## wp1760 - References

## Source details

**Canonical URL:** [wp1760 - References](https://www.imf.org/-/media/files/publications/wp/2017/wp1760.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2017/wp1760.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2017/wp1760.pdf.json)

---

### 1. Introduction: ownership rationales and debates
- Financial globalization since the 1990s: increase in foreign-owned banks and decline in government-owned banks.
- Post-GFC debate:
  - Foreign-owned banks: viewed as channels transmitting crisis from advanced to developing countries.
  - Government-owned banks: systemically important government banks (e.g., Brazil, China, India) recovered quickly, suggesting a potential mitigating role.
- Two benign rationales for government-owned banks:
  - “Social” view: overcome market failures and promote welfare-enhancing investments (Atkinson and Stiglitz, 1980; Stiglitz, 1993).
  - “Development” view: allocate resources to strategically important industries (Gerschenkron, 1962).
- Two critical views of government-owned banks:
  - “Agency” view: agency costs lead to inefficiencies and misallocation (Banerjee, 1997; Hart et al., 1997).
  - “Political” view: politicians pursue personal or political goals, causing misallocation (Shleifer and Vishny, 1994; Shleifer, 1998).
- Political economy prediction: politicians more likely to favor government bank ownership when public accountability and judicial independence are low (Perotti and Vorage, 2010).
- Foreign bank arguments:
  - Support: bring capital, technical skills, product innovation, foster competition and efficiency (Levine, 1996; Goldberg, 2004).
  - Criticism: transmit external shocks, threaten local banks, concentrate on least risky segments reducing access (Stiglitz, 1994; Detragiache et al., 2008).
- Empirical approach: evidence summarized separately for government and foreign ownership.
- Paper organization: Section 2 trends and regional patterns; Section 3 literature on ownership, performance, competition; Section 4 evidence on financial stability; Section 5 impact on access to finance; Section 6 conclusions.

### 2. Trends in bank ownership: aggregate shares and time series
- Aggregate shares:
  - Foreign-owned banks account for 43 percent of total banking assets across all countries.
  - Government-owned banks account for 18 percent of banking system assets.
- Cross-country/regional variation documented for 93 countries (Table A.1).
- Trends (1990s up to GFC):
  - Developing countries foreign-owned asset share:
    - rose from 8 (median) and 16 (average) percent in 1995 to 35 (median) and 40 (average) percent at end of the 1990s and reached 52 (median) and 50 (average) percent by 2010.
  - High-income countries foreign-owned asset share:
    - rose from 5 (median) and 15 (average) percent in 1995 to 12 (median) and 28 (average) percent by end of the 1990s, reaching 27 (median) and 43 (average) percent in 2008.
  - Government-owned banks declined:
    - Developing countries: median (average) fell from 40 (40) percent in 1995 to 8 (17) percent in 2008.
    - High-income countries: median (average) fell from 30 (36) percent in 1995 to 4 (10) percent in 2008.
- Impact of the GFC (2008–2010):
  - Between 2008 and 2010, average (median) share of assets held by foreign-owned banks fell by 0.2 (3.1) percentage points in developing countries and by 1.6 (2.2) percentage points in high-income countries.
  - Repatriation of funds to cover capital losses and meet higher capital requirements observed.
  - South-South banking: banks from non-OECD countries recorded positive net foreign entries every year between 2009 and 2013 and more than doubled their presence abroad, mostly within their regions (Claessens and Van Horen, 2015).
- Post-GFC government involvement:
  - Several countries recapitalized or nationalized troubled banks; governments’ asset shares increased by more than 10 percentage points between 2008 and 2010 in Iceland, Kazakhstan, the United Kingdom, and Venezuela (Table A.1).

### 3. Regional patterns and notable country/region findings
- Latin America:
  - Foreign-owned banks ~40% (1999-2010); government-owned banks ~13% (1999-2010).
  - Spanish banks’ 1990s acquisitions increased foreign presence.
  - Reliance of foreign-owned banks on local funding and independent subsidiaries limited retrenchment at GFC onset (Cull and Martinez Peria, 2013).
- Eastern Europe & Central Asia:
  - Median foreign-owned asset share rose from 5 percent in 1995 to 60 percent by 2008, then fell from 60% in 2008 to 48% in 2010.
  - Government-owned banks’ median asset share increased by 5 percentage points between 2008 and 2010—the highest regional increase.
- Sub-Saharan Africa:
  - Median foreign-owned asset share rose from 34 percent in 1995 to 66 percent in 2008.
  - In 2010 the region had median foreign bank share 73 percent and median government bank share 8 percent (highest foreign, lowest government among developing regions).
  - Expansion of Pan-African banks (South Africa, Nigeria, Morocco, Kenya, Togo) sustained foreign participation after Western retrenchment (Beck et al., 2014).
- South Asia:
  - Lowest foreign-owned share and highest government-owned share among regions.
  - Foreign bank participation rose from 1 percent in 1995 to 15 percent in 2008.
  - Government bank ownership declined from 72 percent in 1995 to 39 percent by 2010.
- East Asia and Pacific:
  - Median foreign-owned asset share rose from 8 percent in 1995 to almost 30 percent before the GFC.
  - Median government-owned asset share fell from 23 percent in 1995 to 13 percent by 2008.
  - Figure 3.1 shows a substantial decline in foreign-bank asset share in 2012-2013 relative to 2010 (2012-2013 numbers come from a different source and cover fewer countries).

