## _wp15203

## Source details

**Canonical URL:** [_wp15203](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15203.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15203.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15203.pdf.json)

---

### I. Introduction — scope and stylized facts
- Definition: Financial dollarization measured as the ratio of foreign exchange (FX) deposits and loans to total deposits and loans.
- Associations: Dollarization linked with unstable and high inflation, exchange rate volatility, and undisciplined monetary policy; complicates monetary transmission and increases balance sheet and liquidity risks.
- CCA countries: Armenia, Azerbaijan, Georgia, Kazakhstan, the Kyrgyz Republic, Tajikistan, Turkmenistan, and Uzbekistan.
- CCA dollarization levels (2013Q4):
  - Deposit dollarization: 46 percent on average.
  - Loan dollarization: 40 percent on average.
- Study timeframe: 2001Q1–2014Q1.

### II. Literature review — mechanisms and prior evidence
- Monetary policy and dollarization:
  - Monetary policy autonomy weakened when intermediation is largely in foreign currency; exchange rate changes gain importance relative to interest rates (Ize and Yeyati; IMF 2006; Fischer and others, 2013).
  - Dollarization increases exchange rate pass-through and can limit countercyclical monetary policy, especially under managed exchange rate regimes.
- Financial risks associated with dollarization:
  - Credit risk from private-sector currency mismatches.
  - Solvency risk from large depreciations.
  - Liquidity risk and divergence between onshore and offshore dollar deposit rates.
- Empirical associations from prior literature:
  - Dollarization linked to higher nonperforming loan ratios, deposit volatility, and higher interest rate spreads (De Nicolo and others, 2005; Honohan, 2007).
  - Dollarization associated with lower growth and higher output volatility (Yeyati, 2006).
  - Persistence of dollarization may result from asymmetric exchange rate policies, misguided prudential regulations, and financial stability concerns (Ize and Yeyati, 2005; Erasmus and others, 2009; Rennhack, 2006).
- Prior regional findings:
  - Exchange rate flexibility and reduced bias against depreciation can reduce deposit dollarization (Rennhack and Nozaki, 2006).
  - Prudential measures, local-currency capital market development, and deposit de-dollarization facilitate credit de-dollarization (García-Escribano and Sosa, 2011).

### III. Empirical analysis — model, data, and core empirical findings
- Model and variables:
  - Dependent variables: fxdeposit_t and fxloan_t (deposit and loan dollarization at time t).
  - Key independent variables: lagged dependent variable (persistence), exchange rate (lexr / log(exr)), exchange rate volatility (exrvol: rolling standard deviation over previous four quarters), inflation rate (inf), inflation volatility (infvol: rolling standard deviation over previous four quarters), financial development proxied by the ratio of credit to the private sector to GDP (lcredit), and a depreciation dummy capturing asymmetric exchange rate policy (exr_dum: value of one for depreciation, zero for appreciation).
  - Data sources: IMF IFS, IMF WEO, Haver, national authorities.
  - For Turkmenistan and Uzbekistan (no quarterly GDP), the Denton method used to extrapolate yearly GDP into quarterly series.
  - Sample period: 2001Q1–2014Q1.
- Estimation approach:
  - Dynamic panel data model estimated using Difference-GMM (D-GMM); robust standard errors used.
  - Example Sargan test detail for FX deposit regression: 98 instruments used to estimate eight parameters, yielding 90 over-identifying restrictions; the null hypothesis that the instruments are valid is not rejected at the 5 percent significance (95 percent confidence) level.
- Core empirical findings (CCA-specific and comparative):
  - Exchange rate effects:
    - Frequent depreciations and high exchange rate volatility are associated with a rise in FX deposits, but not with FX loans.
  - Persistence:
    - FX deposits and FX loans exhibit strong persistence (significant lagged dependent variables).
    - Selected coefficients for persistence:
      - L.lfxdeposit: 0.599 (CCA) — Z-statistic (9.20)***.
      - L.lfxloan: 0.397 (CCA) — Z-statistic (4.80)***.
  - Financial depth:
    - Weak financial depth (low credit-to-private-sector-to-GDP) contributes to higher FX deposits and loans.
    - Selected coefficients for lcredit:
      - Deposit regression: lcredit = -0.119 (CCA) — (5.49)***.
      - Loan regression: lcredit = -0.153 (CCA) — (5.22)***.
  - CCA-unique findings:
    - High levels of FX deposits induce banks to lend in foreign currency to maintain matched balance sheets, promoting loan dollarization (lfxdeposit coefficient in loan regressions: 0.327 (CCA) — (4.65)***).
    - Inflation volatility increases both FX deposits and FX loans in the CCA:
      - infvol in deposit regression: 0.014 (CCA) — (5.51)***.
      - infvol in loan regression: 0.008 (CCA) — (2.68)***.
    - Higher inflation per se affects FX loans but not FX deposits in the CCA:
      - inf in loan regression: 0.495 (CCA) — (4.50)***.
      - inf in deposit regression: 0.145 (CCA) — (1.44) (not significant at conventional levels).
    - Asymmetric exchange rate policy (exr_dum) induces higher FX deposits in the CCA by reducing confidence in the domestic currency:
      - exr_dum in deposit regression: 0.012 (CCA) — (2.22)**.
      - exr_dum in loan regression: -0.003 (CCA) — (0.55) (not significant).
- Robustness checks:
  - Results robust across OLS, FE, RE, static and other dynamic panel methods; D-GMM preferred based on Hausman and Sargan tests.
  - Trimming top and bottom 5 percent of dependent-variable observations yields similar results.
  - Pre- and post-global financial crisis (GFC) splits:
    - Deposit persistence important in both periods; coefficient larger pre-GFC (0.916 pre-GFC vs. 0.547 post-GFC in robustness table).
    - Exchange rate changes more important pre-GFC; exchange rate volatility more important post-GFC.
    - Inflation volatility becomes more relevant for deposits after the GFC.
  - Fixed versus flexible exchange rate regimes:
    - For FX deposits: persistence and inflation volatility effects remain; exchange rate depreciations and asymmetric policy matter mainly under fixed regimes (FixedER lexr and exr_d coefficients larger and more significant).
    - For FX loans: inflation and financial development matter mainly under fixed regimes; FX deposits and lagged FX loans matter under both regimes.

