## Financial Supervision Unification

## Source details

**Canonical URL:** [Financial Supervision Unification](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11261.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11261.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11261.pdf.json)

---

### I. Introduction and Objectives
- Paper objective:
  - Empirically test impact on economic resilience of:
    - changes in supervisory architecture (often toward unification), and
    - supervisory governance improvements.
  - Review proposals to improve supervisory effectiveness and offer a proposal: conducting supervision through two separate agencies (one for macro- and one for micro-prudential supervision) to introduce checks and balances.
- Main empirical findings:
  - (i) the two types of modifications introduced in supervision (unification and better governance arrangements) are negatively associated with economic resilience;
  - (ii) the quality of public sector governance and the degree of financial liberalization are negatively associated with economic resilience in this crisis, with associations even stronger than those of the supervisory features analyzed;
  - (iii) the degree of involvement of the central bank in supervision did not have any significant impact on resilience.

### II. Conceptual Framework and Indices
- Four groups of initiatives to improve supervisory effectiveness:
  - Basel Core Principles (BCP) and related codes (IOSCO, IAIS); major BCP revision in 2006.
  - Revisions of supervisory architecture (trend toward unification; FSA UK example, 1997).
  - Principles of supervisory governance organized around four pillars: independence, accountability, transparency and integrity.
  - Reliance on market discipline (e.g., subordinated debt to reduce moral hazard).
- Measurement constructs:
  - Financial Supervision Herfindahl Hirschman (FSHH) Index:
    - FSHH = Σ i=1..n s_i^2 where s_i is share of supervisory power of authority i.
    - Assumptions: three sectors (banking, securities, insurance) each of equal importance; supervisory power equally divided among authorities in each sector where more than one authority exists.
  - Central Bank as Financial Supervisor (CBFS) / CBSS Index:
    - CBFS/CBSS is the share of the central bank in supervisory responsibilities (range 0 to 1).
  - Supervisory governance ratings:
    - Ratings for independence and accountability: “2” if legal framework satisfies criteria, “1” partial compliance, “0” noncompliance, “-1” for practices undermining independence/accountability; summed and normalized between 0 and 1.

### III. Stylized Patterns (2007 vs. 2009)
- FSHH:
  - before crisis (2007) EU had greater consolidation than industrial countries and Europe;
  - consolidation continued in advanced countries during crisis (2009) while sample-wide slight reduction observed.
- CBFS / CBSS:
  - before crisis (2007) advanced countries averaged lower central bank involvement;
  - during crisis (2009) central bank involvement increased in advanced, European and EU countries (a “Great Reversal”).
- Governance ratings:
  - governance quality highest in EU, Europe, industrial countries pre-crisis (2007); governance quality increased further in 2009.

### IV. Empirical Strategy and Controls
- Dependent variable:
  - average real output growth in years 2008–09.
- Key explanatory variables dated 2007:
  - FSHHI 2007, CBSS 2007 (CBFS index), GOVRATING07.
- Macro controls (dated 1996–2006 or 2004–2006 as specified):
  - log GDP/POP (log level of income per capita 1996–2006),
  - GDP growth 0406 (average growth rate of GDP 2004–2006),
  - log POP (log of population 1996–2006).
- Additional controls (z_p):
  - public sector regulatory quality (regqua, Worldwide Governance Index sub-component 1996–2006),
  - banking regulation/liberalization (CreditMktReg_0406, Fraser Institute 2004–06),
  - Abiad et al. financial liberalization (Finreg7305),
  - indicators of banking/financial system size, depth, performance, and structure (2004–06).

