## 1. Regional Comparison

## Source details

**Canonical URL:** [1. Regional Comparison](https://www.imf.org/-/media/files/publications/cr/2024/english/1kgzea2024002.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2024/english/1kgzea2024002.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2024/english/1kgzea2024002.pdf.json)

---

### Introduction and recent inflation trajectory
- Kyrgyz Republic experienced an inflation surge in 2022:
  - Inflation rose from about 6 percent in 2020 to 16.2 percent by February 2023 (ten-year peak).
  - Surpassed global average of 8.7 percent and CCA average of 13 percent for 2022.
- Headline inflation trend:
  - Began receding from March 2023 and reached single digits in August 2023.
- Core inflation:
  - Remained persistently in double digits (core excludes food and energy).
- Historical spikes:
  - 2007–08 global food crisis, 2014 (substantial Russian ruble depreciation), and post-COVID/war-in-Ukraine episode—each coincided with global food and energy price increases.

### Inflation dynamics and breadth
- Broadening of inflation:
  - 2019: 5 out of 47 CPI components had annual inflation over 10 percent; by August 2023 this increased to 32.
  - CPI components with annual inflation less than 5 percent fell from 45 in 2019 to 6 by August 2023.
- Drivers and deviations:
  - Strong correlation with global food and energy prices due to import shares in the CPI basket.
  - Recent episode deviated: after initial surge from global food prices, domestic inflation remained resilient to global food price declines, indicating increasing influence of domestic demand factors and non-food/services contributions since 2021.

### Policy measures and administratively suppressed prices
- Government interventions to safeguard food security and limit price fluctuations:
  - September 2023: government earmarked 16 goods for price regulation should prices escalate significantly.
  - March 2022: six-month export ban on agricultural goods.
  - May 2022: temporary abolition of a 12 percent import VAT on certain grain crops.
- Effect:
  - Interventions resulted in administratively suppressed inflation for commodities subject to administrative price controls.

### Monetary policy framework and transmission
- Framework and institutional changes:
  - 2014: transition to an interest rate-based monetary policy framework; policy rate set by the central bank Board.
  - Since 2021: central bank increased the policy rate to bring inflation down.
  - Recent changes: introduction of a symmetric interest rate corridor; successful de-dollarization by the central bank.
- Transmission and liquidity:
  - Passthrough from policy rate to interbank rates strengthened; passthrough to market rates subdued by marked influx of liquidity.
  - Liquidity influx driver: authorities’ decision to facilitate sales and export of domestically produced gold through the central bank, requiring significant sterilization operations and foreign exchange interventions.
  - Liquidity defined as the sum of short term NBKR notes, overnight deposits, and excess reserves.
  - Core inflation recently patterns well with the amount of liquidity in the system.

### Model approaches and key quantitative findings
- Methods used:
  - Augmented Phillips curve framework (headline and core inflation separately).
  - Structural VAR (SVAR).
  - Novel comprehensive measure of inflation expectations: blend of adaptive and forward-looking expectations using recursive autoregressive four-quarter forecasts and three-year-forward IMF WEO projections combined via Principal Component Analysis (PCA).
- Augmented Phillips curve key estimates and contributions:
  - Strong positive relationship between headline inflation and inflation expectations:
    - Levels specification: coefficient of 0.3 for inflation expectations.
    - Standardized specification: relative contribution of inflation expectations exceeds 22 percent.
  - Output gap: positive relationship with headline inflation but weak statistical significance; including foreign factors further diminishes association.
  - Exchange rate passthrough: average pass-through from exchange rate depreciation to inflation estimated at 25 percent.
  - Global factors as primary determinants of headline inflation:
    - Global food prices: a unit increase corresponds to a 0.62-point increase in headline inflation.
    - Foreign CPI: a unit increase corresponds to a 0.34-point increase in headline inflation.
    - By comparison:
      - Inflation expectations: a unit increase corresponds to a 0.30-point increase in headline inflation.
      - Output gap: a unit increase corresponds to a 0.12-point increase in headline inflation (Panel 1).
  - Core inflation findings:
    - Strong positive correlation with inflation expectations, coefficient over 0.5, indicating higher persistence than headline inflation.
    - Domestic factors—especially inflation expectations and wages—have stronger impacts on core inflation; foreign factors contribute less.

### Selected augmented Phillips curve empirical estimates (headline CPI, Annex I. Table 1)
- Inflation Expectations (Normal): 3.846* (0.000) ; 2.173* (0.000) ; 1.968 (0.000)
  - Standardized: 0.598* (0.000) ; 0.336* (0.000) ; 0.304* (0.000)
- Output Gap / Lag (Normal): 0.652* (0.000) ; 0.304* (0.007) ; 0.232** (0.042)
  - Standardized: 0.361 (0.000) ; 0.160** (0.011) ; 0.121*** (0.058)
- NEER / Lag (Normal): -0.273* (0.000) ; -0.301* (0.000) ; -0.287* (0.000)
  - Standardized: -0.242 (0.000) ; -0.261* (0.000) ; -0.248* (0.000)
- Global Food Prices / Lag (Normal): 0.213* (0.000) ; 0.235* (0.000)
  - Standardized: 0.567* (0.000) ; 0.623* (0.000)
- Foreign Inflation / Lag (Normal): 1.261* (0.000) ; 1.189* (0.000)
  - Standardized: 0.365* (0.000) ; 0.344* (0.000)
- Public Wage / Lag (Normal): 0.056^ (0.244) ; 0.053^ (0.265)
  - Standardized: 0.080^ (0.197) ; 0.076^ (0.215)
- Global Oil Prices / Lag (Normal): 0.022** (0.045)
  - Standardized: 0.120** (0.047)
- Adjusted R square (Normal): 0.603 ; 0.771 ; 0.779
  - (Standardized): 0.603 ; 0.774 ; 0.783

### Selected core CPI estimates (Annex I. Table 2)
- Inflation Expectations (Normal): 2.417* (0.000) ; 2.096* (0.000) ; 2.132* (0.000)
  - Standardized: 0.0603* (0.000) ; 0.541* (0.000) ; 0.543* (0.000)
- Output Gap / Lag (Normal): 0.246** (0.011) ; 0.051^ (0.508) ; 0.185** (0.044)
  - Standardized: 0.173** (0.038) ; 0.090^ (0.210) ; 0.144*** (0.066)
- NEER / Lag (Normal): -0.213* (0.001) ; -0.237* (0.000) ; -0.246* (0.000)
  - Standardized: -0.225* (0.007) ; -0.305* (0.000) ; -0.343* (0.000)
- Global Food Prices / Lag (Normal): 0.084* (0.001) ; 0.101* (0.000)
  - Standardized: 0.334* (0.001) ; 0.273* (0.007)
- Foreign Inflation / Lag (Normal): 1.125* (0.000) ; 1.135* (0.000)
  - Standardized: 0.458* (0.000) ; 0.463* (0.000)
- Public Wage / Lag (Normal): 0.091* (0.001) ; 0.076** (0.027)
  - Standardized: 0.192* (0.004) ; 0.144** (0.037)
- Global Oil Prices / Lag (Normal): 0.021** (0.014)
  - Standardized: 0.181** (0.019)
- Adjusted R square (Normal): 0.489 ; 0.688 ; 0.708
  - (Standardized): 0.454 ; 0.690 ; 0.679

