Docstoc

The relationship between bank performance and the Net Stable

Document Sample
The relationship between bank performance and the Net Stable Powered By Docstoc
					          The relationship between bank performance and the Net Stable Funding Ratio.

                               Evidence from the recent financial crisis.




            Laura Chiaramonte*               Barbara Casu                 Roberto Bottiglia

           University of Verona            Cass Business School          University of Verona

                   Italy               City University London, UK              Italy




Abstract
Based on a sample of top-tier international banks, this paper examines the main factors that
determine the bank performance and more specifically investigates whether the new structural
liquidity measure, the Net Stable Funding Ratio (NSFR), is a key determinant of ROA and ROE.
The analysis encompasses two time periods: a pre-crisis period (2003-2006) and a crisis and post
crisis period (2007-2010) and focuses on bank-specific characteristics as well as macroeconomic
and industry-specific factors. The results of the empirical analysis indicate that the NSFR become
significantly only during the crisis and post crisis period and it is positively related to bank
performance. This finding shows that, during the crisis and post crisis period, banks characterized
by values of NSFR below the threshold of 100%, are perceived by supervisors as having a less solid
capital structure, lower rating and higher overall funding costs. Hence, the impact of the new
structural liquidity ratio on bank performance is positive.




Keywords: Net Stable Funding Ratio (NSFR); Basel 3; liquidity risk; bank performance; financial

crisis.

JEL Classification: G01; G21



* Corresponding author: Tel. +39 045 8028610; E-mail address: laura.chiaramonte@univr.it
                                                   1
1. Introduction

The recent financial crisis has shown how rapidly and acutely liquidity risk, in terms of both market

liquidity risk and funding risk, can manifest itself in financial markets and how it can affect banks’

stability and the whole financial system. For these reasons, the Basel Committee with the December

2010 final document (so-called Basel 31) has set the gradual introduction of two internationally

harmonized global liquidity standards for banks: the Liquidity Coverage Ratio (LCR), short-term

rule that will be introduced on 1st January 2015, and the Net Stable Funding Ratio (NSFR), longer-

term structural rule that will be introduced on 1st January 2018. Although their introduction in the

banking system is subject to a period of experimentation and development, it is important to analyze

some possible implications of this discipline. Among the various aspects, the potential

consequences that the compliance of a liquidity rule, expressed in quantitative and measurable

terms, might have on future banks’ profitability are relevant. Indeed, whether the new liquidity

rules, on the one hand, will lead to a strengthening financial stability, on the other hand, will

radically modify the functioning of banks, their relationship with the market and their profitability.

To date, the impact of the new liquidity rules on bank profitability is not clear and necessitates

further investigation. For example, Harle et al. (2010), King (2010, 2012) and Resti (2011)

hypothesized a negative impact on bank performance.

Given the importance of profitability for the stability of the banking industry, this paper, on the

basis of retrospective data, aims to shed some light on the impact of the introduction of the new

liquidity standards proposed by Basel 3. Although the Basel Committee outlines two liquidity

standards, this research focus on the NSFR only, since publicity available information does not

allow an evaluation of the LCR.

More specifically, this paper aims to build upon this recent strand of literature (Pasiouras and

Kosmidou, 2007; Athanasoglou et al., 2008; Kosmidou, 2008; Dietrich and Wanzenried, 2011;

Mohd Said and Hanafi Tumin, 2011) by investigating the determinants of performance, defined
                                                  2
using traditional measure like Return On average Assets (ROA) and Return On average Equity

(ROE), of top-tier international banks, over the period from 2003 to 2010, to understand whether

the new structural liquidity measure, the NSFR, is a key determinant of bank performance. For the

investigation of the determinants of bank performance, the empirical analysis considers not only

liquidity risk measure (the NSFR), but also other bank-specific characteristics of bank performance,

as well as macroeconomic and industry-specific characteristics.

Furthermore, we analyze the impacts of the recent financial crisis on the determinants of bank

performance. To evaluate the impact of the recent financial crisis, we separately consider the pre-

crisis period, 2003-2006, and the crisis and “post” crisis period, 2007-2010.

There is a large literature on the determinants of bank performance (see Molyneux and Thorton,

1992; Goddard et al., 2004; Athanasoglou et al., 2006; Pasiouras and Kosmidou, 2007;

Athanasoglou et al., 2008; Dietrich and Wanzenried, 2011, among others). The existing research

focuses mainly on ratio analysis to explain the performance of banks and includes different proxies

for liquidity. None of the more recent academic studies used the Basel 3 indicators of liquidity risk

in order to investigates whether the new liquidity measures themselves are determinants of bank

performance (defined by ROA and/or ROE). To date, only King (2012) focused on the impact of

NSFR on banks’ performance, however defined by bank net interest margins. In particular, King

(2012) in a study in which he analyzed the impact of NSFR on bank net interest margins, showed

that the strategies to use to meet NSFR, reduce net interest margins by 66 basis points on average,

or 28% of their year-end 2009 values, with the biggest absolute declines for Swiss, French and

German banks.

The paper makes three original contributions to the related literature. The first contribution relates

the analysis of the relationship between bank performance (defined by ROA and ROE) and

liquidity. This is the first academic study that uses the new structural liquidity ratio (the NSFR)

calculated on the basis of the December 2010 Final Document. The second contribution relates the
                                                  3
geographical coverage of sample banks. This study is one of very few which concerns exclusively

with international banks in order to analyze. Related studies have focused mainly on banks of a

specific country (see Guru et al., 1999; Shen et al., 2001; Mamatzakis and Remoundos, 2003;

Kosmidou et al., 2005; Kosmidou, 2008; and Naucer and Kandil, 2009), or on a panel of countries

(see Bourke, 1989; and Mohd Said and Hanafi Tumin, 2011), or on a specific geographic area

(Molyneux and Thorton, 1992; Athanasoglou et al., 2006; and Pasiouras and Kosmidou, 2007).

Only Demirgüç-Kunt and Huizinga (1999) and Shen et al. (2009) have focused on a sample of

international banks.

Finally, this study is one of the first contributions that takes also into account the recent financial

crisis (2003-2010). The literature on the impact of the recent financial turmoil on the determinants

of bank profitability is relatively sparse (see Beltratti and Stulz, 2009; Xiao, 2009; Millon Cornett et

al., 2010; and Dietrich and Wanzenried, 2011). However, none of these study considered liquidity

measures in their analysis.

The results of the empirical analysis can be summarized as follow. First, banks’ performance are

determined principally by internal determinants, especially when they are measure by ROA.

Second, the financial crisis has a significant impact on the internal determinants of bank

profitability. Bank performance is explained by different bank-specific characteristics in the pre-

crisis period and in the crisis years. Third, in the pre-crisis period bank profitability is strongly

influenced by the improvement of the operational efficiency, especially when they are measured by

ROA. Fourth, during the crisis and post crisis period bank performance tends to decrease principally

for banks that in recent years showed one or more of these characteristics: poor quality loan

portfolio, higher leverage, decline in operational efficiency, and/or poor liquidity. Finally, the

NSFR become significantly only during the crisis and post crisis period and it is positively related

to bank performance. During the crisis years, more liquid banks (with NSFR more than 100 per

cent) were perceived as less risky by the authorities and this has determined higher rating and lower
                                                   4
overall funding costs. Hence, the impact of the new structural liquidity ratio on bank performance is

positive.

The remainder of the paper is organized as follows. Section 2 presents a literature review on

determinants of banks’ performance. Section 3 describes the data sample and the dependent and

independent variables used in our analysis, with particular attention to the construction of the

NSFR. Section 4 describes the empirical methodology. Sections 5 and 6 present empirical results

and robustness tests, respectively. Finally, Section 7 summarizes the major findings and concludes

the paper.



2. Literature review on determinants of banks’ performance

Empirical research regarding the profitability of a bank can be classified in three categories. The

first consists of studies that focus on the internal and external determinants of banks’ profitability

(see Short, 1979; Bourke, 1989; Molyneux and Thorton, 1992; Berger, 1995; Demirgüç-Kunt and

Huizinga, 1999; Guru et al., 1999; Shen et al., 2001; Mamatzakis and Remoundos, 2003; Goddard

et al., 2004; Kosmidou et al., 2005; Athanasoglou et al., 2006; Pasiouras and Kosmidou, 2007;

Athanasoglou et al., 2008; Bennaceur and Goaied, 2008; Kosmidou, 2008; Naucer and Kandil,

2009; Shen et al., 2009; Dietrich and Wanzenried, 2011; and Mohd Said and Hanafi Tumin, 2011).