### V. Empirical Findings (selected coefficients and significance)
- Table 2 (dependent variable: average real GDP growth 2008-09):
  - Regression I: FSHHI 2007 coefficient = -2.296 (1.68)*; observations = 96; R-squared = 0.09.
  - Regression II: CBSS 2007 coefficient = 2.455 (2.08)**; observations = 96; R-squared = 0.10.
  - Regression III (both FSHHI and CBSS): FSHHI 2007 = -2.93 (2.16)**; CBSS 2007 = 2.936 (2.49)**; observations = 96; R-squared = 0.15.
  - Regression IV: GOVRATING07 = -16.688 (3.52)***; observations = 49; R-squared = 0.36.
  - Regression V (three indicators jointly): FSHHI 2007 = -5.1 (3.00)***; CBSS 2007 = -0.002 (0.05); GOVRATING07 = -15.107 (3.42)***; observations = 49; R-squared = 0.49.
- Inclusion of regulatory quality (Table 3):
  - regqua coefficient ≈ -2.061 (4.56)*** across specifications; inclusion does not materially change negative associations of FSHHI and GOVRATING07.
- Inclusion of banking regulation (CreditMktReg_0406) (Table 4):
  - CreditMktReg_0406 coefficient ≈ -1.387 (3.44)***.
  - FSHHI coefficients remain negative and significant in several specifications (e.g., -2.512 (1.94)*).
- Using Abiad et al. financial liberalization index (Finreg7305) (Table 5):
  - Finreg7305 coefficients ≈ -13.73 (4.96)***.
  - FSHHI 2007 = -3.348 (2.41)** in regression I (observations 71; R-squared 0.41).
- Combined regulatory controls (Table 6):
  - regqua remains consistently negative and significant (e.g., -1.384 (2.51)**).
  - FSHHI 2007 often negative and significant (e.g., -2.12 (1.67)*), GOVRATING07 negative and significant in multiple specifications (e.g., -8.301 (1.74)*).
- Interaction tests (Table 7):
  - Interaction terms (FSHH * regqua, FSHH * finreg, GOVRAT*regqua, GOVRAT*finreg) generally negative but not significant.
- Financial sector size and performance controls (Tables 8–11):
  - Including size indicators (e.g., com bank assets/GDP): FSHHI 2007 remains negative and often significant (e.g., -2.918 (2.20)**); bank assets sometimes negative and significant (e.g., -7.325 (2.22)** in one specification).
  - Including performance indicators: GOVRATING07 remains negative and significant in some specifications (e.g., -8.203 (1.74)*); bank credit/deposits ratio significant and negative in selected regressions.
  - Banking industry structure variables (concentration, internationalization, conglomerates) generally not significant; supervisory unification and governance negative effects persist but often not significant when these structure variables included.
- Empirical summary:
  - Supervisory unification (FSHHI) and better supervisory governance (GOVRATING07) are associated with weaker economic resilience during 2008–09 across multiple regressions.
  - Degree of central bank involvement in supervision (CBSS/CBFS) did not have a robust significant impact on resilience.
  - Quality of public sector regulation (regqua) and degree of financial liberalization (CreditMktReg_0406, Finreg7305) are consistently negatively associated with resilience and sometimes stronger predictors than supervisory features.

### VI. Narrative Evidence on Supervisory Failures (2007–09)
- Documented supervisory architecture failures:
  - fragmented supervisory structures (Leijonhufvud, 2009—US example);
  - absence of an agency in charge of macro-prudential/systemic supervision;
  - UK Northern Rock episode: Bank of England reportedly lacked information on bank state and failed to timely intervene (Buiter, 2008; FSA, 2009).
- Supervisory governance failures:
  - weak independence and accountability;
  - political, industry and cognitive capture;
  - lack of intrusiveness, reluctance to probe or act, insufficient skills to understand sophisticated risks;
  - misalignment of incentives for cross-border cooperation and information sharing.

### VII. Policy Recommendations and Proposed Institutional Design
- Consensus recommendations from literature and policy debates:
  - Distinguish macroprudential supervision from microprudential supervision (twin-peak model).
  - Central banks should play a major role in macroprudential supervision in many proposals.
  - Strengthen supervisory governance: clarify mandates, increase independence and accountability, improve skills and compensation, ensure resources, enhance transparency and market discipline, develop cross-border coordination.
  - Move toward more intrusive, proactive, risk-based and results-oriented supervision.
- Paper’s second-best proposal (institutional separation to generate checks and balances):
  - Structure: two separate agencies — macroprudential likely housed in the central bank; microprudential in an agency at arm’s length.
  - Theoretical rationale: models (Laffont and Martimort 1999; Boyer and Ponce 2010) show separation reduces scope for collusion/capture by dividing information and discretion, introducing Bayesian-Nash dynamics that raise transaction costs of collusion and improve social welfare.
  - Expected advantages:
    - (i) checks and balances better align supervisors’ incentives and reduce capture likelihood;
    - (ii) avoid concentration of all power in one agency;
    - (iii) synergies from central bank analytical scope for macroprudential tasks;
    - (iv) clearer link between macroprudential supervision and central bank liquidity provisioning/lender of last resort function.
  - Expected costs and challenges:
    - (i) need for some double reporting by financial firms to maintain checks and balances;
    - (ii) need for general coordination while preserving checks and balances;
    - (iii) coordination challenges in deciding which agency takes specific measures and in transferring authority to resolution agencies;
    - (iv) potential inter-agency competition (mitigated by distinct mandates).
  - Policy judgment: models indicate costs are lower than potential benefits; several countries are redesigning supervisory architecture and can apply this second-best mechanism in conjunction with governance improvements.