### IV. Tables and key statistics
- Summary statistics for regression variables in the CCA (Table 2):
  - Log(fxloan): Obs 364; Mean 3.574621; Std. Dev 1.061646; Min -1.60944; Max 4.753715.
  - Log(fxdeposit): Obs 339; Mean 3.93807; Std. Dev 0.425402; Min 2.661841; Max 4.466441.
  - Log(exr): Obs 392; Mean 2.915796; Std. Dev 2.573393; Min -0.24296; Max 7.709832.
  - Exrvol: Obs 370; Mean 7.824148; Std. Dev 23.99622; Min 0; Max 235.9098.
  - Exrasymmetry: Obs 396; Mean 0.090909; Std. Dev 0.923295; Min -1; Max 1.
  - Inf: Obs 409; Mean 0.085955; Std. Dev 0.075303; Min -0.06521; Max 0.617309.
  - Infvol: Obs 391; Mean 2.549398; Std. Dev 2.673644; Min 0.207859; Max 21.60189.
  - Log(credit): Obs 323; Mean 1.316759; Std. Dev 2.215232; Min -4.87544; Max 4.076545.
- Selected regression coefficients (Tables 3 and 4):
  - Drivers of deposit dollarization (CCA selected):
    - L.lfxdeposit: 0.599 — (9.20)***.
    - lexr: 0.194 — (4.12)***.
    - exrvol: 0.002 — (2.07)**.
    - exr_dum: 0.012 — (2.22)**.
    - infvol: 0.014 — (5.51)***.
    - lcredit: -0.119 — (5.49)***.
    - N: 255.
  - Drivers of loan dollarization (CCA selected):
    - L.lfxloan: 0.397 — (4.80)***.
    - lfxdeposit: 0.327 — (4.65)***.
    - inf: 0.495 — (4.50)***.
    - infvol: 0.008 — (2.68)***.
    - lcredit: -0.153 — (5.22)***.
    - N: 255.
- Robustness diagnostics:
  - Example diagnostics in robustness tables:
    - Hausman: 30.26*** (deposit table); 39.13*** (loan table).
    - Sargan: 97.68 (DGMM deposit); 89.76 (DGMM loan).
    - # instr: 98 (DGMM deposit) and 97 (DGMM loan) in selected columns.

