Liquidity Assessment and the Use of Liquidity Ratio as Defined in the Basel III Accords to Identify Bank Distress



Similar documents
Transformation Risk and its Determinants: A New Approach based on the Basel III Liquidity Management Framework

LIQUIDITY RISK MANAGEMENT GUIDELINE

Prof Kevin Davis Melbourne Centre for Financial Studies. Managing Liquidity Risks. Session 5.1. Training Program ~ 8 12 December 2008 SHANGHAI, CHINA

CONSULTATION PAPER P October Proposed Regulatory Framework on Mortgage Insurance Business

Chart I.1. Difference between Primary Surplus (PS) and Bond Yield Spreads in Selected EU 1 Countries

Measuring banking sector development

For further information UBI Banca Investor Relations Tel UBI Banca Press relations Tel.

THE EARLY WARNING SYSTEM PRESENTATION DEFINITION

EIOPA Stress Test Press Briefing Frankfurt am Main, 4 July 2011

Basel Committee on Banking Supervision. Consultative Document. Basel III: The Net Stable Funding Ratio. Issued for comment by 11 April 2014

Net Stable Funding Ratio

Bank Recovery and Resolution. Sven Schelo. 01 Wolters Kluwer Law & Business

Lecture 10: International banking

Notation. Entrepreneurs Undertake the project if. Total deposits. Sum of all deposits except bank i. Sum of all loans except bank i

DEPARTMENT OF ECONOMICS CREDITOR PROTECTION AND BANKING SYSTEM DEVELOPMENT IN INDIA

Economic Commentaries

THE IMPACT OF MACROECONOMIC FACTORS ON NON-PERFORMING LOANS IN THE REPUBLIC OF MOLDOVA

Disclosure 17 OffV (Credit Risk Mitigation Techniques)

PRESS RELEASE Amsterdam, 23 July 2010

COMMERZBANK Capital Update - EU Wide Stress Test Results.

Financial Evolution and Stability The Case of Hedge Funds

Funds Transfer Pricing A gateway to enhanced business performance

Liquidity Coverage Ratio

Basel Committee on Banking Supervision

Credit Risk Securitization and Bank Soundness

FINANCIAL STABILITY ISSUES FOR SMALL STATES. Mirko Mallia Assistant Executive Financial Stability Surveillance, Assessment and Data

Contents. About the author. Introduction

Basel Committee on Banking Supervision. Consultative Document. Net Stable Funding Ratio disclosure standards. Issued for comment by 6 March 2015

RISK FACTORS AND RISK MANAGEMENT

Final Assessment 1 of Spain's eligibility for an EFSF/ESM loan to recapitalize certain financial institutions

3/22/2010. Chapter 11. Eight Categories of Bank Regulation. Economic Analysis of. Regulation Lecture 1. Asymmetric Information and Bank Regulation

BRUNEL UNIVERSITY Economics and Finance EC1004 LECTURE 10. Issues in commercial bank risk management. Reading: Madura Chapter 19

Policy on the Management of Country Risk by Credit Institutions

@ HONG KONG MONETARY AUTHORITY

NEED TO KNOW. IFRS 9 Financial Instruments Impairment of Financial Assets

REMARKS ON THE BASEL CAPITAL FRAMEWORK AND TRADE FINANCE, 27 FEBRUARY 2014 SESSION 4. Mr. Andrew CORNFORD Research Fellow Financial Markets Center

Risk Management Programme Guidelines

Key learning points I

CHILE FINANCIAL SECTOR ASSESSMENT PROGRAM

2014 EU-wide Stress Test

mr. M.G.F.M.V. Janssen Secretary to the Managing Board T: I:

Morgan Stanley European Financials Conference

The challenge of liquidity and collateral management in the new regulatory landscape

Basel 3: A new perspective on portfolio risk management. Tamar JOULIA-PARIS October 2011

Bank Liabilities Survey. Survey results 2013 Q3

Investment insight. Fixed income the what, when, where, why and how TABLE 1: DIFFERENT TYPES OF FIXED INCOME SECURITIES. What is fixed income?

Prof Kevin Davis Melbourne Centre for Financial Studies. Current Issues in Bank Capital Planning. Session 4.4

The International Certificate in Banking Risk and Regulation (ICBRR)

Basel II. Tamer Bakiciol Nicolas Cojocaru-Durand Dongxu Lu

FINANCIAL CONTAGION IN THE ERA OF GLOBALISED BANKING

SURVEY OF INVESTMENT REGULATION OF PENSION FUNDS. OECD Secretariat

Press release Press enquiries:

The Business Impact of Basel III

Prudential Lessons from the Global Financial Crisis

Micro and macroeconomic determinants of net interest margin in the Albanian banking system ( )

NEED OF FINANCIAL INSTITUTIONS SUPERVISION THROUGH AN SINGLE FRAMEWORK OF MACRO-PRUDENTIAL SUPERVISION

Financial market integration and economic growth: Quantifying the effects, Brussels 19/02/2003

Understanding Financial Consolidation

Risk & Capital Management under Basel III

Does Deposit Insurance Increase Banking System Stability? An Empirical Investigation. Aslı Demirgüç-Kunt and Enrica Detragiache

RISK MANAGEMENT IN NETTING SCHEMES FOR SETTLEMENT OF SECURITIES TRANSACTIONS

Guiding principles on the temporary funding needed to support the orderly resolution of a global systemically important bank ( G-SIB )

How To Find Out If International Debt Securities And Gdp Are Related

Bank Capital Adequacy under Basel III

Basel Committee on Banking Supervision. Basel III: the net stable funding ratio

The consequences of a low interest rate environment from a supervisory perspective

Glossary & Definitions

Lecture 2: The nature of financial intermediation:

* * * BIS central bankers speeches 1

Chart 9.1 Non-performing loans ratio and structure of non-performing loans (right) 25% 80 06/08 03/11 03/09 12/07 12/08 06/09 09/09 12/09 09/08 06/11

Fourth study of the Solvency II standard approach

FSB launches peer review on deposit insurance systems and invites feedback from stakeholders

18,343 18,308 3 Accumulated other comprehensive income (and other reserves)

STRESS TESTING GUIDELINE

HAS FINANCE BECOME TOO EXPENSIVE? AN ESTIMATION OF THE UNIT COST OF FINANCIAL INTERMEDIATION IN EUROPE

ICRA Lanka s Credit Rating Methodology for Non-Banking Finance Companies

CREDIT RISK MANAGEMENT IN INDIAN COMMERCIAL BANKS

The Goldman Sachs Group, Inc. and Goldman Sachs Bank USA Annual Dodd-Frank Act Stress Test Disclosure

