## _wp1511

## Source details

**Canonical URL:** [_wp1511](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp1511.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp1511.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp1511.pdf.json)

---

### I. Introduction and scope
- Study objective:
  - Document stylized facts of credit booms in low- and lower-middle income countries versus upper-middle countries.
  - Analyze macroeconomic behavior around booms.
  - Identify characteristics of booms that end in banking crises.
  - Examine relationship between credit booms and financial deepening.
- Sample:
  - Comprehensive bank credit database covering 135 developing countries over the period 1960–2011.
  - Alternative sample considered: shorter sample period starting in 1994.
- Credit measure:
  - Sum of claims on the private sector by deposit money banks (line 22d) plus, when available, claims on the private sector by other financial institutions (line 42d).

### II. Methodology for identifying credit booms
- Primary methodology: Mendoza and Terrones (2012) approach.
- Definition:
  - A credit boom is an episode in which the deviation in (log) real credit per capita from its country-specific long-run trend (HP-filter) exceeds the standard deviation of the cyclical component by a factor of 1.65.
  - Note: 1.65 falls in the 5 percent tail of the standardized normal distribution.
  - Sensitivity: factor of 2 also considered; qualitative results largely unchanged.
- Event-study construction:
  - Seven-year event windows of cyclical components centered on the peak of credit booms (t=0).
- Robustness checks:
  - Mendoza and Terrones methodology implemented using credit-to-GDP.
  - Appendix 4 presents Tornell and Westermann methodology results.

### III. Incidence, magnitude, and duration of credit boom episodes (stylized facts)
- Total identified credit boom episodes (full sample): 111.
- Incidence by income group (full sample):
  - Low-income countries: 21 percent of episodes.
  - Lower-middle-income countries: 38 percent of episodes.
  - Upper-middle-income countries: 29 percent of episodes.
- Time distribution:
  - About 60 percent of boom episodes occurred in the past 20 years.
- Magnitude at peak of boom:
  - Average expansion in real credit per capita: around 30 percent above trend across income groups.
  - Normalized by variability: medians of around 2 across income groups.
  - Highest absolute peak expansion observed in low-income countries.
- Duration:
  - Median duration (full sample): 6 years (median and mean roughly 6 years across income groups).
  - Booms since 1994: tended to last one year less (4–5 years).
- Build-up phase growth rates:
  - Preceding 3 years: between 50 to 70 percent of episodes had real credit per capita growth rate of at least 5 percent.
  - Preceding 5 years: between 30 to 40 percent of episodes had real credit per capita growth rate of at least 5 percent.
  - These proportions are higher for low-income and upper-middle-income countries relative to lower-middle-income countries.
- Coincidence across samples:
  - Around 80 percent of credit boom episodes identified for middle-income countries coincide in the two samples (full and 1994–2011).
  - For low-income countries, coincidence is around 60 percent, with many new booms identified after the global financial crisis.

### IV. Comparison: real credit per capita vs. credit-to-GDP identification
- Correlation between booms identified by the two measures (Mendoza and Terrones methodology):
  - Correlation since 1970 (All countries): 0.47
    - Low-income countries: 0.52
    - Lower-middle income countries: 0.55
    - Upper-middle income countries: 0.41
  - Correlation since 1990 (All countries): 0.49
    - Low-income countries: 0.66
    - Lower-middle income countries: 0.53
    - Upper-middle income countries: 0.44

### V. Macroeconomic dynamics around credit booms (event study findings)
- Event study variables: output, investment, private and public consumption, non-tradable output, real exchange rate, broad money, external demand, current account to GDP ratio, capital inflows to GDP ratio.
- Key cross-country patterns:
  - Dynamics more similar in lower- and upper-middle income countries than in low-income countries.
  - Low-income countries: only a subset of macro variables exhibit defined dynamics (economic activity, investment, non-tradable output, money balances, external demand).
  - Middle-income countries: broader cyclical patterns—expansion in upswing followed by steep contraction in downswing.
- Middle-income specifics (medians):
  - Real exchange rate: appreciates by around 4 percent above-trend in build-up then drops to around 3 percent.
  - Capital inflows (gross other investments to GDP): median above-trend increase of around one percentage point of GDP by t= -1.
  - Money balances (M2): expansion up to 10 percentage points above trend in build-up.
  - Non-tradable output: expansion up to 5 percentage points above trend in build-up.
- Low-income specifics:
  - Investment: increase of ~8 percentage points above trend during expansion, followed by contraction.
  - Current account: deterioration mainly at peak (t) and at t=1; cyclical component close to zero during build-up.
  - Capital inflows ratio: no clear dynamics around booms.
- Post-1994 episodes:
  - Qualitative results largely similar; some differences in real exchange rate dynamics and larger deviations around peak for private consumption and current account.

### VI. Drivers and correlates of credit booms (frequency analysis and regressions)
- Frequency analysis:
  - Surges in capital inflows associated with credit booms.
  - Between 30 and 44 percent of credit boom episodes were preceded by an increase in capital inflows (definition: preceding three-year average of ratio of gross investments to GDP above median for each income group).
  - Credit booms most often observed under less flexible exchange rate regimes.
    - Proportion of credit booms under fixed or managed regimes:
      - Lower-middle countries: 62 percent.
      - Upper-middle countries: 73 percent.
      - Low-income countries: 46 percent.
  - Financial liberalization (modified Abiad et al. 2008 index):
    - Above-median preceding increases in financial liberalization:
      - Low-income: around 22 percent of credit booms.
      - Lower-middle income: 24 percent.
      - Upper-middle income: 38 percent.
- Panel probit regressions (1970–2011):
  - Positive associations between incidence of credit booms and:
    - Capital inflows.
    - Favorable international liquidity conditions.
    - Sustained growth in economic activity.
    - Sustained growth in domestic private credit.
    - Initial level of financial depth.
  - Key coefficient estimates (Table 9, marginal significance noted):
    - Gross other investments/GDP: coefficient 0.11; Standard error 0.06; P-value 0.06 *
    - Private sector credit/GDP: coefficient 0.18; Standard error 0.04; P-value 0.00 ***
    - GDP growth rate: coefficient 0.25; Standard error 0.11; P-value 0.02 **
    - Growth rate of private sector credit-to-GDP: coefficient 0.18; Standard error 0.04; P-value 0.00 ***
  - Interpretation example:
    - If ratio of gross investments increases to around 20 percent of GDP from sample median of around 2 percent, probability of a credit boom increases by about 40 percent.
  - Robustness:
    - Second-year lag of capital inflows ratio positive and significant.
    - Results robust to variation in sample, boom criteria, and alternative inflow measures.

