## Spillovers from Russia to Neighboring Countries: Transmission Channels and Policy Options

## Source details

**Canonical URL:** [Spillovers from Russia to Neighboring Countries: Transmission Channels and Policy Options](https://www.imf.org/-/media/files/publications/wp/2023/english/wpiea2023185-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2023/english/wpiea2023185-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2023/english/wpiea2023185-print-pdf.pdf.json)

---

### Executive summary
- Russia’s output fluctuations are an important driver of output fluctuations of countries in the region, especially for oil importers.
- Spillovers are transmitted increasingly via trade and market confidence channels.
- The magnitude of cross-border spillovers is larger for countries with relatively high bilateral trade concentration, low export diversification, and weak external buffers.
- Stronger public institutional quality—especially in the fiscal area—may help insulate countries from volatility in the Russian sovereign debt market.
- Policy implications emphasized: diversify trade and product markets (including through stronger regional integration), reduce liability dollarization, and strengthen public institutions.

### Introduction and scope
- Research questions: How do recessions in Russia affect neighboring countries? How can policies dampen their effects?
- Empirical approach:
  - Quarterly data on 8 neighboring countries over 2004-2021 using VAR and dynamic panel models.
  - Country-specific regressions using quarterly GDP on 32 advanced and emerging market countries from 2004-19 to analyze heterogeneity of spillovers.
- CIS countries covered: Armenia, Azerbaijan, Belarus, Kazakhstan, Moldova, Tajikistan; Georgia and Ukraine included for geographic proximity and large economic ties to Russia.
- Main findings preview:
  - Spillovers from Russia are sizable and larger for oil importers.
  - Trade, remittance, and market confidence channels matter; remittance channel declined over time while trade and market confidence channels rose.
  - Structural factors such as trade concentration in commodities, current account deficits, shares of foreign currency-denominated debt, and weak public institutions amplify spillovers.

### Stylized facts — Trade linkages
- Long-run change in trade with Russia:
  - Exports to Russia fell from over a third of total exports in 1997 to 15 percent as of 2021.
  - Imports from Russia fell from 40 to 30 percent (implicit over comparable period up to 2021).
- Changing partners by 2021:
  - China comprised 15% of the region’s total exports and 12% of its imports as of 2021.
  - The EU is the region’s largest export market and second-largest import market.
- Product composition shifts:
  - Machinery and mechanical goods shares fell for Armenia, Azerbaijan, and Georgia; shares of food, agricultural, and base metals rose.
  - Kazakhstan: base metals rose from 9 percent of bilateral exports in 2000 to 24 percent as of 2020.
  - Belarus: manufactured goods and machinery share fell while food and agricultural products share grew.
- Imports dependence by product and country:
  - Imports from Russia are largely fuels, metals, and food; Armenia, Georgia, Ukraine, and Belarus have large minerals and fuels shares; Kazakhstan, Azerbaijan, and Moldova have more dispersed import linkages as of 2020.

### Financial linkages — FDI, portfolio, sovereign premia, remittances, exchange rates
- FDI and portfolio investment:
  - Largest recipients of Russian FDI (latest available): Armenia: 12.4 percent of GDP; Kyrgyz Republic: 12.3 percent of GDP; Moldova: 7.1 percent of GDP; Belarus: 6.9 percent of GDP.
  - Azerbaijan, Kazakhstan, Georgia, and Ukraine receive less than 4 percent of GDP of Russian FDI.
  - Outward FDI flows to Russia are under 2 percent of GDP where data available.
  - Portfolio investment held by Russian residents in CIS countries is small: Ukraine: 2.3 percent of GDP as of 2014 (largest); Belarus and Kazakhstan next largest; others under 0.5% of GDP.
- Sovereign risk premia:
  - Sovereign bond spreads move closely between Russia and the region; largest convergence with Russia in Ukraine, Kazakhstan, and Belarus, followed by Georgia and Armenia.
  - Correlations computed using monthly data from 2012-2021.
- Remittances (as percent of GDP, 2021):
  - Tajikistan: 19 percent of GDP
  - Armenia: 7 percent of GDP
  - Georgia: 2 percent of GDP
  - Moldova: 2 percent of GDP
  - Remittances from Russia are procyclical with Russia’s non-tradable sector; remittances to the region declined substantially after 2014 and remain well below early-decade levels as of 2021.
- Exchange rate linkages:
  - Ruble/Dollar movements associated with regional currency movements; ruble depreciation (2015 oil price shock) transmitted depreciations to CIS currencies, amplifying debt servicing costs and inflationary pressures.

### Business cycle co-movements and VAR evidence
- Stylized facts: Russian output fluctuations co-move with the region during global shocks (GFC, Covid-19) and Russia-specific shocks (2014 Crimea annexation).
- VAR specification: quarterly q-o-q growth for EU, oil price growth, Russia, and each CIS country; 2004–2019; Cholesky decomposition; two lags; variables stationary in first differences.
- Impulse-response to a 1 standard deviation positive shock to Russia (defined as a 1.8 percentage point increase in Russia’s growth):
  - Kazakhstan: increase of 0.2 percentage points in first quarter and 0.4 percentage points in subsequent quarter.
  - Armenia: spillover peaks at 3 quarters, reaching 0.8 percentage points.
  - Belarus: peaks at 3 quarters, reaching 0.6 percentage points.
  - Ukraine: reaches 1 percentage point in first quarter and 0.7 percentage points in second quarter.
- Comparative EU shock impacts:
  - Kazakhstan: EU shock coefficient peaks at 0.26 percentage points (smaller than Russia’s effect).
  - Georgia: EU spillover peaks after 2 quarters at 0.46 percentage points and lasts up to 7 quarters.
  - Armenia and Ukraine: EU spillovers larger and longer lasting than Russia’s—peaking at 1 and 0.86 percentage points, respectively, and lasting up to 9 quarters.
- FEVD results:
  - Variances of Russia growth explain more forecast error variance of GDP growth than EU for Kazakhstan, Ukraine, Azerbaijan, and Belarus; EU growth variance more important for Armenia, Georgia, and Moldova.

