alexander popov european central bank gregory f. udell indiana university
DESCRIPTION
Cross-border Banking and the International Transmission of Financial Distress During the Crisis of 2007-2008. Alexander Popov European Central Bank Gregory F. Udell Indiana University. On the Credit Crunch in Central and Eastern Europe - PowerPoint PPT PresentationTRANSCRIPT
![Page 1: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/1.jpg)
Cross-border Banking and the International Transmission of Financial Distress During the
Crisis of 2007-2008
Alexander PopovEuropean Central Bank
Gregory F. UdellIndiana University
![Page 2: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/2.jpg)
On the Credit Crunch in Central and Eastern Europe
“There is no credit crunch in Europe and the IMF has been too pessimistic in its growth forecasts for the region.”
Jean-Pierre Landau, deputy governor of Banque de France,
8 April 2008
![Page 3: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/3.jpg)
On the Transmission of Financial Distress by Foreign Banks in Central and
Eastern Europe
“[…] foreign banks have so far exerted a stabilizing influence, as witnessed by the contrast in gradual slowdown in credit in the Baltics and the much sharper contraction in Kazakhstan. ”
Eric Berglof, EBRD Chief Economist, 19 September 2008
![Page 4: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/4.jpg)
Motivation
• Question 1: Was there a credit crunch in central and eastern Europe in the early stages of the crisis?
– We focus on period between August 2007 and September 2008
– Look at one particular channel – bank lending to SMEs
• Question 2: If yes, were foreign banks a stabilizing influence?
- Or, were foreign banks a channel through which this crisis was propagated?
• Question 3: Can Q1 and Q2 can be answered in a satisfactory fashion?
![Page 5: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/5.jpg)
Our Contribution
• We are only paper that simultaneously:
1. Analyses international transmission of the effects of bank financial distress and
2. Accounts for changes in demand and
3. Accounts for contamination due to changing composition of firms demanding credit
![Page 6: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/6.jpg)
The Literature on Cross Border Bank Lending
• Evidence is ambiguous
– Some studies find increased access
• More credit (Clarke, Cull and Peria 2006)
• Higher sales (Giannetti and Ongena 2009)
• Lower rates (Ongena and Popov 2009)
– Some studies find foreign banks cherry pick
• Berger, Klapper and Udell (2001)
• Mian (2006)
• Gormley (2009)
![Page 7: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/7.jpg)
The Literature on the Credit Crunch
• Historical Crises:
– US: e.g., Bernanke and Lown (1991), Berger and Udell (1994), Hancock and Wilcox (1998)
– Japan: e.g., Peek and Rosengren (1997), Woo (1999), Kang and Stulz (2000), Hayashi and Prescott (2002), Watanabe (2006), Taketa and Udell (2007)
– Other Crises: e.g. Bae, Kang and Lim (2002), Jiangli, Unal and Yom (2009), Park, Shin and Udell (2009), Chava and Purnanadam (JFE 2009), Khwaja and Mian (AER 2008)
• Current Crisis:
– Ivashina and Scharfstein (2009), Puri, Rocholl, and Steffen (2009), de Haas and van Horen (2009), Jimenez, Ongena, Peydro, and Saurina (2009)
![Page 8: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/8.jpg)
Key Challenges in Analyzing Credit Crunches:(The Problems Related to Question 3)
• Credit crunches notoriously difficult to identify
• Simultaneity issue at macro level
– Supply and demand for credit can both be affected – and usually are
• Simultaneity issue at micro level
– Demand at worst hit banks can be relatively more affected
– Composition of applicants and non-applicants may be different for different banks
![Page 9: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/9.jpg)
Identifying Demand vs. Supply: Micro Data
• Approach 1. Select a setting where demand didn't change
– Peek and Rosengren (AER 1997) – Japanese banks and US households after Nikkei collapse
– Domestic event - no change in US firms‘ demand
– Not applicable for 2007-2008 – global recession
![Page 10: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/10.jpg)
Identifying Demand vs. Supply: Micro Data • Approach 2. Use application data, make sure
demand changes throughout
- Puri, Rocholl, and Steffen (2009) - US banks and German firms after August 2007
- However, key problem
- Doesn‘t account for composition of firms that self-select out of the application process because they get discouraged
• Approach 3. Our data allows to overcome this problem by using application data that controls for discouraged applicants
![Page 11: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/11.jpg)
Empirical Approach• Calculate financial distress by bank, map into incidence of credit
constraint to identify transmission
- Adjust for discouraged applicants
- Control for common macro factors, common industry factors, local macro factors, and account for soft information
• Compare transmission by foreign and domestic banks
• Use industry characteristics to study differential effect
• Hypothesis 1: Distressed banks have higher probability of rejecting a loan application by an identical firm
• Hypothesis 2: For the same level of distress, foreign banks have higher probability of rejecting a loan application by an identical firm
![Page 12: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/12.jpg)
Data on Firms
• 2005 and 2008 Business Environment and Enterprise Performance Survey (BEEPS) by the World Bank and the EBRD.
