## wp1774

## Source details

**Canonical URL:** [wp1774](https://www.imf.org/-/media/files/publications/wp/2017/wp1774.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2017/wp1774.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2017/wp1774.pdf.json)

---

### I. Research question and context
- Purpose: Assess whether data transparency reforms—IMF Data Standards Initiatives (Special Data Dissemination Standard, SDDS; General Data Dissemination System, GDDS)—reduce sovereign bond spreads in emerging economies, using an event study to mitigate endogeneity.
- Background:
  - SDDS established in April in 1996 in response to the Mexican financial crisis.
  - GDDS introduced in 1997 amid the Asian financial crises.
  - e-GDDS introduced in May 2015; all GDDS participation events in this study occurred before the introduction of the e-GDDS.
  - Special Data Dissemination Standard Plus introduced in 2012 as part of the G-20 Data Gaps Initiative.
- Rationale: Improved data dissemination is expected to increase transparency and mitigate perceived risk by international investors; prior literature provides mixed evidence.

### II. Data, sample, and measurement
- Sample and data sources:
  - 52 emerging market economies where the EMBIG is available.
  - EMBIG: J.P. Morgan’s Emerging Market Bond Index Global; EMBIG spread defined as EMBIG portfolio yield over a theoretical U.S. zero coupon curve.
  - EMBIG downloaded from Bloomberg at quarterly frequency; other macroeconomic variables from Haver Analytics.
- Coverage and counts:
  - As of December 2016, 74 countries were subscribing to the SDDS (including 11 countries subscribing to the SDDS Plus).
  - As of December 2016, 110 countries were participating in the GDDS.
  - For the event analysis, up to 26 countries adopted a reform during the sample period in which EMBIG spreads are available.
- Dependent variable and timing:
  - Baseline dependent variable: end-of-quarter daily EMBIG spreads; robustness check uses the quarterly average of daily EMBIG spreads.
  - Event windows used: 1, 2, 4 (benchmark), and 8 quarters.
- Key macroeconomic predictors examined for reform timing:
  - current-account-to-GDP ratio
  - external-debt-to-GDP ratio
  - government-surplus-to-GDP ratio
  - real GDP growth
  - growth rate of the real effective exchange rate (REER growth)
  - inflation rate
- Overlap and additional events:
  - Identified 10 cases in which data transparency reforms took place within a two-year window of IMF-supported programs.
  - Identified 5 cases in which reforms took place within a two-year window of systematic banking crises (Laeven and Valencia 2012).
  - Nigeria identified as a potential outlier (EMBIG narrowed to 800 basis points from about 1,700 after subscription).

### III. Empirical approach and identification
- Two-step identification strategy:
  1. Test whether macroeconomic variables systematically changed prior to reforms using the model z_{i,t} = α_i + η_t + sum_{s=-k}^{k} β_s δ_{i,s,t} + ε_{i,t}, where δ_{i,s,t} equals one when country i is s periods away from subscription.
  2. Event-study difference-in-means around reform dates to estimate effect on EMBIG spreads using z_{i,t} = α_i + η_t + γ X_{i,t} + λ R_{i,t} + ε_{i,t}, where R_{i,t} is treatment dummy, X_{i,t} is vector of controls.
- Also estimate a full-sample panel regression: z_{i,t} = α_i + η_t + γ X_{i,t} + ε_{i,t}.
- Controls for global conditions:
  - U.S. federal funds rate (quarterly average) and VIX included as global controls in some specifications in lieu of time-fixed effects.

### IV. Main empirical findings
- Exogeneity of reform timing:
  - None of the six selected macroeconomic variables shows a significant deviation around reform dates (plots of β_s with 80 percent and 95 percent confidence intervals), indicating reforms were unlikely to have been driven by those macroeconomic developments.
  - Movements of the six variables before reforms were mixed: current-account-to-GDP ratio, external-debt-to-GDP ratio, and REER growth move toward increasing spreads; government-surplus-to-GDP ratio, real GDP growth, and inflation move in the opposite direction.
- Effect of reforms on sovereign spreads:
  - Event-study estimates of λ are consistently negative and statistically significant for all windows except the one-quarter window.
  - Baseline four-quarter window estimate corresponds to about a 75-basis-point decline from the average of 550 basis points in EMBIG spreads.
  - Placebo test: moving event dates one year ahead shows no systematic decline in EMBIG spreads.
- Interpretation:
  - Observed short-run reduction in secondary-market spreads implies increased investor confidence that can feed into primary-market issuance conditions over time.

