Title: Credibility of Non-Insurance and Governance as Determinants of Market Discipline and Risk-taking in Banking
1Credibility of Non-Insurance and Governance as
Determinants of Market Discipline and Risk-taking
in Banking
- Penny Angkinand
- University of Illinois, Springfield
-
- Clas Wihlborg
- Copenhagen Business School
- The 2006 CFR Fall Workshop, October 25th
2SUMMARY
- The effectiveness of market discipline on
risk-taking depends on - (i) the existence and credibility of
non-insurance of depositor and creditor groups,
and - (ii) governance structures of banks as
reflected in managerial incentives relative to
shareholders and creditors. - Empirical analysis using bank level data for a
sample of industrialized and emerging market
countries during 1995-2005. - Proxies for governance variables (bank-specific
and country level) government ownership, foreign
ownership, concentration of ownership,
shareholder rights, and private monitoring.
3OUTLINE
- The impact of deposit insurance on risk taking
and market discipline the role of Credibility of
Non- Insurance - The effect of governance characteristics on the
relationship between deposit insurance and risk
taking. - Hypotheses
- emphasize the hypothesis with respect to the
quality of bank governance - Empirical Evidence
- emphasize the measures of banks risk taking,
deposit insurance coverage, and proxies of bank
governance
4Starting Points
- Due to fear of bank failures and contagion,
supervisory authorities and governments in all
countries offer some degrees of explicit or
implicit deposit insurance. As a result market
discipline can be weak. - Regulatory and supervisory structures for banks
may not substitute effectively for market
discipline. Excess risk-taking will take place.
5Deposit Insurance and Bank Risk Taking
- Empirical evidence in the literature
- The existence and higher coverage of explicit
deposit insurance schemes increase risk taking
and banking crises. - Implicit insurance is sometimes captured by
proxies of institutional characteristics, legal
system, and strictly regulated environment. - Demirgüç-Kunt and Detragiache (1997, 2002) and
Hutchison and McDill (1999), Barth et al (2004),
Hovakimiam et al (2003), Demirgüç-Kunt and
Huizinga (2004), Nier and Baumann (2006). - Explicit deposit insurance reduces or has no
impact on risk taking. - Eichengreen and Arteta (2002), Hoggarth et al
(2005), Gonzales (2005), Gropp and Vesala (2004)
6Credibility of Non-Insurance and Risk Taking
- Angkinand and Wihlborg (AW, 2005), linking
explicit coverage and implicit protection,
hypothesize and estimate a U-shaped relationship
between explicit DI coverage and banks risk
taking. - The absence of explicit guarantees leads to
strong expectations that governments and
regulators will bail out all creditors in times
of financial distress. Thus, non-insurance of all
creditors is not credible. - It can be expected that the credibility of
non-insurance increases as explicit deposit
insurance coverage expands. - ENABLE CREDIBLE NON-INSURANCE OF SOME CREDITOR
GROUPS AND SHAREHOLDERS Market Discipline
7Figure 1a. Extent of Credible Non-insurance
8Hypotheses (from AW 2005)
- Hypothesis 1
- The relationship between risk-taking incentives
and explicit deposit insurance coverage is likely
to be U-shaped such that (excess) risk-taking is
minimized at a positive but partial deposit
insurance coverage.
9Credibility of Non-Insurance and Risk Taking
- Effective market discipline depends on three
factors - the coverage of explicit deposit insurance
schemes - the credibility of non-insurance of those not
covered by explicit schemes - the relationship between the coverage of explicit
insurance and the credibility of non-insurance,
which depends on - institutional and supervisory environment
- bank governance
10Factors Affecting the Credibility of Non-Insurance
- The institutional and supervisory characteristics
- Powers of Supervisors, Powers and Procedures for
Prompt Corrective Action, Rule of Law, and
Corruption (analyzed in AW, 2005) - Bank corporate governance
11Hypotheses (from AW 2005)
- Hypothesis 2
- Strengthened institutional environment and
supervision enhance the market discipline and
therefore reduce risk taking. This reduction in
risk-taking is relatively large at low levels of
explicit coverage of deposit insurance schemes.