### VIII. Conclusion
- Post-Asian-crisis reforms (BCPs, governance improvements, architectural changes, market discipline) had mixed empirical support pre-crisis and did not prevent the 2007–09 crisis.
- Systematic empirical analysis for about 100 countries finds:
  - consolidation/unification in supervision and better supervisory governance are negatively correlated with economic resilience in 2008–09;
  - central bank involvement in supervision does not have a significant impact on resilience;
  - the impact of supervisory regimes is intertwined with public sector regulatory quality and financial liberalization; each supervisory feature’s impact depends on overall setting.
- Policy implication:
  - improvements in supervisory governance are necessary but limited;
  - combining institutional separation of macro- and micro-prudential supervision (twin-peak with institutional separation) with governance reforms offers a promising second-best approach to better align incentives and enhance supervisory effectiveness.

### IX. Appendix 1 – Data Description (selected)
- Dependent variable:
  - GDP growth (annual %): Annual Growth of Gross Domestic Product in current US dollars (World Development Indicators).
- Independent variables (supervisory and governance):
  - FSHH Index: level of consolidation of the supervisory powers (our calculation).
  - CBSS Index: level of central bank involvement in supervision (our calculation).
  - Govrating: quality of supervisory governance (Quintyn et al. 2004).
- Regulatory variables:
  - Reg Qua (Regulatory Quality): sub-component of the Worldwide Governance Index, 1996–2006.
  - CreditMktReg0406 (Banking Regulation Quality): Fraser Institute, Economic Freedom Network.
  - Finreg7305 (Financial Regulation Quality): Abiad et al. 2008.
- Financial variables (selected definitions and sources):
  - DepMoneyBankAssetsShare: DEPOSIT MONEY BANK ASSETS / (DEPOSIT MONEY + CENTRAL) BANK ASSETS (World Bank Financial Structure Dataset).
  - LiqLiab/GDP: LIQUID LIABILITIES / GDP (World Bank Financial Structure Dataset).
  - CBAssets: CENTRAL BANK ASSETS / GDP (World Bank Financial Structure Dataset).
  - DepMoneyBankAssets: DEPOSIT MONEY BANK ASSETS / GDP (World Bank Financial Structure Dataset).
  - PrivateCreditBanks: PRIVATE CREDIT BY DEPOSIT MONEY BANKS / GDP (World Bank Financial Structure Dataset).
  - BankDep: BANK DEPOSITS / GDP (World Bank Financial Structure Dataset).
  - BankCredit/Dep: BANK CREDIT / BANK DEPOSITS (World Bank Financial Structure Dataset).
  - BankCosts/Assets: BANK OVERHEAD COSTS / TOTAL ASSETS (World Bank Financial Structure Dataset).
  - NetIntMargin: NET INTEREST MARGIN (World Bank Financial Structure Dataset).
  - Concentration: Assets of three largest banks as a share of assets of all commercial banks (World Bank Financial Structure Dataset).

### X. Appendix 2 – Selected Data Statistics (as reported)
- gdpg~9annual:
  - Obs: 99
  - Mean: 1.1966233.724454-11.2834
  - Max: 9.35
- gdpg~6annual:
  - Obs: 102
  - Mean: 5.3600262.63548-4.7
  - Max: 12.82556
- pop:
  - Obs: 102
  - Mean: 48.42103161.42670.281333
  - Max: 1260.827
- logpop:
  - Obs: 102
  - Mean: 1.047930.704866-0.55078
  - Max: 3.100655
- gdp:
  - Obs: 102
  - Mean: 364.51691207.9661.542372
  - Max: 10853.19
- gdppop:
  - Obs: 102
  - Mean: 24.1092182.616920.03812
  - Max: 662.2144
- loggdppop:
  - Obs: 102
  - Mean: 0.6183940.819486-1.41885
  - Max: 2.820999
- regqua9606:
  - Obs: 102
  - Mean: 0.3446080.84381-1.84
  - Max: 1.85
- bankreg0406:
  - Obs: 97
  - Mean: 8.4051551.0328335.1
  - Max: 9.9
- finreg7305:
  - Obs: 73
  - Mean: 0.8215070.1428840.45
  - Max: 1
- fshh2007:
  - Obs: 99
  - Mean: 0.5824240.2847060.15
  - Max: 1
- cbss2007:
  - Obs: 99
  - Mean: 0.3135350.3218550
  - Max: 1
- govrati~2007:
  - Obs: 50
  - Mean: 0.630.1054630.4
  - Max: 0.9
- liqliab0406:
  - Obs: 87
  - Mean: 0.6223350.4494310.146818
  - Max: 3.298681
- fsdep_0406:
  - Obs: 91
  - Mean: 0.5870350.4663160.082607
  - Max: 3.260568
- pricred~0406:
  - Obs: 91
  - Mean: 0.6290190.4973580.055888
  - Max: 2.026815
- cbasset0406:
  - Obs: 86
  - Mean: 0.045960.0610690.000134
  - Max: 0.396019
- banksas~0406:
  - Obs: 96
  - Mean: 0.9025150.1280480.38469
  - Max: 0.999917
- bankdep0406:
  - Obs: 91
  - Mean: 0.580120.4633120.082607
  - Max: 3.260568
- bankcos~0406:
  - Obs: 100
  - Mean: 0.0403840.0220160.009241
  - Max: 0.111793
- netintm~0406:
  - Obs: 100
  - Mean: 0.047450.0402040.007704
  - Max: 0.361475
- bankcre~0406:
  - Obs: 102
  - Mean: 1.0421650.5405370.26653
  - Max: 3.273302
- concentrat~n:
  - Obs: 80
  - Mean: 68.0418819.3013721
  - Max: 100
- internazio~n:
  - Obs: 76
  - Mean: 33.1430329.525740
  - Max: 100