### V. Robustness checks and sensitivity analyses
- Global controls and dependent variable treatment:
  - Replacing time-fixed effects with the U.S. federal funds rate and VIX substantially increases the magnitude and statistical significance of the main coefficient (Table 3, Column I).
  - Replacing end-of-quarter EMBIG with the quarterly average of daily EMBIG spreads does not affect the finding (Table 3, Column II).
- Outlier and overlapping-event treatments:
  - Dropping Nigeria and re-estimating Equation (2) (Column III) does not materially change core results.
  - Dropping countries with overlapping IMF-supported programs within a two-year window (10 cases) and dropping those with overlapping systematic banking crises within a two-year window (5 cases) (Columns IV and V) hardly affect the estimated reform effect.
- Window sensitivity:
  - Results robust across event windows of 1, 2, 4, and 8 quarters, except the one-quarter window where effects are not significant.
- Lagged dependent variable:
  - Inclusion of a lag of the log EMBIG yields statistically significant results (Table 3, Column VI) though Nickell (1981) bias cautions interpretation.
- Controlling for additional macroeconomic factors:
  - Inclusion of government-surplus-to-GDP ratio, external-debt-to-GDP ratio, real GDP growth, REER growth, and inflation in various combinations does not alter the main conclusion (Table 4, Columns I–VII).
- Alternative global/emerging risk controls:
  - Controlling for an emerging market risk factor (weighted average of EMBIG spreads across sample countries, weights by GDP in U.S. dollars) yields coefficient comparable to baseline (Table 5, Columns I–II).
  - Inclusion of sovereign ratings (Fitch converted to 0–19 scale) does not change qualitative results (Table 5, Column III).
- Placebo tests:
  - Event dates artificially moved k+1 quarters forward and backward produce no significant effects across event windows (Table 6).
- Subscription vs. compliance:
  - Replacing SDDS subscription dates with compliance dates still shows effectiveness but with decreased magnitude across every event window (Table 7).
- Non-linearity: SDDS vs. GDDS:
  - Larger effects for SDDS than GDDS in every window (Table 8); interpretation may reflect a non-linear effect or differences in country characteristics across regimes.

### VI. Key quantitative results and statistics
- Headline estimate: data dissemination efforts via IMF platforms (SDDS and GDDS) lead to a 15 percent reduction in sovereign risk premia one year following reforms.
- Baseline event-study estimates (Table 2, dependent variable: Log of the EMBIG spread):
  - k=4: λ = -0.142**
  - k=1: λ = -0.067
  - k=2: λ = -0.109*
  - k=8: λ = -0.189***
  - Observations: 182; 62; 97; 336 respectively.
- Robustness checks (Table 3, k=4, Dependent variable: Log of the EMBIG spread):
  - Column I: λ = -0.424*** (0.106)
  - Column II: λ = -0.110* (0.058)
  - Column III: λ = -0.130** (0.060)
  - Column IV: λ = -0.306*** (0.062)
  - Column V: λ = -0.161** (0.080)
  - Column VI: λ = -0.122** (0.057)
  - Column VII: λ = -0.128* (0.073)
  - VIX coefficient example: VIX = 0.020** (0.008) in Column I.
  - Lag of the log EMBIG reported as 0.245** (0.099) and 0.680*** (0.100) in specified columns.
  - Observations for k=4 robustness: 182 (varying by column).
- Controlling for macro factors (Table 4, k=4):
  - λ values across Columns I–VI: -0.118*** (0.010), -0.214** (0.087), -0.205*** (0.033), -0.132* (0.075), -0.164** (0.068), -0.199*** (0.043).
  - REER Growth: -0.014*** (0.004) and -0.008** (0.004) in relevant columns.
  - Inflation Rate: 0.021* (0.011) in one column.
  - Observations vary: 70, 63, 88, 156, 156, 72.
- Additional factors (Table 5, k=4):
  - λ: -0.141* (0.076), -0.142* (0.073), -0.075 (0.060), -0.188** (0.085) across Columns I–IV.
  - Average EMBIG coefficients: 0.822*** (0.061), 0.785*** (0.075), 0.622*** (0.120) when included.
  - Sovereign rating coefficients: -0.084*** (0.014), -0.092*** (0.060) where included.
  - Observations: 182, 182, 108, 108.
- Placebo tests (Table 6):
  - k=4 forward λ = 0.077 (0.123), backward λ = 0.082 (0.100); Obs 162 and 171 respectively.
- Subscriptions vs. Compliance (Table 7):
  - λ (Compliance) for k=4: -0.089** (0.037)
  - k=1: -0.124 (0.092); k=2: -0.095* (0.056); k=8: -0.159*** (0.054)
  - Observations: 247; 83; 138; 455.
- SDDS vs. GDDS (Table 8, k=4):
  - SDDS λ (k=4): -0.221** (0.086); Observations: 117.
  - GDDS λ (k=4): -0.137* (0.076); Observations: 65.

### VII. Mechanisms, limitations, and scope
- Mechanisms:
  - Adoption of inflation targeting can drive both EMBIG spreads and decisions to undertake reforms by improving central bank credibility and encouraging reforms in data dissemination (Mishkin 2000, Amato and Gerlach 2002).
  - A statistically significant decline in the inflation rate following the reforms (Figure 3) is consistent with the inflation-targeting mechanism.
  - Because adoption of inflation targeting is a one-time irreversible event, countries experiencing a change in inflation targeting during the event window were dropped; Brazil was the only case and deleting it does not alter main results.
- Limitations and scope:
  - Analysis focuses on short-run effects observed in secondary-market spreads; it does not attempt to quantify long-run effects.
  - EMBIG covers only emerging markets; advanced-economy observations are excluded by construction of the dependent variable.
  - Small sample size may limit power to reject some null hypotheses; 80 percent confidence intervals were plotted to reflect this.