### V. Conclusions and policy recommendations
- Shared drivers across CCA and other regions:
  - Frequent depreciations and high exchange rate volatility drive FX deposits.
  - High inflation increases FX lending.
  - Dollarization persistence and weak financial systems exacerbate financial dollarization.
- CCA-specific policy-relevant insights:
  - Inflation volatility and asymmetric exchange rate policy are particularly important drivers of both FX deposits and loans in the CCA.
  - High FX deposit levels mechanically encourage FX lending by banks (balance-sheet matching).
- Recommended policy menu (macro and financial stability measures):
  - Macroeconomic frameworks:
    - Pursue credible monetary and exchange rate frameworks, low and stable inflation, and deeper domestic financial markets.
    - Consider gradual moves toward more flexible exchange rates and inflation-targeting regimes where appropriate.
    - Avoid fiscal dominance; establish or strengthen an explicit price stability mandate.
  - Financial market development:
    - Deepen financial systems: local currency–denominated securities, credible indexation systems, forward and futures FX markets, non-banking institutions and capital markets, improved credit information systems.
    - Strengthen supervision and prudential regulation to internalize currency-risk costs in contracts.
    - Remove administrative controls on interest rates and apply unbiased taxation between FX and local-currency financial income.
    - Promote domestic bond markets and long-term local-currency bond issuance.
  - Prudential and microprudential measures:
    - Consider reserve and liquidity requirements that discourage FX liabilities (e.g., higher reserve requirements on FX deposits; remunerating local-currency reserves at higher rates than FX reserves).
    - Require routine evaluations of currency risks, higher provisioning or reserves against FX lending to unhedged borrowers, currency-specific credit reporting.
    - Limit FX lending to un-hedged borrowers and raise insurance premiums on FX deposits where appropriate.
  - Implementation considerations and sequencing:
    - Favor market-based de-dollarization over forced de-dollarization to avoid capital flight, financial disintermediation, and banking instability.
    - Sequence reforms carefully: widen exchange rate bands gradually, manage liquidity efficiently, improve monetary transmission via effective policy rates, and prioritize deposit de-dollarization to facilitate loan de-dollarization.
    - Emphasize transparent and effective central bank communication to build public trust and monetary credibility.
    - Recognize that de-dollarization is gradual and may take many years; sustain reform momentum and monitor risks of regulatory arbitrage and disintermediation.

*Source: _wp15203 - References (excerpt).*

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

### _wp15203 - References

### I. Introduction — scope and stylized facts
- Financial dollarization defined as the ratio of foreign exchange (FX) deposits and loans to total deposits and loans.
- Dollarization associated with unstable and high inflation, exchange rate volatility, and undisciplined monetary policy; complicates monetary transmission and increases balance sheet and liquidity risks.
- CCA countries defined as Armenia, Azerbaijan, Georgia, Kazakhstan, the Kyrgyz Republic, Tajikistan, Turkmenistan, and Uzbekistan.
- Dollarization in the CCA:
  - Deposit dollarization stood at 46 percent on average in 2013Q4.
  - Loan dollarization stood at 40 percent on average in 2013Q4.
- Timeframe of the study: 2001Q1–2014Q1.

### II. Literature review — mechanisms and prior evidence
- Monetary policy autonomy weakened when intermediation is largely in foreign currency; exchange rate changes gain importance relative to interest rates (Ize and Yeyati; IMF 2006; Fischer and others, 2013).
- Dollarization increases exchange rate pass-through and can limit countercyclical monetary policy, especially under managed exchange rate regimes.
- Financial risks associated with dollarization:
  - Credit risk from private-sector currency mismatches.
  - Solvency risk from large depreciations.
  - Liquidity risk and divergence between onshore and offshore dollar deposit rates.
- Empirical associations reported in the literature:
  - Dollarization linked to higher nonperforming loan ratios, deposit volatility, and higher interest rate spreads (De Nicolo and others, 2005; Honohan, 2007).
  - Dollarization associated with lower growth and higher output volatility (Yeyati, 2006).
  - Persistence of dollarization can result from asymmetric exchange rate policies, misguided prudential regulations, and financial stability concerns (Ize and Yeyati, 2005; Erasmus and others, 2009; Rennhack, 2006).
- Prior regional findings:
  - Exchange rate flexibility and reduced bias against depreciation can reduce deposit dollarization (Rennhack and Nozaki, 2006).
  - Prudential measures, local-currency capital market development, and deposit de-dollarization facilitate credit de-dollarization (García-Escribano and Sosa, 2011).

