## 1. The Impact of Dollarization on Financial Debt

## Source details

**Canonical URL:** [1. The Impact of Dollarization on Financial Debt](https://www.imf.org/-/media/files/publications/wp/2018/wp18200.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2018/wp18200.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2018/wp18200.pdf.json)

---

### I. Introduction and data scope
- Sample: 77 emerging and developing countries over the period 1996–2015 (panel is unbalanced).
- Definition:
  - "Dollarization" refers to use of any foreign currency other than the legal tender (not only the dollar).
  - Deposit dollarization = ratio of dollar-denominated deposits to total broad money deposits.
  - Credit dollarization = ratio of dollar-denominated loans to total loans.
- Empirical focus: impact of partial (unofficial) dollarization on financial development measured along three dimensions:
  - Financial depth: log of private credit to GDP (PrivCred).
  - Financial access: log of bank accounts per 1000 adults (Accounts).
  - Financial efficiency: bank net interest margin (NetIntMarg).
- Stylized facts:
  - Average deposit dollarization across the sample of partially dollarized developing economies was around 30 percent in 2015.
  - Deposit and credit dollarization correlation: 82 percent.
  - High-dollarization examples: Cambodia and Nicaragua with dollarization levels around 90 percent.
  - Low-dollarization examples: China and Bangladesh with close to zero dollarization.
  - Geographic patterns: Eastern European transition economies and Latin America generally more dollarized; MENA and African countries much less dollarized.
  - Deposit dollarization has generally been higher than credit dollarization across the sample.
  - Both deposit and credit dollarization diminished over 1996–2015, with a notable jump between 1996 and 1997 due to inclusion of transition economies.

### II. Main empirical findings
- Overall result: financial dollarization, and deposit dollarization in particular, has a negative impact on financial development.
- Financial depth (PrivCred):
  - Deposit dollarization has a statistically significant negative impact on financial development (log of credit to GDP).
  - Quantitative effects:
    - "A 1 percentage point increase in deposit dollarization reduces credit to GDP by around 0.4 percent in the short run, and 2–2.5 percent in the long run."
    - "The observed 5 percentage point reduction in dollarization since its peak has contributed to an increase in credit to GDP by about 10 percent in the past 15 years."
    - For highly dollarized economies:
      - Cutting dollarization by half (e.g. from about 90 percent to 45 percent) would potentially increase financial depth by close to 20 percent in the short run.
      - In the long run, halving dollarization could theoretically double financial depth in countries with near complete dollarization.
  - Credit dollarization has a much smaller and not statistically significant impact on financial depth.
  - Mechanisms:
    - A share of foreign currency deposits may be transferred overseas rather than returned to the domestic economy as private credit, contributing to a shallower domestic financial sector.
    - Financial assets and liabilities in multiple currencies create additional frictions and costs inhibiting financial deepening.
  - Mismatch evidence:
    - The aggregate difference (DepDol-CredDol) has a negative effect on financial depth, supporting the hypothesis that excess foreign currency deposits are invested abroad.
  - Interaction with inflation history:
    - Deposit dollarization’s negative impact on financial deepening is dampened in countries with past experiences of very high inflation (annual inflation higher than 250 percent between 1980 and 1997; for Eastern European transition economies the period is 1995–1998). In some high-inflation cases deposit dollarization can facilitate financial deepening.
- Financial access (Accounts):
  - System GMM: no evidence of a link between deposit or credit dollarization and financial access (columns 1 and 2 in Table 3).
  - Difference GMM: both deposit and credit dollarization have a statistically significant negative impact on financial access, but difference GMM is given less weight due to performance concerns when the dependent variable is highly persistent.
- Financial efficiency (NetIntMarg):
  - Some evidence that deposit dollarization increases net interest margins, implying lower financial efficiency.
  - Possible explanations: contraction in credit supply when banks shift assets abroad and higher concentration/monopoly power in banking systems of dollarized economies (supported by a positive coefficient on the 3 Bank Conc. variable).
  - Results vary across specifications and estimators; significance can disappear when reducing the number of instruments or using alternative parsimonious specifications.