### 4. Bank ownership, performance, and competition
- Mechanism: efficient foreign banks can pressure domestic banks to reduce spreads and costs, improving efficiency.
- Empirical evidence: at least fifty studies compare foreign and domestic bank performance with mixed results; sensitivity to methods and performance measures noted.
- Evidence on foreign-owned vs domestic banks:
  - High-income country studies often suggest domestic banks outperform foreign banks (Berger et al., 2000; Miller and Parkhe, 2002); U.S. evidence listed (DeYoung and Nolle, 1996; Hasan and Hunter, 1996; Chang et al., 1998; Peek et al., 1999).
  - Developing-country studies largely find foreign-owned banks perform better in cost and profit efficiency (various country and regional studies cited).
  - Mixed/ambiguous evidence in China and India; scant evidence in Sub-Saharan Africa and MENA (Farazi et al., 2013 exception).
- Cross-country large-sample findings:
  - Foreign ownership associated with higher interest margins and profitability, and lower overhead costs, in developing hosts (Demirgüç-Kunt and Huizinga, 1999; Micco et al., 2007).
  - Opposite found in high-income countries (Claessens et al., 2001).
  - Heterogeneity factors: home country income, shared language/regulatory framework, weak host-country regulations, and foreign bank size/market share (Claessens and Van Horen, 2012).
- Foreign entry, competition, and mode of entry:
  - Foreign banking can negatively affect domestic bank performance via competitive pressures (Demirgüç-Kunt et al.,1998; Claessens et al., 2001; Bayraktar and Wang, 2004).
  - Threshold effects: competitiveness impact stronger where foreign-owned banks have higher market share (Claessens and Lee, 2003).
  - Mode matters:
    - Greenfield: tends to increase number of banks and decrease market power (fiercer price competition).
    - Acquisitions: can increase concentration and market power, lowering competition (Lehner and Schnitzer, 2008).
  - Spatial competition and information asymmetry models influence lending rates and competition outcomes (various studies cited).
- Competition measures and findings:
  - Early cross-country evidence suggests foreign ownership positively related to competition (Claessens and Laeven, 2004; Gelos and Roldos, 2004); Levy-Yeyati and Micco (2007) find opposite in Latin America.
  - Positive foreign-competition link stronger when entering banks are more efficient and less risky and when host banking sector is less concentrated.
  - Positive effect of foreign bank penetration on competition stronger with greenfield than with M&A (Jeon et al., 2011).
- Bank-level Lerner index evidence (Delis et al., 2016):
  - Sample: bank-level data for 131 countries over 1997-2009.
  - No direct impact of bank ownership (foreign vs domestic) on bank-level market power.
  - Greater foreign bank presence has a positive effect on overall levels of market power in the banking sector; attributed to foreign entry occurring mainly through mergers and acquisitions rather than greenfield investments.
- Evidence on government (state) bank ownership:
  - Theoretical: state ownership can spur competition if state-owned banks are efficient, but in developing countries high agency costs limit competitive pressure.
  - Relative performance:
    - Regional studies often find government-owned banks perform worse than private domestic and foreign banks; notable Latin America result (Micco et al., 2007).
    - South and South-East Asia: government-owned banks underperform (Williams and Nguyen, 2005; Micco et al., 2007; Cornett et al., 2010).
    - India: opposite findings where government-owned banks represented 78 percent of banking system assets at the turn of the century (Sensarma, 2005).
    - MENA: state-owned banks have lower interest margins, higher cost ratios, larger NPLs, larger loan loss provisions, and lower profitability than private domestic non-listed banks (Farazi et al., 2013).
    - Western Europe (1999–2004): government-owned banks exhibited lower profitability and cost efficiency despite lower funding costs (Ianotta et al., 2007).
    - Germany case studies: government-owned banks have profit and cost advantage; government-owned banks represented 42 percent of banking assets in 2000 (Altunbas et al., 2001; Chakravarty and Williams, 2006).
  - Large-sample studies: government-owned banks underperform in developing countries; differences significant when including country fixed effects (Micco et al., 2007; Mian, 2003).
  - Gap: no cross-country study systematically analyzing how government-bank relative performance varies with home/host country conditions and bank characteristics.
- Bank privatization and efficiency effects:
  - Evidence that privatization raises bank profitability and efficiency over time across various developing-country studies and panels (Boubakri et al., 2005; Williams and Nguyen, 2005; country case studies cited).
  - Berger et al. (2005) methodology applied to Argentina confirms privatizations associated with significant performance improvements.
  - Market implications: privatizations sometimes negatively impacted stock performance of private competitors, consistent with government ownership undermining competition (Otchere and Chan, 2003; Chen et al., 2005; Otchere, 2005).

### 5. Bank ownership and banking sector stability
- Foreign bank ownership: stabilizing or destabilizing depending on shocks
  - Domestic shocks: foreign banks may stabilize via access to parent liquidity and diversification.
  - External shocks: foreign banks can import shocks, amplify transmission, or increase domestic competition that pushes domestic banks into riskier activities.
- Cross-country/panel evidence:
  - 80-country evidence (1988-1995): greater foreign bank presence associated with lower probability of systemic banking crises (Demirguc-Kunt et al., 1998).
  - Broader sample (107 countries): official barriers to foreign bank entry associated with measures of banking system fragility (Barth et al., 2004).
- Crisis-period behavior:
  - Greenfield foreign banks’ credit growth more resilient than private domestic banks during domestic crises (De Haas and Van Lelyveld, 2006; De Haas and Van Lelyveld, 2010).
  - Latin America and Asia (1989-2001): foreign banks’ lending and deposit rates less volatile than domestic banks in crises (Arena et al., 2007).
  - Argentina, Brazil, Mexico (1994–1999): foreign banks did not pull back during economic problems; had higher growth and lower lending volatility in crises (Peek and Rosengren, 2000a; Dages et al., 2000; Crystal et al., 2001, 2002).
  - Malaysian Asian-crisis study: diversified foreign banks played a stabilizing role (Detragiache and Gupta, 2006).
- Evidence that foreign banks can amplify external shocks:
  - Cross-border transmission documented using exogenous shocks (Peek and Rosengren, 1997, 2000b; Chava and Purnanandam, 2011; Schnabl, 2012).
- GFC evidence (2008–2009):
  - Overall, foreign banks reduced lending earlier and faster than domestic banks (Aiyar, 2012; Fungáčová et al., 2013; de Haas and van Lelyveld, 2014; Claessens and van Horen, 2015; Choi et al., 2016), notably in Eastern Europe (Mihaljek, 2011; de Haas et al., 2015).
  - Heterogeneity: stronger transmission where host countries are less regulated (Anginer et al., 2017), more financially open, and more competitive (Jeon et al., 2013).
  - Transmission stronger for foreign-owned banks with lower capital, profitability, retail funding, and less independent management (Popov and Udell, 2012; Jeon et al., 2013; de Haas and Van Lelyveld, 2014; de Haas et al., 2015; Anginer et al., 2017; Choi et al., 2016).
  - Transmission stronger for greenfield entrants than M&A entrants (Jeon et al., 2013).
  - Internal cross-border capital allocation and active internal capital markets documented (Cetorelli and Goldberg, 2012a, 2012b, 2012c; Jeon et al., 2013).
  - Dataset example: 45 multinational banks from 18 home countries with 194 subsidiaries across 46 countries.
- Government bank ownership and stability:
  - Mixed evidence on direct association with banking crises:
    - Caprio and Martinez Peria (2002): government ownership positively associated with likelihood of banking crises (sample: 64 countries, 1980-1997).
    - La Porta, Lopez de Silanes and Schleifer (2002): no significant results.
    - Barth et al. (2004): positive bi-variate correlation present but not robust after controlling for regulatory/supervisory features.
  - More robust associations with other instability measures:
    - Government-owned banks in developing countries have higher fractions of non-performing loans and lower Z-scores (Micco et al., 2004; IADB 2005; Iannotta et al., 2007; Berger et al., 2009; Farazi et al., 2011).
  - Political economy and misallocation evidence:
    - Government banks subject to political manipulation; increase lending to politically strategic sectors/regions near elections and favor politically connected firms (Sapienza, 2004; Khwaja and Mian, 2005; Cole, 2009a,b; Carvalho, 2014; Claessens et al., 2008).
  - Crisis-period behavior:
    - Government bank lending less procyclical than private bank lending in pre-GFC periods (Micco and Panizza, 2006; Foos, 2009; Brei and Schclarek, 2013; Onder and Ozyldirim, 2013; Bertay et al., 2015; Duprey, 2015).
    - Mixed evidence during the GFC; some studies find government banks supportive in crises (Leony and Romeu, 2011; Cull and Martinez Peria, 2013; Coleman and Feler, 2015; De Haas et al., 2015; Chen et al., 2016; Choi et al., 2016; Allen et al., forthcoming).