### Panel regression evidence — aggregate spillover magnitudes and drivers
- Baseline panel OLS (quarterly, 2004–2021):
  - A 1 percentage point increase in Russia’s real GDP growth is associated with, on average, a 0.67 percentage point increase in real GDP growth in CIS countries.
  - For oil importers (Armenia, Belarus, Georgia, Moldova, Tajikistan, Ukraine), the coefficient is 0.79 percentage points.
  - A 1 percentage point increase in EU real GDP growth is associated with, on average, a 0.43 percentage point increase in CIS growth.
  - A 1 percentage point increase in China real GDP growth is associated with, on average, a 0.49 percentage point increase in CIS growth.
- Oil price effects:
  - Oil price movements matter for growth across CIS countries, including oil importers; positive oil price coefficient is smaller after controlling for Russia’s GDP growth, suggesting part of oil price effects operate indirectly via Russia.

### Transmission channels — trade, financial, remittances, combined
- Trade channel:
  - Correlation between bilateral exports and growth in Russia has increased for most countries.
  - Regression split results:
    - 2004-2013: a 10-percentage point increase in bilateral export growth associated with a 0.23 percentage point increase in CIS real GDP growth.
    - 2014-2021: a 10-percentage point increase in bilateral export growth associated with a 0.53 percentage point increase in CIS real GDP growth.
  - Exports to the EU insignificant after controlling for the oil price; trade with China not associated with regional GDP growth.
  - Shift toward commodities in bilateral exports increases sensitivity.
- Financial channel:
  - Larger FDI links tend to increase business cycle synchronization.
  - In panel regressions:
    - Increases in Russian equity prices positively associated with local GDP growth.
    - Increases in Russian government bond spreads (EMBIG) negatively associated with local GDP growth; movements in Russia’s government bond spreads have the strongest association with local GDP growth.
    - Central Bank of Russia policy rate movements not associated with local output fluctuations.
  - Data constraints: limited quarterly bilateral FDI flows and lack of cross-border banking claims data for CIS limit direct estimation of some channels.
- Remittance channel:
  - A 10-percentage point increase in growth of remittances from Russia associated with a 0.2 – 0.6 percentage point increase in regional GDP growth.
  - For oil importers, remittance coefficient ranges from 0.3 to 0.8 percentage points.
  - Controlling for EU GDP growth and oil prices dampens remittance coefficients.
  - Data limitation: absence of quarterly real GDP for Kyrgyz Republic and Uzbekistan may understate remittance channel importance.
- Combined regressions (trade, remittance, financial variables; splits 2010-2016 and 2016-2021):
  - 2010-2016: both trade and remittances positively associated with regional output fluctuations.
  - 2016-2021: remittance coefficient becomes insignificant; coefficients on trade and sovereign risk premia are large and significant.
  - Robustness: standardized growth specifications yield similar conclusions (decline in remittance channel; increase in trade and market confidence channels).
  - Caveat: exclusion of Tajikistan, Kyrgyz Republic, and Uzbekistan may bias results toward underestimating remittance importance.

### Role of diversification, integration, institutions, and fundamentals
- Cross-sectional findings:
  - Stronger bilateral trade ties to Russia → larger spillovers.
  - Partner diversification (Global Connectedness - Breadth Index) is negatively associated with spillovers.
  - Higher Economic Complexity Index associated with smaller spillovers; removing Azerbaijan and Kazakhstan strengthens this relationship.
  - Higher share of commodities trade associated with larger spillovers.
  - Macro buffers reduce spillovers: larger current account deficits associated with larger spillovers; higher shares of foreign-currency sovereign debt associated with larger spillovers.
  - Stronger public governance (Government Effectiveness, Rule of Law) correlates with milder spillovers; weaker public governance → sovereign spreads more sensitive to Russian spreads.
- Event study:
  - Large Russian output contractions associate with output and trade contractions in the region.
  - More diversified economies exhibit weaker output and bilateral trade fluctuations during large Russian output swings.

### Selected cross-sectional regression coefficients (Table 6, selected)
- Partner Diversification: -0.757*** (0.134); Observations = 32; R-squared = 0.509
- Partner Diversification (alt): -0.600** (0.272); Observations = 32; R-squared = 0.526
- Trade with Russia (% of total trade): 0.0309*** (0.00585); Observations = 32; R-squared = 0.628
- Commodities trade with Russia (% of GDP): 0.0530*** (0.00925); Observations = 30; R-squared = 0.072
- Economic Diversification: -0.318 (0.227); Observations = 31; R-squared = 0.247
- Government Effectiveness: -0.585*** (0.195); Observations = 31; R-squared = 0.246
- Rule of Law: -0.500** (0.183); Observations = 32; R-squared = 0.170
- Current Account Balance (% of GDP): -0.0240*** (0.00799); Observations = 32; R-squared = 0.397
- GDP Per Capita (PPP, current international $): -0.316*** (0.0763) (one specification shown)

### Policy implications and recommendations
- Reduce trade concentration with Russia and diversify product markets, including through stronger regional integration, to lower trade-induced spillovers.
- Improve public governance—particularly fiscal institutions—to mitigate confidence effects and reduce sovereign spread co-movement with Russia.
- Strengthen macroeconomic buffers:
  - Lower foreign-currency share of sovereign debt.
  - Reduce current account deficits.
  - Pursue de-dollarization to lessen amplification via liability dollarization.
- Continue monitoring shifts in transmission channels—particularly the rising role of trade and market confidence channels and the declining role of remittances.

