## Dollarization in Sub-Saharan Africa: Experience and Lessons

## Source details

**Canonical URL:** [Dollarization in Sub-Saharan Africa: Experience and Lessons](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/dp/2015/_afr1504.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/dp/2015/_afr1504.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/dp/2015/_afr1504.pdf.json)

---

### Introduction and scope
- Dollarization defined: use of foreign currencies as medium of exchange, store of value, or unit of account.
- Dollarization in SSA remains significant and persistent:
  - "over 30 percent rates for both bank loans and deposits" (general statement).
- Study period and focus: analysis of dollarization in SSA over the past decade to assess vulnerabilities and inform policy options.
- Recent context highlights:
  - Macroeconomic and financial landscape improved in last decade.
  - Financial deepening progressing though systems remain underdeveloped in many countries.

### Stylized facts and key statistics (2001–12)
- Deposit dollarization:
  - Average deposit dollarization (excluding zeros): 29 percent in 2001 and 31.6 percent in 2012.
  - Including all countries but Zimbabwe: 14 percent in 2001 and 16 percent in 2012.
- Loan dollarization:
  - Average loan dollarization: 26.1 percent in 2001 and 34.8 percent in 2012.
  - Including all countries but Zimbabwe: 7 percent in 2001 and 12 percent in 2012.
- Regional and cross‑country positions:
  - SSA was the second most dollarized region after Europe and Central Asia.
  - "In 2001, 26 percent of total bank credit in SSA was denominated in foreign currencies; this ratio increased to 34 percent in 2012."
  - Most dollarized SSA countries: Democratic Republic of the Congo, Liberia, São Tomé and Príncipe, Angola, Mozambique, Zambia.
- Debt currency composition:
  - Average ratio of foreign exchange debt to total debt in SSA: around 78 percent in 2001, decreased to around 60 percent at end-2012.
- Persistence measures:
  - Dollarization rates in the second half 2010s equal to 79 percent of the rates in the early 2000s (rest of world); in SSA this proportion increases to 103 percent (difference significant at 5 percent).
  - Panel unit root test: accepts null that foreign exchange deposit shares have unit roots for all SSA countries; rejects same null for rest of the world at 1 percent significance.
- Selected country-level time series (excerpted):
  - Angola FX Loans to total Loans (selected sequence): 56.96 57.11 57.87 62.23 71.34 71.11 71.09 60.59 64.73 64.10 52.38 43.81
  - Democratic Republic of Congo FX Deposits to total deposits (selected sequence): 81.90 89.73 88.65 90.19 91.64 95.06 93.89 94.26 93.82 90.91 91.11 87.64
  - Net open FX position - private sector (selected Angola sequence): -50.40 -52.97 -27.50 7.70 55.16 57.05 172.35 189.07 63.62 54.63 22.88 19.84
- Sectoral patterns:
  - Firms generally have more foreign currency loans than households.
  - Negative net open positions for other resident sectors in every country with available data in 2012 (definition: foreign currency loans minus foreign currency deposits, divided by total deposits).

### Drivers and econometric findings
- Primary drivers confirmed:
  - Inflation and nominal exchange rate depreciation are key drivers (currency substitution) for deposits and loans.
  - Political instability, dependence on primary commodity exports, and limited financial market development also contribute.
- Portfolio and market features:
  - Higher real deposit rates (domestic) associated with lower deposit dollarization.
  - Higher M2/GDP (financial depth) unambiguously associated with lower deposit dollarization.
  - GDP per capita: higher GDP p.c. associated with marked decline in deposit dollarization, particularly in SSA.
  - Capital account openness: no significant impact on deposit dollarization for SSA sample.
  - External debt to GDP: tends to be associated with higher share of foreign currency deposits.
  - Exports/GDP: in SSA many results point to a significant positive impact on deposits in foreign exchange.
  - Oil exports to GDP: in SSA associated with higher deposit dollarization in many specifications.
- Selected stylized regression relationships (figures/equations reported):
  - Deposits dollarization 2011 versus Average inflation 2001–2011: y = 1.31x + 16.43; R² = 0.159
  - Loans dollarization 2011 versus Average inflation 2001–2011: y = 1.17x + 16.03; R² = 0.139
  - Deposits dollarization 2011 versus Average depreciation 2001–2011: y = 1.87x + 23.81; R² = 0.147
  - Loans dollarization 2011 versus Average depreciation 2001–2011: y = 1.84x + 23.33; R² = 0.131
  - Deposits dollarization 2011 versus Average ratio of M2 to GDP 2001–2011: y = -0.45x + 46.57; R² = 0.3195
  - Loans dollarization 2011 versus Average ratio of M2 to GDP 2001–2011: y = -0.42x + 44.03; R² = 0.254
- Loan versus deposit dollarization contrasts:
  - Current account openness: no impact on deposit dollarization but positive impact on loan dollarization (affects banks’ ability to obtain foreign funds).
  - Currency substitution evidence stronger and more consistent for loan dollarization in Rest of World; for SSA results less conclusive.
  - Portfolio effects differ across deposits and loans and between SSA and RoW.

### Case studies and examples of de‑dollarization outcomes
- Angola (Box 3.1, as of 2014):
  - A decade earlier nearly three-fourths of deposits and two-thirds of credit were in foreign currency; ratios closer to 40 percent in 2014.
  - Measures implemented 2013–early 2014 included oil sector foreign exchange law (mid-2013), tighter bank FX exposure limits, and reserve requirement adjustment in early 2014 (BNA reduced reserve requirements on kwanza deposits below those on dollar deposits).
  - Evidence through 2014: from June to December (2013) dollar deposits fell from 53 to 42 percent; dollar credit fell from 41 to 37 percent.
  - By 2014 inflation reduced to single digits (below 8 percent in 2013) and FX rate stabilized.
- São Tomé and Príncipe (Box 3.2):
  - Deposits in foreign currency rose from 35 percent in 2001 to about 57 percent by 2007; foreign currency loans accounted for almost ¾ of credit.
  - March 2010: required 90 percent of reserve requirements be paid in local currency; lowered required reserve ratio for domestic currency to 18 percent while keeping FX statutory requirements at 21 percent; peg to the euro in January 2010.
  - Result: foreign currency loans fell from 77 percent in 2010 to 58 percent in 2012; banks sought local‑currency funding in 2013.

### What distinguishes successful de‑dollarization (cross‑country evidence)
- Definitions:
  - "Dollarized" if FCD/TD > 30 percent.
  - "De-dollarized (successful)" if: (1) three-year average dollarization declined by at least 20 percent between 2001–03 and 2010–12; and (2) average share of foreign currency loans did not increase between these periods.
- Outcomes across 194 countries:
  - 42 initially dollarized; 11 successful de-dollarizations; 31 remained dollarized.
- Comparative averages (selected numeric contrasts):
  - Initial (2001–03) dollarization: Successful 67.4; Unsuccessful 48.4
  - Initial inflation (percent): Successful 21.6; Unsuccessful 9.4
  - After 2003 average inflation (2004–12): Successful 8.5; Unsuccessful 8.9
  - Real GDP growth (2004–12): Successful 7.3; Unsuccessful 4.5
  - Current account balance (2004–12): Successful 0.3; Unsuccessful -7.6
  - Stock of public debt (2001–03): Successful 60.1; Unsuccessful 54.6
  - Stock of external debt (2001–03): Successful 63.8; Unsuccessful 59.3
  - Net changes (2001–03 to 2010–12):
    - Inflation difference: Successful -14.4; Unsuccessful -1.8
    - Public debt difference: Successful -32.4; Unsuccessful -14.4
    - External debt difference: Successful -27.7; Unsuccessful -4.6
- Stylized predictors of success:
  - Higher initial dollarization (>65 percent) increases probability of success (43 percent higher).
  - Current account surplus or deficit below 3 percent of GDP in 2001–03 increases probability of success by about 30 percent.
  - Initial public or external debt above 40 percent of GDP increased probability of success by 25 percent.
  - Most successful countries experienced a nominal appreciation of at least 15 percent at some point between 2001 and 2012.
  - Fiscal balance above 2 percent of GDP increased probability of success; significant reduction in external debt common among de-dollarizers.

### Policy implications and recommended strategy
- Principal conclusion: de-dollarization is difficult and requires long, consistent, credible stabilization.
- Key ingredients for successful de-dollarization (explicit list from source):
  1. A sustained and credible macroeconomic stabilization.
  2. A proper liberalization effort encouraging de-dollarization by removing distortionary direct controls or constraints that discourage use of the local currency.
  3. Prudential measures that reduce banks’ incentives to borrow and lend in foreign currencies (examples: asymmetric reserve requirements; provisioning for currency-induced credit risks).
- Complementary market-based measures:
  - Development of local-currency securities markets (initially government bonds, later longer maturities).
  - Strengthening payments system and central bank liquidity management.
  - Issuance of medium-term government bonds in local currency and development of a benchmark yield curve.
  - Excluding foreign currency deposits from deposit insurance schemes (as a regulatory option).
- Cautions on direct administrative measures:
  - Mandatory conversion and forced de-dollarization have sometimes backfired (capital flight, disintermediation); administrative measures should be complements to market-based stabilization.
- Role of prudential instruments and sequencing:
  - Asymmetric reserve requirements and higher provisioning for FX loans can reduce FX lending incentives.
  - Credibility of monetary policy generally needs to be reestablished first; sequencing matters.

