## wpiea2019168-print-pdf - References

## Source details

**Canonical URL:** [wpiea2019168-print-pdf - References](https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019168-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2019/wpiea2019168-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2019/wpiea2019168-print-pdf.pdf.json)

---

### I. Introduction and scope
- Research question: What macroeconomic and initial conditions underpin persistent debt dollarization in emerging market economies (EMEs), with focus on nonfinancial firms’ foreign currency borrowing.
- Key metric of financial development: private credit-to-GDP ratio.
- Data: accounting information of over 9,000 nonfinancial firms from 21 major EMEs during 2009–2017.
- Focus currency: US dollars (foreign currency plays dual role as invoicing currency and primary unit of account for international debt contracts).

### II. Analytical framework and testable implications
- Theoretical model:
  - Partial equilibrium, mean-variance utility maximizing firms borrowing fixed amount = 1.
  - Bilateral exchange rate s in units of local currency per US dollar: s=1 in first period; second-period s∼N(1,σ^2_s).
  - Firm income y_i = θ_i + (1−θ_i) s; local-currency debt share α_i; dollar debt share 1−α_i.
  - Second-period profit Π_i = y_i − R^l_i α_i − R^d_i (1−α_i) s.
  - First-order condition: 1 − α^*_i = (R^l_i − R^d_i) / (ψ_i σ^2_s R^{d2}_i) + (1−θ_i) / R^d_i.
- Extended model with inflation:
  - Inflation shock π ∼ N(1,σ^2_π); profit Π_i = y_i − R^l_i α_i π − R^d_i (1−α_i) s π.
  - Optimal dollar share given by equation (5) with V = (R^d_i)^2 σ^2_s + (R^l_i)^2 σ^2_π + 2 R^l_i R^d_i σ_{sπ}.
- Key testable implications:
  - Dollar borrowing increases with interest differential R^l_i − R^d_i and decreases with exchange rate volatility σ^2_s.
  - Firms with dollar income (θ_i < 1) take on more dollar debt for hedging.
  - Dollar borrowing is positively associated with inflation volatility σ^2_π.

### III. Financial development, hedging availability, and de facto de-dollarization
- Market incompleteness implies firms use debt currency composition as primary hedge.
- As domestic financial markets develop:
  - Firms hedge more through foreign exchange derivatives; transaction costs decline.
  - Greater variety and availability of instruments reduce need for dollar funding (original sin hypothesis nuance).
- Covered interest parity and costless hedges:
  - If R^l_i > R^d_i f then firms borrow exclusively in dollars: 1 − α^*_i = 1.
  - If R^l_i = R^d_i f firms indifferent between dollar and local-currency borrowing.
- De facto de-dollarization arises when hedging availability makes exchange rate risk immaterial regardless of nominal debt currency composition.

### IV. Data, sample construction, and descriptive facts
- Data source: Capital IQ, S&P Global Market Intelligence corporate balance sheet database (Debt Capital Structure).
- Final analysis sample: 9,317 firms; 33,905 firm-year observations during 2009–2017 (after discarding top and bottom 1 percent of firm-year observations for each firm-level explanatory variable).
- Sample economies (21): Argentina, Brazil, Chile, China, Colombia, Czech Republic, Hungary, India, Indonesia, Korea, Malaysia, Mexico, Peru, the Philippines, Poland, Romania, Russia, South Africa, Taiwan Province of China (POC), Thailand, and Turkey.
- Panel characteristics: highly unbalanced and Asia focused; about 63 percent of firms located in China, India, and Taiwan POC.
- Treatment of ambiguous currency reporting:
  - If repayment currency reported as “foreign currency” or “multiple currency,” report as unavailable; in this sample about 16 percent of firms in 2015 reported carrying “foreign currency” or “multiple currency” liabilities; these liabilities are treated as part of dollar debt in this paper.
- Timing adjustment: statements filed before July 1 reassigned to previous calendar year; filed after June 30 assigned to same calendar year.

### V. Econometric specification
- Dependent variable: Y_i,j,k,t = ratio of dollar debt to total debt (bounded 0–1).
- Tobit latent-variable specification: Y* = α X_k,t−1 + β F_i,j,k,t−1 + γ Q_k,t−1 + μ_j + μ_k + μ_t + ε_i,j,k,t, with ε iid ∼ N(0, σ^2_ε).
- Regressors include:
  - Macroeconomic variables: interest rate differentials (short-term local currency interest rates minus three-month US dollar LIBOR), exchange rate volatility (annualized standard deviation of monthly real exchange rate changes (y/y) against the US dollar over 12 months), private credit-to-GDP ratio, real GDP growth, real exchange rate depreciation (end-of-year change (y/y) of US dollar per local currency nominal exchange rate multiplied by CPI_local/CPI_USA), inflation rate, inflation volatility.
  - Firm-level variables: log total assets, debt-to-total assets ratio (leverage), tangible assets-to-total assets ratio, cash-to-total assets ratio, return on assets, tradable sector dummy.
  - Additional controls: real GDP per capita (PPP, 2011 international dollars), exports-to-GDP ratio, composite country risk rating.
- Fixed effects: industry (28 industries), country, and time; standard errors clustered at country level; all variables lagged one year.