### VIII. Policy implications
- Empirical evidence supports that participation in IMF data platforms (SDDS and GDDS) materially reduces sovereign bond spreads; headline estimate is a 15 percent reduction in sovereign risk premia one year after reforms.
- Results are robust across multiple specifications: inclusion of lagged dependent variable, additional macroeconomic controls, alternative global/emerging risk controls, sovereign ratings, placebo tests, use of compliance dates, and subgroup analyses (SDDS vs. GDDS).
- Policy implication: Promoting data transparency through IMF Data Standards Initiatives (SDDS and GDDS) can play a significant role in lowering sovereign borrowing costs and strengthening the international financial system.

*Source: wp1774 - Appendix Tables and conclusion (IMF working paper content).*

### References .............................................................................................................

### References

### Figures
- 1. The EMBIG Spreads Prior to and After Data Transparency Policy Reforms ......................... 7
- 2. The EMBIG Spreads Prior to and After Data Transparency Policy Reforms: Placebo Test .. 8
- 3. Conditional Mean of Selected Variables Around Data Transparency Policy Reforms ......... 13

### Tables
- 1. Impact of Macroeconomic Variables on Sovereign Bond Spreads ....................................... 20
- 2. Impact of Data Transparency Policy Reforms on Sovereign Bond Spreads: Baseline ......... 20
- 3. Data Transparency Policy Reforms on Sovereign Bond Spreads: Robustness Checks......... 21
- 4. Impact of Data Transparency Policy Reforms on Sovereign Bond Spreads: Controlling for Macroeconomic Factors ............................................................................................................. 21
- 5. Impact of Data Transparency Policy Reforms on Sovereign Bond Spreads: Controlling for Additional Factors ...................................................................................................................... 22
- 6. Impact of Data Transparency Policy Reforms on Sovereign Bond Spreads: Placebo Test ... 22
- 7. Impact of Data Transparency Policy Reforms: Subscriptions vs. Compliance ..................... 23
- 8. Impact of Data Transparency Policy Reforms: SDDS vs. GDDS ......................................... 23

*Source: wp1774 - References*

### Appendix Tables

### Appendix Tables

### I. Research question and context
- Purpose: Assess whether data transparency reforms—IMF Data Standards Initiatives (Special Data Dissemination Standard, SDDS; General Data Dissemination System, GDDS)—reduce sovereign bond spreads in emerging economies, using an event study to mitigate endogeneity.
- Background:
  - SDDS established in April in 1996 in response to the Mexican financial crisis.
  - GDDS introduced in 1997 amid the Asian financial crises.
  - e-GDDS introduced in May 2015; all GDDS participation events in this study occurred before the introduction of the e-GDDS.
  - Special Data Dissemination Standard Plus introduced in 2012 as part of the G-20 Data Gaps Initiative.
- Rationale: Improved data dissemination is expected to increase transparency and mitigate perceived risk by international investors; prior literature provides mixed evidence.

### II. Data, sample, and measurement
- Sample and data sources:
  - 52 emerging market economies where the EMBIG is available.
  - EMBIG: J.P. Morgan’s Emerging Market Bond Index Global; EMBIG spread defined as EMBIG portfolio yield over a theoretical U.S. zero coupon curve.
  - EMBIG downloaded from Bloomberg at quarterly frequency; other macroeconomic variables from Haver Analytics.
- Coverage and counts:
  - As of December 2016, 74 countries were subscribing to the SDDS (including 11 countries subscribing to the SDDS Plus).
  - As of December 2016, 110 countries were participating in the GDDS.
  - For the event analysis, up to 26 countries adopted a reform during the sample period in which EMBIG spreads are available.
- Dependent variable and timing:
  - Baseline dependent variable: end-of-quarter daily EMBIG spreads; robustness check uses the quarterly average of daily EMBIG spreads.
  - Event windows used: 1, 2, 4 (benchmark), and 8 quarters.
- Key variables examined as potential predictors of reform timing (six macroeconomic determinants of sovereign borrowing costs):
  - current-account-to-GDP ratio
  - external-debt-to-GDP ratio
  - government-surplus-to-GDP ratio
  - real GDP growth
  - growth rate of the real effective exchange rate (REER growth)
  - inflation rate
- Overlap and additional events:
  - Identified 10 cases in which data transparency reforms took place within a two-year window of IMF-supported programs.
  - Identified 5 cases in which reforms took place within a two-year window of systematic banking crises (Laeven and Valencia 2012).
  - Nigeria identified as a potential outlier (EMBIG narrowed to 800 basis points from about 1,700 after subscription).

