## _wp10168 — Appendix I and empirical results summary

## Source details

**Canonical URL:** [_wp10168 — Appendix I and empirical results summary](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10168.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10168.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10168.pdf.json)

---

### Model context, predictions, and mechanisms
- Context and dataset:
  - Focus: currency composition of domestic sovereign debt in emerging market (EM) countries; emphasis on domestic debt to disentangle issuance location from currency and to capture sovereign and domestic investor incentives.
  - Dataset for empirical tests: central government debt in 22 EM countries over 1980–2005 constructed by Jeanne and Guscina (2006).
- Key modelling assumptions and setup:
  - Two-period model; households are the only buyers of financial assets; government bonds are the only financial assets; no investment; no default.
  - Government issues domestic-currency or foreign-currency bonds. Foreign nominal interest rate r* exogenous; domestic nominal rate i determined in period 1.
  - Fiscal rules determine taxes and transfers as function of output; sovereign debt manager chooses currency composition to minimize expected debt service cost less γ/2 Var.
  - Assumption 1: Cov(ŝ, p̂) > 0 (positive covariance between exchange rate and price change).
  - Assumption 2: Cov(ŝ, ε̂) < 0 (exchange rate countercyclicality: exchange rate depreciates during recessions).
- Main theoretical predictions (small-shock approximations summarized):
  - When households and the sovereign debt manager are risk averse and there is no risk of default, the equilibrium currency composition of sovereign debt will be tilted toward instruments that represent a better hedge against exchange rate, inflation and/or output shocks.
  - If the real exchange rate is less volatile than domestic inflation, sovereign domestic debt will be, on average, more dollarized.
    - Intuition: households and the debt manager dislike volatility in yields; households prefer foreign-currency bonds when real exchange rate volatility (determining foreign bond return in consumption units) is lower than inflation volatility (determining local bond real returns).
  - If the exchange rate is countercyclical (β̂ε < 0), a more procyclical fiscal policy (μ > 0) increases the incentive for households and the sovereign debt manager to hold/issue foreign-currency bonds; thus, procyclical fiscal policy combined with exchange-rate countercyclicality raises dollarization of domestic sovereign debt.
  - Two-way link: greater sovereign dollar debt can reinforce fiscal procyclicality (real burden rises in bad times), and conversely procyclical fiscal policy can lead to more dollarization when the exchange rate is countercyclical.
- Equilibrium condition and comparative statics (verbatim structure preserved in source):
  - Household demand and government supply equate (A0 δ = D0 δ′ and A0 = D0 under assumptions).
  - Equilibrium share combines financial hedging (MVP), non-financial income hedging (depends on μ, τ, β̂ε), and debt-to-output ratio (dy).
  - Proposition: If β̂ε < 0, μ > 0 and τ < 1/2, the non-financial income hedging term is positive and δ** increases in both μ and (−β̂ε), and decreases in sovereign debt-to-output ratio.

### Empirical specification, variables and data construction
- Dependent variable:
  - Dollarization ratio = share of foreign-currency denominated or indexed debt (tradable/securitized and loans) issued by central governments in domestic markets; source: Jeanne-Guscina EM Debt Database (2006).
- Key explanatory variables:
  - MVP (minimum variance portfolio) — computed as in equation (15); MVP10 uses rolling 10-year sample of quarterly inflation and real exchange rate growth (CPI and nominal exchange rate from IFS).
  - Exchange rate countercyclicality (E) — estimated as (−1) times rolling OLS beta of nominal exchange rate growth on real GDP growth over 15-year windows (annual data).
  - Fiscal procyclicality (F) — estimated as OLS coefficient of growth rate of real government consumption on growth rate of real GDP over moving 15-year window (annual data).
  - Debt-to-GDP ratio (D) — domestically issued sovereign debt-to-GDP from Jeanne-Guscina EM Debt Database.
  - Interaction term E × F included to capture amplification of exchange-rate countercyclicality by fiscal procyclicality.
  - Controls: size of economy (log real GDP), institutional quality (Polity2), regional dummies.
- Estimation strategy:
  - Tobit (censored) model used to account for limited dependent variable and many zero observations for dollarization ratio.
  - Correct marginal effects for interaction term derived analytically for Tobit (Appendix II).

### Main empirical results (pooled sample of 22 EM countries, 1980–2005)
- Stylized findings (consistent with theoretical predictions):
  - MVP coefficient positive and statistically significant at 1 percent level in benchmark regressions.
  - Exchange rate countercyclicality associated with higher dollarization of domestic sovereign debt.
  - Higher domestic debt-to-GDP ratio associated with lower dollarization of domestic sovereign debt.
  - Fiscal procyclicality coefficient alone often not statistically significant; however, the interaction term E × F is positive and statistically significant — fiscal procyclicality amplifies the effect of exchange-rate countercyclicality on dollarization.
  - Countries with better political institutions are more likely to borrow in local currency.
- Selected reported coefficient values from Table 3 (preserved exactly as in source; dependent variable: Share of Dollar Debt in Domestic Debt; columns (1)–(7) reported in source):
  - Minimum Variance Portfolio: 0.578***; 0.575***; 0.594***; 0.568***; 0.507***; 0.510***; 0.253***
  - Countercyclicality of Exchange Rate: 0.193***; 0.190***; 0.166***; 0.061***; 0.107***; 0.107***; 0.104***
  - Domestic Debt-to-GDP: -0.166**; -0.147**; -0.168**; -0.177**; -0.247***
  - Interaction Term (E × F): 0.024***; 0.030**; 0.030**; 0.021**
  - Size of Economy: -5.471***; -5.448***; -3.573***
  - Institutional Quality: -0.067; -1.349***
  - Regional dummies: Latin America dummy 26.957***; Asian dummy -27.51***
  - Summary statistics (example column entries preserved exactly): Observations: 392; LogLik -1145.9; Pseudo R2 0.047 (varies by column).
- Robustness results (selected values from Table 4, preserved exactly):
  - MVP: 0.253***; 0.238***; 0.245***; 0.233***; 0.268***; 0.255***
  - Countercyclicality of Exchange Rate: 0.104***; 0.105***; 0.103***; 0.106***; 0.096***; 0.105***
  - Interaction Effect (E*F): 0.021**; 0.021**; 0.021**; 0.029***; 0.027***; 0.021***
  - Domestic Debt–to-GDP Ratio: -0.247***; -0.238***; -0.234***; -0.212**; -0.263***; -0.247***
  - Institutional Quality: -1.349***; -1.374***; -1.322***; -1.053***; -1.299***; -1.368***
  - Size of Economy: -3.573***; -3.668***; -3.706***; -4.701***; -3.728***; -3.611**
  - Additional controls: currency crisis dummy 10.963* (in specified columns); debt crisis dummy 4.451 (in specified column).
  - Robustness summary entries (examples preserved): Observations 390; LogLik -1032.2; reported Pseudo R2 ranges 0.138 to 0.154 across specifications.