### VI. Main empirical findings
- General associations (benchmark Tobit estimates):
  - Exchange rate volatility coefficient: negative and statistically significant.
  - Interest differential coefficient: positive and statistically significant.
  - Interaction: exchange rate volatility × financial depth (private credit-to-GDP) positive and statistically significant.
  - Interaction: interest differential × financial depth statistically insignificant.
  - Financial depth alone: highly insignificant across specifications.
- Magnitudes (based on estimates in column (4) and summary):
  - Sample average dollar debt-to-total debt ratio: 10 percent (Dollar debt ratio 0.107 reported in sample summary).
  - A one percentage point increase in the interest differential is associated with a 0.1 percentage point increase in the dollar debt ratio.
  - A one standard deviation increase in exchange rate volatility (0.1) is associated with a 0.1 percentage point decrease in the dollar debt ratio.
- Thresholds and interactions:
  - Estimated threshold: about 100 percent private credit-to-GDP ratio (financial depth = 1.061 sample mean; threshold reported as private credit-to-GDP ratio exceeds 1 or 100 percent).
  - When private credit-to-GDP rises above 1, exchange rate volatility loses statistical significance as determinant of dollar debt ratio (Figure 2 description).
  - Interest differential remains statistically significant independent of financial depth.
- Other empirical relationships:
  - Inflation volatility and tradable sector dummy: positively associated with dollar debt ratio.
  - Real exchange rate depreciation: negative sign.
  - Firm-level variables (log total assets, total debt-to-total assets, tangible assets-to-total assets): positive and highly significant.
- Selected coefficient estimates (Table 3 highlights):
  - Exchange rate volatility: −0.096∗∗∗ (column (1)) [0.032]
  - Exchange rate volatility × Financial depth: 0.230∗∗∗ (column (2)) [0.082]
  - Interest differential: 0.724∗∗ (column (1)) [0.338]
  - Exchange rate depreciation: −0.230∗∗∗ (column (1)) [0.077]
  - Number of observations across regressions in Table 3: 33,905

### VII. Robustness and sensitivity analyses
- Reverse causality check:
  - Alternative exchange rate volatility: average exchange rate volatility of neighboring economies; results retain expected signs and significance.
- Capital account openness (KA) horse-race:
  - Negative correlation observed between financial depth and KA openness post-crisis (Figure 3).
  - Adding KA openness and interactions: interaction effects of financial depth remain robust; interest differential and exchange rate volatility remain significant at 1 percent level.
  - Interaction between interest differential and KA openness: highly significant with negative sign (quantitative importance of interest differentials declines as capital account opens).
  - Using KA openness in 2005 as alternative proxy yields similar results for exchange rate volatility; interaction between interest differential and KA openness becomes no longer statistically significant.
- Additional robustness checks:
  - Subsample splits (high FD: private credit-to-GDP > 1; low FD: private credit-to-GDP < 1); exclusion of largest-economy observations; separate regressions for Asian/non-Asian in low FD group; balanced panel.
  - Alternative dependent variables: foreign currency-denominated debt to-total debt ratio; log(dollar debt + 1); first-differenced dollar debt ratio.
  - Alternative estimators: fixed-effects, fractional probit (Papke and Wooldridge, 2008), two-part model (Cragg, 1971).
  - Conclusion: benchmark results robust across these variants.
- Note: benchmark results do not hold with euro-denominated debt ratio for low FD group (unreported).

### VIII. Taper Tantrum natural experiment and firm investment performance
- Context: May–September 2013 Fed tapering remarks triggered capital outflows, asset price declines, and exchange rate depreciation across EMEs.
- Empirical setup (2012–2014):
  - Sample: firms with positive dollar debt in yeart−1.
  - Dependent variable: I_i,j,k,t = capital expenditure in yeart / total assets in yeart−1.
  - Main regressor: DDR_i,j,k,t−1 × ΔEXR_i,j,k,t (interaction of firm dollar debt ratio and real exchange rate depreciation).
  - Expectation: If depreciation reduces investment via balance sheet effects, coefficient α on DDR × ΔEXR should be positive (α>0).
- Key empirical results (Table 7, fixed-effects; standard errors clustered at country level):
  - DDR × Exchange rate depreciation:
    - All: 0.232∗∗∗ [0.072]
    - Low FD (private credit-to-GDP < 1): 0.246∗∗∗ [0.090]
    - High FD (private credit-to-GDP > 1): 0.455 [0.301]
  - Exchange rate depreciation (level effect):
    - All: −0.107∗∗ [0.042]
    - Low FD: −0.169∗∗∗ [0.057]
    - High FD: −0.241 [0.169]
  - Number of observations:
    - All: 3,549
    - Low FD: 2,072
    - High FD: 1,477
- Interpretation:
  - Positive and significant DDR × ΔEXR in Low FD sample indicates negative balance sheet effect of depreciation on investment where financial markets are less developed.
  - Interaction insignificant in High FD sample; negative level effect of depreciation reflects competitiveness gains on investment.

### IX. Overall interpretation and policy implications
- Main synthesis:
  - Evidence supports that the importance of exchange rate volatility as determinant of firms’ debt currency composition declines with domestic financial market development and disappears when private credit-to-GDP exceeds estimated threshold of 100 percent.
  - Greater availability of foreign exchange hedging instruments in more developed financial markets likely reduces unhedged dollar debt.
- Policy recommendations (differentiated by financial development stage):
  - For EMEs at low or moderate financial development:
    - Priority: deepen and diversify domestic financial markets to reduce reliance on unhedged foreign currency debt.
    - Caution: foreign exchange interventions that dampen exchange rate volatility could encourage firms to increase unhedged foreign currency borrowing.
  - For EMEs with more developed financial markets:
    - Priority: enhance macroprudential measures, particularly when the interest differential is higher.
    - Policy space to intervene in foreign exchange markets appears relatively larger because exchange rate volatility is a less important determinant of firm dollar debt shares.
- Conceptual nuance:
  - Relationship between financial development and debt dollarization is non-linear and more complex than a simple original sin interpretation.
  - Low exchange rate volatility does not necessarily lead to higher foreign currency indebtedness; effect depends on state of financial development.
- Caveats:
  - Financial development is not claimed as a sufficient condition for redemption from original sin; effects may reflect broader structural differences (institutions, policy credibility).
  - Paper does not address whether low exchange rate volatility hinders development of foreign exchange derivative markets.