### SVAR methodology and key impulse-response findings
- SVAR setup:
  - Augments VAR by including exogenous variables (e.g., global food prices and global CPI).
  - Baseline SVAR variables: headline and core inflation, global food prices, nominal effective exchange rate (NEER), central bank policy rate, market lending rates, government borrowing from the domestic banking sector, private sector credit, and the output gap.
  - Alternative specifications: replace interest-rate instruments with quantity variables (monetary aggregates, liquidity measures).
  - Identification: Cholesky decomposition; estimated form A Y_t = B(L) Y_{t-1} + ε_t (A is 8 x 8).
- Key impulse-response findings (standardized variables):
  - One standard deviation shock in global food prices raises headline inflation by nearly 1.5 standard deviations in six quarters.
  - One standard deviation increase in claims on the government (proxy for fiscal policy) leads to roughly a 0.8-unit surge in headline inflation.
- Pass-through and credit:
  - Some pass-through from policy rate to bank lending rates; pass-through to deposit rates is less pronounced and has longer lags.
  - Baseline SVAR: central bank policy rate shows limited direct impact on inflation.
  - Inflation shows limited sensitivity to lending and deposit rates, to changes in private sector credit, and to the output gap.

### Monetary policy transmission, liquidity, and interpretation
- Alternative SVAR with liquidity measures:
  - Kyrgyz inflation responds strongly to monetary policy changes when monetary aggregates / liquidity measures are used instead of the policy rate.
- Interpretation:
  - Evidence points to a misalignment between the key policy rate (increased since 2021) and expanding market liquidity.
  - Market liquidity and the policy rate have moved in tandem (whereas under normal market conditions they tend to exhibit an inverse relationship).
- Policy implication:
  - Effectiveness of policy interest rates can be strengthened by further reducing excess liquidity.

### SVAR-based policy implications and recommendations (inflation focus)
- Reduce excess liquidity and align it with policy interest rates to strengthen monetary policy transmission and the central bank’s ability to manage inflation.
- Prudent fiscal and public wage policies to support disinflation efforts.
- Continued sterilization and foreign exchange interventions are costly but necessary given prior liquidity injections from gold operations.
- Discontinue central bank participation in the domestic gold market, identified as the main source of liquidity injections.

### Trade patterns, external balance, and re-exports since the start of Russia’s war in Ukraine
- Trade pattern changes (2022 vs. 2021 and historical averages):
  - Officially recorded exports decreased by 18 percent while imports increased by 76 percent compared to 2021.
  - Share of imports from China rose from 26 percent in 2021 to almost 42 percent in 2022 (historical average 22 percent, 2005–2018).
  - Russia received almost 48 percent of Kyrgyz total exports in 2022, up from 14.3 percent in 2021 and historical average of 14.1 percent (48 percent excludes unrecorded re-exports).
- External balances:
  - Current account deficit widened to 43.6 percent of GDP in 2022, up from 8 percent in 2021 (historical average 9.2 percent, 2005–2018).
  - Trade deficit widened from 26.1 percent of GDP in 2021 to 58.5 percent in 2022.
- Import surge and composition:
  - Total imports jumped to 77.2 percent of GDP in 2022 from 56.2 percent of GDP in 2021 (regional CCA average 31.7 percent in 2022).
  - Increase in non-oil imports: 81.9 percent in 2022 versus 48.5 percent growth in the previous year; region increase 22.4 percent in 2022.
  - Fast-growing import categories: household electronics and consumption/intermediate goods (dishwashers, sewing and washing machines, apparel, shoes, computers and accessories, construction equipment, cell phones and cars).
  - Official data show fast-growing exports in same product categories, implying part of imports were re-exported.
- Under-recorded re-exports and BoP implications:
  - Errors and omissions in the BoP amounted to nearly 37 percent of GDP in 2022.
  - Staff estimates unrecorded re-exports account for about 30 percent of total imports and are primarily destined to Russia.
  - Using staff estimates:
    - Russia’s share in Kyrgyz trade rises to 43 percent for total trade turnover and 77 percent for exports.
  - Supporting evidence for large unrecorded re-exports:
    - Near doubling of non-oil imports not explained by domestic consumption even after migrant demand.
    - Other CCA countries with import surges also recorded increases in re-exports.
    - No plausible alternative explanation: net money transfers declined by 21 percent; services and income BoP lines consistent with historical trends; limited access to international capital markets; no major FDI in 2022 known to staff.

### Economic costs of fragmentation and trade vulnerability
- Global fragmentation risks elevated (China–U.S. tensions; war in Ukraine; western sanctions on Russia), contributing to greater trade concentration for Kyrgyz Republic.
- Long-term cost estimates of fragmentation:
  - Bolhuis et al. based estimate for the Kyrgyz Republic: 2.7 percent of GDP.
  - Average cost for Central Asia: 2.2 percent of GDP.
  - Caucasus countries in model: 4.4 percent of GDP.
- Model notes:
  - Based on production and trade model using 2019 data for over 100 tradable products, and demand and trade elasticities.
  - Short-term costs typically twice long-term costs (Bolhuis et al.; Boehm et al.).
  - Fragmentation scenarios range from 1 percent of global GDP losses (strategic decoupling) to around 12 percent of GDP in severe disintegration.

### Tax system overview, fragmentation, and reform potential
- Tax collection and structure:
  - Kyrgyz Republic collects roughly 20 percent of GDP in tax revenues; broadly stable between 19 and 21 percent of GDP since 2014.
  - 2022: tax revenues increased to 23.1percent of GDP on the back of higher imports and a one-off payment by Kumtor.
  - Main tax revenue sources:
    - VAT: around 8 percent of GDP and over 40 percent of total tax revenue.
    - Taxes on income (PIT and CIT): around 5 percent of GDP and represent 25–30 percent of tax revenues.
    - Customs duties via EAEU: 3 percent of GDP or 15 percent of total taxes.
- Special tax regimes and fragmentation:
  - Numerous special regimes: Patent regime; Single Tax (turnover-based 0.5–8 percent); 5 Free Economic Zones; High Technology Park (PIT at reduced rate 5 percent); crypto mining regime (10 percent on value of electricity consumed); electronic commerce regime (2 percent turnover tax); gambling and trading zone regimes.
  - Social contributions around 5 percent of GDP (out of scope of paper).
  - EAEU membership since 2015: country applies common external tariff and receives 1.9 percent of all customs duties collected by EAEU.
- Tax potential, effort, and gap:
  - Tax potential (SFA, GTRE model): around 35 percent of GDP.
  - Comparable IMF estimate: 34.9 percent of GDP.
  - Tax effort: around 60 percent.
  - Tax gap (difference between actual revenue and potential): 13 percent of GDP.
  - Peer comparisons: peer tax potential 33 percent of GDP; peer tax effort 54 percent. Kyrgyz tax effort 60 percent.
- Reform scenario and revenue potential:
  - Reducing tax gap by one third over three years would generate 4 percent of GDP in additional revenues.
  - Tax effort would rise from 60 to 75 percent.
  - Tax-to-GDP ratio would sustainably increase from 20 percent of GDP in 2021 to 24 percent of GDP.