The second consists of studies that examine the impact of regulation on banks’ performance (see

Barth et al., 1997; Rime, 2001; Barth et al., 2003, 2004; Heid et al., 2004; Blum, 2008; Brimissis et

al., 2008; Beltratti and Stulz, 2009; Laeven and Levine, 2009; and Allen et al., 2010). The third

consists of studies that investigate the influence of ownership structure on banks’ profitability (see

Saunders et al., 1990; Altunbas et al., 2001; Berger et al., 2005; Iannotta et al., 2007; Micco et al.,

2007; Beltratti and Stulz, 2009; and Millon Cornett et al., 2010).

Despite the three strands of research listed above, this section focuses only on the first category.

Research on the investigation of the determinants of bank performance are numerous. The study on
                                                  5
the determinants began as early as 1979 when Short (1979) examined the relationship between

profit rate and the bank concentration. Then Bourke (1989) has classified the determinants of bank

profitability into factors internal and external to bank. The internal determinants are factors that are

mainly influenced by a bank’s management decisions and policy objectives. In most studies,

variables such as bank size, credit risk, liquidity risk, capital ratio and operational efficiency are

used as internal factors of banking profitability. The external determinants, both industry-related

and macroeconomic, are variables that reflect the economic and legal environment where the credit

institution operates. This second group of determinants can be distinguish between control

variables, such as cyclical output and inflation, and variables that represent characteristics of bank

market, such as the market concentration.

Further study performed by Athanasoglou et al. (2008) classified the determinants to three specific

aspects: bank-specific characteristics (or internal determinants), macroeconomic determinants and

industry-specific characteristics of bank profitability. These last two are external determinants.

Recent studies extend the research of Athanasoglou et al. by using more determinants (see Shen et

al., 2009; Dietrich and Wanzenried, 2011; and Mohd Said and Hanafi Tumin, 2011). A number of

explanatory variables have been proposed for both categories, according to the nature and purpose

of each studies.

Not all the above-mentioned studies that have investigated the determinants of bank performance

have considered also liquidity risk measure among the bank-specific characteristics (or internal

determinants) of bank performance. Among these, Bourke (1989), Molyneux and Thorton (1992),

Demirgüç-Kunt and Huizinga (1999), Guru et al. (1999), Shen et al. (2001), Mamatzakis and

Remoundos (2003), Kosmidou et al. (2005), Athanasoglou et al. (2006), Pasiouras and Kosmidou

(2007), Kosmidou (2008), Naceur and Kandil (2009), Shen et al. (2009), and Mohd Said and Hanafi

Tumin (2011) have analyzed also the relationship between bank liquidity risk and performance. The

respective empirical studies have focused their analyses or on banks of a specific country, including
                                                   6
Malaysia (Guru et al., 1999), Taiwan (Shen et al., 2001), Greece (Mamatzakis and Remoundos,

2003; and Kosmidou, 2008), the UK (Kosmidou et al., 2005), and Egypt (Naucer and Kandil,

2009), or on a specific geographic area: Europe (Molyneux and Thorton, 1992; and Pasiouras and

Kosmidou, 2007), and South Eastern Europe (Athanasoglou et al., 2006), or on a panel of countries:

Europe, North America and Australia (Bourke, 1989), and Malaysia and China (Mohd Said and

Hanafi Tumin, 2011). Only Demirgüç-Kunt and Huizinga (1999), and Shen et al. (2009) have

focused on a sample of international banks.

None of the other above-mentioned studies that have considered also liquidity risk measure in the

analysis of the determinants of bank profitability, have investigated the impacts of the recent

financial crisis on the determinants of bank performance. To the present the only studies that

analyzing bank performance have also considered the impact of the crisis are Beltratti and Stulz

(2009), Xiao (2009), Millon Cornett et al. (2010), and Dietrich and Wanzenried (2011). However,

none of these studies considered liquidity measures in the analysis. In particular, Beltratti and Stulz

find that large banks with more Tier 1 capital and more deposit financing at the end of 2006

exhibited significantly higher returns during the crisis. Xiao runs qualitative and quantitative

analyses to examine the performance of French banks during 2006-2008. The author concludes that

French banks were not immune but proved relatively resilient to the global financial crisis reflecting

their business and supervision features. Millon Cornett et al. looking at internal corporate

governance mechanisms and the performance of publicly traded U.S. banks before and during the

financial crisis, finds that banks of all size groups suffered bank performance decreases. However,

they find that the largest banks faced the largest losses. Finally, Dietrich and Wanzenried have

examined how bank-specific characteristics, industry-specific and macroeconomic factors affect the

profitability of commercial banks in Switzerland over the period from 1999 to 2009. To evaluate the

impact of the recent financial crisis, they separately consider the pre-crisis period, 1999-2006, and

the crisis years of 2007-2009. Their results provide some evidence that the financial crisis did
                                                  7
indeed have significant impact on the Swiss banking industry and on bank profitability in particular.

Dietrich and Wanzenried provide empirical evidence that, for the Swiss market during the financial

crisis, state-owned banks are more profitable than privately owned banks. They believe that, during

this time of turmoil, state-owned banks were thought of as safer and better banks in comparison to

privately owned institutions. The loan loss provisions relative to total loans ratio, which is a

measure of credit quality, did not have a statistically significant effect on bank profitability before

the crisis. However, the loan loss provisions have significantly increased during the crisis, and this

is reflected in its negative impact on profitability during the crisis years. The yearly growth of

deposits has had a significant and negative impact on bank profitability, and this effect is seen

mainly in the crisis years. Banks in Switzerland were not able to convert the increasing amount of

deposit liabilities into significantly higher income earnings during the recent time of turmoil.

With reference to the authors that have considered also liquidity risk measure in analyzing the

determinants of bank performance, they distinguish themselves not only in terms of their datasets,

time periods, and countries, but also in terms of the type of indicator used to estimate bank liquidity

risk. The ratios used in previous studies include the ratio of liquid assets to total assets (Bourke,

1989; Molyneux and Thornton, 1992; and Guru et al., 1999), the ratio of loans to total assets

(Demirgüç-Kunt and Huizinga, 1999; Mamatzakis and Remoundos, 2003; and Athanasoglou et al.,

2006), the ratio of liquid assets to deposits (Shen et al. 2001), the ratio of liquid assets to customer

and short term funding (Kosmidou et al., 2005), the ratio of net loans to customer and short term

funding (Pasiouras and Kosmidou, 2007; Kosmidou, 2008; Nacer and Kandil, 2009; and Shen et al.,

2009), the ratio of financing gap to total assets (Shen et al., 2009), and the ratio of net loans to

deposits and short-term funding (Mohd Said and Hanafi Tumin, 2011). None of the more recent

academic studies cited used the Basel 3 indicators of liquidity risk in order to investigates whether

the new liquidity measures themselves are determinants of bank performance.



                                                   8
The empirical findings of the authors cited above in terms of the impact of bank liquidity, assessed

using ratios different from the NSFR, on bank performance measured only by ROA, or ROA and

ROE, turned out to be rather heterogeneous. In particular, Molyneux and Thornton (1992), and

Guru et al. (1999) showed that there is a negative relationship between the level of liquidity and

ROA. The same finding is reached by Mamatzakis and Remoundos (2003) but with reference to

ROE. While Bourke (1989) Demirgüç-Kunt and Huizinga (1999), Kosmidou et al. (2005),

Kosmidou (2008) and Shen et al. (2009) found that there is a strong and positive relationship

between liquidity and ROA. Shen et al. (2009) found the same result with reference to the impact of

liquidity risk on bank ROE also. Pasiouras and Kosmidou (2007) concluded that the impact of

liquidity risk on ROA was different for domestic and foreign banks, being positive in the first case

and negative in the second. By contrast, other authors demonstrated that bank liquidity risk had no

effect on either ROA and ROE (Athanasoglou et al., 2006; and Mohd Said and Hanafi Tumin,

2011) or that liquidity risk did not significantly determine ROA or ROE (Naceur and Kandil, 2009).