Banking System Stress Testing

Determinants of Non Performing Loans: Case of US Banking Sector

Santander Views on Basel s Liquidity Framework

We believe there are arguments in favour of the inclusion within this section of bonds issued by financial institutions and guaranteed by the State:

Rating Methodology for Domestic Life Insurance Companies

Bank risk taking and liquidity creation following regulatory interventions and capital support

Non-Bank Deposit Taker (NBDT) Capital Policy Paper

Financial Status, Operating Results and Risk Management

As of July 1, Risk Management and Administration

Basel Committee on Banking Supervision

Determinants of Capital Structure in Developing Countries

Real Estate in Europe. Update on regulatory developments. Ad Buisman

DECLARATION ON STRENGTHENING THE FINANCIAL SYSTEM LONDON SUMMIT, 2 APRIL 2009

Basel Committee on Banking Supervision. Net Stable Funding Ratio disclosure standards

Insurance corporations and pension funds in OECD countries

Measurement of Banks Exposure to Interest Rate Risk and Principles for the Management of Interest Rate Risk respectively.

Viral V Acharya NYU Stern School of Business February 2012

Basis of the Financial Stability Oversight Council s Final Determination Regarding General Electric Capital Corporation, Inc.

The global financial crisis which began in the

Zurich, 13 December Introductory remarks by Philipp Hildebrand

Australian Bank Capital and the Regulatory Framework

The Mortgage Market Review and Non-bank mortgage lenders is enhanced Prudential Supervision on the way?

Transcription:

Liquidity Assessment and the Use of Liquidity Ratio as Defined in the Basel III Accords to Identify Bank Distress Alain Angora, Caroline Roulet Université de Limoges, LAPE, 5 rue Félix Eboué, 87031 Limoges Cedex, France This version: April 2011 Preliminary draft - Please do not quote without the permission of the authors Abstract We assess the advantage of using the net stable funding ratio as defined in the Basel III accords to detect bank distress. Considering US and European publicly traded commercial banks over the 2005-2009 period, we specify a Logit model which tests if the net stable funding ratio adds predictive value to models relying on liquidity ratios from the CAMELS approach to explain the bank probability of default. Our main results support the use of the net stable funding ratio. These findings shed light on the advantage to improve the definition of liquidity in order ameliorate the assessment of bank financial conditions. Keywords: Bank Distress, Liquidity, Bank Regulation JEL classification: C25, G01, G21, G28 Corresponding author. Tel: +33-555-32-81-88, c.roulet@jplc.fr (C. Roulet). 1

1. Introduction Since financial globalisation and deregulation have highlighted the potential fragility of banks, prudential policies that strengthen banking system stability have not ceased to be reinforced. Following the Subprime crisis that began in mid-2007, proposals from governments at the Pittsburgh G-20 summit or from the Basel Committee are in favour of improving the current regulation framework. The debate focuses on various aspects of the financial regulation such as the redefinition of core capital, the implementation of liquidity ratios, the improvement of risk valuation methodologies, the extension of the scope of the regulation (i.e., to the shadow financial system ) and the implementation of macroprudential regulation. Without minimising the importance of various regulatory standards, prudential policies are generally based on two main principles that are solvency standards and deposit insurance schemes. Whereas the first principle is based on capital requirements to prevent bank solvency problems, the second is related to liquidity requirements and allows avoiding depositor panics. However, the Basel I and II accords are focused on the first principle and require banks to maintain a given level of capital compared with their risk weighted assets. These banking regulation frameworks have been largely criticized, in particular because of their procyclicality. Besides, they give a particular importance to bank capital standards and minimises several other aspects as the role of liquidity. There is a large consensus in the literature that financial market failures and liquidity shortages are the root causes of the Subprime crisis (Adrian and Shin, 2009). Most of the empirical studies about banking crises, using non linear econometric models, are based on microeconomic indicators that include CAMELS 1 indicators. Through the CAMELS approach, bank liquidity is measured by liquidity ratios based on accounting data such as liquid assets to total assets or total loans to total deposits (Demirguc-Kunt, 1990; Gonzalez-Hermosillo, 1999). However, Poorman and Blake (2005) indicate that using such liquidity ratios is not sensitive enough to measure bank liquidity. For example, a large regional bank such as the Southeast Bank in the US, which presented a ratio of total liquid assets to total assets more than 30%, bankrupts due to its inability to repay some liabilities claimed on demand with only its liquid assets. Thus, it emphasizes the importance to consider the liquidity mismatch of assets and liabilities to evaluate the liquidity profile of banks. Besides, loan portfolios have become an 1 The CAMEL approach refers to five components in order to assess bank financial soundness: C for Capital Adequacy, A for Asset quality, M for Management, E for Earnings and L for Liquidity. Since 1997, a sixth component has been added and the CAMEL approach has become the CAMELS approach. The additional component S corresponds to the sensitivity to market risk (Bohn et al., 2003). 2

important factor in liquidity management. Banks can use loans as collateral to secure borrowings, enter into loan participation agreements, and sell the loans on the secondary market. Moreover, focusing on deposit funding leads to ignore some widely used alternative sources of funding through the issue of commercial paper or covered bonds (Bradley and Shibut, 2006). Besides, Decker (2000) mentions that bank liquidity has bank-specific components but also it is likely to be impacted by market collapses. Thus, given the development of bank market activities (Shleifer and Vishny, 2009) 2, the value of assets that could be monetised and the availability of market funding, which depend on the liquidity of financial markets, are essential to consider for liquidity assessment in banking. In recognition of the need for banks to improve their liquidity management following the Subprime crisis 3, the Basel Committee on Banking Regulation and Supervision has developed an international framework for liquidity assessment in banking (BIS, 2009). Among the several guidelines, the Basel III accords include the implementation of the net stable funding ratio. This indicator shows to what extent a bank is able to meet its liquidity requirements without borrowing money or fire selling its assets. It measures the amount of longer-term, stable sources of funding employed by an institution relative to the amount of assets that could not be monetised through the sale or the use as collateral in a secured borrowing. It also includes the potential contingent calls on funding liquidity arising from offbalance sheet commitments and obligations. This ratio is based on accounting data but it includes the liquidity unbalances of both sides of on and off-balance sheets as a whole. In addition, it considers the impact of the liquidity of financial markets, on the valuation of assets and on the availability of funding, in order to define the liquidity of bank assets and liabilities. Indeed, banks are likely to face too many losses from having to sell some assets at fire sale prices to repay some liabilities that may be claimed at short notice (i.e., the fundings that are likely to become unavailable). These losses may prevent banks to repay this amount of debts as the cash value of assets may be too weak. 2 Financial globalisation and the development of financial innovation have led to increase the connection between banks and financial markets. We can mention the widespread use of securitisation of loans and the issuance of complex debt instruments (such as collateralised debt obligations), the fact that many banks lend directly to highly leveraged institutions like hedge funds, the increasing share of trading activities and the increasing use of market funding. 3 Throughout the global financial crisis which began in mid-2007, many banks struggled to maintain adequate liquidity. Unprecedented levels of liquidity support were required from central banks in order to sustain the financial system and even with such extensive support a number of banks failed, were forced into mergers or required resolution. These circumstances and events were preceded by several years of ample liquidity in the financial system, during which liquidity and its management did not receive the same level of scrutiny and priority as other areas. 3