### VII. Credit booms associated with banking crises
- Incidence:
  - 23 out of 111 credit boom episodes associated with a banking crisis (Laeven and Valencia 2012).
  - Distribution:
    - 50 percent in upper-middle income countries.
    - 25 percent in lower-middle income countries.
    - Around 10 percent in low-income countries.
- Characteristics (Table 11):
  - Duration (median, years):
    - Crisis-associated booms: 6.0 (median), mean 6.4.
    - Non-crisis booms: 6.0 (median), mean 5.6.
  - Credit-to-GDP ratio (%), median before start (average of t-1 and t-2):
    - Crisis-associated: 25 (median), mean 29.
    - Non-crisis: 19 (median), mean 24.
  - Expansion in credit-to-GDP ratio (%) during boom:
    - Crisis-associated: 15 (median), mean 38.
    - Non-crisis: 9 (median), mean 22.
- Macroeconomic dynamics:
  - Crisis-associated booms show larger fluctuations and more abrupt declines than non-crisis booms.
  - Many cyclical components are above trend at least 3 years before the peak (current account below-trend), indicating accumulation of imbalances.

### VIII. Relationship between credit booms and financial deepening
- Outcomes for non-crisis booms:
  - 76 percent of credit booms not associated with banking crises exhibit an increase in level of real credit per capita by end of boom (time t).
  - Proportions over time (Table 12):
    - All income groups: End of boom 76; End of boom + one year 70; End of boom + two years 67.
    - Low-income countries: End of boom 70; End of boom + one year 71; End of boom + two years 63.
    - Excluding low-income countries: End of boom 78; End of boom + one year 72; End of boom + two years 68.
  - Using credit-to-GDP definition (Appendix 4): 70 percent of non-crisis booms show increase in credit-to-GDP two years after end of boom.
- Benchmarking (financial possibility frontier, Barajas et al. 2013 approach):
  - Predicted real credit per capita regressed on structural factors (log GDP per capita, log population, log population density, log age dependence ratio, offshore dummy, transition dummy, fuel-exporter dummy, time effects) using pooled OLS (1970–2011).
  - Regression highlights:
    - Log GDP per capita: 0.4473*** (std. error 0.0650) and 0.4623*** (std. error 0.0740) in two specifications.
    - R_Squared: 0.79 and 0.80; Number of observations: 3864 and 3365.
  - Findings:
    - Large negative changes in gap between actual credit and benchmark associated with higher likelihood of boom episodes.
    - About 80 percent of credit booms associated with banking crises occurred when real credit per capita was well above levels predicted by structural and macroeconomic fundamentals.
    - Barajas et al. (2013) evidence: probability of a crisis surpasses 10 percent for negative gaps (predicted minus actual) of over 50 percent; rapid deepening above structural changes (e.g., by 30 percent or more over a ten year period) further increases likelihood of instability.

### IX. Data, appendixes, and robustness
- Appendixes:
  - Appendix 1: list of 135 developing countries by income groups (2013 World Bank classification).
  - Appendix 2: description and definition of main variables.
  - Appendix 3: boom episodes list.
  - Appendix 4: robustness using Mendoza and Terrones’ methodology and Tornell and Westermann’s approach.
  - Additional appendixes: significance of event window coefficients; robustness—credit booms and net capital inflows; robustness—credit booms, financial reforms and banking supervision.
- Sample composition note (Appendix 1 summary):
  - Out of 135 developing countries:
    - 32 low-income,
    - 45 lower-middle-income,
    - 43 upper-middle-income,
    - 15 high-income.
  - 65 are eligible for the IMF’s Poverty Reduction and Growth Facility Trust (PRGT), Zimbabwe and 69 emerging markets (EM) — as reported in the source text.

### X. Conclusions and policy recommendations
- Stylized facts:
  - Duration and magnitude of credit booms in low-income countries broadly similar to middle-income countries, but aggregate behavior differs.
  - As low-income countries develop, monitor cyclical behavior of output, investment, private and public consumption, real exchange rate, broad money, external demand, current account, and capital inflows to GDP ratios.
- Monitoring signals:
  - Monitor duration and magnitude (deviation from trend) of credit expansion together with cyclical behavior of pertinent variables; large and longer deviations from trend signal unsustainable imbalance build-up.
- Role of capital flows, exchange rate regimes, and liberalization:
  - Capital inflows are important drivers and predictors of credit booms.
  - Less flexible exchange rate regimes can amplify credit expansions during capital inflow surges, increasing vulnerability to reversals.
  - Financial liberalization can trigger credit booms if not accompanied by prudent regulation and supervision.
- Distinguishing benign deepening from risky booms:
  - Credit booms associated with banking crises show distinct macro dynamics and large deviations from fundamentals (about 80 percent above-benchmark).
  - Comparing observed financial deepening with levels predicted by structural fundamentals can help gauge unsustainable expansions.
- Policy measures to mitigate unsustainable credit expansions and support safe financial deepening:
  - Hedge against currency, maturity, and balance sheet mismatches as financial interconnection grows.
  - Impose limits on net foreign open positions consistent with pace of financial deepening.
  - Ensure financial institutions do not rely solely on short-term funding; increase capital risk weights on volatile funding sources.
  - Mitigate high credit risk from rapid expansion via proper provisioning standards and tightening definitions of NPLs.
  - Strengthen banking supervision (increase on-site audits, monitor vulnerabilities related to cross-border funding).

*Source: _wp1511 - Appendix 4 also presents the results using Tornell and Westermann methodology for credit boom (PDF chapter/section).*

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

### _wp1511 - References .............................................................................................................

### I. Introduction and scope
- Study objective: document stylized facts of credit booms in low- and lower-middle income countries versus upper-middle countries; analyze macroeconomic behavior around booms; identify characteristics of booms that end in banking crises; examine relationship between credit booms and financial deepening.
- Sample: comprehensive bank credit database covering 135 developing countries over the period 1960–2011.
- Credit measure: sum of claims on the private sector by deposit money banks (line 22d) plus, when available, claims on the private sector by other financial institutions (line 42d).
- Alternative sample considered: shorter sample period starting in 1994 to account for structural breaks related to late 1980s–early 1990s reforms.