Conversely, studies that analyzed the impact of liquidity risk, assessed using ratios different from

the NSFR, on bank performance measured by Net Interest Margin (which is net interest income

expressed as a percentage of average earnings assets or total assets in some cases) all pointed to the

existence of an inverse relationship between bank liquidity and this indicator (see Demirgüç-Kunt

and Huizinga, 1999; Shen et al., 2001; Kosmidou et al., 2005; Naucer and Kandil, 2009; and Shen

et al., 2009).

The different impact of liquidity (or an indicator of liquidity risk) on bank performance measured

by ROA and/or ROE is puzzling and therefore necessitates further investigation. In addition, the

relationship linking liquidity, particularly in the form of long term liquidity ratio - Net Stable

Funding Ratio, to ROA and/or ROE is uncertain. Thus, loosely following the strand of literature on

determinants of bans’ performance, this paper empirically assess the main factors that determine the

performance and more specifically investigates whether the new structural liquidity measure, the
                                                  9
NSFR, is a key determinants of a bank performance, building upon the works of Pasiouras and

Kosmidou (2007), Kosmidou (2008), and Dietrich and Wanzenried (2011), among others.



3. Data sample and descriptive statistics


3.1 Data description


The study considers only top-tier international banks (with total assets greater than 10 billion US$)

operating in five areas of specialization: Bank Holding & Holding Companies, Commercial Banks,

Cooperative Banks, Savings Banks, and Real Estate & Mortgage Banks. In our study, we use only

consolidated bank statement (Bankscope consolidation codes: C1 and C22). Subsidiaries were

excluded from the analyses to avoid double counting.

We considered only large banks (by total assets), given that large banks create systemic risk and are

most likely to be subject to supplementary rules or rules that differ from those applicable to other

financial intermediaries.

The final sample is composed of 224 top-tier international banks, 137 of which European (Austria:

7, Belgium: 2, Denmark: 18, France: 3, Germany: 10, Ireland: 4, Italy: 14, The Netherlands: 4,

Norway: 21, Portugal: 5, Spain: 26, Sweden: 9, Switzerland: 4, and United Kingdom: 10), 10 are

Australian, 13 are Canadian, and 64 US banks. Since top-tier banks in the Asiatic countries did not

experience liquidity problems during the recent financial crisis, we decided to not consider such

banks in the study. See Table 1 for sample distribution by specialization in each country.

The overall time horizon comprises the period from 2003 to 2010, the last data available at the time

of this study. This time period is then divided into two sub-periods (see Millon Cornett et al., 2010;

and Dietrich and Wanzenried, 2011). The first sub-period is the period prior to the start of the crisis

(from 2003 to 2006), so-called the ‘pre-crisis period’. The second sub-period, so-called the ‘crisis

and “post” crisis period’, is the period from 2007, the year in which the financial crisis starts, to

                                                  10
2010. This sub-period takes into account the outbreak of the recent financial crisis, began in

summer 2007, and the acute and less acute phases of the crisis.



[Insert Table 1]



3.2 Dependent variable


This paper uses as dependent variable two alternative measure of bank profitability: the Return On

average Assets (ROA) and the Return on Average Equity (ROE). These variables were extracted

with annual frequency from the Bankscope database - Universal Bank Model.

In particular, ROA reflects the ability of a bank’s management to generate profit from the bank’s

assets. ROA is the net profits after tax expressed as a percentage of average total assets. While ROE

indicates the return to shareholders on their equity. ROE is the net profits after tax expressed as a

percentage of average total equities.

Table 2 presents descriptive statistics of ROA and ROE for the sample banks in the pre-crisis

period, and in the crisis and “post” crisis period. ROA and ROE showed a drastic reduction from

the first to the second sub-period, principally due to a significant deterioration in the asset quality of

most sample banks. In particular, the average ROA value of sample banks decreased from 0.99% in

the pre-crisis period to 0.38% in the crisis and “post” crisis period, while the average ROE value

decreased from 14.09% to 4.24%. During the crisis years, negative average ROA and ROE values

were recorded principally by UK, Irish and US banks, due principally to the prevalence of the

Originate to Distribute banking model.



[Insert Table 2]




                                                   11
3.3 Explanatory variables

This section describes the explanatory variables that we used to analyze bank profitability. In line

with previous related studies, they include mainly bank-specific characteristics (or internal

determinants). In addition, we have also considered some common macroeconomic and industry-

specific variables (or external determinants).



[Insert Table 3]



3.3.1 Bank-specific characteristics

Seven bank-specific characteristics are used as internal determinants of bank profitability. They

include: bank size (two variables), liquidity risk, capital, credit risk, efficiency, and funding costs.

The data used to calculated this variables were extracted with annual frequency from the Bankscope

database - Universal Bank Model. The seven internal factors and their hypothesized relationship

(irrespective of the time horizon considered) with the dependent variable (alternately ROA and

ROE) are outlined below and summarized in Table 3.



Size: we use the natural logarithm of a bank’s total assets in millions of Euros (SIZE) to proxy size

and their square (SIZE2) to capture the non-linear relationship. Generally, the effect of a growing

size on profitability has been proved to be positive to a certain level of size (see Eichgreen and

Gibson, 2001). However, for banks that become extremely large, the effect of size could be

negative due to agency costs, the overhead of bureaucratic processes, and other costs related to

managing extremely large banks (see Stiroh and Rumble, 2006; Pasiouras and Kosmidou, 2007).

Hence, the size-profitability relationship may be expected to be non-linear.




                                                  12
On the basis of earlier studies cited above, we expected a double relationship between size and bank

performance: positive, where an increase of bank size resulted in scale economies; negative, where

an increase of bank size and complexity resulted in scale diseconomies.



Liquidity risk: we use the net stable funding ratio (NSFR) as a measure of bank’s liquidity. Unlike

previous related studies, this paper uses as a liquidity risk indicator the new Basel 3 structural rule.

Although the Basel Committee proposed two liquidity measures, this study focused only on the

NSFR, given that public information does not make it possible to assess the liquidity coverage ratio

value. Sound management of structural liquidity would avoid negative effects in the short-term.

We calculated the NSFR for each bank in the sample using the information available in the

Bankscope database. Since the Bankscope database does not contain all the detailed information

specified in the Basel Committee December 2010 Final Document, in this paper we assessed NSFR

values using a simplified version. Caution is therefore required in the interpretation of results. In

particular, the NSFR version used in this paper (see equation 1), is similar to the NSFR assessed by

King (2010)3. There are two fundamental differences between the NSFR used in this paper and that

assessed by King. The first difference relates to Available Stable Funding (ASF) and Required

Stable Funding (RSF) factors. Unlike King, we used the ASF and RSF factors of the Basel

Committee December 2010 Final Document rather than the December 2009 Consultative

Document. The second difference relates to the length of the retail loans used in the calculation of

the NSFR. While King used only retail loans under one year, in this paper we used retail loans

under one year and greater. This is in line with the December 2010 Final Document (Basel 3).



         ASF Equity  Liabs  1 yr  ( StableDeposits  1 yr  90 %)  ( LessStableDeposits  1 yr  80 %)
NSFR        
         RSF   (GovtDebt  5%)  (CorpLoans  1 yr  50 %)  (Re tLoans  85 %)  OtherAsset s


                                                                                             (equation 1)
                                                   13
The numerator of equation 1 measures the sources of ASF and is equal to the sum of funding

instruments weighted to penalize less stable funding sources. Equity and Liabilities with effective

maturities of one year or greater (or Long Term Funding), Liabs>1yr, are the most stable forms of

funding and for this reason they are funded at 100%. Equity and Liabs>1yr are followed by deposits:

Stable Deposits (non-maturity or with residual maturity of less than one year) weighted at 90% and

Less Stable Deposits (non-maturity or with residual maturity of less than one year) weighted at

80%. With reference to deposits, in the Bankscope database there is no distinction between Stable

Deposits and Less Stable Deposits. For this reason, the only way to determine the value of NSFR is

to make an assumption (assumption n. 1). In particular, in this paper Stable Deposits and Less

Stable Deposits were determined using one of the supervisors’ assumptions indicated in King. For

the supervisors, 75% of deposits are stable and 25% are less stable.

In the calculation of the numerator it was not possible to consider the Other preferred shares and

capital instruments in excess of Tier 2 allowable amount having an effective maturity of one year or

greater, and Wholesale funding provided by non-financial corporate customers, sovereign central

banks, multilateral development banks and public sector entity (non-maturity or with residual

maturity of less than one year). The Bankscope database did not contain this balance sheet item.