### Concluding summary for SSA
- Dollarization in SSA is more persistent than elsewhere; few successful de-dollarizations have occurred.
- Effective de-dollarization strategies for SSA require:
  - time, persistent coordination, and a mix of sound macroeconomic policies, market-based incentives, and microprudential measures.
- Empirical evidence supports central policy levers: lower inflation, reduced exchange rate volatility, deeper domestic financial markets (higher M2/GDP), improved governance (Polity scores), and fiscal consolidation as core to reducing dollarization.

*Prepared by a staff team led by Mauro Mecagni; Departmental Paper Series (April 2015). International Monetary Fund.*

### 1. Dollarization – Africa, Sub-Saharan.   2. Foreign exchange – Africa, Sub-Saharan.  3. Africa,

### Dollarization in Sub-Saharan Africa: Experience and Lessons

### Introduction: scope and significance
- Dollarization is defined as the use of foreign currencies as a medium of exchange, store of value, or unit of account.
- Dollarization is prominent in sub-Saharan Africa (SSA), remaining significant and persistent at over 30 percent rates for both bank loans and deposits.
- The study analyzes dollarization in SSA over the past decade to assess vulnerabilities and inform policy options.
- Recent context in SSA:
  - Macroeconomic and financial landscape has improved significantly in the last decade.
  - Chronic macroeconomic instability is receding; conflicts and social turmoil are more circumscribed.
  - Institutions and governance are improving; budgetary and monetary policies are becoming more sustainable.
  - Financial deepening is progressing, though the financial system remains underdeveloped in many countries.
- Observed manifestations of dollarization in SSA:
  - Moderate increases in residents’ holding of deposits denominated in foreign currency—mainly the U.S. dollar.
  - Real sector dollarization: use of foreign currencies to pay a large share of purchases of goods and services and wages (examples: Democratic Republic of the Congo; until recently, Angola).
  - Currency substitution: use of foreign currency as a means of payment (example: Zimbabwe).
  - Heterogeneity across SSA: some countries, notably in Western Africa, see little or no dollarization.

### Risks, trade-offs, and potential benefits
- Adverse effects and policy challenges posed by dollarization:
  - Exposes balance sheets of the public sector, private enterprises, and households to exchange rate risks when assets and liabilities in foreign currency are mismatched.
  - Reduces authorities’ capacity to use monetary policy and complicates the central bank’s lender-of-last-resort function.
  - Weakens structural fiscal balance and fiscal flexibility by reducing scope for seigniorage.
  - Reduces governments’ ability to issue medium- and long-term debt in domestic currency—known as the original sin—amplifying vulnerabilities to shocks and macroeconomic/output fluctuations.
- Potential benefits in specific circumstances:
  - In economies with high and volatile inflation, allowing foreign currency deposits may encourage residents to transact through the banking system rather than withdraw savings abroad or hold nonmonetary assets.
  - Use of foreign currency can provide credibility to disinflation efforts, particularly in situations of very high inflation or hyperinflation, via exchange rate–based stabilization programs.
  - In highly dollarized economies the policy debate often centers on whether to fully dollarize, fully de-dollarize, or maintain the status quo.

### The uniqueness of SSA: structural factors
- Key characteristics that set SSA apart:
  - Political instability: higher incidence of conflict and longer-lasting impact due to ethnic fragmentation and weak political institutions; risks of expropriation encouraged capital flight.
  - Limited financial deepening: shallow financial systems with limited local-currency investment opportunities; banks often lend to the sovereign or large resource-exporting companies.
  - Limited private sector development: smaller formal private sector, limited bankable projects, and inadequate demand for formal credit in local currency; pervasive role of the state and state enterprises.
- Historical consequences and recent shifts:
  - Economic take-offs in many SSA countries tended to be short with frequent reversals from external shocks (terms of trade) or domestic turmoil.
  - With tight FX restrictions, small private export sectors, limited monetization, and large informal sectors, foreign currency use manifested mainly via capital flight (residents moving savings overseas).
  - Since the 1990s, political stability and market-friendly reforms improved; debt levels, inflation, and budget deficits were curbed.
  - Last decade: SSA reversed the decline of its share in world trade; foreign direct investment inflows were substantial—particularly in energy and mining—and financial integration is starting to gain ground.
  - Commodity exporters in SSA: most commodities are priced and invoiced in dollars; transactions associated with foreign trade are effectively dollar denominated, and resource-rich countries have registered higher use of foreign currency.
  - In some countries commercial banks have been allowed to collect foreign currency deposits from residents and nonresidents, contributing to increased use of foreign currency, particularly the U.S. dollar.

### Key results and policy implications (summary)
- Main findings:
  - Dollarization has been more persistent in SSA than in the rest of the world.
  - There have been few successful episodes of de-dollarization in SSA.
    - Angola has recorded a downward trend in dollarization.
    - Democratic Republic of the Congo, Liberia, and São Tomé and Príncipe are still experiencing the same levels of dollarization as in 2000.
  - Inflation and nominal exchange rate depreciation are confirmed as key drivers of dollarization, supporting the currency substitution argument that foreign currencies are used to hedge against inflation risk.
  - Political instability, strong dependence on primary commodities exports, and limited financial market development also contribute to SSA’s dollarization profile.
- Policy implications and strategies:
  - Successful de-dollarization requires time and persistent, coordinated efforts combining sound macroeconomic policies, market-based incentives, and microprudential measures.
  - Mandatory measures and direct controls tend to be effective only as complements to a market-based strategy.
  - Countries that significantly reduced foreign currency use implemented sustained disinflation and stabilization processes that increased the attractiveness of using local currency.

### Structure of the study (reader’s map)
- Chapter 1: Stylized facts on dollarization in SSA; dynamics over the last decade from a balance sheet perspective; trends and comparisons with other regions.
- Chapter 2: Determinants of deposit and loan dollarization in SSA; panel-data benchmarking with other regions; comparison of dollarization levels and macroeconomic performance by country groups; analysis of currency mismatches and dynamics.
- Chapter 3: Factors and policy measures promoting de-dollarization; case studies of de-dollarization; policy implications for SSA countries.
- Final chapter: Main conclusions on how dollarized countries in SSA could de-dollarize or mitigate the consequences of dollarization.

*Prepared by a staff team led by Mauro Mecagni; Departmental Paper Series (April 2015).*

### Box 1. A Brief Literature Review

### Box 1. A Brief Literature Review

### Drivers of dollarization
- Large macroeconomic imbalances and high inflation drove dollarization in countries such as Chile, Colombia, and Peru, where substitution of domestic currency with the U.S. dollar followed periods of macroeconomic turbulence and high inflation.
- Financial repression and capital controls contributed to dollarization in many Latin American economies in the 1970s and 1980s and in some SSA countries (for example, Democratic Republic of the Congo, Liberia, and Nigeria).
- Use of the dollar as an anchor for macrostability: countries (for example, Ecuador, El Salvador) adopted the dollar as legal tender to “import” the credibility of U.S. monetary institutions and escape long histories of monetary and financial disorder.
- Underdeveloped financial markets led domestic borrowers to contract debt in foreign currencies due to a lack of domestic currency alternatives in incomplete financial markets.

### Conceptual distinctions and early literature
- Two types of dollarization were distinguished:
  - Currency substitution: use of foreign currency as medium of exchange.
  - Asset substitution: use of foreign currency as store of value (Levy‑Yeyati 2006).
- Early literature (motivated by high inflation in Latin America) emphasized currency substitution and concluded monetary policy is ineffective where foreign currencies substitute domestic currency (Miles 1978; Brillenbourg and Schadler 1980; Girton and Roper 1981; Ortiz 1983).
- Expected inflation, reflected in the nominal exchange rate, underpins currency substitution; elasticity of substitution rises when perceived risk of sharp domestic currency value changes is greater (floating or adjustable predetermined exchange rates).

### Persistence and the shift to asset/portfolio/institutional views
- Persistence of dollarization in the 1990s, despite lower inflation, challenged the currency substitution view; persistence was sometimes attributed to long-lasting memories of past inflation.
- Asset substitution literature: Ghosh and others (1998) showed highly dollarized countries or those with currency boards enjoyed lower inflation, more stable exchange rates, and faster economic growth than nondollarized economies, while losing benefits of flexible monetary policy.
- Post-1990s literature split into three strands:
  - Portfolio view: dollarization as optimal portfolio choice given distribution of real returns in each currency.
  - Market portfolio view: dollarization as a response to market imperfections.
  - Institutional view: country-specific institutional factors explain dollarization.
- Vetlov (2001) identified factors leading to dollarization: high devaluation expectations, inflation rate differentials between domestic and foreign inflation, significant interest spread between domestic and foreign currency deposits, current account deficits, and inadequate levels of international reserves.

### Theoretical models and transaction-cost perspectives
- Giovanni and Turtelboom (1994) classified models of money demand in multicurrency environments into:
  1. Cash‑in‑advance models: real return differentials between domestic and foreign currencies determine currency demand.
  2. Transactions cost models (Marshall 1987; Poloz 1986): currency substitution may emerge when transaction costs are high, leading agents to maintain foreign exchange both as medium of exchange and store of value.
- Bogetic (2000) found higher domestic inflation vis-à-vis foreign inflation increases foreign currency holdings.

### Hysteresis and difficulty of de-dollarization
- Dollarization is often difficult to reverse; memories of macroeconomic instability and hyperinflation cause agents to maintain foreign currency assets even after stabilization and credible policy.
- Explanations of hysteresis:
  - Baliño and others (1999): currency-use practices change slowly and switch only when significant benefits exist.
  - Ize and Yeyati (1998): hysteresis can result from exchange‑based stabilization measures not aligned with de‑dollarization (e.g., reducing exchange rate volatility below that of domestic prices).
  - Kokenyne and others (2010): maintaining trends in exchange rate can entrench expectations of further depreciations and promote persistence of dollarization in some circumstances.