### II. Methodology for identifying credit booms
- Primary methodology: Mendoza and Terrones (2012) approach.
- Definition: a credit boom is an episode in which the deviation in (log) real credit per capita from its country-specific long-run trend (HP-filter) exceeds the standard deviation of the cyclical component by a factor of 1.65.
  - Note: 1.65 falls in the 5 percent tail of the standardized normal distribution.
  - Sensitivity check: a factor of 2 was also considered; qualitative results remained largely unchanged.
- Event-study construction: seven-year event windows of cyclical components centered on the peak of credit booms (t=0).
- Robustness checks: implemented Mendoza and Terrones methodology using credit-to-GDP; Appendix 4 and other robustness appendixes referenced.

### III. Incidence, magnitude, and duration of credit boom episodes (stylized facts)
- Total identified credit boom episodes (full sample): 111.
- Incidence by income group (full sample):
  - Low-income countries: 21 percent of episodes.
  - Lower-middle-income countries: 38 percent of episodes.
  - Upper-middle-income countries: 29 percent of episodes.
- Time distribution: about 60 percent of boom episodes occurred in the past 20 years.
- Magnitude at peak of boom:
  - Average expansion in real credit per capita: around 30 percent above trend across income groups.
  - Normalized by variability (standard deviation of cyclical component): medians of around 2 across income groups.
  - Highest absolute peak expansion observed in low-income countries.
- Duration:
  - Median duration (full sample): 6 years (median and mean roughly 6 years across income groups).
  - Booms since 1994: tended to last one year less (4–5 years).
- Build-up phase growth rates:
  - In the preceding 3 years, between 50 to 70 percent of credit boom episodes exhibited a growth rate of real credit per capita of at least 5 percent.
  - In the preceding 5 years, between 30 to 40 percent of episodes exhibited a growth rate of real credit per capita of at least 5 percent.
  - These proportions are higher for low-income and upper-middle-income countries relative to lower-middle-income countries.
- Coincidence across samples:
  - Around 80 percent of credit boom episodes identified for middle-income countries coincide in the two samples (full and 1994–2011).
  - For low-income countries, proportion is around 60 percent, with many new booms identified after the global financial crisis.

### IV. Comparison of identification using real credit per capita vs. credit-to-GDP
- Correlation between booms identified by the two measures (Mendoza and Terrones methodology):
  - Correlation since 1970 (All countries): 0.47
    - Low-income countries: 0.52
    - Lower-middle income countries: 0.55
    - Upper-middle income countries: 0.41
  - Correlation since 1990 (All countries): 0.49
    - Low-income countries: 0.66
    - Lower-middle income countries: 0.53
    - Upper-middle income countries: 0.44

### V. Macroeconomic dynamics around credit booms (event study findings)
- Event study variables include: output, investment, private and public consumption, non-tradable output, real exchange rate, broad money, external demand, current account to GDP ratio, capital inflows to GDP ratio.
- Key cross-country patterns:
  - Macroeconomic dynamics around credit booms are more similar in lower- and upper-middle income countries than in low-income countries.
  - In low-income countries, only a subset of macroeconomic variables (economic activity, investment, non-tradable output, money balances, and external demand) exhibit a defined dynamics around credit booms.
  - For middle-income countries, broader cyclical patterns are observed (expansion in upswing followed by steep contraction in downswing), consistent with prior literature.

### VI. Drivers and correlates of credit booms (regression and frequency analysis)
- Frequency analysis:
  - Surges in capital inflows are associated with credit booms.
  - A higher proportion of credit booms in middle-income countries are preceded by an increase in capital inflows compared to low-income countries.
  - A significant proportion of credit booms are associated with fixed or managed exchange rates across all income groups.
- Panel probit regressions (1970–2011) indicate positive associations between the incidence of credit booms and:
  - Capital inflows.
  - Favorable international liquidity conditions.
  - Sustained growth in economic activity.
  - Sustained growth in domestic private credit.
  - Initial level of financial depth.

### VII. Credit booms associated with banking crises
- Distinct characteristics:
  - Booms associated with banking crises exhibit distinct macroeconomic dynamics compared with non-crisis booms.
  - Macroeconomic fluctuations in crisis-associated booms are larger and display more abrupt declines.
- Benchmarking analysis (Barajas et al. (2013) approach):
  - Credit growth in the majority of booms associated with banking crises was well above levels consistent with country-specific structural factors and fundamentals.

### VIII. Relationship between credit booms and financial deepening
- Findings:
  - An overshooting of private credit-to-GDP ratio significantly beyond levels predicted by structural fundamentals is associated with credit booms and busts.
  - Barajas et al. (2013) evidence referenced: probability of a crisis surpasses 10 percent for negative gaps (predicted minus actual) of over 50 percent.
  - Rapid deepening above structural changes (e.g., by 30 percent or more over a ten year period) further increases likelihood of instability.
- Assessment approach: benchmarking exercise to assess whether financial depth is in line with structural and policy fundamentals (results summarized above for crisis-associated booms).

### IX. Data, appendixes, and robustness
- Data sources and appendixes:
  - Appendix 1: list of 135 developing countries by income groups (2013 World Bank classification).
  - Appendix 2: description and definition of main variables.
  - Appendix 3: boom episodes list.
  - Appendix 4: robustness using Mendoza and Terrones’ methodology and application of Tornell and Westermann’s approach.
  - Additional appendixes: significance of event window coefficients; robustness—credit booms and net capital inflows; robustness—credit booms, financial reforms and banking supervision.
- Sample composition note (Appendix 1 summary in text):
  - Out of 135 developing countries: 32 low-income, 45 lower-middle-income, 43 upper-middle-income, and 15 high-income (put differently, 65 are eligible for the IMF’s Poverty Reduction and Growth Facility Trust (PRGT), Zimbabwe and 69 emerging markets (EM) — as reported in the source text).

