentation includes sensitivity analysis, alternative scenarios, and probabilistic forecasts of fiscal outcomes.
- Specific Fiscal Risks: main specific risks disclosed in a summary report with magnitude estimates and likelihood where practicable.
- Long-Term Fiscal Sustainability Analysis: multiple scenarios published over at least the next 30 years using a range of assumptions.
- Budgetary Contingencies: budget includes an allocation for contingencies with transparent access criteria and in-year reporting.
- Asset and liability Management: liabilities and significant asset acquisitions/disposals authorized by law; balance-sheet risks managed according to published strategy.
- Guarantees: all government guarantees and gross exposure published at least annually; maximum value authorized by law.
- Public-Private Partnerships: annual publication of total rights, obligations, exposures, expected receipts/payments; legal limit on accumulated obligations.
- Financial Sector Exposure: quantify and disclose explicit support to the financial sector at least annually; regular assessments of financial sector stability.
- Natural Resources: annual estimates of volume and value of major natural resource assets under different price and extraction scenarios.
- Environmental Risks: identify and quantify main fiscal risks from natural disasters and manage them according to a published strategy.
- Sub-national Governments: financial condition/performance published quarterly; limit on liabilities or borrowing.
- Public Corporations: disclose all direct/indirect support and publish an annual report on overall financial performance, including quasi-fiscal activities.

### Existing mitigation tools in Kazakhstan
- Current tools include:
  - ceilings for borrowing and limits on transfers from the NFRK;
  - a budget reserve to finance expenditures not planned when budgets were prepared;
  - annual limits on state guarantees set in the budget.
- Information is produced (scenario-based forecasts by MNE, medium-term ceilings, centralized reporting by MoF) but not systematically analyzed and disclosed as a fiscal risk statement.

### Recommendations and technical assistance
- Produce a single fiscal risk statement:
  - potentially via a newly-established “fiscal risks unit” housed at the MoF or the MNE, in collaboration with other ministries, the NBK, and SOEs;
  - include quantitative macro-fiscal sensitivity and scenario analysis, ideally with balance-sheet analysis;
  - make the statement part of the budget document submitted to parliament and publish it, taking into consideration sensitive information and moral hazard concerns.
- IMF technical support areas:
  - Fiscal Risk Management and Stress Testing: identify sources of risk (macro shocks, guarantees, PPPs, financial sector) and quantify them; examine monitoring, management, mitigation, and reporting arrangements.
  - Fiscal Transparency Evaluation: assessment against the Fiscal Transparency Code; analysis of fiscal vulnerability based on transparency indicators; action plan for reform priorities.
  - Fiscal Oversight of SOEs: focus on institutional responsibilities, monitoring SOE performance and fiscal impact, identification and management of quasi-fiscal activities, and sector-wide reporting.

*References: IMF, 2009, Fiscal Risks. Sources, Disclosure, and Management; IMF, 2014, The Republic of Kazakhstan. Selected Issues in Financial Reporting, Accrual Budgeting, Audit, and Fiscal Risk Management; IMF, 2014, The Fiscal Transparency Code; IMF, 2016, Analyzing and Managing Fiscal Risks: Best Practices.*

---

### Key Elements of the New Regime of Natural Resource Taxation

### Goals and overview
- Revisions to the fiscal regime for extractive industries aim to encourage exploration and incentivize investment.
- Petroleum sector changes: introduction of an optional alternative tax on profits, elimination of the commercial discovery bonus, and more favorable treatment of exploration expenses.
- Mining sector changes: removal of the excess profits tax (EPT) and increased reliance on production-based instruments; a comprehensive review of mining taxation is recommended.