*Source: _wp11261 - References (PDF).*

### 1.  Financial Supervision Unification ..................................................................................

### Financial Supervision Unification

### Major sections
- 1.  Financial Supervision Unification ......................................................................................12
- 2.  Central Bank Involvement in Supervision ..........................................................................13
- 3.  Supervisory Governance Ratings ........................................................................................14

### Appendices
- Appendix 1 – Data Description ................................................................................................44
- Appendix 2 – Data Statistics ....................................................................................................47

*Source: _wp11261 - 1. Financial Supervision Unification (page and section listing as provided).*

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

### _wp11261 - References

### I. Introduction and Objectives
- Paper objective is twofold:
  - Empirically test impact on economic resilience of:
    - changes in supervisory architecture (often toward unification), and
    - supervisory governance improvements.
  - Review proposals to improve supervisory effectiveness and offer a proposal: conducting supervision through two separate agencies (one for macro- and one for micro-prudential supervision) to introduce checks and balances.
- Main empirical findings summarized in the paper:
  - (i) the two types of modifications introduced in supervision (unification and better governance arrangements) are negatively associated with economic resilience;
  - (ii) the quality of public sector governance and the degree of financial liberalization are negatively associated with economic resilience in this crisis, with associations even stronger than those of the supervisory features analyzed;
  - (iii) the degree of involvement of the central bank in supervision did not have any significant impact on resilience.

### II. Background and Conceptual Framework
- Four groups of initiatives to improve supervisory effectiveness:
  - Basel Core Principles (BCP) and related codes (IOSCO, IAIS); major BCP revision in 2006.
  - Revisions of supervisory architecture (trend toward unification; FSA UK example, 1997).
  - Principles of supervisory governance organized around four pillars: independence, accountability, transparency and integrity.
  - Reliance on market discipline (e.g., subordinated debt to reduce moral hazard).
- Prior empirical literature:
  - Mixed evidence on BCP compliance and bank soundness (Sundararajan et al. 2001; Podpiera 2006; Demirgüç-Kunt et al. 2006; Demirgüç-Kunt and Detragiache 2010).
  - Mixed evidence on supervisory architecture; some studies find unified regimes associated with higher compliance (Čihák and Podpiera, 2007; Arnone and Gambini, 2007) or fewer nonperforming loans when independent supervisors are outside central bank (Eichengreen and Dincer, 2011—insignificant when including 2008–09).
  - Das, Quintyn and Chenard (2004) show quality of governance matters for banking soundness and good public governance amplifies supervisory governance effects.

### III. Quantifying Supervisory Features: Indices and Measurement
- Financial Supervision Herfindahl Hirschman (FSHH) Index:
  - FSHH = Σ i=1..n s_i^2 where s_i is share of supervisory power of authority i.
  - Assumptions: three sectors (banking, securities, insurance) each of equal importance; supervisory power equally divided among authorities in each sector where more than one authority exists.
- Central Bank as Financial Supervisor (CBFS) Index:
  - CBFS is the share of the central bank in supervisory responsibilities (range 0 to 1).
- Supervisory governance ratings:
  - Ratings for independence and accountability: “2” if legal framework satisfies criteria, “1” partial compliance, “0” noncompliance, “-1” for practices undermining independence/accountability; summed and normalized between 0 and 1.
- Stylized cross-country patterns (2007 vs. 2009):
  - FSHH: before crisis (2007) EU had greater consolidation than industrial countries and Europe; consolidation continued in advanced countries during crisis (2009) while sample-wide slight reduction observed.
  - CBFS: before crisis (2007) advanced countries averaged lower central bank involvement; during crisis (2009) central bank involvement increased in advanced, European and EU countries (a “Great Reversal”).
  - Governance ratings: governance quality highest in EU, Europe, industrial countries pre-crisis (2007); governance quality increased further in 2009.