*Source: _wp1511 - References .............................................................................................................*

### Appendix 4 also presents the results using Tornell and Westermann methodology for credit boom

### _wp1511 - Appendix 4 also presents the results using Tornell and Westermann methodology for credit boom

### Dynamic behavior of macroeconomic variables around credit boom peaks
- Sample and methodology:
  - All variables measured in constant prices, per-capita terms and de-trended using the HP filter, except current account and capital inflows to GDP ratios (in current prices and not de-trended), and the real exchange rate.
- Middle-income countries (lower-middle and upper-middle):
  - Most variables: increase during expansionary phase and decrease during unwinding.
  - Investment: larger expansion and subsequent contraction compared to private and government consumption.
  - Current account ratio: deteriorates during build-up and improves as boom unwinds (medians).
  - Real exchange rate: appreciates by around 4 percent above-trend in the build-up phase and then drops to around 3 percent.
  - Capital inflows (gross other investments to GDP): median above-trend increase of around one percentage point of GDP by t= -1.
  - Money balances (M2): expansion in build-up phase up to 10 percentage points above trend, followed by a decline.
  - External demand (weighted average GDP growth of trading partners): well-defined pattern, above-trend in build-up.
  - Non-tradable output: expansion up to 5 percentage points above trend in build-up, then decline.
  - Normalization by standard deviations indicates similar amplitude across income groups despite larger absolute amplitude in upper-middle income countries.
  - Association with inflation rates and terms-of-trade: less clear cut (not reported here).
- Low-income countries:
  - Output: defined dynamics around peak similar to middle-income countries.
  - Domestic demand components: only investment exhibits an increase during expansion (~8 percentage point above trend) followed by contraction.
  - Government expenditure: procyclical in middle-income countries; no discernible pattern in low-income countries.
  - Consumption: no clear dynamics (measurement issues).
  - Current account ratio: deterioration mainly at peak (time t) and at t=1; cyclical component close to zero during build-up.
  - Non-tradable output: expansion during build-up, maximum deviation at t=-1 (one period before peak).
  - External demand: above-trend expansion during build-up but declines before peak.
  - Capital inflows ratio: no clear dynamics around booms (heterogeneity in financial integration and depth).
  - Monetary expansions (policy easing, relaxed lending standards, external demand shocks) can fuel significant credit expansions impacting investment and output.

### Differences when focusing on post-1994 episodes
- Qualitative results largely similar, with differences shown in Figure 3:
  - Real exchange rate cyclical component shows clear dynamics in lower- and upper-middle countries when looking at mean values.
  - Real GDP: no clear dynamics for low-income countries, but private consumption and investment do.
  - Larger deviations around peak for private consumption and current account to GDP ratio.

### Potential drivers of credit booms (frequency analysis)
- Triggers examined: (i) surges in capital inflows, (ii) financial sector reform, (iii) exchange rate regimes.
- Capital inflows surge definition: preceding three-year average of ratio of gross investments to GDP above median for each income group.
  - Between 30 and 44 percent of credit boom episodes were preceded by an increase in capital inflows.
  - Proportion increases by income group (more important in middle-income countries than low-income countries).
- Exchange rate regimes (Ilzetzki, Reinhart and Rogoff 2011 classification into five groupings):
  - Credit booms most often observed under less flexible regimes.
  - Proportion of credit booms under fixed or managed regimes:
    - 62 percent for lower-middle countries.
    - 73 percent for upper-middle countries.
    - 46 percent for low-income countries.
  - Caveat: many low-income countries belong to a monetary union.
- Financial liberalization (modified Abiad et al. 2008 index excluding banking supervision):
  - Country identified as having an increase in financial liberalization when preceding three-year change is above income-group median.
  - Results:
    - Around 22 percent of credit booms in low-income countries preceded by above-median increases in the index.
    - 24 percent for lower-middle income countries.
    - 38 percent for upper-middle income countries.
  - Credit booms prior to 1990 largely associated with low quality of banking supervision (lowest value of index).
- Banking supervision (Abiad et al. 2008 index ranges 0–3):
  - Table 8 (index distribution for booms):
    - Number of booms associated with index values 0,1,2,3: 34, 11, 6, 2.
    - Percentage: 64, 21, 11, 4.
    - Number of booms since 1990 associated with index values 0,1,2,3: 11, 11, 6, 2.
    - Percentage since 1990: 21, 21, 11, 4.
  - Note: index data coverage limited (~47 percent of sample; available for 91 countries from 1973–2005).

### Empirical analysis of incidence of credit booms (panel Probit, 1970–2011)
- Dependent variable: dummy =1 if a credit boom episode observed (start to end), 0 otherwise.
- Capital inflow measures: ratio of gross investments (excluding FDI) and gross other investments (excluding portfolio flows and FDI).
- Controls: credit growth, real GDP growth, inflation rates, exchange rate regime flexibility, depth of domestic financial system, external demand, terms-of-trade, U.S. federal funds rate (cyclical component), trade openness. All explanatory variables lagged.
- Key regression findings (Table 9 and marginal effects in Table 10; marginal effects evaluated at medians from specification [1]):
  - Gross other investments/GDP: coefficient 0.11; Standard error 0.06; P-value 0.06 *
  - Private sector credit/GDP: coefficient 0.18; Standard error 0.04; P-value 0.00 ***
  - GDP growth rate: coefficient 0.25; Standard error 0.11; P-value 0.02 **
  - Growth rate of private sector credit-to-GDP: coefficient 0.18; Standard error 0.04; P-value 0.00 ***
- Interpretation example from Figure 4 and text:
  - If ratio of gross investments increases to around 20 percent of GDP from sample median of around 2 percent, probability of a credit boom increases by about 40 percent.
- Robustness:
  - Second-year lag of capital inflows ratio has a positive and significant effect on incidence of credit booms.
  - Results robust to variation in sample, boom criteria, and alternative explanatory variables (not reported here).
  - Appendix 6: using ratio of net other investments to GDP gives similar positive association.
- Financial reforms and supervision indexes (Appendix 7 regressions):
  - Coefficient signs: financial reforms increase probability of booms; better banking supervision and prudential regulations decrease probability.
  - Coefficients not statistically significant.
  - Caveat: indices cover only 60 countries in sample, reducing observations by half.