### III. Theoretical channels and literature synthesis
- Portfolio approach:
  - Dollarization as a response to macroeconomic instability (high inflation, exchange rate volatility) via minimum-variance portfolio choice; expectations and monetary policy credibility are key.
  - Persistence of dollarization explained by lack of credible monetary policy/exchange rate regimes.
- Credit-market frictions:
  - Market frictions, devaluation expectations, moral hazard, and incomplete credit markets can drive dollarization.
  - Presence of foreign banks associated with higher credit dollarization.
- De-dollarization policy lessons from literature:
  - Credible macroeconomic stabilization (lower inflation, exchange rate stability).
  - Differential prudential regulations to reduce incentives for foreign currency transactions (e.g., higher provisions for foreign currency loans, tighter capital requirements against open FX positions, differentiated reserve requirements/remuneration).
  - Development of local currency capital markets to provide alternate vehicles for saving and investment.

### IV. Measurement of variables
- Dollarization measures:
  - Deposit dollarization = ratio of dollar-denominated deposits to total broad money deposits.
  - Credit dollarization = ratio of dollar-denominated loans to total loans.
  - Foreign-currency cash holdings and non-bank financial institutions are not included due to data limitations.
- Financial development measures:
  - Financial depth: log of private credit to GDP (PrivCred).
  - Financial access: log of bank accounts per 1000 adults (Accounts).
  - Financial efficiency: bank net interest margin (NetIntMarg) = net interest revenue / average interest-bearing assets.
- Additional construct:
  - Minimum-variance portfolio (MVP) share: MVP = [Var(π)+Cov(π,s)]/[Var(π)+Var(s)+2Cov(π,s)] where π = inflation and s = change in the real exchange rate, using past five-year variances and covariance.

### V. Empirical strategy and identification
- Baseline dynamic panel specification relates financial development indicator F_{i,t} to lagged F_{i,t-1}, dollarization measure (DepDol or CredDol), policy vector P, structural vector S, dollarization-related vector D, country fixed effects, time fixed effects, and error term. Coefficient of interest is β on dollarization.
- Controls:
  - Policy variables (P): log changes of real GDP per capita (GDP growth), CPI inflation (Inflation), banking crisis dummy (Banking Crisis), market share of three largest banks (3 Bank Conc), capital account openness index (KaOpen), external debt to GNI (ExtDebt), Worldwide Governance Indicators composite (WGI).
  - Structural variables (S): log real GDP per capita (GDP pc), log population (Pop), log age dependency ratio (Age Dep ratio).
  - Dollarization determinants (D): MVP, REER Cyclicality, Foreign Banks, log NEER, Transition dummy.
  - Robustness controls: Imports, Remittances, short-term nominal interest rate differential (i Diff).
- Endogeneity and estimator choice:
  - Dynamic panel with lagged dependent variable induces Nickell bias if estimated by OLS/FE.
  - Preferred estimator: system GMM (difference GMM used for robustness).
  - Endogeneity assumptions:
    - GDP growth treated as endogenous.
    - Inflation allowed endogenous with respect to financial depth.
    - Institutional/regulatory variables treated as predetermined; structural variables treated as exogenous.
  - Instrument proliferation addressed by "collapsing" instruments and by models restricting endogeneity of dollarization determinants; robustness checks with fewer instruments performed.
- Diagnostics:
  - Arellano-Bond AB-AR(2) indicates no second-order autocorrelation.
  - Hansen J-test confirms joint validity of instruments in reported specifications.

### VI. Robustness checks and caveats
- Robustness strategies:
  - Restricting dollarization and its determinants to be predetermined instead of endogenous.
  - Allowing only lagged dependent variable, GDP growth, inflation and dollarization to be endogenous while treating dollarization determinants as exogenous or predetermined.
  - Reducing number of instruments and using more parsimonious models.
  - Alternative estimators: difference GMM and fixed effects.
  - Excluding countries with deposit or credit dollarization below 1 percent (sample of 63 countries): conclusions remain the same.
- Key robustness conclusions:
  - Reducing instrument count does not change main conclusions; in some specifications the impact of deposit dollarization on financial depth is larger when determinants are predetermined.
  - Difference GMM yields similar signs but is less reliable when dependent variable is highly persistent.
  - Fixed effects estimates support sign of coefficients but are subject to Nickell bias.
- Caveats:
  - System GMM is only a partial solution to identification; residual time-varying unobserved heterogeneity may bias results and increase Type 1 error risk.
  - Data limitations: exclusion of foreign-currency cash holdings and non-bank financial institutions.
  - Heterogeneity across countries and specifications, particularly for efficiency results, implies caution in causal interpretation.
  - Further research recommended: study significant de-dollarization events and use more granular sectoral/firm/individual data.