### III. Empirical analysis — model, data, and core empirical findings
- Model and variables:
  - Dependent variables: fxdeposit_t and fxloan_t (deposit and loan dollarization at time t).
  - Key independent variables: lagged dependent variable (persistence), exchange rate, exchange rate volatility (rolling standard deviation over previous four quarters), inflation rate, inflation volatility (rolling standard deviation over previous four quarters), financial development proxied by the ratio of credit to the private sector to GDP, and a depreciation dummy capturing asymmetric exchange rate policy (value of one for depreciation, zero for appreciation).
  - Data sources: IMF IFS, IMF WEO, Haver, national authorities.
  - For countries without quarterly GDP (Turkmenistan and Uzbekistan), the Denton method was used to extrapolate yearly GDP into quarterly series.
  - Sample period: 2001Q1–2014Q1.
- Estimation approach:
  - Dynamic panel data model estimated using Difference-GMM (D-GMM); robust standard errors used.
  - Example Sargan test detail for FX deposit regression: 98 instruments used to estimate eight parameters, yielding 90 over-identifying restrictions; the null hypothesis that the instruments are valid is not rejected at the 5 percent significance (95 percent confidence) level.
- Core empirical findings (CCA-specific and comparative):
  - Exchange rate effects:
    - Frequent depreciations and high exchange rate volatility are associated with a rise in FX deposits, but not with FX loans.
  - Persistence:
    - FX deposits and FX loans exhibit strong persistence (significant lagged dependent variables).
  - Financial depth:
    - Weak financial depth (low credit-to-private-sector-to-GDP) contributes to higher FX deposits and loans.
  - CCA-unique findings:
    - High levels of FX deposits induce banks to lend in foreign currency to maintain matched balance sheets, promoting loan dollarization.
    - Inflation volatility increases both FX deposits and FX loans in the CCA.
    - Higher inflation per se affects FX loans but not FX deposits in the CCA.
    - Asymmetric exchange rate policy (allowing depreciation while resisting appreciation) induces higher FX deposits in the CCA by reducing confidence in the domestic currency.
- Robustness checks:
  - Results robust across OLS, static and other dynamic panel methods; D-GMM preferred based on econometric tests (Hausman, Sargan).
  - Trimming top and bottom 5 percent of dependent-variable observations (to control for outliers) yields similar results.
  - Splitting the sample pre- and post-global financial crisis (GFC) shows broadly comparable results:
    - Deposit persistence important in both periods; coefficient larger pre-GFC.
    - Exchange rate changes more important pre-GFC; exchange rate volatility more important post-GFC.
    - Inflation volatility becomes more relevant for deposits after the GFC.
  - Results for fixed versus flexible exchange rate regimes:
    - For FX deposits: persistence and inflation volatility effects remain; exchange rate depreciations and asymmetric policy matter mainly under fixed regimes.
    - For FX loans: inflation and financial development matter mainly under fixed regimes; FX deposits and lagged FX loans matter under both regimes.

### IV. Conclusions and policy recommendations
- Shared drivers across CCA and other regions:
  - Frequent depreciations and high exchange rate volatility drive FX deposits.
  - High inflation increases FX lending.
  - Dollarization persistence and weak financial systems exacerbate financial dollarization.
- CCA-specific policy-relevant insights:
  - Inflation volatility and asymmetric exchange rate policy are particularly important drivers of both FX deposits and loans in the CCA.
  - High FX deposit levels mechanically encourage FX lending by banks.
- Recommended policy menu (macro and financial stability measures):
  - Macroeconomic frameworks:
    - Pursue credible monetary and exchange rate frameworks, low and stable inflation, and deeper domestic financial markets.
    - Consider gradual moves toward more flexible exchange rates and inflation-targeting regimes where appropriate.
    - Avoid fiscal dominance; establish or strengthen an explicit price stability mandate.
  - Financial market development:
    - Deepen financial systems: local currency–denominated securities, credible indexation systems, forward and futures FX markets, non-banking institutions and capital markets, improved credit information systems.
    - Strengthen supervision and prudential regulation to internalize currency-risk costs in contracts.
    - Remove administrative controls on interest rates and apply unbiased taxation between FX and local-currency financial income.
    - Promote domestic bond markets and long-term local-currency bond issuance.
  - Prudential and microprudential measures:
    - Consider reserve and liquidity requirements that discourage FX liabilities (e.g., higher reserve requirements on FX deposits; remunerating local-currency reserves at higher rates than FX reserves).
    - Require routine evaluations of currency risks, higher provisioning or reserves against FX lending to unhedged borrowers, currency-specific credit reporting.
    - Limit FX lending to un-hedged borrowers and raise insurance premiums on FX deposits where appropriate.
  - Implementation considerations and sequencing:
    - Favor market-based de-dollarization over forced de-dollarization to avoid capital flight, financial disintermediation, and banking instability.
    - Sequence reforms carefully: widen exchange rate bands gradually, manage liquidity efficiently, improve monetary transmission via effective policy rates, and prioritize deposit de-dollarization to facilitate loan de-dollarization.
    - Emphasize transparent and effective central bank communication to build public trust and monetary credibility.
    - Recognize that de-dollarization is gradual and may take many years; sustain reform momentum and monitor risks of regulatory arbitrage and disintermediation.

