## 1pryea2019002

## Source details

**Canonical URL:** [1pryea2019002](https://www.imf.org/-/media/files/publications/cr/2019/1pryea2019002.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2019/1pryea2019002.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2019/1pryea2019002.pdf.json)

---

### What caused the growth acceleration?
- GDP per capita growth:
  - Paraguay: GDP per capita grew at an average annualized rate of 3 percent over the past decade.
  - LAC region comparator: 0.8 percent.
- Drivers of acceleration:
  - Bounce back from late-1990s crisis and previously sub-par growth.
  - Benevolent external environment, notably the commodity price boom.
  - Improved macroeconomic stability.
- Composition of growth:
  - Largely extensive: capital deepening and increasing labor inputs.
  - TFP growth gained importance in the most recent years.

### Historical context and real convergence
- Historical swings:
  - 1970s boom linked to Itaipu dam construction and improved connectivity with Brazil.
  - Post-Itaipu slowdown in the 1980s and 1990s; commodity exports once constituted around 70 percent of total exports.
- Since 2003: growth averaged over 4 percent annually; steady growth despite weak Brazil and Argentina performance since 2010.
- Convergence:
  - GDP per capita catching up to the US level by 0.4 percentage point annually since 2011.
  - Greater convergence in PPP terms during the 2000s associated with exchange rate appreciation.

### Improved macro stability that facilitated growth
- Banking crisis (late 1990s):
  - Crisis cost over 12 percent of GDP in various rescue packages.
  - Loan delinquency ratio rose to over 20 percent; real GDP per capita dropped by over 10 percent from 1995 to 2000.
- Post-crisis stabilization measures and outcomes:
  - Inflation averaged 6 percent since 2000 (versus 17.4 percent in the two decades before).
  - BCP adopted inflation targeting in 2011.
  - Fiscal deficits of about 8 percent in early 1990s turned into surpluses; government debt reduced from over 65 percent of GDP in 1990 to around 20 percent in 2017.
  - Fiscal responsibility law in effect in 2015.
  - NPL ratio declined from over 20 percent in 2003 to 2 percent in 2008.
  - Reforms: risk-based supervision, strengthened prudential standards, comprehensive legal framework; two IMF-supported programs from 2004 to 2007 helped implementation.

### The agricultural commodity price boom and spillovers
- Soybean-driven gains:
  - Paraguay is the fourth largest exporter of soybeans.
  - Soybean export price in dollars tripled between 2003 and 2013.
  - China imported over 60 percent of soybeans in the world during that period.
  - Soybean export values: $363 million in 2000 to $2.7 billion in 2013.
- Sectoral and economy-wide effects:
  - Sowing areas for major crops doubled between 2003 and 2018.
  - Private sector credit: 13 percent of GDP in 2003 to 40 percent in 2017.
  - Exchange rate appreciation increased wages and dollar purchasing power, boosting domestic demand for construction, commerce, and services.

### Sustainability of the current growth path
- Baseline and composition:
  - Recent resurgence began from a low base; real GDP per capita growth since 1995 modest.
  - GDP per capita (without PPP) at par with the late 1970s as a percentage of US level.
- TFP dynamics:
  - Historically growth driven by capital deepening and labor input; TFP mostly negative in 1980s–1990s.
  - Reversal in the last decade with TFP contributions outstripping capital and labor in recent periods.
  - Long-term TFP level over the past half century largely flat.
  - In agriculture: labor productivity growth mostly from capital deepening; output growth largely from agricultural land expansion; average yield growth for major crops limited.
- External environment risks:
  - Commodity boom largely driven by China; China’s growth moderation has tapered demand increases over the past five years.
  - Long-term trend decline in agricultural commodity prices.
  - Exchange depreciation after price drops would harm non-tradable sector via reduced purchasing power.
  - Growth prospects for Brazil and Argentina are lower than in the previous decade.
  - Interest rate normalization in advanced economies raises cost of capital.
  - Paraguay increasingly decoupled from regional performance.

### Structural gaps relative to high-converging peers
- Investment and saving:
  - Investment rate around 20 percent of GDP in Paraguay (noted as low relative to both peer groups).
  - Capital-output ratio close to average level of high-converging peers.
  - Domestic saving and FDI lag behind high-converging peers.
- Sectoral structure and exports:
  - Paraguay much less industrialized; high share of primary sector.
  - Non-agriculture, non-energy exports: only around 7 percent of total exports (and elsewhere noted as only 3 percent of GDP).
  - Employment shifts: agriculture employment declined; employment transitioned mostly into non-tradeable services; industrial employment share declined since 2007.
- Export sophistication:
  - Low economic complexity due to lack of diversification and agriculture-dominated exports.
- Informality and entrepreneurship:
  - High informality relative to high-converging peers; registered entrepreneurial activity substantially lower.
  - Informality detrimental to government revenue and productivity growth.
- Institutional and infrastructure shortfalls:
  - Lower scores on legal system quality, public observance of law, and corruption.
  - Transport infrastructure and education quality lower than peers.
  - Corporate tax level at par with high-converging peers.

### Maquila industry (FDI-driven manufacturing) — Box 1
- Origins: began ~2007; FDI mainly from Brazilian firms.
- Products: labor-intensive goods (auto parts, textiles, plastics, leather).
- Growth and scale:
  - Sector exports average annual growth over 20 percent.
  - Employment: around 15,000 workers (less than 0.5 percent of total employment).
  - Exports: around 4 percent of total exports.
- Scenarios to 2030:
  - 15 percent annual export growth → industry contributes around 10 percent of total exports by 2030.
  - 25 percent annual export growth → industry contributes around 30 percent of total exports by 2030.
- Risks and constraints:
  - Over 80 percent of Maquila investments come from Brazil; continuation uncertain.
  - Lack of connectivity and weak transport infrastructure.

### Electricity endowment and industrial structure — Box 2
- Paraguay consumes 20 percent of its allotted share of power from Itaipu.
- Contrast with Iceland: domestic electricity consumption over eight times the EU average and industry-intensive electricity use reshaped exports.
- Observation: abundant hydro power has not substantially altered Paraguay’s export basket aside from electricity exports.