Based on these facts, it seems relevant to reconsider the broad role of liquidity for the occurrence of bank distress. So far, most empirical studies about the bank probability of default consider indicators from the CAMELS approach that are computed with accounting data. We suggest to consider the net stable ratio as defined in the Basel III accords in addition to the liquidity ratios from the CAMELS approach in the determination of the bank probability of default. The novelty of this ratio is that, in addition of including the information provided by accounting data on the liquidity profile of banks, it considers the information about the liquidity of financial markets to determine the liquidity of bank assets and liabilities. The aim of this paper is to test if the introduction of the net stable ratio in addition to the liquidity ratios from the CAMELS approach can contribute to improve the prediction of bank financial distress. By using standard Logit model, we test if the net stable funding ratio adds predictive value to models relying on liquidity ratios from the CAMELS approach to explain the bank probability of default. In accordance with the proposals made by the BIS (2009) and given the increasing connections between banks and financial markets, we wonder about the advantage to improve the definition of bank liquidity in order to ameliorate the assessment of bank financial conditions. In particular, we deal with the issue of considering a liquidity ratio that is not only based on accounting data but that also considers the impact of the liquidity of financial markets to define the liquidity of bank assets and liabilities. We may contribute to the strand of the empirical literature on the determinants of individual bank failure as well as to the current debate on new banking regulation proposals since this issue is important to assess the accuracy of improving the definition of liquidity ratios to detect bank distress. Our main results support the use of the net stable funding ratio in addition to the liquidity ratios from the CAMELS approach to detect bank distress. These findings highlight the benefit of improving the definition of bank liquidity as it ameliorates the assessment of bank financial soundness. Given the development of bank market activities, these results shed light on the importance to consider in addition to liquidity ratios based on accounting data, a liquidity ratio for which the liquidity of bank assets and liabilities is defined based on the valuation of assets and the availability of funding on financial markets, for liquidity assessment in banking. The remainder of this paper is organised as follows. In section 2, we present our data set, the issue and the methodology. In section 3, we describe our explanatory variables. In 4

section 4, we comment our results. In section 5, we discuss about some robustness checks. Section 6 concludes. 2. Sample and methodology 2.1. Presentation of the sample Our sample consists of US and European 4 publicly traded commercial banks over the 2005-2009 period. We do our empirical analysis in the context of the most recent financial crisis that is the Subprime crisis (that began in mid-2007) marked by important liquidity shortages. We consider a pre-crisis period of two years to capture the changes that can intervene from a calm period to a period of financial distress. We focus on US and European banks because the required data are available on standard databases to ensure an accurate representativeness of our sample of banks in each country. In addition, US and European banks have been impacted by the Subprime crisis. Finally, we only include listed banks because their balance sheet data are more detailed which allows us to compute the net stable funding ratio that is our main variable of interest. Annual financial statements are extracted from Bloomberg. As our objective is to model the bank probability of default, we need informations about bank default until 2009 but we consider financial statements over the 2005-2008 period. Over this period, we identify 870 listed commercial banks (645 in the US and 225 in Europe). However, the breakdown for loans by category and the breakdown for deposits by maturity that are necessary to compute the net stable funding ratio, are not detailed in Bloomberg or in annual reports for 71 US banks and 18 European banks. Thus, the final sample consists of 781 commercial banks (574 in the US and 207 in Europe). In table 1, we present the bank distribution by country. To deal with the issue of sample representativeness, we verify that on average from 2005 to 2008, the final sample constitutes over 75% of the banking assets of US commercial banks and over 63% of the banking assets of European commercial banks. [Insert Table 1] 4 We use data for European banks from the 20 following countries: Austria, Belgium, Cyprus, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Italy, Liechtenstein, Malta, Netherlands, Norway, Portugal, Spain, Sweden, Switzerland and the United Kingdom. 5

Table 2 presents some general descriptive statistics of our final sample. By considering several key accounting ratios, the data show that banks are on average focused on traditional intermediation activities as loans and deposits account for a large share of bank total assets. Indeed, the average share of total loans in total assets is 68.9% and the average share of total deposits in total assets is 69.2%. In addition, on average, interest income accounts for nearly three quarters of total income (73.2%). However, there is a high heterogeneity across banks as shown by the high standard deviation and the extreme values of each ratio 5. Regarding the quality of bank assets, the average share of total provisions for loan losses in total loans is 0.6%. Moreover, considering profitability, the average return on assets is 0.8%. Lastly, in terms of capitalisation, the average risk weighted capital ratio is higher than the minimum regulatory requirement at 12.7% and the average ratio of Tier 1 capital to total assets is 7.9%. [Insert Table 2] 2.2. The issue and the methodology There is a large strand of the empirical literature that focuses on individual bank failure. The seminal studies (developed in the 1970s and 1980s) consider several empirical methodologies and financial ratios computed from bank balance sheet and income statement consistent with the CAMEL rating approach to explain the bank probability of default (Sinkey, 1975; Altman, 1977; Martin, 1977; Avery and Hanweck, 1984; Barth et al., 1985; Benston, 1985; Gajewsky, 1988; Demirguc-Kunt 1990; Gonzalez-Hermosillo, 1999). Since 1997, a sixth component has been added and the CAMEL approach. The CAMEL approach becomes the CAMELS approach. The additional component S corresponds to the sensitivity to market risk (Bohn et al., 2003). Among all components of the CAMELS rating approach, liquidity is one relevant factor to assess the financial soundness of banks. The liquidity ratios are defined according to two different definitions. The first definition used considers the importance of 5 We notice that the average share of total loans to total assets is significantly higher for US than for European banks (respectively, 69% and 65%). The average share of total deposits to total assets is significantly higher for US banks than for European banks (respectively, 77% and 49%). In addition, the average share of interest income is significantly higher for US banks than for European banks (respectively, 78% and 58%). These specificities may be explained by the differences in regulation in the US from Europe. Indeed, in the US, banking groups are submitted to requirements in terms of segmentation of their activities into several subsidiaries. In addition, US banking groups are allowed to carry out activities closely related to banking, such as investment banking and insurance, only if they are considered as well capitalised by the Federal Reserve (i.e., if they meet the Fed s highest risk-based capital rating). It is the reason why most banking groups are focused on banking business, primarily issuing deposits and making loans. However, in Europe, banking groups are not submitted to such requirements and can more easily develop their market activities. 6