King also omitted from his calculation of NSFR balance sheet items for which Basel 3 established a

weighting of 100% and 50% respectively.

The denominator measures the sources of RSF and is calculated as the sum of the value of the assets

(Government Debt, Corporate Loans, Retail Loans, and Other Assets) weighted inversely to their

degree of liquidity (see equation 1).

In particular, Government Debt (GovtDebt) is considered very liquid and must only be funded at

5%. Corporate Loans that mature within one year (CorpLoans<1yr) are funded at 50%.



                                                 14
In Bankscope database there is not distinction between Corporate Loans with less than one year to

maturity and Corporate Loans with more than one year to maturity. For this reason, in order to

determine short loans to corporate clients provided by Basel 3 for the calculation of the NSFR, this

paper used one of the supervisors’ assumptions (assumption n. 1) indicated in King. For the

supervisors, 25% of Corporate Loans are less than one year in maturity.

In this paper, Retail Loans less and more than one year in maturity are funded at 85%; for Retail

Loans, we used a simplified hypothesis. Basel 3, unlike equation 1, established for Retail Loans two

different weightings. In particular, the Unencumbered residential mortgages of any maturity and

other unencumbered loans with Basel 2 standardized risk weights of 35% or below are funded at

65%, while the Other loans to retail clients and small businesses having a maturity of less than one

year are funded at 85%.

On the basis of retail loan information in the Bankscope database, it was not possible to distinguish

between the two types of retail loans. For this reason, such loans could not be weighted by the

respective RSF factor. Although Bankscope database divided retail loans in Residential Mortgage

Loans, Other Mortgage Loans and Other Consumer/Retail Loans, for most of the sample banks only

the overall value of retail loans was available, and seldom - if ever - the value of single types of

retail loans. For this reason, it was decided to determine the NSFR by means of a simplified

assumption. Consequently, the Retail Loans used in equation 1 are Retail Loans of any maturity

(short- and long-term), funded, from a prudential perspective, at 85%, in other words the higher of

the two weightings established for retail loans (65 and 85%).

Other Assets are all the other items not included in the previous categories of the denominator. All

remaining assets must be funded at 100%.

Finally, in the calculation of the denominator it was not possible to consider the Highly rated

corporate and covered bonds, and the Off balance sheet exposures: undrawn amount of committed



                                                 15
credit and liquidity facilities. The Bankscope database does not contain such assets, that Basel 3

would fund at 20 and 5%, respectively.

Table A in the Appendix describes the construction of the NSFR used in this paper on the basis of

Bankscope database information. Table A shows also the two supervisors’ assumptions and the

simplified hypotheses used in this paper. Furthermore, Table A compares the balance sheet items

composing the ASF and RSF described in the Basel Committee December 2010 Final Document

and the Bankscope database items used in this paper to calculate the simplified version of the

NSFR.

For the relationship between NSFR and ROA and NSFR and ROE we expected a double sign:

negative or positive. The relationship can be interpreted negatively when longer funding leads to an

increase in interest costs and a decrease in net interest income, lowering operating profit and net

income. On the other hand, the relationship can be interpreted positively when banks with more

medium-long term than short-term funding are perceived by supervisors as having a more balanced,

solid capital structure, higher rating and lower overall funding costs. In this case, an increase in

NSFR would correspond to an increase in ROA and ROE.

The fact that the relationship between NSFR and ROA and between NSFR and ROE can be

interpreted in two different ways, either negatively or positively, confirms the mixed empirical

results of the related literature (see Section 2).



Capital: we use the ratio of equity to total assets (E/TA) as a measure of bank’s capital, in the same

way as the majority of related studies.

Bourke (1989), Molyneux and Thornton (1992), Demirgüç-Kunt and Huizinga (1999), Barth et al.

(2003), Demirgüç-Kunt et al. (2003), Kosmidou et al. (2005), Pasiouras and Kosmidou (2007),

Athanasoglou et al. (2008), Kosmidou (2008), and Shen et al. (2009) found a positive relationship

between this capital ratio and bank performance. The authors explain this relation with the
                                                     16
observation that a high ratio would imply low leverage, and hence a lower need for external

funding, lower risk, and therefore higher bank profitability. Well-capitalized banks face a lower risk

of going bankrupt, which reduces their cost of funding.

Dietrich and Wanzenried (2011) found a negative relationship between equity to total asset and

bank profitability measure by ROA during the financial crisis 2007-2009. One of the main reasons

for this relation is that safer banks in Switzerland were attracting additional saving deposits during

the crisis. However, they were not able to convert the substantially increasing amount of deposits

into significantly higher income earnings as the demand for lending decreased in the crisis period.

On the basis of the earlier studies cited above, we expected a double relationship: positive or

negative.



Credit risk: we used the loans loss provision to total loans ratio (LLP/TL) to proxy the credit risk in

the same way as some related studies.

With reference to relationship between credit risk and bank profitability, previous studies have

showed a negative relationship (see Bourke, 1989, Molyneux and Thornton, 1992, Athanasoglou et

al., 2008, and Dietrich and Wanzenried, 2011, among other). These authors find a negative and

significant relationship between the level of risk and profitability. This result might reflect the fact

that financial institutions that are exposed to high-risk loans also have a higher accumulation of

unpaid loans. These loan losses lower the returns of the affected banks. Theory suggests that

increased exposure to credit risk is normally associated with decreased firm profitability and, hence,

we expected a negative relationship between loans loss provision to total loans ratio and bank

performance.



Efficiency: we use the cost to income ratio (CIR) as a proxy of banks’ operational efficiency.

Among others, Pasiouras and Kosmidou (2007), and Dietrich and Wanzenried (2011) find that
                                                  17
better efficiency (lower ratio) is associated with higher bank profitability. Thus, a negative sign

between cost income ratio and banks’ profitability is expected.



Funding costs: we use the ratio of interest expenses to average total deposits (IE/ATD) as a proxy

of funding costs of a bank. This ratio has been considered for the first time between the explanatory

variables by Dietrich and Wanzenried (2011). Following prior research of Dietrich and Wanzenried

a negative relationship between interest expenses to average total deposits and bank profitability is

hypothesized.



3.3.2 Macroeconomic variables

In addition to the bank-specific characteristics described above, our analysis includes two

macroeconomic variables (or external determinants): the gross domestic product (GDP) growth and

the inflation growth. This variables were extracted with annual frequency from the World Economic

Outlook Database (International Monetary Fund, IMF). The two macroeconomic factors and their

hypothesized relationship (irrespective of the time horizon considered) with the dependent variable

(alternately ROA and ROE) are outlined below and summarized in Table 3.



GDP growth: GDP is among the most commonly used macroeconomic indicators and it is a

measure of total economic activity within an economy. In this study we used annual percent change

of GDP (GDPC).

A positive relationship between this variable and bank performance was to be expected according to

the literature on the association between economic growth and financial sector performance (see

Bourke, 1989, Molyneux and Thornton, 1992, Demirguc-Kunt and Huizinga, 1999, Athanasoglou

et al., 2008, and Dietrich and Wanzenried, 2011, among others).



                                                 18
Inflation growth: another important macroeconomic variable considered in this study was inflation.

In particular, we considered annual percent change of inflation (INFC).

The relationship between the inflation rate and bank performance is ambiguous and depends on

whether or not inflation is anticipated (see Perry 1992). An inflation rate fully anticipated by the

bank’s management implies that banks can appropriately adjust interest rates in order to increase

their revenues faster than their costs and thus acquire higher performance. On the contrary,

unanticipated inflation could lead to improper adjustment of interest rates and hence to the

possibility that costs could increase faster than revenues. This will consequently have a negative

impact on bank performance. Most related studies (see, for example, Bourke 1989, Molyneux and

Thorton 1992) observe a positive relationship between inflation and bank performance. Thus we

expected a positive relationship between this variable and bank performance: an increase in

inflation should increase bank performance.



3.3.3 Industry-specific characteristic

Finally, in our analysis we considered one industry-specific characteristic (or external determinant):

the bank market concentration.

The data used to calculated this variable were extracted with annual frequency from the Bankscope

database - Universal Bank Model. This factor and its hypothesized relationship (irrespective of the

time horizon considered) with the dependent variable (alternately ROA and ROE) are outlined

below and summarized in Table 3.