### Annex III — alternative panel specifications (selected coefficient highlights)
- Russia Real GDP Growth coefficients (Columns (1)–(4)): 0.790***, 0.691***, 0.913***, 0.787***
- EU Real GDP Growth (Columns (1)–(4)): 0.439**, -0.0664, 0.550**, 0.100
- China Real GDP Growth (t-1) (Columns (1)–(4)): 0.513***, 0.201, 0.574**, 0.0672
- Panel sample sizes and fit:
  - Observations Columns (1)–(4): 602; R-squared: 0.296, 0.213, 0.233, 0.301
  - Observations Columns (5)–(8): 450; R-squared: 0.352, 0.264, 0.266, 0.357
- CIS subsample Russia Real GDP Growth significant positive coefficients: 0.533***, 0.461***, 0.587***, 0.507***
- Combined standardized-variable regressions (selected):
  - Export to Russia Growth: 0.345*** (0.123) in one 2016-2021 CIS subsample specification.
  - Bilateral Remittances Growth: 0.150* (0.0782) in one 2010-2016 CIS specification.
  - Δ log Russia EMBIG Spread (t-1): -0.247*** (0.0815) in one 2016-2021 specification.
  - Oil Price Growth: 0.493*** (0.0629) in one 2010-2016 CIS specification.
- Observations in combined standardized regressions range: 119–189 across columns; number of ifs_code per column: 5–7.

*Source: IMF Working Paper — Annex III. Alternative Panel Specifications, Spillovers from Russia to Neighboring Countries: Transmission Channels and Policy Options (Working Paper No. WP/2023/185).*

### EXECUTIVE SUMMARY ___________________________________________________________________________ 4

### Spillovers from Russia to Neighboring Countries: Transmission Channels and Policy Options

### Executive summary
- Russia’s output fluctuations are an important driver of output fluctuations of countries in the region, especially for oil importers.
- Spillovers are transmitted increasingly via trade and market confidence channels.
- The magnitude of cross-border spillovers is larger for countries with relatively high bilateral trade concentration, low export diversification, and weak external buffers.
- Stronger public institutional quality—especially in the fiscal area—may help insulate countries from volatility in the Russian sovereign debt market.
- Policy implications emphasized: diversify trade and product markets (including through stronger regional integration), reduce liability dollarization, and strengthen public institutions.

### Introduction and scope
- Research questions: How do recessions in Russia affect neighboring countries? How can policies dampen their effects?
- Context: Russia’s invasion of Ukraine and international sanctions renewed interest in transmission from Russia to neighboring countries.
- Empirical approach:
  - Quarterly data on 8 neighboring countries over 2004-2021 using VAR and dynamic panel models.
  - Country-specific regressions using quarterly GDP on 32 advanced and emerging market countries from 2004-19 to analyze heterogeneity of spillovers.
- CIS countries covered in the sample: Armenia, Azerbaijan, Belarus, Kazakhstan, Moldova, and Tajikistan. Georgia and Ukraine are included in the sample due to geographic proximity and large economic ties to Russia.
- Main findings preview:
  - Spillovers from Russia are sizable and larger for oil importers.
  - Trade, remittance, and market confidence channels matter; remittance channel declined over time while trade and market confidence channels rose.
  - Structural factors such as trade concentration in commodities, current account deficits, shares of foreign currency-denominated debt, and weak public institutions amplify spillovers.

### Literature synthesis (selected findings)
- Bilateral trade ties are associated with stronger business cycle co-movements (cited studies: Frankel and Rose (1998), Wincoop (2001), Baxter and Koupiratsas (2004), Imbs (2004, 2006)).
- Trade integration can synchronize output via demand declines or similar-industry specialization, but specialization can also produce asymmetric responses.
- Financial integration has mixed effects: it can synchronize cycles (Kose et al. (2003), Imbs (2006), Morgan et al. (2004)) or create divergence except during financial turmoil (Kalemli-Ozcan et al. (2013a)).
- FDI generally synchronizes business cycles via multinational company linkages, though new FDI can reduce synchronization by promoting specialization (Hsu et al. 2011; Stiblarova 2021).
- Other determinants of co-movement include industrial similarity, convergence in monetary/fiscal policies, level of development, and currency union membership.
- Specific to Russia and the region:
  - Previous work highlights trade, financial, and remittance linkages as important channels (Alturki et al., 2009; Stepanyan et al., 2015; IMF 2022a).
  - Remittances are procyclical with the Russian economy and can both exacerbate downturns in recipients and act as shock absorbers during recipient currency depreciation (Poghosyan 2020; IMF 2022b).
- Contribution of this paper: uses more recent and higher-frequency data (quarterly, 2004–2021) to quantify spillovers from Russia and the evolving importance of transmission channels.

