Bank Performance, Efficiency and Ownership in Transition Countries

Bank Performance, Efficiency and Ownership in Transition Countries

• John P. Bonin

• Wesleyan University

•  Iftekhar Hasan

• Rensselaer Polytechnic Institute

•  Paul Wachtel

• Stern School of Business,

• New York University

Banking Developments in Transition Countries

• Rapid evolution from planned economy era

• Fall in government ownership

• Greenfield banks

• Foreign participation

• High compared to other emerging markets
• Dramatic increase in late 1990s

Hypotheses explored

• Banking sectors are becoming more efficient

• Spillover effects of foreign participation

• Foreign-owned banks perform better

• Ownership matters

• Government vs. Private
• Foreign vs. Domestic

Related literature

• On efficiency improvements in transition banking

• Buch (1997) and (2000), Fries and Taci (2002), Fries, Neven and Seabright (2002), Drakos (2002), Claessens, Demirguc-Kunt, and Huizinga (2001)

• On importance of ownership

• IMF(2002), Nikiel and Opiela (2002), Hasan and Marton (2003)

Data

• Bankscope bank data

• 11 advanced transition countries (Eastern Europe and the Baltics)
• Financial statements 1996-2000
• Ownership information 1999
• Exclude non-bank intermediaries

• Total of 830 bank-year observations with financial data and ownership information

• Country coverage
• Poland, Croatia, and Hungary ~ 45%
• Baltics ~ 13%

Ownership Characteristics (of bank observations)

• Majority Foreign 59%

• Less than one-half in Croatia, Slovenia and Latvia

• Majority government 10%

• Largest in Slovakia 14%

• Majority private domestic 31%

• Only Croatia, Slovenia and Latvia above one-half

• Participation of international institutional investor 10%

• Bulgaria, Estonia, Romania above 20%

Balance Sheet Characteristics

Efficiency Estimation

• Stochastic Frontier analysis

• Profit and Cost efficiency functions

• Standard translog specification

• Efficiency measures

• Raw efficiency – distance from frontier
• Relative efficiency – distance relative to mean for all same country bank observations

Performance compared to overall mean

Performance Measures

Regression analysis of Performance measures

• Independent variables

• Log of asset in constant $
• Fixed effects for year
• Dummy variables for ownership characteristics

• Additional robustness tests

• Balance sheet ratios effects on ROA Ratios to assets of loans, deposits, non-interest expenditures
• Real GDP growth

Explanations of Performance

• ROA declines over time – banking becoming more competitive

• Efficiency improves after 1998

• Better bank performance in countries with higher growth

• Larger banks generally less efficient

• Banks with larger deposit base (retail banks) have lower ROA

Ownership effects on performance

• Govt. banks less efficient than private

• Foreign-owned banks more efficient than other private banks in same country

• Presence of international institutional investors (most frequently EBRD) has important impact

• Higher ROA
• More profit efficient but not more cost efficient

Why do international institutional investors matter?

• Cherry-picking

• They successfully choose the most profitable banks for their investment portfolios

• Technology and Knowledge Transfer

• Transfers facilitate the development of banks

• Signaling or Screening

• Investments provide information about quality of banks
• Imprimatur of ‘official’ investors attracts customers

Size and year effects (Table 6, cols. 1 & 2)

Ownership Effects (Table 6, columns (1)-(3))

Additional Tests (Table 7, columns (1),(3) and (6))

Conclusions

• Increased foreign participation leads to more efficient and competitive banking sectors in advanced transition countries

• Private banks are more efficient than govt. owned banks

• Foreign owned banks are more efficient than private banks

• International institutional investors ‘cherry-pick’

Things to do

• Differentiate foreign greenfield operations from foreign ownership of formerly state owned banks

• Additional dummies for country clusters, e.g., Baltics, Northern-tier (Czech R., Hungary, and Poland)

• Relative to maximum efficiency in country vs. average efficiency

• Random effects estimation