## _wp1417

## Source details

**Canonical URL:** [_wp1417](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp1417.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp1417.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp1417.pdf.json)

---

### Executive summary and purpose
- Objective: determine drivers of sovereign CDS spreads in the CESEE region during the 2007–12 period.
- Specific questions:
  - Role of global risk aversion, macroeconomic fundamentals, CDS market liquidity, and spillovers from other countries in explaining divergent CDS movements.
  - Whether the role of these factors changed between the two main stress episodes: the aftermath of Lehman Brothers (2008/09) and the euro area crisis (May 2010–June 2012).
- Data and methods:
  - Generalized variance decomposition on daily data to detect cross-country influences in the CDS market.
  - Dynamic panel regression framework on monthly data to measure impact of macroeconomic fundamentals on CDS spreads.
- Sample period used in regressions and analysis: January 2007–December 2012.
- Out-of-sample evaluation: forecast of CDS spread developments in the second half of 2012.

### Key empirical findings
- Dominant drivers:
  - Spreads in the CESEE region are primarily driven by changes in global investor sentiment, proxied by the VIX index.
  - Country specific macroeconomic fundamentals and CDS market liquidity conditions also play important roles.
- Fundamentals of particular importance:
  - Growth prospects.
  - Forward-looking fiscal indicators (e.g., one year ahead fiscal deficit forecasts).
- Heterogeneity:
  - The role of fundamentals is particularly strong for high debt and low growth countries.
- Liquidity:
  - Impact of liquidity conditions was very prominent during the global liquidity shock in the 2008/09 crisis and much smaller thereafter.
- Spillovers:
  - Little empirical support for strong direct spillovers in CDS spreads from euro area periphery to CESEE; direct spillovers are quantitatively very small.
  - Spillovers from the euro area crisis to CESEE largely operated through the crisis’s impact on global risk aversion.
- Dynamics and modeling:
  - CDS market dynamics can be captured by an error-correction process where CDS spreads evolve around a level linked to fundamentals and market conditions.
  - Short-term dynamics are driven by frequent adjustments in market perception of fundamentals, with occasional over- and undershooting.

### Comparative assessment across crisis episodes
- Two distinct stress periods observed in CESEE CDS spreads:
  - 2008/09 global financial and economic crisis.
  - Euro area crisis period (May 2010–June 2012).
- Between Spring 2009 and Summer 2011, CDS spreads moderated considerably despite the start of the euro area crisis around May 2010.
- Reasons for relative resilience during the euro area crisis:
  - Reduction in fiscal deficits.
  - Sharp narrowing of current account balances.
  - Gradual acceleration in growth.
  - These improved fundamentals partially offset negative spillovers transmitted via global risk aversion.

### Out-of-sample forecast: second half of 2012
- Forecast summary:
  - The sharp drop of CDS spreads across the board in the second half of 2012 following the ECB’s OMT announcement was to a large extent due to a drop in risk aversion.
  - Country specific fundamentals remained on average broadly unchanged or (for growth prospects) deteriorated somewhat in H2 2012.
- Forecast performance:
  - Model one-step ahead forecasts for July–Dec. 2012 generally tracked actual CDS well.
  - Exceptions: Portugal and Ireland in July 2012 — model under-predicted the large drop by around "200 basis points" for Portugal and Ireland.
  - By late 2012, forecast errors were "below 100 basis points for all countries."

### Policy-relevant implications
- Sovereign CDS spreads reflect both global factors and economic fundamentals:
  - Policies that strengthen fundamentals—improving growth prospects and forward-looking fiscal positions—can materially reduce sovereign risk premia.
- Importance of liquidity management:
  - In severe global liquidity shocks, liquidity conditions in the CDS market can amplify spread movements; addressing market liquidity frictions matters in crisis times.
- Spillover interpretation:
  - Claims that countries are “innocent bystanders” suffering strong direct spillovers from euro area periphery CDS dynamics receive little empirical support; much of the regional transmission operates through common global risk aversion.
- For high-debt countries (debt > 80 percent of GDP), the market penalizes insufficient growth and fiscal outlook more strongly.

### Spillovers — methodology and principal findings
- Methodology:
  - VAR/VEC framework with generalized impulse response/variance decomposition (GIR/GVD) as in Shin and Pesaran (1998).
  - One lag VAR in OLS; 10-day ahead forecast errors computed over a 160 day rolling window.
  - Normalized generalized variance decomposition used to obtain directional spillover shares.
- Principal findings:
  - "The direct spillover impact from CDS spreads in the euro area periphery countries (as a group) has not been the most dominant effect influencing CDS spreads in the CESEE region since 2007."
  - Average share of shocks from the euro area periphery to CESEE countries are generally much lower than that from advanced Europe (including Germany, Netherlands or Austria).
  - Intra-regional CESEE shocks dominated CESEE spillovers in both October 2008–March 2009 and May 2010–June 2012.
  - During 2008–09, Hungary and the Baltic countries and Romania were main shock sources within CESEE.
  - During May 2010–June 2012, Hungary, and to a smaller extent Romania and Bulgaria, were most pronounced shock generators within CESEE.
  - From euro area periphery, Italy and Spain had the largest impact on CESEE but rarely in the top quartile (exception: Croatia vis-à-vis Italy).
  - Ukraine, despite highest CDS levels, had the lowest share of influence from euro area periphery spillovers among CESEE countries.

### Role of global risk aversion and liquidity (selected quantitative findings)
- Global risk aversion (VIX or VSTOXX) effects:
  - For the euro area crisis period, "a 1 percent increase in the VIX would raise CDS spreads by about 5.4 bps compared to 4.6 bps before the euro crisis." (VIX coefficient (euro crisis): 5.444***)
  - VSTOXX and VIX coefficients reported in multiple specifications (examples): VSTOXX 5.556*** (0.209) in full sample; VIX 0.0384*** (0.00166) in log specifications (full sample).
- Liquidity (CDS bid-ask spread) effects:
  - "An increase in the bid-ask spread by one bps would increase the CDS spread by 6 bps for the euro area crisis period (5.6 bps before the euro area crisis)."
  - Correlation between VIX and bid-ask spread in sample: "around 0.2."
  - Bid-ask spread coefficients (examples): Bid-ask spread (adjusted*) 5.989*** (euro crisis table); 7.275*** (full sample, VSTOXX table).

### Role of macroeconomic fundamentals — norm estimation (selected coefficients and interpretation)
- Explanatory groups used: VIX/VSTOXX; CDS bid-ask spread; macro fundamentals via Consensus Economics forecasts: GDP growth (current and next year), fiscal deficit (current and next year), current account balance (current year), previous year public debt/GDP.
- Selected coefficients (examples preserved exactly from tables):
  - VIX coefficient (euro crisis): 5.444***
  - GDP growth (next year, adjusted*): -32.49***
  - Bid-ask spread (adjusted*): 5.989***
  - Deficit*Debt80 (adjusted*): 13.06***
  - Deficit next year (adjusted*): -15.57***
  - Interaction of 2 year growth with debt threshold (adjusted*): -19.62***
  - Ultra-low growth dummy: 66.28***
  - Euro area dummy: -24.56 (not significant in crisis sub-sample)
  - Debt above 80: 105.2***
  - Observations (euro crisis): 703; Number of country = 23; R-Squared = 0.918
- Interpretive highlights (numerical statements preserved):
  - "For the euro area crisis period, one percentage point higher GDP growth forecast for the next year reduces CDS spreads by 32 bps (compared to 42 bps before the euro crisis)."
  - "For countries with debt exceeding 80 percent of GDP, the norm CDS spreads ... higher by 105 bps during the euro crisis (181 bps before the crisis)."
  - "Countries whose average two year growth are below 0.05 percent (advanced euro area) or one percent (CESEE) would see on average 66 bps increase in CDS spreads" during the euro crisis (versus 14 bps before).
  - "A one percentage point increase in the fiscal deficit forecast for the next year raises CDS spreads by 16 bps during the euro crisis period (compared to 2.6 bps before the euro crisis)."
  - Current account balance: statistically insignificant for both periods in crisis in some specifications.