### X. Selected numeric highlights and sample statistics
- Sample and observation counts:
  - Total sample firms: 9,317
  - Final analysis sample observations: 33,905 (2009–2017)
- Key sample means/percentiles:
  - Dollar debt ratio (sample mean reported in various panels): 0.107; sample-level Dollar debt ratio 0.100 and elsewhere 0.107 reported.
  - Interest differential (sample mean): 0.038
  - Exchange rate volatility (sample mean): 0.138
  - Financial depth (private credit-to-GDP sample mean): 1.061
  - Time trend: dollar debt ratio falls from 0.114 in 2011 to 0.098 in 2017.
- Table I.1 (average dollar debt ratio, percent, end-2014) select values (Capital IQ measure, column (1)):
  - Brazil: 12.9 (standard deviation 19.8)
  - China: 2.7 (12.3)
  - India: 8 (13.1)
  - Indonesia: 35.7 (38.1)
  - Korea: 5.5 (14.8)
  - Mexico: 39.2 (35.1)
  - All sample EMEs (6,126 firms): 6.8 (19.3)
- Table 1 firm-level medians/means (selected, end-2017):
  - Total sample firms (Table III.2 Panel A total): Number of firms 9,317; Dollar debt ratio 0.107.
  - Total sample firms (Table 1 selected): Number of firms 5,470; Total assets mean 1,446.46 (USD, millions); Dollar debt ratio mean 0.100; Dollar debt (USD, millions) mean 47.76.

*Source: wpiea2019168-print-pdf - References . . .*

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

### wpiea2019168-print-pdf - References .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .

### I. Introduction and scope
- Research question: What macroeconomic and initial conditions underpin persistent debt dollarization in emerging market economies (EMEs), with focus on nonfinancial firms’ foreign currency borrowing.
- Key metric of financial development: private credit-to-GDP ratio.
- Data: accounting information of over 9,000 nonfinancial firms from 21 major EMEs during 2009–2017.
- Focus currency: US dollars (foreign currency plays dual role as invoicing currency and primary unit of account for international debt contracts).

### II. Analytical framework
- Theoretical approach:
  - Simple portfolio allocation model of firms’ optimal debt currency share, in the spirit of Ize and Yeyati (2003) but using a partial equilibrium setting tailored to nonfinancial firms.
  - Extension: firms earn income in both local and foreign currencies; firms maximize mean-variance utility under exchange rate and inflation risk.
  - The model nests Ize and Yeyati (2003) as a special case when firms earn income entirely in local currency.
- Empirical strategy:
  - Firm-level regressions linking firms’ dollar debt share to macroeconomic variables and measures of financial development.
  - Robustness checks include tests for reverse causality (using average value of neighboring economies’ exchange rate volatility) and a horse-race between financial development and capital account openness.
  - Natural experiment exploited: the Taper Tantrum episode as a stress test for hedging effectiveness across financial development levels.

### III. Main empirical findings
- General associations:
  - Firms’ dollar debt share is negatively associated with exchange rate volatility.
  - Firms’ dollar debt share is positively associated with the local currency–dollar interest rate differential.
- Magnitude of effects (relative to sample averages):
  - Sample average dollar debt-to-total debt ratio: 10 percent.
  - A one percentage point increase in the interest rate differential in a given year is associated with a 0.1 percentage point increase in the dollar debt ratio.
  - A one standard deviation increase in exchange rate volatility is associated with a 0.1 percentage point decrease in the dollar debt ratio.
- Interaction with financial development:
  - The negative effect of exchange rate volatility on dollar debt share weakens as domestic financial markets develop.
  - Financial development per se (private credit-to-GDP) does not affect the level of dollar debt share directly.
  - Estimated threshold: about 100 percent private credit-to-GDP ratio. On average, exchange rate volatility loses statistical significance as an explanatory variable beyond this level (as illustrated in Figure 2).
- Other empirical results:
  - No analogous moderating relationship found between the local currency–dollar interest rate differential and financial development.
  - Horse-race tests indicate main results derive from financial development rather than from capital controls.
  - Taper Tantrum evidence supports that firms in more developed financial markets are better hedged against exchange rate shocks.

### IV. Implications for investment and firm performance
- Literature synthesis:
  - Evidence is mixed on whether depreciation harms investment for firms with high dollar debt; results vary by region, sample, and financial market development.
  - Related findings in the literature point to differing post-depreciation investment and profit responses depending on financial development and firm characteristics.
- Paper-specific evidence:
  - Section V.D presents supporting evidence from the investment performance of sample firms during the Taper Tantrum period (details and results discussed in main text).

### V. Policy conclusions and recommendations
- Policy approaches should be differentiated by stage of financial development:
  - For EMEs at low or moderate financial development:
    - Priority: deepen and diversify domestic financial markets to reduce reliance on unhedged foreign currency debt.
    - Caution: foreign exchange interventions that dampen exchange rate volatility could encourage firms to increase unhedged foreign currency borrowing.
  - For EMEs with more developed financial markets:
    - Priority: enhance macroprudential measures, particularly when the interest differential is higher.
    - Policy space to intervene in foreign exchange markets appears relatively larger because exchange rate volatility is a less important determinant of firm dollar debt shares.
- Conceptual nuance:
  - The relationship between financial development and debt dollarization is non-linear and more complex than a simple original sin interpretation.
  - Low exchange rate volatility does not necessarily lead to higher foreign currency indebtedness; the effect depends on the state of financial development.

*Source: wpiea2019168-print-pdf - References .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  . .*