12This Paper
- We extend the argument developed in AW (2005)
that the relationship between risk taking and
explicit DI coverage is likely to be U-shaped by
using bank-specific data. - Focus on how the governance of banks affects the
disciplinary effect of partial deposit insurance
system.
13Corporate Governance and Bank Risk Taking
- Good governance managers maximize
shareholders wealth. - Governance is influenced by the ownership
structures of banks and the explicit or implicit
contractual relationship between shareholders and
managers. - Ownership of banks and risk taking
- State Ownership
- Foreign ownership
- Bank concentration ( shares of three largest
shareholders)
14Credibility of Non-Insurance, Corporate
Governance, and Risk Taking
- How quality of governance in banks affects the
relationship between explicit DI coverage and
risk taking. - at low and high levels of the DI coverage,
shareholders prefer excessive risks. - at an intermediate range of the DI coverage where
market discipline is effective, risk taking is
reduced. - Therefore, we expect that a higher quality of
bank governance leads to a more pronounced
U-shape relationship.
15Hypothesis With Respect To Quality of Bank
Governance
- Hypothesis 3
- The relationship between explicit deposit
insurance coverage and risk-taking is described
by a flatter curve for banks with relatively low
quality of governance from shareholders point of
view.
16Figure 1b. Impact of Bank Governance
17Model Specification
We estimate the following model specification for
risk-taking in bank j in country i in period t
EC Explicit Deposit Insurance Coverage
The effect of CorpGov on the relationship between
explicit DI coverage and risk taking EC2i,t ?
CorpGovj,i,t
18Data
- The data is an unbalance panel covering a total
of 518 banks in 53 countries of which 16 are
industrial and 37 emerging market countries for
the period 1995-2005.
19Proxies for Risk Taking
- NPL/Capital (Natural Logarithm)
- Adjusted NPL/Total Capital
- Standard Deviation of NPL/Average of Capital
- Bankscope database
- Proxies for the bank willingness to expose the
bank to the possibility that loan losses exceed
equity capital. - Adjusted an adjustment factor
20Additional Proxies for Risk Taking
- NPL/Total Assets
- Adjusted NPL/Total Assets
- Risk-Weighted Assets/Total Assets
- Standard Deviation of Capital/Average of Capital
- Volatility of Stock Price
- Estimation Methodology
- Random Effects Model
- OLS (when standard deviation of NPL or capital is
used.)
21Deposit Insurance Coverage
- The interval variable of the ratio of deposit
insurance coverage per deposit per capita
(covdep) - (authors calculation based on the data from
Demirgüç-Kunt et al., 2005.) - Covdepint 0 if there is no explicit deposit
insurance - 1 if 0 lt covdep lt 1
- 2 if 1 covdep lt 2
- 3 if 2 covdep lt 3
- 4 if 3 covdep lt 5
- 5 if 5 covdep lt 10
- 6 if 10 covdep lt 15
- 7 if 15 covdep lt 20
- 8 if 20 covdep lt 50
- 9 if covdep 50
- 10 for the full coverage.
22Deposit Insurance Coverage
- Additional variables used for the robustness
check - The natural logarithm of (1covdep)
- The interval variable of the ratio of deposit
insurance coverage per GDP per capita (covgdp)
23Proxies for Governance
- 1. Bank-Specific Governance or Ownership
Variables (Reuters) - Foreign-Owned
- Government-Owned
- Three Largest Shareholders (Bank Concentration)
- Institution-Owned
- 2. Country-Level Governance Variables
- Private Monitoring (Barth et al, 2004)
- Shareholder and Creditor Rights (La Porta et al,
1998)
24Macro and Bank-Specific Control Variables
- Bank-specific control variables
- Net Interest Margin/Total Assets
- Capital/Total Assets
- Cost/Income
- Total deposit of each bank/total deposits of all
banks in a country - Macroeconomic variables
- Real GDP Growth Rate
- Inflation
- Real Interest Rate
25Empirical Result I Proxies of Risk Taking
- First, compare regressions using alternative
proxies for risk taking (Table 3) - Focus on ln(NPL/Cap) and StdNPL/AvgCapita
26Empirical Result II Credibility of
Non-Insurance and Governance
- Include bank-specific and country-specific
corporate governance variables (table 4). - Add the interaction terms between corporate
governance variables and insurance coverage
(Covdepint) (table 5).