### IV. Empirical Strategy and Controls
- Dependent variable: average real output growth in years 2008–09.
- Key explanatory variables dated 2007: FSHHI 2007, CBSS 2007 (CBFS index), GOVRATING07.
- Macro controls (dated 1996–2006 or 2004–2006 as specified):
  - log GDP/POP (log level of income per capita 1996–2006),
  - GDP growth 0406 (average growth rate of GDP 2004–2006),
  - log POP (log of population 1996–2006).
- Additional controls (z_p): public sector regulatory quality (regqua, Worldwide Governance Index sub-component 1996–2006), banking regulation/liberalization (CreditMktReg_0406, Fraser Institute 2004–06), Abiad et al. financial liberalization (Finreg7305), and multiple indicators of banking/financial system size, depth, performance, and structure (2004–06).

### V. Empirical Findings (selected regression coefficients and significance)
- Table 2 (dependent variable: average real GDP growth 2008-09; sample sizes reported):
  - Regression I: FSHHI 2007 coefficient = -2.296 (1.68)*; observations = 96; R-squared = 0.09.
  - Regression II: CBSS 2007 coefficient = 2.455 (2.08)**; observations = 96; R-squared = 0.10.
  - Regression III (both FSHHI and CBSS): FSHHI 2007 = -2.93 (2.16)**; CBSS 2007 = 2.936 (2.49)**; observations = 96; R-squared = 0.15.
  - Regression IV: GOVRATING07 = -16.688 (3.52)***; observations = 49; R-squared = 0.36.
  - Regression V (three indicators jointly): FSHHI 2007 = -5.1 (3.00)***; CBSS 2007 = -0.002 (0.05); GOVRATING07 = -15.107 (3.42)***; observations = 49; R-squared = 0.49.
- Inclusion of regulatory quality (Table 3):
  - regqua coefficient ≈ -2.061 (4.56)*** across specifications; inclusion does not materially change negative associations of FSHHI and GOVRATING07.
- Inclusion of banking regulation (CreditMktReg_0406) (Table 4):
  - CreditMktReg_0406 coefficient ≈ -1.387 (3.44)***; indicates countries that liberalized banking more were more affected.
  - FSHHI coefficients remain negative and significant in several specifications (e.g., -2.512 (1.94)*).
- Using Abiad et al. financial liberalization index (Finreg7305) (Table 5):
  - Finreg7305 coefficients ≈ -13.73 (4.96)***; confirms stronger hits for more liberalized systems.
  - FSHHI 2007 = -3.348 (2.41)** in regression I (observations 71; R-squared 0.41).
- Combined regulatory controls (Table 6):
  - regqua remains consistently negative and significant (e.g., -1.384 (2.51)**).
  - FSHHI 2007 often negative and significant (e.g., -2.12 (1.67)*), GOVRATING07 negative and significant in multiple specifications (e.g., -8.301 (1.74)*).
- Interaction tests (Table 7):
  - Interaction terms (FSHH * regqua, FSHH * finreg, GOVRAT*regqua, GOVRAT*finreg) generally negative but not significant.
- Financial sector size and performance controls (Tables 8–11):
  - Including size indicators (e.g., com bank assets/GDP): FSHHI 2007 remains negative and often significant (e.g., -2.918 (2.20)**); bank assets sometimes negative and significant (e.g., size indicator -7.325 (2.22)** in one specification).
  - Including performance indicators: GOVRATING07 remains negative and significant in some specifications (e.g., -8.203 (1.74)*); bank credit/deposits ratio significant and negative in selected regressions.
  - Banking industry structure variables (concentration, internationalization, conglomerates) generally not significant; supervisory unification and governance negative effects persist but often not significant when these structure variables included.
- Summary of empirical conclusions:
  - Across regressions, supervisory unification (FSHHI) and better supervisory governance (GOVRATING07) are associated with weaker economic resilience during 2008–09.
  - The degree of involvement of the central bank in supervision (CBSS/CBFS) did not have a robust significant impact on resilience.
  - Quality of public sector regulation (regqua) and degree of financial liberalization (CreditMktReg_0406, Finreg7305) are consistently negatively associated with resilience; sometimes stronger predictors than supervisory features.