### VII. Policy implications and recommendations
- Justification for de-dollarization efforts:
  - Dollarization is associated with lower financial depth and banking sector efficiency, alongside lower monetary policy effectiveness, limits to fiscal flexibility and heightened financial stability risks.
  - Policy efforts to increase domestic currency use in financial transactions are warranted, especially where financial dollarization is high.
- Components of successful de-dollarization strategies:
  - Credible macroeconomic stabilization to lower inflation and stabilize exchange rate.
  - Strengthening economic institutions and monetary policy frameworks to enhance credibility.
  - Prudential measures to lower incentives for foreign currency transactions:
    - Raising provisions for foreign currency loans.
    - Tighter capital requirements against open foreign exchange positions.
    - Differentiated reserve requirements and remuneration on foreign currency deposits.
  - Development of local currency financial markets to provide alternate vehicles for savings and longer-term investment.

### VIII. Key dataset and summary statistics (selected exact values)
- Sample period: 1996–2015.
- Sample size: 77 emerging market and developing countries (panel is unbalanced).
- Countries after excluding deposit or credit dollarization below 1 percent on average: 63.
- Selected summary statistics (from Table 1):
  - Deposit Dollarization: Mean 0.291; Std. Dev 0.245; Min 0; Max 0.984; Obs 1427.
  - Credit Dollarization: Mean 0.252; Std. Dev 0.240; Min 0; Max 0.982; Obs 1070.
  - Private Credit to GDP: Mean 34.82; Std. Dev 28.09; Min 1.61; Max 165.7; Obs 1427.
  - Net Interest Margin: Mean 5.755; Std. Dev 2.931; Min 0.17; Max 25.49; Obs 1406.
  - Accounts per 1000 adults: Mean 942.4; Std. Dev 915.7; Min 2.39; Max 5342; Obs 605.
  - GDP p.c., USD: Mean 4739.1; Std. Dev 6266.4; Min 160.3; Max 49015.9; Obs 1427.
  - Inflation: Mean 0.134; Std. Dev 1.146; Min -0.09; Max 41.45; Obs 1425.
  - Foreign Banks (%): Mean 38.84; Std. Dev 25.66; Min 0; Max 100; Obs 1357.
  - 3 Bank concentration ratio: Mean 0.626; Std. Dev 0.189; Min 0.146; Max 1; Obs 1398.
  - NEER: Mean 405.7; Std. Dev 8200; Min 33.2; Max 304944; Obs 1427.
  - KaOpen: Mean 0.476; Std. Dev 0.330; Min 0; Max 1; Obs 1426.
  - WGI: Mean -0.324; Std. Dev 0.567; Min -1.67; Max 1.25; Obs 1427.
  - REER Cyclicality: Mean 0.104; Std. Dev 0.544; Min -0.99; Max 1.00; Obs 1383.
  - External Debt/GNI: Mean 0.502; Std. Dev 0.335; Min 0.03; Max 2.26; Obs 1394.
  - MVP: Mean 0.045; Std. Dev 0.934; Min -2.69; Max 9.59; Obs 1338.
  - Population (mil.): Mean 51.0; Std. Dev 157.7; Min 0.47; Max 1371.2; Obs 1427.
  - Age Dependency Ratio: Mean 4.101; Std. Dev 0.288; Min 3.47; Max 4.73; Obs 1427.
  - Transition Economy dummy: Mean 0.222; Std. Dev 0.416; Min 0; Max 1; Obs 1427.
  - Banking Crisis Dummy: Mean 0.056; Std. Dev 0.229; Min 0; Max 1; Obs 1423.