### Appendix I — derivations of theoretical expressions (summary of contents preserved verbatim)
- Appendix headings and structure:
  - A. The Household’s Optimization Problem:
    - Source reproduces nominal and real expressions for household wealth in period 1, first-order linear approximations for small shocks, variable redefinitions, expressions for expected value and variance of real consumption, first-order condition for household choice, and the optimal share of household financial wealth held in foreign currency bonds (presented in the source as algebraic expressions reproduced verbatim).
  - B. The Debt Manager’s Optimization Problem:
    - Source reproduces government nominal budget constraint, division by price level, linear approximations, real budget constraint, definitions for parameters, expected value and variance of fiscal buffer, first-order condition from debt manager’s objective and the optimal share of dollar debt (presented as algebraic expressions reproduced verbatim).
  - C. Domestic Bond Market Clearing Condition:
    - Source reproduces household demand for domestic dollar debt, government supply of domestic dollar debt, equilibrium condition for market-clearing interest rate, expression for equilibrium share δ**, and alternative re-writing combining MVP, fiscal parameters, and debt-to-output ratio (expressions reproduced verbatim).
- Note: Equations and expressions in Appendix I are reproduced verbatim in the source; algebraic lines in the PDF are preserved in their original form in the Appendix I section.

### Appendix II — corrected interaction effect in Tobit models (summary)
- Problem statement:
  - The magnitude of the interaction effect in Tobit models does not equal the marginal effect of the interaction term; standard software does not calculate its value or statistical significance directly.
- Model and re-expression:
  - Tobit panel data model considered with censoring: yit = max(0, xitβ + uit), uit | x ~ Normal(0, σ2) (as in source).
  - Index re-expressed to include interacting variables x1 (procyclicality), x2 (countercyclicality), coefficients β1 and β2, controls X and coefficients Β (expression as in source).
- Marginal effects derived (expressions reproduced verbatim from source):
  - Marginal effect of a change in procyclicality (x1) on conditional expectation E(y | x) — expression provided in Appendix II and reduced in the source to a form involving β1, β2, σ and Φ.
  - Marginal effect of countercyclicality (x2) analogous.
  - Marginal effect of the interaction term derived explicitly (complex expression involving Φ and φ functions and β coefficients), reproduced verbatim in source.

### Empirical conclusions and policy implications (as stated in source)
- Empirical confirmations (22 EMs, 1980–2005):
  - Countries where the real exchange rate is less volatile than domestic inflation have, on average, a more dollarized domestic sovereign debt.
  - Countries where the exchange rate is countercyclical have, on average, a more dollarized domestic sovereign debt.
  - In countries with exchange-rate countercyclicality, more procyclical fiscal policy leads, on average, to more dollarized domestic sovereign debt.
  - Countries with smaller public debt-to-GDP ratios tend to have higher levels of foreign-currency domestic sovereign debt.
- Policy implications highlighted in the source:
  - A combination of flexible exchange rate regime and inflation targeting would reduce the relative volatility of inflation to exchange rate changes and hence reduce the share of foreign-currency denominated instruments in domestic debt markets.
  - If the exchange rate is highly countercyclical, reducing fiscal procyclicality may help reduce incentives to dollarize domestic sovereign debt.
  - Deepening domestic debt markets (increasing domestic debt-to-GDP) is associated with lower dollarization of sovereign debt.

*Source: _wp10168 — Appendix I and Sections III–V, with empirical results and Appendices reproduced from the supplied content.*

### Appendix I. Derivations of Theoretical Expressions ...............................................................36

### Appendix I. Derivations of Theoretical Expressions

### A. The Household’s Optimization Problem
- Heading listed in the source as: "A. The Household’s Optimization Problem:"
- No derivations or equations for this subsection are present in the supplied content.

### B. The Debt Manager’s Optimization Problem
- Heading listed in the source as: "B. The Debt Manager’s Optimization Problem:"
- No derivations or equations for this subsection are present in the supplied content.

### C. Domestic Bond Market Clearing Condition
- Heading listed in the source as: "C. Domestic Bond Market Clearing Condition:"
- No derivations or equations for this subsection are present in the supplied content.

### Model context, predictions, and empirical testing (extracted from the supplied text)
- Context:
  - The paper focuses on the currency composition of domestic sovereign debt in emerging market (EM) countries, arguing advantages of studying domestic debt (disentangling issuance location from currency, focusing on sovereign and domestic investor incentives, and greater cross-country heterogeneity).
  - Dataset used for empirical tests: central government debt in 22 EM countries over 1980–2005 constructed by Jeanne and Guscina (2006).