liquid assets such as cash and near cash items, interbank assets, government bonds and trading assets (Bourke, 1989; Molyneux and Thornton, 1992; Barth et al., 2003; Demirgüç-Kunt et al., 2003; Kosmidou et al., 2005; Hadley and Touhey, 2007; Chen et al., 2010). The second definition used considers the importance of bank loan portfolio (Demirgüç-Kunt and Huizinga, 1999; Athanasoglou et al., 2006; Kosmidou et al., 2007; Kosmidou, 2008; Naceur and Kandil, 2009). In recognition of the need for banks to improve their liquidity management following the Subprime crisis, the Basel Committee has developed an international framework for liquidity assessment in banking (BIS, 2009). Among the several guidelines, the Basel III accords include the implementation of the net stable funding ratio. This ratio is defined based on the recent evolutions of the banking industry with the increasing connection between banks and financial markets. It considers the information provided by accounting data on the liquidity profile of banks by including the liquidity mismatch of both sides of bank balance sheet as a whole. Besides, it considers the impact of the liquidity of financial markets, on the valuation of assets and on the availability of funding, to evaluate the liquidity of bank assets and liabilities. Indeed, banks are likely to face too many losses from having to sell some assets at fire sale prices to repay some liabilities that may be claimed at short notice. Thus, banks may be unable to repay such amount of debts, the cash value of assets being too weak. The novelty of this paper consists in considering the net stable ratio as defined in the Basel III accords in addition to the liquidity ratios traditionally used in the CAMELS approach to explain the bank probability of default. We question if the introduction of the net stable ratio in addition to the liquidity ratios from the CAMELS approach can contribute to improve the prediction of bank financial distress. We may contribute to the strand of the empirical literature on the determinants of individual bank failure as well as to the current debate on new banking regulation proposals. In particular, this issue seems relevant in order to outline the opportunity of improving the definition of liquidity indicators to detect bank distress. To address this empirical issue, we use as dependent variable a binary variable that takes the value of 1 at time t if the bank is bankrupt or quasi bankrupt at time t+1 and 0 otherwise 6 (Gonzalez-Hermosillo, 1999). We adopt this approach as most of default events are identified ex post. Furthermore, as the values of some explanatory variables at time t+1 are likely to be impacted by the crisis itself, we delete from the panel all observations at time t+1 6 We consider that a bank has ever defaulted at time t when it is bankrupt or quasi bankrupt at time t+1. Indeed, we assume that its fundamentals have been largely damaged at time t and public intervention has been necessary at time t+1. 7

and so on for banks in default or quasi default at time t+1. The purpose is to avoid feedback effects that are likely to disturb the relationship. Following Gonzalez-Hermosillo (1999) and Bongini et al. (2001, a), we define a bank as bankrupt or as quasi bankrupt at time t+1 if (i) the bank failed; (ii) the bank was acquired by other financial institution on last resort; (iii) the bank s operations were temporarily suspended by the government; (iv) the bank was recapitalized by either the central bank or by an agency specifically created to address the crisis, required a liquidity injection from the monetary authorities, or was nationalised in order to prevent its default (because of its too big to fail status) 7. Table 3 contains the name, the nature and the date of default for each bank that failed or quasi bankrupt in the US and in Europe over the 2005-2009 period. More precisely in Europe, 20 commercial banks over 207 banks are bankrupt or quasi bankrupt (12 banks in 2008 and 8 banks in 2009). In the US, 17 commercial banks (4 in 2008 and 13 in 2009) are bankrupt or quasi bankrupt over 574 banks. [Insert Table 3] Then, we estimate the bank probability of default at time t by using a standard Logit model. We regress the binary dependent variable on a set of explanatory variables that correspond to time t if the bank is bankrupt or quasi bankrupt at time t+1. The purpose is to identify the main factors that have contributed to increase bank financial distress before its bankruptcy or quasi bankruptcy. In this perspective, it is the deterioration of bank fundamentals that explains bank failure and not bank distress that explains the deterioration of bank fundamentals. Hence, the endogeneity problem is minimized. 3. Determinants of bank distress According to our empirical issue and based on previous studies, we consider a set of indicators consistent with the CAMELS approach that are likely to impact the bank probability 7 In practice, we consider information for individual banks from Bloomberg archives on ownership, merger and acquisition history. Following Bongini et al. (2002, b), we use the specific terms like bankrupt, failed, closed, recapitalized, suspended and majority purchase as key words in Bloomberg to identify banks that are failed or quasi bankrupt since the beginning of the Subprime crisis. 8