### VAT, sales tax, and income tax specifics
- VAT:
  - Average revenue since 2014: 8.5 percent of GDP.
  - VAT rate: 12 percent.
  - VAT c-efficiency: 70 percent.
  - Registration threshold: KGS 30 million (US$340,000).
  - Authorities and IMF estimate VAT tax expenditures at around 2 percent of GDP.
  - IMF VAT modelling attributes majority of VAT tax expenditures to exemptions on agricultural goods and services, passenger transportation, and pharmaceutical products.
  - Additional compliance gap could be as high as 1 percent of GDP.
  - VAT collection dynamics:
    - Domestically collected VAT stable at 2 percent of GDP over 2018–2022.
    - VAT collected at the border increased from 7 percent of GDP before the pandemic to 9 percent of GDP in 2022.
  - VAT refunds:
    - Paid refunds increased nominally from KGS 333 million in 2018 to around KGS 500 million in 2022.
    - Ratio of refunds to net VAT collected: ~0.6 percent pre-pandemic; jumped to over 1 percent in 2020; dropped to 0.4 percent in 2022.
- Sales tax:
  - Revenue productivity: about 0.8 percent of GDP.
  - Sales tax estimated tax expenditures: 1.5 percent of GDP.
  - Sales tax breaches neutrality and incentivizes vertical integration; domestically produced goods, including exports, taxed more heavily than imported products not subject to sales tax.
  - Policy alternative: elimination of sales tax could be compensated by raising VAT to around 15 percent.
- PIT and CIT:
  - Flat 10 percent rate for both PIT and CIT.
  - 2014–2022: PIT and CIT generated around 5 percent of GDP annually.
  - About 2/3 of income taxes paid by corporates and 1/3 by individuals (includes revenues from special regimes).
  - Authorities estimate tax expenditures: PIT 1.5 percent of GDP; CIT 0.3 percent of GDP.
  - Staff estimates combined PIT and CIT tax gap: 2–2.5 percent of GDP.
  - Staff-derived tax potential for PIT and CIT combined: around 7 to 7.5 percent of GDP under existing policies, implying a tax gap of 2 to 2.5 percent of GDP.
- Reform options and revenue-raising PIT scenario:
  - Modern anti-avoidance rules: limit interest deductions; strengthen transfer pricing.
  - Simplify presumptive regimes: two streamlined regimes recommended (patent for micro-businesses; presumptive turnover-based tax for small businesses); apply general regime broadly.
  - Progressive PIT staff scenario:
    - 10 percent rate on all income,
    - 20 percent rate on income above KGS 177 thousand (mean of household income),
    - 35 percent top rate applicable to income above KGS 229 thousand (household income of 7th decile),
    - Could yield up to 1.5 percent of GDP in additional revenue.
  - Consider a dual income tax where capital income taxed at a lower flat rate and labor income taxed progressively.

### Tax expenditures (authorities' estimates) — total ~6 percent of GDP
- PIT (percent of GDP): 2016 1.50%; 2017 1.39%; 2018 1.35%; 2019 1.29%; 2020 1.47%; 2021 1.50%; 2022 1.33%
- CIT (percent of GDP): 2016 0.16%; 2017 0.11%; 2018 0.13%; 2019 0.26%; 2020 0.22%; 2021 0.31%; 2022 0.41%
- VAT (percent of GDP): 2016 2.30%; 2017 2.01%; 2018 1.80%; 2019 1.72%; 2020 1.89%; 2021 2.62%; 2022 2.62%
- Sales tax (percent of GDP): 2016 0.70%; 2017 1.04%; 2018 1.08%; 2019 1.12%; 2020 1.63%; 2021 1.64%; 2022 1.76%
- Property tax (percent of GDP): 2016 0.02%; 2017 0.02%; 2018 0.02%; 2019 0.02%; 2020 0.02%; 2021 0.02%; 2022 0.02%
- Land tax (percent of GDP): 2016 0.07%; 2017 0.06%; 2018 0.06%; 2019 0.05%; 2020 0.05%; 2021 0.04%; 2022 0.07%
- Total tax expenditures (percent of GDP): 2016 4.75%; 2017 4.64%; 2018 4.43%; 2019 4.46%; 2020 5.28%; 2021 6.13%; 2022 6.20%

### Key tax policy recommendations and priorities
- Reduce fragmentation: rationalize special regimes and limit derogations from the general tax regime.
- Strengthen VAT policy:
  - Reassess VAT exemptions and zero-rating (major expenditures: agriculture, passenger transport, pharmaceuticals).
  - Address unexplained decline in refunds relative to VAT collected; ensure refund system aligns with destination principle.
  - Consider replacing sales tax with a modest VAT rate increase (around 15 percent).
- Strengthen income tax base and design:
  - Introduce modern anti-avoidance rules (interest limitation, transfer pricing).
  - Simplify presumptive regimes into two streamlined regimes and broadly apply general regime.
  - Consider progressive PIT or dual income tax to raise up to 1.5 percent of GDP and improve redistribution.
- Target tax expenditures with better alternatives:
  - Replace poorly targeted VAT exemptions with targeted means-tested cash transfers to protect poorest households at lower fiscal cost.
- Improve revenue administration:
  - Update taxpayer registry, segment taxpayers, strengthen risk-based compliance, continue e-filing rollout and analytical capacity building.

### Kumtor gold mine taxation and fiscal significance
- Ownership and macro significance:
  - Kumtor is now fully state-owned.
  - 2021 sectoral contribution: Kumtor accounted for 8 percent of GDP and 15 percent of GDP in exports.
- Kumtor direct levies (percent of gross proceeds):
  - 13 percent gross proceeds tax.
  - 1 percent contribution to the Issyk-Kul Region Development Fund.
  - 4 percent for replenishment of the mineral reserves fund.
  - 0.4 percent contribution to the Fund of Social Partnership for Regional Development.
  - 0.6 percent contribution to the Fund for Development of the Naryn Region.
  - Aggregate effect: royalty-type taxation of 19 percent of gross proceeds.
  - With full government ownership, remaining profits should be distributed to the state in the form of dividends.
- General mining regime and investor attractiveness:
  - Production-based royalty: 3–5 percent of gross proceeds.
  - Variable price-based income tax: 1–30 percent of gross proceeds (income tax does not allow cost deduction; functions as production-based royalty).
  - Fiscal modeling (FARI tool) with baseline gold price US$1,900 per ounce:
    - General regime’s average effective tax rate (AETR) is close to 90 percent.
    - Peer countries commonly have AETRs between 30 and 50 percent.
    - Regime is highly regressive (AETR decreases with increasing profitability).
    - Breakeven price for investor to achieve a 15 percent pre-tax hurdle rate: US$ 1,550 per ounce.