### Short-run dynamics (error-correction estimates, representative values)
- Euro crisis period (difference specifications, representative coefficients):
  - Δcds t-1: 0.380*** (examples across tables near 0.430*** or 0.277** depending on specification)
  - Δcds t-2: around 0.06*
  - Error correction term ec(t-1): ≈ -0.380*** (or -0.196*** in logs, depending on specification)
  - ec(t-2): ≈ 0.231*** (or 0.0954*** in logs)
  - Interpretation: "During the euro area crisis period, over 35 percent of such changes are carried over in the current month, and the lagged impact of CDS change appears not lasting beyond two month."
- Short-run effects (selected point estimates preserved):
  - During the euro crisis period (Table 8a):
    - GDP growth (next year, forecast) (difference): -7.422* (i.e., a 1 percentage point increase reduces ΔCDS by 7.422 bps)
    - VIX (difference): 1.400***
    - CDS bid-ask spread (difference): 3.338***
    - Interaction avg. growth & Debt threshold (difference): -9.446*
  - Before the euro crisis (Table 8b):
    - GDP growth (next year, forecast) (difference): -16.00***
    - VIX (difference): 2.541***
    - CDS bid-ask spread (difference): 2.155***
- Magnitudes summary preserved:
  - Short-run, during the euro crisis, changes in next-year growth forecasts reduce CDS by about 6–7 bps per 1 percentage point versus about 16 bps before the euro crisis.
  - VIX and bid-ask spread changes affect CDS by around 1.5 and 3 bps respectively during the euro crisis (about 2.5 bps each before).

### Country-level decompositions — illustrative cases (numbers preserved)
- Latvia:
  - March 2009: CDS spread norm = 1043bps; actual CDS = 1044bps.
    - Contributions: Latvia dummy = 230; GDP growth forecasts* = 280; Fiscal forecasts** = 101; VIX = 211; bid-ask spread = 311; current account = 7.
  - May 2010: CDS norm = 429bps; actual CDS = 357bps.
  - June 2012: CDS norm = 182bps; actual CDS = 296bps.
- Slovenia:
  - March 2009: CDS norm = 232bps; actual CDS = 155bps.
    - Contributions: Euro area & Slovenia dummy = -585; VIX = 211; bid-ask spread = 11.
  - May 2010: CDS norm = 325bps; actual CDS = 78bps.
  - June 2012: CDS norm ≈ 418bps; actual CDS ≈ 408bps.
    - Fiscal forecasts explained "128 basis points" of CDS level deterioration.

### Cross-country variability snapshots (selected exact numbers)
- March 2009: Max. CDS = 4008, Min. CDS = 16, Max.-Min. = 384, Standard Dev. = 963
- May 2010: Max. CDS = 860, Min. CDS = 4, Max.-Min. = 356, Standard Dev. = 138
- July 2011: Max. CDS = 747, Min. CDS = 69, Max.-Min. = 678, Standard Dev. = 102
- June 2012: Max. CDS = 861, Min. CDS = 17, Max.-Min. = 844, Standard Dev. = 195
- Country examples (June 2012 CDS levels):
  - Ukraine: 861bps
  - Hungary: 557bps
  - Latvia: 300bps
  - Estonia: 111bps (range indicated elsewhere as 111bps–251bps for Baltics)
- By June 2012, most CESEE countries had lower CDS spreads than any of the euro area periphery countries.
  - Sovereign CDS spreads of the Czech Republic and Estonia were "even lower than that of France and Austria."
  - CESEE countries with lower CDS spreads than Belgium: "the Czech Republic, Estonia, Turkey, Slovakia, Russia and Poland."

### Market reaction to ECB OMT and late-2012 developments (selected exact movements)
- Drop in VSTOXX was around 0.8 standard deviation.
- Growth forecasts for 2013 on average deteriorated by almost 0.4 standard deviations.
- Many other indicators moved about 0.1 standard deviation in the second half of 2012.
- Some countries with larger macroeconomic imbalances experienced undershooting (possible under-pricing of risk) after the OMT announcement.

### Annex highlights (selected methodological and estimation notes)
- ANNEX III: Evolution of bid-ask spreads of 5 year sovereign CDS spreads (in basis points) for sample dates from "1/1/2007" through "9/1/2012" (countries shown include Germany, Czech Republic, Bulgaria, Croatia, Ukraine, Portugal).
- ANNEX IV: Country abbreviations used are reported in source (concatenated list present in source).
- ANNEX V and VI: Extensive estimation tables presented for multiple specifications:
  - Estimation variants include CDS levels and CDS in logs; with VIX, VSTOXX; end-of-month CDS and monthly full-sample; euro crisis and pre-euro crisis sub-samples.
  - Representative statistical outputs preserved in source tables (examples):
    - Observations: full sample 1,492; euro crisis 703; pre-euro crisis 789 (VSTOXX table).
    - R-Squared values: e.g., 0.878 (full sample), 0.926 (euro crisis) in some specifications.
    - Structural break tests: Test statistic for no structural break, χ2 (33), Prob. of no structural break = 0 in multiple reported tests.
  - Arellano-Bond tests and robustness diagnostics are reported in source tables (selected p-values preserved in annex tables).

*Source: Executive Summary and selected sections and annexes of _wp1417 (analysis covering January 2007–December 2012, with emphasis on crisis episodes in 2008/09 and May 2010–June 2012).*

### Executive Summary ......................................................................................................

### _wp1417 - Executive Summary

### I. Purpose and scope
- Objective: determine drivers of sovereign CDS spreads in the CESEE region during the 2007–12 period.
- Specific questions:
  - Role of global risk aversion, macroeconomic fundamentals, CDS market liquidity, and spillovers from other countries in explaining divergent CDS movements.
  - Whether the role of these factors changed between the two main stress episodes: the aftermath of Lehman Brothers (2008/09) and the euro area crisis (May 2010–June 2012).
- Data and methods:
  - Generalized variance decomposition on daily data to detect cross-country influences in the CDS market.
  - Dynamic panel regression framework on monthly data to measure impact of macroeconomic fundamentals on CDS spreads.
- Sample period used in regressions and analysis: January 2007–December 2012.
- Out-of-sample evaluation: forecast of CDS spread developments in the second half of 2012 to gauge model performance.

### II. Key empirical findings
- Dominant drivers:
  - Spreads in the CESEE region are primarily driven by changes in global investor sentiment (global risk aversion), proxied by the VIX index.
  - Country specific macroeconomic fundamentals and CDS market liquidity conditions also play important roles.
- Fundamentals of particular importance:
  - Growth prospects.
  - Forward-looking fiscal indicators (e.g., one year ahead fiscal deficit forecasts).
- Heterogeneity:
  - The role of fundamentals is particularly strong for high debt and low growth countries.
- Liquidity:
  - Impact of liquidity conditions was very prominent during the global liquidity shock in the 2008/09 crisis.
  - Liquidity effects were much smaller after the 2008/09 episode.
- Spillovers:
  - Little empirical support for strong direct spillovers in CDS spreads from euro area periphery to CESEE.
  - Quantitative effect of direct spillovers from the euro area periphery on CESEE CDS spreads is very small.
  - Spillovers of the euro area crisis to the region largely occurred through the impact of the crisis on global risk aversion.
- Dynamics and modeling:
  - CDS market dynamics can be captured relatively well by an error-correction process where CDS spreads evolve around a level linked to economic fundamentals and specific market conditions (such as CDS market liquidity).
  - Short-term dynamics are driven by frequent adjustments in market perception of fundamentals, with occasional over- and undershooting.

### III. Comparative assessment across crisis episodes
- Two distinct stress periods observed in CESEE CDS spreads:
  - 2008/09 global financial and economic crisis.
  - Euro area crisis period (May 2010–June 2012).
- Between Spring 2009 and Summer 2011, CDS spreads moderated considerably despite the start of the euro area crisis around May 2010.
- Reasons for relative resilience of CESEE CDS spreads during the euro area crisis:
  - Improvement in CESEE country-specific fundamentals, including:
    - Reduction in fiscal deficits.
    - Sharp narrowing of current account balances.
    - Gradual acceleration in growth.
  - These improved fundamentals partially offset negative spillovers transmitted via global risk aversion.