### Section VI discusses the policy implications of the findings.

### Section VI discusses the policy implications of the findings.

### III. AN ILLUSTRATIVE MODEL: setup and optimal debt currency choice
- Economy:
  - Continuum of firms indexed by i∈[0,1]; two periods.
  - Initial investment financed by borrowing fixed amount = 1.
  - Bilateral exchange rate s in units of local currency per US dollar: s=1 in first period; second-period s∼N(1,σ^2_s).
  - Firm i’s second-period income (local currency terms): y_i = θ_i + (1−θ_i) s, where θ_i is share of income in local currency.
- Debt instruments:
  - Share of loan in local currency: α_i; remainder in dollars: (1−α_i).
  - Local-currency gross interest rate: R^l_i > 1; dollar gross interest rate: R^d_i > 1.
- Second-period profit (local currency terms):
  - Π_i = y_i − R^l_i α_i − R^d_i (1−α_i) s. (Equation (1))
- Objective:
  - Firms risk averse; maximize mean-variance utility U(Π_i) = E(Π_i) − (ψ_i/2) Var(Π_i), subject to 0 ≤ α_i ≤ 1; ψ_i > 0.
- First-order condition — optimal dollar debt share:
  - 1 − α^*_i = (R^l_i − R^d_i) / (ψ_i σ^2_s R^{d2}_i) + (1−θ_i) / R^d_i. (Equation (2))
  - Rearranged expression linking currency mismatch:
    - (1/ψ_i)[(R^l_i − R^d_i) σ^2_s / R^{d2}_i] = (1−α^*_i) − (1−θ_i)/R^d_i. (Equation (3))
- Testable implications drawn from equation (2):
  - Firms borrow relatively more in dollars if interest differential R^l_i − R^d_i is high and expected exchange rate volatility σ^2_s is low.
  - Firms with dollar income (θ_i < 1) take on more dollar debt for hedging against exchange rate risk.

### Extended model with inflation: additional shocks and implications
- Inflation shock π ∼ N(1,σ^2_π); firm profit in real local-currency terms:
  - Π_i = y_i − R^l_i α_i π − R^d_i (1−α_i) s π. (Equation (4))
  - R^l_i and R^d_i now denote nominal gross interest rates.
- Optimal dollar share with inflation:
  - 1 − α^*_i =
    (R^l_i − R^d_i) / (ψ_i V)
    + (1−θ_i)(R^d_i σ^2_s + R^l_i σ_{sπ}) / V
    + R^l_i (R^l_i σ^2_π + R^d_i σ_{sπ}) / V. (Equation (5))
  - Where V = (R^d_i)^2 σ^2_s + (R^l_i)^2 σ^2_π + 2 R^l_i R^d_i σ_{sπ}, and σ_{sπ} = Cov(s,π).
- Key comparative statics:
  - Dollar borrowing remains negatively associated with exchange rate volatility σ^2_s (provided R^l_i − R^d_i > 0).
  - Dollar borrowing is positively associated with interest differential (R^l_i − R^d_i) and inflation volatility σ^2_π.

### Financial market development: hedging availability and de facto de-dollarization
- Market incompleteness baseline: firms adjust debt currency composition as primary hedge due to limited alternative instruments.
- As domestic financial markets develop and deepen (financial development and financial deepening used interchangeably):
  - Firms hedge currency exposure more through foreign exchange derivatives as transaction costs decline.
  - Figure 1 (described) shows positive relationship between domestic financial depth and post-crisis average daily turnover of foreign exchange derivatives (forwards, swaps, options).
  - Greater availability and variety of instruments (for example, long-term maturity bonds) reduce firms’ need to tap dollar funding markets (original sin hypothesis).
- Hypothetical costless derivative hedge that locks second-period exchange rate at f:
  - If R^l_i > R^d_i f, firms borrow exclusively in dollars: 1 − α^*_i = 1.
  - If R^l_i < R^d_i f, firms borrow in local currency.
  - If covered interest parity holds (R^l_i = R^d_i f), firms indifferent between dollar and local-currency borrowing.
  - Financially developed economy with available hedging can be considered de facto de-dollarized: firms no longer face exchange rate risk regardless of nominal debt currency composition (hedged dollar debt ≈ local-currency debt).

### IV. Firm-level data: data source and advantages
- Data source: Capital IQ, S&P Global Market Intelligence corporate balance sheet database.
  - Debt Capital Structure database provides individual debt instrument details: principal amount due, currency of denomination, and instrument type (bank loan, bond).
  - Information collected from company financial reports filed to national regulatory agencies, typically in supplementary notes to financial statements.
- Advantages relative to other commercial databases:
  - More detailed information on firms’ outstanding debt held in balance sheets than Worldscope and Orbis.
  - Direct, comprehensive information on firms’ liability exposure to exchange rate risks.
  - Example observation (from database): Ayala Land had an outstanding US$1.5 million variable interest rate bank loan from a local bank as of the end of fiscal year (textual example in source; no further years or dates provided).

*Source: wpiea2019168-print-pdf - Section VI discusses the policy implications of the findings.*

### 2016. Such information is unlikely to be included in international debt issuance databases but

### wpiea2019168-print-pdf - 2016. Such information is unlikely to be included in international debt issuance databases but

### Data and sample construction
- Sample composition:
  - Nonfinancial sector firms owned by the private sector, including both listed and nonlisted firms.
  - Accounting information is on a consolidated basis at the ultimate corporate parent level and converted from the local currency to millions of US dollars using the historical exchange rate at the end of each fiscal year.
  - Currency breakdown of outstanding total debt is obtained by aggregating information on individual debt instruments for each firm-year pair; aggregated debt amount is cross-checked against total principal due on the balance sheet.
  - Firms included are those that reported carrying a positive amount of foreign currency-denominated debt for at least one year during 2002–2017 to mitigate bias from firms that consistently do not report debt currency denomination.
  - Final analysis sample: 9,317 firms or 33,905 firm-year observations during 2009–2017 (after discarding the top and bottom 1 percent of the firm-year observations for each firm-level explanatory variable).
  - Sample economies (21): Argentina, Brazil, Chile, China, Colombia, Czech Republic, Hungary, India, Indonesia, Korea, Malaysia, Mexico, Peru, the Philippines, Poland, Romania, Russia, South Africa, Taiwan Province of China (POC), Thailand, and Turkey.
  - Panel characteristics: highly unbalanced and Asia focused, with about 63 percent of firms located in China, India, and Taiwan POC; includes both operating and non-operating firms.
- Treatment of ambiguous currency reporting:
  - Capital IQ criteria: (1) report stated repayment currency; (2) if repayment currency reported as “foreign currency” or “multiple currency,” report as unavailable; (3) if unspecified, assign financial statement reporting currency (usually local currency).
  - In this sample, about 16 percent of firms in 2015 reported carrying “foreign currency” or “multiple currency” liabilities; these liabilities are treated as part of dollar debt in this paper.
  - Dollar debt ratio estimates based on this definition are comparable with macro-level statistics and earlier studies (see Table I.1 in Appendix I).
- Timing adjustment:
  - Statements filed before July 1 in any given calendar year are reassigned to the previous calendar year; those filed after June 30 are assigned to the same calendar year (minimizes timing mismatch up to six months).