### Revenue administration diagnosis and reforms (STS)
- 2016 TADAT assessment: mixed results
  - Strengths: taxpayer services and tax audit planning.
  - Weaknesses: limited accuracy of taxpayer registration database; lack of intelligence on revenue losses; outdated business processes; lack of full automation; absence of risk-based compliance and institutional risk management; limited rollout of e-filing.
- STS reform initiatives underway:
  - Improve taxpayer registry accuracy by developing criteria for inactive taxpayers.
  - Establish monitoring and analytical department for compliance monitoring.
  - Large-scale training program to enhance staff capacity.
  - Introduce mandatory e-filing for payroll withholding tax.
- Priority administrative actions:
  - Update taxpayer registry.
  - Segment taxpayers.
  - Continue reforms to improve taxpayer compliance and reduce tax losses.

### Conclusions and medium-term fiscal opportunity
- Current tax revenue collection: roughly 20 percent of GDP; estimated tax potential around 34 percent of GDP.
- Substantial tax gaps across VAT, CIT, and PIT.
- Medium-term revenue opportunity:
  - Authorities could increase tax revenue by up to 4 percent of GDP over the medium-term through policy and administrative reforms.
- Recommended priority reforms:
  1. Eliminate inefficient VAT exemptions, operationalize refund mechanism, remove sales tax, rationalize special tax regimes.
  2. Increase progressivity of PIT by introducing graduated tax rates.
  3. Improve revenue administration by strengthening taxpayer registry and risk management.
- Role of the Fund:
  - The Fund can support reforms with targeted technical assistance.

*Source: INFLATION: DEVELOPMENTS AND DRIVERS — Kyrgyz Republic, International Monetary Fund (January 19, 2024); IMF staff country report excerpts and Annex I estimations.*

### 1. Regional Comparison _______________________________________________________________ 3

### 1. Regional Comparison

### Introduction and recent inflation trajectory
- The Kyrgyz Republic experienced a notable inflation surge in 2022, with inflation rising from about 6 percent in 2020 to 16.2 percent by February 2023, representing a ten-year peak and surpassing the global average of 8.7 percent and the Caucasus and Central Asia (CCA) average of 13 percent for 2022.
- Headline inflation began receding from March 2023 and reached single digits in August 2023; however, core inflation (excluding food and energy) has remained persistently in double digits.
- Historical inflation spikes in the Kyrgyz Republic occurred during the 2007-08 global food crisis, in 2014 after a substantial depreciation of the Russian ruble, and most recently after the COVID pandemic and Russia’s war in Ukraine—each episode coinciding with notable increases in global food and energy prices.

### Inflation dynamics and breadth
- Inflation has become more broad-based:
  - In 2019, 5 out of 47 CPI components had annual inflation over 10 percent; by August 2023 this number increased to 32.
  - The number of CPI components with annual inflation less than 5 percent fell from 45 in 2019 to 6 by August 2023.
- Inflation in the Kyrgyz Republic has been strongly correlated with global food and energy prices because imports of these goods account for a significant share of the CPI basket.
- The recent episode deviated from the established pattern: after the initial surge prompted by global food prices, domestic inflation showed resilience to their subsequent decline, suggesting increasing influence of domestic demand factors and non-food and services contributions since 2021.

### Policy measures and administratively suppressed prices
- Government interventions to safeguard food security and limit price fluctuations include export bans, direct price regulations, and tax incentives on imports.
  - In September 2023, the government earmarked 16 goods for price regulation should their prices escalate significantly.
  - A six-month export ban on agricultural goods was instituted in March 2022.
  - A 12 percent import VAT was temporarily abolished on certain grain crops in May 2022.
- These interventions have resulted in administratively suppressed inflation for commodities subject to administrative price controls (Figure 5).

### Monetary policy framework and transmission
- In 2014, the Kyrgyz Republic transitioned to an interest rate-based monetary policy framework where the policy rate is established by the Board of the central bank.
- Since 2021, the central bank increased the policy rate to bring inflation down.
- Recent institutional changes:
  - Introduction of the symmetric interest rate corridor earlier this year.
  - Successful de-dollarization by the central bank.
- These changes strengthened the passthrough from the policy rate to interbank rates (Figure 6), but further passthrough to market rates has been subdued by a marked influx of liquidity.
  - The liquidity influx was driven by authorities’ decision to facilitate sales and export of domestically produced gold through the central bank, complicating monetary policy and requiring significant sterilization operations and foreign exchange interventions.
- For the purposes of this paper, liquidity is defined as the sum of short term NBKR notes, overnight deposits, and excess reserves.
- Recently, core inflation seems to pattern well with the amount of liquidity in the system.

### Model approaches and key quantitative findings
- Two complementary approaches are used: an augmented Phillips curve framework (for headline and core inflation separately) and a structural VAR (SVAR).
- Novelty in methods:
  - A comprehensive measure of inflation expectations blending adaptive and forward-looking expectations is constructed using recursive autoregressive four-quarter forecasts and three-year-forward projections from the IMF’s World Economic Outlook, combined via Principal Component Analysis (PCA).
- Augmented Phillips curve results:
  - Strong positive relationship between headline inflation and inflation expectations:
    - Levels specification yields a coefficient of 0.3 for inflation expectations.
    - Standardized specification indicates the relative contribution of inflation expectations exceeds 22 percent.
  - The relationship between the output gap and headline inflation is positive but exhibits weak statistical significance; inclusion of foreign factors further diminishes this association.
  - Average pass-through from exchange rate depreciation to inflation is estimated at 25 percent.
  - Global food prices and foreign CPI are primary determinants of headline inflation:
    - A unit increase in global food prices corresponds to a 0.62-point increase in headline inflation.
    - A unit increase in foreign CPI corresponds to a 0.34-point increase in headline inflation.
    - By comparison, a unit increase in inflation expectations corresponds to a 0.30-point increase and a unit increase in the output gap corresponds to a 0.12-point increase in headline inflation (Panel 1).
  - For core inflation:
    - Strong positive correlation with inflation expectations, with a coefficient over 0.5, indicating even higher persistence than for headline inflation.
    - Domestic factors—especially inflation expectations and wages—have stronger impacts on core inflation, while foreign factors contribute less to core inflation.