### IV. Out-of-sample forecast: second half of 2012
- An out-of-sample forecast based on the panel results suggests:
  - The sharp drop of CDS spreads across the board in the second half of 2012 following the ECB’s OMT announcement was to a large extent due to a drop in risk aversion.
  - Country specific fundamentals remained on average broadly unchanged or (in the case of growth prospects) deteriorated somewhat during that period.

### V. Policy-relevant implications
- Sovereign CDS spreads reflect both global factors and economic fundamentals:
  - Policies that strengthen fundamentals—improving growth prospects and forward-looking fiscal positions—can materially reduce sovereign risk premia.
- Importance of liquidity management:
  - In severe global liquidity shocks, liquidity conditions in the CDS market can amplify spread movements; addressing market liquidity frictions matters in crisis times.
- Spillover interpretation:
  - Claims that countries are “innocent bystanders” suffering strong direct spillovers from euro area periphery CDS dynamics receive little empirical support; much of the regional transmission operates through common global risk aversion.

*Source: Executive Summary of _wp1417 (analysis covering January 2007–December 2012, with emphasis on crisis episodes in 2008/09 and May 2010–June 2012).*

### 2010. In second half of 2011, CDS spreads have increased sharply following the

### _wp1417 - 2010. In second half of 2011, CDS spreads have increased sharply following the

### CDS spreads dynamics and regional comparison
- In the second half of 2011, CDS spreads "have increased sharply following the intensification of the euro area crisis", but "overall they remained much below their peak levels during the first crisis."
- In 2012, "CDS spreads were moderating in the region, with a sharp decline in the second half of the year."
- By June 2012, "most CESEE countries had lower CDS spreads than any of the euro area periphery countries."
- Specific comparisons in June 2012:
  - Sovereign CDS spreads of the Czech Republic and Estonia were "even lower than that of France and Austria."
  - CESEE countries with lower CDS spreads than Belgium: "the Czech Republic, Estonia, Turkey, Slovakia, Russia and Poland."
- Cross-country variability:
  - March 2009: Max. CDS = 4008, Min. CDS = 16, Max.-Min. = 384, Standard Dev. = 963
  - May 2010: Max. CDS = 860, Min. CDS = 4, Max.-Min. = 356, Standard Dev. = 138
  - July 2011: Max. CDS = 747, Min. CDS = 69, Max.-Min. = 678, Standard Dev. = 102
  - June 2012: Max. CDS = 861, Min. CDS = 17, Max.-Min. = 844, Standard Dev. = 195
- Country examples (June 2012 CDS levels):
  - Ukraine: 861bps
  - Hungary: 557bps
  - Latvia: 300bps
  - Estonia: 111bps (range indicated elsewhere as 111bps–251bps for Baltics)

### Spillovers: methodology and principal findings
- Methodology:
  - VAR/VEC framework with generalized impulse response/variance decomposition (GIR/GVD) as in Shin and Pesaran (1998).
  - One lag VAR in OLS; 10-day ahead forecast errors computed over a 160 day rolling window.
  - Normalized generalized variance decomposition used to obtain directional spillover shares.
- Key findings:
  - "The direct spillover impact from CDS spreads in the euro area periphery countries (as a group) has not been the most dominant effect influencing CDS spreads in the CESEE region since 2007."
  - For CESEE countries, "the average share of shocks from the euro area periphery countries are generally much lower than that for advanced Europe (including Germany, Netherlands or Austria)."
  - Ukraine, despite highest CDS levels, had "the lowest" share of influence from euro area periphery spillovers among CESEE countries.
  - Bilateral results: "none of the euro area periphery countries were among the most important sources of direct shock for any of the CESEE countries."
  - Intra-regional CESEE shocks dominated CESEE spillovers in both October 2008–March 2009 and May 2010–June 2012.
  - During 2008–09, Hungary and the Baltic countries and Romania were main shock sources within CESEE.
  - During May 2010–June 2012, Hungary, and to a smaller extent Romania and Bulgaria, were most pronounced shock generators within CESEE. From euro area periphery, Italy and Spain had the largest impact on CESEE but rarely in the top quartile (exception: Croatia vis-à-vis Italy).

### Role of global risk aversion and liquidity
- Global risk aversion measure: VIX (Chicago Board Options Exchange Market Volatility Index).
  - For the euro area crisis period, "a 1 percent increase in the VIX would raise CDS spreads by about 5.4 bps compared to 4.6 bps before the euro crisis."
- Liquidity proxied by CDS bid-ask spread:
  - "An increase in the bid-ask spread by one bps would increase the CDS spread by 6 bps for the euro area crisis period (5.6 bps before the euro area crisis)."
  - Correlation between VIX and bid-ask spread in sample: "around 0.2."
- Interpretation:
  - Liquidity had particularly large impact during the 2008/09 crisis; thereafter liquidity shocks were less synchronized and country differentiation increased.

### Role of macroeconomic fundamentals (panel norm estimation)
- Explanatory groups: (1) VIX; (2) CDS bid-ask spread; (3) macro fundamentals via monthly Consensus Economics forecasts: GDP growth (current and next year), fiscal deficit (current and next year), current account balance (current year), previous year public debt/GDP.
- Key quantitative findings from the norm (equation (6.1), Table 6):
  - VIX coefficient (euro crisis): 5.444***
  - GDP growth (next year, adjusted*): -32.49***
  - Bid-ask spread (adjusted*): 5.989***
  - Deficit*Debt80 (adjusted*): 13.06***
  - Deficit next year (adjusted*): -15.57*** (note: negative sign reported in table; interpretation discussed in text)
  - Interaction of 2 year growth with debt threshold (adjusted*): -19.62***
  - Ultra-low growth dummy: 66.28***
  - Euro area dummy: -24.56 (not significant in crisis sub-sample)
  - Debt above 80: 105.2***
  - Observations (euro crisis): 703; Number of country = 23; R-Squared = 0.918
- Interpretive highlights:
  - Growth prospects matter strongly: "For the euro area crisis period, one percentage point higher GDP growth forecast for the next year reduces CDS spreads by 32 bps (compared to 42 bps before the euro crisis)."
  - High-debt effect: "For countries with debt exceeding 80 percent of GDP, the norm CDS spreads ... higher by 105 bps during the euro crisis (181 bps before the crisis)."
  - Ultra-low growth effect: "countries whose average two year growth are below 0.05 percent (advanced euro area) or one percent (CESEE) would see on average 66 bps increase in CDS spreads" during the euro crisis (versus 14 bps before).
  - Fiscal outlook: "A one percentage point increase in the fiscal deficit forecast for the next year raises CDS spreads by 16 bps during the euro crisis period (compared to 2.6 bps before the euro crisis)." (textual statement)
  - Current account balance: statistically insignificant for both periods in crisis.

### Short-run dynamics (error-correction dynamics, equation (6.2))
- Lagged dynamics and error correction:
  - Euro crisis period: Δcds(t-1) coefficient ≈ 0.380***; Δcds(t-2) ≈ 0.06*
  - Error correction term ec(t-1) ≈ -0.380***; ec(t-2) ≈ 0.231***
  - Interpretation: "During the euro area crisis period, over 35 percent of such changes are carried over in the current month, and the lagged impact of CDS change appears not lasting beyond two month."
- Short-run effects of changes:
  - During the euro crisis period (Table 8a, specification):
    - GDP growth (next year, forecast) (difference): -7.422* (i.e., a 1 percentage point increase reduces ΔCDS by 7.422 bps)
    - VIX (difference): 1.400***
    - CDS bid-ask spread (difference): 3.338***
    - Interaction avg. growth & Debt threshold (difference): -9.446*
  - Before the euro crisis (Table 8b, specification):
    - GDP growth (next year, forecast) (difference): -16.00***
    - VIX (difference): 2.541***
    - CDS bid-ask spread (difference): 2.155***
- Magnitudes summary (text):
  - During the euro crisis, changes in next-year growth forecasts reduce CDS by about 6–7 bps per 1 percentage point (short-run), versus about 16 bps before the euro crisis.
  - VIX and bid-ask spread changes affect CDS by around 1.5 and 3 bps respectively during the euro crisis (about 2.5 bps each before).