### Econometric specification
- Estimated model (Tobit specification):
  - Dependent variable Y_i,j,k,t = ratio of dollar debt to total debt, bounded between 0 and 1.
  - Latent variable specification: Y*_i,j,k,t = α X_k,t−1 + β F_i,j,k,t−1 + γ Q_k,t−1 + μ_j + μ_k + μ_t + ε_i,j,k,t, with ε_i,j,k,t iid ∼ N(0, σ^2_ε).
  - Regressors:
    - Macroeconomic variables (X_k,t−1): interest rate differentials (short-term local currency interest rates minus three-month US dollar LIBOR), exchange rate volatility (annualized standard deviation of monthly real exchange rate changes (y/y) against the US dollar over 12 months), private credit-to-GDP ratio (financial depth), real GDP growth, real exchange rate depreciation (end-of-year change (y/y) of US dollar per local currency nominal exchange rate multiplied by CPI_local/CPI_USA), inflation rate, inflation volatility (annualized standard deviation of monthly CPI inflation (y/y) over 12 months).
    - Firm-level variables (F_i,j,k,t−1): logarithm of total assets, debt-to-total assets ratio (leverage), tangible assets-to-total assets ratio, cash-to-total assets ratio, return on assets, tradable sector dummy.
    - Additional controls (Q_k,t−1): real GDP per capita (PPP, 2011 international dollars), exports-to-GDP ratio, composite country risk rating (International Country Risk Guide Database).
  - Fixed effects: industry (μ_j; 28 industries at industry-sector level based on Capital IQ classification), country (μ_k), and time (μ_t).
  - Standard errors clustered at the country level.
  - All variables are lagged by one year.

### Data description (summary statistics highlighted)
- Regional averages:
  - On average, Latin American firms hold the highest share of dollar-denominated debt at 29 percent; firms in other regions hold between 8 and 13 percent.
  - Percentile statistics indicate firms carrying dollar-denominated debt are only a small fraction of sample firms in all regions except Latin America.
- Correlations (2009–2017, Table 2 panel B):
  - Dollar debt ratio positively correlated with interest rate differentials and inflation volatility.
  - Dollar debt ratio negatively associated with financial depth (private credit-to-GDP).
  - Dollar debt ratio positively correlated with exchange rate volatility (contradicting model prediction); examined further via multivariate analyses.

### Hypotheses tested
- Hypothesis 1: Dollar debt ratio is positively associated with interest differentials and negatively associated with exchange rate volatility.
- Hypothesis 2: Greater financial market depth weakens the relationship between dollar debt ratio and exchange rate volatility; relationship between dollar debt ratio and interest differentials is not affected by financial development.
  - Empirical test: coefficients on interaction terms between exchange rate volatility and financial depth, and between interest differentials and financial depth.

### Main results (Tobit regressions, Table 3 and Figure 2)
- Sign and significance:
  - Exchange rate volatility coefficient: negative and statistically significant in benchmark specifications.
  - Interest differential coefficient: positive and statistically significant in benchmark specifications.
  - Interaction term between exchange rate volatility and financial depth: positive and statistically significant.
  - Interaction term between interest differential and financial depth: statistically insignificant.
  - Estimated coefficient of financial depth alone: highly insignificant across specifications.
- Economic magnitudes (based on estimates in column (4)):
  - A one percentage point increase in the interest differential is associated with a 0.1 percentage point increase in the dollar debt ratio.
  - A one standard deviation increase in exchange rate volatility (0.1) is associated with a 0.1 percentage point decrease in the dollar debt ratio.
- Figure 2 (Average marginal effects of exchange rate volatility on dollar debt ratio):
  - Financial depth (private credit-to-GDP) reduces average marginal effects of exchange rate volatility on dollar debt ratio when private credit-to-GDP is below 1.
  - When private credit-to-GDP rises above threshold of 1, exchange rate volatility loses statistical significance in determining dollar debt ratio.
  - Interest differential remains statistically significant independent of financial depth.
- Other estimated relationships:
  - Inflation volatility and tradable sector dummy: positively associated with dollar debt ratio.
  - Real exchange rate depreciation: negative sign, likely reflecting competitiveness effect on export earnings.
  - Firm-level variables (log total assets, total debt-to-total assets, tangible assets-to-total assets): positive and highly significant.

### Robustness tests
- Reverse causality check (alternative exchange rate volatility):
  - Alternative measure: average exchange rate volatility of neighboring economies (subgroups defined; South Africa uses entire sample).
  - Results (Table 4): coefficients for alternative exchange rate volatility and its interaction with financial depth, and interest rate differentials, retain expected signs and statistical significance; interaction between interest differentials and financial depth remains insignificant.
- Surrogate financial intermediaries (capital account openness tests):
  - Observed negative correlation between financial depth and KA openness post-crisis (Figure 3).
  - Horse-race tests adding KA openness and its interactions to regressions (Table 5):
    - Interaction effects of financial depth remain robust; coefficients for interest differential and exchange rate volatility remain significant at 1 percent level with expected signs.
    - Interaction between interest differential and KA openness: highly significant with negative sign (implies quantitative importance of interest differentials declines as capital account opens).
    - Using KA openness in 2005 as alternative proxy yields similar results for exchange rate volatility; interaction between interest differential and KA openness becomes no longer statistically significant.
  - Conclusion: benchmark results not driven by omission of capital account openness.
- Additional robustness checks (Table 6):
  - Subsample splits: high FD (private credit-to-GDP > 1) and low FD (private credit-to-GDP < 1); exclusion of largest-observation economies in each group; separate regressions for Asian and non-Asian firms in low FD group; balanced panel subsample.
  - Alternative dependent variables: foreign currency-denominated debt to-total debt ratio; logarithm of amount of dollar-denominated debt plus 1; first-differenced dollar debt ratio (to address serial correlation).
  - Alternative estimators: fixed-effects, fractional probit (Papke and Wooldridge, 2008), two-part model (Cragg, 1971).
  - Overall: benchmark results in Table 3 are robust to these subsamples, alternative dependent variables, and estimation methods.
  - Additional unreported robustness variants found consistent results: (a) restricting sample to firms carrying positive foreign currency-denominated debt in year t; (b) using imports-to-GDP instead of exports-to-GDP; (c) clustering standard errors at firm level instead of country level.
- Note: benchmark results do not hold with euro-denominated debt ratio for the low FD group (unreported).