*Source: wp18200 — 1. The Impact of Dollarization on Financial Debt (PDF chapter/section).*

### 1. The Impact of Dollarization on Financial Debt ________________________________18

### 1. The Impact of Dollarization on Financial Debt

### Contained Sections
- 1. The Impact of Dollarization on Financial Debt ________________________________18
- 2. Dollarization and Financial Development in Countries with a History of High  
  Inflation _______________________________________________________________21
- 3. Dollarization and Financial Access and Efficiency _____________________________23

### Figures Listed
- 1. Deposit and Credit Dollarization  ____________________________________________9
- 2. Regional Average Deposit Dollarization ______________________________________9
- 3. Regional Average Credit Dollarization  ______________________________________10
- 4. Deposit Dollarization in a Number of Countries _______________________________10
- 5. Deposit vs. Credit Dollarization ____________________________________________10
- 6. Deposit Dollarization and Financial Development ______________________________11
- 7. Credit Dollarization and Financial Development _______________________________11
- 8. Aggregate Mismatch between Deposit and Credit Dollarization and Financial  
  Development ___________________________________________________________12
- 9. Deposit Dollarization in Foreign Asset Ratios of Deposit Taking Banks  ____________19

### Appendix and Data Structure
- APPENDIX I: DATA _______________________________________________________30
  - Country coverage _______________________________________________________30
  - Dollarization ___________________________________________________________30
  - Financial development ___________________________________________________30
  - Control and additional variables ____________________________________________31
- TABLE 5
- 1. Summary Statistics ______________________________________________________33
- APPENDIX II: ADDITIONAL RESULTS ______________________________________34
  - TABLES
    - 1. Dollarization and Financial Development with Dollarization and Its Determinants  
      Predetermined __________________________________________________________34
    - 2. Dollarization and Financial Development Relationship Estimated with Fewer  
      Instruments ____________________________________________________________35
    - 3. Dollarization and Financial Development Using More Parsimonious Models  ________36
    - 4. Deposit Dollarization and Financial Depth, Access and Efficiency with FE and Difference 
      GMM Estimators  _______________________________________________________37

*Source: wp18200 - 1. The Impact of Dollarization on Financial Debt — PDF chapter/section (pages and items as listed).*

### 5. The Relationship between Dollarization and Financial Development in Countries with

### 5. The Relationship between Dollarization and Financial Development in Countries with Deposit and Credit Dollarization Above 1 Percent

### I. Introduction and data scope
- Sample: 77 emerging and developing countries over the period 1996–2015.
- Definition: "Dollarization" refers to use of any foreign currency other than the legal tender (not only the dollar). Deposit dollarization = ratio of dollar-denominated deposits to total broad money deposits. Credit dollarization = ratio of dollar-denominated loans to total loans.
- Empirical focus: impact of partial (unofficial) dollarization on financial development measured along three dimensions: financial depth (log of private credit to GDP), financial access (log of bank accounts per 1000 adults), and financial efficiency (bank net interest margin).
- Stylized facts from the sample:
  - Average deposit dollarization across the sample of partially dollarized developing economies was around 30 percent in 2015.
  - Deposit and credit dollarization correlation: 82 percent.
  - High-dollarization examples: Cambodia and Nicaragua with dollarization levels around 90 percent.
  - Low-dollarization examples: China and Bangladesh with close to zero dollarization.
  - Geographic patterns: Eastern European transition economies and Latin America generally more dollarized; MENA and African countries much less dollarized.
  - Deposit dollarization has generally been higher than credit dollarization across the sample.
  - Both deposit and credit dollarization diminished over 1996–2015, with a notable jump between 1996 and 1997 due to inclusion of transition economies.

### II. Main empirical findings
- Overall result: financial dollarization, and deposit dollarization in particular, has a negative impact on financial development.
- Financial depth:
  - Dollarization slows down financial deepening (negative impact on credit-to-GDP).
  - Results robust to alternative specifications and estimation methods.
  - Possible mechanisms:
    - A share of foreign currency deposits may be transferred overseas rather than returned to the domestic economy as private credit, contributing to a shallower domestic financial sector.
    - Additional costs in markets with multi-currency assets/liabilities that inhibit further financial deepening.
  - Interaction with inflation history:
    - The negative impact of dollarization on financial deepening is dampened somewhat in countries with past experiences of very high inflation—i.e., dollarization may mitigate the negative impact of past macroeconomic instability on financial development in some country cases.
- Financial efficiency:
  - Evidence that dollarization is associated with higher net interest margins (NetIntMarg), implying lower financial efficiency.
  - Net interest margins are positively related to levels of dollarization in the estimations, but results vary across specifications, warranting caution.
- Financial access:
  - No evidence of an association between financial dollarization and financial access in the data.
- Consistency with prior literature:
  - Broadly consistent with De Nicolo, Honohan and Ize (2005) and Court, Ozsoz and Rengifo (2012) on negative effects of dollarization on depth; complements mixed findings in the literature (e.g., Levy Yeyati 2006; Reinhart, Rogoff and Savastano 2014).