### Credit booms, banking crises, and financial deepening
- Incidence of banking crises among credit booms:
  - 23 out of 111 credit boom episodes associated with a banking crisis (Laeven and Valencia 2012).
  - Most of these episodes:
    - 50 percent in upper-middle income countries.
    - 25 percent in lower-middle income countries.
    - Around 10 percent in low-income countries.
- Characteristics of booms associated with banking crises (Table 11):
  - Duration (median, years):
    - Booms associated with banking crises: 6.0 (median), mean 6.4.
    - Booms not associated with banking crises: median 6.0, mean 5.6.
    - Note: median for crisis-associated booms smaller than 6 years when excluding the longest-duration boom.
  - Credit-to-GDP ratio (%), median before start (average of t-1 and t-2):
    - Crisis-associated: 25 (median), mean 29.
    - Non-crisis: 19 (median), mean 24.
  - Expansion in credit-to-GDP ratio (%) during boom:
    - Crisis-associated: 15 (median), mean 38.
    - Non-crisis: 9 (median), mean 22.
- Macroeconomic dynamics in crisis vs non-crisis booms:
  - Event study (Figure 5, medians): macroeconomic variables show well-defined patterns in crisis-associated booms with larger fluctuations and more abrupt declines than in non-crisis booms.
  - For crisis-associated booms, many cyclical components are above trend at least 3 years before the peak (current account below-trend), indicating accumulation of imbalances (real exchange rate appreciation, current account deterioration, capital inflow increases).
- Financial deepening outcomes for non-crisis booms:
  - Three quarters (76 percent) of credit booms not associated with banking crises exhibit an increase in level of real credit per capita by end of boom (time t).
  - Proportions over time (Table 12):
    - All income groups: End of boom 76; End of boom + one year 70; End of boom + two years 67.
    - Low-income countries: End of boom 70; End of boom + one year 71; End of boom + two years 63.
    - Excluding low-income countries: End of boom 78; End of boom + one year 72; End of boom + two years 68.
  - Using credit-to-GDP definition (Appendix 4): 70 percent of non-crisis booms show increase in credit-to-GDP two years after end of boom.
  - Majority of booms not associated with crises or deepening occurred in contexts of downward financial intermediation trends (about 70 percent before 1993).
- Benchmarking/financial possibility frontier (Barajas et al. 2013 approach):
  - Predicted level of financial deepening estimated by regressing real credit per capita on structural factors (GDP per capita, population, population density, age dependency ratio, dummy variables for off-shore center, transition country, oil-exporter, and time effects) using pooled OLS (1970–2011).
  - Findings:
    - Large negative changes in gap between actual credit and benchmark associated with higher likelihood of boom episodes.
    - About 80 percent of credit booms associated with banking crises occurred when real credit per capita was well above levels predicted by structural and macroeconomic fundamentals.

### Conclusions and policy recommendations
- Stylized facts:
  - Duration and magnitude of credit booms in low-income countries share broadly similar characteristics with middle-income countries, but macroeconomic aggregate behavior differs.
  - As low-income countries develop, monitor cyclical behavior of output, investment, private and public consumption, real exchange rate, broad money, external demand, current account, and capital inflows to GDP ratios—these tend to increase during expansion and decrease during unwinding in middle-income contexts.
- Monitoring signals:
  - Monitor duration and magnitude (deviation from trend) of credit expansion together with cyclical behavior of pertinent variables; large and longer deviations from trend signal unsustainable imbalance build-up.
- Role of capital flows, exchange rate regimes, and liberalization:
  - Capital inflows are important drivers and predictors of credit booms.
  - Less flexible exchange rate regimes can amplify credit expansions during capital inflow surges, increasing vulnerability to reversals.
  - Financial liberalization can trigger credit booms if not accompanied by prudent banking regulation and supervision.
- Distinguishing benign deepening from risky booms:
  - Credit booms associated with banking crises show distinct macro dynamics and large deviations from fundamentals (about 80 percent above-benchmark).
  - Comparing observed financial deepening with levels predicted by structural fundamentals can help gauge unsustainable expansions.
- Policy measures to mitigate unsustainable credit expansions and support safe financial deepening:
  - Hedge against currency, maturity, and balance sheet mismatches as financial interconnection grows.
  - Impose limits on net foreign open positions consistent with pace of financial deepening.
  - Ensure financial institutions do not rely solely on short-term funding; increase capital risk weights on volatile funding sources.
  - Mitigate high credit risk from rapid expansion via proper provisioning standards and tightening definitions of NPLs.
  - Strengthen banking supervision (increase on-site audits, monitor vulnerabilities related to cross-border funding).
  - These policies can help low-income countries achieve financial deepening aligned with macroeconomic fundamentals.

*Italic: Source — content unit from _wp1511 - Appendix 4 also presents the results using Tornell and Westermann methodology for credit boom (PDF chapter/section).*

### REFERENCES

### _wp1511 - REFERENCES

### References
- Abiad, Abdul, Erica Detragiache, and Thierry Tressel, 2008, “A New Database of Financial Reforms,” IMF Working Paper No. 08/266.
- Barajas, Adolfo, Giovanni Dell’Ariccia, and Andrei Levchenko, 2009, “Credit Booms: The Good, the Bad, and the Ugly,” Unpublished manuscript, International Monetary Fund (Washington, DC).
- Barajas, Adolfo, Thorsten Beck, Era Dabla-Norris, and Seyed Reza Yousefi, 2013, “Too Cold, Too Hot, Or Just Right? Assessing Financial Sector Development Across the Globe,” IMF Working Paper No. 13/81.
- Borio, Claudio, and Philippe Lowe, 2002, “Assets Prices, Financial and Monetary Stability: Exploring the Nexus,” BIS Working Paper No. 114 (Basel: Bank for International Settlements).
- Borio, Claudio and Mathias Drehmann, 2009, “Assessing the Risk of Banking Crises – Revisited,” BIS Quarterly Review, March, pp. 29–46.
- Calderón, Cesar and Megumi Kubota, 2012, “Gross inflows gone wild: gross capital inflows, credit booms and crises,” World Bank Policy Research Working Paper No. 6270.
- Chinn, Menzie and Hiro Ito, 2006, “"What Matters for Financial Development? Capital Controls, Institutions, and Interactions,” Journal of Development Economics, Volume 81, Issue 1, Pages 163–192 (October). Database updated until 2011 http://web.pdx.edu/~ito/Chinn-Ito_website.htm
- Claessens, Stijn, Mayhan Kose, and Marco Terrones, 2011, “Financial Cycles: What? How? When?” IMF Working Paper No. 11/76.
- Collyns, Charles and Abdelhak Senhadji. 2002. “Lending Booms, Real Estate Bubbles, and the Asian Financial Crisis.” IMF Working Paper No. 02/20.
- Decressin, Jörg and Marco Terrones, 2011, “Credit Boom-Bust Cycles: Their Triggers and Policy Implications,” World Economic Outlook September 2011: Slowing Growth, Rising Risks, Chapter 1, Box 1.2.
- Dell'Ariccia, Giovanni, Deniz Igan, Luc Laeven, and Hui Tong, 2012, “Policies for Macrofinancial Stability: How to Deal with Credit Booms,” IMF Staff Discussion Note, 12/06.
- Gourinchas, Pierre-Olivier, Rodrigo Valdes and Oscar Landerretche, 2001, “Lending Booms: Latin America and the World,” Economia, Volume 1, No 2, Spring, pp. 47–99.
- IMF, 2012, “Enhancing Financial Sector Surveillance in Low-Income Countries: Financial Deepening and Macro-Stability.”
- Ilzetzki, Ethan, Carmen Reinhart, and Kenneth Rogoff, 2011, ”The Country Chronologies and Background Material to Exchange Rate Arrangements into the 21st Century: Will the Anchor Currency Hold?” http://personal.lse.ac.uk/ilzetzki/index.htm/
- Laeven, Luc and Fabian Valencia. 2012, “Resolution of Banking Crises: The Good, the Bad, and the Ugly,” IMF Working Paper 12/163.
- Mitra Srobona, Jaromír Benes, Silvia Iorgova, Kasper Lund-Jensen, Christian Schmieder, and Tiago Severo. 2011, “Toward Operationalizing Macroprudential Policies: When to Act?” Chapter 3 in Global Financial Stability Report, September (Washington, DC: International Monetary Fund)
- Magud, Nicolas, Carmen Reinhart, and Esteban Vesperoni. 2012, “Capital Inflows, Exchange Rate Flexibility, and Credit Booms.” IMF Working Paper 12/41.
- Mendoza, Enrique and Marco Terrones, 2008, “An Anatomy of Credit Booms: Evidence from Macro Aggregates and Micro Data,” NBER Working Paper No. 14049.
- Mendoza, Enrique and Marco Terrones, 2012, “An Anatomy of Credit Booms and their Demise,” NBER Working Paper No. 18379.
- _____, 2008, “An Anatomy of Credit Booms: Evidence from Macro Aggregates and Micro Data,” NBER Working Paper No. 14049.
- Reinhart, Carmen and Vincent Reinhart, 2008, “Capital Flow Bonanzas: An Encompassing View of the Past and Present,” NBER Working Paper No. 14321.
- Reinhart, Carmen and Kenneth Rogoff, 2009, “This Time is Different: Eight Centuries of Financial Folly,” Princeton: Princeton University Press.
- Schularick, Moritz and Alan Taylor. 2012, “Credit Booms Gone Bust: Monetary Policy, Leverage Cycles, and Financial Crises, 1870–2008,” American Economic Review, vol. 102, issue 2, pp.1029–61.
- Tornell, Aaron and Frank Westermann, 2002, “Boom-Bust Cycles in Middle Income Countries: Facts and Explanation,” IMF Staff Papers, Vol. 49, Special Issue.