*Source: _wp15203 - References (excerpt).*

### REFERENCES

### _wp15203 - REFERENCES

### References list (selected)
- Acosta-Ormaechea, Santiago, and David Coble (2011) “The Monetary Transmission in Dollarized and Non-Dollarized Economies: The Cases of Chile, New Zealand, Peru and Uruguay,” IMF Working Paper WP/11/87.
- Anderson, T.W., and C. Hsiao (1981) “Estimation of Dynamic Models with Error Components,” Journal of the American Statistical Association 76 (375): 589606.
- Anderson, T.W., and C. Hsiao (1982) “Formulation and Estimation of Dynamic Models Using Panel Data,” Journal of Econometrics 18 (1): 4782.
- Arellano, M., and S. Bond (1991) “Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations,” The Review of Economic Studies 58 (2): 27797.
- Cayazzo, J., A. Pascual, E. Gutierrez, and S. Heysen (2006) “Toward an Effective Supervision of Partially Dollarized Banking Systems,” IMF Working Paper WP/06/32.
- De Nicoló, Gianni, Patrick Honohan, and Alain Ize (2005) “Dollarization of Bank Deposits: Causes and Consequences,” Journal of Banking and Finance 29 (7): 1697–1727.
- Di Fonzo, Tommaso, and Marco Marini (2012) “On the Extrapolation with the Denton Proportional Benchmarking Method,” IMF Working Paper WP/12/169.
- Erasmus, L., J. Leichter, and J. Menkulasi (2009) “De-dollarization in Liberia: Lessons from Cross-country Experience,” IMF Working Paper WP/09/37.
- Fischer, Felix, Charlotte Lundgren, and Samir Jahjah (2013) “Making Monetary Policy More Effective: The Case of the Democratic Republic of the Congo,” IMF Working Paper WP/13/226.
- García-Escribano, M. (2010) “Peru: Drivers of De-dollarization,” IMF Working Paper WP/10/169.
- García-Escribano, M. and S. Sosa (2011) “What is Driving Financial De-dollarization in Latin America?” IMF Working Paper WP/11/10.
- Honohan, Patrick (2007) “Dollarization and Exchange Rate Fluctuations,” World Bank Policy Research Working Paper No. 4172.
- IMF (2015) “Azerbaijan: Selected Issues Paper,” forthcoming IMF Country Report.
- IMF (2014) “Kazakhstan: Selected Issues Paper,” IMF Country Report 14/243.
- IMF (2007) “Peru: Selected Issues Paper,” IMF Country Report 07/53.
- IMF (2006) “Inflation Targeting and the IMF,” IMF Board Paper SM/06/33.
- Ize, A., and E. L. Yeyati (2005) “Financial Dollarization: Is It for Real?” IMF Working Paper WP/05/187.
- Ize, A., and E. L. Yeyati (2003) “Financial Dollarization,” Journal of International Economics 59 (2): 323347.
- Ize, A., and E. L. Yeyati (1998) “Dollarization of Financial Intermediation: Causes and Policy Implications?” IMF Working Paper WP/98/28.
- Kokenyne, A., J. Ley, and R. Veyrune (2010) “De-dollarization,” IMF Working Paper WP/10/188.
- Leiderman, L., R. Maino, and E. Parrado (2006) “Inflation Targeting in Dollarized Economies,” IMF Working Paper WP/06/157.
- Luca, Alina and Iva Petrova (2008) “What Drives Credit Dollarization in Transition Economies?,” Journal of Banking and Finance 32 (5): 858869.
- Mwase, Nkunde and Francis Y. Kumah (2015) “Revisiting the Concept of Dollarization: The Global Financial Crisis and Dollarization in Low-Income Countries,” IMF Working Paper WP/15/12.
- Neanidis, Kyriakos C., Christos S. Savva (2009) “Financial Dollarization: Short-run Determinants in Transition Economies,” Journal of Banking and Finance 33 (10): 1860–1873.
- Rennhack, R. and M. Nozaki (2006) “Financial Dollarization in Latin America,” IMF Working Paper WP/06/7.
- Sahay, Ratna and Carlos Végh (1995) “Dollarization in Transition Economies: Evidence and Policy Implications,” IMF Working Paper WP/95/96.
- Yeyati, Eduardo Levy (2006) “Financial Dollarization: Evaluating the Consequences,” Economic Policy 21 (45): 61118.