### Exchange rate volatility and de-dollarization
- Several studies find higher exchange rate volatility can encourage de-dollarization if two‑way exchange rate movements are allowed (Kokenyne and others (2010); Garcia‑Escribano (2010)).
- Examples where greater exchange rate flexibility contributed to de-dollarization: Lao P.D.R. (1995), Poland (1995–2000), Turkey (2001).
- Local currency appreciation trends contributed to deposit de-dollarization in Bolivia, Peru, Paraguay, and Uruguay (2001–10).
- Other studies (for example, Berkmen and Cavallo 2010) argue the causal relationship between exchange rate volatility and de-dollarization is generally not strong.

### Dollarization in Africa: existing evidence
- No comprehensive study focuses on dollarization in the African context, but notable country studies include:
  - Tanzania (Kessy 2011): financial sector liberalization (1992) allowing banks to open foreign currency deposit accounts led to foreign currency deposits rising to about 15 percent of broad money within a year and peaking at 33.5 percent in 2003; Tanzania became one of the most dollarized countries within the EAC, though dollarization has been declining in recent years.
  - Olalekan (2009): examined macroeconomic fluctuations and deposit dollarization in 18 SSA countries (1980–2004), highlighting the role of expectations about exchange rate changes and capital account restrictions.

### Stylized facts (overview for SSA, 2001–12)
- Data period: 2001−12.
- Two measures used: bank deposit dollarization and bank loan dollarization, defined as the ratio of bank deposits (loans) denominated in foreign currency to total bank deposits (loans).
- Data sources: IMF International Financial Statistics (IFS) database and African Department database (AFR).
- Coverage: IFS provides information for an average of 100 countries during 2001–11; for SSA, IFS has 30 countries for deposits and 29 for loans. AFR database used to expand SSA coverage; AFR does not decompose by economic sector. Broad definition used: all sectors except central bank and central government.

### Key SSA dollarization statistics and patterns
- Average deposit dollarization (excluding countries with zero values): 29 percent in 2001 and 31.6 percent in 2012.
- Average loan dollarization: 26.1 percent in 2001 and 34.8 percent in 2012.
- Including all countries but Zimbabwe, the average deposits dollarization was 14 percent in 2001 and 16 percent in 2012; the average for loans was 7 percent in 2001 and 12 percent in 2012.
- The most dollarized SSA countries include resource‑dependent economies: Democratic Republic of the Congo, Liberia, São Tomé and Principe, Angola, Mozambique, and Zambia.
- Number of countries with high deposit dollarization (above 30 percent) remained the same between 2001 and 2012, but composition changed: Tanzania and Mozambique reduced below 30 percent; Sierra Leone increased above 30 percent; Zimbabwe became fully dollarized.
- Countries with high loan dollarization increased from four (Angola, Mozambique, São Tomé and Principe, Tanzania) to six (Angola, Democratic Republic of the Congo, Liberia, São Tomé and Príncipe, Tanzania, Uganda).
- Correlation: for most SSA countries, higher deposit dollarization moves pari passu with higher loan dollarization.

### Trends over 2001–12
- Deposit and loan dollarization did not increase significantly overall from 2001–12; mean values for the first and second half of the decade are close.
- Country‑level notable changes:
  - Seychelles: largest statistically significant increase in deposits dollarization, associated with loss of peg credibility, near‑depletion of reserves, and a significant parallel market.
  - Democratic Republic of the Congo: similar dynamics for loan dollarization.
  - Angola and Mozambique: significant reductions in deposits and loans dollarization, respectively, during the second half of the decade.
- Country classification by trends (sample restrictions noted):
  - Deposits sample: 23 SSA countries with positive deposit dollarization values—excluding Zimbabwe, Lesotho, Swaziland, and CFA zone countries—divided almost equally into downward, upward, and stable trend groups.
  - Loans sample: 15 countries with positive loan dollarization values—excluding Zimbabwe—concentrated in the upward trend group.
- Of the five most dollarized economies (Angola, Democratic Republic of the Congo, Liberia, São Tomé and Príncipe, Zambia):
  - Angola recorded a downward trend in both deposits and loans.
  - Democratic Republic of the Congo, Liberia, and São Tomé and Príncipe recorded upward trends.
  - Zambia exhibited a downward trend in deposits but an upward trend in loans.

### Debt currency composition and “original sin”
- Average ratio of foreign exchange debt to total debt in SSA:
  - Started around 78 percent in 2001.
  - Decreased rather steadily to around 60 percent at end-2012.
- Implication: increasing ability of governments to issue debt in domestic currency, enhancing resilience to external shocks.
- Recent developments: several SSA economies have resumed tapping international capital markets; some debut issuers have issued second‑time bonds (Ghana, Gabon, Nigeria), and others (Kenya, Senegal, Tanzania) consider international issuance in the near future.
- Overall reduction of “original sin” is associated with differentiation of external financing sources for some frontier SSA economies.

*International Monetary Fund*

### 25.6 percent in 2012.

### _afr1504 - 25.6 percent in 2012.

### Overview: regional dollarization trends (2001–12)
- In 2001, 26 percent of total bank credit in SSA was denominated in foreign currencies; this ratio increased to 34 percent in 2012.
- As a result, SSA is today the second most dollarized region after Europe and Central Asia.
- Figure reference: "Figure 1.7. Evolution of Loans Dollarization by Region, 2001–12 (Loans in foreign currency as percent of total deposits)".

### Regional deposit and loan dollarization (selected reported figures)
- Reported regional aggregates (table fragment as presented):
  - East and South Asia and Pacific20.918.120.117.8
  - Europe and Central Asia59.745.858.737.3
  - Latin America & the Caribbean29.926.526.724.0
  - Middle East and North Africa19.5*11.711.4*13.1
  - Sub-Saharan Africa27.631.626.234.8
- Note from text: p25 and p75 show, respectively, the first and third quartile.

### Comparison of observed dollarization with benchmark estimates
- For most SSA countries, observed dollarization tracks benchmark levels estimated on the basis of per capita GDP, financial depth, inflation, and openness (constellation around the 45 degree line in Figure 1.8).
- When only SSA countries are considered, the gap between observed dollarization and the benchmarks narrows significantly.
- Chart references: "World: Deposit Dollarization (Foreign currency deposits as percent of total deposits)" and "World: Credit Dollarization (Foreign currency loans as percent of total loans)"; "Sub-Saharan Africa: Deposit Dollarization" and "Sub-Saharan Africa: Loans Dollarization".

### Macroeconomic associations and risks
- Dollarized economies tend to have higher and more volatile inflation and exchange rate depreciation; this tendency is more pronounced in SSA (Figure 1.10).
- Currency mismatches represent a major financial risk:
  - Large foreign currency liabilities with domestic-currency assets create solvency risk if the national currency depreciates sharply.
  - Public fear of such scenarios could trigger capital flight and self-fulfilling expectations.
- The analysis highlights that macroeconomic volatility typical of dollarized economies is more pronounced in SSA than in other regions; currency mismatches are significant, and persistence is higher in SSA compared to other regions.

### Persistence of dollarization in SSA
- Simple regression result reported in text:
  - Dollarization rates in the second half 2010s are on average equal to 79 percent of the rates in the early 2000s (rest of world).
  - In SSA this proportion increases to 103 percent, implying more persistent dollarization in SSA.
  - The difference is statistically significant at 5 percent level.
- Panel unit root test summary (text excerpt):
  - A panel unit root test accepts the null hypothesis that foreign exchange deposit shares have unit roots for all SSA countries whereas it rejects the same null for the rest of the world at 1 percent significant level.
- Table reporting test statistics as presented in text:
  - Inverse chi-squared: P41.21.0000324.70.0000
  - Inverse normal: Z3.91.0000-2.40.0090
  - Inverse logit t: L*3.70.9998-4.80.0000
  - Modified inverse chi-squared: Pm-3.40.99979.40.0000
  - Sample size / Number of countries / Average periods per country: 4379 1211.3 516891 SSA The Rest of the World

### Households and firms: sectoral dollarization dynamics (Annex 1.1)
- Sectoral patterns are highly country specific; sample from IMF’s IFS database allows analysis by households, firms, other financial corporations, and other resident sectors.
- General observations:
  - Firms tend to have more loans denominated in foreign currency than households, partly because some firms earn foreign currencies through exports.
  - Firms in natural resource-based economies may match liabilities to hard currency-generating assets.
  - Households and firms have high rates of foreign currency borrowing in countries with historically high inflation (Angola, Democratic Republic of the Congo, and São Tomé and Príncipe).
  - Domestic inflation appears generally to be a factor inducing firms and households to hold foreign currency financial assets.
- Salient country-specific points (as reported):
  - For SSA countries in the east in 2012, firms were more likely to borrow in foreign currency than households in the sampled countries.
  - Democratic Republic of the Congo: percent of foreign currency loans to total loans greatly increased for firms and households over the period.
  - Mozambique reduced dollarization in both deposits and loans in households and firms.
  - Uganda: firms increased both their deposits and loans dollarization.
  - Tanzania: firms reduced deposits dollarization but increased share of loans in foreign currency from 2001–12.
  - Firms in east SSA, notably Nigeria, increased foreign currency deposits over the past five years in the sample.
  - Households in São Tomé and Príncipe, Seychelles, Sierra Leone, and Tanzania increased foreign currency deposits in recent years; households in Mauritius and Mozambique gradually switched to local currency deposits.
- Net open currency positions:
  - Defined as foreign currency–denominated loans minus foreign currency–denominated deposits, divided by total deposits.
  - Every country with available data shows a negative net open position with other resident sectors in 2012 (Table A.1.1).