### Structural reform priorities — methodology and empirical findings
- Priority dimensions:
  1. Potential impact on convergence.
  2. Potential private sector support (measured via World Bank Enterprise Survey responses).
- Empirical approach:
  - Cross-country growth regression: g_{c,t} = ∆ ln( Y_{c,t} / Y_{US,t} ), includes lagged relative GDP gap and structural indicator (SI).
  - β̂3 measures additional growth-acceleration effect from SI improvement.
  - Estimation via OLS and fixed-effect panel on “all countries” and “peer group” samples; SIs in log form.
- Regression results: selected significant β̂3 estimates (with standard errors)
  - Burdens of customs procedure: β̂3 = 0.0379* (0.0225); observations 1,385; R-squared 0.026.
  - Quality of education: β̂3 = 0.0449** (0.0204); observations 1,385; R-squared 0.029.
  - Government effectiveness: β̂3 = 0.0103* (0.00529); observations 3,275; R-squared 0.017.
  - Law and order: β̂3 = 0.0362*** (0.00823); observations 4,120; R-squared 0.009.
  - New business density: β̂3 = 0.00284* (0.00168); observations 256; R-squared 0.334.
  - Quality of transport infrastructure: β̂3 = 0.0607*** (0.0219); observations 993; R-squared 0.021.

### Measuring private sector support for reforms
- Data: World Bank Enterprise Survey firm-level survey for Paraguay (2017).
- Proxy: percentage of firms answering “yes” to “Is factor X a major constraint on business activities?” used as proxy for potential popularity.
- For export-oriented factors, the higher percentage between exporting-firm sample and whole sample is used.

### Priority quadrants and top-ranked reforms
- Highest-priority (upper-right quadrant): transportation infrastructure; quality of education; rule of law.
- Rationale: these have large regression coefficients, substantial empirical gaps with high-converging peers, and are identified by the largest percentage of firms as major constraints.
- Other priorities:
  - Customs and trade regulations: second-priority (upper-left quadrant); close to 30 percent of exporters identify as major constraint.
  - Business facilitation (new formal business density): third-priority (lower-right quadrant); high impact on convergence speed driven by large gap in new business density.

### Quantitative facts and constraints
- Government spending on education: around 5 percent of GDP.
- Non-agriculture, non-energy exports: only 3 percent of GDP.
- Close to 30 percent of private sector exporter firms identify customs and trade regulations as a major constraint.
- Business facilitation impact driven by large gap in new business density between Paraguay and high-converging peers.

### Policy implications and recommended sequencing
- Short-to-medium term priorities:
  - Improve rule of law and customs/trade regulations: lower fiscal cost, concrete implementation paths, relatively faster payoff.
- Medium-to-long term priorities:
  - Improve transport infrastructure and quality of education: require substantial financial resources and persistent strategic plans.
- Specific considerations:
  - Transport infrastructure: high potential impact but high fiscal cost and need for long-term consistent implementation.
  - Quality of education: spending around 5 percent of GDP but unsatisfactory outcomes; reform should diagnose missing links.
  - Rule of law: lower fiscal cost but reforms take time and face cultural/political inertia—ideal for short-to-medium term prioritization.
  - Customs and trade regulation: relatively low cost and concrete roadmaps—suitable for short-term action.
  - Business facilitation: high potential impact but requires further investigation into root causes of low new business creation.

### Credit growth dynamics and banking-sector findings (Appendix)
- Long-run determinants of credit:
  - Four external factors explain 67 percent of variation in guaraní-denominated credit and 64 percent in dollar-denominated credit.
  - Growth of major trading partners explains 54 percent of variation in guaraní-denominated credit.
  - Monetary policy rate is the most important domestic driver for credit growth long-run.
- Short-term responses to external shocks:
  - Positive shock to trading-partner growth:
    - Guaraní credit: raises on impact by 1 percentage point; peaks at 6 percent in the fourth quarter.
    - Dollar credit: raises on impact by 2 percentage point; peaks at 6 percent in the third quarter.
  - Positive shock to Latin America EMBI spread:
    - Dollar credit: drops on impact by 1 percentage point; peaks at -4 percent in the second quarter.
    - Guaraní credit: impulse response not statistically significant at 10 percent.
  - Positive shock to soy bean price:
    - Dollar credit: increases by 2 percentage point in the first quarter; to 4 percentage point in the fifth quarter.
    - Guaraní credit: no statistically significant short-term impact.
  - Positive shock to Fed Funds Rate: no statistically significant impact on either credit series.
- Short-term responses to domestic shocks:
  - Monetary policy tightening:
    - Guaraní credit: falls 1 percentage point in first quarter; widens to 3 percentage point in third quarter.
    - Dollar credit: increases by 2 percentage point on impact; dissipates quickly.
  - Real effective exchange rate appreciation:
    - Dollar credit: raises growth by 1.5 percentage point on impact; effect dissipates after several quarters.
    - Guaraní credit: response not statistically significant.
  - Positive shock to GDP growth: no statistically significant impact on either credit series.
- Drivers of rapid credit growth in last decade:
  - Favorable external environment: strong trading-partner growth and high agricultural prices; partially offset by higher Latin America EMBI spreads.
- Bank-level credit cycles:
  - Two credit booms:
    - 2010Q3–2011Q3 peak: credit growth 60 percent y/y.
    - 2014Q4–2015Q4 peak: credit growth 67 percent y/y.
  - Booms shorter than cross-country average: about one year vs. average three years.
  - During boom years, credit-to-GDP grew on average by 15 percent per year.
  - Distressed loans rose during boom endings; second boom followed by a sharper credit decline.
- Funding and sector concentration:
  - Banks relied more on non-core funding and non-resident funding during booms, stronger in the second boom.
  - Domestic banks increased credit to agriculture by close to 165 percent over 2011–2015; soy price increased by 150 percent over same period.
  - Credit to soy explains about 49 percent of total agriculture sector credit on average 2008–2016.
- Heterogeneous bank behavior:
  - Domestic banks repaired balance sheets more after soy price collapse and reduced soy exposure more than non-domestic banks.
  - Domestic banks rely less on core funding and more on non-core funding (e.g., securities) after shocks.
  - Domestic banks respond more strongly to regional EMBI spreads than foreign banks.
  - Among foreign banks: Argentinian banks tended to shrink Paraguayan balance sheets under stress; Brazilian banks increased assets and loans in Paraguay when EMBI spreads increased in Brazil.
  - End-2017 market shares: Brazilian banks around 19 percent of total assets; Argentinian banks around 11 percent; domestic banks close to 50 percent.
- Monetary and macroprudential policy response:
  - Monetary policy rate not actively used to target credit cycle because inflation remained low.
  - BCP measures in 2010: leverage limits and tightened FX sales.
  - December 2015: BCP introduced transitory measures that effectively relaxed NPL definition for agriculture and livestock (Resolution No. 33 and Resolution No.34).
  - Macroprudential toolkit by end-2016: CCB (), LVR (), Capital requirements (), LFC (), RR (), LFX (); other items listed with limited availability.