- Main theoretical predictions (as presented in the supplied content):
  - When households and the sovereign debt manager are risk averse and there is no risk of default, the equilibrium currency composition of sovereign debt will be tilted toward instruments that represent a better hedge against exchange rate, inflation and/or output shocks.
  - Countries where the real exchange rate is less volatile than the domestic inflation rate would have, on average, a more dollarized domestic sovereign debt.
    - Rationale: Households and the debt manager dislike volatility in yields; households prefer foreign currency bonds if the real exchange rate (determining return volatility of foreign currency bonds in local consumption units) is less volatile than domestic inflation (determining volatility of real return on local currency bonds). The debt manager similarly dislikes volatility in real sovereign bond yields.
  - In countries where the exchange rate is countercyclical (exchange rate depreciates during recessions), a more procyclical fiscal policy (expansionary in good times and contractionary in bad times) would lead, on average, to a more dollarized domestic sovereign debt.
    - Rationale: Households and the sovereign debt manager value assets that hedge output shocks affecting non-financial income and sovereign tax revenues. Foreign currency bonds represent a better hedge against output shocks when the exchange rate is countercyclical. A more procyclical fiscal policy increases procyclicality of households’ non-financial incomes and thus incentives to hold foreign currency bonds; strong fiscal procyclicality can also create incentives for the debt manager to issue more foreign currency debt, reinforcing households’ preferences.
  - Two-way link observation: While greater sovereign dollar debt can reinforce procyclical fiscal policy (because real debt burden rises more in bad times), the converse holds when the exchange rate is countercyclical: procyclical fiscal policy may lead to more dollarized sovereign debt.

- Empirical findings summarized in the supplied content:
  - The empirical analysis using the 22 EM countries over 1980–2005 supports the model predictions.
  - Specific empirical confirmations reported in the supplied content:
    - Countries with relatively high inflation uncertainty tend to have higher levels of dollarization of sovereign debt.
    - In countries where the exchange rate is countercyclical, a more procyclical fiscal policy tends to be associated with more dollarized domestic sovereign debt.
  - The empirical results are described as robust under a range of specifications and estimation methods.

- Ancillary points and literature connections included in the supplied content:
  - Prior literature emphasized external sovereign debt and lender preferences (the "original sin" literature), while this paper emphasizes domestic debt and borrower/investor incentives.
  - Financial dollarization persistence relates to the relative volatility of inflation versus the real exchange rate (cited similarity with Ize and Levy-Yeyati (2003) result).
  - Historical EM crises (Mexico, 1994; Russia, 1998; Brazil, 1998) illustrate the role of domestic sovereign debt in distress episodes.

*Source: Appendix I heading and accompanying introductory and summary text from the supplied content.*

### Section IV presents the results of the empirical analysis, and Section V concludes the paper.

### _wp10168 - Section IV presents the results of the empirical analysis, and Section V concludes the paper.

### II. Cyclical patterns in macroeconomic and policy variables in EMs
- Key empirical regularities assumed:
  - Nominal exchange rate changes and domestic price changes have positive covariance.
  - Nominal exchange rate depreciations (appreciations) tend to be associated with downturns (upturns) in economic activity — labeled exchange rate countercyclicality.
  - Fiscal policies tend to be expansionary (contractionary) during upturns (downturns) — labeled fiscal procyclicality.
- Supporting literature highlights:
  - Fiscal procyclicality documented extensively in EM and developing countries (e.g., Riascos and Vegh (2003); Ilzetzki and Vegh (2008)).
  - Theoretical channels include incomplete financial markets, political economy (“voracity effect”), and large fluctuations in the tax base (Talvi and Vegh (2005)).

### III. The model — setup and mechanisms
- Structure:
  - Two-period model; households are only buyers of financial assets; government bonds are the only financial assets; no investment; no default.
  - Government issues domestic-currency or foreign-currency bonds. Foreign nominal interest rate *r is exogenous; domestic nominal rate *i determined in period 1.
  - Fiscal rules determine taxes and transfers as function of output; sovereign debt manager chooses currency composition to minimize debt service cost and its variability.
- Timeline:
  - Period 0: initial household wealth A00 and government debt D00; sovereign debt manager chooses composition of D0; households choose composition of A0.
  - Period 1: output shock realized; exchange rate and inflation shocks observed; domestic bond market rate *i determined; sovereign debt repaid; consumption occurs.
- Household problem and assumptions:
  - Nominal consumption is geometric average of tradable and non-tradable consumption with tradable share α (equation (1)).
  - Aggregate price level defined by tradable and non-tradable prices (equation (2)).
  - Price change p̂ has mean E(p̂) and variance V(p̂); unexpected nominal exchange rate depreciation ŝ has mean E(ŝ) and variance V(ŝ).
  - Assumption 1: Cov(ŝ, p̂) > 0 (positive covariance).
  - Households’ non-financial income shock ε̂: E(ε̂) = 0, Var(ε̂) = σε̂2; transfers equal μ times fiscal revenue overperformance/underperformance; μ > 0 implies procyclical transfers.
  - Assumption 2: Cov(ŝ, ε̂) < 0 (negative contemporaneous correlation between exchange rate and output).
- Optimal household foreign-currency share (small-shock approximation) (equation (5)):
  - Components: interest rate differential; financial portfolio hedging (MVP) — weighted average of pass-through to tradable and non-tradable prices; non-financial income hedging (depends on β̂ε, μ, τ).
  - Definitions:
    - β̂T = Cov(ŝ, p̂T)/V(ŝ) = exchange rate pass-through in tradable sector.
    - β̂NT = Cov(ŝ, p̂NT)/V(ŝ) = exchange rate pass-through in non-tradable sector.
    - β̂ε = Cov(ŝ, ε̂)/V(ŝ) = correlation between exchange rate changes and output.