### Stylized facts — Trade linkages
- Long-run change in trade with Russia:
  - Exports to Russia fell from over a third of total exports in 1997 to 15 percent as of 2021.
  - Imports from Russia fell from 40 to 30 percent (implicit: over comparable period up to 2021).
- Changing partners by 2021:
  - China comprised 15% of the region’s total exports and 12% of its imports as of 2021.
  - The EU is the region’s largest export market and second-largest import market (no percentage provided).
- Product composition shifts:
  - Since the early 2000s, the share of machinery and mechanical goods in bilateral exports has fallen for Armenia, Azerbaijan, and Georgia; the share of food, agricultural, and base metals has risen.
  - Kazakhstan: the share of base metals in bilateral exports rose from 9 percent of bilateral exports in 2000 to 24 percent as of 2020.
  - Belarus: the share of manufactured goods and machinery in bilateral exports has fallen while the share of food and agricultural products has grown (no percentages provided).
- Imports dependence by product and country:
  - Imports from Russia are largely concentrated in fuels, metals, and food, but vary across countries.
  - For Armenia, Georgia, Ukraine, and Belarus, imports of minerals and fuels from Russia comprise a large share of bilateral imports.
  - Kazakhstan, Azerbaijan, and Moldova have relatively more dispersed import linkages with Russia as of 2020.

### Empirical results and transmission channels (summary)
- VAR evidence (section 4.1): spillovers from Russia’s output fluctuations to CIS economies are sizable, especially for oil importers.
- Panel regressions (sections 4.2 and 5) identify main channels:
  - Trade channel: increased role over time; higher bilateral trade concentration and commodities trade with Russia are associated with larger spillovers.
  - Remittance channel: important historically, but its size has declined over the 2015–2021 period relative to 2010–2014.
  - Market confidence / financial channel: risk-premia convergence and sovereign debt market linkages amplify spillovers; stronger public institutional quality—particularly fiscal institutions—can help insulate countries from volatility in the Russian sovereign debt market.
- Heterogeneity drivers:
  - Larger spillovers associated with higher trade concentration with Russia, low export diversification, larger current account deficits, high shares of foreign currency-denominated debt, and weak public institutions.
- Comparative magnitude:
  - Spillovers from Russia’s output fluctuations are large, including compared to output fluctuations in the EU and China (no exact numeric comparison provided in the excerpt).

### Policy implications
- Diversify trade and product markets, including through stronger regional integration.
- Reduce liability dollarization.
- Strengthen public institutions, with emphasis on fiscal institutional quality to help insulate against sovereign debt market volatility.
- Continue monitoring shifts in transmission channels—particularly the rising role of trade and market confidence channels and the declining role of remittances.

*Source: IMF Working Paper — Executive Summary, Introduction, Literature, and Stylized Facts sections of "Spillovers from Russia to Neighboring Countries: Transmission Channels and Policy Options" (selected excerpts).*

### 3.2 Financial Linkages

### 3.2 Financial Linkages

### FDI and portfolio investment links
- Russia is a large source of Foreign Direct Investment (FDI) for several CIS countries, though for most countries FDI from Russia has fallen since the Global Financial Crisis (GFC).
- Largest recipients of Russian FDI (latest available data):
  - Armenia: 12.4 percent of GDP
  - Kyrgyz Republic: 12.3 percent of GDP
  - Moldova: 7.1 percent of GDP
  - Belarus: 6.9 percent of GDP
- Azerbaijan, Kazakhstan, Georgia, and Ukraine receive less than 4 percent of GDP of Russian FDI.
- Outward FDI flows to Russia are limited: existing data suggest it is small—accounting for under 2 percent of GDP for countries with available data.
- Portfolio investment held by Russian residents in CIS countries is small:
  - Ukraine: 2.3 percent of GDP as of 2014 (largest recipient)
  - Belarus and Kazakhstan: next largest recipients (exact figures implied smaller than Ukraine)
  - Other CIS countries receive under 0.5% of GDP of Russian portfolio investment.

### Other financial linkages and sovereign risk premia
- Sovereign bond spreads move closely between Russia and the region, indicating convergence in perceived government default risk.
- Drivers of common movement likely include common exposures such as geopolitical events, commodity price movements, and global uncertainty; investors may anticipate spillovers from adverse shocks affecting Russia.
- Evidence from 2015: during the oil price collapse, currencies of CIS countries depreciated against the dollar following the Ruble’s depreciation, reflecting anticipation of adverse spillovers from Russia (Stepanyan et al., 2015).
- Countries exhibiting the largest convergence in sovereign risk premia with Russia: Ukraine, Kazakhstan, and Belarus, followed by Georgia and Armenia.
- Correlations of Russian and domestic sovereign bond spreads are computed using monthly data from 2012-2021.

### Remittances (related financial channel)
- Remittances are a large source of external financing and income for several CIS countries; Russia accounts for the vast majority.
- Remittances from Russia as percent of GDP (as of 2021):
  - Tajikistan: 19 percent of GDP
  - Armenia: 7 percent of GDP
  - Georgia: 2 percent of GDP
  - Moldova: 2 percent of GDP
- Remittance flows from Russia tend to be procyclical with Russia’s non-tradable sector where migrant workers from the region tend to work; contractions in Russia propagate via declines in outward remittances.
- Empirical evidence: personal remittance flows from Russia to neighboring countries vary with the Russian business cycle; remittances to the region declined substantially after 2014 and remain well below early-decade levels as of 2021.

### Exchange rate linkages
- Ruble/Dollar exchange rate movements have important implications for CIS countries’ exchange rates.
- Periods of ruble depreciation against the US dollar are associated with depreciations of regional currencies against the dollar.
- During the 2015 oil price shock, the Ruble’s depreciation was an important transmission channel to CIS countries; countries with larger remittance and trade linkages with Russia experienced more severe depreciation, reflecting expected deterioration in current account balances (Stepanyan et al., 2015).
- Synchronized currency depreciation can amplify the impact of Russian output contractions by increasing debt servicing costs and fueling inflationary pressures.