### III. Theoretical channels and literature synthesis
- Portfolio approach:
  - Dollarization as response to macroeconomic instability (high inflation, exchange rate volatility) via minimum-variance portfolio choice; expectations and monetary policy credibility are key.
  - Persistence of dollarization explained by lack of credible monetary policy/exchange rate regimes.
- Credit-market frictions:
  - Market frictions, devaluation expectations, moral hazard, and incomplete credit markets can drive dollarization.
  - Presence of foreign banks associated with higher credit dollarization.
- De-dollarization policy lessons from literature:
  - Credible macroeconomic stabilization (lower inflation, exchange rate stability).
  - Differential prudential regulations to reduce incentives for foreign currency transactions (e.g., higher provisions for foreign currency loans, tighter capital requirements against open FX positions, differentiated reserve requirements/remuneration).
  - Development of local currency capital markets to provide alternate vehicles for saving and investment.

### IV. Measurement of variables
- Dollarization measures used:
  - Deposit dollarization: ratio of dollar-denominated deposits to total broad money deposits.
  - Credit dollarization: ratio of dollar-denominated loans to total loans.
  - Note: foreign-currency cash holdings and non-bank financial institutions are not included due to data limitations.
- Financial development measures:
  - Financial depth: log of private credit to GDP (PrivCred).
  - Financial access: log of bank accounts per 1000 adults (Accounts).
  - Financial efficiency: bank net interest margin (NetIntMarg) = net interest revenue / average interest-bearing assets.
- Additional constructs:
  - Dollar share of the minimum-variance portfolio (MVP) defined as MVP=[Var(π)+Cov(π,s)]/[Var(π)+Var(s)+2Cov(π,s)] where π = inflation and s = change in the real exchange rate, using past five-year variances and covariance.
  - Aggregate mismatch between deposit and credit dollarization considered in descriptive analysis.

### V. Empirical strategy and identification
- Baseline dynamic panel specification (equation (1) in source) relates financial development indicator F_{i,t} to lagged F_{i,t-1}, dollarization measure (DepDol or CredDol), policy vector P, structural vector S, dollarization-related vector D, country fixed effects, time fixed effects, and error term. The coefficient of interest is β on dollarization.
- Controls:
  - Policy variables (P): log changes of real GDP per capita (GDP growth), CPI inflation (Inflation), banking crisis dummy (Banking Crisis), market share of three largest banks (3 Bank Conc), capital account openness index (KaOpen), external debt to GNI (ExtDebt), Worldwide Governance Indicators composite (WGI).
  - Structural variables (S): log real GDP per capita (GDP pc), log population (Pop), log age dependency ratio (Age Dep ratio).
  - Dollarization determinants (D): MVP, correlation between real GDP growth and real exchange rate changes (REER Cycl), share of foreign banks (Foreign Banks), log nominal effective exchange rate (NEER), Transition dummy for Eastern European transition economies. Robustness checks include Imports to GDP (Imports), Remittances to GDP (Remittances), and short-term nominal interest rate differential (i Diff).
- Endogeneity and estimator choice:
  - Dynamic panel with lagged dependent variable induces Nickell bias if estimated by OLS/fixed effects.
  - Preferred estimator: system GMM (and difference GMM for robustness) to address persistence and endogeneity.
  - Endogeneity assumptions:
    - Log changes in GDP (GDP growth) treated as endogenous.
    - Inflation allowed endogenous with respect to financial depth.
    - Institutional, regulatory and market-structure policy variables treated as predetermined; structural variables treated as exogenous. Predetermined/exogenous variables entered as beginning-of-period values (one-year lags) to reduce simultaneity.
  - Instrument concerns:
    - Instrument proliferation risk addressed by "collapsing" instruments into smaller sets; authors also check robustness with fewer instruments.
    - Acknowledge risk of residual time-varying unobserved heterogeneity leading to bias even with lag-based instruments.