### Petroleum sector: alternative tax vs. standard regime
- Main features of the alternative tax:
  - optional “alternative” tax on profits with a sliding rate scale varying with price;
  - below US$50/bbl, the alternative tax rate is zero;
  - replaces sector-specific production and profit-based instruments under the standard regime.
- Comparative simulation (illustrative stylized project assumptions and outcomes):
  - Project description data shown in figure: Size: 1,000 MMBbl; Costs: $30.0/Bbl; Oil price: $60.0/Bbl (real); IRR pre tax: 20%; Discount Rate: 10.0%.
  - Textual simulation highlights:
    - Assumed stylized oil field: 1 billion barrels, unit cost approximately US$30/bbl.
    - At an assumed US$60/bbl price level, and pre-tax return of 18.3 percent:
      - Standard regime yields a discounted average effective tax rate (AETR) of over 100 percent.
      - Alternative regime yields an AETR around 70 percent.
      - Using a “breakeven price” approach (to meet a hurdle post-tax real rate of 12.5 percent):
        - Standard regime breakeven price is over US$120/bbl.
        - Alternative regime breakeven price is US$56.5/bbl.
  - Trade-off: reliance on profit-based instruments under the alternative regime shifts revenue risk to the government and is more progressive (government share of total benefits rises with profitability but bears more downside for less profitable projects).

### Mining sector: removal of EPT and implications
- The removal of the excess profits tax (EPT) followed observation that very few companies (aside from coal producers) paid the EPT, due largely to tax planning and avoidance.
- Concerns noted:
  - companies with refining operations or non-resource activities create possibilities for transfer pricing, especially if ringfencing rules are not clear.
  - recommendation for careful review of ringfencing and transfer pricing rules to support profit-based instruments.
- Compensatory changes: to offset loss of EPT from coal producers, the export rent tax was increased from 2.1 to [text truncated in source].

### Transparency and regulatory environment
- Super-giant fields (Karachaganak, Tengiz, and Kashagan) are governed by project-specific contracts with confidential fiscal terms, complicating assessment of expected revenues and fiscal risks.
- Recommendations:
  - further improvements in the investment and regulatory environment to secure new investment offshore;
  - in line with international norms, authorities and oil companies are encouraged to disclose all contractually agreed fiscal terms, particularly for the “super-giant” fields, to allow clear public understanding of fiscal benefits.

*Prepared by Alpa Shah.*

### 4.7 percent, implying greater reliance on production-based instruments. The authorities should

### cr18278 - 4.7 percent, implying greater reliance on production-based instruments. The authorities should

### Interest-rate rules: role and cautions
- Interest-rate rules can provide useful guidance to monetary policy, serve as a benchmark for the policy stance, and be a communication tool.
- They should not be followed mechanically and have limitations (e.g., in contexts where central banks lack full independence).
- Modern central banks (example: U.S. Federal Reserve) present outcomes from several rules to inform decision-making.

### The NBK's current approach
- The NBK uses a complex interest-rate rule as part of its quarterly projection model (QPM).
- The initial QPM rule combined the world interest rate (approximated by the US Federal funds rate) and a Kazakhstan premium linked to the terms of trade and trade balance.
- With the transition to inflation targeting, a new complex rule was adopted that is essentially a weighted average of:
  - a balance-of-payments-based rule; and
  - a rule depending on lagged policy rate, the equilibrium real rate, deviation of inflation from its target, and the output gap.

### Simpler rules and small-scale models
- Simpler Taylor-type rules and small-scale two-equation models (output gap and inflation equations, with lags) can provide additional, more tractable and communicable insights.
- A basic model of inflation and output gap dynamics was estimated using quarterly data for 1Q 2003-4Q 2017.
  - Sample: 2003Q1–2017Q4
  - Included observations: 60
  - Inflation defined as the log difference over 4 quarters minus the inflation target—set at 6 percent.
  - Output gap is the log difference of actual and potential GDP extracted from an HP-filter.
  - Equilibrium real rate assumed equal to 4 percent.
- Unit root tests were overall inconclusive, with p values around 0.06-0.07 (Phillips-Perron); the KPSS test did not reject stationarity.

### Estimated model results (selected)
- Output gap equation (GAP) estimation highlights:
  - GAP(-1) coefficient: 0.823801 (Std. Error 0.070914, Prob. 0.0000)
  - RR(-1) coefficient: -0.245767 (Std. Error 0.084206, Prob. 0.0051)
  - RR(-2) coefficient: 0.234006 (Std. Error 0.083058, Prob. 0.0067)
  - R-squared: 0.752174; Adjusted R-squared: 0.734151
- Inflation equation (P) estimation highlights:
  - GAP(-1) coefficient: 0.017046 (Std. Error 0.055494, Prob. 0.7599) — statistically insignificant
  - P(-1) coefficient: 1.261463 (Std. Error 0.055060, Prob. 0.0000)
  - P(-2) coefficient: -0.280772 (Std. Error 0.054406, Prob. 0.0000)
  - R-squared: 0.949228; Adjusted R-squared: 0.943480