### III. Empirical approach
- Identification strategy:
  - Two-step approach to address endogeneity and reverse causality:
    1. Test whether macroeconomic variables systematically changed prior to reforms (following methodology from Gourinchas and Obstfeld (2012) and Catao and Milesi-Ferretti (2014)).
       - Model: z_{i,t} = α_i + η_t + sum_{s=-k}^{k} β_s δ_{i,s,t} + ε_{i,t}, where δ_{i,s,t} equals one when country i is s periods away from subscription.
    2. Event-study difference-in-means around reform dates to estimate the effect on EMBIG spreads.
       - Model: z_{i,t} = α_i + η_t + γ X_{i,t} + λ R_{i,t} + ε_{i,t}, where R_{i,t} is treatment dummy, X_{i,t} is vector of controls.
  - Also estimate a full-sample panel regression to relate macroeconomic factors to EMBIG spreads:
    - Model: z_{i,t} = α_i + η_t + γ X_{i,t} + ε_{i,t}.
- Controls for global conditions:
  - U.S. federal funds rate (quarterly average) and VIX (Chicago Board Options Exchange Market Volatility Index) included as global controls in some specifications in lieu of time-fixed effects.

### IV. Main empirical findings
- Exogeneity of reform timing:
  - None of the six selected macroeconomic variables shows a significant deviation around reform dates (plots of β_s with 80 percent and 95 percent confidence intervals), indicating reforms were unlikely to have been driven by those macroeconomic developments.
  - Movements of the six variables before reforms were mixed: current-account-to-GDP ratio, external-debt-to-GDP ratio, and REER growth move toward increasing spreads; government-surplus-to-GDP ratio, real GDP growth, and inflation move in the opposite direction.
- Effect of reforms on sovereign spreads:
  - Event-study estimates of λ are consistently negative and statistically significant for all windows except the one-quarter window.
  - Baseline four-quarter window estimate corresponds to about a 75-basis-point decline from the average of 550 basis points in EMBIG spreads.
  - Placebo test: moving event dates one year ahead shows no systematic decline in EMBIG spreads, reducing concern that observed declines simply reflect a preexisting downward trend.
- Interpretations:
  - The observed short-run reduction in secondary-market spreads implies increased investor confidence that can feed into primary-market issuance conditions over time.

### V. Robustness checks and sensitivity analyses
- Global controls and dependent variable treatment:
  - Replacing time-fixed effects with the U.S. federal funds rate and VIX substantially increases the magnitude and statistical significance of the main coefficient (Table 3, Column I).
  - Replacing end-of-quarter EMBIG with the quarterly average of daily EMBIG spreads does not affect the finding (Table 3, Column II).
- Outlier and overlapping-event treatments:
  - Dropping Nigeria (outlier with EMBIG narrowing from about 1,700 to 800 basis points) and re-estimating Equation (2) (Column III) does not materially change core results.
  - Dropping countries with overlapping IMF-supported programs within a two-year window (10 cases) and dropping those with overlapping systematic banking crises within a two-year window (5 cases) (Columns IV and V) hardly affect the estimated reform effect.
- Window sensitivity:
  - Benchmarked on a four-quarter window; results robust across event windows of 1, 2, 4, and 8 quarters, except the one-quarter window where effects are not significant.

### VI. Limitations and scope
- The analysis focuses on short-run effects observed in secondary-market spreads; it does not attempt to quantify long-run effects.
- EMBIG covers only emerging markets; advanced-economy observations are lost by construction of the dependent variable.
- Small sample size may limit power to reject some null hypotheses; 80 percent confidence intervals were plotted to reflect this.

*Source: wp1774 - Appendix Tables (IMF working paper content).*

### conclusion (Table 3, Column III-V).

### conclusion (Table 3, Column III-V)

### Mechanisms and related events
- Adoption of inflation targeting can drive both EMBIG spreads and decisions to undertake reforms by improving central bank credibility and encouraging reforms in data dissemination and mitigation of inflation-financing concerns (Mishkin 2000, Amato and Gerlach 2002).
- A statistically significant decline in the inflation rate following the reforms (Figure 3) is consistent with the inflation-targeting mechanism.
- Because adoption of inflation targeting is a one-time irreversible event (as is data transparency reform), the authors drop countries experiencing a change in inflation targeting during the event window; Brazil is the only case and deleting it does not alter main results.
- Subscriptions to IMF data platforms (SDDS and GDDS) signal willingness to provide high-quality macroeconomic data but do not guarantee successful reforms; compliance dates are available from IMF Statistics Department monitoring.

### Robustness checks and sensitivity analyses
- Lagged dependent variable
  - A lag of the log EMBIG is included despite potential Nickell (1981) bias in a dynamic panel with fixed effects and short time dimension.
  - Results remain statistically significant when including the lag (Table 3, Column VI), though the test is presented as suggestive rather than definitive.

- Controlling for additional macroeconomic factors
  - Additional macro variables included in Equation (2): government-surplus-to-GDP ratio, external-debt-to-GDP ratio, real GDP growth, real effective exchange rate growth (REER growth), and the inflation rate.
  - Results shown by adding each macro variable in turn (Table 4, Columns I-VI) and by including three macro variables simultaneously (Table 4, Column VII).
  - All statistically significant variables enter with the predicted sign and inclusion does not alter the main conclusion.