### III.B Fiscal policy and sovereign debt management
- Government fiscal balance (equation (6)): F1 = τY1 + τ ε̂ + μ τ Y1 ε̂ − τY1 ??? (expressed in paper as 1F = ˆˆ1Y Y τ ε μ τ ε — see text for exact algebraic expression).
- Value of debt at end of period 1 (equation (7)): D1 = (1)Drsi ˆ(1) δ terms (exact expression in source).
- Government’s budget constraint (equation (8)): G1 + F1 ≤ D1.
- Debt manager objective (equation (10)): maximize expected real fiscal buffer less γ/2 Var; chooses δ′ (share foreign currency debt).
- Optimal share of foreign-currency government debt (small-shock approximation) (equation (11)):
  - Components: financial portfolio hedging; interest rate differential; fiscal hedging (depends on μ, τ, β̂ε, and exchange rate countercyclicality).
  - Key comparative statics:
    - Supply of foreign-currency bonds decreases in interest rate differential.
    - Financial hedging increases incentive to issue foreign-currency debt.
    - Fiscal hedging generally reduces incentive to issue foreign-currency debt unless μ > 1 (very strong procyclicality), in which case fiscal rule can make foreign-currency debt more attractive.

### III.C Equilibrium in domestic bond market and main proposition
- Equilibrium condition: household demand for dollar debt equals government supply (A0 δ = D0 δ′ and A0 = D0 under model assumptions).
- Equilibrium share (small-shock approximation) (equation (12)) combines:
  - Financial portfolio hedging (MVP),
  - Non-financial income hedging (depends on μ, τ, β̂ε),
  - Debt-to-output ratio (dy).
- Proposition:
  - If exchange rate is countercyclical (β̂ε < 0), fiscal policy is procyclical (μ > 0) and τ < 1/2, then the non-financial income hedging term in (12) is positive and δ** increases in both μ and (-β̂ε), and decreases in sovereign debt-to-output ratio.
  - Prediction: countries with exchange-rate countercyclicality and fiscal procyclicality have, on average, more dollarized domestic sovereign debt. Higher sovereign debt-to-output ratio implies lower share of dollar debt.

### IV. Testing empirical implications — stylized facts and empirical specification
- Stylized facts:
  - Domestic sovereign debt structures across EM and developing countries are characterized by significant heterogeneity.
  - Dollarization of domestic sovereign debt and financial dollarization are correlated.
  - Foreign-currency debt appears more commonly used than inflation-linked debt; there is a threshold of inflation inducing CPI-indexed debt.
- Benchmark empirical specification starts from equation (12) and approximates via Taylor expansion to estimate direct effects.
- Variables included:
  - MVP (minimum variance portfolio) — computed as in equation (15), MVP10 uses a rolling 10-year sample.
  - Exchange rate countercyclicality (E) — estimated as (-1) times rolling OLS beta of nominal exchange rate growth on real GDP growth over 15-year windows.
  - Fiscal procyclicality (F) — estimated as OLS coefficient of growth rate of real government consumption on growth rate of real GDP over moving 15-year window.
  - Debt-to-GDP ratio (D) — domestically issued sovereign debt-to-GDP from Jeanne-Guscina EM Debt Database.
  - Interaction term E × F included to capture amplification effect.
  - Controls: size of economy (log real GDP), institutional quality (Polity2), regional dummies.
- Estimation technique:
  - Tobit model used to account for limited dependent variable and many zero observations for dollarization ratio.
  - Correct marginal effects for interaction term derived analytically for Tobit (Appendix II referenced).

### IV.D Data notes
- Dependent variable: dollarization ratio = share of foreign-currency denominated or indexed debt (tradable/securitized and loans) issued by central governments in domestic markets.
- Dataset: Jeanne-Guscina EM Debt Database (2006) — annual data on central government debt structure in 22 emerging markets over 25 years.
- MVP computation:
  - CPI and nominal exchange rate (local currency per US dollar) from IFS.
  - Real exchange rate = nominal exchange rate / CPI.
  - Inflation and real exchange rate growth computed as log differences at quarterly frequency.
  - MVP10 computed with variances/covariances over rolling 10-year periods as in equation (15).
- Procyclicality and countercyclicality measures constructed using annual data with rolling 15-year OLS regressions; results typically positive for fiscal procyclicality and positive for exchange rate countercyclicality (after multiplying beta by -1).

### IV.E Main empirical results and robustness
- Sample: pooled sample of 22 emerging market countries over 1980–2005.
- Main findings (consistent with theoretical predictions):
  - MVP has expected positive coefficient; statistically significant at 1 percent level.
  - Exchange rate countercyclicality associated with higher dollarization of debt.
  - Higher debt-to-GDP ratio associated with lower dollarization of debt.
  - Fiscal procyclicality coefficient alone not statistically significant; however the interaction term between fiscal procyclicality and exchange rate countercyclicality is positive and statistically significant — fiscal procyclicality amplifies effect of exchange rate countercyclicality on dollarization.
  - Countries with better political institutions are more likely to borrow in local currency.
- Robustness checks:
  - Controlling for currency/debt crisis dummies or restricting sample to tranquil times does not alter main results.
  - Restricting sample to 1987–2005 yields very similar results.
  - Using one-year lags of fiscal procyclicality and exchange rate countercyclicality (and their interaction) produces consistent results.
  - Results robust to alternative measures of institutional quality (ICRG index, Kaufmann et al. regulatory quality index).