### Business cycle co-movements and empirical evidence
- Stylized facts: Russian output fluctuations tend to move in line with those of the region, during global shocks (GFC, Covid-19) and Russia-specific shocks (2014 Crimea annexation).
- During the GFC, synchronized output contractions in the region were most drastic; Armenia, Georgia, and Ukraine experienced the deepest recessions, while Azerbaijan, Belarus, and Kazakhstan had milder recessions.
- The 2015 oil price shock exacerbated slowdown, especially for oil exporters Azerbaijan and Kazakhstan, but oil importers (Armenia, Belarus, Ukraine) also experienced output contractions.
- Kazakhstan and Russia have particularly strongly correlated business cycles, likely due to common exposure to oil price fluctuations.

### VAR analysis (dynamic transmission)
- VAR specification includes quarterly quarter-on-quarter growth rates for EU, oil price growth, Russia, and each CIS country; quarterly data span 2004–2019.
- Identification: Cholesky decomposition; two lags based on Akaike and Schwarz tests; variables stationary in first differences per ADF tests.
- Key VAR impulse-response findings from a 1 standard deviation positive shock to Russia (defined as a 1.8 percentage point increase in Russia’s growth):
  - Kazakhstan: increase of 0.2 percentage points in the first quarter and 0.4 percentage points in the subsequent quarter.
  - Armenia: spillover peaks at 3 quarters after the initial shock, reaching 0.8 percentage points.
  - Belarus: spillover peaks at 3 quarters, reaching 0.6 percentage points.
  - Ukraine: reaches 1 percentage point in the first quarter and 0.7 percentage points in the second quarter.
- Spillovers from EU GDP growth tend to be milder than those from Russia for several countries:
  - Kazakhstan: EU shock coefficient peaks at 0.26 percentage points (smaller than Russia’s effect).
  - Georgia: EU spillover peaks after 2 quarters at 0.46 percentage points and lasts up to 7 quarters.
  - Armenia and Ukraine: spillovers from EU are larger and longer lasting than those from Russia—peaking at 1 and 0.86 percentage points, respectively, and lasting up to 9 quarters.
- Alternative VAR ordering (oil prices precede EU GDP) yields consistent impulse-response results.

### Forecast Error Variance Decomposition (FEVD) results
- FEVD assesses relative importance of Russia and EU in explaining forecast error variance of GDP growth in CIS countries.
- Variances of growth in Russia are more important than variances of growth in the EU for: Kazakhstan, Ukraine, Azerbaijan, and Belarus.
- Variance of EU growth is more important for: Armenia, Georgia, and Moldova.

### Panel regression results (quarterly, 2004–2021)
- Baseline panel OLS growth regression controls for lagged domestic growth and matrix X of explanatory variables (including EU growth, China growth, oil price fluctuations, US 10-year treasury rate, headline CPI, REER).
- Key coefficients and findings:
  - A 1 percentage point increase in Russia’s real GDP growth is associated with, on average, a 0.67 percentage point increase in real GDP growth in CIS countries.
  - For oil importers (Armenia, Belarus, Georgia, Moldova, Tajikistan, Ukraine), the association is stronger: coefficient of 0.79 percentage points.
  - A 1 percentage point increase in EU real GDP growth is associated with, on average, a 0.43 percentage point increase in CIS growth.
  - A 1 percentage point increase in China real GDP growth is associated with, on average, a 0.49 percentage point increase in CIS growth.
  - When controlling for Russia and China growth, the coefficient on EU growth is not always significant; possible reasons include dispersed links to the EU, oil-dominated trade with the EU for exporters, and indirect effects of EU growth via Russia.
- Oil price movements matter for growth across CIS countries, including oil importers; the positive oil price coefficient is smaller after controlling for Russia’s GDP growth, suggesting part of oil price effects operate indirectly via Russia.

*Source: wpiea2023185-print-pdf - 3.2 Financial Linkages*

### 5. Transmission Channels

### 5. Transmission Channels

### 5.1 Trade Channel
- Empirical exercises (VAR and panel) indicate Russia remains an important driver of regional output fluctuations via trade.
- Correlation between bilateral exports and growth in Russia has increased for most sample countries (Figure 13).
- Regression results (split 2004-2013 and 2014-2021) find:
  - A 10-percentage point increase in bilateral export growth was associated with a 0.23 percentage point increase in CIS real GDP growth over 2004-2013.
  - A 10-percentage point increase in bilateral export growth was associated with a 0.53 percentage point increase in CIS real GDP growth over 2014-2021 (Table 2, columns 4 and 8).
- Exports to the EU are insignificant after controlling for the oil price; significance of EU exports in 2014-2021 disappears when controlling for trade with Russia.
- Trade with China is not associated with GDP growth in the region (historically small trade ties).
- Possible drivers of the increased trade sensitivity:
  - Shift in the share of non-commodities toward commodities in bilateral exports, with commodity goods more sensitive to shocks.
- Caveat: regional, non-Russia trade interlinkages may amplify the association between exports to Russia and local output (e.g., contraction in Russia → decline in imports from Armenia → contraction in Armenia–Georgia trade).