### VI. Robustness and caveats
- Robustness checks: alternative specifications, estimation methods (difference GMM, fixed effects), and fewer instruments confirm core results (dampening caveats noted).
- Caveats:
  - Potential remaining bias from dynamic latent sources of heterogeneity (Bellemare et al. 2017) may increase risk of Type 1 errors.
  - Data limitations: foreign-currency cash holdings and non-bank financial institutions not included.
  - Heterogeneity across countries and model specifications (particularly for efficiency results) implies caution in causal interpretation.

*Source: wp18200 - 5. The Relationship between Dollarization and Financial Development in Countries with Deposit and Credit Dollarization Above 1 Percent.*

### Appendix II).

### Appendix II). V. RESULTS

### Financial Depth
- Deposit dollarization has a statistically significant negative impact on financial development (log of credit to GDP, PrivCred).
- A "1 percentage point increase in deposit dollarization reduces credit to GDP by around 0.4 percent in the short run, and 2–2.5 percent in the long run."
- The observed "5 percentage point reduction in dollarization since its peak has contributed to an increase in credit to GDP by about 10 percent in the past 15 years."
- For highly dollarized economies:
  - Cutting the level of dollarization by half (e.g. from about 90 percent to 45 percent) would potentially increase financial depth by close to 20 percent in the short run.
  - In the long run, halving dollarization could theoretically double financial depth in countries with near complete dollarization.
- Credit dollarization has a much smaller and not statistically significant impact on financial depth.
- The lagged dependent variable is statistically significant in all models, validating the dynamic specification.
- The magnitude difference between models treating dollarization and its determinants as predetermined versus endogenous is small.
- Robustness of financial depth results:
  - Same conclusion from a more parsimonious model with fewer instruments (Table 2 in Appendix II).
  - Same conclusion from a model with only significant control variables (Table 3 in Appendix II).
  - Results produced using the difference GMM estimator (Table 4 in Appendix II) are consistent in sign, though difference GMM generally performs poorly when the dependent variable is highly persistent.
- Diagnostics:
  - Arellano-Bond tests for order 2 serial correlation in the residuals, AB-AR(2), indicate no second-order autocorrelation.
  - Hansen J-test of over-identifying restrictions confirms joint validity of the instruments.
- Possible mechanisms and additional findings:
  - One hypothesis (De Nicolo, Honohan and Ize (2005)) is that part of foreign currency deposits are exported rather than returned to the domestic economy as private credit, leading to a shallower domestic financial sector.
  - A simple correlation plot suggests a positive correlation between deposit dollarization and the share of assets that banks hold abroad, consistent with banks exporting foreign currency deposits instead of extending new loans.
  - Financial assets and liabilities in multiple currencies may create additional frictions and costs in credit markets inhibiting financial deepening.
- Mismatch between deposit and credit dollarization:
  - The aggregate difference (DepDol-CredDol) has a negative effect on financial depth (column 5, Table 1), supporting the Honohan and Shi (2001) hypothesis that excess foreign currency deposits are invested abroad.
  - This result holds when using the difference between foreign currency deposits and credit scaled by GDP (not shown).
- Role of high-inflation history:
  - Interaction analysis with a dummy for countries that experienced annual inflation higher than 250 percent between 1980 and 1997 (for Eastern European transition economies the period is 1995–1998) shows some evidence that deposit dollarization has a less negative or even positive impact on financial depth in economies with a history of very high inflation (see Table 2).
  - This suggests deposit dollarization can, in some country cases with high inflation and macroeconomic instability, facilitate financial deepening.
  - Results in Table 2 confirm that credit dollarization does not have a statistically significant impact on financial deepening.
- Note on alternative approaches:
  - An exploration using financial possibility frontiers (Beck et al. (2008) framework) produced only insignificant or non-robust results (not presented).