### Conclusions (summary)
- External factors play a dominant role in long-run credit dynamics; growth of major trading partners is the single most significant driver for guaraní-denominated credit.
- Short-run: both guaraní- and dollar-denominated credit respond to trading-partner growth; regional risk aversion (EMBI spreads) lowers dollar credit but not guaraní credit; soy price improvements spur dollar credit growth.
- Banking-sector resilience improved since the 1990s crisis due to supervisory modernization, higher capital, and better provisioning; fall in soybean price in 2015 triggered a credit decline but did not cause a banking crisis.
- For sustained convergence and higher labor productivity, Paraguay needs higher TFP growth, further capital deepening, industrial diversification, and structural reforms prioritized on transport infrastructure; quality of education; rule of law; customs and trade regulations.

*Source: International Monetary Fund (content unit 1pryea2019002 — Paraguay report excerpts).*

### References  ________________________________________________________________________________ 12

### WHAT CAUSED THE GROWTH ACCELERATION?

### A. Introduction
- Paraguay experienced a growth acceleration over the past decade:
  - GDP per capita grew at an average annualized rate of 3 percent, compared to 0.8 percent for the LAC region.
  - In PPP-terms, growth was even more rapid.
- The acceleration over the past 15 years reflects a combination of factors:
  - a bounce back from the crisis in the late 1990s and sub-par growth of the two decades prior;
  - a benevolent external environment, the commodity price boom in particular;
  - improved macroeconomic stability.
- Composition of growth:
  - Largely extensive: mostly from capital deepening and increasing labor inputs.
  - TFP growth has played a bigger role in the most recent years.

### B. Why Has Growth Picked Up?
- Historical swings and turning points:
  - 1970s boom linked to Itaipu dam construction and improved connectivity with Brazil; soybean and cotton price surge aided agriculture.
  - Growth slowed in the 1980s and 1990s after Itaipu completion; commodity exports had constituted around 70 percent of total exports.
- Since 2003:
  - Growth averaged over 4 percent annually.
  - Paraguay maintained steady growth despite weak performance in Brazil and Argentina since 2010.
- Real convergence:
  - GDP per capita has been catching up to the US level by 0.4 percentage point annually since 2011.
  - Convergence in PPP terms was even greater during the 2000s due to exchange rate appreciation associated with the agricultural commodity price boom.

### C. Improved Macro Stability Facilitated Growth
- Banking crisis in the late 1990s:
  - Crisis resulted from rapid financial liberalization, lax entry requirements, and weak banking supervision.
  - The crisis cost over 12 percent of GDP in various rescue packages.
  - Loan delinquency ratio increased to over 20 percent; real GDP per capita dropped by over 10 percent from 1995 to 2000.
- Post-crisis stabilization:
  - IMF-supported programs aided restoration of macroeconomic stability.
  - Monetary policy: BCP implemented restrictive policy and open market operations to absorb liquidity.
  - Inflation outcomes:
    - Inflation has averaged 6 percent since 2000, compared to 17.4 percent for the two decades before.
    - The BCP adopted an inflation targeting framework in 2011, further reducing inflation and price volatility.
    - Exchange rate volatility has also come down (defined as the rolling standard deviation over five-year windows).
  - Fiscal outcomes:
    - Fiscal deficits that had amounted to 8 percent in the early 1990s turned into fiscal surpluses.
    - Government debt reduced from over 65 percent of GDP in 1990 to around 20 percent in 2017.
    - A fiscal responsibility law in effect in 2015 prescribed definitive rules for government spending and deficit.
  - Banking sector recovery:
    - NPL ratio came down from over 20 percent in 2003 to 2 percent in 2008 after bank closures/downsizing and strengthened supervision.
    - Reforms included a move towards risk-based supervision, strengthened prudential standards, and a more comprehensive legal framework.
    - Two IMF-supported programs from 2004 to 2007 facilitated implementation of new banking legislation and prudential regulations.
- Result:
  - Improved macroeconomic management helped Paraguay avoid boom-bust cycles and decouple its performance from the rest of the region.

### D. The Agricultural Commodity Price Boom
- Soybean-driven export gains:
  - Paraguay is the fourth largest exporter of soybeans; soybean export price in dollars tripled between 2003 and 2013.
  - China imported over 60 percent of soybeans in the world during that period.
  - Soybean export values grew from $363 million in 2000 to $2.7 billion in 2013.
- Sectoral and economy-wide effects:
  - Sowing areas for major crops doubled between 2003 and 2018.
  - Exchange rate appreciation increased wages and purchasing power in dollar terms, boosting domestic demand for tertiary sectors (construction, commerce, services).
  - Private sector credit expanded from 13 percent of GDP in 2003 to 40 percent in 2017, with credits to agriculture and services as main components.
  - The commodity boom had positive spillovers to non-tradable sectors via increased domestic demand.

### E. How Sustainable is the Current Growth Path?
- Growth baseline and composition:
  - Recent resurgence began from a low base; real GDP per capita growth since 1995 has been modest.
  - GDP per capita as a percentage of the US level (without PPP adjustment) is at par with the late 1970s.
- TFP dynamics:
  - Growth accounting indicates growth historically driven by capital deepening and labor input growth.
  - Throughout the 1980s and 1990s, TFP growth was mostly negative; reversal in the last decade with TFP contributions outstripping capital and labor in recent periods.
  - Long-term TFP level over the past half century has been largely flat.
  - In agriculture, labor productivity growth driven mostly by capital deepening; output growth largely from expansion of agricultural land; average yield growth for major crops limited.
- External environment risks:
  - The agricultural commodity price boom was largely driven by demand from China; China’s growth moderation implies that demand increases have tapered off in the past five years.
  - Over the medium term, growth benefit from external terms of trade is likely limited.
  - Long-term data show a general trend decline in agricultural commodity prices.
  - Exchange depreciation following commodity price drops would negatively impact non-tradable sector growth through reduced purchasing power.
  - Growth prospects in main trading and investment partners (e.g., Brazil, Argentina) are lower than in the previous decade.
  - Interest rate normalization in advanced economies is raising the cost of capital and making investments more expensive.
  - However, Paraguay’s economy is increasingly decoupled from the region, reducing the trade-partner impact compared to the past.