- Alternative control variables
  - Global control concern: VIX may not capture emerging-market-specific risks; authors control for an emerging market risk factor measured by the weighted average of EMBIG spreads across sample countries (weights determined by GDP in U.S. dollars).
  - After controlling for the emerging market risk factor, the coefficient size is comparable to baseline (Table 5, Columns I-II).
  - Sovereign ratings: Fitch letter ratings converted to a 0–19 scale (AAA highest, D “Default” lowest). Inclusion of sovereign rating does not change qualitative results; credit ratings show predicted sign (Table 5, Column III).

- Placebo tests: false dates
  - Event dates are artificially moved k+1 quarters forward and backward (no overlap with true windows).
  - Coefficients on false reform dummies show no significant effects across event windows (Table 6), supporting identification.

- Subscription vs. compliance with SDDS
  - Replacing SDDS subscription dates with compliance dates (where IMF Statistics reports actual compliance) yields continued effectiveness of Data Standards Initiatives in reducing sovereign bond spreads (Table 7).
  - Magnitude of coefficients decreases across every event window when using compliance dates, suggesting the effect of initiating a reform is greater than completing it.

- Non-linearity: SDDS vs. GDDS
  - Splitting sample shows larger effects for SDDS than GDDS in every window (Table 8), consistent with Cady and Pellechio (2006).
  - Possible interpretations: a non-linear effect where only high standards (SDDS) change investor perceptions, or inherent differences between SDDS (middle-/high-income) and GDDS (developing/low-income) countries. Limited graduations from GDDS to SDDS in sample prevent definitive separation.

### Key quantitative results and statistics
- Main headline: data dissemination efforts via IMF platforms (SDDS and GDDS) lead to a 15 percent reduction in sovereign risk premia one year following reforms.
- Baseline event-study estimates (Table 2, dependent variable: Log of the EMBIG spread):
  - k=4: λ = -0.142**
  - k=1: λ = -0.067
  - k=2: λ = -0.109*
  - k=8: λ = -0.189***
  - Observations: 182; 62; 97; 336 respectively.
- Robustness checks (Table 3, k=4, Dependent variable: Log of the EMBIG spread):
  - Column I: λ = -0.424*** (0.106)
  - Column II: λ = -0.110* (0.058)
  - Column III: λ = -0.130** (0.060)
  - Column IV: λ = -0.306*** (0.062)
  - Column V: λ = -0.161** (0.080)
  - Column VI: λ = -0.122** (0.057)
  - Column VII: λ = -0.128* (0.073)
  - VIX and U.S. FFR coefficients reported in Table 3 where applicable (e.g., VIX = 0.020** (0.008) in Column I).
  - Lag of the log EMBIG appears in some columns: 0.245** (0.099) and 0.680*** (0.100) in specified columns.
  - Obs for k=4 robustness: 182 (varying by column).
- Controlling for macro factors (Table 4, k=4):
  - λ values across Columns I–VI: -0.118*** (0.010), -0.214** (0.087), -0.205*** (0.033), -0.132* (0.075), -0.164** (0.068), -0.199*** (0.043).
  - REER Growth: -0.014*** (0.004) and -0.008** (0.004) in relevant columns.
  - Inflation Rate: 0.021* (0.011) in one column.
  - Observations vary across columns: 70, 63, 88, 156, 156, 72.
- Additional factors (Table 5, k=4):
  - λ: -0.141* (0.076), -0.142* (0.073), -0.075 (0.060), -0.188** (0.085) across Columns I–IV.
  - Average EMBIG coefficients: 0.822*** (0.061), 0.785*** (0.075), 0.622*** (0.120) when included.
  - Sovereign rating coefficients: -0.084*** (0.014), -0.092*** (0.060) where included.
  - Observations: 182, 182, 108, 108.
- Placebo tests (Table 6):
  - k=4 forward λ = 0.077 (0.123), backward λ = 0.082 (0.100); Obs 162 and 171 respectively.
  - Other windows reported with no significant placebo effects.
- Subscriptions vs. Compliance (Table 7):
  - λ (Compliance) for k=4: -0.089** (0.037)
  - k=1: -0.124 (0.092); k=2: -0.095* (0.056); k=8: -0.159*** (0.054)
  - Observations: 247; 83; 138; 455.
- SDDS vs. GDDS (Table 8, k=4):
  - SDDS λ (k=4): -0.221** (0.086); Observations: 117.
  - GDDS λ (k=4): -0.137* (0.076); Observations: 65.
  - Other k windows reported with corresponding λ and Obs.

### Conclusion and policy implications
- Empirical evidence supports that participation in IMF data platforms (SDDS and GDDS) materially reduces sovereign bond spreads; headline estimate is a 15 percent reduction in sovereign risk premia one year after reforms.
- Results are robust across multiple specifications: inclusion of lagged dependent variable, additional macroeconomic controls, alternative global/emerging risk controls, sovereign ratings, placebo tests, use of compliance dates, and subgroup analyses (SDDS vs. GDDS).
- Policy implication: Promoting data transparency through IMF Data Standards Initiatives (SDDS and GDDS) can play a significant role in lowering sovereign borrowing costs and strengthening the international financial system.