### Detailed time-series excerpts (selected entries from Table A.1.1)
- FX Loans to total Loans - Other Nonfinancial Corporations (selected rows):
  - Angola56.9657.1157.8762.2371.3471.1171.0960.5964.7364.1052.3843.81
  - Democratic Republic of Congo29.8946.2974.9180.8481.7779.2088.0390.3390.1677.3886.9291.46
  - Mozambique42.2643.8554.6359.4549.5231.3428.6137.4134.9233.7627.3130.35
  - Uganda20.5022.8920.3522.4526.6524.8123.1125.7420.9838.9341.2152.71
- FX Deposits to total deposits - Other Resident Sectors (selected rows):
  - Democratic Republic of Congo81.9089.7388.6590.1991.6495.0693.8994.2693.8290.9191.1187.64
  - Mozambique53.1145.5041.3534.4136.5235.5232.8629.9429.1628.6921.4420.88
  - Tanzania30.4133.4133.6330.4830.6533.1728.7026.7624.0334.3738.0035.19
  - Zambia34.2136.4532.0937.1531.1127.9538.9928.6926.8122.7322.1118.31
- Net open FX position - private sector (selected rows):
  - Angola-50.40-52.97-27.507.7055.1657.05172.35189.0763.6254.6322.8819.84
  - Democratic Republic of Congo-63.92-61.57-42.75-49.24-45.85-36.84-29.79-10.36-25.08-42.79-36.89-28.04
  - Sierra Leone-23.49-20.07-20.60-23.56-28.25-28.37-34.24-31.64-35.96-37.45-42.79-39.82
  - Zambia-54.43-48.24-43.66-49.22-39.96-37.81-44.68-39.77-41.99-42.53-42.04-29.82

### Legal and regulatory environment (Table A.1.2)
- Summary of reported legal restrictions on FX deposits/loans (excerpted as presented):
  - Use of foreign exchange among residents / Approval Required / Foreign exchange accounts permitted — resident and nonresident accounts listed per country.
  - Selected country entries as presented:
    - Angola yesyesnoyesyes
    - Benin yesyesyesyesyes
    - Botswana noy esnoyesno
    - Burkina Faso yesyesyesyesyes
    - Burundi yesyesnoyesyes
    - Cameroon yesyesyesnon.r.
    - Cape Verde yesyesnoyesyes
    - Congo, DR yesyesnoyesyes
    - Kenya noy esnoyesno
    - South Africa noy esyesyesno
    - Tanzania noy esnonono
    - Uganda noy esnoyesno
    - Zambia noy esnoyesno
    - Zimbabwe yesyesnoyesyes
- Note: Source listed as "IMF’s Annual Report on Exchange Arrangements and Exchange Restrictions."

*Source: INTERNATIONAL MONETARY FUND.*

### 2. Modeling the Determinants of

### 2. Modeling the Determinants of Dollarization in SSA

### Overview and methodology
- Scope: panel of data covering all Sub‑Saharan Africa (SSA) countries over the period 2001–12; stationarity tested over 2000–11 using the Levin‑Lin‑Chu test.
- Core question: correlation between degree of deposit dollarization and average inflation and average exchange rate depreciation over the past few years, contrasting SSA with the rest of the world.
- Main empirical approach: feasible generalized least squares (GLS) panel estimates for household, firm, and overall deposit and loan dollarization to exploit time and cross‑country variation.
- Regression framework (schematic):
  - DDit = β1·CurrSubsit + β2·Portfolioit + β3·MktFailureit + β4·Accessit + β5·Institit + β6·Controlsit + εit
  - DD is measured by deposit (loan) dollarization of the overall economy, firms, and households.

### Stylized correlations (figures)
- Deposits dollarization 2011 versus Average inflation 2001–2011:
  - y = 1.31x + 16.43
  - R² = 0.159
- Loans dollarization 2011 versus Average inflation 2001–2011:
  - y = 1.17x + 16.03
  - R² = 0.139
- Deposits dollarization 2011 versus Average depreciation 2001–2011:
  - y = 1.87x + 23.81
  - R² = 0.147
- Loans dollarization 2011 versus Average depreciation 2001–2011:
  - y = 1.84x + 23.33
  - R² = 0.131
- Deposits dollarization 2011 versus Average ratio of M2 to GDP 2001–2011:
  - y = -0.45x + 46.57
  - R² = 0.3195
- Loans dollarization 2011 versus Average ratio of M2 to GDP 2001–2011:
  - y = -0.42x + 44.03
  - R² = 0.254

Notes from figures:
- Countries excluded from some charts: a) inflation charts: Angola (46 percent average inflation) and Dem. Rep. of Congo (50 percent); b) depreciation charts and equations: Zimbabwe and Iraq with average depreciations greater than 1,000 percent; c) M2 to GDP charts and equations: Liberia with an average ratio greater than 1,000 percent.

### Key econometric findings — general
- Currency substitution (inflation and nominal depreciation) generally enters significantly with expected positive sign for both deposits and loans; nominal exchange rate depreciation is particularly relevant and significant in nearly all cases.
- Portfolio effects: a positive domestic return differential reduces deposit dollarization. A higher real deposit rate is associated (in most cases significantly) with lower levels of dollarization. Firms are more sensitive to interest differentials than households.
- Financial development: higher M2/GDP is unambiguously associated with lower deposit dollarization.
- Income per capita: higher GDP per capita is associated with a marked decline in deposit dollarization, particularly in SSA.
- Capital account openness: no significant impact on deposit dollarization for the SSA sample.
- External debt to GDP: tends to be associated with a higher share of foreign currency deposits, though significance varies.
- Access to foreign exchange finance (BIS deposits+loans to GDP, exports to GDP): results are mixed; exports/GDP show a tendency in SSA for export earners to keep proceeds in foreign currency.
- Oil exports: in SSA, higher oil exports to GDP are associated with higher deposit dollarization; in the rest of the world higher oil exports are associated with lower foreign currency deposits.
- Exchange rate regime: floating regimes in SSA tend to be associated with higher dollarization ratios while fixed regimes are generally associated with lower dollarization ratios (base category = intermediate regimes). This likely reflects floating regimes in SSA being associated with higher and more volatile inflation.

### Deposit dollarization — summarized results and numeric highlights
- Sample considerations: baseline excludes countries with zero deposit dollarization due to legal restrictions (WAEMU, CEMAC, Lesotho, Swaziland); some specifications exclude all zero observations.
- Currency substitution:
  - Inflation effects: mixed significance; generally positive where significant.
  - Nominal depreciation to USD often significant and positive.
- Portfolio model:
  - Deposit rate spread to US: negative and significant coefficients across regressions (example coefficients from Table 2.1: -0.06*** (RoW), -0.03* (SSA), -0.08***, -0.06***, -0.11***, -0.30***, -0.07***, -0.14***; robust standard errors reported in table).
  - Real deposit rate (i(t) − infl(t+1)): negative and often significant (example coefficients: -0.08**, -0.12***, -0.07**, -0.09*, -0.11**, -0.30, -0.05, 0.04).
- Market failure / macro proxies:
  - CA index: generally not significant for SSA deposit dollarization in baseline.
  - External debt to GDP: positive association (example coefficients: 0.09***, 0.03**, 0.08***, 0.08***, -0.00, 0.02, 0.11***, 0.04).
  - M2 to GDP ratio: negative and significant (example coefficients: -0.36***, -0.09**, -0.40***, -0.15***, -0.35***, 0.17, -0.46***, -0.02).
  - GDP p.c.: negative association in SSA (example coefficients: -0.00, -0.78*, -0.01, -1.80***, -0.14, -3.64***, -0.24**, -3.77***).
- Access to FX finance:
  - BIS deposits+loans to GDP: inconclusive for deposit dollarization (example coefficients: -0.02, 0.01, 0.00, 0.02**, 0.08**, -0.03, -0.00, 0.06***).
  - Exports to GDP: for SSA, many results point to a significant positive impact on deposits in foreign exchange (example coefficients: -0.04, -0.03, -0.05, 0.23***, -0.05, 1.44***, -0.08*, 0.33***).
  - Oil exports to GDP: SSA overall economy coefficients suggest positive association (example coefficients: -0.20***, 0.49***, -0.22***, 0.27***, -0.19***, -0.57, -0.17*, -0.70***).
- Institutions and controls:
  - De jure peg: coefficients often large and negative for deposit dollarization relative to intermediate regimes (example coefficients: -12.78***, 0.57, -7.92***, -20.05***, -12.50***, -18.85***, -12.22***, -7.98***).
  - Polity index: negative and significant in several regressions (example coefficients: -0.30***, 0.06, -0.21**, -0.77***, 0.15, -2.86***, -0.29**, -0.34).
  - Population control: coefficients negative and significant in many specifications (example coefficients: -0.21***, -0.22***, -0.20***, -0.23***, -0.20***, 0.09, -0.19***, 0.17***).
- Table sample sizes (selected): Observations: 447, 214, 428, 164, 370, 106, 391, 112; Number of countries: 44, 21, 43, 16, 39, 12, 41, 13. Significance notation: *** p<0.01, ** p<0.05, * p<0.1.