Bank market concentration: in order to measure each country’s degree of banking system

concentration, we determined the concentration ratio 3 (CR3). This ratio is calculated as the total

assets held by the three largest banks (operating in five areas of specialization: Commercial Banks,

Cooperative Banks, Savings banks, Bank Holding & Holding Companies, and Real Estate &
                                                 19
Mortgage Banks) divided by the total assets of all banks (Commercial Banks, Cooperative Banks,

Savings banks, Bank Holding & Holding Companies, and Real Estate & Mortgage Banks)

operating in each country.

The previous studies showed that the relationship between bank performance and bank

concentration is not univocal. The structure-conduct-performance (SCP) hypothesis, showed that

banks in highly concentrated markets tend to collude and therefore earn monopoly profits. Bourke

(1989) and Molyneux and Thornton (1992) showed a positive and statistically significant

relationship between the bank concentration ratio and the profitability of a bank. The results of

these authors are consisted with the traditional structure-conduct-performance paradigm. In

contrast, the results of Demirguc-Kunt and Huizinga (1999) and Staikouras and Wood (2004)

indicated a negative but statistically insignificant relationship between bank concentration and bank

profits. Likewise, the estimations by Berger (1995) and Mamatzakis and Remoundos (2003)

contradict the structure-conduct-performance hypothesis.

On the basis of the earlier studies cited above, we expected a double relationship: positive or

negative.



Table 4 reports descriptive statistics relating to the internal and external determinants of bank

performance of the sample banks for the pre-crisis period (2003-2006), and the crisis and “post”

crisis period (2007-2010).

Unlike the two profitability indicators (ROA and ROE), explanatory variable values did not change

significantly from the pre-crisis period to the crisis and “post” crisis period. The only exception was

the average value of LLP/TL, CIR and GDPC.

In particular, with reference to bank-specific characteristics, the average values of SIZE, NSFR,

EQ/TA, and IE/ATD, remained substantially unchanged for almost all sample banks from the pre-

crisis period to crisis and “post” crisis period. Table 4 indicates that SIZE grew moderately from
                                                  20
11.72 in the pre-crisis period to 11.79 in the crisis and “post” crisis period. Therefore, on average

the total assets of sample banks showed a slight tendency to increase. The average NSFR value of

sample banks increased from 85.89% to 87.19%. Significant differences emerge looking at the

average value of NSFR of the various bank specializations considered. Table B in the Appendix

shows that the banks with structural liquidity furthest from minimum requirements, regardless of

their geographical area, belonged to the category of Bank Holding & Holding Companies or

Commercial Banks; in comparative terms, the total assets of these banks were higher, well

diversified and internationally active. Underlying the NSFR values significantly below minimum

requirements typically observed in banks with the asset configurations described above, there

appears to be a tendency towards a high degree of maturity transformation. Banks that are below the

100% required minimum have until 2018 to meet the standard and can take a number of measures

to do so, including lengthening the term of their funding, reducing maturity mismatching, or scaling

back activities which are most vulnerable to liquidity risk in periods of stress.

Conversely, the sample banks with liquidity structures in line or very close to the new liquidity

requirement NSFR, are those belonging to the specializations Cooperative Banks, Savings Banks,

or Real Estate & Mortgage Banks; these banks operate on a smaller geographical scale, and so have

lower total assets, high capitalization and capital requirements in line with traditional standards.

Unlike the banks described in the preceding paragraph, they are generally characterized by NSFR

values above minimum requirements due to lower maturity transformation.

The average value of E/TA showed a slight increase from 7.28 per cent in the pre-crisis period to

7.64% in the crisis and “post” crisis period. The reduction of financial leverage of sample banks is

due to the deleveraging phenomenon showed during the crisis years.

The average IE/ATD value of sample banks decreased from 3.41% in the pre-crisis period to 2.80%

in the crisis and “post” crisis period.



                                                   21
Conversely, the average value of LLP/TL and CIR showed a significant variation from the sub-

period to the second sub-period. In particular, LLP/TL grew from 0.03% to 1.23%. This growth was

due to the deterioration of credit quality in a number of banking systems during the recent financial

crisis. The operational efficiency showed a worsening during the crisis years due more to the

decline in operating income than the concomitant increase in operating costs. The average value of

CIR grew from 60.10% in the pre-crisis period to 65.09% in the crisis and “post” crisis period.

With reference to the macroeconomic factors, the average value of GDPC showed a considerable

decline from the pre-crisis period (+2.79%) to the crisis and “post” crisis period (+0.21%) due to

the hoarding effect of a contraction in the main determinants of demand. Conversely, the average

value of INFC showed a tendency to decrease from the first sub-period (+2.37%) to the second sub-

period (+1.98%).

Finally, the macroeconomic variable CR3 showed a limited tendency to increase, from 42.80% in

the pre-crisis period to 43.97% per cent in the crisis and “post” crisis period. The increase in

banking system concentration in the period 2007-2010 was principally due to M&A operations

justified by the crisis to avoid bailouts.



[Insert Table 4]




Finally, correlation coefficients were calculated between internal and external variables and ROA

and ROE respectively to show the relationship between the dependent variable and each

explanatory variable (see Table 5). In particular, the results of correlations made it possible to verify

whether the hypothesized relationships between dependent variables and each explanatory variable

were correct (see Table 3). Table 5 shows that all sign hypothesized in Table 3 are respected.



                                                   22
[Insert Table 5]



4. Empirical methodology

To determine the main factors of performance of top-tier international banks, and in particular

whether the Net Stable Funding Ratio is a key determinant, we follow panel data regression. We

specify the following generic model:



BankPerfit = α + β1(BankCh)it + β2(MacroVar)it + β3(IndCh)it +d crisis +εit                (2)



where BankPerf is bank performance of ith bank at time t. In our study, it is Return On average

Assets (ROA) and Return On average Equity (ROE). Each regression was assessed using alternately

ROA and ROE as the dependent variable. BankCh, MarcoVar, BankSp are bank-specific

characteristics, macroeconomic variables and industry-specific characteristic, respectively. Banks

characteristics include bank size (SIZE and SIZE2), liquidity risk (NSFR), capital (E/TA), credit

risk (LLP/TL), efficiency (CIR), and funding costs (ID/ACD). Macroeconomic variables include

annual percent change of gross domestic product (GDPC), and annual percent change of inflation

(INFC). Industry-specific characteristic is three-bank concentration ratio (CR3). dcrisis is the

dummy variable that identifies the outbreak of the recent financial crisis (from 2007) and εit is the

error.

In the first instance, the regressions were conducted covering the entire time horizon (2003-2010); a

first regression including only the variables; a second one including the variables plus the dummy

crisis.

Subsequently, to determine whether the relationship between bank performance and the variables

considered changes with varying macroeconomic and financial conditions, two further panel


                                                 23
regressions were performed: one on the period preceding the crisis (2003-2006), and one on the

crisis and “post” crisis period (2007-2010).




5. Results

Table 6 summarizes the empirical results for our two profitability measured considered: ROA and

ROE. The first two columns report the results obtained considering the entire time horizon (2003-

2010) without and with the dummy crisis. In order to investigate the impact of the recent financial

crisis on the banks’ profitability determinants, we further split up the sample: columns three refers

to the period before the crisis (2003-2006), while columns four reports the estimates for the crisis

and “post” crisis period (2007-2010).

With reference to the results for our first profitability measure ROA, in both the first two

regressions (without and with the dummy crisis) the final sample consisted of 1166 observations for

224 banks. Table 6 indicates that the variables considered explain 60% of bank profitability (in

terms of adjusted R2); and that the factors considered and the dummy crisis together explain nearly

61% of bank profitability (in terms of adjusted R2).

From the first panel regression of table 6 it emerges that the bank profitability measured by ROA is

explained mainly by the following factors: the liquidity ratio (NSFR), the capital ratio (E/TA), the

measure of credit quality (LLP/TL) and the operational efficiency measure (CIR). The annual

percent change of GDP (GDPC) is also a significant variable but it is the least significant among all.

All these significant explanatory variables have the expected sign.

From the second regression panel it emerges that the dummy crisis is significant and therefore

indicates that the crisis was a relevant event in the relationship between bank performance,

measured by ROA, and the explanatory variables, as expected.