### 6. Bank ownership and access to credit
- Foreign bank ownership: information asymmetry and access implications
  - Theoretical: foreign banks face informational disadvantage vs domestic banks for opaque borrowers; rely on "hard information" unless they acquire local resources/expertise via M&A.
  - Theoretical prediction: foreign bank participation associated with reduced access to credit for informationally opaque firms (Detragiache et al., 2008; Gormley, 2014).
  - Cross-country patterns:
    - Negative association between foreign bank presence and private credit to GDP and private credit growth rates in low-income countries with severe information barriers (Detragiache et al., 2008; Claessens and Van Horen, 2014).
    - Negative associations in countries with costlier contract enforcement and limited credit information (Claessens and Van Horen, 2014).
    - SME lending prevalence/pricing tied to institutional and legal environment (Beck et al., 2011).
  - Entry context matters:
    - Negative relation driven by non-random entry into crisis-hit markets or markets with prior credit declines; foreign banks often acquired failed domestic banks and recapitalized sectors, temporarily reducing credit (Cull and Martinez Peria, 2008).
  - Empirical evidence: foreign banks favor larger, less informationally opaque borrowers in Argentina, Mexico, Pakistan, India, and CEE (Berger et al., 2001; Beck and Martinez Peria, 2010; Mian 2003, 2006; Berger et al., 2008; Gormley, 2010; Beck and Brown, 2015).
  - Countervailing forces:
    - Legal/regulatory improvements enable new lending technologies facilitating arms-length SME lending (Berger and Udell, 2006).
    - Foreign banks adopting credit scoring, standardized risk-rating, asset-based lending, factoring, fixed-asset lending, leasing (De la Torre et al., 2010).
    - Large foreign banks in four Latin American countries lent as much or more to SMEs as large domestic banks (Argentina, Chile, Colombia, Peru) (Clarke et al., 2005b).
    - Bolivia credit registry: foreign banks can serve same clientele by requiring collateral, shorter maturities, pricing based on credit ratings/collateral (Beck et al., forthcoming).
  - Competitive dynamics:
    - Foreign competition for large transparent borrowers can induce some domestic banks to “go down-market,” increasing lending to opaque borrowers (Clarke et al., 2006).
    - Eastern Europe: foreign bank presence not associated with reduced SME access; many SMEs incorrectly anticipated rejection (De Haas et al., 2010; Giannetti and Ongena, 2012; Brown et al., 2011).
  - Business-cycle adaptability:
    - Relationship lenders better able to relax credit constraints in downturns; charge higher spreads in normal times to lend favorably in crises (Bolton et al., 2016).
    - Survey: CEOs of 400 banks across 21 Eastern Europe and Caucasus countries found areas with more relationship lenders saw less severe financial constraints during the 2008 crisis (Beck et al., 2014).
    - Traditional depiction (foreign = transaction lenders, domestic = relationship lenders) does not uniformly hold; heterogeneity exists across ownership types.
    - Heterogeneity by foreign bank type: ‘Big Six’ multinational European banks in Eastern Europe maintained lending growth comparable to domestic banks during crises; non-‘Big Six’ foreign banks had higher pre-crisis loan growth but declined during crises (Bonin and Louie, forthcoming).
- Government bank ownership: access and allocation implications
  - Rationale: state banks motivated to correct credit market failures and serve segments underserved by private banks (small, opaque borrowers).
  - Cross-country findings:
    - Higher government ownership associated with fewer bank branches and ATMs; not significantly linked to indicators of prevalence of loans and deposit accounts (Beck et al., 2007).
    - Proxies for barriers to access/use of banking services increase with government ownership and decrease with foreign participation.
    - Bank concentration’s negative effect on access more pronounced with higher state ownership (Beck et al., 2004).
  - Political influence and allocation:
    - Political influences undermine state banks’ ability to deepen access: Italy (lower rates and more lending to politically affiliated firms in election years, Sapienza, 2004); India (state banks maintain lending to government-owned and rural firms rather than smaller opaque firms, Berger et al., 2008); China (disproportionate credit to government-owned firms and those with Communist Party affiliation, Cull et al., 2015).
  - Crisis-stabilizing role:
    - Theoretical models: state banks better able to lend in crises because they are less profit-maximizing and less exposed to deposit withdrawals due to recapitalization prospects (Brei and Schclarek, 2015).
    - Empirical: state bank lending substantially less procyclical where governance is good; state banks expand non-deposit liabilities more slowly during downturns and show more stable NPL ratios over the cycle (Bertay et al., 2015).
    - In countries with good governance, state banks may help stabilize credit flows during crises.
- Synthesis (ownership trade-offs and overall implications)
  - Ownership changes since mid-1990s: foreign participation increased while government-owned presence declined until the GFC.
  - Empirical summary:
    - Foreign-owned banks: tend to be more efficient than domestic banks in developing countries, promote competition, and help stabilize credit under idiosyncratic shocks.
    - Trade-offs: foreign banks can transmit external shocks and may not always expand access to credit; effects depend on entry timing, mode (greenfield vs M&A), competition level, and institutional development. Negative relations between foreign presence and credit levels sometimes reflect foreign banks entering to recapitalize failing systems.
    - Government-owned banks: little evidence of substantial benefits relative to other ownership types for the banking sector, real economy, or users of banking services, especially in developing countries. Some evidence government banks can stabilize credit growth in crises, but overall record mixed and often associated with reduced competition and weaker performance.

### 7. References, data notes, and figure/table annotations (selected numeric and dataset details)
- Datasets, panels, and figure notes:
  - "2012-2013 data are available for 80% of the sample." (Figure 1.1 notes)
  - Figure 1.2: "The map reflects 2010 data, except for the MENA region where 2008 figures are used instead." (Figure 1.2 notes)
  - Ownership definition in figures/appendix: "Ownership is defined as the asset share of banks that are more than 50% controlled by the foreign owners, except for 1995 where ownership reflects the actual share owned by foreign banks."
  - Comparable government-ownership definition: "Ownership is defined as the asset share of banks that are more than 50% controlled by the government, except for 1995 where ownership reflects the actual share owned by the government."
- Time-series percentage sequences (verbatim from figures):
  - "16% 40% 50% 49% 43% 15% 28% 43% 42% 36% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 1995 1999-2001 2008 2008-2010 2012-2013 1995 1999-2001 2008 2008-2010 Developing countriesHigh-income countries"
  - "40% 23% 17% 18% 36% 15% 10% 12% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 1995 1999-2001 2008 2010 1995 1999-2001 2008 2010 Developing countriesHigh-income countries"
  - Regional median blocks (verbatim selections):
    - "8% 5% 6% 1% 34% 5% 16% 28% 38% 10% 67% 12% 28% 60% 41% 15% 66% 27% 28% 48% 39% 14% 73% 25% 12% 63% 30% 3% 51% 18% East Asia & Pacific Europe & Central Asia Latin America & Caribbean South AsiaSub-Saharan Africa High income (all regions) 1995 1999-2001 2008 2010 2012-2013"
    - "23% 48% 43% 72% 22% 30% 6% 15% 14% 65% 6% 14% 13% 9% 12% 45% 8% 4% 13% 14% 13% 39% 8% 5% East Asia & Pacific Europe & Central Asia Latin America & Caribbean South AsiaSub-Saharan Africa High income (all regions) 1995 1999-2001 2008 2010"
- Table A.1 notes (verbatim):
  - "Notes: Income groups follow the 2010 World Bank analytical classification. Government (respectively foreign) ownership is defined as the asset share of banks that are more than 50% controlled by the government (respectively foreign owners), except for 1995 where ownership reflects the actual share owned by the government (respectively foreign owners)."
  - "Sources: Micco, Panizza and Yanez (2007) for 1995 data, Regulation and Supervision surveys (World Bank) for 1999-2010 data, Claessens and Van Horen (2015) for 2012-2013 data."