### V. Concluding remarks — substantive findings and policy implications
- Empirical confirmations (using 22 EMs, 1980–2005):
  - Countries where the real exchange rate is less volatile than the domestic inflation rate have, on average, a more dollarized domestic sovereign debt.
  - Countries where exchange rate is countercyclical (exchange rate depreciates during recessions) have, on average, a more dollarized domestic sovereign debt.
  - In countries with exchange-rate countercyclicality, more procyclical fiscal policy leads, on average, to more dollarized domestic sovereign debt.
  - Countries with smaller public debt-to-GDP ratios tend to have higher levels of foreign-currency domestic sovereign debt.
- Policy implications highlighted:
  - A combination of flexible exchange rate regime and inflation targeting would reduce the relative volatility of inflation to exchange rate changes and hence reduce the share of foreign-currency denominated instruments in domestic debt markets.
  - If the exchange rate is highly countercyclical, reducing fiscal procyclicality may help reduce incentives to dollarize domestic sovereign debt.
  - Deepening domestic debt markets (increasing domestic debt-to-GDP) is associated with lower dollarization of sovereign debt.

*Source: _wp10168 - Section IV presents the results of the empirical analysis, and Section V concludes the paper.*

### References

### References

### Key cited works
- Adler, Gustavo, 2008, “Original Sin and Procyclical Fiscal Policy: Two Sides of the Same Coin?” IMF Working Papers, 08/209, pp. 1–27.
- Alesina, Alberto, Silvia Ardagna, and Francesco Trebbi, 2006, “Who Adjusts and When? On the Political Economy of Reforms,” NBER Working Papers 12049.
- Bordo, Michael D., and Christopher M. Meissner, 2005, “The Role of Foreign Currency Debt in Financial Crises: 1880–1913 vs. 1972–1997,” NBER Working Paper No. 11897.
- Bordo, Michael D., and Angela Redish, 2005, “How ‘Original Sin’ Was Overcome: The Evolution of External Debt Denominated in Domestic Currencies in the United States and the British Dominions 1800–2000,” in Other People’s Money, pp. 122–153.
- Borensztein, Eduardo, Marcos Chamon, Olivier Jeanne, Paolo Mauro, and Jeromin Zettelmeyer, 2004, “Sovereign Debt Structure for Crisis Prevention”, IMF Occasional Paper No.237.
- Calvo, Guillermo, 1980, “Tax-Financed Government Spending in Neoclassical Model with Sticky Wages and Rational Expectations,” Journal of Economic Dynamics and Control, February 1980.
- Calvo, Guillermo, and Pablo Guidotti, 1990, “Indexation and Maturity of Government Bonds: An Exploratory Model,” in Capital Markets and Debt Management.
- Chang, Roberto and Velasco, Andrés, 1997, “Financial Fragility and the Exchange Rate Regime,” FRB Atlanta Working Paper No. 97–16 and NBER Working Paper No. 6469.
- Chamon, Marcos, 2003, "Why Can't Developing Countries Borrow from Abroad in Their Currency?" Mimeo.
- Claessens, Stijn, Daniela Klingebiel, and Sergio Schmukler, 2007, "Government Bonds in Domestic and Foreign Currency: The Role of Macroeconomic and Institutional Factors," Review of International Economics, 15:2, pp. 370–413.
- Eichengreen, Barry, Ricardo Hausmann, and Ugo Panizza, 2003, “Currency Mismatches, Debt Intolerance and Original Sin: Why They are Not the Same and Why it Matters,” NBER No.10036.
- Guscina, Anastasia, 2008, “Impact of Macroeconomic, Political, and Institutional Factors on the Structure of Government Debt in Emerging Market Countries,” IMF Working Paper 08/205.
- Hausmann, Ricardo and Ugo Panizza, 2003, “The Mystery of Original Sin,” IMF Research Paper No. 2003164B.
- Ilzetzki, Ethan and Carlos Vegh, 2008, “Procyclical Fiscal Policy in Developing Countries: Truth of Fiction?,” NBER Working Papers 14191.
- IMF and World Bank, 2003, Guidelines for Public Debt Management: Accompanying Document and Selected Case Studies.
- Ize, Alain, 2005, “Financial Dollarization Equilibria: A Framework for Policy Analysis,” IMF Working Paper 05/186.
- Ize, Alain, and Eduardo Levy-Yeyati, 2003, “Financial Dollarization,” Journal of International Economics, Vol. 59, pp. 32347.
- Jaggers, Keith and Ted Robert Gurr, 1995, “Tracking Democracy’s Third Wave with the Polity III Database,” Journal of Peace Research, Vol. 32, No. 4, pp. 469–482.
- Jeanne, Olivier, 2000, “Foreign Currency Debt and the Global Financial Architecture,” European Economic Review, Vol. 44, No. 46 (May), 719–727.
- Jeanne, Olivier, 2004, “Debt Maturity and the International Financial Architecture,” IMF Working Paper 04/137.
- Jeanne, Olivier, and Anastasia Guscina, 2006, “Government Debt in Emerging Market Countries: A New Data Set,” IMF Working Paper 06/98.
- Kaminsky, G., C. Reinhart, and C. Vegh, 2004, “When It Rains, It Pours: Procyclical Capital Flows and Macroeconomic Policies”, NBER Working Paper No. 10780.
- Kaufmann, Daniel, Aart Kray, and Massimo Matruzzi, 2005, “Governance Matters IV: Governance Indicators for 1996–2004,” Policy Research Working Paper No. 3630.
- Lane, Philip and Aaron Tornell, 1999, "The Voracity Effect", American Economic Review, 89, 22–46, March 1999.
- Lane, Philip R. and Shambaugh, Jay C., 2010, “The Long or Short of It: Determinants of Foreign Currency Exposure in External Balance Sheets", Journal of International Economics, Volume 80, Issue 1, pp. 33–44.
- Marshall, M. and K. Jaggers, 2002, “Polity IV Project: Political Regime Characteristics and Transitions, 1800–2002,” http://www.bsos.umd.edu/cidcm/polity/
- Mehl, Arnaud and Julien Reynaud, 2008, “The Domestic Debt Structures in Emerging Markets: New Empirical Evidence,” Documents de travail du Centre d’Economie de la Sorbonne, Universite Pantheone-Sorbonne (Paris 1).
- Norton, E., H. Wang, and C. Ai, 2004, “Computing Interaction Effects and Standard Errors in Logit and Probit Models,” The Stata Journal, Vol. 4, No. 2, pp. 154–167.
- Perotti, E., and Volpin, P., 2007, “Politics, Investor Protection and Competition,” Amsterdam University Business School Working Paper.
- Rajan R., and L. Zingales, 2003, “The Great Reversals: The Politics of Financial Development in the Twentieth Century,” Journal of Financial Economics, Vol. 69 (July) pp. 5–50.
- Riascos, Alvaro and Carlos Vegh, 2003, “Procyclical Fiscal Policy and Developing Countries: The Role of Incomplete Markets” (mimeo).
- Reinhart, C., and K. Rogoff, 2008, “The Forgotten History of Domestic Debt,” NBER Working Paper No. 13946.
- Talvi, E., and C. Vegh, 2005, “Tax Base Variability and Procyclical Fiscal Policy in Developing Countries,” Journal of Development Economics, Vol.78, No.1, pp.156–90.
- Tirole, Jean, 2002, Financial Crises, Liquidity and the International Monetary System (Princeton: Princeton University Press).
- Tornell, Aaron and Andres Velasco, 2000, “Fixed Versus Flexible Exchange Rates: Which Provides More Fiscal Discipline?” Journal of Monetary Economics, Vol. 45, Issue 2, April 2000, pp. 399–436.
- Velasco, Andres, 2000, “Debt and Deficits under Fragmented Fiscal Policymaking,” Journal of Public Economics, Vol. 76, Issue 1, April 2000, pp. 105–125.
- Wheeler, G., 1997, “New Zealand’s Experience with Autonomous Sovereign Debt Management,” Risk Management Fore Sovereign Countries (Washington: International Monetary Fund).

