Michael Koetter Frankfurt SFM Kasper Roszbach Riksbank U

Michael Koetter (Frankfurt SFM) Kasper Roszbach (Riksbank, U. Groningen) Giancarlo Spagnolo (SSE-SITE, EIEF) Financial Stability and Central Bank Governance DNB conference on Corporate Governance of Financial Institutions 8 -9 November 2012 The views expressed in this presentation are solely the responsibility of the authors and should not be interpreted as reflecting the views of the Executive Board of Sveriges Riksbank. 1

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance • Origin of project • Interest in the governance (characteristics) of central banks • Why is central banking organized as it is? • Enormous variation in way in which central banks are organized; know relatively little • Perceived objectives, institutional features, wages, other measures of independence • Supervisory incentives, checks and balances • Survey 2

3

4

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance • The central banking agenda • Financial crisis has economists and policy makers thinking about: Institutional organization of micro-prudential (bank)supervision Spillovers from monetary policy to financial stability policy (incl. bank resolution) What the main objective of a CB should be Toolbox Whethere is scope for macro-prudential supervision as a complement to micro-prudential supervision What MPS is and how it should be conducted • • • 5

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance • Future institutional arrangements • • • Should new tools and mandates for prudential regulation and systemic financial stability should be given to Central Banks? Does this have consequences for the allocation of micro-prudential supervision? If so, then how to regulate the inevitable conflicts between monetary policy, macrosystemic stability, and micro-supervisory objectives UK recently "reversed" the 1997 separation between a micro-prudential supervisor and the Bo. E More or less with same arguments: supervision of Northern Rock vs Barings/BCCI considered ineffective. Perception: assigning bank supervision to the Central Bank is a poor institutional arrangement was widespread before the crisis 6

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance • But opposing views • • Some notable commentators recently suggested that this widespread perception was wrong Other at least as prominent observers (Goodhart, Buiter): suggest opposite Independent bank supervision authority preferable Only the new macro-systemic stability tasks and tools should be assigned to Central Banks" • • • Reasonable theoretical arguments exist in support of both views => • Important question: which institutional arrangement performs better in • terms of preventing financial distress, particularly in wake of a financial crisis Micro supervision to a specialized authority or to the Central Bank? 7

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance • This paper • • Do differences across Central Banks in terms of their objectives, supervisory mandate, and relevant governance features help to explain banking system fragility prior and during the crisis? Use a new dataset with detailed information on central banks' governance characteristics 8

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance Ingredients • Survey data on central bank governance characteristics (objectives, budgetary freedom, allocation of profits) • Measure of financial fragility: average fraction of non-performing loans in jurisdiction • Can CB governance characteristics, institutional features for supervision, country controls and CB objectives explain variation in fragility? • Pre-crisis and during-crisis 9

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance • Limitations! • • • No panel data on governance and institutional structure Number of jurisdictions (homogeneity) Measure of financial stability 10

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance • Findings • • Country controls, like legal origin, income dispersion, historical levels of credit risk, and OECD membership explain most of fragility Objectives: only currency stability exhibits a consistently negative relationship with credit risk prior to the financial crisis • Possibly some relation to IMF programs? • Institutional separation (or unification) of prudential and monetary policy has no significant impact on financial fragility • Governance characteristics of central banks do not influence NP • For the crisis period, most links break down! 11

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance Data • Survey on central bank governance • Collected information on Board composition, appointment rules, meeting frequency, investment rules, budget rules, et cetera • Conducted between 2004 and 2006 • BIS, World Bank, Laeven and Valencia (crisis variable), Shleifer (legal origin, creditor protection), World Value Survey (trust), market structure variables HHI, Panzar-Rosse) 12

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance Variables of interest • Objectives of central bank: MP, FS, Forex • Separated or integrated bank supervision • Age of CB • Budgetary freedom • Freedom to allocate profits 13

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance 14

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance 15

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance 16

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance • Select regressors from very large set of country controls, CB governance characteristics and institutional features • Keep variables that display univariate relation and add a few other governance controls 17

18

19

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance Pre-crisis results • Macroeconomic differences across countries most relevant to explain financial fragility • Neither more frequent crises in the past nor banking market structure affect NPL • Higher income dispersion and high historical levels of NPL explain most of cross-country variation; correlate positively with pre-crisis NPL • OECD Membership and a UK style legal system reduce NPL • Monetary policy objectives exhibit no significant relation • Central bank governance characteristics have no impact • Pursuing external stability as a prime objective correlates most significantly, negatively, with pre-crisis NPL • Differences in the prudential supervisory architecture have no significant impact 20

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance Mean NPL and country controls pre-crisis (2004 -2006) 21

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance Mean NPL and CB characteristics pre-crisis (2004 -2006) 22

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance Mean NPL and CB characteristics pre-crisis (2004 -2006) 23

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance Crisis results • Ability of covariates to explain NPL shares vanishes almost entirely. • Pre-crisis monitoring equally (in)effective prudential architecture • Number of currency crises post 1970, higher banking market concentration, and more recent foundation of the Central Bank exhibit statistically weakly evidence in favor of lower NPL ratios 24

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance Mean NPL and CB characteristics in crisis (2007 -2009) 25

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance Mean NPL and CB characteristics in crisis (2007 -2009) 26

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance Mean NPL and CB characteristics pre-crisis (2007 -2009) 27

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance Can we explain role of foreign exchange objectives? 28

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance Do foreign exchange objectives reflect learning from crises? 29

Koetter Roszbach Spagnolo: Financial Stability and Central Bank Governance Conclusions • Country specific conditions, past experiences a best at explaining different fragility levels • Consistent with the negative findings of Hasan and Mester (2008) • Only governance characteristic that matters is foreign currency objective; potentially related to previous currency crisis experience • Relationships break down in crisis • Debate on whether to centralize prudential supervision, the potential role of Central Banks versus alternative authorities, appear less relevant • For future research: do countries learn from experiences and/or IMF programs? 30