### F. Conclusion and Implications for Sustainable Growth
- Long-term convergence challenges:
  - Despite strong recent growth, over a longer horizon Paraguay has not attained significant convergence with advanced economies.
  - Historically, growth has been driven by increased labor and capital inputs; TFP contribution has increased only recently.
- Requirements for sustainable convergence:
  - Long-run convergence will need higher labor productivity growth, requiring higher TFP growth and further capital deepening.
  - Empirical evidence shows a strong linkage between GDP per capita and composite structural indicators such as the World Competitiveness Index, where Paraguay ranks poorly.
  - Identifying and correcting structural deficiencies that hamper productivity growth and capital accumulation is crucial for sustainable growth.

*Source: International Monetary Fund (Paraguay report excerpt).*

### 11.      It’s interesting to note the diverging trend in macro stability between the two peer

### 1pryea2019002 - 11.      It’s interesting to note the diverging trend in macro stability between the two peer

### Diverging macro stability across peer groups
- Macro stability indicators for the high-converging country group do not seem to be significantly better than those of the low-convergence group during the starting years of their high-convergence episodes.
- Over subsequent periods:
  - Government debt to GDP ratio drops by an average of 2.4 percentage points over a 10-year high-converging period (high-converging group).
  - Overall fiscal balance deteriorates by 2.3 percentage points over 10 years for the low-converging group.

### Investment, capital, and FDI in Paraguay
- Investment rate:
  - Around 20 percent of GDP in Paraguay (noted as low relative to both high-converging and low-converging peers).
- Capital-output ratio:
  - Close to the average level of high-converging peers.
  - High capital-output ratio combined with low investment suggests marginal return on investment may be low due to subdued productivity growth; consistent with historical TFP trajectory in Paraguay.
- Domestic saving and FDI:
  - Domestic saving rate and foreign direct investment (FDI) lag behind high-converging peers.
- Implication:
  - Diversification into a more productive industrial sector is needed to increase investment despite already high capital stock.

### Industrialization, export composition, and labor reallocation
- Sectoral structure:
  - Paraguay is much less industrialized compared to high-converging countries; structure resembles low-converging peers with a high share of the primary sector.
- Export composition:
  - Agricultural commodities dominate the export basket; low share of industrial exports.
  - Share of non-agriculture, non-energy export in total exports is only around 7 percent.
- Labor shifts:
  - Labor share of agriculture declined significantly due to increased mechanization/labor substitution.
  - Employment mostly transitioned into the non-tradeable service sector.
  - Share of employment in the industrial sector declined since 2007.
- Productivity implication:
  - Non-tradeable sector generally less productive; to boost overall labor productivity, higher industrialization and expansion of the tradeable sector are needed.
  - Development of emerging export industries with higher TFP will also help increase investment level.

### Export sophistication and diversification
- Paraguay significantly lags behind high-converging peers in economic complexity of exports due to lack of export diversification and agricultural-dominated exports (lower technological value-added).
- Going forward, development of new industries and export categories leveraging Paraguay’s comparative advantages is crucial for sustainable growth and convergence.

### Box 1 — Maquila Industry in Paraguay (FDI‑driven manufacturing)
- Origins and drivers:
  - Maquila industry started around 2007; FDI mainly from Brazilian companies seeking lower labor and electricity costs and lower tax rates.
  - Exports primarily labor-intensive goods: auto parts, textiles and apparels, plastics, leather.
- Growth and scale:
  - Average annual growth of the sector’s exports has been over 20 percent.
  - Current employment: around 15,000 workers (less than 0.5 percent of total employment).
  - Current exports: around 4 percent of total exports.
- Hypothetical scenarios for industry export growth to 2030:
  - If industry’s exports grow at 15 percent annual, the industry would contribute to around 10 percent of total exports by 2030.
  - If industry’s exports grow at 25 percent annual, the industry would contribute to around 30 percent of total exports by 2030.
- Risks:
  - Over 80 percent of investments in the Maquila industry come from Brazil; uncertain whether Brazil-sourced FDIs will continue to grow given Brazilian growth outlook.
- Constraints:
  - Lack of connectivity and weak transport infrastructure hinder leveraging of comparative advantages.
  - Need to improve infrastructure and institutions to widen international appeal and encourage development of new industries.

### Box 2 — Electricity cost and industrial structure: Paraguay vs Iceland
- Comparative advantage:
  - Paraguay endowed with abundant hydro power; low-cost electricity is a natural comparative advantage.
- Current utilization:
  - Paraguay currently consumes 20 percent of its allotted share of power generated from Itaipu.
  - Construction of new hydroelectric facilities has not substantially changed Paraguay’s export basket aside from electricity exports.
- Iceland contrast:
  - Iceland’s domestic electricity consumption is over eight times higher than the EU average, largely by industry (notably aluminum).
  - Iceland’s export structure evolved from fish/agricultural products to electricity-intensive outputs and ICT services, transport, tourism — illustrating how power endowment can shape industrialization.

### Institutional quality, human capital, and infrastructure
- Institutional indicators (International Country Risk Guide):
  - Paraguay scores lower on the quality of legal system, public observance of law, and corruption relative to both peer groups.
- Other business environment aspects:
  - Corporate tax level: at par with high-converging peers (and much lower than in the low-convergence group).
  - Qualities of transport infrastructure and education: lower than peers.
- Firm-level survey consistency:
  - Transportation infrastructure, labor force, and public institutions are reported among the biggest deterring factors to business operation.

### Informality and entrepreneurial activity
- Private sector composition:
  - High percentage of informality relative to high-converging peers.
  - Registered entrepreneurial activity substantially lower compared to both peer groups.
- Consequences of informality:
  - Detrimental to government revenue.
  - Detrimental to productivity growth: informal businesses tend to be smaller, less productive, and grow slower.

### Cross-country comparison summary
- High-converging peers share characteristics:
  - Higher investment and FDI rates.
  - More industrialized sectoral structure with higher product sophistication.
  - Higher-quality institutions, infrastructure, education, and a more entrepreneurial business sector.
- Paraguay lags behind high-converging peers on most parameters examined, except macroeconomic stability.
- Conclusion: Structural reforms on many levels are needed for Paraguay to join high-converging countries.