*Content derived from wp1760 - References (PDF) as provided.*

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

### wp1760 - References

### 1. Introduction
- Financial globalization since the 1990s led to significant changes in bank ownership structures: an increase in foreign-owned banks and a decline in government-owned banks.
- The Global Financial Crisis (GFC) of 2008–2009 reignited debate on ownership structure: foreign-owned banks were viewed as a channel transmitting the crisis from advanced to developing countries, while countries with systemically important government-owned banks (e.g., Brazil, China, India) recovered quickly, suggesting a potential mitigating role for government banks.
- Two benign rationales for government-owned banks:
  - “Social” view: help overcome market failures and promote welfare-enhancing investments (Atkinson and Stiglitz, 1980; Stiglitz, 1993).
  - “Development” view: allocate resources to strategically important industries the private sector will not finance (Gerschenkron, 1962).
- Two critical views of government-owned banks:
  - “Agency” view: agency costs of government bureaucracy lead to operational inefficiencies and misallocation (Banerjee, 1997; Hart et al., 1997).
  - “Political” view: politicians use government banks to pursue personal or political goals, causing misallocation (Shleifer and Vishny, 1994; Shleifer, 1998).
- Political economy models: politicians more likely to favor government bank ownership when public accountability and judicial independence are low (Perotti and Vorage, 2010).
- Arguments in support of foreign bank ownership: bring capital, technical skills, and product innovation, foster competition and efficiency (Levine, 1996; Goldberg, 2004).
- Criticisms of foreign-owned banks: can transmit external shocks, threaten local banks via competition, and concentrate on least risky segments reducing access to finance (Stiglitz, 1994; Detragiache et al., 2008).
- Empirical work often mixes private domestic, government, and foreign categories; this paper summarizes evidence for government and foreign ownership separately.
- Paper organization: Section 2 trends and regional patterns; Section 3 literature on ownership, performance, competition; Section 4 evidence on financial stability; Section 5 impact on access to finance; Section 6 conclusions.

### 2. Trends in bank ownership
- Aggregate shares:
  - Foreign-owned banks account for 43 percent of total banking assets across all countries.
  - Government-owned banks account for 18 percent of banking system assets.
- Cross-country and regional variation is substantial; Table A.1 reports country-level shares for 93 countries.
- Trends from the 1990s up to the GFC:
  - Developing countries: median (average) share of assets owned by foreign banks rose from 8 (16) percent in 1995 to 35 (40) percent at end of the 1990s and reached 52 (50) percent by 2010.
  - High-income countries: median (average) foreign-owned asset share rose from 5 (15) percent in 1995 to 12 (28) percent by end of the 1990s, reaching 27 (43) percent in 2008.
  - Government-owned banks declined:
    - Developing countries: median (average) government-owned asset share fell from 40 (40) percent in 1995 to 8 (17) percent in 2008.
    - High-income countries: median (average) fell from 30 (36) percent in 1995 to 4 (10) percent in 2008.
- Impact of the GFC:
  - Between 2008 and 2010, the average (median) share of assets held by foreign-owned banks fell by 0.2 (3.1) percentage points in developing countries and by 1.6 (2.2) percentage points in high-income countries.
  - Many financial institutions repatriated funds to cover capital losses and meet higher capital requirements.
  - South-South banking investment: banks from non-OECD countries recorded positive net foreign entries every year between 2009 and 2013 and more than doubled their presence abroad, mostly within their regions (Claessens and Van Horen, 2015).
- Post-GFC government involvement:
  - Several countries recapitalized or nationalized troubled banks after the GFC; notable increases of more than 10 percentage points in government asset shares between 2008 and 2010 occurred in Iceland, Kazakhstan, the United Kingdom, and Venezuela (see Table A.1).

### 3. Regional patterns and notable country/region findings
- Latin America:
  - Foreign-owned banks represent roughly 40% over 1999-2010.
  - Government-owned banks roughly 13% over 1999-2010.
  - Spanish banks’ acquisitions in the 1990s increased foreign presence.
  - Reliance of foreign-owned banks on local funding and independence of subsidiaries limited retrenchment at GFC onset (Cull and Martinez Peria, 2013).
- Eastern Europe & Central Asia:
  - Rapid increase in foreign bank presence in the 1990s and early 2000s.
  - Median foreign-owned asset share rose from 5 percent in 1995 to 60 percent by 2008, then fell from 60% in 2008 to 48% in 2010.
  - Government-owned banks’ median asset share increased by 5 percentage points between 2008 and 2010—the highest regional increase.
- Sub-Saharan Africa:
  - Median foreign-owned asset share rose from 34 percent in 1995 to 66 percent in 2008.
  - In 2010, the region had the highest median foreign bank share (73 percent) and lowest median government bank share (8 percent) among developing regions.
  - Expansion of Pan-African banks (South Africa, Nigeria, Morocco, Kenya, Togo) sustained foreign participation after Western retrenchment (Beck et al., 2014).
- South Asia:
  - Lowest share of foreign-owned banks and highest share of government-owned banks among regions.
  - Foreign bank participation rose from 1 percent in 1995 to 15 percent in 2008.
  - Government bank ownership declined from 72 percent in 1995 to 39 percent by 2010.
- East Asia and Pacific:
  - Median foreign-owned asset share rose from 8 percent in 1995 to almost 30 percent before the GFC.
  - Median government-owned asset share fell from 23 percent in 1995 to 13 percent by 2008.
  - Despite sudden stops during the crisis, global banks did not reverse brick-and-mortar operations to the same extent as other regions between 2008 and 2010; however, Figure 3.1 shows a substantial decline in foreign-bank asset share in 2012-2013 relative to 2010 (2012-2013 numbers come from a different source and cover fewer countries).

### 4. Bank ownership, performance, and competition
- General mechanism:
  - If foreign banks are more efficient, their entry can pressure domestic banks to reduce spreads and costs, improving overall efficiency.
- Empirical evidence:
  - At least fifty studies compare foreign and domestic bank performance with mixed results.
  - Results are sensitive to estimation methods (data envelopment, distribution free approaches, stochastic frontier analysis) and to performance measures (profitability ratios, cost and profit efficiency measures, stability indicators).

