纽约联储-银行组织的复杂性与风险性：来自国际银行研究网络的证据（英文）-2021.5-35正式版.ppt

Complexity and Riskinessof Banking Organizations:Evidence from theInternational BankingResearch NetworkNO.9 6 6MAY 2 0 2 1Claudia M.Buch|Lind a GoldbergComplexity and Riskiness of Banking Organizations:Evidence from the International Banking Research NetworkClaudia M.Buch and Linda GoldbergFederal Reserve Bank of New York Staff Reports,no.966May 2021JEL classification:G21,G28,G32AbstractComplexity of banks can have important ramifications for the performance and the risks of the bankingsystem.Financial sector reforms that were implemented in the past decade have thus aimed to reduce andto better manage the risk implications of bank complexity.Yet,surprisingly little is known about changesin complexity across countries,its drivers,and its effects.The International Banking Research Network(IBRN)used data and analytical advances to generate rich cross-country insights on the complexity andriskiness of banking organizations.The initiative has yielded four key findings.First,the largest banks incountries tend to be the more complex ones.Even controlling for size,there is substantial diversity acrossbanking organizations in terms of complexity choices.Second,over the past decade,bankingorganizations have tended to reduce complexity by limiting the number of affiliates in domestic andforeign locations.Generally,however,complexity patterns are fairly persistent.Third,regulatory changescan alter both banking organization complexity and the associated risk profiles.Fourth,the link betweencomplexity and risks involves trade-offs:diversification benefits and reductions in liquidity risk mayweigh against agency problems,monitoring costs,and systemic risk contributions arising from highercomplexity.Key words:bank complexity,financial regulation,international banking,risks in banking_Goldberg(corresponding author):Federal Reserve Bank of New York(email:linda.goldbergny.frb.org).Buch:Deutsche Bundesbank.The authors thank the International BankingResearch Network and in particular Iaki Aldasoro,Isabel Argmon,Diana Bonfim,Sonia Felix,Krysztoph Gajewksi,Bryan Hardy,Andres Murcia Pabon,Francesco Palazzo,Maria Rodrguez-Moreno,Alejandra Rosado Cuervo,Esther Segalla,and Ursula Vogel for thoughtful discussions of content,methodology,and data,as well as for the empirical results and research that serve as inputs into the meta-analysis of this paper.Excellent research support was provided by Benedikt Fritz,Sarah Hamerling,Janavi Janakiraman,and Kevin Lai.This paper presents preliminary findings and is being distributed to economists and other interestedreaders solely to stimulate discussion and elicit comments.The views expressed in this paper are those ofthe author(s)and do not necessarily reflect the position of the Federal Reserve Bank of New York or theFederal Reserve System.Any errors or omissions are the responsibility of the author(s).To view the authors disclosure statements,visithttps:/www.newyorkfed.org/research/staff_reports/sr966.html.I.MotivationBanking organizations are quite heterogeneous:they can be simple comprised of traditionalbanks that mainly provide basic banking functions of taking deposits and extending loans orcan be complex in their organizational structures,types of businesses conducted,and theirgeographic span.Many banking organizations are corporate conglomerates that contain banks,but also can contain dozens,hundreds,or even thousands of nonbank legal entities.Theirbusiness scope can span financial and non-financial activities,and their geography can spanmultiple countries.Despite the clear relevance,the complexity patterns and their implications for the activities andthe risks of banking organizations are under-researched.Understanding these patterns and theimplications for banking organization risks are the focus of this paper.While complexity oftenhas a negative connotation,we find that it entails tradeoffs.Complexity can reduce exposure tosome risks as it allows banks to exploit synergies across activities.It can yield benefits in termsof risk diversification and reduced liquidity risk.However,complexity can also increase risks dueto the stronger challenges that occur around risk containment and management,and it canincrease the costs and feasibility of resolution when the organization fails.The global financial crisis of 2007/08 demonstrated the dark side of bank complexity.Thebalance sheet frailties of large and complex financial institutions had been underestimated,aswere the negative externalities that were imposed on other institutions,governments,and thereal economy.Particularly relevant are costs that arise in times of stress,with recovery andresolution of gone concern banking organizations impeded by a high degree of complexity,including in cross-border contexts.Far-reaching post crisis reforms have thus aimed at making financial institutions more resilientand at reducing their systemic risk externalities.The regulatory community agreed to a commonapproach