### Structural VAR findings and policy implications
- SVAR analysis uncovers a misalignment between the central bank’s policy rate and market liquidity, which weakens interest rate pass-through and monetary policy effectiveness.
- Policy implications and recommendations emphasized in the analysis:
  - Reduce excess liquidity and align it with policy interest rates to strengthen monetary policy transmission and the central bank’s ability to manage inflation.
  - Prudent fiscal and public wage policies would support the central bank’s efforts to lower inflation.
  - Continued focus on sterilization and foreign exchange interventions is costly but necessary given prior liquidity injections from gold operations.

*Source: INFLATION: DEVELOPMENTS AND DRIVERS — Kyrgyz Republic, International Monetary Fund (January 19, 2024).*

### 13.      SVAR allows to expand the analysis of inflation in the Kyrgyz Republic by introducing

### 13.      SVAR allows to expand the analysis of inflation in the Kyrgyz Republic by introducing 

### SVAR methodology and model setup
- SVAR augments regular VAR by allowing inclusion of exogenous variables (e.g., global food prices and global CPI) that are unaffected by domestic factors.
- Baseline SVAR variables: headline and core inflation, global food prices, nominal effective exchange rate (NEER), central bank policy rate, market lending rates, government borrowing from the domestic banking sector, private sector credit, and the output gap.
- Alternative SVAR specifications: replace interest-rate (price) instruments with quantity variables (monetary aggregates, liquidity measures) to probe pass-through mechanisms.
- Identification: Cholesky decomposition (Stock and Watson, 2001) used to assign different contemporaneous endogeneity levels; A is lower triangular with ones on the diagonal (8 x 8 system).
- Estimated form: A Y_t = B(L) Y_{t-1} + ε_t with A being 8 x 8, Y_t an 8 x 1 vector, B(L) an 8 x 8 polynomial in lag operator L, and ε_t an 8 x 1 vector of structural shocks.

### Key impulse-response findings
- Global food prices and NEER strongly affect headline inflation, consistent with augmented Phillips curve results.
- Quantified effects (standardized variables):
  - A one standard deviation shock in global food prices raises headline inflation by nearly 1.5 standard deviations in six quarters.
  - A one standard deviation increase in claims on the government (proxy for fiscal policy) leads to roughly a 0.8-unit surge in headline inflation.
- Pass-through from policy rate to bank lending and deposit rates:
  - There is some pass-through from the policy rate to bank lending rates; the pass-through to deposit rates is less pronounced and has longer lags than for lending rates.
  - Baseline SVAR: central bank policy rate shows limited direct impact on inflation.
- Credit and output:
  - Inflation shows limited sensitivity to lending and deposit rates, to changes in private sector credit, and to the output gap (see Figure 8 and Figure 9 in the source).

### Monetary policy transmission and liquidity
- Alternative SVAR specification using liquidity conditions (excess liquidity over broad money or reserve money) finds:
  - Kyrgyz inflation responds strongly to changes in monetary policy when monetary aggregates / liquidity measures are used instead of the policy rate.
- Interpretation:
  - Evidence points to a likely misalignment between the key policy rate (which was increased considerably since 2021) and expanding market liquidity.
  - Market liquidity and the policy rate have moved in tandem, whereas under market conditions they tend to exhibit an inverse relationship.
- Policy implication:
  - Effectiveness of policy interest rates can be strengthened by further reducing excess liquidity.

### Conclusions and policy recommendations
- Methodology: the paper uses two complementary approaches—an augmented Phillips Curve framework and a structural VAR—and constructs a comprehensive measure of inflation expectations combining adaptive and forward-looking components.
- Determinants of inflation:
  - Global food prices remain a historical driver.
  - Domestic factors now play a more prominent role: the exchange rate, public sector wages, and inflation expectations underpin inflation persistence.
  - Incomplete pass-through from central bank policy rates to inflation: policy rates affect market rates to some degree, but transmission to credit growth, output, and ultimately inflation is limited.
  - Liquidity conditions contribute strongly to inflation, indicating a disconnect between policy interest rates and market liquidity.
- Recommended policy actions:
  - Drain excess liquidity and align market liquidity with the policy rate to strengthen monetary policy transmission and support the National Bank of the Kyrgyz Republic’s transition to an inflation targeting framework.
  - Discontinue central bank participation in the domestic gold market, identified as the main source of liquidity injections.
  - Implement supportive fiscal and public wage policies.
  - Reduce reliance on price controls, which distort markets and undermine transmission channels.

### Annex I — Model specifications, data, and robustness
- Sample coverage: quarterly data from 2001 through the first half of 2023.
- Variable standardization: estimates presented from standardized variables (zero mean, unit standard deviation); headline and core consumer price inflation used as dependent variables.
- Inflation expectations: constructed to reflect both forward-looking and historical/adaptive perspectives of households and firms.
- Addressing collinearity among global factors:
  - Two-step approach: (i) isolate components of global food prices unaffected by global oil prices; (ii) identify components of global CPI unrelated to global oil and food; use isolated components to avoid model overidentification.
- Monetary and fiscal policy proxies:
  - Monetary policy: central bank policy rate, market lending and deposit rates, monetary aggregates.
  - Fiscal policy: credit to the government from the banking system.
- Output gap: difference between actual output and trend using Hodrick-Prescott filter with a 1600 smoothing parameter.
- Robustness checks:
  - Recursive model with a flexible window across the 2001–first half of 2023 sample validates coefficient consistency.
  - Notable coefficient magnitude changes around 2014 (strong Russia currency depreciation) and 2021 (COVID pandemic); Chow breakpoint tests failed to reject the null of “no break” for both 2014 and 2021.
- Selected empirical estimates (headline CPI, Annex I. Table 1) — Normal specification and standardized equivalents (Baseline / Aug1 / Aug2):
  - Inflation Expectations: 3.846* (0.000) ; 2.173* (0.000) ; 1.968 (0.000) — Standardized: 0.598* (0.000) ; 0.336* (0.000) ; 0.304* (0.000)
  - Output Gap / Lag: 0.652* (0.000) ; 0.304* (0.007) ; 0.232** (0.042) — Standardized: 0.361 (0.000) ; 0.160** (0.011) ; 0.121*** (0.058)
  - NEER / Lag: -0.273* (0.000) ; -0.301* (0.000) ; -0.287* (0.000) — Standardized: -0.242 (0.000) ; -0.261* (0.000) ; -0.248* (0.000)
  - Global Food Prices / Lag: 0.213* (0.000) ; 0.235* (0.000) — Standardized: 0.567* (0.000) ; 0.623* (0.000)
  - Foreign Inflation / Lag: 1.261* (0.000) ; 1.189* (0.000) — Standardized: 0.365* (0.000) ; 0.344* (0.000)
  - Public Wage / Lag: 0.056^ (0.244) ; 0.053^ (0.265) — Standardized: 0.080^ (0.197) ; 0.076^ (0.215)
  - Global Oil Prices / Lag: 0.022** (0.045) — Standardized: 0.120** (0.047)
  - Adjusted R square (Normal): 0.603 ; 0.771 ; 0.779 — (Standardized): 0.603 ; 0.774 ; 0.783
- Core CPI estimates (Annex I. Table 2) — selected coefficients (Normal / Standardized):
  - Inflation Expectations: 2.417* (0.000) ; 2.096* (0.000) ; 2.132* (0.000) — Standardized: 0.0603* (0.000) ; 0.541* (0.000) ; 0.543* (0.000)
  - Output Gap / Lag: 0.246** (0.011) ; 0.051^ (0.508) ; 0.185** (0.044) — Standardized: 0.173** (0.038) ; 0.090^ (0.210) ; 0.144*** (0.066)
  - NEER / Lag: -0.213* (0.001) ; -0.237* (0.000) ; -0.246* (0.000) — Standardized: -0.225* (0.007) ; -0.305* (0.000) ; -0.343* (0.000)
  - Global Food Prices / Lag: 0.084* (0.001) ; 0.101* (0.000) — Standardized: 0.334* (0.001) ; 0.273* (0.007)
  - Foreign Inflation / Lag: 1.125* (0.000) ; 1.135* (0.000) — Standardized: 0.458* (0.000) ; 0.463* (0.000)
  - Public Wage / Lag: 0.091* (0.001) ; 0.076** (0.027) — Standardized: 0.192* (0.004) ; 0.144** (0.037)
  - Global Oil Prices / Lag: 0.021** (0.014) — Standardized: 0.181** (0.019)
  - Adjusted R square (Normal): 0.489 ; 0.688 ; 0.708 — (Standardized): 0.454 ; 0.690 ; 0.679
- Data sources (Annex I. Table 3) include: National Statistics Committee (NSC), National Bank of Kyrgyz Republic (NBKR), UN FAO Food Price Index, Bloomberg (Brent), IMF WEO (for foreign inflation and inflation expectations), with notes on construction (e.g., foreign inflation constructed using import weights of US (Rest of the World), CHN, RUS and KZ; core inflation excludes food and energy).