### Table 1: Exchange Rate and Monetary Frameworks in the CCA (high-level)
- Conventional peg: Comoros, South Sudan, TKM, Venezuela.
- Stabilized arrangement: KAZ, Macedonia, Vietnam; Bangladesh, Congo, TJK; Angola, AZE.
- Crawl-like arrangement: Croatia, Jamaica, China, UZB; ARM, Dominican Rep.; Belarus.
- Other managed arrangement: Nigeria, Rwanda; Czech Rep.; KGZ, Russia.
- Floating: Brazil, GEO, Romania, South Africa, Turkey; India.
- Free-floating: Chile, Mexico, Poland.
- Note: 1 These countries maintain a de facto exchange rate anchor to the U.S. dollar.

### Table 2: Summary statistics for regression variables in the CCA
- Log(fxloan): Obs 364; Mean 3.574621; Std. Dev 1.061646; Min -1.60944; Max 4.753715.
- Log(fxdeposit): Obs 339; Mean 3.93807; Std. Dev 0.425402; Min 2.661841; Max 4.466441.
- Log(exr): Obs 392; Mean 2.915796; Std. Dev 2.573393; Min -0.24296; Max 7.709832.
- Exrvol: Obs 370; Mean 7.824148; Std. Dev 23.99622; Min 0; Max 235.9098.
- Exrasymmetry: Obs 396; Mean 0.090909; Std. Dev 0.923295; Min -1; Max 1.
- Inf: Obs 409; Mean 0.085955; Std. Dev 0.075303; Min -0.06521; Max 0.617309.
- Infvol: Obs 391; Mean 2.549398; Std. Dev 2.673644; Min 0.207859; Max 21.60189.
- Log(credit): Obs 323; Mean 1.316759; Std. Dev 2.215232; Min -4.87544; Max 4.076545.

### Table 3: Regression estimates — Drivers of deposit dollarization (selected coefficients)
- Dependent variable context: Columns labeled CCA, All, EUR, LAC, DEVEMASIA, EMDE.
- L.lfxdeposit: 0.599 (CCA) — Z-statistic (9.20)***; 0.426 (All) — (7.01)***; 0.184 (EUR) — (6.14)***; 0.289 (LAC) — (3.15)***; 0.525 (DEVEMASIA) — (3.83)***; 0.429 (EMDE) — (7.33)***.
- lexr: 0.194 (CCA) — (4.12)***; 0.162 (All) — (2.45)**; -0.051 (EUR) — (0.93); -0.336 (LAC) — (1.25); 0.535 (DEVEMASIA) — (1.92)*; 0.161 (EMDE) — (2.43)**.
- exrvol: 0.002 (CCA) — (2.07)**; -0.000 (All) — (0.81); 0.000 (EUR) — (0.55); 0.001 (LAC) — (1.70)*; 0.023 (DEVEMASIA) — (0.39); -0.000 (EMDE) — (0.85).
- exr_dum: 0.012 (CCA) — (2.22)**; -0.004 (All) — (0.84); -0.006 (EUR) — (1.18); -0.038 (LAC) — (2.22)**; 0.005 (DEVEMASIA) — (0.54); -0.004 (EMDE) — (0.85).
- inf: 0.145 (CCA) — (1.44); 0.072 (All) — (0.61); -0.068 (EUR) — (0.49); 0.779 (LAC) — (0.80); 0.648 (DEVEMASIA) — (0.93); 0.075 (EMDE) — (0.63).
- infvol: 0.014 (CCA) — (5.51)***; 0.000 (All) — (0.02); -0.006 (EUR) — (1.96)*; -0.044 (LAC) — (2.05)**; -0.009 (DEVEMASIA) — (0.51); 0.000 (EMDE) — (0.07).
- lcredit: -0.119 (CCA) — (5.49)***; -0.063 (All) — (2.46)**; -0.234 (EUR) — (12.44)***; -0.037 (LAC) — (0.24); 0.230 (DEVEMASIA) — (1.35); -0.064 (EMDE) — (2.45)**.
- _cons: 1.158 (CCA) — (4.17)***; 1.606 (All) — (5.98)***; 2.642 (EUR) — (19.86)***; 3.314 (LAC) — (1.99)**; -0.630 (DEVEMASIA) — (0.64); 1.604 (EMDE) — (6.10)***.
- N: 255 (CCA); 886 (All); 278 (EUR); 83 (LAC); 55 (DEVEMASIA); 882 (EMDE).
- Significance notes: * p<0.1; ** p<0.05; *** p<0.01. Z-statistics in parenthesis.