### Loan dollarization — summarized results and numeric highlights
- Commonalities with deposit dollarization but also differences owing to longer horizon and institutional factors required for loan dollarization.
- Currency substitution:
  - Inflation and nominal depreciation are significantly positively related to loan dollarization in the rest of the world; results for SSA are less conclusive.
  - Example inflation coefficients (Table 2.2): 0.30***, -0.02, 0.31***, -0.19***, 0.39***, -0.09, -0.01, 1.09***.
  - Example nominal depreciation coefficients: 0.04**, 0.02*, 0.06***, 0.05*, -0.03, 0.08*, 0.00, 0.20***.
- Portfolio effects:
  - Lending rate spread to US: generally negative coefficients (example: -0.05***, 0.00, -0.06***, -0.06, -0.02, -0.07, -0.01, -0.12).
  - Real lending rate (i(t) − infl(t+1)): mixed signs; for SSA higher real lending rates associated with higher loan dollarization in some specifications (example coefficients: -0.14***, 0.08*, -0.12***, 0.04, -0.11**, 0.31**, -0.01, -0.36***).
- Market features:
  - CA index shows positive and significant coefficients for loan dollarization in many specifications (example coefficients: 1.14**, 3.70***, 1.04**, 0.48, 2.59***, 1.10, 1.99***, 4.01***).
  - External debt to GDP: positive and significant (example coefficients: 0.15***, 0.06***, 0.20***, 0.19***, 0.19***, 0.08, 0.12***, 0.20***).
  - M2 to GDP: negative and significant (example coefficients: -0.34***, -0.18***, -0.30***, -0.58***, -0.24***, -0.05, -0.35***, -0.63***).
- Access to FX finance:
  - BIS exposure inconclusive for loan dollarization (example coefficients: -0.02, -0.00, -0.03, 0.14, -0.07*, 0.45, -0.03, 0.23*).
  - Exports to GDP: negative relation in rest of world with loan dollarization (example coefficients: -0.21***, 0.10**, -0.23***, 0.21, -0.28***, 0.10, -0.16***, -0.18*).
  - Oil exports to GDP: SSA shows positive and often large coefficients in some household regressions (example coefficients: -0.26***, 0.63***, -0.33***, 0.74***, 0.05, 7.38***, -0.14**, 12.61***).
- Institutions and controls:
  - De jure float: large and sometimes extreme coefficients in loan regressions for SSA/households in Table 2.2 (example coefficients: 1.84, 10.37***, 0.93, 10.43**, 3.15**, -41.16**, 0.17, -112.29***), reflecting small sample and sparse observations.
  - Polity and population controls show varied significance (example polity coefficients: 0.01, -0.07, 0.09, -0.46*, 0.26**, 0.65, 0.05, 6.55***; population: -0.17***, -0.00, -0.19***, 0.72***, -0.15***, 1.19***, -0.14***, 1.78***).
- Table sample sizes (selected): Observations: 376, 183, 346, 87, 299, 58, 331, 39; Number of countries: 41, 20, 38, 10, 35, 8, 37, 7.

### Loan versus deposit dollarization — contrasts
- Many common drivers (institutional weaknesses, external debt, financial development) explain both loan and deposit dollarization.
- Differences:
  - Current account openness: no impact on deposit dollarization but a significant positive impact on loan dollarization (consistent across regressions), suggesting CA openness affects banks’ ability to obtain foreign funds to lend domestically.
  - Currency substitution: stronger and more consistent for loan dollarization in the rest of the world; for SSA results are less conclusive.
  - Portfolio effects differ: higher real lending rates in SSA associated with higher loan dollarization (demand factors); for the rest of the world signs can be opposite.

### Household versus corporate dollarization
- Granular IFS data (limited country coverage) permit analysis of household and enterprise deposit dollarization.
- Findings largely consistent with aggregate results.
- Firms are substantially more sensitive than households to portfolio considerations for deposit dollarization — reflecting firms’ greater access to alternative investment and treasury options.

### Policy implications (drawn from empirical findings)
- Pursuing macroeconomic stability — keeping inflation low and stable and reducing nominal exchange rate volatility — could gradually diminish dollarization in SSA given the prominent role of currency substitution.
- Deepening domestic financial markets (raising M2/GDP and expanding domestic currency investment opportunities) should reduce deposit and loan dollarization.
- Raising domestic returns on local‑currency deposits (narrowing negative return differentials vis‑à‑vis foreign currency) reduces deposit dollarization.
- Improvements in institutions, governance, and income per capita are associated with lower dollarization and should be part of long‑term de‑dollarization strategies.
- Prudential and regulatory frameworks (though not fully tested here due to data limitations) remain an important area for policy given evidence from other regions that prudential measures and safety‑net design can affect dollarization incentives.

*International Monetary Fund*

### Annex 2.1. Econometric Output

### Annex 2.1. Econometric Output

### Key econometric findings (tables A.1–A.4)
- Dependent variables examined include deposit dollarization and loan dollarization for Sub-Saharan Africa (SSA) versus Rest of the World (RoW) over 2001–2012.
- Estimation methods reported: Pooled, FGLS - AR(1), Panel Tobit, Blundell-Bond Panel.
- Sample sizes and panel dimensions (selected):
  - Observations: 4320, 4216, 4019, 3915 (various specifications reported).
  - Number of countries: 44, 42, 42, 41 (various specifications reported).
  - Time dummies: XX (included where indicated).
- Goodness-of-fit (selected R-squared values reported):
  - 0.689, 0.986, 0.725, 0.933, 0.642, 0.690 (from Table A.1).
  - 0.680, 0.720, 0.642, 0.821 (from Table A.2).
  - 0.653, 0.721, 0.582, 0.571 (from Table A.3).
  - 0.656, 0.725, 0.580, 0.788 (from Table A.4).
- Robust standard errors reported in parentheses; significance levels: *** p<0.01, ** p<0.05, * p<0.1.
- Consistent predictor categories across specifications:
  - Currency Substitution: Inflation; Nominal depreciation to USD.
  - Portfolio model variables: Deposit rate spread to US; Real deposit rate i(t) − infl(t+1); Lending rate spread to US; Real lending rate i(t) − infl(t+1).
  - Market failure indicators: CA index; External debt to GDP ratio; M2 to GDP ratio; GDP p.c.
  - Access to FX finance: BIS deposits+loans to GDP ratio; Export to GDP ratio; Oil export to GDP ratio.
  - Institutions (Monetary): Defacto Peg; Defacto Float.
  - Institutions (Political): Polity Score.
  - Controls: Population; Lag dependent; Constant.
- Representative coefficient patterns (sign and statistical significance preserved as reported):
  - M2 to GDP ratio: predominantly negative and often statistically significant (e.g., -0.40**, -0.46**, -0.41**, -0.46***, -0.25***, -0.28*** across tables).
  - External debt to GDP ratio: mixed but several positive and significant coefficients (e.g., 0.43**, 0.19***, 0.14, 0.17***, 0.29***, 0.29***).
  - Defacto Peg: large negative and often highly significant coefficients in multiple specifications (e.g., -38.19***, -38.02***, -30.78**, -30.93**).
  - Polity Score: generally negative and often significant (e.g., -2.07**, -2.03**, -2.07**, -2.01**, -1.29***).
  - Population: negative and often significant (e.g., -0.16***, -0.15***, -0.17**).
  - BIS deposits+loans to GDP ratio and Export to GDP ratio show mixed signs and significance depending on specification.
- Note on interpretation: tables report many specification-specific coefficients and standard errors; statistical significance indicated by stars as printed.

### Unit-root tests (Annex 2.2)
- Levin-Lin-Chu unit-root test for Deposits in Foreign Currency:
  - Number of panels = 94
  - Number of periods = 12
  - AR parameter: Common; Asymptotics: N/T -> 0
  - Panel means: Included
  - Time trend: Not included
  - ADF regressions: 1.36 lags average (chosen by AIC)
  - LR variance: Bartlett kernel, 7.00 lags average (chosen by LLC)
  - Unadjusted t = -50.5829
  - Adjusted t* = -49.3483; p-value = 0.0000
- Levin-Lin-Chu unit-root test for Credit in Foreign Currency:
  - Number of panels = 88
  - Number of periods = 12
  - AR parameter: Common; Asymptotics: N/T -> 0
  - Panel means: Included
  - Time trend: Not included
  - ADF regressions: 1.55 lags average (chosen by AIC)
  - LR variance: Bartlett kernel, 7.00 lags average (chosen by LLC)
  - Unadjusted t = -5.4932
  - Adjusted t* = -2.8166; p-value = 0.0024

### Effective de-dollarization: summary of analysis and policy recommendations (Chapter 3)
- Empirical diagnosis:
  - Dollarization in SSA is relatively high and, unlike other regions, largely failed to decline over the past decade (except in a few countries).
  - Financial dollarization increases: (i) hampers the effectiveness of monetary policy, (ii) limits fiscal policy flexibility, and (iii) increases financial sector vulnerabilities.
- Principal policy conclusion:
  - Dollarization is difficult to reverse and can only be reduced through long, consistent, and credible stabilization efforts.
- Key ingredients for a successful de-dollarization strategy (as stated):
  - (i) a sustained and credible macroeconomic stabilization;
  - (ii) a proper liberalization effort encouraging de-dollarization by removing distortionary direct controls or constraints that discourage the use of the local currency; and
  - (iii) prudential measures that reduce banks’ incentives to borrow and lend in foreign currencies—such as asymmetric reserve requirements and the provisioning for currency-induced credit risks.
- Additional guidance on policy instruments:
  - Measures that provide market-based incentives are most successful.
  - Administrative measures aimed at forcing de-dollarization can easily backfire, potentially encouraging capital flight and reducing financial sector intermediation, ultimately hindering growth.
  - A mix of sound macroeconomic policies, microprudential measures, and sustained efforts to create conditions for longer-term domestic capital market development is required.
  - Prudential measures cited include asymmetric reserve requirements and provisioning for currency-induced credit risks.
  - Development of local-currency securities markets (initially government bonds, later longer maturities) is an important complement.