to measure complexity using then available data(BIS 2013).This measure usesspecific balance sheet categories associated with informational opacity and illiquidity inassessing the need for additional liquidity and capital requirements,as well as proxies forcomplexity.2 However,the crisis and subsequent policy responses revealed the need for abetter understanding of the complexity of banking organizations,both in terms of determinantsand implications for risk.2The Basel Committee on Banking Supervision adopted an assessment methodology for global systemicallyimportant banks,and higher loss absorbency requirements,with the updated methodology is at BIS(2013).2Today,we are in a much better position to assess the determinants and patterns of bankcomplexity than prior to the global financial crisis.This includes assessing relationships withorganizational incentives and risk outcomes,and the effects of the regulatory reforms that havebeen implemented post crisis.One simple reason is the passage of time.More than tenadditional years of data enable meaningful analytics,comparing developments over time andaround significant policy actions that occurred during this period.In addition,the datainfrastructure has improved significantly:more granular bank-level data allows creation ofcomplexity measures and studying trends across different banking organizations.Another factoris that the research community has developed tools to analyze the efficiencies and incentiveissues within banking organizations,including on how moral hazard,organizational design,andcorporate cultures influence risk outcomes.The International Banking Research Network(IBRN)advanced this agenda by generating richcross-country insights on the complexity and riskiness of banking organizations.Researchersfrom thirteen central banks and the Bank for International Settlements(BIS)worked withconfidential data collected by their regulators to provide comprehensive new evidence onbanking organization complexity.Research papers written span perspectives of home and hostcountries of complex banking organizations.Studies consider the mechanisms through whichthe complexity of banking organizations affects risks associated with these institutions,as wellas drivers of such complexity.This comprehensive new evidence on banking complexity from the vantage point oforganizational,business,and geographic dimensions yields four key contributions:First,structural features of bank complexity are quite persistent over time.High complexitytends to be concentrated in a relatively limited number of institutions,with the largest assetsize tier of banking organizations also having the greatest degree of complexity.The relationshipbetween(asset)size and complexity strongest amongst larger banking organizations.However,even controlling for size,there is considerable diversity across organizations in their complexitychoices so that size is not a sufficient proxy for banking organization complexity.Second,over time,banks have tended to reduce organizational complexity by limiting thenumber of affiliates located in domestic and foreign locations.Aggregate indicators of businesscomplexity exhibited more modest changes over time,with specific changes arising in thecomposition of businesses rather than complexity across businesses.Geographic complexityincreased for banking organizations from some countries while declining for others.Third,regulatory changes can alter both complexity and the risk profiles of complex bankingorganizations.Several studies measure the response of complexity and risk to the Basel III3regulatory framework,including its criteria for the designation of G-SIBs.German banks affectedby a tightening of regulations reduced geographic and business complexity,but at the sametime managed to increase their diversification and thereby reduce theirrisk(Martynova andVogel 2021).Foreign banking affiliates of G-SIBs hosted by Hong Kong saw a larger decline inrisks than their counterparts as a result of the reduced business complexity of G-SIBs(Ho,Wongand Tan 2021).Implementation of Basel III regulations has been associated with a change inequity portfolios and divested financial holdings in Austrian banking organizations(Ehrlich,Elsinger,Lindner,Segalla and Sigmund 2020).Meanwhile,Norwegian bank balance sheetopacity declined in response to Basel III(Cao and Juelsrud 2021).US regulatory changesspecifically targeted at organizational complexity,such as the Living Will provisions of the DoddFrank Act,significantly changed complexity and risk outcomes.Organizational complexitydeclines were associated with reduced systemic risk,but increased liquidity risk exposures(Correa and Goldberg 2021).In Spain,the introduction of Institutional Protection Schemes(IPS)as a consolidation mechanism allowed increases in organizational complexity without affectingidiosyncratic risk(Argimn and Rodrguez-Moreno 2021).Fourth,specific tests of the mechanisms through which complexity influences risk outcomesreinforce