• 2008 wave interviewed in April 2008, asked about experience with banks during “fiscal year 2007”
– For all countries, firms extend fiscal year to end of March
– 1.5 non-crisis and 2.5 crisis quarters (bias goes against finding anything)
– No match to specific bank
![Page 13: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/13.jpg)
Data on Firms (cont.)
• 4,421 firms from 14 central and eastern European countries
– Albania, Bulgaria, Croatia, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Macedonia, Montenegro, Poland, Romania, Slovakia, and Slovenia
• 1,266 localities
• Firm level characteristics
– Size (74% <100 workers, 3% >500 workers), Age
– Ownership (private/state/foreign), competition, exporter, subsidized, audited
– 18 Industries
![Page 14: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/14.jpg)
Data on Financial Distress• Balance sheet data from Bankscope for 2005-2008
1) Equity capital / total assets ratio
2) Tier 1 capital ratio
3) Gain (loss) on financial assets
Also:
– Mortgage lending, deposits, MM funding, profits, securities, problem loans, etc.
• 141 banks present in the 1,266 localities
– 27 domestic, 117 subsidiaries and branches of foreign banks
– 291 localities with more than 1 firms, rest matched manually to closest locality
![Page 15: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/15.jpg)
Data on Financial Distress (cont.)
• Don’t have direct matching between bank and borrower
• Calculate a locality-specific measure of “financial distress” by weighting balance sheet data for all banks present
– 1) equally
– 2) by number of branches
For foreign-owned, use the balance sheet data on the mother (group)
![Page 16: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/16.jpg)
The Ideal Data to Study the Credit Crunch• Application data including discouraged applicants
• Firm characteristics• Bank characteristics• Loan characteristics• Identification of firm’s bank• Firm’s banking relationship• Panel data• Cross-country data• Third party mercantile data• Lending technology deployed, e.g.:
- Financial statement lending- Relationship lending- Real estate-based lending- Equipment lending- Leasing- Factoring- Asset-based lending- Trade credit
No one has all this!
![Page 17: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/17.jpg)
![Page 18: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/18.jpg)
Fraction Foreign Bank Ownership (2008)
0.00
0.20
0.40
0.60
0.80
1.00
1.20
Albania
Bulgar
ia
Croat
ia
Czech
Rep
ublic
Estonia
Hunga
ry
Latvi
a
Lithu
ania
Mac
edon
ia
Mon
tene
gro
Poland
Roman
ia
Slovak
ia
Sloven
ia
![Page 19: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/19.jpg)
Key Survey Questions• K16: “Did the establishment apply for any loans or lines
of credit in the fiscal year 2007?”
• If “No” to K16, go to K17: “What was the main reason?”