*Source: wp1774 - conclusion (Table 3, Column III-V).*

### 55. Peterson Institute, 1998

### 55. Peterson Institute, 1998

### Key bibliographic references
- Gonzalez‐Rozada, Martín, and Eduardo Levy Yeyati. “Global factors and emerging market spreads.” Economic Journal 118.533 (2008): 1917-1936.
- Gourinchas, Pierre-Olivier, and Maurice Obstfeld. “Stories of the twentieth century for the twenty-first.” American Economic Journal: Macroeconomics 4.1 (2012): 226-265.
- Hammond, Gill. “State of the Art of Inflation Targeting.” Bank of England. CCBS Handbook. 2012.
- Hashimoto, Yuko, and Konstantin M. Wacker. “The role of risk and information for international capital flows: New evidence from the SDDS.” Review of World Economics, Volume 152, Issue 3 (2016), pp 529–557.
- Hollyer, James R., B. Peter Rosendorff, and James Raymond Vreeland. “Measuring transparency.” Political analysis (2014).
- International Monetary Fund, World Economic Outlook, April 2016; Chapter 3: Time for a Supply-Side Boost? Macroeconomic Effects of Labor and Product Market Reforms in Advanced Economies (2016a).
- International Monetary Fund, Staff Background Paper for the G20 Surveillance Note - Priorities for Structural Reforms in G20 Countries (2016b).
- Johnson, Simon, et al. “Corporate governance in the Asian financial crisis.” Journal of Financial Economics 58.1 (2000): 141-186.
- Kamin, Steven B., and Karsten Von Kleist. “The Evolution and Determinants of Emerging Market Credit Spreads in the 1990s.” International Finance Working Paper 653 (1999).
- Kaminsky, Graciela, and Sergio L. Schmukler. “Emerging market instability: do sovereign ratings affect country risk and stock returns?” World Bank Economic Review 16.2 (2002): 171-195.
- Laeven, Luc, and Fabian Valencia. “Systemic banking crises: a new database.” IMF working papers (2012): 1-78.
- Lane, Philip R., and Gian Maria Milesi-Ferretti. “International investment patterns.” Review of Economics and Statistics 90.3 (2008): 538-549.
- Marques, Luís Brandão, Gaston Gelos, and Natalia Melgar. “Country transparency and the global transmission of financial shocks.” No. 13-156. International Monetary Fund, 2013.
- Mishkin, Frederic S. “Inflation Targeting in Emerging-Market Countries.” American Economic Review 90.2 (2000): 105-109.
- Mora, Nada. “Sovereign credit ratings: Guilty beyond reasonable doubt?” Journal of Banking & Finance 30.7 (2006): 2041-2062.
- Moretti, Laura “Transparency and emerging market bond spreads.” Univ.-Bibliothek Frankfurt am Main, 2012.
- Morris, Stephen, and Hyun Song Shin. “Social value of public information.” American Economic Review 92.5 (2002): 1521-1534.
- Mrkaic, Mico. “Data dissemination standards and the statistical quality of the IMF’s World Economic Outlook Forecasts.” No. 10-203. International Monetary Fund, 2010.
- Nickell, Stephen. “Biases in Dynamic Models with Fixed Effects.” Econometrica 49.6 (1981): 1417-26.
- Portes, Richard, Hélène Rey, and Yonghyup Oh. “Information and capital flows: The determinants of transactions in financial assets.” European Economic Review 45.4 (2001): 783-796.
- Ranciere, Romain G., and Aaron Tornell. “Why Do Reforms Occur in Crises Times?” USC-INET Research Paper 16-19 (2016).
- Reinhart, Carmen M., and Christoph Trebesch. “The International Monetary Fund: 70 Years of Reinvention.” Journal of Economic Perspectives 30.1 (2016): 3-27.
- Sy, Amadou NR. “Emerging market bond spreads and sovereign credit ratings: reconciling market views with economic fundamentals.” Emerging Markets Review 3.4 (2002): 380-408.
- Tapsoba, Sampawende J., Robert C. York, and Neree CGM Noumon. “Can Statistical Capacity Building Help Reduce Procyclical Fiscal Policy in Developing Countries?” No. 16/209. International Monetary Fund, 2016.
- Tesar, Linda L., and Ingrid M. Werner. “Home bias and high turnover.” Journal of International Money and Finance 14.4 (1995): 467-492.
- Tiffin, Andrew, Christian B. Mulder, and Charalambos Christofides. “The Link Between Adherence to International Standards of Good Practice, Foreign Exchange Spreads, and Ratings.” No. 03/74. International Monetary Fund, 2003.
- Tong, Hui. “Disclosure standards and market efficiency: Evidence from analysts’ forecasts.” Journal of International Economics 72.1 (2007): 222-241.
- Walsh, Carl E. “Optimal economic transparency.” International Journal of Central Banking 3.1 (2007).
- Wei, Shang-Jin, and Yi Wu. “Negative alchemy? Corruption, composition of capital flows, and currency crises.” Preventing currency crises in emerging markets. University of Chicago Press, 2002. 461-506.