### Country-level decompositions and illustrative cases
- Latvia (March 2009, May 2010, June 2012):
  - March 2009: CDS spread norm = 1043bps; actual CDS = 1044bps.
    - Contributions: Latvia dummy = 230; GDP growth forecasts* = 280; Fiscal forecasts** = 101; VIX = 211; bid-ask spread = 311; current account = 7.
  - May 2010: CDS norm = 429bps; actual CDS = 357bps.
    - Improvement driven mainly by better growth forecasts and liquidity improvement.
  - June 2012: CDS norm = 182bps; actual CDS = 296bps.
    - Further improvement driven by fundamentals and lower global risk aversion.
- Slovenia (March 2009, May 2010, June 2012):
  - March 2009: CDS norm = 232bps; actual CDS = 155bps.
    - Contributions: Euro area & Slovenia dummy = -585; VIX = 211; bid-ask spread = 11; growth/fiscal contributions small/positive.
  - May 2010: CDS norm = 325bps; actual CDS = 78bps.
    - Fiscal deterioration raised norm; euro area membership discount weakened.
  - June 2012: CDS norm ≈ 418bps; actual CDS ≈ 408bps.
    - Deterioration driven mainly by worse growth prospects and fiscal/banking sector problems; fiscal forecasts explained "128 basis points" of CDS level.

### Out-of-sample forecasts and the post-OMT period (July–Dec. 2012)
- Context: sharp improvement in CDS spreads following ECB President Mario Draghi’s announcements and OMT in summer 2012.
- Forecast performance:
  - Model one-step ahead forecasts for July–Dec. 2012 generally tracked actual CDS well.
  - Exceptions: Portugal and Ireland in July 2012 — model under-predicted the large drop by around "200 basis points" for Portugal and Ireland.
  - By late 2012, forecast errors were "below 100 basis points for all countries."
- Drivers of late-2012 decline:
  - "An analysis ... suggested a primary role for a drop in risk aversion in that period."
  - On average in H2 2012, "country specific macroeconomic fundamentals was negligible (below ...)" (text truncated in source).

### Policy-relevant implications (derived from findings in the text)
- Country-specific fundamentals matter: fiscal consolidation credible plans and improved growth prospects materially lower sovereign CDS premia.
- Liquidity conditions amplify stress: more liquid CDS markets reduce spread penalties, especially in crisis episodes.
- Global risk aversion matters: large swings in VIX translate into measurable CDS spread movements.
- Intra-regional shock transmission within CESEE is often more important than direct spillovers from euro area periphery; policy attention to domestic vulnerabilities (banking sector, fiscal imbalances) is warranted.
- For high-debt countries (debt > 80 percent of GDP), the market penalizes insufficient growth and fiscal outlook more strongly.

*Source: IMF staff analysis and empirical results as presented in the provided chapter.*

### 0.1 standard deviation) for most indicators in the second half of 2012, although growth

### _wp1417 - 0.1 standard deviation) for most indicators in the second half of 2012, although growth

### Market reaction to ECB OMT and late-2012 developments
- The ECB’s OMT announcement and coordinated euro-area policy actions contributed to stabilizing markets.
- Risk-aversion and market-implied volatility fell markedly:
  - Drop in VSTOXX (the European VIX) was around 0.8 standard deviation.
  - Growth forecasts for 2013 on average deteriorated by almost 0.4 standard deviations.
  - Many other indicators moved about 0.1 standard deviation in the second half of 2012.
- The improvement in European-region fundamentals (as a whole) in the second half of 2012 reflected tangible progress in resolving the euro area crisis, even though country-specific macroeconomic fundamentals on balance remained broadly unchanged or deteriorated somewhat.
- Some countries—particularly those with larger macroeconomic imbalances—experienced undershooting (possible under-pricing of risk) after the OMT announcement as investor risk appetite shifted toward a renewed search for yields.

### Empirical findings on CDS spillovers and determinants
- Direct spillover impact of CDS spreads from euro area periphery countries on CESEE CDS dynamics was limited over the past several years; CESEE countries were less affected by periphery spillovers than euro-area core countries.
- During the euro area crisis:
  - Within the periphery, CDS developments in Italy and Spain had the highest impact on other euro-area country CDS spreads.
  - Ireland had the highest impact during the 2008/09 crisis.
- Key determinants of CESEE countries’ CDS spreads:
  - Global investor sentiment (risk aversion).
  - Country-specific economic fundamentals (especially growth prospects and forward-looking fiscal indicators).
  - CDS market liquidity conditions.
- Among fundamentals:
  - Growth prospects and forward-looking fiscal indicators are particularly important.
  - Current account forecasts are significant but play a smaller role than fiscal and growth forecasts.
  - Fundamentals matter especially for countries with very weak growth prospects and countries with high public debt.

### Time variation in the role of fundamentals and market pricing behavior
- The relative importance of fundamentals changes over time and tends to strengthen during crisis periods; markets placed greater weight on future growth and fiscal consolidation effort during the euro-area crisis.
- The pre-crisis period (2000–2007) showed macroeconomic fundamentals explaining little of euro-area sovereign pricing, a pattern that changed during the crisis.
- Market pricing generally follows a gradual error-correction process:
  - Markets may under- or overestimate risk temporarily relative to fundamentals.
  - Eventually markets re-price to align with fundamentals, potentially causing sharp repricing and overshooting.
- Evidence for late-2012:
  - Large decreases in CDS spreads mainly occurred in countries with weaker fundamentals.
  - The fundamentals-based model suggested higher CDS spreads for a number of countries than market pricing at end-2012, consistent with possible under-pricing amid search-for-yield behavior.
- The upward drift in emerging-market CDS spreads in 2013 (noted in the text) partly reflected renewed differentiation based on fundamentals as market concerns about FED normalization emerged.

### Country examples and notable numeric observations
- Model estimation and forecast windows:
  - Error-correction model estimated for July 2007-June 2012; out-of-sample forecasts based on estimation from July 2007 to June 2012.
- Croatia:
  - June 2012: actual CDS broadly in line with fundamentals-based norm estimate.
  - December 2012: actual CDS spreads were well below fundamentals-based norm prediction, signaling potential under-pricing if the model is correct.
- CDS peaks noted in time series plots:
  - Ukraine's CDS spread peaks around 6000 in early 2009.
  - Greece's CDS spread shoots above 1500 during the Summer of 2011 and remains at levels well above 1000 bps thereafter.

### Policy implications and recommendations
- Countries with low CDS spreads despite weak macroeconomic fundamentals should implement adjustment policies proactively to avoid losses from an inevitable re-pricing of risk.
- Countries already facing a sharp re-pricing should balance:
  - Maintaining market confidence in fiscal consolidation.
  - Considering the growth impact of consolidation measures to avoid excessive short-term damage.
- Monitor market liquidity conditions and global investor sentiment as central channels transmitting euro-area stress to CESEE sovereign risk premia.

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp1417.pdf*

### ANNEX III

### _wp1417 - ANNEX III

### ANNEX III — Evolution of bid-ask spreads of 5 year sovereign CDS spreads (in basis points)
- Chart covers sample dates shown from "1/1/2007" through "9/1/2012".
- Countries shown (chart labels in source): Germany, Czech Republic, Bulgaria, Croatia, Ukraine, Portugal.
- Axis/tick labels shown in source include sequences: "1 3 5 7 9 11 13" and repeated date ticks "1/1/2007 5/1/2007 9/1/2007 1/1/2008 5/1/2008 9/1/2008 1/1/2009 5/1/2009 9/1/2009 1/1/2010 5/1/2010 9/1/2010 1/1/2011 5/1/2011 9/1/2011 1/1/2012 5/1/2012 9/1/2012".
- Additional numeric axis labels visible for some country charts: e.g., "1 6 11 16 21 26 31 36", "1 11 21 31 41 51 61 71 81 91", "1 51 101 151 201 251 301 351 401", "1 11 21 31 41 51 61 71 81 91 101".