### VI. Narrative Evidence on Supervisory Failures in 2007–09 Crisis
- Supervisory architecture failures highlighted in narrative literature include:
  - fragmented supervisory structures (Leijonhufvud, 2009—US example);
  - no agency in charge of macro-prudential/systemic supervision;
  - UK Northern Rock episode: Bank of England reportedly lacked information on bank state and failed to timely intervene (Buiter, 2008; FSA, 2009).
- Supervisory governance failures widely documented:
  - weak independence and accountability;
  - political, industry and cognitive capture;
  - lack of intrusiveness, reluctance to probe or act, insufficient skills to understand sophisticated risks;
  - misalignment of incentives for cross-border cooperation and information sharing.

### VII. Policy Recommendations and Proposal
- Consensus emerging in literature and policy debates:
  - Distinguish macroprudential supervision from microprudential supervision (twin-peak model).
  - Central banks should play a major role in macroprudential supervision in many proposals.
  - Strengthen supervisory governance: clarify mandates, increase independence and accountability, improve skills and compensation, ensure resources, enhance transparency and market discipline, develop cross-border coordination.
  - Move toward more intrusive, proactive, risk-based and results-oriented supervision.
- Paper’s second-best proposal to address incentive misalignment:
  - Use supervisory architecture to generate checks and balances: institutional separation of macroprudential and microprudential supervision (two separate agencies), with macroprudential likely housed in the central bank and microprudential in an agency at arm’s length.
  - Theoretical rationale: Laffont and Martimort (1999) and Boyer and Ponce (2010) models show that separation reduces scope for collusion/capture by dividing information and discretion, introducing Bayesian-Nash dynamics that raise transaction costs of collusion and improve social welfare.
  - Expected advantages:
    - (i) checks and balances better align supervisors’ incentives and reduce capture likelihood;
    - (ii) avoid concentration of all power in one agency;
    - (iii) synergies from central bank analytical scope for macroprudential tasks;
    - (iv) clearer link between macroprudential supervision and central bank liquidity provisioning/lender of last resort function.
  - Expected costs and challenges:
    - (i) need for some double reporting by financial firms to maintain checks and balances;
    - (ii) need for general coordination while preserving checks and balances;
    - (iii) coordination challenges in deciding which agency takes specific measures and in transferring authority to resolution agencies;
    - (iv) potential inter-agency competition (mitigated by distinct mandates).
  - Models indicate costs are lower than potential benefits; several countries are redesigning supervisory architecture and can apply this second-best mechanism in conjunction with governance improvements.

### VIII. Conclusion (paper’s distilled conclusions)
- Post-Asian-crisis reforms (BCPs, governance improvements, architectural changes, market discipline) had mixed empirical support pre-crisis and did not prevent the 2007–09 crisis.
- Systematic empirical analysis for about 100 countries finds:
  - consolidation/unification in supervision and better supervisory governance are negatively correlated with economic resilience in 2008–09;
  - central bank involvement in supervision does not have a significant impact on resilience;
  - the impact of supervisory regimes is intertwined with public sector regulatory quality and financial liberalization; each supervisory feature’s impact depends on overall setting.
- Policy implication: improvements in supervisory governance are necessary but limited; combining institutional separation of macro- and micro-prudential supervision (twin-peak with institutional separation) with governance reforms offers a promising second-best approach to better align incentives and enhance supervisory effectiveness.

*Italic: Source — _wp11261 - References (PDF).*

### Appendix 1 – Data Description

### Appendix 1 – Data Description

### Dependent variable
- GDP growth (annual %)
  - Definition: Annual Growth of Gross Domestic Product in current US dollars
  - Description: World Development Indicators

### Independent variables (supervisory and governance)
- FSHH Index (Financial Supervision Herfindahl Hirschman Index)
  - Definition: level of consolidation of the supervisory powers
  - Description: our calculation
- CBSS Index (Central Bank Supervisory Share Index)
  - Definition: level of central bank involvement in supervision
  - Description: our calculation
- Govrating (Supervisory Governance)
  - Definition: quality of supervisory governance
  - Description: Quintyn et al. 2004

### Macro variables
- GDP per capita (current US$)
  - Definition: Gross per capita Domestic Product in current US dollars
  - Source: World Development Indicators
- log_GDP
  - Definition: Logarithmic transformation of GDP per capita
- Population, total
  - Definition: Population
  - Source: World Development Indicators
- log_pop
  - Definition: Logarithmic transformation of Population