### Other data and methodological sources cited
- World Economic Outlook (figures for exchange rate and fiscal series).
- International Financial Statistics (fiscal procyclicality data source).
- Jeanne-Guscina EM Debt Database (2006) (dollarization and domestic sovereign debt structure).
- Authors’ calculations and estimations (tables and regression results).
- Polity III/IV databases (political regime variables).

### Figures (captions and data sources)
- Figure 1: Exchange Rate Countercyclicality
  - Axes labeled: Real GDP Growth and Change in Nominal Exchange Rate.
  - Source: World Economic Outlook.

- Figure 2: Fiscal Procyclicality
  - Axes labeled: Real Government Consumption Growth and Real GDP Growth.
  - Source: World Economic Outlook and International Financial Statistics.

- Figure 3: Dollarization of Sovereign Debt and Dollarization of Bank Loans & Deposits
  - Country labels include ARG, CHL, HUN, ISR, MEX, POL, RUS, THA, TUR, URY, VEN and ARG, CHL, COL, DOM, HUN, ISR, MEX, POL, RUS, THA, TUR, URY, VEN.
  - Axes labeled: Average Dollarization of Domestic Sovereign Debt; Average Loan Dollarization; Average Deposit Dollarization (scale 0 to 1).
  - Source: Jeanne-Guscina EM Debt Database (2006).

- Figure 4a: Latin America–Domestic Sovereign Debt Structure–Selected Countries (In percent of total domestic debt, in USD)
  - Countries: ARGENTINA, BRAZIL, CHILE, MEXICO; years plotted include 1980 through 2004.
  - Debt categories: Foreign Currency debt, Indexed Debt, Local Currency Debt.
  - Source: Jeanne-Guscina EM Debt Database (2006).

- Figure 4b: EM Europe–Domestic Sovereign Debt Structures–Selected Countries (In percent of total domestic debt, in USD)
  - Countries: HUNGARY, POLAND, RUSSIA, TURKEY; years range 1980–2005 for various series.
  - Debt categories: Foreign Currency Debt, Indexed Debt, Local Currency Debt.
  - Source: Jeanne-Guscina EM Debt Database (2006).

- Figure 5: Effect of Countercyclicality of the Exchange Rate on Dollarization of Sovereign Debt, for Different Levels of Fiscal Procyclicality
  - Axes labeled: Effective parameter on Exchange rate countercyclicality and Procyclicality of Fiscal Policy (range -4 -2 0 2 4 and -0.2 to 0.6).
  - Source: Authors’ calculations.

### Tables (summary statistics and regression results)
- Table 1. Countercyclicality of the Exchange Rate: Summary Statistics (Country-level Obs, Std. Deviation, Min, Max)
  - Examples of entries (preserved exactly as in source):
    - Argentina: Obs 224, Std. Deviation 1.326, Min 4.272, Max 124.889
    - Brazil: Obs 212, Std. Deviation 3.701, Min 5.758, Max 77.031
    - Chile: Obs 22, Std. Deviation 5.017, Min 0.711, Max 3.733
    - (Other country rows follow in the table; Source: Authors’ calculations.)