### ANNEX IV — Country abbreviations used
- AUTAustriaGRCGreecePOLPoland
- BELBelg
- iumHRVCroatiaPRTPortugal
- BGRBulgariaHUNHungaryROMRomania
- CZ
- ECzech R.IRLIrelandRUSRussia
- DEUGerman
- yITAItalySVKSlovakia
- ESPS
- painLTULithuaniaSVNSlovenia
- ESTEstoniaLV
- ALatviaTURTurkey
- FRAFranceNLD    NetherlandsUKRUkraine

### ANNEX V — Additional Estimation Results for CDS

CDS with VSTOXX — CDS Spreads: Estimation the Norm (with Country Dummies)
- Columns: (1) full sample, (2) Euro Crisis, (3) Pre Euro Crisis.
- VSTOXX: 5.556*** (0.209); 7.077*** (0.383); 4.834*** (0.209)
- GDP growth forecast: -8.519*** (0.888); -2.370 (2.327); -10.77*** (0.978)
- GDP growth (next year, adjusted*): -40.10*** (2.672); -23.32*** (4.120); -42.11*** (3.144)
- Bid-ask spread (adjusted*): 7.275*** (0.307); 5.005*** (0.596); 5.987*** (0.358)
- Deficit*Debt80 (adjusted*): 29.07*** (2.359); 13.25*** (2.995); 18.31*** (3.787)
- Deficit next year (adjusted*): -3.771*** (1.370); -16.63*** (2.372); -3.640*** (1.311)
- CA balance (adjusted*): 1.503** (0.731); 0.0914 (1.776); 0.0682 (1.102)
- Interaction of 2 year growth with debt threshold (adjusted*): -41.10*** (4.050); -17.80*** (6.500); -11.48*** (3.433)
- Ultra-low growth dummy: 57.14*** (6.906); 68.19*** (12.02); 16.89** (7.945)
- Euro area dummy: -92.17*** (22.10); -12.01 (19.38); -212.2*** (15.28)
- Debt above 80: 157.8*** (8.449); 92.44*** (11.45); 173.9*** (19.42)
- Constant: -30.96 (23.14); -69.53*** (17.97); 127.3*** (15.21)
- Observations: 1,492; 703; 789
- Number of country: 23; 23; 23
- R-Squared: 0.878; 0.926; 0.926
- Note: "*variables adjusted for colinearity with other variables. Data source: Bloomberg, Consensus Forecast, and WEO. Note: Standard errors in parentheses. Country dummies and their interaction terms between the crisis dummy and the explanatory variables are not shown. Data limited to episodes where CDS was below 1000."

CDS Spreads: Structural Break Test Results (with VSTOXX)
- Dummy*VSTOXX: 3.356*** (0.497)
- Dummy*GDP growth forecast: 12.41*** (3.088)
- Dummy*GDP growth (next year, adjusted*): 22.21*** (5.583)
- Dummy*Bid-ask spread (adjusted*): 1.058 (0.719)
- Dummy*Deficit*Debt80 (adjusted*): -1.729 (6.141)
- Dummy*Deficit next year (adjusted*): -17.72*** (3.111)
- Dummy*CA balance (adjusted*): -1.016 (2.329)
- Dummy*Interaction of 2 year growth with debt threshold (adjusted*): -1.179 (8.759)
- Dummy*Ultra-low growth dummy: 66.26*** (15.14)
- Dummy*Euro area dummy: -17.21 (21.02)
- Dummy*Debt above 80: -74.24* (38.77)
- Euro crisis dummy: -22.95 (21.96)
- Constant: -64.93* (38.19)
- Test statistic for no structural break, χ2 (33): 1765
- Prob. of no structural break: 0
- Note: "*variables adjusted for colinearity with other variables. Data source: Bloomberg, Consensus Forecast, and WEO. Note: Standard errors in parentheses. Country dummies and their interaction terms with the crisis dummy are not shown. Data limited to episodes where CDS was below 1000."

Estimation Results for Equation (6.2), Various Specification for the Euro Crisis Period (VSTOXX version)
- Common variables across columns (1)-(4):
  - Δcds t-1: 0.430*** (0.0659); 0.432*** (0.0662); 0.431*** (0.0672); 0.429*** (0.0658)
  - Δcds t-2: 0.0558 (0.0345); 0.0564 (0.0344); 0.0559 (0.0346); 0.0548 (0.0345)
  - VSTOXX (difference): 2.193*** (0.266); 2.187*** (0.268); 2.197*** (0.266); 2.180*** (0.275)
  - CDS bid ask spread (bps) (difference): 3.124*** (0.708); 3.127*** (0.708); 3.105*** (0.722); 3.126*** (0.709)
  - lag of diff(ec): -0.468*** (0.0610); -0.469*** (0.0604); -0.465*** (0.0614); -0.466*** (0.0629)
  - 2nd lag of diff(ec): 0.312*** (0.0603); 0.314*** (0.0588); 0.311*** (0.0609); 0.310*** (0.0620)
  - Constants: 1.737** (0.808); 1.765** (0.789); 1.934* (1.002); 1.708** (0.798)
- Other coefficients shown with larger standard errors or insignificance are listed in source table (e.g., interaction terms, GDP growth differences).
- Observations: 627 (all columns)
- Number of country: 23 (all columns)
- R-Squared: 0.535; 0.535; 0.535; 0.536
- Arellano-Bond test for AR(1): 0.000300; 0.000296; 0.000297; 0.000311
- Arellano-Bond test for AR(2): 0.311; 0.308; 0.308; 0.305
- Note: "Robust standard errors in parentheses. Data source: Bloomberg, Consensus Forecast, and WEO."

Estimation Results for Equation (6.2), Various Specification (before the Euro Crisis Period) (VSTOXX version)
- Representative coefficients:
  - Δcds t-1: 0.300*** (0.0374); 0.300*** (0.0377); 0.310*** (0.0391); 0.302*** (0.0380)
  - Δcds t-2: -0.116*** (0.0327); -0.116*** (0.0307); -0.117*** (0.0322); -0.120*** (0.0351)
  - GDP growth (next year, forecast) (difference): -14.14*** (3.390); -14.38*** (3.304); -14.32*** (3.452); -12.85*** (3.197)
  - VSTOXX (difference): 2.264*** (0.564); 2.261*** (0.566); 2.255*** (0.545); 2.211*** (0.533)
  - CDS bid ask spread (bps) (difference): 2.273*** (0.358); 2.271*** (0.354); 2.316*** (0.341); 2.274*** (0.345)
  - ec t-1: -0.358*** (0.0767); -0.358*** (0.0764); -0.361*** (0.0745); -0.354*** (0.0752)
- Observations: 691 (all columns)
- Number of country: 23 (all columns)
- R-Squared: 0.618; 0.618; 0.623; 0.619
- Arellano-Bond test for AR(1): 0.000954; 0.000951; 0.000940; 0.000905
- Arellano-Bond test for AR(2): 0.974; 0.999; 0.706; 0.830

CDS in logs — with VIX — CDS Spreads: Estimation the Norm (with Country Dummies)
- Columns: (1) full sample, (2) Euro Crisis, (3) Pre Euro Crisis.
- VIX: 0.0384*** (0.00166); 0.0252*** (0.00183); 0.0443*** (0.00217)
- GDP growth forecast: -0.0846*** (0.00527); -0.0442*** (0.0110); -0.0933*** (0.00743)
- GDP growth (next year, adjusted*): -0.133*** (0.0176); -0.103*** (0.0179); -0.199*** (0.0248)
- Bid-ask spread (adjusted*): 0.00222*** (0.000309); 0.00131*** (0.000317); 0.00424*** (0.000880)
- Deficit*Debt80 (adjusted*): 0.0756*** (0.0206); -0.0118 (0.0190); -0.0472 (0.0606)
- Deficit next year (adjusted*): 0.0115 (0.00973); -0.110*** (0.0111); 0.0367** (0.0151)
- CA balance (adjusted*): -0.0247*** (0.00483); -0.000894 (0.00744); -0.0114 (0.00841)
- Interaction of 2 year growth with debt threshold (adjusted*): -0.358*** (0.0254); -0.0907** (0.0368); -0.517*** (0.0533)
- Ultra-low growth dummy: 0.473*** (0.0509); 0.237*** (0.0556); 0.337*** (0.0842)
- Euro area dummy: -0.309** (0.124); 0.238** (0.102); -2.013*** (0.197)
- Debt above 80: 1.169*** (0.0734); 0.332*** (0.0680); 1.783*** (0.326)
- Constant: 2.690*** (0.166); 3.879*** (0.129); 3.982*** (0.111)
- Observations: 1,541; 736; 805
- Number of country: 23; 23; 23
- R-Squared: 0.888; 0.928; 0.928