*Italicized source attribution: Content derived from wp1760 - References (PDF) as provided.*

### 2010. We cannot therefore reliably claim that there was a substantial retrenchment in foreign bank presence in East

### wp1760 - 2010. We cannot therefore reliably claim that there was a substantial retrenchment in foreign bank presence in East

### Evidence on foreign-owned vs domestic banks
- Regional and country-level studies focusing on high-income countries mostly suggest that domestic banks outperform foreign banks (Berger et al., 2000; Miller and Parkhe, 2002), in particular in the United States (DeYoung and Nolle, 1996; Hasan and Hunter, 1996; Chang et al., 1998; Peek et al., 1999).
- The majority of studies focusing on developing countries find that foreign-owned banks perform better than other types of banks in terms of cost and profit efficiency:
  - Eastern Europe and Central Asia: Bonin, Hasan and Wachtel, 2005; Fries and Taci, 2005; Grigorian and Manole, 2006; Havrylchyk and Jurzyk, 2011.
  - Country studies: Poland and Czech Republic (Weill, 2003; Matousek and Taci, 2004; Havrylchyk, 2006); Hungary (Hasan and Marton, 2003; Majnoni, Shankar and Varhegyi, 2003); Croatia (Jemric and Vujcic, 2002; Kraft et al., 2006); Turkey (Isik and Hassan, 2002).
  - Asia: Malaysia (Detragiache and Gupta, 2006; Matthews and Ismail, 2006), Thailand (Chantapong, 2005), Indonesia (Mulyaningsih, 2015), Pakistan (Bonaccorsi di Patti and Hardy, 2005).
- Findings are more ambiguous in China (Wu et al., 2007; Berger et al., 2009) and India (Bhattacharya et al. 1997; Sensarma, 2006), where foreign bank participation remains limited.
- Evidence from Latin America in the 1990s varies with the measure of efficiency employed (Barajas et al., 2000; Goldberg et al., 2000; Crystal et al., 2001; Berger et al., 2005).
- Evidence from Sub-Saharan Africa and MENA is scant, with Farazi et al. (2013) noted as one exception.

### Cross-country and large-sample findings on foreign ownership
- Large-sample studies find:
  - Foreign ownership is associated with higher interest margins and profitability, and lower overhead costs, in developing host countries (Demirgüç-Kunt and Huizinga, 1999; Micco et al., 2007).
  - The opposite is true in high-income countries (Claessens et al., 2001).
- Factors reconciling heterogeneous results:
  - Foreign banks are relatively more profitable when they are from a high-income country, when they share the same language and a similar regulatory and supervisory framework with the host country, when regulations in the host country are weak, and when the foreign bank is larger in size and market share (Claessens and Van Horen, 2012).

### Foreign entry, competition, and mode of entry
- Foreign ownership can affect domestic bank performance through competitive pressures; studies find a negative relation between foreign banking and the performance of domestic banks (Demirgüç-Kunt et al.,1998; Claessens et al., 2001; Bayraktar and Wang, 2004).
- Claessens and Lee (2003) highlight threshold effects: the competitiveness impact is stronger where foreign-owned banks have a higher market share.
- Models and empirical results on competition:
  - Spatial competition models show foreign bank presence can lower lending rates of local banks (Kaas, 2004; Park and Pennacchi, 2009).
  - Specific market characteristics may lead both foreign and domestic banks to charge higher margins (Delis et al., 2016).
  - Mode of entry matters:
    - Greenfield investments tend to increase the number of banks and decrease market power, leading to fiercer price competition.
    - Acquisitions can increase concentration and market power, lowering competition (Lehner and Schnitzer, 2008).
  - Asymmetric information models show competition depends on information distribution about incumbent customers (Van Tassel and Vishwasrao, 2007; Claeys and Hainz, 2014).
- Competition measured by H-statistic and other approaches:
  - Early cross-country evidence suggests a positive relation between foreign ownership and competition (Claessens and Laeven, 2004; Gelos and Roldos, 2004); Levy-Yeyati and Micco (2007) find opposite results for Latin America.
  - Controlling for banking system characteristics reconciles findings: positive link stronger when entering banks are more efficient and less risky and when host banking sector is less concentrated.
  - Positive effect of foreign bank penetration on competition is stronger with greenfield investments than with mergers and acquisitions (Jeon et al., 2011).
- Bank-level Lerner index evidence (Delis et al., 2016):
  - Uses bank-level data for 131 countries over 1997-2009.
  - Finds no evidence of a direct impact of bank ownership (foreign vs domestic) on bank-level market power.
  - Finds that greater foreign bank presence has a positive effect on overall levels of market power in the banking sector; authors attribute this to foreign entry occurring mainly through mergers and acquisitions rather than greenfield investments.

### Evidence on government (state) bank ownership
- Theoretical possibilities:
  - State ownership can spur competition if state-owned banks are more efficient and push private banks to lower prices.
  - In practice in developing countries, high agency costs in state-owned banks lead to operational inefficiencies and low intermediation quality, limiting competitive pressure.
- Relative performance of government-owned banks:
  - Regional studies often find government-owned banks perform worse than private domestic and foreign banks; Latin America shows strong differences (Micco et al., 2007), though Costa Rica is an exception (Levy Yeyati et al., 2007).
  - South and South-East Asia: government-owned banks underperform (Williams and Nguyen, 2005; Micco et al., 2007; Cornett et al., 2010); confirmed in Pakistan and China case studies (Bonaccorsi di Patti and Hardy, 2005; Berger et al., 2009).
  - India: opposite findings where government-owned banks represented 78 percent of banking system assets at the turn of the century (Sensarma, 2005).
  - MENA: state-owned banks have lower interest margins, higher cost ratios, larger NPLs, larger loan loss provisions, and lower profitability than private domestic non-listed banks (Farazi et al., 2013).
  - Central and Eastern Europe in the 1990s: government-owned banks less cost efficient than private banks but reversed in profit efficiency (Bonin et al. 2005a, 2005b; Yildirim and Philippatos, 2007).
  - Western Europe (1999 to 2004): government-owned banks exhibited lower profitability and cost efficiency than privately-owned banks despite lower costs of funding (Ianotta et al., 2007).
  - Germany case studies: government-owned banks have a profit and cost advantage; government-owned banks represented 42 percent of banking assets in 2000 (Altunbas et al. 2001; Chakravarty and Williams 2006).
- Large-sample studies:
  - Government-owned banks underperform in developing countries, consistent with political economy models and the development-mandate objective of government banks (Micco et al., 2007).
  - Mian (2003) finds government-owned banks had significantly lower return on loans and higher rates of loan default despite lower deposit costs in a large panel of emerging countries.
  - Micco et al. (2007) confirms differences in profitability, nonperforming loans and overhead costs are significant only in developing countries when including country fixed effects.
  - No cross-country study to date systematically analyzes how relative performance of government banks varies with home and host country conditions and bank characteristics.

### Bank privatization and efficiency effects
- Cross-country and regional evidence on privatization:
  - Studies in South East Asia and a panel of 22 developing countries suggest bank privatization raises bank profitability and efficiency over time to levels above pre-privatization performance, even when the acquirer is a foreign bank (Boubakri et al., 2005; Williams and Nguyen, 2005).
  - Country case studies of Nigeria, Pakistan and China corroborate these results (Beck et al., 2005; Bonaccorsi di Patti and Hardy, 2005; Berger et al., 2009).
  - Berger et al. (2005) proposes a methodology addressing misspecification issues and, applying it to Argentina data, confirms privatizations are associated with significant improvements in bank performance.
- Other evidence on market implications:
  - Case studies of two public banks in Hong Kong and Australia show privatizations had a negative impact on the stock market performance of private-sector competitors, consistent with the hypothesis that government bank ownership undermines competition (Otchere and Chan, 2003; Chen et al., 2005).
  - Cross-country analysis of abnormal returns reaches similar conclusions (Otchere, 2005).