### Case studies of successful de-dollarization (selected facts and outcomes)
- Israel:
  - Inflation reached about 445 percent in 1984.
  - Share of dollarized deposits peaked at 39 percent in 1984.
  - Fiscal consolidation cut the fiscal deficit from 19 percent of GDP in 1985 to about 10 percent in the late 1990s.
  - Policy mix: macro stabilization (inflation targeting, greater exchange rate flexibility), prudential measures (higher remuneration on domestic-currency reserve requirements than on foreign-currency reserves; stronger borrower requirements for those without FX income), and development of domestic-currency securities issuance.
- Poland:
  - Monthly inflation reached 55 percent in October 1989.
  - Fiscal deficit increased to 7 percent of GDP in the late 1980s.
  - Real GDP contracted by more than 11 percent in 1990; growth resumed in 1992 and reached about 4 percent in 1993.
  - Average inflation reduced from 585 percent in 1990 to 35 percent in 1993.
  - Deposit dollarization dynamics reported: only about 20 percent of bank deposits were zloty-denominated (late 1980s); by 1993 the level of dollarization had declined to about 35 percent; by 1999 it reached 4.5 percent.
  - Key policy actions: comprehensive reform program, use of the exchange rate as a nominal anchor, issuance of local-currency treasury bills perceived as high-yield safe investments.
- Bolivia:
  - In 2001 the banking sector was almost completely dollarized: only 6 percent of deposits and 3 percent of loans were denominated in local currency.
  - By 2012 the share of deposits denominated in foreign currencies had declined to 25 percent and loans to 21 percent.
  - Policy lessons: a previous forced mandatory conversion in 1982 led to severe instability; market-driven strategy in the 2000s used prudent macro policies, higher reserve requirements on foreign-currency deposits, higher provisioning on foreign-currency loans, and development of a local-currency bond market with longer maturities (9 to 15 years).
- Peru:
  - Historical peak inflation exceeded 7,600 percent a year in 1990.
  - Deposit and loan dollarization declined from about 79 percent in 2001 to about 50 percent in 2010.
  - Earlier forced conversion in the mid-1980s led to capital flight and a decline in financial intermediation.
  - Successful strategy since the early 2000s combined: (i) strengthened macroeconomic stability (fiscal surpluses, lower public debt, inflation targeting, higher international reserves), enabling an investment-grade rating; (ii) prudential measures to reflect currency mismatch risks; and (iii) development of a securities market with longer maturities in domestic currency.

*International Monetary Fund*

### Box 3.1. De-Dollarization in Angola (as of 2014)

### Box 3.1. De-Dollarization in Angola (as of 2014)

### Background and context
- The U.S. dollar remained the legal tender, like the local currency, the kwanza.  
- The dollar was extensively used as a unit of account and for transaction purposes, and banks were heavily dollarized.  
- A decade earlier nearly three-fourths of deposits and two-thirds of credit to the private sector were in foreign currency; both ratios were closer to 40 percent in 2014.  
- Dollarization drivers:
  - Long-lasting civil conflict and significant uncertainty about Angola’s macroeconomic conditions and the value of the kwanza.
  - Inflation at hyperinflation levels in the early 2000s: it topped 100 percent when the civil war ended in 2002 and still exceeded 30 percent in 2004.
  - Since 2001 the kwanza has depreciated by nearly 1,600 percent vis-à-vis the dollar.
  - High dependence on oil revenues: the oil sector used dollars extensively to settle payments to local suppliers; these dollars were often retained rather than sold to commercial banks, producing severe foreign currency shortages in the retail market.
  - Higher reserve requirements on kwanza than on dollar deposits (until recently), encouraging banks to offer more attractive rates on dollar deposits and loans.
  - Capital flight incentives: tight capital controls and nonconvertibility of the kwanza outside the country led to disguised capital flight, especially through over-invoicing of foreign payment obligations; over-invoicing, combined with BNA tailoring foreign currency sales to estimated valid import demand, produced perceived retail foreign currency shortages.

### Impact of dollarization on policy
- High dollarization reduced the effectiveness of monetary policy and obliged the Banco Nacional de Angola (BNA) to rely on the exchange rate as the nominal anchor.
- The BNA’s policy rate remained a relatively weak monetary policy instrument.

### Measures implemented to reduce dollarization (BNA package, 2013–early 2014)
- Oil sector foreign exchange law phased in during 2013:
  - Since mid-2013 the law requires the oil sector to settle all domestic transactions, including taxes and payments to local suppliers, in local currency purchased from domestic banks.
  - This measure channeled the supply of dollars entering the nonoil sector through commercial banks.
- Mid-2013 measures to improve accuracy of registered foreign currency demand:
  - Made it more difficult for importers to overstate foreign currency obligations.
  - Aimed to make the demand registered by banks at the primary auctions a more accurate reflection of import needs.
- Tighter bank foreign exchange exposure limits:
  - Obliged banks to sell surplus dollars into the retail or the interbank market.
  - Limited extension of foreign exchange–denominated consumer credit to foreign exchange earners only.
- Reserve requirement adjustment in early 2014:
  - BNA reduced the reserve requirements on kwanza deposits below those on dollar deposits, allowing banks to offer more attractive rates on kwanza deposits and loans.

### Evidence of impact (through 2014)
- Foreign exchange shortages had been eliminated and an interbank market had emerged.
- Dollarization trends:
  - Dollar deposit and loan ratios had trended down through the past decade and the decline accelerated since the oil sector foreign exchange law in mid-2013.
  - From June to December (2013), dollar deposits fell from 53 to 42 percent of total deposits.
  - From June to December (2013), dollar credit fell from 41 to 37 percent of total credit to the economy.
  - Anecdotal evidence suggested the oil sector foreign exchange law impacted payment dollarization and that the kwanza was rapidly becoming the primary unit of account and medium of exchange.
- Macroeconomic stabilization contributing to de-dollarization:
  - By 2014 Angola had reduced inflation to single digits (below 8 percent in 2013).
  - The foreign exchange rate had been stabilized.

*International Monetary Fund — Box 3.1, De-Dollarization in Angola (as of 2014).*

### Box 3.2.   São Tomé and Príncipe

### Box 3.2.   São Tomé and Príncipe

### Dollarization in São Tomé and Príncipe: evolution and policy actions
- Inflation rose from 9 percent in 2001 to about 28 percent in 2008, prompting high currency substitution.
- The ratio of foreign currency–denominated deposits to broad money rose from 35 percent in 2001 to about 57 percent by 2007.
- Commercial banks operated largely with foreign currency; foreign currency loans accounted for almost ¾ of the credit portfolio.
- Policies and outcomes:
  - In March 2010, the central bank updated its policy to require 90 percent of the reserve requirements be paid in local currency.
  - The central bank lowered the required reserve ratio for domestic currency to 18 percent, while keeping the statutory requirements for foreign exchange at 21 percent.
  - A peg to the euro was implemented in January 2010, achieving a slow decrease in inflation and dollarization.
  - Efforts, including prudential regulations regarding reserve requirements, led to a reversal in the share of foreign currency loans from 77 percent in 2010 to 58 percent in 2012.
  - In 2013, banks actively sought funding in local currency, causing an increase in the spread between foreign currency and domestic currency interest rates and steering an increase in domestic currency–denominated deposits.

*Note: São Tomé and Príncipe is identified in the cross-country analysis as an outlier with a stock of debt well in excess of GDP and as one of the SSA countries that recorded an upward trend in dollarization during the decade.*

### Cross-country study: scope, definitions, and sample
- Dollarized: “dollarization ratio” (foreign currency deposits (FCD) / total deposits (TD)) exceeds 30 percent.
- Initially dollarized: dollarization rate exceeded 30 percent on average between 2001 and 2003, or maximum dollarization rate between 2001 and 2003 exceeded 30 percent.
- De-dollarized (successful) if:
  1. Three-year average dollarization ratio declined by at least 20 percent between 2001–03 and 2010–12; and
  2. Average proportion of foreign currency loans to total bank loans did not increase between these two three-year periods.
- Data coverage and outcomes:
  - Out of 194 countries with data, 42 were initially dollarized (nine in sub-Saharan Africa).
  - Of these 42, 11 countries were successful (de-dollarized) and 31 were unsuccessful (remained dollarized).
  - Four outliers excluded from certain analyses: Qatar; Democratic Republic of the Congo; Liberia; São Tomé and Príncipe. All these outliers were unsuccessful.