### Structural reform priorities: methodology
- Priority evaluation dimensions (inspired by Biljanovska & Sandri (2018)):
  1. Potential impact of the reform on convergence.
  2. Potential support of the reform among private sector agents (private sector “popularity” measured via firm survey responses).
- Interpretation of priority quadrants:
  - Upper right quadrant: highest priority (high growth benefit and likely popular).
  - Upper left: second-highest priority (popular and some growth benefit).
  - Lower left: lower priority relative to others but still empirically strongly associated with growth and important.
- Empirical approach:
  - Cross-country growth regression where dependent variable g_{c,t} is annual growth of the gap in real GDP between country c and the United States (g_{c,t} = ∆ ln( Y_{c,t} / Y_{US,t} )).
  - Regression includes lagged relative GDP gap and structural indicator (SI).
  - Classical convergence predicts negative coefficient on lagged relative GDP per capita.
  - Coefficient β_3 (parameter of interest) measures additional growth-acceleration effect from improvement in SI.
- Estimation:
  - OLS and fixed-effect panel methods used on two samples: “all countries” and “peer group” (countries at a similar convergence stage as Paraguay).
  - Structural indicators are in log form.

### Selected regression estimates and statistics (reported SI variables with positive and significant β̂3)
- Lagged relative GDP per capita coefficients (with standard errors in parentheses):
  - Burdens of customs procedure: -0.0156*** (0.00298)
  - Quality of education: -0.0163*** (0.00288)
  - Government effectiveness: -0.0149*** (0.00317)
  - Law and order: -0.00972*** (0.00210)
  - New business density: -0.305*** (0.0312)
  - Quality of transport infrastructure: -0.0162*** (0.00411)
- Structural factor coefficients (β̂3) (with standard errors):
  - Burdens of customs procedure: 0.0379* (0.0225)
  - Quality of education: 0.0449** (0.0204)
  - Government effectiveness: 0.0103* (0.00529)
  - Law and order: 0.0362*** (0.00823)
  - New business density: 0.00284* (0.00168)
  - Quality of transport infrastructure: 0.0607*** (0.0219)
- Constants (with standard errors):
  - Burdens of customs procedure: -0.0828** (0.0372)
  - Quality of education: -0.0902*** (0.0322)
  - Government effectiveness: -0.0156* (0.00855)
  - Law and order: -0.0593*** (0.0142)
  - New business density: -0.798*** (0.0850)
  - Quality of transport infrastructure: -0.126*** (0.0373)
- Observations and R-squared:
  - Burdens of customs procedure: Observations 1,385; R-squared 0.026
  - Quality of education: Observations 1,385; R-squared 0.029
  - Government effectiveness: Observations 3,275; R-squared 0.017
  - Law and order: Observations 4,120; R-squared 0.009
  - New business density: Observations 256; R-squared 0.334
  - Quality of transport infrastructure: Observations 993; R-squared 0.021

### Measuring private sector support for reforms
- Data source:
  - World Bank Enterprise Survey firm-level survey for Paraguay (2017).
- Proxy for potential popularity:
  - Percentage of firms answering “Is factor X a major constraint on business activities?” with “yes” is used as proxy for potential private sector support for reform on factor X.
- Sample handling:
  - Given importance of export orientation, the higher percentage between exporting-firm sample and whole sample is used for the priority ranking.
- Note on interpretation:
  - Survey question captures both business resonance with reform areas and perceived impact of reforms on private business growth; treated here solely as estimate of potential popularity.

### Variable matching between regression indicators and firm survey
- Customs and trade regulation:
  - Regression variable: Burdens of customs procedure (GCP)
  - Survey variable: Customs and trade regulation
- Quality of education:
  - Regression variable: Quality of education (GCR)
  - Survey variable: Inadequately educated workforce
- Government effectiveness:
  - Regression variable: Government effectiveness (WGI)
  - Survey variable: Average of tax administration & labor regulations
- Rule of law:
  - Regression variable: Law and order (CRG)
  - Survey variable: Average of corruption, courts system, crime & theft
- Business facilitation:
  - Regression variable: New business density (WB)
  - Survey variable: Business licensing and permits
- Transportation infrastructure:
  - Regression variable: Quality of transport infrastructure (GCR)
  - Survey variable: Transportation
- Caveats:
  - Matching is imperfect; some regression variables (e.g., government effectiveness) cover a broader concept than the corresponding firm-survey items.

*International Monetary Fund — Content unit 1pryea2019002 (excerpts presented).*

### 35.       The chart below presents the results for structural reform priority quadrants,

### 1pryea2019002 - 35.       The chart below presents the results for structural reform priority quadrants,

### Structural reform priority quadrants: top-ranked factors
- Highest-priority (upper-right) quadrant: transportation infrastructure; quality of education; rule of law.
- Rationale:
  - These three areas have large coefficients in the cross-country convergence regression and substantial empirical gaps between Paraguay and the high-converging peers.
  - They are also identified by the largest percentage of firms as major business constraints.

### Key empirical findings and quantitative facts
- Government spending on education in Paraguay: around 5 percent of GDP.
- Non-agricultural, non-energy exports: only 3 percent of GDP.
- Close to 30 percent of private sector firms that are exporters identify customs and trade regulations as a major constraint.
- Business facilitation (new formal business density) shows a very high impact on convergence speed, driven mostly by the large gap in new business density between Paraguay and high-converging peers.
- Many structural regressions show significant roles for lagged relative GDP and specific structural indicators (see regression tables for detailed coefficient estimates and sample sizes).

### Policy implications and recommended prioritization
- Transport infrastructure
  - Impact and constraint: high potential impact on growth but requires substantial financial resources.
  - Implementation requirement: consistent implementation of strategic focus over an extended period of time given limited fiscal room.
- Quality of education
  - Fiscal context: government spending on education in Paraguay is around 5 percent of GDP and is higher than the averages in both its high-converging and low-converging peers.
  - Policy emphasis: given high investment but unsatisfactory results, the government should carefully examine where the missing links are to devise an effective reform plan.
- Rule of law (including courts reform, corruption reduction, public security)
  - Fiscal cost: requires less financial commitment than infrastructure or education.
  - Timing and difficulty: reforms can take a long time to take effect because the status quo has deep cultural and political roots.
  - Priority: an ideal area to prioritize in the short to medium term due to lower fiscal burden and potential growth benefits.
- Customs and trade regulations (second-priority, upper-left quadrant)
  - Private sector constraint: close to 30 percent of exporter firms identify this area as a major constraint.
  - Impact: estimated to have a significantly positive—albeit relatively small in magnitude—impact on convergence speed.
  - Cost and implementation: does not cost substantial financial resources; roadmaps and action items tend to be concrete with international benchmarks to borrow from; ideal candidate for short-term prioritization.
- Business facilitation (third-priority, lower-right quadrant)
  - Impact: shown to have a very high impact on convergence speed, primarily due to the gap in new (formal) business density.
  - Private sector proxy: percentage of businesses identifying business licensing and permits as a major constraint (noted as an imperfect proxy).
  - Knowledge gap: existing data do not clarify why new business creation is so low in Paraguay or what specific government actions would raise it; this requires further investigation and is beyond the scope of the current paper.