### Additional analysis previewed
- Evidence from the Taper Tantrum:
  - Empirical analyses indicate influence of exchange rate volatility on firm-level dollar debt composition diminishes with domestic financial market development and becomes statistically insignificant when private credit-to-GDP exceeds estimated threshold of 1.
  - If this finding is driven by greater use of currency hedging derivatives by firms in more developed financial markets, further analysis follows (text continues beyond provided excerpt).

*Source: IMF working paper excerpt (wpiea2019168-print-pdf).*

### Section III, these firms should also be relatively less affected by exchange rate shocks than

### Section III, these firms should also be relatively less affected by exchange rate shocks than

### Taper Tantrum as a natural experiment
- Context: During May–September 2013, remarks by the U.S. Federal Reserve Chairman on likely tapering of the Fed’s asset purchase program triggered large capital outflows from EMEs, sharp declines in asset prices, and exchange rate depreciation across EMEs.
- Mechanism examined: For firms with significant unhedged dollar debt, exchange rate depreciation inflates the local currency value of dollar debt and can weaken balance sheets, negatively affecting investment (balance sheet channel). A competing channel is real depreciation’s positive effect on investment via improved competitiveness.

### Empirical specification for the balance sheet channel
- Sample period: 2012–2014.
- Sample restriction: Firms that hold dollar debt in yeart−1.
- Dependent variable: I_i,j,k,t = ratio of firm i’s capital expenditure in yeart to total assets in yeart−1.
- Main regressor: DDR_i,j,k,t−1 × ΔEXR_i,j,k,t (interaction between firm dollar debt ratio in yeart−1 and real exchange rate depreciation against the dollar in yeart).
- Interpretation: If exchange rate depreciation reduces investment via balance sheet effects, the coefficient α on DDR × ΔEXR should be positive (α>0). The coefficient should be larger/significant where dollar debt is less hedged (expected in less-developed financial markets).
- Controls and fixed effects:
  - Lagged firm-level variables: logarithm of total assets; debt-to-total assets ratio; tangible assets-to-total assets ratio; cash-to-total assets ratio; dollar debt-to-total debt ratio; sales growth (all lagged one year).
  - Contemporaneous macro variables: real GDP growth; inflation; inflation volatility; country risk rating; logarithm of real GDP per capita.
  - Fixed effects: firm, year, and industry-year.

### Empirical results (impact of exchange rate shocks on investment)
- Table 7 key coefficients (fixed-effects panel regressions; firms with positive dollar debt in yeart−1; standard errors clustered at country level):
  - Dollar debt ratio × Exchange rate depreciation:
    - All: 0.232∗∗∗ [0.072]
    - Low FD (private credit-to-GDP ratio less than 1): 0.246∗∗∗ [0.090]
    - High FD (private credit-to-GDP ratio greater than 1): 0.455 [0.301]
  - Exchange rate depreciation (level effect):
    - All: −0.107∗∗ [0.042]
    - Low FD: −0.169∗∗∗ [0.057]
    - High FD: −0.241 [0.169]
  - Number of observations:
    - All: 3,549
    - Low FD: 2,072
    - High FD: 1,477
- Interpretation of results:
  - The positive and statistically significant interaction coefficient for the Low FD sample indicates a negative balance sheet effect of exchange rate depreciation (higher dollar debt share amplifies investment declines) in less-developed financial markets.
  - The interaction term is insignificant in high FD economies (column (3) of Table 7), suggesting the negative balance sheet effect was not significant there.
  - The coefficient on exchange rate depreciation is negative (statistically significant for All and Low FD), implying real depreciation per se had a beneficial competitiveness effect on investment; the DDR × ΔEXR interaction isolates the balance sheet cost for firms with dollar debt.

### Overall findings and interpretation (Concluding remarks)
- Evidence supports the original sin hypothesis: the importance of exchange rate volatility as a determinant of firms’ debt currency composition declines as domestic financial markets develop and statistically disappears when the private credit-to-GDP ratio exceeds an estimated threshold level of 100 percent.
- Explanation: Greater availability of foreign exchange hedging instruments in more developed financial markets likely reduces unhedged dollar debt.
- Taper Tantrum evidence: Investment performance during the episode supports the hedging-availability explanation — negative balance sheet effects concentrated in less-developed financial markets.
- Caveats:
  - Financial development is not claimed to be a sufficient condition for redemption from original sin; estimated effects could reflect deeper structural differences (institutions, credibility of macroeconomic policy regimes).
  - The paper does not address whether low exchange rate volatility hinders financial development, particularly foreign exchange derivative markets. Existing studies cited (Gadanecz and Mehrotra, 2013; Mihaljek and Packer, 2010) do not find a robust link between exchange rate flexibility and foreign exchange derivative market development compared with trade, financial openness, and size of domestic bond and equity markets.

### Policy implications and recommendations
- Need for differentiated policy approach to debt dollarization:
  - In economies at relatively low stages of financial development:
    - Policies aimed at developing domestic financial markets could promote de-dollarization by lowering unhedged dollar debt on firms’ balance sheets.
    - Maintaining low exchange rate volatility could be costlier than in more developed markets because it may worsen and prolong unhedged debt dollarization.
  - In economies at relatively advanced stages of financial development:
    - Priority should be strengthening oversight and macroprudential policies to safeguard financial stability, especially when interest rate differentials are high.