*Source: Authors’ estimations and calculations, Annex I of the chapter.*

### 1.      Since the start of Russia’s war in Ukraine, the Kyrgyz Republic’s external trade patterns

### Since the start of Russia’s war in Ukraine, the Kyrgyz Republic’s external trade patterns

### Changes in trade patterns and concentration
- Officially recorded exports decreased by 18 percent while imports increased by 76 percent as compared to 2021.
- The share of imports from China increased from 26 percent in 2021 to almost 42 percent in 2022, compared with the historical average of 22 percent observed from 2005 to 2018.
- Russia received almost 48 percent of the Kyrgyz total exports in 2022, up from 14.3 percent in 2021 and the historical average of 14.1 percent.
- Staff note: the 48 percent does not include unrecorded re-exports to Russia.

### External balance and trade deficit
- The current account deficit widened to 43.6 percent of GDP in 2022, up from the preceding year's 8 percent and well above the historical average of 9.2 percent observed from 2005 to 2018.
- The trade deficit widened from 26.1 percent of GDP in 2021 to 58.5 percent in 2022.
- Kyrgyz Republic’s 2022 current account and trade balance evolution contrasted with other CCA countries, which either shifted to surpluses or significantly reduced their deficits.

### Import surge and composition
- In 2022, total imports jumped to 77.2 percent of GDP from 56.2 percent of GDP in 2021, considerably higher than the regional average of 31.7 percent for the CCA region in 2022.
- The increase in non-oil imports reached 81.9 percent in 2022 as compared to the last year’s 48.5 percent growth and a 22.4 percent increase for the region in 2022.
- Fast-growing import categories include household electronics and consumption and intermediate goods: dishwashers, sewing and washing machines, apparel, shoes, computers and accessories, construction equipment, cell phones and cars.
- Official data show fast-growing exports in the same product categories, implying part of imports were re-exported.

### Under-recorded re-exports and balance of payments implications
- Staff analysis suggests re-exports are substantially under-recorded in official external accounts.
- Errors and omissions in the BoP amounted to nearly 37 percent of GDP in 2022, far exceeding historical levels.
- Because trade within the Eurasian Customs Union (EACU) is not subject to mandatory Customs declarations and inspections, trade is likely significantly under-reported.
- Staff estimates unrecorded re-exports account for about 30 percent of total imports and are primarily destined to Russia.
- Using these estimates, Russia’s share in Kyrgyz trade rises to 43 percent for total trade turnover and 77 percent for exports.
- Supporting factors for the staff assessment:
  - Near doubling of non-oil imports in 2022 cannot be explained by domestic consumption even after accounting for additional demand from migrants.
  - Other CCA countries with import surges also recorded increases in re-exports.
  - No plausible alternative explanation for the unrecorded inflow of foreign exchange: net money transfers, including remittances, declined by 21 percent; services and income BoP lines remained consistent with historical trends; limited access to international capital markets; staff unaware of major FDI in 2022.

### Economic costs of fragmentation and vulnerability
- Global fragmentation risks have increased (heightened China–U.S. tensions; war in Ukraine and western sanctions on Russia), contributing to greater trade concentration for the Kyrgyz Republic.
- The Kyrgyz Republic ranks higher on the human development index than its connectedness level would imply, but stands to gain from better integration into the global economy.
- Long-term cost estimates of fragmentation:
  - A Bolhuis et al. based long-term cost estimate for the Kyrgyz Republic is 2.7 percent of GDP.
  - The average cost for Central Asia is 2.2 percent of GDP.
  - Caucasus countries in the model experience a larger cost of fragmentation at 4.4 percent of GDP.
- Model notes:
  - Estimates are based on a production and trade model using 2019 (pre-Covid) data for over 100 tradable products, and demand and trade elasticities.
  - Short-term costs of fragmentation are typically twice long-term costs (Bolhuis et al.; Boehm et al.).
  - Fragmentation scenarios can range from 1 percent of global GDP losses (strategic decoupling) to around 12 percent of GDP in severe disintegration scenarios.

### Tax potential and fiscal context
- The Kyrgyz Republic collects roughly 20 percent of GDP in tax revenues; tax revenues have been broadly stable between 19 and 21 percent of GDP since 2014.
- In 2022, tax revenues increased to 23.1percent of GDP on the back of higher imports and a one-off payment by Kumtor.
- Main tax revenue sources:
  - VAT generates around 8 percent of GDP and over 40 percent of total tax revenue.
  - Taxes on income (PIT and CIT) generate around 5 percent of GDP and represent between 25 and 30 percent of tax revenues.
  - Customs duties collected through the Eurasian Economic Union (EAEU) contribute 3 percent of GDP or 15 percent of total taxes.
- Fiscal mobilization potential:
  - The analysis finds authorities could sustainably increase tax revenues by 4 percent of GDP over the next three years while maintaining attractiveness for new investors by adopting the following measures:
    1) Eliminating inefficient Value Added Tax (VAT) exemptions, adequately funding and operationalizing the refund mechanism, rationalizing special tax regimes and removing the sales tax,
    2) Increasing the progressivity of the Personal Income Tax (PIT) by introducing graduated tax rates, and
    3) Improving the efficiency of the revenue administration.