### Implementation sequencing and trade-offs
- Short-to-medium term priorities:
  - Improving rule of law and quality of customs and trade regulations: lower fiscal cost, more concrete implementation paths, and relatively faster payoff.
- Medium-to-long term priorities:
  - Improving transport infrastructure and quality of education: paramount for sustainable convergence but require more financial investments and persistent, gradual, focused strategic plans over extended periods.

### Summary conclusions (macro and structural context)
- Over the past 15 years Paraguay has experienced relatively stable and rapid growth; especially in the past 5 years its performance stands out as one of the fastest growing economies in South America.
- External demand is the main driver of growth for Paraguay as a small open economy with limited domestic market size.
- After the 2000s commodity boom, agricultural commodity prices have declined and may decline further; this affects both the export sector and the non-tradable sector through income effects from exchange rate depreciation.
- Development of new industries and faster growth of the non-agriculture, non-energy tradable sector are crucial for long-run growth and convergence; sustainable growth calls for increases in TFP growth and higher private-sector driven investments.
- Compared to high-converging peers, Paraguay:
  - Fares well on macro stability indicators (inflation, government balance, debt level).
  - Lags in structural measures: rates of investment and FDI, degree of industrialization, level of economic complexity, quality of institutions, infrastructure, and education.
- Empirical priority list for structural reforms: transport infrastructure; quality of education; rule of law; customs and trade regulations.

*Source: 1pryea2019002 - 35.       The chart below presents the results for structural reform priority quadrants,*

### appendix follows.

### Long-Run Determinants and Credit Growth Dynamics (Appendix)

### Long-run determinants
- The four external factors together explain 67 percent of the variation in the guaraní-denominated credit and 64 percent of the variation in the Dollar-denominated credit.
- The growth of major trading partners accounts for 54 percent of the total variation in the guaraní-denominated credit.
- For dollar-denominated credit, the role of each external factor is more balanced, indicating a less concentrated influence across external drivers.
- Among domestic factors, the monetary policy rate is the most important driver in explaining the dynamics of credit growth in Paraguay in the long-run.

### Short-term dynamics: response to external shocks
- Positive shock to growth of major trading partners:
  - Guaraní-denominated credit: raises on impact by 1 percentage point; impact peaks at 6 percent in the fourth quarter after the shock; increase is immediate and persistent.
  - Dollar-denominated credit: raises on impact by 2 percentage point; impact peaks at 6 percent in the third quarter after the shock; immediate but less persistent than guaraní credit.
- Positive shock to Latin America EMBI spread:
  - Dollar-denominated credit: growth drops on impact by 1 percentage point; impact peaks at -4 percent in the second quarter after the shock; immediate and persistent.
  - Guaraní-denominated credit: impulse response is not statistically significant at the 10 percent confidence level (model cannot reject null of no impact).
- Positive shock to soy bean price:
  - Dollar-denominated credit: increases by 2 percentage point in the first quarter after the shock; continues to increase to 4 percentage point in the fifth quarter after the shock.
  - Guaraní-denominated credit: no statistically significant impact in the short-term (at the 10 percent confidence level).
- Positive shock to the Fed Funds Rate: no statistically significant impact on guaraní-denominated credit growth or dollar-denominated credit growth.

### Short-term dynamics: response to domestic shocks
- Positive shock to the domestic monetary policy rate (tightening):
  - Guaraní-denominated credit: falls 1 percentage point in the first quarter after the shock; drop widens to 3 percentage point in the third quarter after the shock and then starts to dissipate.
  - Dollar-denominated credit: increases by 2 percentage point on impact; impact dissipates very quickly.
- Shock to the real effective exchange rate (appreciation):
  - Dollar-denominated credit: raises growth by 1.5 percentage point on impact and remains around this level for a few quarters; impact starts to dissipate in the seventh quarter after the shock.
  - Guaraní-denominated credit: response not statistically significant.
- Positive shock to GDP growth: no statistically significant impact on guaraní-denominated credit growth or dollar-denominated credit growth.

### What drove rapid credit growth in the last decade (aggregate findings)
- Rapid credit growth was largely due to a favorable external environment:
  - Strong growth of major trading partners.
  - High prices of agricultural products (which helped support growth of dollar-denominated credit).
- These favorable factors were partly offset by the negative impact of increases in the Latin America EMBI spread.

### Bank-level analysis: credit cycles and behavior
- Two credit booms identified (methodology per Box 1):
  - First credit boom peaked in 2010Q3 – 2011Q3 with credit growth of 60 percent, y/y.
  - Second credit boom identified in 2014Q4 – 2015Q4 with credit growth of 67 percent, y/y.
  - Note: These credit booms were typically shorter than traditional booms — about a year compared to the cross-country average of three years.
  - During boom years, credit-to-GDP grew on average by 15 percent per year.
- NPL dynamics:
  - Distressed loans increased between late 2010 and 2012, and between 2014 and 2016.
  - Both booms saw rising NPLs, increasing provisioning costs, falling profits, and increasing distressed loans at the end of the cycle.
  - The second boom was followed by a sharper credit decline.
- Bank risk-on/off behavior:
  - At beginning of booms banks had plenty of capital and took more risk; credit growth accelerated and loan quality deteriorated.
  - Near end of booms rising NPLs led to higher provisions, reduced profits, greater risk aversion, and a shift from loans to risk-free assets, contributing to the credit bust.
  - Mid-2016: agriculture prices fell; banks tightened credit and credit growth collapsed from 30 percent (y/y) in mid-2015 to close to zero percent in late 2016.
- Funding and sector concentration:
  - During booms banks relied more on non-core funding sources, including non-residents; trend stronger during the second boom.
  - Domestic banks increased credit to the agriculture sector by close to 165 percent over 2011- 2015, while soy price increased by 150 percent over the same period.
  - Credit to the soy sector explains about 49 percent of total agriculture sector credit, on average from 2008-2016.
- Heterogeneous responses by bank ownership:
  - Domestic banks repaired balance sheets more after the soy price collapse and reduced exposure to the soy sector more than non-domestic banks.
  - Domestic banks relied less on core funding and more on non-core funding (e.g., securities) after the shock.
  - Domestic banks respond more strongly to rises in regional EMBI spreads than foreign banks, driving reductions in dollar-denominated credit growth.
  - Among foreign banks: Brazilian and Argentinean banks behaved differently — Argentinian banks tended to shrink Paraguayan balance sheets when stress rose; Brazilian-owned banks increased assets and loans in Paraguay when EMBI spreads increased in Brazil, suggesting relocation of funds.
  - As of end-2017: Brazilian banks make up around 19 percent of total assets, Argentinian banks around 11 percent, and domestic banks close to 50 percent.