of default 8. In addition, we consider the net stable funding ratio as defined in the Basel III accords. Furthermore, we include a set of other potentially explanatory variables. 3.1. CAMELS indicators As a proxy of bank capitalisation, we consider the ratio of Tier 1 and 2 capital to total risk weighted assets (T12_RWA). A bank could be more vulnerable when its capital is weaker compared with the volume of its risky assets (Martin, 1977). In this context, bank security buffer could be too weak to absorb losses from bad quality assets. Consequently, we can expect a negative sign for the coefficient of this variable in the determination of the bank probability of default. As a proxy of the quality of bank assets, we consider the ratio of total provisions for loan losses to total loans (PLL_TLO). The higher the ratio, the worst is the quality of bank assets because bank holds provisions as it expects to face losses following defaults on its credit portfolio (Gajewski, 1988; Gonzalez-Hermosillo, 1999; Arena, 2005; Cihak and Poghosyan, 2009). Thus, we can expect a positive sign for the coefficient of this variable in the determination of the bank probability of default. Management efficiency is proxied by the cost to income ratio (M_EFCY) that corresponds to the ratio of total operating expenses to net income as in Sinkey (1975), Gajewski (1988), Cihak and Pogosyhan (2009). Based on this accounting measure, management efficiency corresponds to the ability of managers for minimising costs. Consequently, we conjecture that the deterioration in bank financial soundness with higher production costs is likely to increase bank vulnerability to shocks. Thus, we can expect a positive sign for the coefficient of this variable in the determination of the bank probability of default. As a proxy of bank earnings, we consider the return on assets (ROA) that is the ratio of net income to total assets (Altman, 1977; Arena, 2005). We can expect that the deterioration in profitability could increase the bank probability of default. Consequently, we can expect a negative sign for the coefficient of this variable in the determination of the bank probability of 8 In the existing literature, there is no consensus about the definition of the indicators related to the component S of the CAMELS that proxies bank sensitivity to market risk (i.e., including exposure to interest rate risk, currency risk and equity risk). Thus, we do not introduce an indicator of market risk. However in our study, bank sensitivity to market risk is to some extent proxied by our net stable funding ratio. This indicator considers the state of financial markets to approximate the cash value of assets that is likely to fall following a market collapse. In addition, this indicator considers the liquidity of funding markets to assess the amount of available funding. Following a shock, some fundings are likely to become more volatile. Consequently, the bank faces the risk to be unable to repay some fundings that are suddenly claimed on demand. 9

default. Nevertheless, higher profitability may results from higher risk taking and captures possible gamble for resurrection behaviour. Thus, we can also expect a positive sign for the coefficient of this variable in the determination of the bank probability of default. Consequently, the expected sign for the coefficient of this variable is ambiguous. We further consider alternatively different measures of bank liquidity. Bank liquidity can be measured using liquid assets ratios. We consider the liquid assets to total assets ratio (LA_TA). It is a measure of the maturity structure of the asset portfolio that can reflect excessive maturity unbalances (Cihak and Poghosyan, 2009). The higher is the ratio, the more liquid an institution is considered. Furthermore, the liquid assets to total customer deposits ratio (LA_DEPO) shows to what extent a bank is able to meet unexpected deposit withdrawals with the liquid assets from its balance sheet (Calomiris and Mason, 1997; Gonzalez-Hermosillo, 1999). The higher is the ratio, the more a bank is able to meet unexpected deposit withdrawals with its own liquid assets. In addition, the liquid assets to total customer deposits and short term market funding ratio (LA_DP_STMD) shows the ability of a bank to repay its liabilities that can be claimed at short notice with its cushion of cash and with the assets that can be readily monetised (Said and Saucier, 2003). The higher is the ratio, the more a bank is able to repay its short term liabilities (following unexpected deposit withdrawals or commercial paper roll-offs) with the liquid assets from its balance sheet. Consequently, we can expect a negative sign for the coefficients of these variables in the determination of the bank probability of default. Bank liquidity can also be proxied using loan ratios. We consider the total loans to total assets ratio (LO_TA). It is a measure of the illiquidity of the asset portfolio that can reflect excessive illiquidity and higher exposure to default risk (Arena, 2005). The higher is the ratio, the more illiquid an institution is considered. Moreover, the purpose behind the total loans to total customer deposits ratio (LO_DEPO) is that loans are illiquid, and any deposit run-off would be funded through the sale of securities (Gonzalez-Hermosillo, 1999). The higher is the ratio, the greater may be bank difficulties to face unexpected deposit withdrawals as illiquid loans cannot be monetised. In addition, the total loans to total customer deposits and short term market funding ratio (LO_DP_STMD) shows to what extent a bank holds illiquid loans but has to fund any deposit run-off or commercial paper roll-off through the sale of securities (Gonzalez-Hermosillo, 1999). The higher is the ratio, the more a bank may have difficulties to meet its liquidity requirements at short notice. Consequently, we can expect a positive sign for the coefficients of these variables in the determination of the bank probability of default. 10

3.2. The net stable funding ratio The net stable funding ratio corresponds to the ratio of the required amount of stable funding to the available amount of stable funding. The required amount of stable funding corresponds to the amount of a particular asset that could not be monetised through the sale or the use as collateral in a secured borrowing on an extended basis during a liquidity event lasting one year. The available stable funding corresponds to the total amount of an institution s: (i) capital; (ii) liabilities with effective maturities of one year or greater; and (iii) a portion of stable non-maturity deposits and / or term deposits with maturities of less than one year that would be expected to stay with the institution for an extended period in an idiosyncratic stress event. To calculate the net stable funding ratio, a specific required stable funding factor is assigned to each particular type of asset and a specific available stable funding factor is assigned to each particular type of liability. In appendix 1 (see table A.1), we briefly summarize the composition of assets and liabilities categories and related stable funding factor as defined in the Basel III accords. In table 4, we show the breakdown of bank balance sheet as provided by Bloomberg and its weighting in accordance with the proposals made by the BIS (2009) to calculate the net stable funding ratio. On the asset side, we consider the type and the maturity of assets in line with the definition of the BIS (2009) to put the corresponding weights. On the liability side, we consider the maturity of the several fundings to put the corresponding weights. However, as we have only the breakdown for deposits according to their maturity and not according to the type of depositors, we consider the intermediate weight of 0.7 for stable demand and saving deposits (including all deposits with a maturity of less than one year). [Insert Table 4] Our net stable funding ratio (NSFR) is calculated as follows: In our study, the net stable funding ratio corresponds to the reverse ratio as defined in the Basel III accords. The higher is our ratio, the higher is the required amount of stable 11