### Supporting descriptive statistics and robustness (selected numeric highlights)
- Estimated threshold where exchange rate volatility’s importance disappears: private credit-to-GDP ratio exceeds 100 percent.
- Table 3 (macro determinants of dollar debt ratio) notable coefficients:
  - Exchange rate volatility: −0.096∗∗∗ (column (1)) [0.032]
  - Exchange rate volatility × Financial depth: 0.230∗∗∗ (column (2)) [0.082]
  - Interest differential: 0.724∗∗ (column (1)) [0.338]
  - Exchange rate depreciation: −0.230∗∗∗ (column (1)) [0.077]
  - Number of observations across regressions in Table 3: 33,905
- Sample-level summary statistics (selected):
  - Total sample firms (Table III.2, Panel A total): Total number of firms 9,317; Dollar debt ratio 0.107; Interest differential 0.038; Exchange rate volatility 0.138; Financial depth 1.061.
  - Time trend (Table III.2, Panel B): dollar debt ratio falls from 0.114 in 2011 to 0.098 in 2017.
- Table I.1 (average dollar debt ratio in select EMEs, percent, end-2014) — select values from Capital IQ measure (column (1)):
  - Brazil: 12.9 (standard deviation 19.8)
  - China: 2.7 (12.3)
  - India: 8 (13.1)
  - Indonesia: 35.7 (38.1)
  - Korea: 5.5 (14.8)
  - Mexico: 39.2 (35.1)
  - All sample EMEs (6,126 firms): 6.8 (19.3)
- Table 1 firm-level medians/means (selected, end-2017):
  - Total sample firms: Number of firms 5,470; Total assets mean 1,446.46 (USD, millions); Dollar debt ratio mean 0.100; Dollar debt (USD, millions) mean 47.76.

*Source: wpiea2019168-print-pdf - Section III, these firms should also be relatively less affected by exchange rate shocks than*

### REFERENCES

### REFERENCES

### Foreign currency borrowing and corporate dollar debt
- Acharya, Viral V, and Siddharth Vij, 2016, “Foreign currency borrowing of corporations as carry trades: Evidence from india,” inNSE-NYU Conference on Indian Financial Markets, pp. 21–22.
- Allayannis, George, Gregory W. Brown, and Leora F. Klapper, 2003, “Capital Structure and Financial Risk: Evidence from Foreign Debt Use in East Asia,”The Journal of Finance, Vol. 58, No. 6, pp. 2667–2710.
- Allayannis, George, and Eli Ofek, 2001, “Exchange rate exposure, hedging, and the use of foreign currency derivatives,”Journal of International Money and Finance, Vol. 20, No. 2, pp. 273 – 296.
- Bleakley, Hoyt, and Kevin Cowan, 2008, “Corporate Dollar Debt and Depreciations: Much Ado About Nothing?”The Review of Economics and Statistics, Vol. 90, No. 4, pp. 612–626.
- Brown, Martin, Steven Ongena, and Pinar Ye ̧sin, 2011, “Foreign currency borrowing by small firms in the transition economies,”Journal of Financial Intermediation, Vol. 20, No. 3, pp. 285 – 302.
- Bruno, Valentina, and Hyun Song Shin, 2017, “Global Dollar Credit and Carry Trades: A Firm-Level Analysis,”The Review of Financial Studies, Vol. 30, No. 3, pp. 703–749.
- Caballero, Ricardo J., and Arvind Krishnamurthy, 2003, “Excessive Dollar Debt: Financial Development and Underinsurance,”The Journal of Finance, Vol. 58, No. 2, pp. 867–893.
- Dao, Mai Chi, Camelia Minoiu, and Jonathan David Ostry, 2017, “Corporate Investment and the Real Exchange Rate,” IMF Working Paper 17/183, International Monetary Fund.
- Galindo, Arturo, Ugo Panizza, and Fabio Schiantarelli, 2003, “Debt composition and balance sheet effects of currency depreciation: a summary of the micro evidence,”Emerging Markets Review, Vol. 4, No. 4, pp. 330–339.
- Kedia, Simi, and Abon Mozumdar, 2003, “Foreign Currency Denominated Debt: An Empirical Examination,”The Journal of Business, Vol. 76, No. 4, pp. 521–546.
- Keloharju, Matti, and Mervi Niskanen, 2001, “Why Do Firms Raise Foreign Currency Denominated Debt? Evidence from Finland,”European Financial Management, Vol. 7, No. 4, pp. 481–496.
- Martinez, Lorenza, and Alejandro Werner, 2002, “The exchange rate regime and the currency composition of corporate debt: the Mexican experience,”Journal of Development Economics, Vol. 69, No. 2, pp. 315–334.
- Rosenberg, Christoph B., and Marcel Tirpàk, 2008, “Determinants of Foreign Currency Borrowing in the New Member States of the EU,” IMF Working Paper 08/173, International Monetary Fund.
- Salomao, Juliana, and Liliana Varela, 2018, “Exchange Rate Exposure and Firm Dynamics,” CEPR Discussion Paper 12654, C.E.P.R. Discussion Paper.
- Jiang, Bo, and Tahsin Saadi Sedik, 2019, “The Turning Tide: How Vulnerable are Asian Corporates?” IMF Working Paper 19/93, International Monetary Fund.