### Optimal interest-rate rule (derived under quadratic loss with equal weights)
- The resulting optimal rule is reported as:
  ݎݎ
  ௢
  ൌ0.07݌ܽ݃൅0.053݌െ0.015݌
  ିଵ
  ൅0.020ݎݎ
  ିଵ
  .
  (rule presented in source text)

### Robust rule (invariant set approach)
- Accounting for parametric and additive uncertainty via an invariant set approach yields the following rule:
  ݎݎ
  ௥
  ൌ0.678݌ܽ݃൅1.859݌െ0.536݌
  ିଵ
  ൅0.140ݎݎ
  ିଵ
  .
  (rule presented in source text)
- The uncertainty sets can be calibrated to confidence intervals of estimated parameters and regression variances.

### Simulation scenarios (illustrative) and key assumptions
- Four scenarios simulated using the estimated model:
  1. No uncertainty.
  2. Parametric uncertainty only: coefficient on output gap in inflation equation = point estimate + one standard deviation; coefficient on lagged real rate = point estimate − one standard deviation.
  3. Parametric uncertainty plus a persistent negative shock to output of 0.25 percent per quarter.
  4. Scenario 3 plus a temporary positive shock to inflation of 0.5 per quarter for four quarters, starting from the second quarter.
- Rules compared: (i) optimal rule; (ii) robust rule; (iii) classical Taylor rule.

### Simulation findings (summary)
- Scenario 1 (no shocks/uncertainty):
  - The optimal rule performs best overall; robust and Taylor rules yield outcomes close on inflation.
  - Robust rule produces more aggressive interest-rate adjustments, lower output volatility, but larger instantaneous loss due to penalization of interest-rate changes.
- Scenario 2 (parametric uncertainty):
  - Robust rule brings inflation back to target faster than optimal and Taylor rules.
  - Higher loss under robust rule due to more aggressive interest-rate adjustments and somewhat higher output gap volatility.
- Scenario 3 (parametric uncertainty + persistent negative output shock):
  - Robust rule exhibits superior performance and stabilizes the system.
  - Optimal rule leads to instability as its timing and magnitude of interest-rate adjustments are ill-suited for the widening output gap.
- Scenario 4 (adds temporary positive inflation shock to Scenario 3):
  - Robust rule again achieves significantly better stabilization.
  - It prescribes a relatively steep interest-rate increase during the inflation shock and larger cuts thereafter.

### Quantitative results from Table 2 (variability measures)
- Scenario 1
  - Inflation: Optimal rule 0.0561; Robust rule 0.0523; Taylor rule 0.0541
  - Output gap: Optimal rule 0.0039; Robust rule 0.0014; Taylor rule 0.0026
- Scenario 2
  - Inflation: Optimal rule 0.0675; Robust rule 0.0225; Taylor rule 0.0424
  - Output gap: Optimal rule 0.0015; Robust rule 0.0043; Taylor rule 0.0016
- Scenario 3
  - Inflation: Optimal rule 5.3905; Robust rule 0.9902; Taylor rule 2.5171
  - Output gap: Optimal rule 1.5054; Robust rule 0.1899; Taylor rule 0.5942
- Scenario 4
  - Inflation: Optimal rule 2.4897; Robust rule 1.0038; Taylor rule 1.5543
  - Output gap: Optimal rule 1.5740; Robust rule 0.8622; Taylor rule 1.0158
- Note: Variability measured as the sum of squared deviations of inflation from target and of output gap from zero, divided by the length of the time period.

### Policy recommendation and toolkit implications
- The NBK could benefit from integrating interest-rate projections implied by an additional set of rules into its decision-making process:
  - Continue using the QPM rule.
  - Add optimal and robust rules as useful benchmarks.
  - Retain the original Taylor rule in the toolkit given its broad acceptance and reasonable performance across circumstances.
  - Consider other rules and variants (e.g., variations of the Taylor rule, difference rules) as further cross-checks.