### 5.2 Financial Channel
- Several CIS countries maintain large financial ties with Russia, particularly in FDI; larger FDI links tend to increase business cycle synchronization.
- Sovereign risk premia co-move: CIS sovereign credit spreads move in line with Russia’s spreads, affecting growth through constrained fiscal space or balance sheet effects.
- Panel regressions replacing bilateral exports with Russian financial market variables show:
  - Increases in Russian equity prices are positively associated with local GDP growth (supports FDI linkage transmission hypothesis, though could reflect global risk sentiment).
  - Increases in Russian government bond spreads (EMBIG) are negatively associated with local GDP growth; movements in Russia’s government bond spreads have the strongest association with local GDP growth.
  - Movements in the Central Bank of Russia’s policy rate do not appear associated with local output fluctuations.
- Historical example: the 2015 oil price collapse amplified spillovers via sovereign spread widening and currency pressures.
- Data and identification constraints:
  - Lack of quarterly bilateral FDI flows data (annual data from 2009 only) limits estimation of direct FDI effects.
  - Potential omitted variable bias: other global risk factors could drive output in both Russia and the region, leading to overestimation of Russia-specific coefficients.
  - Absence of comparable cross-border banking claims data for CIS countries (BIS data do not cover CIS comprehensively) prevents assessment of banking integration as a propagation channel.

### 5.3 Remittance Channel
- Panel regression replacing Russian GDP growth with quarterly bilateral remittance flows indicates remittances are an important transmission channel.
- Estimated impacts:
  - A 10-percentage point increase in growth of remittances from Russia is associated with a 0.2 – 0.6 percentage point increase in regional GDP growth (Table 4).
  - For oil importers, the remittance coefficient ranges from 0.3 to 0.8 percentage points.
- Controlling for EU real GDP growth dampens the remittance coefficient (EU growth affects regional growth and remittances via Russia).
- Controlling for global oil prices also dampens the remittance coefficient (many migrant workers from the region work in Russia’s non-tradable sector, sensitive to oil price fluctuations).
- Data limitation: quarterly real GDP data are not available for some large remittance receivers (Kyrgyz Republic and Uzbekistan), so the remittance channel may be underestimated.
- Sample note: remittance regressions sample consists of Armenia, Azerbaijan, Belarus, Georgia, Kazakhstan, Moldova, Tajikistan, and Ukraine; Georgia data reflect money transfers as remittance data are not published.

### 5.4 Combined
- Combined regressions include trade, remittance, and financial variables with controls, split into 2010-2016 and 2016-2021.
- Main results:
  - 2010-2016: both trade and remittances are positively associated with regional output fluctuations.
  - 2016-2021: remittance coefficient becomes insignificant, while coefficients on trade and sovereign risk premia are large and significant (pattern holds for overall sample and oil importers).
- Interpretation:
  - The remittance channel’s narrowing likely driven by large decline in remittances from Russia after 2014.
  - The trade channel’s emergence likely driven by a shift toward commodity goods in bilateral trade, which are more volatile.
- Robustness: results replicated using standardized growth rates yield similar conclusions (decline in remittance channel; increase in trade and market confidence channels).
- Caveat: exclusion of Tajikistan, Kyrgyz Republic, and Uzbekistan (large remittance recipients) may bias results toward underestimating remittance importance.

### Role of Diversification, Integration, and Institutions (summary of empirical cross-sectional findings)
- Countries with stronger bilateral trade ties to Russia experience larger spillover effects.
- Higher partner diversification (Global Connectedness - Breadth Index) is negatively associated with spillovers—more dispersed ties reduce vulnerability.
- Greater economic diversification (higher Economic Complexity Index) is associated with smaller spillovers; removing large oil exporters (Azerbaijan and Kazakhstan) strengthens this relationship.
- Higher share of commodities trade is associated with larger spillovers (commodities more susceptible to demand and price shocks).
- Macroeconomic buffers reduce spillovers:
  - Larger current account deficits are associated with larger spillovers.
  - Higher shares of sovereign debt denominated in foreign currency are associated with larger spillovers (amplifies market confidence channels).
- Stronger public governance reduces spillovers:
  - Higher government effectiveness and rule of law correlate with milder spillovers.
  - Sovereign spreads of countries with weaker public governance are more sensitive to Russian spreads, amplifying contagion.

*Source: IMF working paper chapter “5. Transmission Channels”*

### 2.5 to 2.5, where higher = better)

### wpiea2023185-print-pdf - 2.5 to 2.5, where higher = better)

### Key findings on spillovers from Russia to neighboring countries
- Spillovers from Russia’s output fluctuations to the CIS region are substantial, especially for oil importers.
- Trade and market confidence factors have become more relevant transmission channels since the first half of the past decade; remittances were more important earlier.
- Higher trade concentration with Russia, especially in commodities, contributes to larger cross-border spillovers.
- Strong macroeconomic buffers and higher quality of public institutions are associated with smaller spillovers.
- Strong currency and risk-premia co-movements with Russia amplify spillovers for countries with relatively high liability dollarization and large current account deficits.

### Cross-sectional analysis and determinants of passthrough
- Country-specific spillover coefficients (훽2푖 from equation 2) capture the impact of a one-pp increase in Russia’s real GDP growth on domestic growth, controlling for non-Russia factors. These are regressed on a matrix of country-specific factors 훼푖 using:
  - 훽2푖 = 휃1(훼푖) + 휖푖
- When country factors are combined, partner diversification and trade intensity with Russia emerge as the strongest determinants of business cycle passthrough.
- Governance variables (e.g., Government Effectiveness, Rule of Law) have the expected negative sign (stronger governance → smaller spillovers), but coefficients become insignificant after controlling for bilateral trade size and structure, implying trade patterns play a stronger role.