### Regulatory Variables
- Reg Qua (Regulatory Quality)
  - Definition: This is a sub-component of the Worldwide Governance Index computed by the World Bank. Regulatory quality is a measure of the ability of the government to formulate and implement sound policies and regulations that permit and promote private sector development. It is based on surveys of firms and industries and on the assessment of commercial risk rating agencies, non-governmental organizations and various multilateral aid agencies and public sector organizations. For example, it includes the assessment of the World Economic Forum global competitiveness report. It considers price liberalization, competition policies in various sectors, discriminatory taxes and tariffs, trade and exchange rate controls, access to capital markets.
  - Source: Worldwide Governance Index, World Bank
- CreditMktReg0406 (Banking Regulation Quality)
  - Definition: This includes ownership of banks (percentage of deposits held in privately owned banks), competition (the extent to which domestic banks face competition from foreign banks), extension of credit (percentage of credit extended to the private sector) and presence of interest rate controls.
  - Source: Fraser Institute, Economic Freedom Network
- Finreg7305 (Financial Regulation Quality)
  - Definition: This includes seven different dimensions: credit controls and reserve requirements, interest rates controls, entry barriers, state ownership, policies on securities markets, banking regulations, capital account restrictions
  - Source: Abiad et al. 2008

### Financial Variables
- DepMoneyBankAssetsShare (DEPOSIT MONEY BANK ASSETS / (DEPOSIT MONEY + CENTRAL) BANK ASSETS)
  - Definition: Ratio of deposit money bank claims on domestic nonfinancial real sector (as defined above) to the sum of deposit money bank and Central Bank claims on domestic nonfinancial real sector (as defined above)
  - Source: World Bank Financial Structure Dataset
- LiqLiab/GDP (LIQUID LIABILITIES / GDP)
  - Definition: Ratio of liquid liabilities to GDP, calculated using the following deflation method: {(0.5)*[Ft/P_et + Ft-1/P_et-1]}/[GDPt/P_at] where F is liquid liabilities, P_e is end-of period CPI, and P_a is average annual CPI
  - Source: World Bank Financial Structure Dataset
- CBAssets (CENTRAL BANK ASSETS / GDP)
  - Definition: Claims on domestic real nonfinancial sector by the Central Bank as a share of GDP, calculated using the following deflation method: {(0.5)*[Ft/P_et + Ft-1/P_et-1]}/[GDPt/P_at] where F is Central Bank claims, P_e is end-of period CPI, and P_a is average annual CPI
  - Source: World Bank Financial Structure Dataset
- DepMoneyBankAssets (DEPOSIT MONEY BANK ASSETS / GDP)
  - Definition: Claims on domestic real nonfinancial sector by deposit money banks as a share of GDP, calculated using the following deflation method: {(0.5)*[Ft/P_et + Ft-1/P_et-1]}/[GDPt/P_at] where F is deposit money bank claims, P_e is end-of period CPI, and P_a is average annual CPI
  - Source: World Bank Financial Structure Dataset
- PrivateCreditBanks (PRIVATE CREDIT BY DEPOSIT MONEY BANKS / GDP)
  - Definition: Private credit by deposit money banks to GDP, calculated using the following deflation method: {(0.5)*[Ft/P_et + Ft-1/P_et-1]}/[GDPt/P_at] where F is credit to the private sector, P_e is end-of period CPI, and P_a is average annual CPI
  - Source: World Bank Financial Structure Dataset
- PrivateCreditTotal (PRIVATE CREDIT BY DEPOSIT MONEY BANKS AND OTHER FINANCIAL INSTITUTIONS / GDP)
  - Definition: Private credit by deposit money banks and other financial institutions to GDP, calculated using the following deflation method: {(0.5)*[Ft/P_et + Ft-1/P_et-1]}/[GDPt/P_at] where F is credit to the private sector, P_e is end-of period CPI, and P_a is average annual CPI
  - Source: World Bank Financial Structure Dataset
- BankDep (BANK DEPOSITS / GDP)
  - Definition: Demand, time and saving deposits in deposit money banks as a share of GDP, calculated using the following deflation method: {(0.5)*[Ft/P_et + Ft-1/P_et-1]}/[GDPt/P_at] where F is demand and time and saving deposits, P_e is end-of period CPI, and P_a is average annual CPI
  - Source: World Bank Financial Structure Dataset
- FinSystemDep (FINANCIAL SYSTEM DEPOSITS / GDP)
  - Definition: Demand, time and saving deposits in deposit money banks and other financial institutions as a share of GDP, calculated using the following deflation method: {(0.5)*[Ft/P_et + Ft-1/P_et-1]}/[GDPt/P_at] where F is demand and time and saving deposits, P_e is end-of period CPI, and P_a is average annual CPI
  - Source: World Bank Financial Structure Dataset
- BankCredit/Dep (BANK CREDIT / BANK DEPOSITS)
  - Definition: Private credit by deposit money banks as a share of demand, time and saving deposits in deposit money banks.
  - Source: World Bank Financial Structure Dataset
- LiqLiab (LIQUID LIABILITIES (IN MIL. 2000 USD))
  - Definition: Absolute value of liquid liabilities in 2000 US dollars
  - Source: World Bank Financial Structure Dataset
- BankCosts/Assets (BANK OVERHEAD COSTS / TOTAL ASSETS)
  - Definition: Accounting value of a bank's overhead costs as a share of its total assets.
  - Source: World Bank Financial Structure Dataset
- NetIntMargin (NET INTEREST MARGIN)
  - Definition: Accounting value of bank's net interest revenue as a share of its interest-bearing (total earning) assets.
  - Source: World Bank Financial Structure Dataset
- Concentration (BANK CONCENTRATION INDEX)
  - Definition: Assets of three largest banks as a share of assets of all commercial banks.
  - Source: World Bank Financial Structure Dataset
- Internationalization (BANK INTERNAZIONALIZATION INDEX)
  - Source: World Bank Financial Structure Dataset
- Conglomerates (CONGLOMERATES INDEX)
  - Source: World Bank Financial Structure Dataset