27Table 5
Bank- and country-specific control variables are
included, but not reported
28Empirical Result II Credibility of
Non-Insurance and Governance
- Evidence of the U-shaped relationship between
risk taking and deposit insurance coverage - The coefficient of Covdep is negative, and the
coefficient of (Covdep)2 is positive, reflecting
a U-shaped relationship. - reflecting that the non-insurance of banks
creditors is not credible - less robust when using the bank specific data
- less significance for emerging market economies.
Possibly explained by the significance of other
variables capturing Too Big To Fail.
29Empirical Result II Credibility of
Non-Insurance and Governance
- The County-Level Governance Variables (from Table
4) - Shareholder rights, creditor rights, and private
monitoring have negative effects on risk taking. - Shareholder rights are significant consistently.
30Empirical Result II Credibility of
Non-Insurance and Governance
- Bank-Specific Governance Variables
- The coefficients of individual variable and its
interaction with deposit insurance coverage are
significant, indicating that external and
internal factors influence market discipline
interactively as predicted (Table 5). - Hypothesis 3 risk taking is expected to be
higher at very low and very high levels of DI
coverage (where the coverage is not at optimal)
in banks with relatively high quality of
governance.
31Empirical Result II Credibility of
Non-Insurance and Governance
- Supporting the Hypothesis 3?
- Compare the signs for the squared covdepint and
its interaction term with governance variable.
For Low quality of governance, the neg. sign of
interaction term a flatter U-shaped, supporting
the hypothesis. - Draw figures describing the relationship between
risk taking and DI coverage for different levels
of ownership.
32The Effect of Foreign Ownership
- Individual effect Foreign ownership reduces risk
taking. - Interactive effect a more pronounced U-shaped
relationship at a high level of foreign
ownership, supporting the Hypothesis 3 if foreign
ownership represents higher quality of
governance. - The impact of foreign ownership is strong in
countries with relatively low explicit DI
coverage/lack of credibility of non-insurance.
33Foreign Ownership Affecting the Relationship
between Risk Taking and DI Coverage
34The Effect of Ownership Concentration
- Individual Effect ambiguous
- Interactive effect a more pronounced U-shaped
relationship at a high level of foreign
ownership, supporting the Hypothesis 3, for a
sample of industrial countries.
35Bank Concentration Affecting the Relationship
between Risk Taking and DI Coverage
36The Effect of State Ownership
- Individual effect State ownership increases risk
taking. - Interactive effect The effect of state ownership
in increasing risk taking is smaller where
deposit insurance coverage is high (indicated by
a negative sign of the interaction term). - The Hypothesis 3 is confirmed if government
ownership is associated with low quality of
governance.
37Sensitivity Analysis
Two Stage Least Square, assuming Capital/TA or
Cost/Income is endogeneous.
38CONCLUSION
- There is the U-shaped relationship between risk
taking and deposit insurance, reflecting that
extensive non-insurance of banks creditors is
not credible. This relationship is more robust
for industrial countries. - For the impact of ownership variables, government
ownership increases risk-taking, foreign
ownership reduces risk-taking, ownership
concentration is ambiguous and has opposite
effects in industrial and emerging market
countries. - The impact of bank governance on risk taking
seems to be strong in countries with low DI
coverage or high implicit coverage caused by lack
of credibility of non insurance.