### What distinguishes successful and unsuccessful de-dollarization (summary of empirical findings)
- Broad patterns (2001–03 initial conditions vs. 2004–12 developments):
  - Successful countries generally had low or declining inflation, comparatively high growth, and experienced a nominal appreciation at some point between 2002 and 2012.
  - Successful countries started with higher initial dollarization and more external debt but had a better current account balance and a higher Polity IV democracy index in 2001–03.
- Key numerical differences (Table 3.2; values are averages unless stated):
  - Initial conditions (average 2001–03):
    - Dollarization: Successful 67.4; Unsuccessful 48.4
    - Real GDP growth (in percent): Successful 4.8; Unsuccessful 4.3
    - Inflation (in percent): Successful 21.6; Unsuccessful 9.4
    - Current account balance: Successful -5.0; Unsuccessful -5.6
    - Fiscal balance: Successful -2.8; Unsuccessful -3.3
    - Stock of public debt: Successful 60.1; Unsuccessful 54.6
    - Stock of external debt: Successful 63.8; Unsuccessful 59.3
    - Democracy: Successful 5.0; Unsuccessful 2.7
  - Conditions after 2003 (average 2004–12):
    - Real GDP growth (in percent): Successful 7.3; Unsuccessful 4.5
    - Inflation (in percent): Successful 8.5; Unsuccessful 8.9
    - Current account balance: Successful 0.3; Unsuccessful -7.6
    - Fiscal balance: Successful 1.4; Unsuccessful -2.6
    - Stock of public debt: Successful 33.6; Unsuccessful 39.4
    - Stock of external debt: Successful 41.1; Unsuccessful 48.4
    - Democracy: Successful 5.2; Unsuccessful 1.5
  - Net change between 2001–03 and 2010–12:
    - Inflation (difference): Successful -14.4; Unsuccessful -1.8
    - Public debt (difference): Successful -32.4; Unsuccessful -14.4
    - External debt (difference): Successful -27.7; Unsuccessful -4.6
    - Public debt (percentage change): Successful -53.5; Unsuccessful -14.9
    - External debt (percentage change): Successful -43.7; Unsuccessful 15.2
    - Democracy (difference): Successful 0.2; Unsuccessful -2.8
  - Other indicators:
    - CPIA (average): Successful 3.55; Unsuccessful 3.45
    - Exchange rate appreciation (average 2004–12 compared to average 2001–03): Successful 2.2; Unsuccessful -33.8
    - Exchange rate appreciation (average 2010–12 compared to average 2001–03): Successful 2.0; Unsuccessful -15.0
    - Exchange rate appreciation (maximum consecutive appreciation during the period): Successful 18.5; Unsuccessful 14.1

- Stylized empirical correlations:
  - Countries with an average dollarization rate higher than 65 percent in 2001–03 had a probability of success 43 percent higher than less dollarized countries.
  - A current account surplus or deficit below 3 percent of GDP between 2001 and 2003 increased probability of success by about 30 percent.
  - Initial public or external debt above 40 percent of GDP increased probability of success by 25 percent.
  - Most countries that de-dollarized experienced a nominal appreciation of at least 15 percent at some point between 2001 and 2012.
  - A current account deficit above 3 percent of GDP between 2001 and 2012 significantly increased the probability of de-dollarizing (by almost 70 percent).
  - A fiscal balance above 2 percent of GDP increased the probability of success.
  - Countries that de-dollarized implemented a significant reduction in external debt.

### Econometric evidence and model performance
- Pooled cross-sectional regressions relate net change in dollarization to initial conditions, post-2003 developments, and exchange rate appreciation.
- Key regression findings:
  - Initial dollarization rate has a significant positive impact on de-dollarization.
  - After 2003, real GDP growth, inflation, and the fiscal balance significantly affect de-dollarization with expected signs (negative for inflation; positive for growth and fiscal balance).
  - De-dollarization is significantly positively correlated with nominal exchange rate appreciation, but appreciation becomes insignificant when other policy variables are included—suggesting appreciation proxies for low inflation and sound fiscal balance.
  - The included variables explain about 70 percent of cross-country variance in the dependent variable in the main specifications.
- Predictive performance:
  - The model correctly predicts success in most actual de-dollarizations; in different specifications it returns between 28 and 30 correct predictions out of 38 cases considered for OLS specifications.
  - Probit specifications return between 28 and 33 correct predictions out of 38 cases under different specifications.
  - Notable model misses: failed to predict Turkey’s and Mozambique’s de-dollarization in some specifications; incorrectly predicts success for Cambodia and Nicaragua in some specifications.

### Key ingredients for successful de-dollarization (policy implications)
- Macroeconomic policies
  - Sustained and credible macroeconomic stabilization is crucial.
  - Low inflation reduces need to hold foreign currency for purchasing-power preservation.
  - Fiscal consolidation reduces government borrowing in foreign currency and inflationary financing; a stronger fiscal balance encourages local investors to purchase longer-term local currency bonds.
- Financial liberalization and market development
  - Liberalizing the domestic banking system and developing domestic financial markets facilitate de-dollarization.
  - Financial innovations (hedging instruments, indexed local currency bonds) reduce need for foreign currencies to hedge currency/inflation risk.
  - Efficient central bank liquidity management, issuance of medium- and long-term local currency government bonds, development of interbank markets, and a well-functioning foreign exchange market backed by adequate official reserves all support de-dollarization.
- Appropriate supervision and prudential regulation
  - Prudential measures can reduce banks’ incentives to borrow and lend in foreign currencies (e.g., asymmetric reserve requirements, provisioning for currency-induced credit risks).
  - Examples: Bolivia, Peru, Paraguay, Uruguay used such measures successfully; within SSA, Angola and São Tomé and Príncipe show the usefulness of prudential measures to incentivize domestic currency deposits and curb foreign currency loan growth.
- Sequencing and market-based approaches
  - Credibility of monetary policy generally needs to be reestablished first.
  - Market-based policies are preferable to administrative forcings; mandatory measures can be useful only as complements to market-based stabilization and carry risks if used alone (capital flight, disintermediation, banking instability).

### Sub-Saharan Africa (SSA) record and specific observations
- Mixed record: dollarization has been more persistent in SSA than elsewhere; few successful de-dollarization episodes.
- Of the five most dollarized SSA economies (Democratic Republic of the Congo, Liberia, São Tomé and Príncipe, Angola, Zambia), only Angola recorded a downward trend in both deposits and loans denominated in foreign currency in the period considered; Democratic Republic of the Congo, Liberia, and São Tomé and Príncipe recorded upward trends.
- Stylized SSA facts:
  - Deposit dollarization in SSA hovers around 30 percent—similar or above other regions except transition economies of Eastern Europe.
  - Loan dollarization in SSA increased from 26 percent in 2001 to 34 percent in 2012.
- Drivers in SSA:
  - Inflation and/or nominal exchange rate depreciation are key drivers of dollarization (currency substitution).
  - Portfolio considerations: low local-currency deposit rates and high local-currency loan rates lead customers to hold foreign currency deposits and contract foreign currency loans.
  - Political instability, dependence on primary commodities exports, and limited financial market development also contribute to persistent dollarization.
- Conclusion for SSA:
  - Successful de-dollarization requires sound macroeconomic policies, market-based incentives, and microprudential measures; it requires time and persistent, coordinated efforts.
  - Mandatory measures/direct controls have been useful only as complements to market-based strategies and can be ineffective or counterproductive otherwise.
- Key elements of a successful de-dollarization strategy (listed in the source):
  1. Restoring macroeconomic stability—reduce inflation and make fiscal policy sustainable.
  2. Strengthening the prudential framework to reflect risks from currency mismatches (examples: tax on U.S. dollar transactions, higher reserve requirements for foreign currency deposits, higher provisions for foreign currency loans).
  3. Pursuing market-based rather than administrative approaches (Bolivia and Peru illustrate superiority of market-based strategies).
  4. Addressing the “original sin” challenge by developing domestic capital markets with medium- and long-term maturities in domestic currency to provide attractive local-currency investment opportunities.

*Source: Box 3.2 and associated analysis on de-dollarization from the provided IMF document.*