- Table 2. Procyclicality of Fiscal Policy: Summary Statistics (Country-level Obs, Std. Deviation, Min, Max)
  - Examples of entries (preserved exactly as in source):
    - Argentina: Obs 221, Std. Deviation 1.182, Min 0.211, Max 4.239
    - Brazil: Obs 170, Std. Deviation 0.880, Min 0.636, Max 3.224
    - Chile: Obs 220, Std. Deviation 0.265, Min 0.339, Max 1.372
    - (Other country rows follow in the table; Source: Authors’ calculations.)

- Table 3. Results of Censored Tobit Estimation– Pooled Data: 1980–2005 (Coefficient and standard error)
  - Dependent Variable: Share of Dollar Debt in Domestic Debt
  - Selected reported coefficients (columns labeled (1) to (7); values preserved exactly):
    - Minimum Variance Portfolio: 0.578***; 0.575***; 0.594***; 0.568***; 0.507***; 0.510***; 0.253***
    - Portfolio: -0.071; -0.072; -0.071; -0.072; -0.053; -0.055; -0.041
    - Procyclicality of Fiscal Policy: 0.687; 0.986; -0.033; 0.19; 0.236; -0.361 (with associated standard errors as in source)
    - Countercyclicality of Exchange Rate: 0.193***; 0.190***; 0.166***; 0.061***; 0.107***; 0.107***; 0.104*** (with listed standard errors)
    - Domestic Debt-to-GDP: -0.166**; -0.147**; -0.168**; -0.177**; -0.247*** (with listed standard errors)
    - Interaction Term: 0.024***; 0.030**; 0.030**; 0.021** (with listed standard errors)
    - Size of Economy: -5.471***; -5.448***; -3.573*** (and other column entries)
    - Institutional Quality: -0.067; -1.349***
    - Latin America dummy: 26.957***
    - Asian dummy: -27.51***
    - Constants, Observations, LogLik, F, Pseudo R2 reported as in source:
      - Observations: 392; LogLik -1145.9; Pseudo R2 0.047 (varies by column)
  - Source: Authors’ estimations.

- Table 4. Robustness Checks for Censored Tobit Estimation: 1980–2005 (Coefficient and standard error)
  - Columns (1) Baseline specification through (6) w/ E regime dummy; selected reported coefficients preserved exactly:
    - Minimum Variance Portfolio: 0.253***; 0.238***; 0.245***; 0.233***; 0.268***; 0.255***
    - Procyclicality of Fiscal Policy: -0.361; -0.356; -0.436; -0.601; -0.398; -0.353
    - Countercyclicality of Exchange Rate: 0.104***; 0.105***; 0.103***; 0.106***; 0.096***; 0.105***
    - Interaction Effect (E*F): 0.021**; 0.021**; 0.021**; 0.029***; 0.027***; 0.021***
    - Domestic Debt–to-GDP Ratio: -0.247***; -0.238***; -0.234***; -0.212**; -0.263***; -0.247***
    - Size of Economy: -3.573***; -3.668***; -3.706***; -4.701***; -3.728***; -3.611**
    - Institutional Quality: -1.349***; -1.374***; -1.322***; -1.053***; -1.299***; -1.368***
    - Currency crisis dummy: 10.963* (in specified columns)
    - Debt crisis dummy: 4.451 (in specified column)
    - Latin America dummy: 26.957***; 26.468***; 25.810***; 20.860***; 25.605***; 26.790***
    - Asia dummy: -27.51***; -28.03***; -27.697***; -26.020***; -27.692***; -27.705***
    - Observations and LogLik by column preserved exactly (e.g., Observations 390; LogLik -1032.2; Pseudo R2 0.138 to 0.154).
  - Source: Authors’ estimations.

*Source: _wp10168 - References (PDF chapter/section).*

### APPENDIX I. DERIVATIONS OF THEORETICAL EXPRESSIONS

### APPENDIX I. DERIVATIONS OF THEORETICAL EXPRESSIONS

### A. The Household’s Optimization Problem
- Nominal value of household’s wealth in period 1 (as given):  
  (1) (1)(1) (1)11ArsiAYY ˆ ˆ (1) (1)(1) (1)11 … (expression as in source)
- Division by 1/P and rewriting yields (source expressions):  
  (1) (1)(1) (1)11 AP AY Y rsi PPPPP ˆ ˆ (1) (1)(1) (1)11 … (expression as in source)
- First-order linear approximation for sufficiently small shocks ˆs and ˆp (expression in square brackets approximated):  
  ***
  ˆ
  (1) (1)(1) (1)
  ˆˆ
  1(  )(1)()
  ˆ
  (1)
  rsi
  rsip
  p
  
  
   
           
  