### Monetary and macroprudential policy response
- Monetary policy:
  - The monetary policy rate has not been actively used to target the credit cycle because credit booms were not associated with price pressures and inflation remained low.
  - During the second credit boom, monetary policy remained accommodative given low inflationary pressures and uncertain international context.
- Macroprudential toolkit and usage:
  - The macroprudential toolkit in Paraguay is limited.
  - At the height of the credit boom in 2010, BCP introduced two prudential policies: (i) introduce leverage limits to contain banks’ lending and (ii) regulations tightening FX sales.
  - Instead of tightening NPL regulations in December 2015, BCP introduced transitory measures that effectively relaxed the definition of NPLs for credits to the agriculture and livestock sectors (Resolution No. 33 and Resolution No.34, December 2015).
- Available macroprudential policies through end-2016 (selected, as listed):
  - CCB: 
  - LVR: 
  - Capital requirements: 
  - LFC (Limits on foreign currency lending & rules/recommendations on FC loans): 
  - RR (Reserve requirements): 
  - LFX (Limits on net or gross open FX positions, FX exposures and FX funding): 
  - Other items listed in the table include LLP, LTV, DSTI, Tax, SIFI, Liquidity, LTD, LCG, Conservative, LoanR, Other — several marked "..." indicating not available/blank.

### Conclusions (summary of empirical results)
- Long-run: External factors play a more important role than domestic factors in driving credit growth. Growth of major trading partners is the most significant external factor for guaraní-denominated credit (explaining more than half of total variation) and is also significant for dollar-denominated credit (close to one-third of total variation). EMBI spreads and soy bean prices each explain about 18 percent of movement in dollar-denominated credit growth.
- Short-run: Both guaraní- and dollar-denominated credit growth respond in tandem to growth of major trading partners. Regional risk-aversion (EMBI spreads) can lower dollar-denominated credit growth but not guaraní-denominated credit growth. Soy bean price improvements spur dollar-denominated credit growth but not guaraní-denominated credit growth. Monetary tightening reduces guaraní-denominated credit growth but increases dollar-denominated credit growth in the short-term; real appreciation raises dollar-denominated credit growth.
- Bank-level: Banks' risk on-off behavior contributes to credit cycles; behavior is heterogeneous across banks and depends on origin of capital.

*Source: BCP and IMF staff calculations.*

### References

### 1pryea2019002 - References

### Bibliographic and cited works
- Caruana, J., 2011, “Why central bank balance sheets matter”, keynote address at the Bank of Thailand-BIS Research Conference entitled “Central bank balance sheets in Asia and the Pacific: the policy challenges ahead”, Chiang Mai, Thailand, 12–13 December, www.bis.org/speeches/sp111216.pdf
- Chen, Q., A. Filardo, D. He, and F. Zhu, 2012, “International spillovers of central bank balance sheet policies,” BIS Papers, No. 66p
- Cook, D and J. Yetman (2012): “Expanding central bank balance sheets in Emerging Asia: a compendium of risks and some evidence”, BIS Papers, No 66d
- Dell’Ariccia, G, Igan D., Laeven, L., and Tong, H., with Bakker, B. and Vandenbussche, J., (2012), “Policies for Macrofinancial Stability: How to Deal with Credit Booms,” IMF Staff Discussion Note 12/06
- Franks, J., Mercer-Blackman, V., Sab, R. and Benelli, R. (2003), “Paraguay – Corruption, Reform and the Financial System.”
- Gadanecz, B, A Mehrotra and M S Mohanty (2014): “Foreign exchange intervention and the banking system balance sheet in emerging market economies”, BIS Working Paper no 445
- Garcia-Herrero, Alicia (1997): “Banking Crises in Latin Amercia in the 1990s: Lessions from Argentina, Paraguay, and Venezuela,”, IMF Working Paper 11/3/97
- International Monetary Fund, 1989-2017, Paraguay: Staff Reports for the Article IV Consultation

### Annex I. Estimation Result of the Structural VAR — model specification and data
- Model specification (vector form): y_t A + u_t = sum_{i=1}^p Φ_i y_{t-i} + B ε_t + μ  (notation as in source)
- y_t is a (9 X 1) vector consisting of:
  - Fed Fund Rate (dyfedpolicy)
  - Trading Partner Growth (worldgdpA)
  - Soy Bean Price (dypsoy)
  - Latin America EMBI Spread (dyembi_lac)
  - Domestic Monetary Policy Rate (dypolicy)
  - Real GDP growth (dyngdpr)
  - Real Effective Exchange Rate (dyreer)
  - Guaraní-denominated credit (dycredit_lc)
  - Dollar-denominated credit (dycreditfcx)
- Sample: quarterly data from 2006Q1 to 2018Q3.
- Measurement: All variables are year-over-year growth rates except Fed Fund Rate, Latin America EMBI Spread and Domestic Monetary Policy Rate, which are year-over-year difference in percentage point.
- Shock vector ε_t is (9 X 1), assumed independently normally distributed.
- Restrictions: imposed on matrices A and B to reflect prior that domestic variables do not affect external variables.

### Annex I. Estimation Results — Matrix A (selected coefficients, standard errors, and significance)
- Notes: "const" indicates constrained to zero in estimation. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1.
- Representative lag-1 coefficients (row variable at t-1 → column variable at t), with standard errors:
  - dyfedpolicy (t-1) → dyfedpolicy (t): 0.93*** (0.05)
  - worldgdpA (t-1) → worldgdpA (t): 0.75*** (0.06)
  - dypsoy (t-1) → dypsoy (t): 0.80*** (0.11)
  - dyembi_lac (t-1) → dyembi_lac (t): 0.70*** (0.12)
  - dypolicy (t-1) → dypolicy (t): 0.74*** (0.10)
  - dyngdpr (t-1) → dyngdpr (t): 0.21*** (0.07)
  - dyreer (t-1) → dyreer (t): 0.46*** (0.12)
  - dycredit_lc (t-1) → dycredit_lc (t): 0.73*** (0.09)
  - dycreditfcx (t-1) → dycreditfcx (t): 0.84*** (0.09)
- Selected cross-variable significant effects (lag-1):
  - worldgdpA (t-1) → dyembi_lac (t): 33.59*** (8.36)
  - dypolicy (t-1) → dyreer (t): 0.70** (0.32)
  - dypolicy (t-1) → dycredit_lc (t): -0.95*** (0.32)
  - dycredit_lc (t-1) → dycreditfcx (t): 0.21** (0.10)
- Constants (selected):
  - dyfedpolicy constant: 0.40*** (0.09)
  - worldgdpA constant: 0.76*** (0.22)
  - dyembi_lac constant: -95.47*** (28.98)
  - dypolicy constant: -1.55*** (0.53)
- Observations: 50 (reported across columns)
- R-squared (by equation): 0.93, 0.88, 0.66, 0.62, 0.80, 0.37, 0.76, 0.93, 0.94