### Table 6 (selected cross-sectional regression results)
- Note: A negative sign indicates that an increase in the variable contributes to a smaller magnitude of spillovers. Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1.
- Columns and selected coefficient estimates:
  - (1) Partner Diversification: -0.757*** (0.134); Observations = 32; R-squared = 0.509
  - (2) Partner Diversification (alt): -0.600** (0.272); Observations = 32; R-squared = 0.526
  - (3) Trade with Russia (% of total trade): 0.0309*** (0.00585); Observations = 32; R-squared = 0.628
  - (4) Commodities trade with Russia (% of GDP): 0.0530*** (0.00925); Observations = 30; R-squared = 0.072
  - (5) Economic Diversification: -0.318 (0.227); Observations = 31; R-squared = 0.247
  - (6) Government Effectiveness: -0.585*** (0.195); Observations = 31; R-squared = 0.246
  - (7) Rule of Law: -0.500** (0.183); Observations = 32; R-squared = 0.170
  - (8) Current Account Balance (% of GDP): -0.0240*** (0.00799); Observations = 32; R-squared = 0.397
  - (9–11) GDP Per Capita (PPP, current international $): coefficients: -0.316*** (0.0763), -0.154 (0.170), -0.214 (0.160); Observations vary 31–31–31; R-squared = 0.525, 0.582, 0.685 respectively
- Table-wide notes: Observations range from 30 to 32 across columns; R-squared values reported per column as above.

### Event study: partner diversification dampens passthrough
- Two key findings from the event analysis:
  - Large contractions in Russia’s output are associated with output and trade contractions in the region.
  - More diversified economies experience weaker output and bilateral trade fluctuations during large Russian output swings.
- Identification: Upswings and Downswings correspond to large Russian output fluctuations defined as periods where Russian real q-o-q GDP growth is one-SD above or below its 2001–19 average. Countries are grouped by partner diversification into top (High), medium (Medium), and bottom (Low) thirds; average responses computed for Upswings and Downswings.
- Graphical results (described):
  - GDP Growth Response to Large Russian Cyclical Output Fluctuations (Percent): High, Medium, Low Diversification shown separately for Upswing and Downswing (range illustrated between -2.5 and 2.5 percent).
  - Bilateral Export Growth Response to Large Russian Cyclical Output Fluctuations (Percent): High, Medium, Low Diversification shown separately for Upswing and Downswing (range illustrated between -2.5 and 2.5 percent).
- Interpretation: Partner diversification mitigates spillovers primarily via constraining bilateral trade fluctuations over the Russian business cycle.

### Role of trade structure, institutions, and fundamentals
- Trade channel:
  - Bilateral trade concentration (exports and commodities trade with Russia) increases vulnerability and magnifies passthrough.
  - A pickup in trade with Russia could heighten the region’s vulnerability to future shocks.
- Institutions:
  - Improving public institutional quality—especially in fiscal areas—can contain spillovers via market confidence channels by reducing sovereign credit spreads and their synchronization with Russia’s.
- Macroeconomic fundamentals:
  - Stronger buffers (smaller shares of FX sovereign liabilities; smaller current account deficits) reduce vulnerability to international spillovers.
  - De-dollarization (reducing liability dollarization) is highlighted as beneficial given currency and risk-premia co-movements amplify spillovers.

### Policy implications and recommendations
- Reduce trade concentration with Russia and diversify product markets, including through stronger regional integration, to lower trade-induced spillovers.
- Improve public governance (particularly fiscal institutions) to mitigate adverse confidence effects and reduce sovereign spread co-movement with Russia.
- Strengthen macroeconomic buffers to reduce vulnerability:
  - Lower foreign-currency share of sovereign debt.
  - Reduce current account deficits.
  - Pursue de-dollarization to lessen amplification via liability dollarization.
- Pursue partner diversification (global connectedness and broader export destinations) to dampen passthrough via trade channels and to limit synchronized market responses.

### Conclusion (as in source)
- Russia’s economic footprint in the CIS has declined over three decades but remains large.
- Trade is a dominant channel of business cycle transmission from Russia to neighboring countries; sovereign spreads and market confidence can amplify shocks.
- Policy mixes that enhance partner diversification, public governance, and macroeconomic fundamentals can contain spillovers and create positive feedback loops improving resilience to Russia’s business cycle.

*Source: IMF staff estimates (from "Spillovers from Russia to Neighboring Countries: Transmission Channels and Policy Options").*

### Annex III. Alternative Panel Specifications

### Annex III. Alternative Panel Specifications

### Overall Panel Specification with Fixed Effects / with Standardized Variables
- Russia Real GDP Growth:
  - Columns (1)–(4): 0.790***, 0.691***, 0.913***, 0.787***
- EU Real GDP Growth:
  - Columns (1)–(4): 0.439**, -0.0664, 0.550**, 0.100
- China Real GDP Growth (t-1):
  - Columns (1)–(4): 0.513***, 0.201, 0.574**, 0.0672
- Headline CPI Inflation:
  - Columns (1)–(8): -0.0483, 0.00159, -0.0170, -0.0477*, -0.0603**, -0.0119, -0.0350, -0.0530**
- REER Growth:
  - Columns (1)–(8): 0.0311, 0.0598*, 0.0673*, 0.0372, 0.00966, 0.0483, 0.0661, 0.0155
- Oil Price Growth:
  - Columns (1)–(8): 0.00898, 0.0423***, 0.0350***, 0.00725, 0.00700, 0.0430**, 0.0388**, 0.00604
- Δ US 10-year Treasury Yield:
  - Columns (1)–(8): 0.142, -0.318, -0.0518, 0.208, -0.0185, -0.543, -0.278, -0.0252
- Constant (selected columns):
  - 0.727***, 0.986***, 0.123, 0.406*, 0.582***, 0.883***, -0.0634, 0.483**
- Country Fixed Effects: Yes (all specifications)
- Observations:
  - Columns (1)–(4): 602
  - Columns (5)–(8): 450
- R-squared:
  - Columns (1)–(4): 0.296, 0.213, 0.233, 0.301
  - Columns (5)–(8): 0.352, 0.264, 0.266, 0.357
- Number of ifs_code:
  - Columns (1)–(4): 8
  - Columns (5)–(8): 6
- Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1