- Variable redefinitions used in derivations:  
  1 1 1 A a P = , 0 0 0 A a P = , 1 1 1 Y y P =  
- Real expression for household wealth in period 1 (after substitutions and approximations):  
  ***
  11 10 00 ˆ ˆˆ 1(  )(1)()11 yy aarsip aa   (expression as in source)
- Further substitutions: 1 01 1(1 )/ l ay  = , (1) 1 01 1/ k ay  = lead to compact expression:  
  ***
  1011 ˆ ˆˆ 1(  )(1)()aarsipkl  (expression as in source)
- At optimum (11 ca = ) implying 1 c  , leading to:  
  ***
  011 ˆ ˆˆ 1(  )(1)()arsipkl  (expression as in source)
- Expected value and variance of real consumption:  
  - Expected value:  
         *** 101001 ˆ ˆˆ (1) ( )Ec akaarEsilEEp             (expression as in source)
  - Variance:  
                     2 ** 1011 2 22 *** 0111 ˆˆ ˆˆ ˆˆ Var2, ˆˆˆ ˆˆˆˆˆ caVslVpCovslp aV sl VlCovsCovsplCov p            (expression as in source)
- First-order condition from maximizing household expected utility 11 ()() 2 Ec Vc   and differentiating wrt *  yields the f.o.c.:  
   *2* 0001 ˆ ˆˆˆˆˆ ()()() (,) (,)0 o a ria E saV sCov s pl Cov s    (expression as in source)
- Optimal share of household financial wealth held in foreign currency bonds (as given):  
    * * 1 0 ˆ ˆ ˆ () , ˆˆ (, ) ˆˆ ˆ ()()() o rEsi Cov s Cov s p l aV sV sV s      (expression as in source)
- Alternative representation using ˆˆˆ 1 TNT pp p p  leads to:  
   * * 1 0 ˆ () ˆ ˆˆˆˆˆ ( , )(,)(, ) (1) ˆˆˆ ˆ ()()()() o TNT rEsi Cov s pCov s pCov s l aV sV sV sV s                  (expression as in source)
- Decomposition of ˆˆ (, ) ˆ Cov s p Vs (analogous to MVP term in Ize and Levy-Yeyati (2003)):  
   ˆˆ , ˆ ()(,) ˆˆ ˆˆˆ ()()() 2( , ) ˆˆ Cov s p VCove pp VsVVeCov e pp    (expression as in source)  
  and related equalities given in the text.

### B. The Debt Manager’s Optimization Problem
- Government nominal budget constraint (as given):  
    * 1110 ˆˆ ˆ 1(1)1(1)(1)GYYrsiD            (expression as in source)
- Division by 1/P and rewriting (as in source):  
   * 00 111 11101 ˆˆ ˆ 1(1)1(1)(1) DP GYY rsi PPPPP                (expression as in source)
- First-order linear approximation for small shocks ˆs and ˆp (expression in square brackets approximated):  
  ***
  ˆ
  (1) (1)(1) (1)
  ˆˆ
  1(  )(1)()
  ˆ
  (1)
  rsi
  rsip
  p
  
  
   
           
  
- Letting 1 1 1 G g P = , 0 0 0 D d P = and 1 1 1 Y y P = gives the budget constraint in real terms:  
    * 1011 ˆ ˆˆ 1(1)()gdk frsi p      (expression as in source)
- Definitions used: 1 01 / k dy  = and (1) 1 01 1 / f dy  =  
- Expected value and variance of fiscal buffer 1 g :  
  - Expected value:  
       * 110001 ˆ ˆˆ () (1)()( )Egkd d dr Esi fEEp           (expression as in source)
  - Variance (as given):  
             2 1011 222 0111 ˆˆ ˆˆˆˆ Var2, ˆˆˆ ˆˆˆˆˆ gdVf sVpCovf sp dV sfVf Cov sCov s pf Covp             (expression as in source)
- First-order condition from the debt manager’s objective (differentiating wrt  ):  
      2 * 001 ˆ ˆˆˆˆˆ ,,0d rE sidV sf Cov sCov s p     (expression as in source)
- Optimal share of dollar debt (as given):  
         * 1 0 ˆ ˆ ˆˆˆ , , ˆˆˆ rEsi Cov s pCov s f dV sV sV s       =     (expression as in source)

### C. Domestic Bond Market Clearing Condition
- Households’ demand for domestic dollar debt (as given):  
      * * 0010 ˆ () ˆˆˆ (1) ˆ () TNT rEsi aala Vs        (expression as in source)
- Government’s supply of domestic dollar debt (as given):  
        * 0010 ˆ ˆˆ ˆ (1) ˆ TNT rEsi ddfd Vs          (expression as in source)
- Equilibrium condition: since in equilibrium * 00 AD   = and 00 AD = , the market clearing interest rate r is implicitly given by:  
     * 11 0 ˆ ˆ () ˆ ()2 lf rEsi aV s       (expression as in source)
- Setting ***    , equilibrium share of foreign currency debt is:  
     11 ** ˆˆˆ (1) 2 TNT fl         (expression as in source)
- Alternative re-writing:  
    ** ˆ 2(1) 1 ˆˆ (1) 2 01 TNT dy                  (expression as in source)

*Italic note: Equations and expressions above are reproduced verbatim from the source content.*

### APPENDIX II. DERIVATIONS OF CORRECTED INTERACTION EFFECT
- Statement of issue: The magnitude of the interaction effect in Tobit models does not equal the marginal effect of the interaction term; its value and statistical significance is not calculated by standard software or derived in the literature.
- Tobit panel data model considered:  
  Ttuy ititit ....,,2,1),,0(maxβx , with u it | x ~ Normal(0, 2) (as in source)
- Re-expression of index xβ in terms of interacting variables 1 x (procyclicality of fiscal policy), 2 x (countercyclicality of the exchange rate), coefficients 1  and 2 , and controls X with coefficients :  
  uXxxxxuy 21122211 * xβ (expression as in source)
- Conditional expectation identity used (from Wooldridge (2005)) and its expression (as in source): long-form expression involving  and  functions and xβ terms (reproduced in source).
- Marginal effects on conditional expectation E(x | y) (expressions reproduced from source):
  - Marginal effect of a 1% change in procyclicality of fiscal policy (1 ): (long multiline expression as presented in source) concluding after cancellations with:  
      2121 1 1 | x x yE                xβ x (expression as in source)
  - Marginal effect of countercyclicality of the exchange rate:  
      1122 2 2 | x x yE                xβ x (expression as in source)
  - Marginal effect of the interaction term (as given):  
      1221211122 2 1 3 |                                xβxβ x xx x yE (expression as in source)

*Italic source attribution: Content reproduced from _wp10168 - APPENDIX I. DERIVATIONS OF THEORETICAL EXPRESSIONS (PDF chapter/section).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10168.pdf_