the focus on trade-offs:Diversification benefits tend to reduce idiosyncratic risk,whileagency problems and monitoring costs in complex institutions might increase risk.For US banks,trade-offs differ by type of complexity considered:higher organizational,business,andgeographic complexity generated diversification benefits.Geographic complexity also reducesliquidity risk exposure.All three types of complexity contribute to increased systemic risk.Idiosyncratic risk also declines with complexity,in particular geographic complexity,for Spanishbanks and also for business complexity,for Polish and Portuguese banking organizations.Suchdiversification benefits appear to be outweighed by higher agency costs of complexorganizations:Results for banking organizations in Colombia,France,and Hong Kong show thatidiosyncratic risks tend to increase with complexity.German systemically important banksmanaged to increase their diversification benefits while also reducing complexity in response toregulatory tightenings.Italian banking organizations active in more markets were found to bemore selective,with reduced exposure to riskier borrowers;the opposite holds forintermediaries with a higher degree of diversification of fee income.This is consistent withmixed idiosyncratic risk outcomes.Various studies point to better risk frontier outcomes whenbanking organizations are ex ante better capitalized or better governed.In a cross-countrysetting,the geographic complexity of foreign banking affiliates of the largest international bankstended to mitigate the effects of local shocks on bank risk(z-score)while weakening the positiveeffects of prudential regulation on capitalization(Aldasoro,Hardy and Jager 2021).4The remainder of the paper is organized as follows.In Part two,we compile new evidence onthe patterns and determinants of bank complexity.Part three focuses on the drivers of bankcomplexity.Part four presents evidence on the link between complexity and risk.Part fiveconcludes with policy insights and research suggestions.II.Evidence on Bank ComplexityWhile the simple,stand-alone commercial bank has traditionally been the perspectiveembedded in research and policy,the banking landscape has changed.Many banks are complexcombinations of different businesses,sometimes in many locations around the world,andconducted through a host of legal entities.Countries can also serve as hosts to foreign bankingorganizations,with subsidiaries or branches of complex global bank holding companies(BHCs)being important providers of local credit and financial services.Below,we describe recentanalytical advances that provide lenses on these forms of complexity,and then we turn to newcross-country evidence.1.Measuring Bank ComplexityRegulators use a number of criteria to designate global systemically important financialinstitutions(G-SIFIs)and to assess systemically important banks on the national level.Thesecriteria include size,cross-jurisdictional activity,inter-connectedness,and complexity.Complexity,in the regulatory context,is related to the opacity of balance sheet and off balancesheet assets of banks.It increases with holdings of assets that are hard to understand and toprice,such as notional amounts of over-the-counter(OTC)derivatives,trading and available-for-sale(AFS)securities,and level 3 assets from the classification method of the Basel Committeefor Banking Supervision(BCBS)for global systemically important banks(G-SIBs)(BCBS 2013).Other relevant dimensions of complexity include the extent to which there is broad businessscope,wide geographic scope,and more organizational complexity through multiple legalentities.These concepts can be partially informed using information on banks balance sheets.Some are also informed by using data on the industries and locations of the bank,non-bankfinancial,and non-financial affiliates of the banking organizations.Recent research links suchcomplexity measures to economic or financial outcomes.Cetorelli and Goldberg(2014),forexample,distinguish between organizational complexity,business complexity,and geographiccomplexity as for 170 foreign banking organizations with US branches,with measuresconstructed using a cross-country database covering full organizations(Bankscope).Similar5metrics and data sources are used by Carmassi and Herring(2016)for the very largest globalbanks.3Yet,publicly available databases reach limits when trying to describe relevant features of bankcomplexity.Hence,papers in this initiative of the International Banking Research Network(IBRN)draw on complementary supervisory data or other sources of micro-level data availablein central banks to document changing patterns in the complexity of banks.The main conceptsof complexity that initiative participants created using regulatory data include:4Organizational complexity,which measures the number of entities within the fullbanking organization.Business complexity,which captures the span and concentration of affiliates across typesof broad business cate