funding compared with the available amount of stable funding, and the more a bank is illiquid. The bank may face greater difficulties to meet its current engagements with its current internal liquidity, and it has to immediately obtain unsecured funding or be recapitalised or be rescued by national authorities. Consequently, we can expect a positive sign for the coefficient of this variable in the determination of the bank probability of default. 3.3. Other explanatory variables We introduce a set of other explanatory variables as control variables. We control for bank size due to the too big to fail status of large banks that could lead to moral hazard behaviour and excessive risk exposure. In addition, it captures the impact of complexity in large organisations (i.e., governance conflicts, origination of sophisticated products and complex transactions) that is likely to impact bank stability. As a proxy of bank size, we consider the log of total assets (LN_TA). Consequently, we can expect a positive sign for the coefficient of this variable in the determination of the bank probability of default. Besides, we consider the impact of bank business model through the revenue diversification. It is an alternative measure of bank risk absorption capability that is likely to impact the bank probability of default (Stiroh, 2002; Lepetit et al., 2008). According to the financial theory, lower diversification leads to increase the bank probability of default (Santomero and Chung, 1992; Saunders and Walters, 1994). However, other studies show that higher diversification leads to increase the bank probability of default (Demsetz and Strahan, 1997; De Young and Roland, 2001; Stiroh and Rumble, 2006; Lepetit et al., 2008). As a proxy of bank revenue diversification, we introduce a normalised Herfindalh Hirschman index of concentration on interest versus non interest income (HHI_INC) 9. Normalised Herfindalh Hirschman index varies between 0 and 1. The more the index is closed to 0, the higher is the diversification. Consequently according to the first view, we can expect a positive sign for the coefficient of this variable in the determination of the bank probability of default. According to 9 Following Stiroh (2002), we calculate a Herfindalh Hirschman index to proxy the level of concentration of bank revenue. We split bank revenue into interest and non interest income. Herfindalh Hirschman index (HHI_I) computed as follows: HHI_I = (total interest income / total income)² + (total non interest income / total income)² We calculate normalised HHI_INC as follows: 1 HHI _ INC HHI _ INC = 2 ) 1 1 2 12

the second view, we can expect a negative sign. Thus, the expected sign for the coefficient of this variable is ambiguous. Moreover, the existing empirical literature about individual bank failure outlines the relevance of macroeconomic indicators complementary to microeconomic indicators to explain the bank probability of default (Kaminsky and Reinhart, 1996; Gonzalez-Hermosillo, 1999; Shen, 2004; Festic et al., 2010). All macroeconomic data are extracted from Bloomberg. Economic downturn is considered by many as an important factor to explain the bank probability of default because the quality of bank loans deteriorates when the business cycle is in a downtrend. To highlight the impact of macroeconomic context on bank vulnerability, we consider the annual growth rate of real GDP (GDP_GWT). Consequently, we can expect a negative sign for the coefficient of this variable in the determination of the bank probability of default. In addition, we consider the impact of liquidity pressures on the interbank market as liquidity shortages are likely to disturb the management of bank liquidity and may lead to acute liquidity problems. As a proxy of the liquidity pressures on the interbank market, we consider the spread of the one month interbank rate and the policy rate of the central bank (IBK1M_CB). The higher is the spread, the higher is the one month interbank rate compared with the policy rate of the central bank. In this context, banks charge a high risk premium for lending to other banks at short notice compared with the policy rate of the lender on last resort. Hence, the higher cost of interbank funding may prevent banks to access to these sources of liquidity and increase their probability of default. Consequently, we can expect a positive sign for the coefficient of this variable in the determination of the bank probability of default. Finally, we control for the impact of supervisory regime. Laeven and Levine (2008) and Shehzad et al. (2010) show that it can impact bank risk taking behaviour. Besides, as banking regulation is likely to vary across countries, this variable enables us to control for possible country effects. Based on Shehzad et al. (2010), we compute an index measuring supervisory control (CONTROL) from the World Bank s 2007 Regulation and Supervisory Database (Barth et al., 2007) 10. We expect that under strong supervisory oversight, banks are encouraged to 10 To compute our proxy of supervisory regime, we combine two indicators. The first indicator refers to supervisory agency control and is the total number of affirmative answers to the following questions: (i) Is the minimum capital adequacy requirement greater than 8%? (ii) Can the supervisory authority ask banks to increase minimum required capital in the face of higher credit risk? (iii) Can the supervisory authority ask banks to increase minimum required capital in the face of higher market risk? (iv) Can the supervisory authority ask banks to increase minimum required capital in the face of higher operational risk? (v) Is an external audit compulsory obligation for banks? (vi) Can the supervisory authority force a bank to change its internal organization structure? (vii) Can the supervisory authority legally declare that a bank is insolvent? (viii) Can the supervisory authority intervene and suspend some or all ownership rights of a problem bank? (ix) Can the supervisory authority supersede shareholders rights? (x) Can the supervisory authority remove and replace managers? (xi) Can the supervisory authority remove and replace directors? The second indicator of the supervisory regime measures deposit insurance agency control and is the total number of affirmative answers to the following questions: (i) Can 13

control their risk exposure. Thus, we can expect a negative sign for the coefficient of this variable in the determination of the bank probability of default. In table 5, we present some descriptive statistics of our main explanatory variables for our panel of US and European publicly traded commercial banks over the 2005-2008 period. In table 6, we provide summary statistics of the main determinants of bank distress for US and European commercial banks in default or quasi default versus non failed banks 11. From our mean tests, we can notice that banks in default or quasi default have significantly lower average total risk weighted capital ratio and average return on assets but higher average cost income ratio than the other banks. In addition, banks in default or quasi default have significantly lower ratios of liquid assets to total customer deposits than non failed banks. Moreover, banks in default or quasi default have significantly higher ratios of total loans to total customer deposits and total loans to total customer deposits and short term market funding than the other banks. Last, banks in default or quasi default have significantly higher average net stable funding ratio (NSFR) than non failed banks. [Insert Tables 5 and 6] According to our empirical issue and considering the indicators of bank distress as described above, the bank probability of default is defined by the following equation (subscripts i and t denoting bank and period respectively): α + β cci,t + β aai,t + β mm i,t + β eei,t + β lcli,t Pr ob ( Y = ) = Φ K it 1 (1) + β lbnsfr i,t + β kcvki,t + ε i,t k= 1 Where Φ is the logistic cumulative distribution. C, A, M and E are respectively proxies of bank capital adequacy, quality of assets, management efficiency, earnings and liquidity. L the deposit insurance agency legally declare that a bank is insolvent? (ii) Can the deposit insurance agency intervene and suspend some or all ownership rights of a problem bank? (iii) Can the deposit insurance agency remove and replace managers? (iv) Can the deposit insurance agency remove and replace directors? (v) Can the deposit insurance agency supersede shareholders rights? 11 In our study, we consider the year just before the bank default or quasi default in order to outline what indicators are significant to explain bank default. To compute statistics for banks in default or quasi default, we only keep the figures corresponding to the year just before their default (i.e., 2007 if the bank has defaulted in 2008 or 2008 if the bank has defaulted in 2009) and we delete the figures corresponding to the year of the default and to the years after. Then, over this period (i.e., including the years 2007 and 2008), we compare these figures to those of banks that have never defaulted from 2005 to 2009. 14