### Appendix: Data description (Table A.1)
- EMBIG spreads: J.P. Morgan’s Emerging Market Bond Index Global (in basis point) — Bloomberg
- Inflation rate: y-o-y growth rate of CPI — IFS, Haver Analytics
- GDP growth: y-o-y growth rate of real GDP — IFS, Haver Analytics
- REER growth: y-o-y growth rate of real effective exchange rate — BIS, Haver Analytics
- Current account to GDP ratio: The ratio of current account to nominal GDP — IFS, Haver Analytics
- External debt to GDP ratio: The ratio of external debt to nominal GDP — IFS, Haver Analytics
- Government surplus to GDP ratio: The ratio of general government surplus to nominal GDP — IFS, Haver Analytics
- Federal Funds rate: The effective Federal Funds rate — Federal Reserve Economic Data
- VIX: The CBOE Volatility Index — Chicago Board Options Exchange
- Reform dates: SDDS subscription/compliance, GDDS participation — http://dsbb.imf.org/pages/sdds/home.aspx
- IMF program dates: Beginning and the ending dates of historical IMF programs — http://www.imf.org/en/data/imf-finances
- Banking crisis dates: Systemic banking crises: a new database — Laeven and Valencia (2012)
- Sovereign rating history: The history of changes in Fitch’s sovereign ratings — https://www.fitchratings.com/site/sovereigns
- Inflation targeting adoption dates: The adoption dates of an inflation targeting regime — Hammond (2012)
- Note: The unbalanced sample of macroeconomic variables spans from 1994Q1 to 2015Q3.