### Tables and Figures (titles and key labels)
- Table 1. Number of Credit Booms Episodes: Using Real Credit per capita
- Table 2. Credit Boom Magnitudes: Income Group Summary Statistics
  - Methodology: Mendoza and Terrones (2012)
  - Income groups listed: Low Income; Lower Middle Income; Upper Middle Income
  - De-trending full sample (1960-2011) observations: 234 232 234 232
  - For episodes in the 1990s observations: 91 59
  - For episodes since 2000 observations: 41 19
  - De-trending sample since 1994 observations: 132 42 1?
  - Peak definition note: Peak are defined when the deviation from trend is at least 1.65 times the standard deviation of the cyclical component. Boom parameter=1.65
  - Sample statistics examples: Mean 0.338 0.300 0.321 2.007 2.057 2.066; Median 0.327 0.323 0.283 1.967 2.000 1.974; Std. Dev 0.154 0.124 0.164; Obs. 234 232 234 232
  - De-trending sample since 1994 examples: Mean 0.211 0.283 0.272 1.469 1.807 1.701; Median 0.163 0.261 0.266 1.533 1.901 1.762; Std. Dev 0.131 0.120 0.165; Obs. 132 42 1 132 42 1
  - Deviation from trend at peak / Std. Dev. of credit (column header)
- Table 3. Duration of Credit Boom Episodes (years): Using Real Credit per capita
  - Methodology: Mendoza and Terrones (2012)
  - Duration parameter and boom parameter permutations presented:
    - De-trending full sample (1960-2011) A. Mean: 6.3 5.8 5.6 5.8 5.4 5.1 6.0 6.0 6.0 6.0 5.0 5.0
    - For episodes in the 1990s A. Mean: 6.3 5.9 5.9 6.1 5.5 5.2 6.0 5.0 6.0 6.0 5.0 5.0
    - For episodes since 2000 A. Mean: 5.0 5.3 5.3 4.3 4.7 5.2 5.0 5.0 5.0 4.0 5.0 5.0
    - De-trending sample since 1994 A. Mean: 4.5 4.6 5.1 4.0 4.0 4.6 4.0 4.5 5.0 4.0 4.0 4.0
    - For episodes since 2000 A. Mean: 4.4 4.7 5.3 3.9 4.0 5.0 5.0 5.0 5.0 4.0 4.0 5.0
  - Notes:
    - 1/ Start and end of the boom episodes are defined when the deviation from trend is 0.25 times the standard deviation of the cyclical component.
    - 2/ Start and end of the boom episodes are defined when the deviation from trend is 0.75 times the standard deviation of the cyclical component.
    - Duration parameter=0.25, Boom parameter=1.65; Duration parameter=0.75, Boom parameter=1.65
- Table 4. Credit Boom Episodes and Periods of Sustained Credit Growth
  - Boom parameter=1.65
  - Low income: De-trending full sample (1960-2011) counts: 2 3 1 5 5 9 2 (format in table reflects multiple counts by threshold and period)
    - For episodes in the 1990s counts: 9 4 1 3 1
    - For episodes since 2000 counts: 4 3 1 1 0
  - Lower-middle income: De-trending full sample (1960-2011) counts: 4 2 2 1 10 14 7
    - For episodes in the 1990s counts: 15 6 2 4 2
    - For episodes since 2000 counts: 11 9 4 7 4
  - Upper-middle income: De-trending full sample (1960-2011) counts: 3 2 2 2 15 12 4
    - For episodes in the 1990s counts: 9 6 4 5 1
    - For episodes since 2000 counts: 9 8 5 5 2
  - Row labels indicate: Number of Booms Episodes where the growth rate of real credit per capita was at least 5 percent / at least 10 percent for the last 3 years / 5 years (and repeated for other income groups)
- Figure 2. Credit Booms, Economic Activity and Domestic Demand
  - Panels by income group: Low-income; Lower-middle income; Upper-middle income
  - Variables shown (cross-country means and medians of cyclical components): Real GDP; Real Consumption; Real Government Consumption; Real Investment
  - Horizontal axes in each small chart range annotations include: -3 -2 -1 0 1 2 3; vertical scales vary (examples: -1 to 1.5; -4 to 6; -15 to 20)
- Figure 2 (cont’d). Credit Booms, Non-Tradable Output, Real Exchange Rate, and External Demand
  - Panels by income group: Low-income; Lower-middle income; Upper-middle income
  - Variables shown: Non-tradables Output; Real Exchange Rate; External Demand
  - Example vertical ranges: -2 to 2; -4 to 6; -0.6 to 1
- Figure 2 (cont’d). Credit Booms, Capital Inflows, Current Account, and Money
  - Panels by income group: Low-income; Lower-middle income; Upper-middle income
  - Variables shown: Other investments to GDP; Capital Inflows (1/ footnote: Gross capital inflows include portfolio inflows, foreign direct investment, and other investments.); Current Account-GDP Ratio; Real Money (M2)
  - Example chart ranges: -0.6 to 0.6; -1 to 3; -8 to 16
- Figure 3. Credit Booms, Economic Activity and Real Exchange Rate
  - Based on de-trending of sample since 1994; income panels as above
  - Variables: Real GDP; Real Consumption; Real Exchange Rate
  - Example chart ranges: -6 to 6; -3 to 5
- Figure 3 (cont’d). Credit Booms, Capital Inflows, and Current Account
  - Note 1/ Gross capital inflows include portfolio inflows, foreign direct investment, and other investments.
  - Variables: Other Investments to GDP; Capital Inflows to GDP; Current Account-GDP Ratio
  - Example chart ranges: -3 to 4; -6 to 6
- Table 6. Capital Inflows Surges before the Peak of Credit Boom Episodes (title only; numeric details in Table 7 and related tables)
- Table 7. Credit Booms and Exchange Rate Regimes
  - Using credit booms episodes identified when the de-trending is done for the full sample
  - Capital inflows surge (above median) counts and percentages (columns grouped by income):
    - Foreign direct investment, portfolio inflows and other investments: 93 91 74 0 15 47 (table formatting indicates counts for groups)
    - Portfolio inflow and other investments: 73 01 63 81 44 4
    - Other investments: 73 01 43 31 44 4
    - Observations: 234 232
  - Using credit booms episodes identified when de-trending is done for the sample since 1994:
    - Foreign direct investment, portfolio inflows and other investments: 53 81 04 21 36 2
    - Portfolio inflow and other investments: 32 39 38 12 57
    - Other investments: 32 38 33 10 48
    - Observations: 132 421
  - Exchange rate regime association (credit booms identified when de-trending full sample (1960-2011)) — counts and percentages (percentage with respect to number of credit boom episodes in each income group):