### Annex I. Estimation Results — Matrix B (impact coefficients, selected entries)
- "const" indicates constrained to zero in estimation.
- Selected impact coefficients (row shock → column variable):
  - dyfedpolicy (shock) → dyfedpolicy (t): 0.34
  - worldgdpA (shock) → worldgdpA (t): 0.80 with cross-impact from dyfedpolicy: 0.14
  - dypsoy (shock) → dypsoy (t): 14.88 with cross-impacts from dyfedpolicy: 1.34 and worldgdpA: 8.44
  - dyembi_lac (shock) → dyembi_lac (t): 71.19 with cross-impacts from dyfedpolicy: -28.19, worldgdpA: -56.86, and dypsoy: -27.89
  - dycreditfcx (shock) row (impacts on all columns): 0.65, 1.90, 0.49, -0.79, 2.04, 0.90, 1.41, 0.43, 3.87

### Annex II. The Banking Sector Between 1990 and 2006 — narrative findings and key statistics
- Early liberalization and reforms in early 1990s:
  - Reforms included unification of exchange rates, floating of the guarani, liberalization of interest rates, introduction of market-based monetary instruments, partial elimination of selective credit controls, authorization of foreign currency loans by local banks for selected purposes, harmonization and gradual reduction of reserve requirements, and open market operations by BCP using its own short-term debt instrument.
- Expansion of banking sector and entry:
  - Number of financial intermediaries rose to close to 100 in late 1994.
  - Number of banks by year:
    - 1995: 35
    - 1996: 32
    - 1997: 33
    - 1998: 23
    - 1999: 22
    - 2000: 20
    - 2001: 18
    - 2002: 14
    - 2003: 8
  - Wholly foreign owned banks by year:
    - 1995: 13
    - 1996: 14
    - 1997: 9
    - 1998: 9
    - 1999: 9
    - 2000: 8
    - 2001: 7
    - 2002: 6
    - 2003: (noted as 6)
  - Number of finance companies by year:
    - 1995: 68
    - 1996: 54
    - 1997: 50
    - 1998: 36
    - 1999: 33
    - 2000: 32
    - 2001: 25
    - 2002: 21
    - 2003: 17
  - Source: Chapter 2 of Paraguay Corruption Reform and the Financial System, Table 2.2
- Drivers of credit expansion and crisis vulnerabilities:
  - High inflation and a cap on lending interest rate produced negative real lending rates, fueling credit expansion.
  - Policy measures in 1992 (subsidized credit through commercial banks to agricultural sector) further contributed to credit growth.
  - Weak banking practices: excessive concentration, connected lending without adequate credit-risk analysis, inadequate accounting practices, and inefficient supervisory framework.
  - Lax enforcement of capital requirements and minimum entry capital; many new small banks and finance companies had very low capital.
  - Political constraints limited the Superintendence of Banks from applying sanctions.
- Banking crisis dynamics (mid-1995 onset):
  - Trigger: publication of an accounting discrepancy in BCP’s books and failure of large banks identified with capital deficiencies.
  - BCP interventions: provided large-scale liquidity support; BCP window became main liquidity source as healthier banks were reluctant to channel excess liquidity through interbank market.
  - Uses of BCP borrowing by distressed banks: ~half to offset deposit withdrawals; ~half to cover call loans and short-term external obligations.
  - Crisis culminated with waves of bank failures ending in late 1998 after closure of three weak banks.
  - By late 1997–1998, about half of financial institutions had failed.
  - Direct cost of intervention led by central bank estimated at around 5 percent of GDP, composed of:
    - liquidity support amounting to 4 percent of GDP provided by the BCP at end-1995 to intervened and distressed banks;
    - additional liquidity support of 1 percent of GDP in mid-1996.
- Structural change and concentration after crisis:
  - By end-December 2002:
    - 85 percent of total deposits were in branches of foreign banks or majority foreign-owned banks.
    - 83 percent of total assets were held by foreign banks or majority foreign-owned banks.
  - Increased share of foreign-owned banks slowed credit recovery for small local firms.
- Post-crisis period and supervisory reform:
  - Recession between 1998 and 2002 due to Argentinean financial crisis, drought, and foot-and-mouth disease; non-performing loan ratio increased and credit to private sector fell from 1997 peak.
  - Mid-2002: deposit run on Banco Aleman (third-largest bank by assets, about 11 percent of banking system assets) after parent institutions ran into difficulties; authorities intervened, paid out small deposits with a deposit guarantee provided by the central bank, and transferred most other assets and liabilities to other banks.
  - Between 2003 and 2008: macroeconomic improvement under IMF programs; central bank shifted to inflation targeting; prudent fiscal policy reduced government debt significantly; global agricultural commodity price boom led to rapid growth in real incomes and a credit boom.
  - Improvements to banking supervision: major revisions to banking laws aligned supervision closer to Basel Core Principles; supervisors granted more resources, more independence and more sanctioning capacity; management capacity and corporate governance strengthened.
  - Bank-level improvements: individual banks generally held more capital than legally required; provisioning for NPLs increased considerably.
- Lessons and later resilience:
  - Causes of the early 1990s credit boom: financial liberalization, favorable economic conditions, and loose entry requirements that allowed weak banks to enter; poor enforcement of banking laws.
  - Outcome: banking crisis with about half of banks failing and collapse of credit; credit growth resumed around 2006 and continued until 2015.
  - Fall of soybean price in 2015 triggered decline in credit growth but did not produce a banking crisis thanks to improved banking-sector resilience from supervisory modernization and stronger banks adopting safer practices.

*International Monetary Fund — content unit: 1pryea2019002 - References*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2019/1pryea2019002.pdf_