*Source: IMF staff calculations and analysis as presented in the supplied content.*

### References

### References

### Bibliographic citations
- Asso, P.F., Kahn, G. and Leeson, R. (2010), “The Taylor Rule and the Practice of Central Banking”, The Federal Reserve Bank of Kansas City RWP 10-05.
- Chernyavskii, D. and Mukanov, N. (2017), “Implementation of Monetary Policy Rules in the Quarterly Projection Model for Kazakhstan”, Den’gi I Kredit, No.5 (in Russian).
- Hansen, L. and Sargent, T. (2008), “Robustness”, Princeton University Press.
- Laubach, T. and J. Williams (2003), “Measuring the Natural rate of Interest”, The Review of Economics and Statistics, Vol. 85, No. 4, pp. 1063-1070.
- Orphanides, A. (2003), “Historical Monetary Policy Analysis and the Taylor Rule," Journal of Monetary Economics, Vol. 50, pp. 983-1022.
- Rudebusch, G. and Svensson, L. (1999), “Policy Rules for Inflation Targeting”, in Taylor (ed.) Monetary Policy Rules, NBER Studies in Business Cycles 31, The University of Chicago Press, pp. 203-252.
- Rozenov, R. (2017), “Public Debt Sustainability Under Uncertainty: An Invariant Set Approach”, IMF Working Paper WP/17/57, Washington DC
- Taylor, J. and Williams, J. (2010), “Simple and Robust Rules for Monetary Policy”, Handbook of Monetary Economics, Vol. 3B, Elsevier B.V.
- Walsh, C. (2017), “The Challenges with Rule-based Policy Implementation”, Working paper prepared for the 2017 Annual Conference of the Federal Reserve Bank of Boston, available at https://www.bostonfed.org/-/media/Images/discretionmonpol2017/walsh-paper.pdf

### Reforming reserve requirements — context and operational features
- The National Bank of Kazakhstan (NBK) operates a complex system of reserve requirements (RRs) with different rates based on residency, maturity, and currency of denomination of liabilities.
- Short-term liabilities are defined as liabilities up to one year.
- The maintenance and calculation periods are set at 28 days and fulfilment of RRs is defined in terms of averages; the average amount of the eligible reserve assets during the maintenance period should be not less than the average RR for the calculation period.
- Currently, all cash holdings in domestic currency are recognized as eligible assets.
- RR are not remunerated.
- Table of current RR rates (percent of liabilities) as reported (Source: National Bank of Kazakhstan):
  - Liabilities in national currency / Liabilities in foreign currency
  - Short-term residents non-residents long-term residents non-residents short-term residents non-residents long-term residents non-residents
  - 2 4 0 2 2 6 0 2

- Prepared by Ivan Luis de Oliveira Lima, Vassili Prokopenko and Rossen Rozenov.

### Objectives and role of reserve requirements
- Principal functions considered: prudential, monetary control, and liquidity management; liquidity management is likely the most relevant at present.
- NBK does not target reserve money; prudential function is covered by liquidity requirements, deposit insurance, and the NBK’s credit facility.
- RRs can be useful for absorbing surplus liquidity and help maintain a more stable and predictable monetary base.
- In absence of RRs, volatility of reserve money would likely track volatility of payments, as banks minimize holdings of NBK reserves that do not bear interest.

### Key issues for reform design
- Base:
  - Currently banks are not required to hold reserves against long-term liabilities of residents in domestic currency or FX.
  - Recommendation: extend RR to all liabilities (exclude liabilities to other financial institutions subject to the same RR regime) because distinction between demand and term deposits is largely formal and creates incentives for window-dressing.

- Rates:
  - Current structure employs residency, maturity, and currency, resulting in eight categories with four different rates.
  - Recommendation: simplify by eliminating residency differentiation as a minimum and retain differentiation by currency (tenge vs FX) to help reduce dollarization.
  - Suggested two-rate structure: one for tenge and one for FX.
  - If reserves are not remunerated, suggested illustrative rates: 4 percent for tenge and 6 percent for FX.
  - If higher rates are chosen, the difference between the new rates and 4 and 6 percent, respectively, should be remunerated.