### References

### _afr1504 - References

### References (bibliography)
- Baliño, T.J.T., A Bennett, and Eduardo Borensztein. 1999. “Monetary Policy in Dollarized Economies,” IMF Occasional Paper 12/171, International Monetary Fund, Washington.
- Berkmen, P., and E. Cavallo. 2010. “Exchange Rate Policy and Liability Dollarization: What does the Data Reveal about Causality?” Review of International Economics 18(5): 781–95.
- Bogetic, Zeljko. 2000. “Official Dollarization: Current Experiences and Issues.” Cato Journal 20(2): 179–213.
- Brillenbourg, Arturo, and Susan Schadler. 1980. “A Model of Currency Substitution in Exchange Rate Determination,” IMF Staff Paper 3 : 513–42, International Monetary Fund, Washington.
- De Nicolò, Gianni, Patrick Honohan, and Alain Ize. 2003. “Dollarization of the Banking System: Good or Bad?” IMF Working Paper 03/146, International Monetary Fund, Washington.
- de Zamaróczy, Mario, and S. Sa. 2003. “Economic Policy in a Highly Dollarized Economy, The Case of Cambodia,” IMF Occasional Paper 219, International Monetary Fund, Washington.
- Ebrill, Liam P. 1994. “Poland, the Path to a Market Economy,” IMF Occasional Paper 113, International Monetary Fund, Washington.
- Eichengreen, Barry, Ricardo Hausmann, and Ugo Panizza. 2002. Original Sin: The Pain, the Mystery and the Road to Redemption, paper presented at a conference on Currency and Maturity Matchmaking: Redeeming Debt from Original Sin, Inter-American Development Bank.
- Garcia-Escribano, M. 2010. “Peru: Drivers of De-Dollarization,” IMF Working Paper 10/169, International Monetary Fund, Washington.
- Giovannini, Alberto, and Bart Turtelboom. 1994. “Currency Substitution.” In The Handbook of International Macroeconomics, edited by F. Van der Ploeg. Cambridge: Blackwell: 390–436.
- Girton, Lance, and Don Roper. 1981. “Theory and implications of currency substitution,” Journal of Money, Credit and Banking 13: 12–30.
- Gulde-Wolf, Anne-Marie, David Hoelscher, Alain Ize, David Marston, and Gianni De Nicolò. 2004. “Financial Stability in Dollarized Economies,” IMF Occasional Paper 230, International Monetary Fund, Washington.
- International Monetary Fund. 2012. Regional Economic Outlook for sub-Saharan Africa. Washington.
- Ize, A., and E. Levy Yeyati. 1998. “Dollarization of Financial Intermediation: Causes and Policy Implications,” IMF Working Paper 98/28, International Monetary Fund, Washington.
- Kessy, Panteleo. 2011. “Dollarization in Tanzania: empirical evidence and cross-country experience,” International Growth Centre Working Paper 11/0251, International Monetary Fund, Washington.
- Kokenyne, A., J. Ley, and R. Veryune. 2010. “Dedollarization,” IMF Working Paper 10/188, International Monetary Fund, Washington.
- Levy-Yeyati, Eduardo. 2006. “Financial dollarization: evaluating the consequences,” Economic Policy 21(45): 61–118.
- Marshall, D.A. 1987. Inflation and Asset Returns in a Monetary Economy with Transactions Costs, Doctoral Dissertation, Carnegie-Mellon University.
- Miles, Marc. 1978. “Currency substitution, flexible exchange rates, and monetary independence,” American Economic Review 3: 428–36.
- Neanidis, Kyriakos C., and Christos S. Savva. 2009. “Financial dollarization: Short-run determinants in transition economies,” Journal of Banking & Finance 33(10): 1860–73.
- Offenbacher, Akiva Edward, and Roy Stein. 2003. "Dollarisation and Indexation in Israel's Inflation and Disinflation: ‘There's more than one way to skin a cat’." Comparative Economic Studies 45(3): 278–305.
- Olalekan, Yinusa. 2009. “Macroeconomic fluctuations and deposit dollarization in sub-Saharan Africa: evidence from panel data,” MPRA Paper 16259, Munich Personal RePEc Archive, Munich.
- Ortiz, Guillermo. 1983. "Dollarization in Mexico: Causes and consequences." In Financial Policies and the World Capital Market: The Problem of Latin American Countries, edited by Pedro Aspe Armella, Rudiger Dornbusch, and Maurice Obstfeld. Chicago: University of Chicago Press: 71–160.
- Poloz, S.S. 1986. “Currency Substitution and the Precautionary Demand for Money,” Journal of International Money and Finance 5(1): 115–24.
- Reinhart, Carmen, Roger Rogoff, and Michael Savastano. 2003. “Addicted to Dollars,” NBER Working Paper 10015, National Bureau of Economic Research, Cambridge.
- Vetlov, Igor. 2001. “Dollarization in Lithuania: An Econometric Approach,” Discussion Papers 1, Institute for Economies in Transition, Bank of Finland.

### Annex 1 — World Dollarization Picture (figures and tables)
- Figure A.1. Deposits Dollarization, 2001 and 2012 (Deposits in Foreign Currency over Total Deposits). Source: International Financial Statistics (IFS) database, IMF.
- Figure A.2. Loans Dollarization, 2001 and 2012 (Loans in Foreign Currency over Total Loans). Source: International Financial Statistics (IFS) database, IMF.
- Table A.1. Dollarization, 2001–11 (Average ratios of deposits/loans in foreign currency over totals). (Table includes country-level averages for 2001-06 and 2007-11 with student t-statistics and significance levels; example entries as presented:)
  - Angola 74.0 57.6 (4.78)*** 63.1 60.8 (0.52)
  - Burundi 10.4 16.8 (-5.01)*** ...... 
  - Comoros 0.9 0.2 (2.09)* 1.2 0.5 (1.27)
  - Congo, Dem. Rep. of 83.2 88.2 (-2.01)* 65.0 88.5 (-2.33)**
  - Eritrea 17.9 16.1 ( 2.54)** 17.1 0.0 ( 9.64)***
  - Ghana 29.2 26.4 (2.09)* 19.0 27.3 (-1.21)
  - Liberia 86.8 87.7 (-0.59) 72.3 83.2 (-6.69)***
  - Mozambique 45.7 37.6 (2.64)** 47.0 28.1 (3.81)***
  - Tanzania 40.4 33.4 (5.62)*** 30.6 31.2 ( -0.40)
  - Zimbabwe 26.5 94.9 (-10.38)*** ...... 
  - Albania 32.5 44.9 (-5.83)*** 76.1 68.8 (3.30)**
  - Armenia 68.7 54.3 (1.99)* 64.8 48.4 (2.94)**
  - Azerbaijan, Rep. of 78.4 48.5 (6.45)*** 67.5 40.4 (5.31)***
  - Bolivia 83.5 44.9 (6.68)*** 93.6 57.8 (4.55)***
  - Cambodia 96.1 97.0 (-1.15) 94.7 97.6 (-2.74)**
  - Canada 26.9 41.3 (-6.79)*** 33.7 38.8 (-1.51)
  - Chile 10.5 14.4 (-2.71)** 10.3 9.9 (0.37)
  - Peru 71.4 24.6 ( 6.39)*** ...... 
  - Nicaragua 93.9 89.1 (7.79)*** 100.0 100.0 (1.33)
  - Paraguay 58.8 34.5 (6.20)*** 45.1 34.1 (3.90)***
  - Uruguay 85.0 71.7 (6.23)*** 63.7 54.7 (3.75)***
  - Uzbekistan 30.6 22.4 (5.09)*** 76.2 40.8 (8.95)***

- Notes accompanying tables:
  - "In parenthesis student t-stadistic for the null hyphothesis of difference between means equal zero"
  - "***, ** and * correspond to 0.01, 0.05 and 0.1 significance level respectively"

- Table A.2. Dollarization Levels by Region — regional averages and standard deviations (as presented):
  - Sub-Saharan Africa 19.9 24.6 21.8 26.7 (-0.74) 14.2 23.2 16.6 26.2 (-0.89)
  - Latin America & the Caribbean 29.8 30.6 25.2 24.8 (1.31) 26.3 29.1 24.6 25.0 (0.48)
  - East Asia and Pacific 21.6 27.4 19.0 25.5 (0.60) 14.1 25.2 14.5 24.8 (-0.10)
  - Europe and Central Asia 50.1 18.6 45.7 14.9 (1.69)* 50.6 23.6 40.7 20.1 (2.86)***
  - High-income OECD members 5.6 10.3 7.2 12.1 (-0.53) 9.9 14.3 5.4 11.3 (1.17)
  - High-income nonOECD members 24.5 21.2 21.0 18.7 (0.74) 7.2 9.8 12.6 10.9 (-2.19)**
  - Middle East and North Africa 12.4 10.6 11.4 9.4 (0.35) 4.7 9.2 7.5 11.3 (-0.93)
  - South Asia 17.1 25.1 24.0 28.1 (-1.00) 14.6 26.6 23.3 32.5 (-1.14)
- Chart labels: Avg 2001-06, Avg 2007-11, S.D 2001-06, S.D 2007-11 (for deposits and loans by region)

### Annex 2 — Measures to Mitigate Dollarization (policy measures and instruments)
- Durable de-dollarization relies on a credible disinflation plan and specific microeconomic (market-based) measures that:
  - facilitate financial sector development,
  - increase the attractiveness of the local currency, and
  - internalize risks associated with the use of foreign currencies.

- Measures aimed at facilitating the development of the local financial market:
  - Strengthening the payments system and enhancing the usability of the local currency. "Dollarization often reflects the inadequacy of the domestic payments system in local currency."
  - Strengthening the central bank liquidity management and instruments, including:
    - "(1) imposing higher reserve requirements for foreign currency than for local currency;"
    - "(2) strengthening the monetary transmission channels by introducing longer maturities, medium-term government bonds in local currency, and developing a benchmark yield curve;"
    - "(3) developing a foreign exchange market that reduces the need to hold precautionary foreign currency balances."
  - Developing domestic financial market and retail banking. "Issuance of local currency-denominated securities or assets should help develop domestic liquid monetary and capital and bond markets." Encourage retail banking to leverage most retail transactions being denominated in local currency.
  - Strengthening fiscal and public debt management. "Fiscal prudence and consolidation, coupled with the issuance of local currency-denominated bonds, would help de-dollarize the government’s balance sheet." "Foreign aid ought to be absorbed in local currency and taxation designed so that it does not discriminate in favor of foreign currencies."

- Measures aimed at increasing the attractiveness of the local currency:
  - Creating an interest rate wedge. "The substitution between foreign currency and local currency can be encouraged by positive interest differentials that provide a higher remuneration to deposits in local currency; this has, however, in some cases, encouraged excessive and potentially destabilizing capital inflows."

- Measures aimed at internalizing the risks associated with the use of foreign currencies:
  - Implementing effective supervision and prudential regulations. Example: "banks lending in foreign currency to nonexporters have to abide by stricter regulatory requirements and to set up higher provisions."
  - Excluding foreign currency deposits from deposit insurance schemes.

- Direct and administrative controls (examples and country mentions as presented):
  - Imposing higher reserve requirements on foreign currency. Examples cited: "Bolivia, Honduras, Israel, and Nicaragua."
  - Mandating the use of local currency in domestic transactions. Examples cited: "Israel, Lao P.D.R., and Peru."
  - Introducing regulatory discrimination against the use of foreign currency. Examples cited: "Angola, Argentina, Israel, Turkey, and Vietnam."

- Forced de-dollarization (example and cautionary note):
  - Mandating the conversion of foreign currency obligations and balances into domestic currency. Examples cited: "Bolivia, Mexico and Peru" — "Some countries ... have attempted to use this measure but reversed it after it triggered capital flight and reduced financial intermediation."

*Source: _afr1504 - References (PDF)._

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/dp/2015/_afr1504.pdf_