corresponds to a liquidity ratio computed based on the CAMELS approach. NSFR corresponds to the net stable funding ratio. CV k is the k th control variable. We estimate the model jointly for US and European banks since they have been impacted by the Subprime crisis 12. The coefficients are estimated by the maximum likelihood using Huber-White robust covariance method. To avoid colinearity problems, we orthogonalise the correlated variables if their introduction disturbs the results of our regressions 13.The quality of the model specification is assessed by Mc Fadden R-square and the likelihood ratio test (i.e., we test the joint significance of regressors by comparing the likelihood of the model with that of a model with only an intercept). Besides, we consider another likelihood ratio test for the contribution of the net stable funding ratio to the predictive value of models relying only on liquidity ratios based on the CAMELS approach (i.e., we test the joint significance of regressors by comparing the likelihood of the model with that of a model without the NSFR as explanatory variable). 4. Results In this paper, we test if the introduction of the net stable ratio as defined in the Basel III accords in addition to the liquidity ratios from the CAMELS approach can contribute to improve the prediction of the bank probability of default. In the CAMELS approach, bank liquidity is proxied by different measures such as liquid asset ratios and loan ratios. Thus, we estimate a standard Logit model (as defined by equation 1) by introducing the net stable funding ratio with alternatively each liquidity ratio from the CAMELS approach (model 1.a to 1.f). Table 7 presents our regression results. [Insert Table 7] Among the liquidity ratios from the CAMELS approach, the ratio of total loans to total assets (LO_TA) and the ratio of total loans to total deposits and short term debts (LO_DP_STMD) are significant to explain the bank probability of default, the other ratios 12 We check the robustness of our results by estimating our model separately for US and European banks. The main conclusions are consistent with those obtained by including all the banks of our sample. For further details, see section 5. 13 In all regressions, we orthogonalise PLL_TLO with ROA and M_EFCY with ROA. In addition, NSFR is correlated at least -0.19 and at most -0.56 with the liquidity ratios from the CAMELS approach. These relatively weak coefficients of correlation emphasize that NSFR includes additional information compared with the liquidity ratios from the CAMELS approach. To check the robustness of our results and deal with potential colinearity problems, in all regressions, we orthogonalise NSFR with each liquidity ratio from the CAMELS approach. The results are consistent with those obtained without orthogonalisation. For further details, see section 5. 15

being not significant. The coefficients of LO_TA and LO_DP_STMD are significantly positive. These results suggest that, in line with the economic theory, the bank probability of default is positively related with the level of illiquid assets (i.e., the size of the loan portfolio). Higher ratios of total loans make banks more illiquid as they have greater difficulties to meet their liquidity requirements at short notice. Any deposit run-offs and commercial paper roll-offs would be funded through the sale of securities. Besides, the coefficient of the net stable funding ratio (NSFR) is significantly positive. This finding highlights the relevance of the liquidity indicator as defined in the Basel III accords to detect bank distress. It emphasizes the importance to consider the liquidity of financial markets to approximate the amount of assets that can be effectively monetised and the availability of external funding to assess bank liquidity. These findings particularly matter in a context where banks and financial markets are more and more connected. More generally, these results outline that liquidity pressures on banks are significantly damaging and tend to make them significantly more fragile following an exogenous and unexpected shock. These findings confirm the relevance of monitoring bank liquidity in order to strengthen the stability of individual banks. Besides, we consider a likelihood ratio test for the contribution of the net stable funding ratio to the predictive value of models relying only on liquidity ratios based on the CAMELS approach (see table 7, LR2). From this test, the introduction of the net stable funding ratio significantly adds predictive value to models that only rely on liquidity ratios from the CAMELS approach. These findings confirm the necessity of improving the definition of bank liquidity as it significantly improves the assessment of bank financial conditions. Given the increasing connection between banks and financial markets, these results shed light on the advantage to consider in addition to liquidity ratios based on accounting data, a liquidity ratio that includes the impact of the liquidity of financial markets - on the valuation of assets and on the availability of funding - to assess the liquidity of bank assets and liabilities and deal with the issue of liquidity assessment in banking. Regarding the additional determinants of the bank probability of default, the coefficient of the total risk weighted capital ratio (T12_RWA) is significantly negative. This result suggests that, in line with the economic theory, the bank probability of default is negatively related with the level of bank capitalization at risk. This finding outlines that the deterioration of bank capitalisation relative to the risk profile of bank assets could be one of the root causes of the Subprime crisis. It confirms the necessity to define a stronger capital base and improve risk valuation models in order to reinforce bank ability to effectively absorb losses when a crisis 16

occurs 14. In addition, the coefficient of the ratio of total provisions for loan losses to total loans is (PLL_TLO) is significantly positive. Consequently, the bank probability of default is inversely related to the quality of bank assets. Assuming that the higher are loan loss provisions, the riskier is the loan portfolio, this result indicates that the deterioration of the quality of the loan portfolio tends to increase the bank probability of default. Furthermore, the coefficient of the cost income ratio (M_EFCY) is significantly positive. Consequently, the bank probability of default is inversely related to the efficiency of bank managers. This finding suggests that lower operating costs and better management efficiency indicate a better likelihood of preventing bank distress. Furthermore, the coefficient of the return on assets (ROA) is significantly negative. Thus, the bank probability of default is negatively related with the level of bank profitability. This finding suggests that banks with good earning profiles are less likely to experience financial distress. Moreover, the coefficient of the proxy of bank size (LN_TA) is significantly positive. Consequently, the bank probability of default is positively related to the size of the bank. This result confirms the necessity to consider the size of banks in order to mitigate moral hazard behaviour of large banks that benefit from their to big to fail status to take excessive risk exposures (US Department of the Treasury, 2009; BIS, 2009). In addition, the coefficient of the spread of the one month interbank rate and the policy rate of the central bank (IBK1M_CB) is significantly positive. Consequently, the bank probability of default is inversely related to the liquidity of the interbank market. The positive sign for the coefficient of this variable indicates that the deterioration of the liquidity of the interbank market tends to increase the bank probability of default. This finding highlights the importance of considering the state of the interbank market for the analysis of individual bank failure. Furthermore, it sheds light on the importance of the effective transmission of monetary policy to the interbank market. Besides, perhaps surprisingly, the proxy of revenue diversification (HHI_INC) does not come out significant in the baseline estimations. In addition, the annual growth rate of real GDP (GDP_GWT) and the index of supervisory regime (CONTROL) are not significant. The low predictive power of these two macroeconomic indicators illustrates to some extent the high degree of economic integration within US and European countries and the fact that many of the banks have operations in more than one country, which limits the ability of country-level macroeconomic variables to explain individual bank distress. 14 For further details about the improvement of the definition of bank capital and of the risk valuation methodologies, see Strengthening the resilience of the banking sector (BIS, 2009) and Financial regulatory reform, a new foundation: rebuilding financial supervision and regulation (US Department of the Treasury, 2009). 17