CDS Spreads: Structural Break Test Results (with VIX, logs)
- Dummy*VIX: -0.0200*** (0.00375)
- Dummy*GDP growth forecast: 0.0517** (0.0208)
- Dummy*GDP growth (next year, adjusted*): 0.0864** (0.0376)
- Dummy*Bid-ask spread (adjusted*): -0.00333*** (0.000785)
- Dummy*Deficit next year (adjusted*): -0.139*** (0.0224)
- Dummy*Interaction of 2 year growth with debt threshold (adjusted*): 0.425*** (0.0745)
- Dummy*Euro area dummy: 0.947*** (0.225)
- Dummy*Debt above 80: -1.419*** (0.286)
- Euro crisis dummy: 1.219*** (0.135)
- Constant: 2.717*** (0.254)
- Test statistic for no structural break, χ2 (33): 2195
- Prob. of no structural break: 0

Estimation Results for Equation (6.2), Various Specification for the Euro Crisis Period (VIX/logs)
- Representative coefficients (columns (1)-(4)):
  - Δcds t-1: 0.277** (0.119); 0.276** (0.120); 0.277** (0.120); 0.277** (0.119)
  - VIX (difference): 0.00698*** (0.000754); 0.00702*** (0.000798); 0.00696*** (0.000745); 0.00698*** (0.000747)
  - CDS bid ask spread (bps) (difference): 0.000143** (1.92e-05); 0.000145*** (1.92e-05); 0.000144*** (1.90e-05); 0.000141*** (1.99e-05)
  - ec t-1: -0.196*** (0.0372); -0.197*** (0.0362); -0.198*** (0.0360); -0.196*** (0.0380)
  - ec t-2: 0.0954*** (0.0266); 0.0944*** (0.0272); 0.0971*** (0.0264); 0.0954*** (0.0270)
- Observations: 667 (all columns)
- Number of country: 23
- R-Squared: 0.342; 0.343; 0.342; 0.342
- Arellano-Bond test for AR(1): 0.00165; 0.00158; 0.00166; 0.00151
- Arellano-Bond test for AR(2): 0.0450; 0.0490; 0.0454; 0.0521

Estimation Results for Equation (6.2), Various Specification (before the Euro Crisis Period) (VIX/logs)
- Representative coefficients (columns (1)-(4)):
  - Δcds t-1: 0.248*** (0.0353); 0.247*** (0.0352); 0.246*** (0.0355); 0.247*** (0.0356)
  - Δcds t-2: -0.0210 (0.0371); -0.0212 (0.0371); -0.0199 (0.0374); -0.0206 (0.0374)
  - GDP growth (next year, forecast) (difference): -0.0485*** (0.00889); -0.0495*** (0.00861); -0.0483*** (0.00900); -0.0449*** (0.00867)
  - VIX (difference): 0.0187*** (0.00133); 0.0187*** (0.00133); 0.0187*** (0.00134); 0.0185*** (0.00126)
  - CDS bid ask spread (bps) (difference): 0.00220*** (0.000273); 0.00220*** (0.000271); 0.00218*** (0.000274); 0.00218*** (0.000279)
  - ec t-1: -0.266*** (0.0349); -0.265*** (0.0350); -0.266*** (0.0348); -0.267*** (0.0352)
  - ec t-2: 0.0872** (0.0352); 0.0865** (0.0354); 0.0889** (0.0349); 0.0881** (0.0351)
- Observations: 713 (all columns)
- Number of country: 23
- R-Squared: 0.615; 0.615; 0.615; 0.617
- Arellano-Bond test for AR(1): 1.24e-05; 1.25e-05; 1.23e-05; 1.26e-05
- Arellano-Bond test for AR(2): 0.109; 0.121; 0.105; 0.101

CDS in logs — with VSTOXX — CDS Spreads: Estimation the Norm (with Country Dummies)
- VSTOXX: 0.0463*** (0.00169); 0.0336*** (0.00155); 0.0513*** (0.00225)
- GDP growth forecast: -0.0806*** (0.00511); -0.0402*** (0.00975); -0.0849*** (0.00718)
- GDP growth (next year, adjusted*): -0.111*** (0.0173); -0.0605*** (0.0161); -0.184*** (0.0241)
- Bid-ask spread (adjusted*): 0.00218*** (0.000301); 0.00129*** (0.000302); 0.00355*** (0.000829)
- Interaction of 2 year growth with debt threshold (adjusted*): -0.349*** (0.0245); -0.0790** (0.0331); -0.460*** (0.0514)
- Ultra-low growth dummy: 0.446*** (0.0490); 0.245*** (0.0495); 0.325*** (0.0812)
- Euro area dummy: -0.376*** (0.121); 0.240*** (0.0900); -1.997*** (0.188)
- Debt above 80: 1.135*** (0.0700); 0.317*** (0.0602); 1.687*** (0.316)
- Constant: 2.469*** (0.160); 3.507*** (0.114); 3.710*** (0.111)
- Observations: 1,541; 736; 805
- Number of country: 23
- R-Squared: 0.897; 0.940; 0.940

CDS Spreads: Structural Break Test Results (with VSTOXX, logs)
- Dummy*VSTOXX: -0.0175*** (0.00359)
- Dummy*GDP growth forecast: 0.0419** (0.0202)
- Dummy*GDP growth (next year, adjusted*): 0.110*** (0.0368)
- Dummy*Bid-ask spread (adjusted*): -0.00275*** (0.000753)
- Dummy*Deficit next year (adjusted*): -0.129*** (0.0212)
- Dummy*Interaction of 2 year growth with debt threshold (adjusted*): 0.379*** (0.0712)
- Dummy*Euro area dummy: 0.925*** (0.213)
- Dummy*Debt above 80: -1.333*** (0.275)
- Euro crisis dummy: 1.110*** (0.140)
- Constant: 2.422*** (0.246)
- Test statistic for no structural break, χ2 (33): 2108
- Prob. of no structural break: 0

Estimation Results for Equation (6.2), Various Specification for the Euro Crisis Period (VSTOXX/logs)
- Representative coefficients:
  - Δcds t-1: 0.331*** (0.122); 0.330*** (0.123); 0.331*** (0.122); 0.332*** (0.122)
  - VSTOXX (difference): 0.0100*** (0.000725); 0.0101*** (0.000750); 0.0101*** (0.000732); 0.0101*** (0.000718)
  - CDS bid ask spread (bps) (difference): 0.000128** (1.75e-05); 0.000131*** (1.84e-05); 0.000129*** (1.78e-05); 0.000125*** (1.74e-05)
  - ec t-1: -0.291*** (0.0503); -0.294*** (0.0487); -0.299*** (0.0502); -0.292*** (0.0510)
  - ec t-2: 0.179*** (0.0366); 0.178*** (0.0367); 0.184*** (0.0370); 0.179*** (0.0370)
- Observations: 667 (all columns)
- Number of country: 23
- R-Squared: 0.423; 0.425; 0.425; 0.423

Estimation Results for Equation (6.2), Various Specification (before the Euro Crisis Period) (VSTOXX/logs)
- Representative coefficients:
  - Δcds t-1: 0.283*** (0.0328); 0.283*** (0.0327); 0.282*** (0.0329); 0.281*** (0.0331)
  - Δcds t-2: -0.0639* (0.0337); -0.0641* (0.0337); -0.0629* (0.0340); -0.0626* (0.0342)
  - GDP growth (next year, forecast) (difference): -0.0409*** (0.00758); -0.0420*** (0.00701); -0.0407*** (0.00770); -0.0370*** (0.00719)
  - VSTOXX (difference): 0.0199*** (0.00122); 0.0199*** (0.00124); 0.0199*** (0.00123); 0.0197*** (0.00117)
  - CDS bid ask spread (bps) (difference): 0.00201*** (0.000246); 0.00202*** (0.000244); 0.00200*** (0.000250); 0.00199*** (0.000252)
  - ec t-1: -0.283*** (0.0298); -0.283*** (0.0300); -0.284*** (0.0295); -0.285*** (0.0303)
  - ec t-2: 0.0859*** (0.0266); 0.0862*** (0.0268); 0.0879*** (0.0258); 0.0872*** (0.0266)
- Observations: 713 (all columns)
- Number of country: 23
- R-Squared: 0.626; 0.626; 0.626; 0.627