                                                 24
To understand whether the relationship between bank profitability and the variables considered

changes with varying macroeconomic conditions, two further panel regressions were conducted,

one on the sub-period 2003-2006 and the other on the sub-period 2007-2010. The final sample

consisted of 418 observations for 166 banks in the pre-crisis period and 748 observations for 219

banks in the crisis and “post” crisis period (see table 6).

Table 6 shows some significant differences between the estimation results of the different time

periods. In the period preceding the crisis, the bank profitability, measured by ROA, is explained

principally by the cost to income ratio (CIR). In the first sub-period, this ratio is the most

explanatory variable. This result is consistent with those of much of the previous empirical banking

literature (see Goddard et al., 2004; Pasiouras and Kosmidou, 2007; Athanasoglou et al., 2008,

among others), suggesting that efficiency is likely to be a more important determinant of

performance. Moreover, the annual percent change of Inflation (INFC) and concentration ratio three

(CR3) are significant variables but they are less significant. The three significant variables explain

89% of the bank profitability (in terms of adjusted R2). All explanatory variables respect the sign

predicted. In particular, table 6 shows a negative relationship between CIR and ROA. The more

efficient a bank is the higher is its profitability. Operationally efficient banks are more profitable

than banks that are less operationally efficient. This results meets our expectation and stands in line

with the results of Dietrich and Wanzenried (2011). In the pre-crisis period bank profitability is

strongly influenced by the improvement of the operational efficiency. In particularly favorable

macroeconomic conditions, such as that which characterized the years 2003-2006, CIR decreased

mainly due to an increase in operating income. In presence of a highly competitive banking market,

banks to survive have reduced the practice of traditional intermediation in favor of those with

higher value added. Table 6 shows also a positive relationship between INFC and ROA. This

implies that during the pre-crisis period the levels of inflation were anticipated by our sample banks.

This gave them the opportunity to adjust the interest rates accordingly and consequently to earn
                                                    25
higher profits. This finding is consistent with the results of Molyneux and Thorton (1992),

Pasiouras and Kosmidou (2007) and Athanasoglou et al. (2008).

The impact of the market structure, approximated by the CR3, shows a positive effect on bank

profitability in the pre-crisis period (see table 6). This results confirms those of Bourke (1989),

Molyneux and Thorton (1992) and Dietrich and Wanzenried (2011), supporting the SCP

hypothesis.

It also emerges that, as expected, in the period before the crisis, the liquidity risk did not determine

ROA. The NSFR is not a significant variable in the pre-crisis period. Banks have been affected by

liquidity problems, both from the asset side (market liquidity risk) and the liability side (funding

risk), once the financial crisis occurred.

In the crisis and “post” crisis period, the number of significant explanatory variables significantly

increased with respect to the previous period. This implies that the financial crisis did indeed have a

significant impact on the determinants of bank profitability. Therefore, the determinants of bank

profitability, measured by ROA, tends to vary as economic and financial conditions vary. This is in

line with the finding of Dietrich and Wanzenried (2011), who find that the determinants of bank

profitability vary strongly across time.

Table 6 also showed that in both periods the internal factors are the main determinants of banks’

performance, measured by ROA. Indeed, from the panel regression regarding the ROA, it emerges

that the internal determinants have the highest degree of significant. This result confirms those

reported by Kosmidou (2007), who finds that individual bank characteristics explain a substantial

part of the variation in bank ROA. The importance of bank’s management decisions and policy

objectives (related to internal determinants) are confirmed by the new prudential regulation (so-

called Basel 3) issued by the Basel Committee after the acute phase of the crisis. These rules are

finalized to radically modify the functioning of banks, their relationship with the market and their

profitability.
                                                  26
In particular, table 6 shows that, during the second sub-period (2007-2010), the bank profitability,

measured by ROA, is explained mainly by the following factors: equity to total asset (E/TA), loan

loans provision to total loans (LLP/TL), cost to income ratio (CIR) and net stable funding ratio

(NSFR). CIR is the only explanatory variable significant in all periods considered. The bank market

concentration variable (CR3) is also a significant variable but it is the least significant among all.

The significant explanatory variables together explain nearly 58% of the bank profitability (in terms

of adjusted R2). All the variables showed the expected sign. In particular, as predicted, E/TA and

NSFR had a positive relationship with ROA. LLP/TL, CIR and the CR3 showed a negative

relationship with ROA. An analysis of the signs of LLP/TL and E/TA reveals that bank

performance tends to decrease principally for banks that in recent years had a poor quality loan

portfolio and/or higher leverage. Both ratios were significant determinants of the ROA only during

the financial crisis 2007-2010. This outcome confirms some of the findings of the study by Dietrich

and Wanzenried (2011), according to which E/TA and LLP/TL do not impact on bank profitability

measured by ROA before the crisis. However, in the crisis years, Dietrich and Wanzenried (2011)

found a negative relationship between E/TA and ROA rather than positive as in our case.

During the crisis and “post” crisis period, CIR showed a significant negative impact on ROA as in

the previous sub-period. However, during the crisis years (2007-2010) the deterioration of bank

performance is also attributable to a decline in operational efficiency due more to the decline in

operating income than the concomitant increase in operating costs. After the worst of the crisis,

with the decreasing profitability, banks have been urged to implement policies to improve

efficiency in the cost structure.

Furthermore, our results show that the liquidity measured considered, the NSFR, is a key

determinants of the bank profitability only in the crisis and “post” crisis period. This may reflect the

increased attention paid by regulators to bank liquidity only in the crisis years. Before the recent

financial crisis, the liquidity risk has long been underestimated since they wrongly considered it
                                                  27
unable to threaten the stability of the banks. Table 6 shows a positive relationship between NSFR

and ROA. It means that decreasing values of NSFR determines a decrease and not an increase of

performance. This result shows that, during the crisis and “post” crisis period, less liquid banks, that

is characterized by values of NSFR below the threshold of 100% due to medium-long term funding

less pronounced, were perceived by supervisors as having a less solid capital structure, lower rating

and higher overall funding costs. And vice versa, during the crisis years, more liquid banks (with

NSFR more than 100%) were perceived as less risky by the authorities and this has determined

lower funding costs. Hence, the impact of the new structural liquidity ratio on bank performance is

positive. This result seems to clarify the doubts on the impact on the impact of the NSFR on the

bank performance, since, for example, according to Harle et al. (2010) and Resti (2011) the

introduction of the new structural liquidity ratio, whether, on the one hand, it will lead to a

strengthening if financial stability, on the other hand, could put banking profitability under pressure.

King (2010), in a study in which he analyzed the possible impact of higher liquidity requirements

on bank lending spreads, reached the same conclusion. In particular, King point out that banks may

wish to offset the decline in ROE by raising lending spreads.

In the crisis and “post” crisis period, CR3 shows a negative relationship with ROA, unlike the

previous sub-period. The concentration affects bank profitability negatively in the crisis years.

Hence, this result finds no evidence to support the SCP hypothesis. This outcome is in accordance

with Dietrich and Wanzenried (2011), which claim that the impact of the market structure seems to

have a significant and positive effect on bank profitability before the crisis, but not thereafter.

With reference to SIZE, table 6 shows that bank size variable is not significant factor of ROA in all

periods considered. This outcome confirms those of Athanasoglou et al. (2008) and Mohd Said and

Hanafi Tumin (2011).

Table 6 reports the panel regression results for our second profitability measure Return On average

Equity (ROE). Again, we estimate the model for the entire time period considered (without and
                                                   28
with the dummy crisis), and then separately for the two subsamples pre-crisis and crisis and “post”

crisis period.

Overall, the results of these panel regressions confirm to a large extent the above-discussed key

results. Table 6 shows that ROE, like ROA, is for the most part explained by the operational

efficiency, the credit quality measure and the liquidity ratio. However, in the three time periods

considered there are some differences from ROA. The capital ratio (E/TA) is a determinant of ROE

also in the pre-crisis period, but not in the whole period (without and with the dummy crisis). With

reference to ROE, E/TA shows a negative sign in the pre-crisis period and a positive sign in the

crisis and “post” crisis period. In a favorable context, such as the pre-crisis period, high leverage

levels had no negative effect on bank performance. Low values of E/TA (hence high leverage) had

a negative impact on ROE only when the financial crisis occurred.