*Prepared by IMF staff (Nasir Rao, Jean van Houtte, Thomas Benninger) based on Country Authorities and IMF Staff calculations.*

### 4.      The Kyrgyz tax system is fragmented. Apart from the major national taxes (PIT, CIT, VAT,

### The Kyrgyz tax system is fragmented

### Special tax regimes and non-tax burdens
- Numerous special regimes coexist with major national taxes (PIT, CIT, VAT, excises and sales tax):
  - Patent regime: replaces CIT, sales tax and VAT for specific activities; available for self-employed persons and companies not engaged in importing or exporting goods with turnover of less than KGS 8 million (US$ 90,000) over a 12-month period.
  - Single Tax: turnover-based tax with rates between 0.5 and 8 percent; replaces CIT, sales tax and VAT for eligible taxpayers. Until 2023, eligibility was turnover below KGS 30 million (US$ 340,000) and all companies in the clothing and textile industry; the threshold was removed and the regime is now available to all companies.
  - Free Economic Zones (5 FEZs): Bishkek FEZ, Maimak FEZ, Naryn FEZ, Karakol FEZ, and Leilek FEZ. Provide special customs privileges; companies may import, store, produce or sell goods in the zone without paying taxes or customs duties; entities may sell goods from FEZs to the domestic market if products have undergone significant processing in the zone.
  - High Technology Park: special tax zone for software development, exports, and IT service centers; exempt from CIT, VAT, and sales tax; PIT is paid at the reduced rate of 5 percent.
  - Crypto mining regime: applicable to companies mining crypto currencies; applied in lieu of income tax, VAT on taxable supplies and sales tax; levied at 10 percent on the value of electricity consumed.
  - Electronic commerce regime: special regime for companies supplying electronic services; subject to a 2 percent turnover tax in lieu of income tax, VAT on taxable supplies and sales tax.
  - Gambling activity regime: applied in lieu of income tax, VAT on taxable supplies and sales tax; levied as a specific tax based on number of slot machines and gaming tables for activities with physical presence and at 4 percent of turnover for online casinos.
  - Trading zones regime: applicable to trading zones selling goods; applied in lieu of income tax, VAT on taxable supplies and sales tax; levied at variable rates at around 0.5 percent of the expected turnover.
- Social contributions are around 5 percent of GDP; these are out of scope of this paper but important for the effective tax and non-tax burden on corporates and individuals.
- The Kyrgyz Republic joined the EAEU in 2015; the country applies the common external tariff and receives 1.9 percent of all custom duties collected by the EAEU.

### Tax frontier analysis: potential, effort, and gaps
- Estimated tax potential and effort:
  - Tax potential (SFA, GTRE model): around 35 percent of GDP.
  - Comparable IMF estimate: 34.9 percent of GDP.
  - Tax effort: around 60 percent.
  - Tax gap (difference between actual tax revenue and potential): 13 percent of GDP.
- Peer comparisons:
  - Average tax potential for peer countries (2014-2021): 33 percent of GDP.
  - Peer tax effort: 54 percent of their tax potential.
  - Kyrgyz Republic tax effort: 60 percent (above regional peers).
- Reform scenario:
  - Reducing the tax gap by one third over three years would generate 4 percent of GDP in additional revenues.
  - This would increase tax effort from 60 to 75 percent.
  - Tax-to-GDP ratio would sustainably increase from 20 percent of GDP in 2021 to 24 percent of GDP.

### Estimated tax gaps and VAT / Sales Tax analysis
- VAT performance:
  - VAT on average generated 8.5 percent of GDP in revenue since 2014.
  - VAT tax rate: 12 percent.
  - VAT c-efficiency: 70 percent.
  - Registration threshold: KGS 30 million (equivalent to US$340,000).
  - Authorities and IMF estimate VAT tax expenditures at around 2 percent of GDP.
  - IMF VAT modelling indicates majority of VAT tax expenditures relate to exemption of supply of agricultural goods and services, passenger transportation and pharmaceutical products.
  - IMF modelling also points to an additional compliance gap which could be as high as 1 percent of GDP.
- VAT collection dynamics:
  - Domestically collected VAT remained stable at 2 percent of GDP over 2018-2022.
  - VAT collected at the border increased from 7 percent of GDP before the pandemic to 9 percent of GDP in 2022, linked to increased imports.
  - Staff believes part of additional imports is reexported, implying higher VAT refunds under the destination principle.
- VAT refunds and ratios:
  - Paid refunds increased nominally from KGS 333 million in 2018 to around KGS 500 million in 2022.
  - Ratio of refunds to net VAT collected: stable at around 0.6 percent before the pandemic; jumped to over 1 percent in 2020 when VAT collected collapsed; dropped to 0.4 percent in 2022 when VAT collected increased to 12 percent of GDP.
- Sales tax assessment:
  - Sales tax revenue productivity: about 0.8 percent of GDP.
  - Sales tax estimated tax expenditures: 1.5 percent of GDP.
  - Sales tax breaches internal and external neutrality and incentivizes vertical integration; domestically produced goods, including exports, are taxed more heavily than imported products not subject to the sales tax.
  - Policy alternative: elimination of the sales tax could be compensated by bringing the VAT rate to around 15 percent (still below regional and income group peers).

### Income taxes (PIT and CIT): structure, revenues, and gaps
- Rates and yield:
  - PIT and CIT apply a flat 10 percent rate.
  - For 2014-2022, PIT and CIT generated around 5 percent of GDP in revenues annually.
  - About ⅔ of income taxes were paid by corporates and ⅓ by individuals (includes revenues from special regimes).
- Tax expenditures and estimated gaps:
  - Authorities estimate tax expenditures at 1.5 percent of GDP for the PIT and 0.3 percent for the CIT.
  - Staff estimates the combined tax gap (PIT and CIT) could be closer to 2–2.5 percent of GDP.
  - Using national accounts data, staff estimates tax potential for PIT and CIT combined around 7 to 7.5 percent of GDP under existing policies, implying a tax gap of 2 to 2.5 percent of GDP.
- Risks and reform options:
  - Multiple derogatory regimes create fragmentation and risk base erosion and profit shifting as taxpayers shift profits to regimes with lowest marginal tax rates.
  - Modern anti-avoidance rules recommended: limit interest deductions; strengthen transfer pricing rules to protect the CIT base.
  - Simplify presumptive regimes: system is overly complex with various thresholds and rates; informality is estimated at over 30 percent of GDP. Prior IMF technical assistance recommended two simplified regimes: a patent system for micro-businesses and a presumptive turnover-based tax for small businesses; apply the general regime with few or no exceptions.
- Personal Income Tax reform scenario:
  - A more progressive PIT could generate additional revenue and enhance redistribution.
  - Staff estimates that a progressive rate schedule with:
    - 10 percent rate on all income,
    - 20 percent rate on income above KGS 177 thousand (mean of household income),
    - 35 percent top rate applicable to income above KGS 229 thousand (household income of 7th decile),
    - could yield up to 1.5 percent of GDP in additional revenue.
  - Consideration: move to a dual income tax where capital income is taxed at a lower flat rate and labor income is taxed progressively.