*Data source: Bloomberg, Consensus Forecast, and WEO. Standard errors shown in parentheses in source tables; significance indicators: *** p<0.01, ** p<0.05, * p<0.1.*

### ANNEX VI. Estimation Results for end of month CDS

### ANNEX VI. Estimation Results for end of month CDS

### CDS with VIX — CDS Spreads: Estimation the Norm (with Country Dummies)
- Full sample (column (1)):
  - VIX 4.761*** (0.196)
  - GDP growth forecast -9.559*** (0.876)
  - GDP growth (next year, adjusted*) -39.55*** (2.610)
  - Bid-ask spread (adjusted*) 6.955*** (0.305)
  - Deficit*Debt80 (adjusted*) 27.19*** (2.378)
  - Deficit next year (adjusted*) -3.856*** (1.373)
  - CA balance (adjusted*) 1.741** (0.723)
  - Interaction of 2 year growth with debt threshold (adjusted*) -44.44*** (4.003)
  - Ultra-low growth dummy 53.02*** (6.871)
  - Euro area dummy -80.29*** (21.99)
  - Debt above 80 168.3*** (8.802)
  - Constant -11.33 (22.95)
  - Observations 1,494
  - Number of country 23
  - R-Squared 0.879
- Euro Crisis (column (2)) and Pre-Euro Crisis (column (3)) samples report analogous coefficients with their own standard errors and:
  - Observations 705 and 789 respectively
  - R-Squared 0.921 and 0.921 respectively
- Note: Standard errors in parentheses. Country dummies and their interaction terms between the crisis dummy and the explanatory variables are not shown. Data limited to episodes where CDS was below 1000.

### CDS with VIX — CDS Spreads: Structural Break Test Results
- Dummy*VIX 2.561*** (0.498)
- Dummy*GDP growth forecast 12.19*** (3.055)
- Dummy*Bid-ask spread (adjusted*) 2.774*** (0.678)
- Dummy*Deficit next year (adjusted*) -18.57*** (3.236)
- Dummy*Ultra-low growth dummy 74.58*** (15.20)
- Dummy*Debt above 80 -91.93** (43.31)
- Euro crisis dummy 36.55* (21.48)
- Constant -46.04 (37.82)
- Test statistic for no structural break, χ2 (33) 1871
- Prob. of no structural break 0

### Estimation Results for Equation (6.2) — Various Specifications for the Euro Crisis Period (with VIX)
- Common reported statistics across columns (1)–(4):
  - Δcds t-1: coefficients near -0.00701 to -0.00981 with standard errors (0.0656)
  - Δcds t-2: 0.138*** (0.0278) to 0.139*** (0.0268)
  - GDP growth (next year, forecast) (difference): -20.44*** (4.501) to -20.83*** (4.260)
  - VIX (difference): 4.367*** (0.477) to 4.351*** (0.478)
  - CDS bid ask spread (bps) (difference): 2.767*** (0.466) to 2.780*** (0.457)
  - ec t-1: -0.101* (0.0601) to -0.0952 (0.0614)
  - Observations 630; Number of country 23; R-Squared 0.592 to 0.594
  - Arellano-Bond test for AR(1) ≈ 0.00149–0.00155; AR(2) ≈ 0.0218–0.0282

### Estimation Results for Equation (6.2) — Various Specifications (before the Euro Crisis Period) (with VIX)
- Common reported statistics across columns (1)–(4):
  - Δcds t-1: 0.288*** (0.0458) to 0.298*** (0.0477)
  - Δcds t-2: 0.118*** (0.0358) to 0.115*** (0.0369)
  - GDP growth (next year, forecast) (difference): -10.14*** (3.488) to -9.440*** (3.369)
  - VIX (difference): 2.971*** (0.487) to 2.940*** (0.471)
  - CDS bid ask spread (bps) (difference): 1.005*** (0.247) to 1.051*** (0.233)
  - ec t-1: -0.341*** (0.0739) to -0.340*** (0.0720)
  - Observations 691; Number of country 23; R-Squared 0.600 to 0.606
  - Arellano-Bond test for AR(1) ≈ 0.00098–0.00102; AR(2) ≈ 0.00139–0.0282

---

### CDS with VSTOXX — CDS Spreads: Estimation the Norm (with Country Dummies)
- Full sample (column (1)):
  - VSTOXX 5.618*** (0.203)
  - GDP growth forecast -8.588*** (0.862)
  - GDP growth (next year, adjusted*) -36.29*** (2.603)
  - Bid-ask spread (adjusted*) 6.652*** (0.297)
  - Deficit*Debt80 (adjusted*) 27.89*** (2.403)
  - Deficit next year (adjusted*) -3.370** (1.358)
  - CA balance (adjusted*) 1.879*** (0.708)
  - Interaction of 2 year growth with debt threshold (adjusted*) -44.67*** (4.062)
  - Ultra-low growth dummy 51.88*** (6.795)
  - Euro area dummy -89.68*** (21.84)
  - Debt above 80 164.7*** (8.815)
  - Constant -34.72 (22.80)
  - Observations 1,494; Number of country 23; R-Squared 0.882
- Euro Crisis and Pre-Euro Crisis columns report corresponding estimates; Observations 705 and 789; R-Squared 0.931 and 0.931

### CDS with VSTOXX — CDS Spreads: Structural Break Test Results
- Dummy*VSTOXX 4.010*** (0.469)
- Dummy*GDP growth forecast 11.97*** (2.917)
- Dummy*GDP growth (next year, adjusted*) 18.71*** (5.303)
- Dummy*Bid-ask spread (adjusted*) 1.599** (0.658)
- Dummy*Deficit next year (adjusted*) -16.56*** (2.985)
- Dummy*Ultra-low growth dummy 84.44*** (14.26)
- Dummy*Debt above 80 -94.57** (43.32)
- Euro crisis dummy -36.58* (21.13)
- Constant -66.03* (37.55)
- Test statistic for no structural break, χ2 (33) 1991
- Prob. of no structural break 0

### Estimation Results for Equation (6.2) — Various Specifications for the Euro Crisis Period (with VSTOXX)
- Selected common coefficients across columns (1)–(4):
  - Δcds t-1: 0.0813 to 0.0791 (standard errors ≈ 0.0570–0.0583)
  - Δcds t-2: 0.117*** (0.0311) to 0.117*** (0.0305)
  - GDP growth (next year, forecast) (difference): -12.42*** (4.319) to -12.74*** (4.222)
  - VSTOXX (difference): 4.947*** (0.480) to 4.932*** (0.481)
  - CDS bid ask spread (bps) (difference): 2.751*** (0.464) to 2.760*** (0.461)
  - ec t-1: -0.239*** (0.0606) to -0.234*** (0.0617)
  - Observations 630; Number of country 23; R-Squared 0.651–0.652
  - Arellano-Bond test for AR(1) ≈ 0.00159–0.00166; AR(2) ≈ 0.0316–0.0394

### Estimation Results for Equation (6.2) — Various Specifications (before the Euro Crisis Period) (with VSTOXX)
- Selected common coefficients across columns (1)–(4):
  - Δcds t-1: 0.348*** (0.0400) to 0.358*** (0.0411)
  - Δcds t-2: 0.0906*** (0.0313) to 0.0886*** (0.0299)
  - GDP growth (next year, forecast) (difference): -7.534** (3.002) to -6.800** (2.913)
  - VSTOXX (difference): 2.874*** (0.463) to 2.842*** (0.449)
  - CDS bid ask spread (bps) (difference): 1.071*** (0.254) to 1.075*** (0.244)
  - ec t-1: -0.366*** (0.0738) to -0.365*** (0.0718)
  - Observations 691; Number of country 23; R-Squared 0.600–0.606
  - Arellano-Bond test for AR(1) ≈ 9.17e-06–0.000979; AR(2) ≈ 0.0122–0.00245