### CIS countries (excluding Russia) and CIS oil importers (subsample results)
- Own Growth (t-1):
  - Selected coefficients across columns: 0.00167, 0.105**, 0.0877, 0.0151, -0.0252, 0.0842, 0.0594, -0.00277
- Russia Real GDP Growth (CIS subsample):
  - Significant positive coefficients: 0.533***, 0.461***, 0.587***, 0.507***
- EU Real GDP Growth (CIS subsample):
  - Coefficients: 0.340***, -0.0500, 0.398***, 0.0622
- China Real GDP Growth (t-1) (CIS subsample):
  - Coefficients: 0.378***, 0.150*, 0.397***, 0.0469
- Headline CPI Inflation (CIS subsample):
  - Coefficients across columns: -0.0705***, -0.0163, -0.0321, -0.0684***, -0.0697***, -0.0115, -0.0333, -0.0621***
- REER Growth (CIS subsample):
  - Coefficients: 0.0261, 0.0489, 0.0640*, 0.0328, 0.00747, 0.0330, 0.0624*, 0.00961
- Oil Price Growth (CIS subsample):
  - Coefficients: 0.0668*, 0.255***, 0.218***, 0.0568, 0.0535, 0.248***, 0.234***, 0.0481
- Δ US 10-year Treasury Yield (CIS subsample):
  - Coefficients: 0.0200, -0.0354, -0.00369, 0.0278, -0.00212, -0.0570, -0.0287, -0.00180
- Constant (CIS subsample): -0.00178, -0.00574, -0.00680, -0.00305, -0.00363, -0.00794, -0.00892, -0.00472
- Observations:
  - Full CIS columns: 594 (columns 1–4)
  - CIS oil importers columns: 444 (columns 5–8)
- Number of ifs_code reported: combinations including 8, 8, 8, 8, 6, 6, 6, 6
- Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1

### Combined Panel Specification with Standardized Variables (Table A.5)
- Sample splits: CIS countries (excluding Russia) and CIS oil importers; time periods 2010-2016 and 2016-2021 are used across columns.
- Own Growth (t-1):
  - Coefficients and standard errors: 0.0258 (0.110), 0.0292 (0.109), -0.116 (0.0714), -0.0309 (0.0629), 0.110 (0.0702), 0.113* (0.0633), -0.208*** (0.0464), -0.0812 (0.0707)
- Export to Russia Growth:
  - Coefficients and standard errors: 0.0625 (0.0401), 0.0645 (0.0399), 0.345*** (0.123), 0.274** (0.125), 0.0960** (0.0450), 0.0967** (0.0453), 0.175* (0.102), 0.0896 (0.113)
- Bilateral Remittances Growth:
  - Coefficients and standard errors: 0.150* (0.0782), 0.146* (0.0767), 0.0917 (0.0877), -0.00107 (0.0738), 0.115 (0.0849), 0.113 (0.0827), 0.119 (0.102), 0.0119 (0.0651)
- Δ log Russia EMBIG Spread (t-1):
  - Coefficients and standard errors: -0.0596 (0.0573), -0.0596 (0.0574), -0.247*** (0.0815), -0.250*** (0.0649), -0.0728 (0.0634), -0.0729 (0.0634), -0.267** (0.109), -0.251*** (0.0854)
- EU Real GDP Growth:
  - Coefficients and standard errors: -0.162 (0.285), 0.322*** (0.106), -0.0651 (0.332), 0.383*** (0.0674)
- Headline CPI Inflation:
  - Coefficients and standard errors: -0.0992* (0.0580), -0.0975* (0.0581), -0.0662 (0.113), -0.0615 (0.128), -0.0786*** (0.0203), -0.0780*** (0.0197), 0.119 (0.102), 0.178* (0.0980)
- REER Growth:
  - Coefficients and standard errors: 0.105** (0.0531), 0.105** (0.0526), 0.0796* (0.0454), 0.0975 (0.0667), 0.108 (0.0828), 0.108 (0.0829), -0.0184 (0.0447), -0.0639 (0.0556)
- Oil Price Growth:
  - Coefficients and standard errors: 0.126* (0.0648), 0.123* (0.0660), 0.493*** (0.0629), 0.256*** (0.0494), 0.127* (0.0711), 0.125* (0.0731), 0.554*** (0.0622), 0.268*** (0.0456)
- Δ US 10-year Treasury Yield:
  - Coefficients and standard errors: -0.108 (0.103), -0.104 (0.108), -0.0889 (0.0572), -0.00708 (0.0537), -0.118 (0.137), -0.116 (0.145), -0.149** (0.0643), -0.0511 (0.0578)
- Constant (with standard errors):
  - 0.00261 (0.0315), -1.44e-05 (0.0313), -0.138*** (0.0453), -0.137*** (0.0418), -0.0341 (0.0247), -0.0351 (0.0272), -0.0859* (0.0444), -0.0722* (0.0403)
- Observations by column:
  - 189, 189, 167, 167, 135, 135, 119, 119
- Number of ifs_code by column:
  - 7, 7, 7, 7, 5, 5, 5, 5
- Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1
- Subsamples labeled: CIS countries (excluding Russia) and CIS oil importers; periods indicated as 2010-2016 and 2016-2021 across columns.

*IMF Working Paper — Annex III. Alternative Panel Specifications, Spillovers from Russia to Neighboring Countries: Transmission Channels and Policy Options (Working Paper No. WP/2023/185).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2023/english/wpiea2023185-print-pdf.pdf_