- Remuneration:
  - Unremunerated RR act as a tax on financial intermediation and widen interest spreads.
  - Given structural surplus liquidity sterilized mainly through NBK notes, recommendation: move to higher, remunerated RR combined with gradual reduction of RR meetable with cash in vaults.
  - Remuneration rate should reflect opportunity cost; in current environment the appropriate rate would be that at which the NBK carries out sterilization operations.
  - Offering interest equal to the deposit facility rate would help drain liquidity at little additional cost for the NBK.
  - For FX, remuneration could be set below LIBOR to discourage dollarization.

- Eligible reserve assets (vault cash and currency denomination):
  - Current regulation allows RR to be fulfilled with cash in vault.
  - Arguments for inclusion: cash is a central bank liability; inclusion supports banks in rural areas that operate with more cash.
  - Arguments against: definitional and measurement difficulties, potential misreporting for banks with large branch networks.
  - Recommendation: consider phasing out vault cash as eligible asset if bank-by-bank analysis shows moderate impact; phase-out could be staged.
  - Recommendation: in stable environment, prefer reserve assets denominated in domestic currency as automatic stabilizer; in periods of large ER fluctuations this may complicate bank liquidity management.
  - Given banks already hold large FX deposits at the NBK, allowing use of part of these deposits to meet FX RR could relax the current regime—advise doing this only with elimination or significant reduction of cash-in-vault eligibility.

- Other operational issues:
  - Averaging: NBK applies averaging, but concerns exist banks may use intraday balances to avoid immobilizing liquidity; nevertheless, eliminating averaging could increase short-term interest rate volatility.
  - Carry-overs: some central banks allow small carry-overs between maintenance periods to smooth overnight rate volatility; consideration recommended.
  - Penalties: most central banks impose penalties for non-compliance set higher than the credit-facility rate; Ordinance 38 regulating RRs in Kazakhstan does not seem to include such provisions—gap should be filled unless regulated elsewhere.

### Summary of recommended reform directions
- Simplify the RR system by eliminating residency differentiation.
- Maintain differentiation by currency denomination; set higher RR rates for FX liabilities to facilitate dedollarization.
- In an environment of structural surplus liquidity, raising RR rates from current levels would support NBK sterilization efforts.
- Consider remunerating RR above the current maximum levels of 4 percent for domestic and 6 percent for FX liabilities; remuneration rate could be set equal to the deposit facility rate for domestic liabilities and below LIBOR for FX liabilities.
- Exclude cash in vaults from eligible reserve assets, or at least significantly limit its use; if excluded, allow banks to meet FX RR with balances on their dollar accounts at the NBK.
- Maintain averaging and consider small carry-over provisions between consecutive maintenance periods.
- Clearly specify penalties for non-compliance with RR.

*Republic of Kazakhstan — IMF staff analysis contained in the referenced material.*

### References

### References

### Works cited in this content unit
- Gray, S. (2011), “Central Bank Balances and Reserve Requirements”, IMF Working Paper WP/11/36, Washington DC.
- Heller, D. and Lengwiler, Y. (2003), “Payment Obligations, Reserve Requirements and the Demand for Central Bank Balances”, Journal of Monetary Economics 50, pp. 419-32.
- Hausmann, R., Hidalgo, C., Bustos, S., Coscia, M., Simoes, A., Yildirim, M., (2013), “The Atlas of Economic Complexity: Mapping Paths to Prosperity”, 2nd ed., Cambridge: MIT Press
- International Monetary Fund (2012), “Macroeconomic Policy Frameworks for Resource-Rich Countries”, International Monetary Fund, Washington, DC
- International Monetary Fund, World Bank, World Trade Organization (2017), “Making Trade an Engine of Growth for All : The Case for Trade and for Policies to Facilitate Adjustment”, https://www.imf.org/en/Publications/Policy-Papers/Issues/2017/04/08/making-trade-an-engine-of-growth-for-all
- Restuccia, D., Rogerson, R., (2008), “Policy Distortions and Aggregate Productivity with Heterogeneous Establishments”, Review of Economic Dynamics No. 11, pp. 707-720
- U. S. Geological Survey (2007), “2005 Minerals Yearbook: Commonwealth of Independent States”, U.S. Department of the Interior, U.S. Geological Survey, Reston, Virginia
- World Bank (2017), “Creating Markets in Kazakhstan: Country Private Sector Diagnostic”, World Bank Group, Washington DC.

*Content unit: cr18278 - References (PDF chapter/section)*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2018/cr18278.pdf_