---

### CDS in logs — with VIX — CDS Spreads: Estimation the Norm (with Country Dummies)
- Full sample (column (1)):
  - VIX 0.0399*** (0.00162)
  - GDP growth forecast -0.0837*** (0.00514)
  - GDP growth (next year, adjusted*) -0.126*** (0.0172)
  - Bid-ask spread (adjusted*) 0.00227*** (0.000311)
  - Deficit*Debt80 (adjusted*) 0.0769*** (0.0203)
  - Deficit next year (adjusted*) 0.0147 (0.00952)
  - CA balance (adjusted*) -0.0234*** (0.00470)
  - Interaction of 2 year growth with debt threshold (adjusted*) -0.353*** (0.0252)
  - Ultra-low growth dummy 0.416*** (0.0496)
  - Euro area dummy -0.313** (0.125)
  - Debt above 80 1.153*** (0.0722)
  - Constant 2.682*** (0.163)
  - Observations 1,541; Number of country 23; R-Squared 0.890
- Euro Crisis and Pre-Euro Crisis columns report analogous log-scale estimates with Observations 736 and 805; R-Squared 0.929 and 0.929

### CDS in logs — Structural Break Test Results (with VIX)
- Dummy*VIX -0.0189*** (0.00362)
- Dummy*GDP growth forecast 0.0457** (0.0202)
- Dummy*GDP growth (next year, adjusted*) 0.0660* (0.0364)
- Dummy*Bid-ask spread (adjusted*) -0.00285*** (0.000744)
- Dummy*Deficit next year (adjusted*) -0.138*** (0.0216)
- Dummy*Interaction of 2 year growth with debt threshold (adjusted*) 0.408*** (0.0721)
- Dummy*Euro area dummy 0.943*** (0.217)
- Dummy*Debt above 80 -1.565*** (0.270)
- Euro crisis dummy 1.170*** (0.130)
- Constant 2.730*** (0.250)
- Test statistic for no structural break, χ2 (33) 2312
- Prob. of no structural break 0

### Estimation Results for Equation (6.2) — Various Specifications for the Euro Crisis Period (CDS logs, with VIX)
- Selected coefficients across columns (1)–(4):
  - Δcds t-1: 0.0269 to 0.0271 (std. errors ≈ 0.0204–0.0215)
  - GDP growth (next year, forecast) (difference): -0.0304** (0.0146) to -0.0331*** (0.0128)
  - VIX (difference): 0.0186*** (0.000770) to 0.0186*** (0.000775)
  - CDS bid ask spread (bps) (difference): 0.000204** (2.98e-05) to 0.000208*** (3.03e-05)
  - Interaction of avg. growth (two year) and Debt threshold (difference): -0.0585* (0.0325) to -0.0607* (0.0335)
  - ec t-2: -0.0935* (0.0518) to -0.0929* (0.0524)
  - Observations 667; Number of country 23; R-Squared 0.509–0.511
  - Arellano-Bond test for AR(1) ≈ 0.000592–0.000681; AR(2) ≈ 0.0156–0.0163

### Estimation Results for Equation (6.2) — Various Specifications (before the Euro Crisis Period) (CDS logs, with VIX)
- Selected coefficients across columns (1)–(4):
  - Δcds t-1: 0.138*** (0.0301) to 0.138*** (0.0302)
  - Δcds t-2: 0.0763*** (0.0295) to 0.0752** (0.0295)
  - GDP growth (next year, forecast) (difference): -0.0423*** (0.00796) to -0.0397*** (0.00766)
  - VIX (difference): 0.0240*** (0.000707) to 0.0239*** (0.000700)
  - CDS bid ask spread (bps) (difference): 0.00146*** (0.000308) to 0.00145*** (0.000304)
  - ec t-1: -0.169*** (0.0356) to -0.169*** (0.0357)
  - Observations 713; Number of country 23; R-Squared 0.656–0.658
  - Arellano-Bond test for AR(1) ≈ 8.36e-06–9.49e-06; AR(2) ≈ 0.305–0.412

---

### CDS in logs — with VSTOXX — CDS Spreads: Estimation the Norm (with Country Dummies)
- Full sample (column (1)):
  - VSTOXX 0.0480*** (0.00164)
  - GDP growth forecast -0.0796*** (0.00495)
  - GDP growth (next year, adjusted*) -0.102*** (0.0168)
  - Bid-ask spread (adjusted*) 0.00221*** (0.000302)
  - Deficit*Debt80 (adjusted*) 0.0771*** (0.0193)
  - Deficit next year (adjusted*) 0.0211** (0.00915)
  - CA balance (adjusted*) -0.0260*** (0.00452)
  - Interaction of 2 year growth with debt threshold (adjusted*) -0.342*** (0.0242)
  - Ultra-low growth dummy 0.387*** (0.0473)
  - Euro area dummy -0.381*** (0.121)
  - Debt above 80 1.119*** (0.0683)
  - Constant 2.452*** (0.157)
  - Observations 1,541; Number of country 23; R-Squared 0.900
- Euro Crisis and Pre-Euro Crisis columns report analogous log-scale estimates with Observations 736 and 805; R-Squared 0.942 and 0.942

### CDS in logs — Structural Break Test Results (with VSTOXX)
- Dummy*VSTOXX -0.0169*** (0.00342)
- Dummy*GDP growth forecast 0.0358* (0.0195)
- Dummy*GDP growth (next year, adjusted*) 0.0910*** (0.0352)
- Dummy*Bid-ask spread (adjusted*) -0.00223*** (0.000705)
- Dummy*Deficit next year (adjusted*) -0.125*** (0.0202)
- Dummy*Interaction of 2 year growth with debt threshold (adjusted*) 0.355*** (0.0682)
- Dummy*Euro area dummy 0.920*** (0.203)
- Dummy*Debt above 80 -1.474*** (0.257)
- Euro crisis dummy 1.063*** (0.133)
- Constant 2.427*** (0.241)
- Test statistic for no structural break, χ2 (33) 2247
- Prob. of no structural break 0

### Estimation Results for Equation (6.2) — Various Specifications for the Euro Crisis Period (CDS logs, with VSTOXX)
- Selected coefficients across columns (1)–(4):
  - Δcds t-1: 0.115** (0.0466) to 0.115** (0.0457)
  - Δcds t-2: small coefficients near 0.00811 to 0.0101 (std. errors ≈ 0.0538–0.0555)
  - GDP growth (next year, forecast) (difference): coefficients near zero and not significant in many specifications
  - VSTOXX (difference): 0.0213*** (0.000831) consistently
  - CDS bid ask spread (bps) (difference): 0.000170** (2.17e-05) to 0.000173*** (2.21e-05)
  - ec t-1: -0.131*** (0.0284) to -0.130*** (0.0280)
  - Observations 667; Number of country 23; R-Squared 0.586–0.587
  - Arellano-Bond test for AR(1) ≈ 0.00153–0.00174; AR(2) ≈ 0.0317–0.0361

### Estimation Results for Equation (6.2) — Various Specifications (before the Euro Crisis Period) (CDS logs, with VSTOXX)
- Selected coefficients across columns (1)–(4):
  - Δcds t-1: 0.224*** (0.0256) to 0.223*** (0.0258)
  - Δcds t-2: ≈ 0.0150–0.0157 (std. errors ≈ 0.0263)
  - GDP growth (next year, forecast) (difference): -0.0237*** (0.00737) to -0.0208*** (0.00704)
  - VSTOXX (difference): 0.0246*** (0.000768) to 0.0244*** (0.000782)
  - CDS bid ask spread (bps) (difference): 0.00128*** (0.000339) to 0.00127*** (0.000333)
  - ec t-1: -0.222*** (0.0349) to -0.223*** (0.0352)
  - Observations 713; Number of country 23; R-Squared 0.670–0.671
  - Arellano-Bond test for AR(1) ≈ 8.34e-06–8.85e-06; AR(2) ≈ 0.102–0.161

*Data source: Bloomberg, Consensus Forecast, and WEO.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp1417.pdf_