    - Fixed and Managed (2): 45.83 61.90 72.73 61.62
    - Dirty Floating (3): 12.50 16.67 9.09 13.13
    - Floating (4): 8.33 0.00 3.03 3.03
    - Other (5): 8.33 2.38 9.09 6.06
    - Mixed (6): 25.00 19.05 6.06 16.16
  - Credit booms identified when de-trending the sample since 1994 — counts and percentages:
    - Fixed and Managed (2): 61.54 62.50 66.67 63.79
    - Dirty Floating (3): 7.69 16.67 9.52 12.07
    - Floating (4): 7.69 0.00 4.76 3.45
    - Other (5): 0.00 0.00 4.76 1.72
    - Mixed (6): 23.08 20.83 14.29 18.97
  - Notes:
    - 1/ Association between peak of credit booms and exchange rate regimes in place the preceding 3 years.
    - 2/ Fixed and managed includes regimes from Reinhart-Rogoff (2010) classification: no separate legal tender, pre-announced peg or currency board arrangement, pre-announced horizontal band that is narrower than or equal to +/-2%, de facto peg, pre-announced crawling peg, pre-announced crawling band that is narrower than or equal to +/-2%, de facto crawling peg, and de facto crawling band that is narrower than or equal to +/-2%.
    - 3/ Dirty floating includes regimes from Reinhart-Rogoff (2010) classification: pre-announced crawling band that is wider than or equal to +/-2%, de facto crawling band that is narrower than or equal to +/-2%, Moving band that is narrower than or equal to +/-2% (i.e., allows for both appreciation and depreciation over time), managed floating.
    - 4/ Floating includes regime from Reinhart-Rogoff (2010) classification: Freely floating.
    - 5/ Other includes regimes from Reinhart-Rogoff (2010) classification: Freely falling, dual market in which parallel market data is missing.
    - 6/ Mixed includes cases where there was more than one exchange rate regime in the preceding 3 years of the credit boom.
- Table 9. Credit Booms and Capital Inflows
  - Dependent variable: Binary variable that takes the value of 1 when there is a credit boom
  - Column headers [1][2][3][4][5][6] with coefficient estimates and standard errors below:
    - Capital inflows (t-1): -0.7590 -0.6019 -0.6079 -1.1350 -0.9061 -1.0016
      - Standard errors: 0.6622 0.7003 0.6681 0.6806 0.7049 0.6872
    - Capital inflows (t-2): 0.8814** 0.9377** 0.8884** 0.7429* 0.7864* 0.7586*
      - Standard errors: 0.4468 0.4509 0.4433 0.4517 0.4549 0.4504
    - Private sector credit/GDP (t-1): 1.4324*** 1.4574*** 1.4680*** 1.5375*** 1.4813*** 1.5864***
      - Standard errors: 0.2333 0.2402 0.2417 0.2460 0.2451 0.2570
    - GDP growth rate (t-1): 1.9954** 1.9262** 2.0464** 1.8303** 1.7828** 1.9489**
      - Standard errors: 0.8073 0.8304 0.8205 0.8437 0.8657 0.8570
    - Growth rate of private sector credit-to-GDP (t-1): 1.4998*** 1.5038*** 1.4981*** 1.7591*** 1.7747*** 1.7771***
      - Standard errors: 0.2230 0.2318 0.2285 0.2451 0.2550 0.2517
    - Trade openness (t-1): -0.4211** -0.5974*** -0.4519** -0.3655** -0.5044** -0.4046**
      - Standard errors: 0.1776 0.1887 0.1788 0.1848 0.1942 0.1867
    - Growth rate of external demand (t-1): 2.2453 2.3213 1.3859 1.9454 1.2696 1.0454
      - Standard errors: 2.3041 2.4106 2.3453 2.4017 2.4937 2.4512
    - Cyclical component of U.S. federal funds rate (t-3): -0.0513** -0.0523** -0.0585*** -0.0588*** -0.0631*** -0.0660***
      - Standard errors: 0.0218 0.0230 0.0223 0.0227 0.0238 0.0233
    - Exchange rate flexibility (t-1): 0.0015 0.0056 -0.0008 -0.0048 0.0028 -0.0073
      - Standard errors: 0.0163 0.0171 0.0163 0.0174 0.0182 0.0175
    - Exchange rate flexibility (t-2): -0.0061 -0.0060 -0.0079 -0.0039 -0.0006 -0.0058
      - Standard errors: 0.0160 0.0168 0.0160 0.0172 0.0181 0.0173
    - Growth rate of terms of trade (t-1): 0.2210 0.3507
      - Standard errors: 0.2660 0.2931
    - Inflation (t-1): -0.1148 -0.1180
      - Standard errors: 0.0734 0.0741
  - Number of observations by column: 2448 2294 2408 2246 2114 2224
  - Number of countries by column: 120 112 119 118 111 118
  - Note: Credit boom episodes are identified using the Mendoza and Terrones methodology (2012). Panel Probits estimated with random-effects. Numbers below the coefficient estimates are standard errors. ***,**,* indicates significance at the 1, 5, or 10 percent level.
  - Measure of capital inflows: Gross other investments-to-GDP; Measure of capital inflows: (Gross other investments plus portfolio investment)-to-GDP
- Figure 5. Credit Booms and Banking Crises (cross-country means and medians of cyclical components)
  - Panels: Booms not associated with banking crisis; Booms associated with banking crises
  - Variables shown: Real Credit Per-Capita; Real GDP; Real Consumption; Real Investment; Real Money; Current Account-GDP Ratio; Real Exchange Rate; Other Investments to GDP
  - Example chart ranges: Real Credit Per-Capita 0 to 40; Real GDP -4 to 6; Real Investment -20 to 30
- Table 9. Benchmarking Exercise: Financial Possibility Frontier
  - Regression results [1][2] (pooled OLS):
    - Log GDP per capita: 0.4473*** 0.4623*** (std. errors 0.0650 0.0740)
    - Log GDP per capita^2: 0.0181*** 0.0188*** (std. errors 0.0028 0.0032)
    - Log population: 0.0961*** 0.0958*** (std. errors 0.0088 0.0097)
    - Log population density: 0.0147 0.0407*** (std. errors 0.0112 0.0128)
    - Log age dependence ratio: -1.3106*** -1.3548*** (std. errors 0.0821 0.0868)
    - Dummy offshore financial center: 0.5488*** 0.6163*** (std. errors 0.0481 0.0522)
    - Dummy fuel exporters: 0.3653*** 0.5233*** (std. errors 0.0657 0.0737)
    - Dummy transition countries: -0.5979*** -0.6516*** (std. errors 0.0588 0.0599)
    - Inflation: -0.0002*** (std. error 0.0001) [appears in column 1 or 2 depending on specification]
    - Exchange rate regime indicator: 0.0083 (std. error 0.0053)
    - GDP growth: -0.0166*** (std. error 0.0037)
    - Time dummies: YES YES
    - Number of observations: 3864 3365
    - R_Squared: 0.79 0.80
  - Note: Regression estimated with pooled OLS. Numbers below the coefficient estimates are standard errors. ***,**,* indicates significance at the 1, 5, or 10 percent level.