### Appendix 2 – Data Statistics (selected variables)
- gdpg~9annual
  - Obs: 99
  - Mean: 1.1966233.724454-11.2834
  - Max: 9.35
- gdpg~6annual
  - Obs: 102
  - Mean: 5.3600262.63548-4.7
  - Max: 12.82556
- pop
  - Obs: 102
  - Mean: 48.42103161.42670.281333
  - Max: 1260.827
- logpop
  - Obs: 102
  - Mean: 1.047930.704866-0.55078
  - Max: 3.100655
- gdp
  - Obs: 102
  - Mean: 364.51691207.9661.542372
  - Max: 10853.19
- gdppop
  - Obs: 102
  - Mean: 24.1092182.616920.03812
  - Max: 662.2144
- loggdppop
  - Obs: 102
  - Mean: 0.6183940.819486-1.41885
  - Max: 2.820999
- regqua9606
  - Obs: 102
  - Mean: 0.3446080.84381-1.84
  - Max: 1.85
- bankreg0406
  - Obs: 97
  - Mean: 8.4051551.0328335.1
  - Max: 9.9
- finreg7305
  - Obs: 73
  - Mean: 0.8215070.1428840.45
  - Max: 1
- fshh2007
  - Obs: 99
  - Mean: 0.5824240.2847060.15
  - Max: 1
- cbss2007
  - Obs: 99
  - Mean: 0.3135350.3218550
  - Max: 1
- govrati~2007
  - Obs: 50
  - Mean: 0.630.1054630.4
  - Max: 0.9
- liqliab0406
  - Obs: 87
  - Mean: 0.6223350.4494310.146818
  - Max: 3.298681
- fsdep_0406
  - Obs: 91
  - Mean: 0.5870350.4663160.082607
  - Max: 3.260568
- pricred~0406
  - Obs: 91
  - Mean: 0.6290190.4973580.055888
  - Max: 2.026815
- cbasset0406
  - Obs: 86
  - Mean: 0.045960.0610690.000134
  - Max: 0.396019
- banksas~0406
  - Obs: 96
  - Mean: 0.9025150.1280480.38469
  - Max: 0.999917
- bankdep0406
  - Obs: 91
  - Mean: 0.580120.4633120.082607
  - Max: 3.260568
- bankcos~0406
  - Obs: 100
  - Mean: 0.0403840.0220160.009241
  - Max: 0.111793
- netintm~0406
  - Obs: 100
  - Mean: 0.047450.0402040.007704
  - Max: 0.361475
- bankcre~0406
  - Obs: 102
  - Mean: 1.0421650.5405370.26653
  - Max: 3.273302
- concentrat~n
  - Obs: 80
  - Mean: 68.0418819.3013721
  - Max: 100
- internazio~n
  - Obs: 76
  - Mean: 33.1430329.525740
  - Max: 100

*Source: Appendix 1 – Data Description (from the provided PDF content).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11261.pdf_