– If “No need for a loan”, classify firm as not desiring credit
– If “Interest rates too high” or “Collateral requirements too strict” or “Did not think it would be approved”, classify firm as constrained
• If “Yes” to K16, go to K18a: “Was any loan or line of credit rejected?”
– If “Yes”, classify firm as constrained
• Grouping of rejected and discouraged firms standard
– Cox and Japelli (JMCB 1993)
• Accounting for discouraged firms crucial in the CEE context
– Up to 2/3 of constrained firms are discouraged – Brown, Ongena, Popov, and Yesin (2009)
![Page 20: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/20.jpg)
![Page 21: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/21.jpg)
Basic Empirical Model
• Express probability of constraint as a two-equation
Y* = f(bank locality-specific distress, firm characteristics, other controls)
– Y* = 1 if the firms is constrained
• Estimate probability firm desires credit and employ a Heckman selection procedure
Prob(Desire Credit) = f(W)– where W contains a vector of firm-specific characteristics and locality-
specific bank distress characteritics
– Probit equation contains at least one more variable than main model
![Page 22: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/22.jpg)
International Transmission of Financial Distress
• Add international dimension to basic model
• Two approaches
1. Look at foreign dominated markets – i.e., repeat tests on just foreign dominated markets where 2/3 of branches are foreign
2. Examine foreign effect by interacting Foreign bank variable with Finance
- where Foreign is the share of foreign branches
![Page 23: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/23.jpg)
(Affected = Tier 1 capital decreased)
Rejectionsincreased
![Page 24: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/24.jpg)
Key Results I
• Was there a credit crunch related to bank distress?
- Evidence that the probability of being credit constrained affected by Tier 1 capital ratio
- Also interesting:
- Small firms and unaudited more constrained
![Page 25: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/25.jpg)
Small firms very vulnerable!
Also, transparency important
![Page 26: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/26.jpg)
Only for Tier 1 related financial distress affects financial constraints
![Page 27: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/27.jpg)
Still only for Tier 1 related financial distress affects financial constraints, although Equity negative
![Page 28: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/28.jpg)
Pooling firms applying in both periods also controls
for demand
![Page 29: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/29.jpg)
Key Results II
• Was there cross-border transmission of the credit crunch?
- Foreign-dominated markets: Evidence of bank distress affecting credit stronger, i.e., more robust to alternative measures of bank distress
- Interaction between distress and foreign: Some evidence that effect of Tier 1 capital and leverage exacerbated by foreign bank presence
- i.e., some evidence that foreign banks contracted more
![Page 30: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/30.jpg)
Foreign-bank domination matters
![Page 31: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/31.jpg)
Economic Significance
• A two standard deviation decrease in equity capital, Tier 1 capital or losses on financial assets leads respectively to:
– 30% increase in rejection rate
– 55% increase in rejection rate
– 32% increase in rejection rate
![Page 32: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/32.jpg)
Foreign effect almost always negative and often significant
![Page 33: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/33.jpg)
Other Results
• Some evidence that the interaction of foreign banks presence and bank distress affects opaque firms more
- Firms in opaque industries more likely to have loan applications rejected
- i.e., firms in most (Rajan/Zingales) opaque industries based on
a. Access to external finance
b. Asset tangibility
c. Capital intensity
![Page 34: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/34.jpg)
Opaque firms clearly hurt most
![Page 35: Alexander Popov European Central Bank Gregory F. Udell Indiana University](https://reader035.vdocument.in/reader035/viewer/2022062518/56814039550346895daba849/html5/thumbnails/35.jpg)
Conclusion
• Firms in localities dominated by distressed banks have higher probability of being rejected
– After accounting for self-selection
– After eliminating common macro, local, and sector unobservables
– Strongest evidence for Tier 1 capital ratio
• Foreign banks transmit to the real sector more of the same financial shock than domestic banks
• Transmission stronger when more opaque firms and firms with less tangible assets involved