### SDDS subscription, GDDS participation, and EMBIG coverage (Table A.2) — selected rows preserved exactly
- Argentina*: Date of SDDS subscription 1996Q3; Date of SDDS compliance 1999Q4; EMBIG coverage 1994Q1-2015Q3
- Belarus: Date of SDDS subscription 2004Q4; Date of SDDS compliance 2004Q4; EMBIG coverage 2010Q3-2015Q3
- Belize: Date of SDDS subscription (blank); Date of SDDS compliance 2006Q3; EMBIG coverage 2007Q1-2015Q3
- Bolivia: Date of SDDS subscription (blank); Date of SDDS compliance 2000Q4; EMBIG coverage 2012Q4-2015Q3
- Brazil*: Date of SDDS subscription 2001Q1; Date of SDDS compliance 2001Q1; EMBIG coverage 1994Q2-2015Q3
- Chile*: Date of SDDS subscription 1996Q2; Date of SDDS compliance 2000Q1; EMBIG coverage 1999Q2-2015Q3
- China*: Date of SDDS subscription 2015Q3; Date of SDDS compliance 2015Q3; Date of GDDS participation 2002Q2; EMBIG coverage 1994Q1-2015Q3
- Colombia*: Date of SDDS subscription 1996Q2; Date of SDDS compliance 2000Q2; EMBIG coverage 1997Q1-2015Q3
- Côte d'Ivoire*: Date of SDDS subscription (blank); Date of SDDS compliance 2000Q2; EMBIG coverage 1998Q2-2015Q3
- Croatia*: Date of SDDS subscription 1996Q2; Date of SDDS compliance 2001Q1; EMBIG coverage 1996Q3-2015Q3
- Ecuador*: Date of SDDS subscription 1998Q1; Date of SDDS compliance 2000Q3; EMBIG coverage 1995Q1-2015Q3
- Egypt*: Date of SDDS subscription 2005Q1; Date of SDDS compliance 2005Q1; EMBIG coverage 2001Q3-2015Q3
- El Salvador: Date of SDDS subscription 1998Q2; Date of SDDS compliance 1999Q4; EMBIG coverage 2002Q2-2015Q3
- Gabon: Date of SDDS subscription (blank); Date of SDDS compliance 2002Q4; EMBIG coverage 2007Q4-2015Q3
- Ghana: Date of SDDS subscription (blank); Date of SDDS compliance 2005Q3; EMBIG coverage 2007Q4-2015Q3
- Guatemala: Date of SDDS subscription (blank); Date of SDDS compliance 2004Q4; EMBIG coverage 2012Q2-2015Q3
- Honduras: Date of SDDS subscription (blank); Date of SDDS compliance 2005Q3; EMBIG coverage 2013Q2-2015Q3
- Hungary*: Date of SDDS subscription 1996Q2; Date of SDDS compliance 2000Q1; EMBIG coverage 1999Q1-2015Q3
- India: Date of SDDS subscription 1996Q4; Date of SDDS compliance 2001Q4; EMBIG coverage 2012Q4-2015Q3
- Indonesia: Date of SDDS subscription 1996Q3; Date of SDDS compliance 2000Q2; EMBIG coverage 2004Q2-2015Q3
- Iraq*: Date of SDDS subscription (blank); Date of SDDS compliance 2009Q4; EMBIG coverage 2006Q1-2015Q3
- Jamaica: Date of SDDS subscription (blank); Date of SDDS compliance 2003Q1; EMBIG coverage 2007Q4-2015Q3
- Jordan: Date of SDDS subscription 2010Q1; Date of SDDS compliance 2010Q1; Date of GDDS participation 2000Q3; EMBIG coverage 2011Q2-2015Q3
- Kazakhstan: Date of SDDS subscription 2003Q1; Date of SDDS compliance 2003Q1; Date of GDDS participation 2001Q1; EMBIG coverage 2007Q2-2015Q3
- Latvia: Date of SDDS subscription 1996Q3; Date of SDDS compliance 1999Q3; EMBIG coverage 2012Q3-2015Q3
- Lebanon*: Date of SDDS subscription (blank); Date of SDDS compliance 2003Q1; EMBIG coverage 1998Q2-2015Q3
- Lithuania: Date of SDDS subscription 1996Q2; Date of SDDS compliance 1999Q3; EMBIG coverage 2009Q4-2015Q3
- Malaysia*: Date of SDDS subscription 1996Q3; Date of SDDS compliance 2000Q3; EMBIG coverage 1996Q4-2015Q3
- Mexico*: Date of SDDS subscription 1996Q3; Date of SDDS compliance 2000Q2; EMBIG coverage 1993Q4-2015Q3
- Mongolia: Date of SDDS subscription (blank); Date of SDDS compliance 2000Q3; EMBIG coverage 2012Q2-2015Q3
- Morocco*: Date of SDDS subscription 2005Q4; Date of SDDS compliance 20005Q4; EMBIG coverage 1997Q4-2015Q3
- Namibia: Date of SDDS subscription (blank); Date of SDDS compliance 2002Q4; EMBIG coverage 2011Q4-2015Q3
- Nigeria*: Date of SDDS subscription (blank); Date of SDDS compliance 2003Q2; EMBIG coverage 1993Q4-2015Q3
- Pakistan: Date of SDDS subscription (blank); Date of SDDS compliance 2003Q4; EMBIG coverage 2004Q2-2015Q3
- Panama*: Date of SDDS subscription (blank); Date of SDDS compliance 2000Q4; EMBIG coverage 1996Q3-2015Q3
- Paraguay: Date of SDDS subscription (blank); Date of SDDS compliance 2001Q3; EMBIG coverage 2013Q1-2015Q3
- Peru*: Date of SDDS subscription 1996Q3; Date of SDDS compliance 1999Q3; EMBIG coverage 1997Q1-2015Q3
- Philippines*: Date of SDDS subscription 1996Q3; Date of SDDS compliance 2001Q1; EMBIG coverage 1993Q4-2015Q3
- Russia*: Date of SDDS subscription 2005Q1; Date of SDDS compliance 2005Q1; EMBIG coverage 1997Q4-2015Q3
- Senegal: Date of SDDS subscription (blank); Date of SDDS compliance 2001Q3; EMBIG coverage 2011Q2-2015Q3
- Slovak Republic: Date of SDDS subscription 1996Q3; Date of SDDS compliance 1999Q4; EMBIG coverage 2013Q3-2015Q3
- South Africa*: Date of SDDS subscription 1996Q3; Date of SDDS compliance 2000Q3; EMBIG coverage 1994Q4-2015Q3
- Sri Lanka: Date of SDDS subscription (blank); Date of SDDS compliance 2000Q3; EMBIG coverage 2007Q4-2015Q3
- Thailand*: Date of SDDS subscription 1996Q3; Date of SDDS compliance 2000Q2; EMBIG coverage 1997Q3-2015Q3
- Trinidad and Tobago: Date of SDDS subscription (blank); Date of SDDS compliance 2004Q3; EMBIG coverage 2007Q2-2015Q3
- Tunisia: Date of SDDS subscription 2001Q2; Date of SDDS compliance 2001Q2; EMBIG coverage 2002Q2-2015Q3
- Turkey*: Date of SDDS subscription 1996Q3; Date of SDDS compliance 2001Q2; EMBIG coverage 1996Q1-2015Q3
- Ukraine*: Date of SDDS subscription 2003Q1; Date of SDDS compliance 2003Q1; EMBIG coverage 2000Q2-2015Q3
- Uruguay*: Date of SDDS subscription 2004Q1; Date of SDDS compliance 2004Q1; EMBIG coverage 2001Q2-2015Q3
- Venezuela*: Date of SDDS subscription (blank); Date of SDDS compliance 2001Q1; EMBIG coverage 1993Q4-2015Q3
- Vietnam: Date of SDDS subscription (blank); Date of SDDS compliance 2003Q3; EMBIG coverage 2005Q4-2015Q3
- Zambia: Date of SDDS subscription (blank); Date of SDDS compliance 2002Q4; EMBIG coverage 2012Q4-2015Q3
- Note: All countries are used in the first event study of estimating Eq. (1). Countries with * are used in the second event study of estimating Eq. (2), because the EMBIG spreads are continuously observed before and after events only for these countries.

### Assigning numerical values to Fitch’s letter credit ratings (Table A.3)
- AAA: 19
- AA+: 18
- AA: 17
- AA-: 16
- A+: 15
- A: 14
- A-: 13
- BBB+: 12
- BBB: 11
- BBB-: 10
- BB+: 9
- BB: 8
- BB-: 7
- B+: 6
- B: 5
- B-: 4
- CCC: 3
- CC: 2
- C: 1
- D (Default): 0

*Source: wp1774 - 55. Peterson Institute, 1998 (pdf: wp1774 - 55. Peterson Institute, 1998).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2017/wp1774.pdf_