### Financial Access and Efficiency
- Financial access measured by number of bank accounts per adult (Accounts):
  - System GMM estimations show no evidence of a link between deposit or credit dollarization and financial access (columns 1 and 2 in Table 3).
  - Difference GMM estimations (Table 4 in Appendix II) find both deposit and credit dollarization have a statistically significant negative impact on financial access, but difference GMM is given less weight due to performance concerns when the dependent variable is highly persistent.
- Financial efficiency approximated by aggregate net interest margin (NetIntMarg):
  - Some evidence that deposit dollarization increases the net interest margin charged by banks (columns 3 and 4 in Table 3).
  - A higher net interest margin signals lower financial sector efficiency, suggesting deposit dollarization may negatively influence banking sector efficiency.
  - Possible explanations include contraction in credit supply when banks shift assets abroad (economies of scale in banking) and higher concentration/monopoly power in banking systems of dollarized economies (supported by a positive coefficient on the 3 Bank Conc. variable).
  - These results vary across specifications and estimators:
    - Negative impact on efficiency is statistically significant in more parsimonious models (Table 3 in Appendix II), but coefficient magnitudes vary.
    - Reducing the number of instruments can render the deposit dollarization coefficient no longer significant (Table 2 in Appendix II).
  - Few control variables are statistically significant, suggesting the model may not fully explain drivers of financial efficiency.
  - Similar inconclusive results are found when analyzing the spread between the deposit and lending rate instead of Net Interest Margin.

### Robustness and Additional Diagnostics
- Robustness checks performed include:
  - Reducing the number of instruments.
  - Using more parsimonious models.
  - Using alternative estimators (difference GMM and fixed effects).
  - Excluding countries with low levels of dollarization.
- Caveats on estimators:
  - Difference GMM generally performs poorly when the dependent variable is highly persistent (Roodman 2009b).
  - Fixed effects (FE) estimates are subject to Nickell bias and should only be regarded as a check on the sign of coefficients.
- Additional control variable findings (baseline):
  - External debt, inflation, foreign banks, banking crises, the nominal effective exchange rate (NEER), low institutional quality, and a high concentration of bank market power (3 Bank conc) have a negative, although not always very significant, impact on financial sector depth.

_ Source: wp18200 - Appendix II)._

### Appendix II. First, we confirm that our results are not driven by instrument proliferation. One

### Appendix II. First, we confirm that our results are not driven by instrument proliferation. One

### Robustness checks on estimation strategy and instruments
- System GMM concern: the number of instruments grows rapidly in the time dimension, especially when dollarization and its determinants are treated as endogenous.
- Strategies to reduce instrument proliferation:
  - Restrict dollarization and its determinants to be predetermined instead of endogenous (Appendix II, Table 1).
  - Allow only the lagged dependent variable, GDP growth, inflation and dollarization to be endogenous and restrict the dollarization determinants to be exogenous or predetermined (Table 2 in Appendix II).
- Key finding:
  - Reducing the instrument count does not change the main conclusions (Appendix II, Table 2).
  - When dollarization determinants and inflation are restricted to be predetermined rather than endogenous, the impact of deposit dollarization on financial depth is much larger and more negative.

### Additional robustness checks
- Excluding insignificant control variables:
  - Results hold when insignificant control variables are excluded from the models (Appendix II, Table 3).
- Alternative estimators:
  - Difference GMM estimations (Appendix II, Table 4) yield results similar to the preferred system GMM.
  - Fixed effects (FE) estimations also support the general conclusions.
- Sample composition check:
  - Excluding countries with very low levels of dollarization (deposit or credit dollarization below 1 percent on average over the sample period) yields a sample of 63 countries; conclusions remain the same (Appendix II, Table 5).

### Summary of core empirical conclusions (from main text reproduced in Appendix II)
- Sample and data scope:
  - Annual data for a sample of 77 emerging market and developing countries over the period 1996–2015.
  - Dollarization levels in the sample range between 0 and 98 percent (for both credit and deposit dollarization).
  - Fully dollarized countries and countries with currencies pegged to the USD (close to zero exchange rate volatility) are excluded.
- Main empirical findings:
  - Deposit dollarization has a negative impact on financial deepening (Private credit by deposit money banks and other financial institutions to GDP).
  - The negative impact may reflect that a share of foreign currency deposits are transferred overseas rather than returned to the domestic economy as private credit.
  - The negative impact of dollarization on financial development is dampened somewhat in countries with past experiences of high inflation.
  - No statistically significant effect of dollarization on financial inclusion found in the data.
  - Deposit dollarization drives the results; no consistent evidence that credit dollarization as such has an impact on financial development.
  - An aggregate mismatch measure (difference between deposit and credit dollarization) is associated with lower levels of financial development, consistent with currency mismatch as a source of financial sector instability.
- Limitations and further research:
  - System GMM is only a partial solution to identification concerns; further research needed to better establish causality from dollarization to financial development.
  - Suggested alternative empirical approach: examine significant de-dollarization events for evidence of their impact on financial development.
  - Further research using more granular data (sectoral, firm, or individual) is needed to better understand how currency mismatch impacts financial development.