### Table 4: Regression estimates — Drivers of loan dollarization (selected coefficients)
- L.lfxloan: 0.397 (CCA) — (4.80)***; 1.070 (All) — (14.28)***; 0.998 (EUR) — (22.89)***; 0.815 (LAC) — (6.84)***; 0.546 (DEVEMASIA) — (5.78)***; 0.767 (EMDE) — (30.28)***.
- lfxdeposit: 0.327 (CCA) — (4.65)***; -0.014 (All) — (0.46); 0.001 (EUR) — (1.57); -0.006 (LAC) — (0.14); -0.108 (DEVEMASIA) — (0.52); 0.011 (EMDE) — (0.47).
- lexr: 0.043 (CCA) — (0.70); 0.144 (All) — (1.78)*; -0.026 (EUR) — (1.34); 0.073 (LAC) — (1.18); 0.182 (DEVEMASIA) — (1.80)*; -0.008 (EMDE) — (0.07).
- exrvol: 0.001 (CCA) — (0.91); -0.000 (All) — (1.68)*; -0.000 (EUR) — (1.54); 0.000 (LAC) — (0.83); 0.007 (DEVEMASIA) — (2.09)**; -0.000 (EMDE) — (1.33).
- exr_dum: -0.003 (CCA) — (0.55); -0.003 (All) — (0.39); 0.022 (EUR) — (3.80)***; -0.001 (LAC) — (0.18); 0.011 (DEVEMASIA) — (4.74)***; 0.002 (EMDE) — (0.46).
- inf: 0.495 (CCA) — (4.50)***; 0.289 (All) — (2.01)**; 0.086 (EUR) — (1.39); -0.653 (LAC) — (1.32); 0.488 (DEVEMASIA) — (2.71)***; 0.285 (EMDE) — (1.82)*.
- infvol: 0.008 (CCA) — (2.68)***; 0.002 (All) — (0.57); -0.000 (EUR) — (0.31); 0.007 (LAC) — (0.53); -0.003 (DEVEMASIA) — (1.10); 0.005 (EMDE) — (0.85).
- lcredit: -0.153 (CCA) — (5.22)***; -0.070 (All) — (2.07)**; -0.018 (EUR) — (2.20)**; 0.104 (LAC) — (2.52)**; -0.085 (DEVEMASIA) — (1.17); -0.101 (EMDE) — (1.84)*.
- _cons: 0.959 (CCA) — (2.56)**; -0.409 (All) — (1.09); 0.071 (EUR) — (0.49); -0.092 (LAC) — (0.13); 2.085 (DEVEMASIA) — (1.34); 0.987 (EMDE) — (2.50)**.
- N: 255 (CCA); 620 (All); 89 (EUR); 29 (LAC); 31 (DEVEMASIA); 616 (EMDE).
- Significance notes: * p<0.1; ** p<0.05; *** p<0.01. Z-statistics in parenthesis.