### Financial dollarization, currency mismatches, and exchange rate exposure
- Aguiar, Mark, 2005, “Investment, devaluation, and foreign currency exposure: The case of Mexico,”Journal of Development Economics, Vol. 78, No. 1, pp. 95–113.
- Arteta, Carlos O., 2005, “Exchange Rate Regimes and Financial Dollarization: Does Flexibility Reduce Currency Mismatches in Bank Intermediation?”The B.E. Journal of Macroeconomics, Vol. 5, No. 1, pp. 1–30.
- Barajas, Adolfo, Sergio Restrepo, Roberto Steiner, Juan Camilo Medellìn, and Cèsar Pabòn, 2017, “Currency Mismatches and Vulnerability to Exchange Rate Shocks: Nonfinancial Firms in Colombia,” IMF Working Paper 17/263, International Monetary Fund.
- Basso, Henrique S., Oscar Calvo-Gonzalez, and Marius Jurgilas, 2011, “Financial dollarization: The role of foreign-owned banks and interest rates,”Journal of Banking and Finance, Vol. 35, No. 4, pp. 794 – 806, crete Conference 2010: The Future of Universal Banking.
- Chui, Michael, Ingo Fender, and Vladyslav Sushko, 2014, “Risks related to EME corporate balance sheets: the role of leverage and currency mismatch,”BIS Quarterly Review, pp. 35–47.
- Chui, Michael, Emese Kuruc, and Philip Turner, 2016, “A new dimension to currency mismatches in the emerging markets: non-financial companies,”BIS Working Paper, No.550.
- Ize, Alain, and Eduardo Levy Yeyati, 2003, “Financial dollarization,”Journal of International Economics, Vol. 59, No. 2, pp. 323 – 347.
- Luca, Alina, and Iva Petrova, 2008, “What drives credit dollarization in transition economies?”Journal of Banking and Finance, Vol. 32, No. 5, pp. 858 – 869.
- Mihaljek, Dubravko, and Frank Packer, 2010, “Derivatives in emerging markets,”BIS Quarterly Review, pp. 43–58.
- McCauley, Robert N, Patrick McGuire, and Vladyslav Sushko, 2015, “Dollar credit to emerging market economies,”BIS Quarterly Review, pp. 27–41.
- Gadanecz, Blaise, and Aaron Mehrotra, 2013, “The exchange rate, real economy and financial markets,” in Bank for International Settlements (ed.),Sovereign risk: a world without risk-free assets?, Vol. 73, pp. 11–23 (Bank for International Settlements).
- Galindo, Arturo, Ugo Panizza, and Fabio Schiantarelli, 2003, “Debt composition and balance sheet effects of currency depreciation: a summary of the micro evidence,”Emerging Markets Review, Vol. 4, No. 4, pp. 330–339.

### Exchange rate regimes, fear of floating, and original sin
- Calvo, Guillermo A., and Carmen M. Reinhart, 2001, “Fixing for your life,” in Susan Collins and Dani Rodrik (eds.),Brookings Trade Forum 2000, pp. 1–39 (Brookings Institution).
- Calvo, Guillermo A., and Carmen M. Reinhart, 2002, “Fear of Floating,”The Quarterly Journal of Economics, Vol. 117, No. 2, pp. 379–408.
- Eichengreen, Barry, and Ricardo Hausmann, 1999, “Exchange Rates and Financial Fragility,” NBER Working Paper 7418, National Bureau of Economic Research, Inc.
- Eichengreen, Barry, Ricardo Hausmann, and Ugo Panizza, 2005, “The Mystery of Original Sin,” in Barry Eichengreen and Ricardo Hausmann (eds.),Other People’s Money (University of Chicago Press).
- Hausmann, Ricardo, and Ugo Panizza, 2011, “Redemption or Abstinence? Original Sin, Currency Mismatches and Counter Cyclical Policies in the New Millennium,”Journal of Globalization and Development, Vol. 2, No. 1, pp. 1–35.
- Reinhart, Carmen, 2000, “Mirage of Floating Exchange Rates,”American Economic Review, Vol. 90, No. 2, pp. 65–70.
- Burnside, Craig, Martin Eichenbaum, and Sergio Rebelo, 2001, “Hedging and financial fragility in fixed exchange rate regimes,”European Economic Review, Vol. 45, No. 7, pp. 1151 – 1193, international Seminar On Macroeconomics.

### Global liquidity, capital markets, and macrofinancial risks
- Avdjiev, Stefan, Michael Chui, and Hyun Song Shin, 2014, “Non-financial corporations from emerging market economies and capital flows,”BIS Quarterly Review, pp. 67–77.
- Caballero, Juliàn, Ugo Panizza, and Andrew Powell, 2016, “The Second Wave of Global Liquidity: Why Are Firms Acting Like Financial Intermediaries?”Inter-American Development Bank, , No. 7638.
- McCauley, Robert N, Patrick McGuire, and Vladyslav Sushko, 2015, “Dollar credit to emerging market economies,”BIS Quarterly Review, pp. 27–41.
- Obstfeld, Maurice, 1998, “The Global Capital Market: Benefactor or Menace?”Journal of Economic Perspectives, Vol. 12, No. 4, pp. 9–30.
- Jeanne, Olivier, 2003, “Why do emerging economies borrow in foreign currency?” International Monetary Fund, , No. 3-177.
- International Monetary Fund, 2018, “World Economic Outlook, October 2018: Challenges for Monetary Policy in Emerging Economies as Global Financial Conditions Normalize,” Washington, DC.

### Methods, econometrics, and empirical techniques
- Cragg, John G, 1971, “Some Statistical Models for Limited Dependent Variables with Application to the Demand for Durable Goods,”Econometrica, Vol. 39, No. 5, pp. 829–44.
- Papke, Leslie, and Jeffrey Wooldridge, 2008, “Panel data methods for fractional response variables with an application to test pass rates,”Journal of Econometrics, Vol. 145, No. 1-2, pp. 121–133.
- Harvey, Campbell R., and Andrew H. Roper, 1999, “The Asian Bet,” in Alison Harwood, Michael Pomerleano, and Robert E. Litan (eds.),The Crisis in Emerging Financial Markets: A World Bank Group-Brookings Conference Report, pp. 29–116 (Brookings Institution Press).
- Mishkin, Frederic S., 1996, “Understanding Financial Crises: A Developing Country Perspective,” NBER Working Papers 5600, National Bureau of Economic Research, Inc.
- Kalemli-Ozcan, Sebnem, Herman Kamil, and Carolina Villegas-Sanchez, 2016, “What Hinders Investment in the Aftermath of Financial Crises: Insolvent Firms or Illiquid Banks?”The Review of Economics and Statistics, Vol. 98, No. 4, pp. 756–769.

*Source: wpiea2019168-print-pdf - REFERENCES*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019168-print-pdf.pdf_