Moreover, the measure of funding costs (IE/ATD) is determinant only of ROE. In particular,

funding costs show a significantly negative impact on ROE both in the whole period and in the

crisis and “post” crisis period. Banks that raise cheaper funds are more profitable. This outcome

confirms the findings of the study by Dietrich and Wanzenried (2011) only with reference to the

whole period.

Finally, bank size variables (SIZE and SIZE2) are determinants only of ROE and only in the crisis

and “post” crisis period. The two variables show positively and negatively signs, respectively. This

outcome confirms those of Dietrich and Wanzenried (2011), according to which larger banks in

Switzerland were less profitable than smaller banks during the past three years of the financial

crisis.



[Insert Table 6]



6. Robustness Tests
                                                 29
In order to identify the stability of the coefficients and their significance, we test the model also

only on the bank-specific determinants (see Kosmidou, 2008). To test this hypothesis, a number of

further regressions were carried out (see Table 7). Both regressions with ROA and ROE

alternatively used as dependent variable and balance sheet data as explanatory variables were

conducted on the overall period considered (2003-2010) and on all sub-periods.

Comparison of the results of the panel regressions in Table 7 with those of the panel regressions of

Table 6, which considered both internal and external determinants, reveals that the coefficients are

stable in all periods considered. Indeed, the significant explanatory variables in Table 7 are the

same of the Table 6. Moreover, adjusted R-square values in Table 7 are very similar to those of

Table 6. This means that banks’ performance are explained mainly by bank-specific characteristics.



[Insert Table 7]



7. Conclusions

This paper investigates the relationship between bank performance (defined using measure like

ROA and ROE) and the Net Stable Funding Ratio. Based on a sample of top-tier international

banks, we examines the main factors that determine the bank performance and more specifically

investigates whether the new structural liquidity measure, the NSFR, is a key determinant of ROA

and ROE. We focuses on bank-specific characteristics as well as industry-specific and

macroeconomic factors. The analysis was conducted over the period from 2003 to 2010. This was

then subdivided into two sub-periods: the pre-crisis period (2003-2006) and the crisis and “post”

crisis period (2007-2010).

Results indicate that in both periods, bank-specific characteristics are the main determinants of

banks’ performance, given that they have the highest degree of significant, especially when bank

performance are measured by ROA. The internal determinants that explain banks’ performance
                                                 30
tends to vary as economic and financial conditions vary. Indeed, the financial crisis was a relevant

event in the relationship between bank performance and internal determinants. In the pre-crisis

period bank profitability is strongly influenced by the improvement of the operational efficiency,

especially when they are measured by ROA. Moreover, cost to income ratio is the only explanatory

variable significant in all periods considered.      During the crisis and post crisis period bank

performance tends to decrease principally for banks that in recent years showed one or more of

these characteristics: poor quality loan portfolio, higher leverage, decline in operational efficiency,

and/or poor liquidity. As expected, in the period before the crisis, the NSFR did not determine

banks’ profitability. The new structural liquidity ratio become significantly only during the crisis

and post crisis period and it is positively related to bank performance. It means that decreasing

values of NSFR determines a decrease and not an increase of performance. This result shows that,

during the crisis and “post” crisis period, less liquid banks, that is characterized by values of NSFR

below the threshold of 100 per cent due to medium-long term funding less pronounced, were

perceived by supervisors as having a less solid capital structure, lower rating and higher overall

funding costs. The compliance of the NSFR threshold does not appear to put banks’ profitability

under pressure. Hence, the impact of the strengthening of the financial stability, deriving from the

introduction of the new structural liquidity measure of Basel 3, on the banks’ performance seems to

be positive. This result of the analysis is of particular interest to both academic and policy makers as

they contribute to the current debate on banking sector reforms and how to reconcile the needs for

financial stability without imposing too high a cost on banks’ profitability.




Acknowledgements



Notes
                                                  31
1. See Basel Committee on Banking Supervision (BCBS, 2010).

2. C1 is the Bankscope code that identifies a consolidated statement without an unconsolidated

statement. C2 is the Bankscope code that identifies a consolidated statement with an unconsolidated

statement.

3. See King (2010, pp. 10 and 23).



References

Allen, W., K.K. Chan, A. Milne, and S. Thomas. 2010. Basel III: Is the Cure Worse Than the

Disease. SSRN Working Papers, September.

Altunbas, Y., L. Evans, and P. Molyneux. 2011. Bank Ownership and Efficiency. Journal of

Money, Credit and Banking 33, no. 4: 926-954.

Athanasoglou, P.P., M.D. Delis, and C.K. Staikouras. 2006. Determinants of Bank Profitability in

the South Eastern European Region. Bank of Greece Working Paper no. 47.

Athanasoglou, P.P., S.N. Brissimis, and D.D. Matthaios. 2008. Bank-specific, industry-specific and

macroeconomic determinants of bank profitability. Journal of International Financial Markets,

Institutions & Money 18, no. 2: 121-136.

Barth, J.R. D.E. Nolle, and T.N. Rice. 1997. Commercial Banking Structure, Regulation and

Performance: An International Comparison. Managerial Finance 23, no. 11: 1-39.

Barth, J.R., D.E. Nolle, T. Phumiwasana, and G. Yago. 2003. A Cross-Country Analysis of the

Bank Supervisory Framework and Bank Performance. Financial Markets, Institutions &

Instruments 12: 67-120.

Barth, J.R., G. Caprio Jr., and R. Levine. 2004. Bank regulation and supervision: what works best?

Journal of Financial Intermediation 13, no. 2: 205-248.

Basel Committee on Banking Supervision (BCBS, 2010) Basel III: International frame work for

liquidity risk measurement, standards and monitoring, December.

                                                32
Beltratti, A., and R. Stulz. 2009. Why Did Some Banks Perform Better During the Credit Crisis? A

Cross-Country Study of the Impact of Governance and Regulation. ECGI’s Finance Working Paper

no. 254.

Bennaceur, S., and M. Goaied. 2008. The determinants of commercial bank interest margin and

profitability: Evidence from Tunisia. Frontiers in Finance and Economics 5, 106-130.

Berger, A. 1995. The profit-structure relationship in banking: tests of market-power and efficient-

structure hypotheses. Journal of Money, Credit and Banking 27, no. 2: 404-431.

Berger, A.N., G.R. Clarke, R. Cull, L. Klapper, and G.F. Udell. 2005. Corporate governance and

bank performance: a joint analysis of the static, selection, and dynamic effects of domestic, foreign,

and state ownership. Journal of Banking and Finance 29, no. 9: 2179-2221.

Blum, J.M. 2008. Why ‘Basel II’ may need a leverage ratio restriction. Journal of Banking &

Finance 32, no. 8: 1699-1707.

Bourke, P. 1989. Concentration and Other Determinants of Bank Profitability in Europe, North

America and Australia. Journal of Banking and Finance 13, no. 1: 65-79.

Brissimis, S.N., M.D. Delis, and N.I. Papanikolaou. 2008. Exploring the nexus between banking

sector reform and performance: evidence from newly acceded EU countries. Journal of Banking

and Finance 32, no. 12: 2674-2683.

Demirgüç-Kunt, A., and H. Huizinga. 1999. Determinants of Commercial Bank Interest Margins

and Profitability: Some International Evidence. World Bank Economic Review 13, no. 2: 379-408.

Dietrich, A., and G. Wanzenried. 2011. Determinants of bank profitability before and during the

crisis: Evidence from Switzerland. Journal of International Financial Markets, Institutions &

Money 21: 307-327.

Eichngreen, B., and H.D. Gibson. 2001. Greek banking at the dawn of the new millennium. CERP

Discussion Paper 2791, London.



                                                 33
Goddard, J., P. Molyneux, and J. Wilson. 2004. The profitability of European Banks: a cross-

sectional and dynamic panel analysis. Manchester School 72, 3: 363-381.

Guru, B.K., J. Staunton, and B. Balashanmugam. 2002. Determinants of commercial bank

profitability in Malaysia. Working Paper, Multimedia University.

Harle, P., E. Luders, T. Pepanides, S. Pfetsch, T. Pappensieker, and U. Stegemann. 2010. Basel III

and European banking: Its impact, how banks might respond, and the challenges of implementation.

EMEA Banking – McKinsey&Company, November.