### Table 5: Robustness checks — Determinants of deposit dollarization (selected)
- Estimation methods and columns: OLS, OLS, FE, RE, DGMM, SGMM, Trimmed, Pre-GFC, Post-GFC, FixedER, FlexibleER.
- L.lfxdep coefficients: 0.846 (OLS) — (16.57)***; 0.786 (OLS) — (12.67)***; 0.846 (FE) — (25.78)***; 0.992 (RE) — (57.12)***; 0.599 (DGMM) — (9.20)***; 0.972 (SGMM) — (67.07)***; 0.583 (Trimmed) — (8.75)***; 0.916 (Pre-GFC) — (6.34)***; 0.547 (Post-GFC) — (6.49)***; 0.690 (FixedER) — (11.34)***; 0.681 (FlexibleER) — (8.03)***.
- lexr coefficients: 0.047, 0.046, 0.047, 0.001, 0.194, -0.000, 0.237, 0.490, 0.201, 0.179, -0.016 with corresponding Z-statistics shown in table.
- exrvol coefficients: 0.001, 0.000, 0.001, -0.001, 0.002, -0.001, 0.002, 0.001, 0.003, 0.002, 0.002.
- exr_d coefficients: 0.022, 0.017, 0.022, 0.016, 0.012, 0.018, 0.014, 0.014, 0.008, 0.025, 0.012.
- inf coefficients: 0.266, 0.359, 0.266, 0.197, 0.145, 0.193, 0.138, 0.227, 0.081, 0.045, 0.111.
- infvol coefficients: 0.007, 0.005, 0.007, 0.005, 0.014, 0.005, 0.014, -0.004, 0.014, 0.006, 0.014.
- lcredit coefficients: -0.012, 0.017, -0.012, -0.003, -0.119, -0.004, -0.113, 0.048, -0.189, -0.093, -0.040.
- _cons: 0.028, 0.162, 0.453, -0.000, 1.158, 0.084, 1.164, -0.760, 1.416, 0.970, 1.269.
- Model diagnostics: Ctryd Y in relevant columns; Timed N or Y noted per column. R2 values reported (example: 0.97, 0.98, 0.79, 0.96 in table). Hausman 30.26***. Sargan 97.68 and 176.86**. # instr 98 and 145 in DGMM/SGMM columns.
- N values by column: 263, 263, 263, 263, 255, 263, 217, 75, 174, 106, 149.
- Significance notes: * p<0.1; ** p<0.05; *** p<0.01.

### Table 6: Robustness checks — Determinants of loan dollarization (selected)
- L.lfxloa coefficients across methods: 0.812 (OLS) — (9.95)***; 0.806 (OLS) — (17.49)***; 0.812 (FE) — (24.92)***; 0.997 (RE) — (111.08)***; 0.397 (DGMM) — (4.80)***; 0.957 (SGMM) — (48.24)***; 0.437 (Trimmed) — (7.58)***; 0.526 (Pre-GFC) — (4.85)***; 0.423 (Post-GFC) — (3.82)***; 0.452 (FixedER) — (6.85)***; 0.430 (Flexible-ERR) — (3.87)***.
- lfxdep: 0.272, 0.272, 0.272, 0.061, 0.327, 0.107, 0.360, 0.197, 0.331, 0.344, 0.249 (Z-statistics shown in table).
- lexr: -0.042, -0.029, -0.042, 0.002, 0.043, 0.009, -0.064, 0.002, 0.096, 0.040, -0.100.
- exrvol: 0.001, 0.000, 0.001, 0.001, 0.001, 0.001, 0.001, 0.001, 0.002, 0.001, 0.001.
- exr_d: 0.008, 0.004, 0.008, 0.011, -0.003, 0.010, 0.003, 0.007, -0.008, 0.012, -0.010.
- inf: 0.394, 0.384, 0.394, 0.322, 0.495, 0.305, 0.318, 0.138, 0.484, 0.341, 0.206.
- infvol: -0.003, -0.006, -0.003, -0.004, 0.008, 0.002, 0.006, -0.005, 0.007, -0.000, 0.003.
- lcredit: -0.005, 0.010, -0.005, -0.001, -0.153, -0.015, -0.159, -0.091, -0.259, -0.180, -0.010.
- _cons: -0.353, -0.308, -0.299, -0.274, 0.959, -0.321, 1.013, 1.158, 0.838, 0.835, 1.396.
- Model diagnostics: Ctryd Y in relevant columns; Timed N or Y noted per column. R2 values reported (example: 0.99, 0.99, 0.87, 0.99). Hausman 39.13***. Sargan 89.76. # instr 97 and 144 in DGMM/SGMM columns.
- N values by column: 263, 263, 263, 263, 255, 263, 219, 75, 174, 107, 148.
- Significance notes: * p<0.1; ** p<0.05; *** p<0.01.

*Content adapted from _wp15203 - REFERENCES (IMF).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15203.pdf_