### Tax expenditures table (authorities' estimates)
- Authorities estimate total tax expenditures at around 6 percent of GDP. [in percent of GDP]
  - PIT: 2016 1.50%; 2017 1.39%; 2018 1.35%; 2019 1.29%; 2020 1.47%; 2021 1.50%; 2022 1.33%
  - CIT: 2016 0.16%; 2017 0.11%; 2018 0.13%; 2019 0.26%; 2020 0.22%; 2021 0.31%; 2022 0.41%
  - VAT: 2016 2.30%; 2017 2.01%; 2018 1.80%; 2019 1.72%; 2020 1.89%; 2021 2.62%; 2022 2.62%
  - Sales tax: 2016 0.70%; 2017 1.04%; 2018 1.08%; 2019 1.12%; 2020 1.63%; 2021 1.64%; 2022 1.76%
  - Property tax: 2016 0.02%; 2017 0.02%; 2018 0.02%; 2019 0.02%; 2020 0.02%; 2021 0.02%; 2022 0.02%
  - Land tax: 2016 0.07%; 2017 0.06%; 2018 0.06%; 2019 0.05%; 2020 0.05%; 2021 0.04%; 2022 0.07%
  - Total: 2016 4.75%; 2017 4.64%; 2018 4.43%; 2019 4.46%; 2020 5.28%; 2021 6.13%; 2022 6.20%

### Key policy recommendations and priorities
- Reduce fragmentation by rationalizing special regimes and limiting derogations from the general tax regime.
- Strengthen VAT policy:
  - Reassess VAT exemptions and zero-rating (major tax expenditures relate to agriculture, passenger transport, pharmaceuticals).
  - Address unexplained decline in refunds relative to VAT collected; ensure refund system aligns with destination principle.
  - Consider replacing sales tax with a modest VAT rate increase (around 15 percent) to improve neutrality and revenue productivity.
- Strengthen income tax base and design:
  - Introduce modern anti-avoidance rules (interest limitation, transfer pricing).
  - Simplify presumptive regimes: implement two streamlined regimes (patent for micro, turnover-based presumptive for small businesses) and apply general regime broadly.
  - Consider a progressive PIT or a dual income tax to raise revenue (up to 1.5 percent of GDP in staff scenario) and improve progressivity.
- Target tax expenditures with better alternatives:
  - Replace poorly targeted VAT exemptions with targeted means-tested cash transfers to protect poorest households at lower fiscal cost.

*Source: IMF staff report chapter on the Kyrgyz Republic tax system.*

### 23.      Kumtor—the largest mining company—operates under a special tax regime. The Kyrgyz

### 23. Kumtor—the largest mining company—operates under a special tax regime.

### Kumtor taxation and macroeconomic significance
- Kumtor gold mine is now fully state-owned.
- Sectoral contribution (2021):
  - Kumtor accounted for 8 percent of GDP.
  - Kumtor accounted for 15 percent of GDP in exports.
- Main direct taxes and contributions paid by Kumtor (all expressed as percent of gross proceeds):
  - 13 percent gross proceeds tax.
  - 1 percent contribution to the Issyk-Kul Region Development Fund.
  - 4 percent for the replenishment of the mineral reserves fund.
  - 0.4 percent contribution to the Fund of Social Partnership for Regional Development.
  - 0.6 percent contribution to the Fund for Development of the Naryn Region.
- Aggregate effect: royalty-type taxation of 19 percent of gross proceeds.
- With full government ownership, remaining profits should be distributed to the state in the form of dividends.

### General mining tax regime and investor attractiveness
- Structure of the general mining regime:
  - Production-based royalty between 3 and 5 percent of gross proceeds.
  - Variable price-based income tax ranging between 1 and 30 percent of gross proceeds.
  - Income tax for the mining sector does not allow cost deduction and therefore functions effectively as a production-based royalty.
- Fiscal modeling (IMF’s Fiscal Analysis of Resource Industries (FARI) tool) findings:
  - Baseline scenario gold price: US$1,900 per ounce.
  - The general regime’s average effective tax rate (AETR) is close to 90 percent.
  - Peer countries commonly have AETRs between 30 and 50 percent.
  - The regime is highly regressive (AETR decreases with increasing profitability over the life of the project).
  - Breakeven price for an investor to achieve a hurdle rate of 15 percent pre-tax: US$ 1,550 per ounce.

### Revenue administration—diagnosis and reforms
- 2016 TADAT assessment of the State Tax Service (STS) — mixed results:
  - Strengths: taxpayer services and tax audit planning.
  - Weaknesses:
    - Limited accuracy of the taxpayer registration database.
    - Lack of intelligence on potential revenue losses.
    - Outdated tax administration business processes.
    - Lack of full automation of core tax administration.
    - Absence of a risk-based approach to compliance and institutional risk management.
    - Limited rollout of e-filing.
- STS reform initiatives underway:
  - Improving accuracy of the taxpayer registration database by developing criteria for inactive taxpayers.
  - Establishing a monitoring and analytical department for compliance monitoring.
  - Large-scale training program to enhance staff capacity.
  - Introduction of mandatory e-filing for payroll withholding tax.
- Priority actions recommended for administration:
  - Update the taxpayer registry.
  - Segment taxpayers.
  - Continue reforms to improve taxpayer compliance and reduce tax losses.

### Conclusions and policy recommendations
- Fiscal context and potential:
  - Current tax revenue collection: roughly 20 percent of GDP.
  - Estimated tax potential: around 34 percent of GDP.
  - Substantial tax gaps exist across VAT, CIT, and PIT.
- Medium-term revenue opportunity:
  - The authorities could increase tax revenue by up to 4 percent of GDP over the medium-term.
- Recommended policy and administrative reforms to realize revenue gains:
  1. Eliminate inefficient VAT exemptions, operationalize the refund mechanism, remove the sales tax, and rationalize special tax regimes.
  2. Increase progressivity of the PIT by introducing graduated tax rates.
  3. Improve revenue administration by strengthening the taxpayer registry and risk management.
- Role of the Fund:
  - The Fund can support these reforms with targeted technical assistance.

*Source: IMF staff country report excerpt.*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2024/english/1kgzea2024002.pdf_