### Policy implications and recommendations (as presented)
- Justification for de-dollarization efforts:
  - In addition to concerns about lower monetary policy effectiveness, limits to fiscal flexibility and heightened financial stability risks, dollarization is associated with lower financial depth and banking sector efficiency.
  - Policy efforts to increase the use of the domestic currency in financial transactions are warranted, particularly in countries with a high degree of financial dollarization where gains from de-dollarization could be substantial.
- Components of successful de-dollarization strategies (cited studies and empirical experience):
  - Credible macroeconomic stabilization policies to lower inflation and stabilize the exchange rate.
  - Strengthening economic institutions, particularly monetary policy frameworks that enhance credibility in the face of external shocks.
  - Lowering incentives for financial institutions and economic agents to transact in foreign currencies via prudential policies, including:
    - Raising provisions for foreign currency loans.
    - Tighter capital requirements against open foreign exchange positions.
    - Differentiated reserve requirements and remuneration on foreign currency deposits.
  - Development of local currency financial markets to provide alternate vehicles for longer-term investment and savings.

### Key dataset and summary statistics (selected exact values)
- Sample period: 1996–2015.
- Sample size: 77 emerging market and developing countries (panel is unbalanced).
- Countries retained after excluding deposit or credit dollarization below 1 percent on average: 63.
- Selected summary statistics (from Table 1):
  - Deposit Dollarization: Mean 0.291; Std. Dev 0.245; Min 0; Max 0.984; Obs 1427.
  - Credit Dollarization: Mean 0.252; Std. Dev 0.240; Min 0; Max 0.982; Obs 1070.
  - Private Credit to GDP: Mean 34.82; Std. Dev 28.09; Min 1.61; Max 165.7; Obs 1427.
  - Net Interest Margin: Mean 5.755; Std. Dev 2.931; Min 0.17; Max 25.49; Obs 1406.
  - Accounts per 1000 adults: Mean 942.4; Std. Dev 915.7; Min 2.39; Max 5342; Obs 605.
  - GDP p.c., USD: Mean 4739.1; Std. Dev 6266.4; Min 160.3; Max 49015.9; Obs 1427.
  - Inflation: Mean 0.134; Std. Dev 1.146; Min -0.09; Max 41.45; Obs 1425.
  - Foreign Banks (%): Mean 38.84; Std. Dev 25.66; Min 0; Max 100; Obs 1357.
  - 3 Bank concentration ratio: Mean 0.626; Std. Dev 0.189; Min 0.146; Max 1; Obs 1398.
  - NEER: Mean 405.7; Std. Dev 8200; Min 33.2; Max 304944; Obs 1427.
  - KaOpen: Mean 0.476; Std. Dev 0.330; Min 0; Max 1; Obs 1426.
  - WGI: Mean -0.324; Std. Dev 0.567; Min -1.67; Max 1.25; Obs 1427.
  - REER Cyclicality: Mean 0.104; Std. Dev 0.544; Min -0.99; Max 1.00; Obs 1383.
  - External Debt/GNI: Mean 0.502; Std. Dev 0.335; Min 0.03; Max 2.26; Obs 1394.
  - MVP: Mean 0.045; Std. Dev 0.934; Min -2.69; Max 9.59; Obs 1338.
  - Population (mil.): Mean 51.0; Std. Dev 157.7; Min 0.47; Max 1371.2; Obs 1427.
  - Age Dependency Ratio: Mean 4.101; Std. Dev 0.288; Min 3.47; Max 4.73; Obs 1427.
  - Transition Economy dummy: Mean 0.222; Std. Dev 0.416; Min 0; Max 1; Obs 1427.
  - Banking Crisis Dummy: Mean 0.056; Std. Dev 0.229; Min 0; Max 1; Obs 1423.

*Source: wp18200 - Appendix II. First, we confirm that our results are not driven by instrument proliferation. One*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2018/wp18200.pdf_