*Source: _wp1511 - REFERENCES (PDF)._

### Appendix I. Sample of Countries

### Appendix I. Sample of Countries

### Low-income
- Bangladesh
- Benin
- Burkina Faso
- Burundi
- C.A.R.
- Cambodia
- Chad
- Comoros
- Eritrea
- Ethiopia
- Gambia, The
- Guinea
- Guinea-Bissau
- Haiti
- Kenya
- Kyrgyz Republic
- Liberia
- Madagascar
- Malawi
- Mali
- Mauritania
- Mozambique
- Myanmar
- Nepal
- Niger
- Rwanda
- Sierra Leone
- Tajikistan
- Tanzania
- Togo
- Uganda
- Zimbabwe
- São Tomé and Príncipe
- Senegal
- Solomon Islands
- Sudan
- Swaziland
- Syria
- Vanuatu
- Vietnam
- Yemen
- Zambia

### Lower-middle-income
- Albania
- Armenia
- Belize
- Bhutan
- Bolivia
- Cameroon
- Cape Verde
- Côte d’Ivoire
- Djibouti
- Egypt
- El Salvador
- Fiji
- Georgia
- Ghana
- Guatemala
- Guyana
- Honduras
- India
- Indonesia
- Laos
- Lesotho
- Moldova
- Mongolia
- Morocco
- Nicaragua
- Nigeria
- Pakistan
- Papua New Guinea
- Paraguay
- Philippines
- Samoa
- Senegal
- Solomon Islands
- Sri Lanka
- St. Lucia
- St. Vincent & Grenadines
- Suriname
- Thailand
- Tonga
- Tunisia
- Ukraine
- Uruguay
- Vanuatu
- Vietnam

### Upper-middle-income
- Algeria
- Angola
- Antigua & Barbuda
- Argentina
- Azerbaijan
- Belarus
- Belize
- Botswana
- Brazil
- Bulgaria
- Chile
- China
- Colombia
- Costa Rica
- Dominica
- Dominican Rep.
- Ecuador
- Gabon
- Grenada
- Jamaica
- Jordan
- Kazakhstan
- Latvia
- Libya
- Lithuania
- Macedonia, FYR
- Malaysia
- Maldives
- Mauritius
- Mexico
- Namibia
- Peru
- Romania
- Russia
- Seychelles
- South Africa
- St. Lucia
- St. Vincent & Grenadines
- Suriname
- Thailand
- Turkey
- Uruguay
- Venezuela

### High-income
- Bahrain
- Barbados
- Brunei Darussalam
- Croatia
- Equatorial Guinea
- Hong Kong
- Hungary
- Kuwait
- Oman
- Poland
- Qatar
- Saudi Arabia
- St. Kitts and Nevis
- Trinidad and Tobago
- U.A.E.

### Source: World Bank Classification
- 1 Low-income economies (GNI per capita of $1,025 or less).
- 2 Lower-middle-income economies (GNI per capita of $1,026 to $4,035).
- 3 Upper-middle-income economies (GNI per capita of $4,036 to $12,475).
- 4 High-income economies (GNI per capita of $12,475 or more).

*Source: World Bank Classification*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp1511.pdf_