*Italic: Source — wp1760 (PDF chapter/section).*

### 4. Bank ownership and banking sector stability

### 4. Bank ownership and banking sector stability

### The evidence on foreign bank ownership
- Foreign banks can be stabilizing or destabilizing depending on the nature of shocks: if shocks are domestic, foreign banks may stabilize because they have access to liquidity and capital from their parents and are typically more diversified than domestic banks; conversely, foreign banks can import shocks from abroad or increase domestic competition that may push domestic banks to riskier segments or out of business.
- Cross-country and panel evidence:
  - Cross-country evidence from 80 developed and developing countries over 1988-1995 suggests that greater foreign bank presence is associated with a lower probability of systemic banking crises in the host country (Demirguc-Kunt et al., 1998).
  - Subsequent research on a broader sample of 107 countries shows that official barriers to foreign bank entry are associated with measures of banking system fragility (Barth et al., 2004).
- Evidence on crisis-period behavior:
  - Bank-level studies find that in Eastern European countries (De Haas and Van Lelyveld, 2006), and in a broader sample of countries (De Haas and Van Lelyveld, 2010), credit growth by greenfield foreign banks is more resilient than that by (private) domestic banks during domestic crisis episodes.
  - In Latin America and Asia over 1989-2001, foreign banks’ lending and deposits rates are less volatile than those of domestic banks in times of crisis (Arena et al, 2007).
  - Case studies of Argentina, Brazil, and Mexico from 1994 to 1999 indicate that foreign banks do not pull back from host countries during economic problems and in Argentina and Mexico foreign banks had higher growth rates and lower volatility of lending than domestic banks during the crises of the mid to late 1990s (Peek and Rosengren, 2000a; Dages et al., 2000).
  - In Latin American countries (Argentina, Brazil, Chile, Mexico, Peru, and Venezuela) foreign banks showed more robust loan growth, a more aggressive response to asset deterioration, and greater ability to absorb losses than domestic banks during this period (Crystal et al., 2001, 2002).
  - Study of Malaysian banks around the Asian crisis: diversified foreign banks played a stabilizing role in Malaysia; banks not specialized in Asia continued to lend and had substantially higher profits and less non-performing loans than domestic banks and foreign banks mainly active in the region (Detragiache and Gupta, 2006).
- Evidence that foreign banks can amplify external shocks:
  - Peek and Rosengren (1997, 2000b) show transmission of shocks across borders through global banks using the collapse in Japanese stock prices as an exogenous shock to the U.S. banking sector; U.S. branches of Japanese banks reduced lending in the US in response.
  - Chava and Purnanandam (2011) and Schnabl (2012) use the 1998 Russian crisis as an exogenous shock and find effects on bank lending to borrowers in the United States and Peru, respectively.
- Evidence from the Global Financial Crisis (GFC):
  - Overall, foreign banks reduced their lending earlier and faster than domestic banks during the crisis (Aiyar, 2012; Fungáčová et al., 2013; de Haas and van Lelyveld, 2014; Claessens and van Horen, 2015; Choi et al., 2016), particularly within Eastern Europe (Mihaljek, 2011; de Haas et al., 2015).
  - Heterogeneous effects depend on country- and bank-level characteristics:
    - Transmission of shocks is stronger in host countries that are less regulated (Anginer et al., 2017), more financially open, and where there is more competition in the banking sector (Jeon et al., 2013).
    - Transmission is stronger for foreign-owned banks with lower levels of capital, profitability, retail funding, and less independent management (Popov and Udell, 2012; Jeon et al., 2013; de Haas and Van Lelyveld, 2014; de Haas et al., 2015; Anginer et al., 2017; Choi et al., 2016).
    - Transmission of shocks is stronger when banks entered host markets via greenfield investment rather than M&A (Jeon et al., 2013).
  - Research documenting internal cross-border capital allocation:
    - Cetorelli and Goldberg (2012a, 2012b, 2012c) and Jeon et al. (2013) provide evidence of an active cross-border internal capital market where global banks reallocate funds across branches and subsidiaries to buffer shocks to the parent bank balance sheet.
- Dataset note:
  - One dataset includes 45 multinational banks from 18 home countries with 194 subsidiaries across 46 countries.

### The evidence on government bank ownership
- Mixed and limited evidence on direct association with banking crises:
  - In a sample of 64 countries over 1980-1997, Caprio and Martinez Peria (2002) find that government bank ownership is positively associated with the likelihood of banking crises.
  - La Porta, Lopez de Silanes and Schleifer (2002) do not find significant results on this correlation.
  - Barth et al. (2004) find a positive bi-variate correlation between banking crises and government ownership, but the association does not survive after controlling for regulatory and supervisory environment features; Barth et al. (2004) argue that multi-collinearity induced by the close association between bank regulations/supervisory practices and government ownership produces insignificant coefficients on government ownership.
- More robust associations with other measures of financial instability:
  - Government-owned banks in developing countries have a higher fraction of non-performing loans than privately-owned banks (Micco et al., 2004; IADB 2005; Iannotta et al., 2007; Berger et al., 2009; Farazi et al., 2011).
  - Government-owned banks tend to display a higher likelihood of default as captured by a lower Z-score.
- Evidence on political economy and misallocation:
  - Government-owned banks can be subject to political manipulation and tend to increase lending to politically strategic sectors or regions near elections (Sapienza, 2004; Khwaja and Mian, 2005; Cole, 2009a,b; Carvalho, 2014).
  - Government-owned banks target firms with political ties to the detriment of other firms (Khwaja and Mian, 2005; Claessens et al., 2008; Carvalho, 2014).
- Evidence that government bank lending can be less procyclical and potentially beneficial in downturns:
  - Several studies show government bank lending is less procyclical than private bank lending in pre-GFC periods (Micco and Panizza, 2006; Foos, 2009; Brei and Schclarek, 2013; Onder and Ozyldirim, 2013; Bertay et al., 2015; Duprey, 2015).
  - Studies analyzing government-owned bank behavior during the GFC find supportive results in some contexts (Leony and Romeu, 2011; Cull and Martinez Peria, 2013; Coleman and Feler, 2015; De Haas et al., 2015; Chen et al., 2016; Choi et al., 2016; Allen et al., forthcoming), but evidence is mixed for Eastern Europe (Cull and Martinez Peria, 2013; De Haas et al., 2015; Allen et al., forthcoming).