5. Robustness checks We perform three robustness checks. The results are shown in appendix 1. First, we consider the impact of the colinearity of NSFR with the liquidity ratios from the CAMELS approach. NSFR is correlated at least -0.19 and at most -0.56 with the liquidity ratios from the CAMELS approach. Consequently, we run regressions by orthogonalising NSFR with alternatively each liquidity ratio from the CAMELS approach (see table A2.1). In all cases, the coefficient of NSFR is significantly positive. In addition, the conclusions of the likelihood ratio test for the contribution of the NSFR to the predictive value of models relying only on liquidity ratios based on the CAMELS approach are identical to those previously obtained. Regarding the additional determinants of the bank probability of default, results are consistent with those previously obtained. Second, we run regressions by separating banks according to their economic area in order to examine whether the results are driven by US banks alone as they account for a large share of our sample. For US banks, we remove from the model all macroeconomic indicators (such as GDP_GWT, IBK1M_CB and CONTROL) as their cross-sectional variances are null (see table A2.2 for US banks and table A2.3 for European banks). Our results are consistent with those previously obtained for all liquidity ratios based on the CAMELS approach for European banks. Regarding US banks, only the coefficient of the ratio of total loans to total deposits (LO_DEPO) becomes significantly positive. In addition, our findings are consistent with those previously obtained for the net stable funding ratio for all banks whatever their location. Besides, the conclusions of the likelihood ratio test for the contribution of the NSFR to the predictive value of models relying only on liquidity ratios based on the CAMELS approach are consistent with those previously obtained whatever the location of banks. Third, we examine the robustness of our findings considering an alternative definition of the net stable funding ratio (NSFR). Harvey and Spong (2001) and Saunders and Cornett (2006) emphasise the importance of core deposits for US banks (including demand, saving deposits and time deposits lower than 100 000 USD) as a stable source of funding. Consequently, we consider an alternative definition for stable deposits only for US banks as it is a concept specific to US banks. In this perspective, we modify the denominator of NSFR for US banks by considering the sum of the core deposits and the other stable funding to proxy the available amount of stable funding 15. For European banks, we keep the definition of stable 15 The average share of core deposits to total deposits over the 2000-2008 period is 77% for the US banks included in our sample. However, there is a high heterogeneity (the standard deviation of this ratio being 14%). 18

deposits from the BIS (2009). This alternative ratio corresponds to the core funding ratio (CFR) and is computed as follows for US banks: CFR = 0 * (cash + interbank assets + short term marketable assets) + 0.5 * (long term marketable assets + customer acceptances) + 0.85 * consumer loans required amount of stable funding = + 1 * (commercial loans + other loans + other assets + net fixed assets) core deposits + stable funding 1 * core deposits + 0 * (short term market debts + other short term liabilities) + 1 * (long term liabilities + equity) The higher is CFR, the higher is the required amount of stable funding compared with the core deposits and the other stable funding. The bank may face greater difficulties to meet its current engagements with its current internal liquidity. Consequently, we can expect a positive sign for the coefficient of this variable in the determination of the bank probability of default. We run regressions by replacing NSFR by CFR (see table A2.4). In addition, as for NSFR, we consider the impact of the colinearity of CFR with the liquidity ratios from the CAMELS approach. CFR is correlated at least -0.18 and at most -0.43 with the liquidity ratios from the CAMELS approach. Consequently, we run regressions by orthogonalising CFR with alternatively each liquidity ratio from the CAMELS approach (see table A2.5). Finally, we run regressions separately for US banks as CFR is defined based on the specificities of US banks (see table A2.6). In all cases, the main conclusions are consistent with those previously obtained by considering NSFR. These findings confirm the advantage of improving the definition of bank liquidity to assess bank financial conditions. 6. Concluding remarks Our objective in this paper is to assess the advantage of using the net stable funding ratio as defined in the Basel III accords to detect bank distress. We question if the introduction of the net stable ratio in addition to the liquidity ratios from the CAMELS approach can contribute to improve the prediction of bank financial distress. By implementing a Logit model, we test if the net stable funding ratio adds predictive value to models relying on liquidity ratios from the CAMELS approach to explain the bank probability of default. We consider US and European publicly traded commercial banks over the 2005-2009 period. We may contribute to the strand of the empirical literature on the determinants of individual bank failure as well as to the current debate on new banking regulation proposals since this issue is important to assess the accuracy of improving the definition of liquidity ratios to detect bank distress. 19

Our main results highlight the relevance of considering liquidity to explain the bank probability of default. Besides, our findings point out the relevance of the liquidity indicator as defined in the Basel III accords to detect bank distress. This result emphasizes the importance of considering the liquidity of financial markets to approximate the amount of assets that can be effectively monetised and the availability of external funding to assess bank liquidity. This finding particularly matters in a context where banks and financial markets are more and more connected. More generally, these findings suggest that liquidity pressures on banks are significantly damaging. They tend to make banks more fragile following an exogenous and unexpected shock. These results confirm the relevance of monitoring bank liquidity in order to strengthen their stability. Moreover, by testing for the contribution of the net stable funding ratio to the predictive value of models relying only on liquidity ratios based on the CAMELS approach, we show that the use of the net stable funding ratio adds predictive value to models relying on liquidity ratios from the CAMELS approach. These results shed light on the benefit to consider in addition to liquidity ratios based on accounting data, the net stable funding ratio as defined in the Basel III accords, since it significantly ameliorates the assessment of bank financial conditions. Finally, our findings confirm the necessity of improving the definition of bank liquidity to detect bank distress. In line with the proposals made by the BIS (2009) and given the increasing share of bank market activities, our findings support the recognition of the liquidity of the financial markets to deal with the issue of liquidity assessment in banking. 20