*Source: wp1760 - 4. Bank ownership and banking sector stability (IMF working paper).*

### 5. Bank ownership and access to credit

### 5. Bank ownership and access to credit

### The evidence on foreign bank ownership
- Theoretical assumption: foreign banks are at an informational disadvantage relative to domestic banks for lending to local, informationally opaque borrowers because domestic banks possess superior "soft information" (knowledge of borrowers’ intangible traits and hard to quantify local business conditions). Unless foreign banks acquire local resources or expertise (for example by entering through mergers or acquisitions), foreign banks are assumed to rely more heavily on "hard information" contained in credit histories, collateral registries, and audited balance sheets and income statements.
- Theoretical prediction: foreign bank participation can be associated with reduced access to credit for informationally opaque firms (Detragiache et al., 2008 and Gormley, 2014). Foreign banks’ focus on lending to the largest and most profitable borrowers could be destabilizing for domestic banks and reduce overall credit levels, or credit access for some market segments.
- Cross-country empirical patterns:
  - Negative association between foreign bank presence and private credit to GDP and private credit growth rates in low-income countries where information barriers are more severe (Detragiache et al., 2008; Claessens and Van Horen, 2014).
  - Negative association between private credit and foreign bank presence found in countries where contract enforcement is costlier and access to credit information is limited (Claessens and Van Horen, 2014).
  - Prevalence, pricing, and conditions on SME loans tied to cross-country differences in institutional and legal environment (Beck et al., 2011).
- Entry context matters:
  - Negative relation between foreign bank presence and credit levels driven by non-random entry of foreign banks into banking markets that were in crisis or experienced large drops in credit levels prior to their entry (Cull and Martinez Peria, 2008). Foreign banks often acquired failed domestic banks and re-capitalized banking sectors, which could temporarily reduce overall credit levels.
- Evidence consistent with foreign banks favoring larger, less informationally opaque borrowers documented for Argentina, Mexico, Pakistan, India, and Central and Eastern Europe (Berger et al., 2001; Beck and Martinez Peria, 2010; Mian 2003, 2006; Berger et al. 2008; Gormley, 2010; Beck and Brown, 2015).
- Countervailing forces and adaptability:
  - Improvements in legal and regulatory environment can facilitate lending to small businesses using different lending technologies (Berger and Udell, 2006).
  - Foreign banks are developing new business models and technologies (credit scoring, standardized risk-rating tools, asset-based lending, factoring, fixed-asset lending, leasing) that facilitate arms-length lending to small firms (De la Torre et al., 2010).
  - Large foreign banks in four Latin American countries lent as much or more to SMEs as large domestic banks (Argentina, Chile, Colombia, Peru) (Clarke et al., 2005b).
  - Detailed credit registry data from Bolivia show foreign banks can serve the same clientele as domestic banks by requiring collateral, imposing shorter maturities, and basing pricing on credit ratings and collateral pledges (Beck et al., forthcoming).
- Competitive dynamics:
  - Competition from foreign banks for large, informationally transparent borrowers can induce some domestic banks to "go down-market", increasing lending to informationally opaque borrowers (Clarke et al., 2006).
  - In Eastern Europe, foreign bank presence was not associated with reductions in access to credit for SMEs, though foreign banks tended to lend more to larger, often multinational, businesses; many SMEs incorrectly anticipated rejection by foreign banks (De Haas et al., 2010; Giannetti and Ongena, 2012; Brown et al., 2011).
- Business-cycle adaptability of lending approaches:
  - Relationship lenders better able to relax credit constraints during downturns than banks relying on hard information lending approaches; they charge higher intermediation spreads in normal times to lend on more favorable terms during crises (Bolton et al., 2016).
  - Face-to-face interviews with CEOs of 400 banks across 21 countries in Eastern Europe and the Caucasus show firms in areas with more relationship lenders faced less severe financial constraints during the crisis (2008), but not during the credit boom (2005). Relationship lending especially eased constraints for young, small, non-exporters lacking tangible assets and other external finance sources (Beck et al., 2014).
  - Traditional depiction of foreign-owned banks as transaction lenders and domestic banks as relationship lenders does not hold uniformly; both relationship and transaction lenders exist among both ownership types.
  - Heterogeneity by foreign bank type: the ‘Big Six’ multinational European banks in Eastern Europe maintained lending growth rates comparable to domestic banks during recent crisis periods; non-‘Big Six’ foreign banks had higher loan growth prior to the global crisis but declined relative to others during the crisis (Bonin and Louie, forthcoming). Evidence consistent with some large foreign banks treating host countries as a "second home market."

### The evidence on government bank ownership
- Rationale and theoretical role: government-owned banks are motivated by correcting credit market failures where information asymmetries and weak contracting environments deter private banks from serving some market segments (e.g., small, informationally opaque borrowers).
- Cross-country empirical findings:
  - Higher share of government ownership in banking is negatively associated with the presence of bank branches and ATMs, and not significantly linked to indicators of prevalence of loans and deposit accounts (Beck et al., 2007).
  - Proxies for barriers to access and use of banking services increase with the level of government ownership in banking (and decrease with foreign bank participation).
  - Negative effects of bank concentration on widespread access to credit are more pronounced in countries with higher shares of state bank ownership (Beck et al., 2004).
- Political influence and allocation:
  - Political influences on lending decisions undermine state banks’ ability to deepen access to credit. Examples:
    - Government-owned banks in Italy charge lower interest rates to otherwise similar borrowers and lend more to politically affiliated firms in election years (Sapienza, 2004).
    - In India, government-owned banks more likely to maintain lending relationships with government-owned and rural firms rather than smaller, informationally opaque firms (Berger et al., 2008).
    - In China, government-owned firms and individuals with Communist Party affiliation have long received credit disproportionately to profitability (reviewed in Cull et al., 2015).
- Crisis-stabilizing role:
  - Theoretical models suggest state banks better able to provide loans during crises while private banks cut back lending and increase liquidity holdings because state banks are less profit-maximizing and less likely to suffer deposit withdrawals due to likelihood of recapitalization (Brei and Schclarek, 2015).
  - Empirical evidence: lending by state banks is substantially less pro-cyclical than lending by private banks in countries with good governance; state banks expand non-deposit liabilities more slowly during downturns and report more stable non-performing loan ratios over the business cycle (Bertay et al., 2015).
  - In countries with good governance, state banks may be effective in stabilizing credit flows during crises.

### Synthesis and implications
- Ownership changes since mid-1990s: foreign bank participation has increased while, until the recent GFC, government-owned bank presence declined across most regions.
- Empirical summary:
  - Foreign-owned banks tend to be more efficient than domestic banks in developing countries, promote competition in host country banking sectors, and help stabilize credit when host countries face idiosyncratic shocks.
  - Trade-offs: foreign-owned banks can transmit external shocks and might not always expand access to credit; impacts depend on context, including the level of competition when foreign banks entered and a country’s level of institutional development. Negative relations between foreign bank presence and credit levels in some studies may reflect foreign banks being brought in to recapitalize failing banking sectors.
  - Little evidence that government bank ownership provides substantial benefits relative to other ownership types for the banking sector, the real economy, or users of banking services, especially in developing countries. Some recent evidence indicates government-owned banks can help stabilize credit growth during crises, but overall they have negatively impacted competition and banking performance and have at best a mixed record in expanding access to credit.

*Source: IMF Working Paper — "5. Bank ownership and access to credit" (chapter from wp1760)*

### References

### wp1760 - References

### Major topical coverage in the references
- Foreign bank entry, foreign-owned banks, and international banking integration (e.g., Aiyar (2012); Anginer, Cerutti, & Pería (2017); Cetorelli & Goldberg (2012a, 2012b, 2012c); Claessens & Van Horen (2012, 2014, 2015); De Haas & Van Lelyveld (2006, 2010, 2014); Giannetti & Ongena (2012); Popov & Udell (2012)).
- Bank ownership and performance: foreign, state-owned, privatized, and domestic (e.g., Altunbas, Evans, & Molyneux (2001); Barth, Caprio Jr, & Levine (2001, 2004); Beck et al. (various); Berger et al. (2000, 2005); Micco, Panizza & Yanez (2007); La Porta, Lopez-de-Silanes & Shleifer (2002); Sapienza (2004)).
- Bank privatization, liberalization, and competition effects (e.g., Barajas, Steiner, & Salazar (2000); Bonaccordi di Patti & Hardy (2005); Megginson (2005); Martinez-Diaz (2009); Stein (2010)).
- Bank efficiency and productivity in transition and emerging markets (e.g., Bhattacharyya, Lovell, & Sahay (1997); Fries & Taci (2005); Weill (2003); Havrylchyk (2006); Kraft, Hofler, & Payne (2006)).
- Bank lending channels, credit access (SMEs and households), and relationship vs transaction lending (e.g., Ashcraft (2006); Beck, Demirgüç-Kunt, & Martinez Peria (2007, 2011); Berger & Udell (2006); Bolton, Freixas, & Gambacorta (2016); De la Torre, Pería, & Schmukler (2010)).
- Crisis transmission, internal capital markets, and liquidity allocation across multinational banks (e.g., Campello (2002); Cetorelli & Goldberg (2012b); Iannotta, Nocera, & Sironi (2007); Hale, Kapan, & Minoiu (2016); Schnabl (2012)).
- Political economy, governance, and connected lending influencing banking outcomes (e.g., Banerjee (1997); Khwaja & Mian (2005); Shleifer & Vishny (1994); Rajan & Zingales (2003); Epstien (2008, 2014)).

### Representative studies and focal questions (by study title wording)
- From financial crisis to great recession: The role of globalized banks (Aiyar, 2012).
- Foreign bank subsidiaries' default risk during the global crisis: what factors help insulate affiliates from their parents? (Anginer, Cerutti, & Pería, 2017).
- Bank ownership and efficiency (Altunbas, Evans, & Molyneux, 2001).
- The lending channel in Emerging economics: Are foreign banks different? (Arena, Reinhart & Vazquez, 2007).
- When arm's length is too far. Relationship banking over the business cycle (Beck et al., 2014).
- Bank privatization and performance: Empirical evidence from Nigeria (Beck, Cull, & Jerome, 2005).
- Bank financing for SMEs: Evidence across countries and bank ownership types (Beck, Demirgüç-Kunt, & Martinez Peria, 2011).
- Bank ownership and credit over the business cycle: Is lending by state banks less procyclical? (Bertay, Demirgüç-Kunt, & Huizinga, 2015).
- Follow the money: Quantifying domestic effects of foreign bank shocks in the great recession (Cetorelli & Goldberg, 2012a).
- Internal capital markets and lending by multinational bank subsidiaries (De Haas & Van Lelyveld, 2010).
- Multinational banks and the global financial crisis: Weathering the perfect storm? (De Haas & Van Lelyveld, 2014).
- The international transmission of bank liquidity shocks: Evidence from an emerging market (Schnabl, 2012).
- The dynamics of foreign bank ownership: Evidence from Hungary (Majnoni, Shankar, & Varhegyi, 2003).
- Has foreign bank entry led to sounder banks in Latin America? (Crystal, Dages, & Goldberg, 2002).

### Key numeric and data notes extracted from the figures and appendix material
- "2012-2013 data are available for 80% of the sample." (Figure 1.1 notes)
- Figure 1.2: "The map reflects 2010 data, except for the MENA region where 2008 figures are used instead." (Figure 1.2 notes)
- Ownership definition applied in figures and appendix: "Ownership is defined as the asset share of banks that are more than 50% controlled by the foreign owners, except for 1995 where ownership reflects the actual share owned by foreign banks." (Figure 2.1 note)
- Comparable government-ownership definition: "Ownership is defined as the asset share of banks that are more than 50% controlled by the government, except for 1995 where ownership reflects the actual share owned by the government." (Figure 2.2 note)
- Time-series percentage sequences shown in the figures (preserved verbatim as presented):
  - Sequence block (from figure area): "16% 40% 50% 49% 43% 15% 28% 43% 42% 36% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 1995 1999-2001 2008 2008-2010 2012-2013 1995 1999-2001 2008 2008-2010 Developing countriesHigh-income countries" (as presented).
  - Sequence block (from figure area): "40% 23% 17% 18% 36% 15% 10% 12% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 1995 1999-2001 2008 2010 1995 1999-2001 2008 2010 Developing countriesHigh-income countries" (as presented).
  - Regional median blocks (verbatim selections from figure area): 
    - "8% 5% 6% 1% 34% 5% 16% 28% 38% 10% 67% 12% 28% 60% 41% 15% 66% 27% 28% 48% 39% 14% 73% 25% 12% 63% 30% 3% 51% 18% East Asia & Pacific Europe & Central Asia Latin America & Caribbean South AsiaSub-Saharan Africa High income (all regions) 1995 1999-2001 2008 2010 2012-2013" (as presented).
    - "23% 48% 43% 72% 22% 30% 6% 15% 14% 65% 6% 14% 13% 9% 12% 45% 8% 4% 13% 14% 13% 39% 8% 5% East Asia & Pacific Europe & Central Asia Latin America & Caribbean South AsiaSub-Saharan Africa High income (all regions) 1995 1999-2001 2008 2010" (as presented).
- Table A.1 notes (exact wording preserved):
  - "Notes: Income groups follow the 2010 World Bank analytical classification. Government (respectively foreign) ownership is defined as the asset share of banks that are more than 50% controlled by the government (respectively foreign owners), except for 1995 where ownership reflects the actual share owned by the government (respectively foreign owners)."
  - "Sources: Micco, Panizza and Yanez (2007) for 1995 data, Regulation and Supervision surveys (World Bank) for 1999-2010 data, Claessens and Van Horen (2015) for 2012-2013 data."

### Implicit empirical building blocks documented across the references
- Cross-country and region-specific empirical evidence on foreign and state bank shares (1995, 1999-2001, 2008, 2010, 2012-2013 panels are repeatedly used).
- Use of bank-level performance metrics (efficiency, profitability, lending behavior), country-level regulatory and supervision surveys, and panel/balanced-sample estimation strategies (explicitly referenced in figure notes and multiple working papers and journal articles).
- Recurrent focus on crisis episodes (2008–2009 global financial crisis, regional crises, and the Great Recession) and their interaction with bank ownership structure.

_Italic: References and figures as listed in wp1760 - References (source PDF: wp1760 - References)._

---


_Source: https://www.imf.org/-/media/files/publications/wp/2017/wp1760.pdf_