Heid, F., D. Porath, and S. Stolz. 2004. Does capital regulation matter for bank behavior? Evidence

for Germany savings banks. Frankfurt am Main: Deutsche Bundesbank.

Iannotta, M.C., G. Nocera, and A. Sironi. 2007. Ownership structure, risk and performance in the

European banking industry. Journal of Banking and Finance 31, no. 7: 2127-2149.

King, M.R. 2010. Mapping capital and liquidity requirements to bank lending spreads. Bank

International Settlements (BIS) Working Papers no. 324, November.

King, M.R. 2012. The Basel III net stable funding ratio and bank net interest margins. International

Settlements (BIS) Working Papers (unpublished).

Kosmidou, K. 2008. The Determinants of Banks’ Profits in Greece during the Period of EU

Financial Integration. Managerial Finance 34: 146-159.

Kosmidou, K. 2008. The determinants of banks’ profits in Greece during the period of EU financial

integration. Managerial Finance 34, no. 3: 146-159.

Kosmidou, K., S. Tanna, and F. Pasiouras. 2005. Determinants of Profitability of Domestic UK

Commercial Banks: Panel Evidence from the Period 1995-2002. Money Macro and Finance (MMF)

Research Group Conference.

Laeven, L. and R. Levine. 2009. Bank governance, regulation and risk taking. Journal of Financial

Economics 93, no. 2: 259-275.



                                                34
Mamatzakis, E., and P. Remoundos. 2003. Determinants of Greek Commercial Banks Profitability,

1989-2000. Spoudai 53, no. 1: 84-94.

Micco, A., U. Panizza, and M. Yanez. 2007. Bank ownership and performance. Does politics

matter? Journal of Banking and Finance 31, no. 1: 219-241.

Millon Cornett, M., M. McNutt, and J.J. Tehranian. 2010. The financial crisis, internal corporate

governance, and the performance of publicy-traded U.S. bank holding companies. Working paper.

Mohd Said, R., and M. Hanafi Tumin. 2011. Performance and Financial Ratios of Commercial

Banks in Malaysia and China. International Review of Business Research Papers 7, no. 2: 157-169.

Molyneux, P., and J. Thornton. 1992. Determinants of European Bank Profitability: A Note.

Journal of Banking and Finance 16, no. 6: 1173-1178.

Naceur, S.B., and M. Kandil. 2009. The Impact of Capital Requirements on Banks' Cost of

Intermediation and Performance: The Case of Egypt. Journal of Economics and Business 61: 70-89.

Pasiouras, F., and K. Kosmidou. 2007. Factors Influencing the Profitability of Domestic and

Foreign Commercial Banks in the European Union. Research in International Business and

Finance 21: 222-237.

Perry, P. 1992. Do banks gain or lose from inflation. Journal of Retail Banking 14: 25-40.

Resti, A. 2011. Liquidità e capitale delle banche: le nuove regole, i loro impatti gestionali. Bancaria

no. 11: 14-23.

Rime, B. 2001. Capital requirements and bank behavior: Empirical evidence for Switzerland.

Journal of Banking and Finance 25, no. 4: 789-805.

Saunders, A., E. Strock, and N.G. Travlos. 1990. Ownership Structure, Deregulation, and Bank

Risk Taking. The Journal of Finance 45, no. 2: 643-654.

Shen, C.H., Y.K. Chen, L.F. Kao, and C.Y. Yeh. 2009. Bank Liquidity Risk and Performance.

SSRN Working Papers, June.



                                                  35
Short, B.K. 1979. The relationship between commercial bank profit rate and banking concentration

in Canada, Western Europe and Japan. Journal of Banking and Finance 3, no. 3: 209-219.

Stiroh, K.J., and A. Rumble. 2006. The dark side of diversification: the case of US financial holding

companies. Journal of Banking and Finance 30, no. 8: 2131-2161.

Xiao, Y. 2009. French Banks Amid the Global Financial Crisis. IMF Working Paper.




                                                 36
Tables


Table 1. Sample distribution by specialization in each country.
                             Bank
                                                                                          Real Estate
                           Holding &       Commercial         Cooperative    Savings
Country/Specialization                                                                    & Mortgage        Total
                            Holding          Banks              Banks         Banks
                                                                                            Banks
                           Companies
Austria                        4                2                 1                                           7
Belgium                        2                                                                              2
Denmark                                         11                1             6                            18
France                          1               1                 1                                           3
Germany                         3               5                 1             1                            10
Ireland                                         3                                              1              4
Italy                           1               5                 8                                          14
Netherlands                     1               2                 1                                           4
Norway                          1               1                               18             1             21
Portugal                        1               2                 1             1                             5
Spain                                           7                 3             16                           26
Sweden                          3               3                               2              1              9
Switzerland                     2               1                 1                                           4
United Kingdom                  4               2                                              4             10
Australia                       1               7                 1                            1             10
Canada                          1               9                 2             1                            13
USA                             60              4                                                            64
Total                           85              65                21            45             8             224

Source: Bankscope Database, authors’ calculations.



Table 2. Summary statistics on dependent variable for sample banks.
                            Pre-crisis period                         Crisis and “post” crisis period
                        Mean                                             Mean
Variable             (Std. Dev.)           Min. – Max.                (Std. Dev.)                  Min. – Max.
ROA                     0.009              -0.016-0.043                  0.003                     -0.224-0.031
                       (0.006)                                          (0.013)
ROE                     0.140              -0.218-0.446                  0.042                     -0.458-0.438
                       (0.066)                                          (0.262)
Notes: This table reports summary statistics on ROA and ROE for the sample banks per for the pre-crisis period (2003-
2006) and for the crisis and “post” crisis period (2007-2010). The sample banks is composed of 224 top-tier
international banks.
The dependent variables are defined in Section 3.
Std. Dev. denotes Standard Deviation, Min. denotes Minimum and Max. denotes Maximum.
Source: Bankscope Database, authors’ calculations.




                                                         37
Table 3. Explanatory variables description and predicted sign.
                       Variable             Measure                                 Notation             Expected sign
                                            Ln (total assets) and
                       Size                                                      SIZE and SIZE2         Positive/negative
                                            Ln (total assets) 2
       Bank-specific




                       Liquidity risk       Net Stable Funding Ratio                 NSFR               Positive/negative

                       Capital              Equity to total assets                  EQ/TA               Positive/negative

                       Credit risk          Loan loss provision to total loans      LLP/TL                  Negative

                       Efficiency           Cost income ratio                         CIR                   Negative

                       Funding costs        Interest expenses to average total      IE/ATD                  Negative
                                            deposits
  Economic




                       Gross Domestic       Annual percent change of GDP             GDPC                   Positive
   Macro-




                       Product

                       Inflation            Annual percent change of Inflation       INFC                   Positive
  Industry-
  specific




                       Concentration        Concentration Ratio 3                     CR3               Positive/negative

Source: Bank-specific and industry-specific variables are available from Bankscope Database. Macroeconomic
variables are available from World Economic Outlook Database (International Monetary Fund, IMF).


Table 4. Summary statistics on explanatory variables for sample banks.
                                                Pre-crisis period                      Crisis and “post” crisis period
                                           Mean                                         Mean
Variable                                (Std. Dev.)            Min.-Max.             (Std. Dev.)             Min.-Max.

Bank-specific
SIZE                                     11.725              9.412-16.832              11.793             9.516-17.089
                                         (1.635)                                       (1.725)
NSFR                                      0.858               0.176-1.364               0.871              0.180-1.559
                                         (0.228)                                       (0.221)
E/TA                                      0.072               0.018-0.215               0.076             -0.003-0.327
                                         (0.034)                                       (0.035)
LLP/TL                                    0.003              -0.013-0.122               0.012             -0.007-0.468
                                         (0.009)                                       (0.024)
CIR                                       0.601               0.147-0.243               0.650              0.100-5.479
                                         (0.151)                                       (0.320)
IE/ATD                                    0.034              -0.000-0.218               0.028              0.001-0.275
                                         (0.023)                                       (0.024)

Macroeconomic
GDPC                                      0.027              -0.385-0.053               0.002             -0.069-0.056
                                         (0.985)                                       (0.027)
INFC                                      0.023               0.046-0.035               0.019             -0.017-0.049
                                         (0.007)                                       (0.013)


Industry-specific
CR3                                        0.42               0.047-0.839               0.439              0.246-0.724
                                         (0.178)                                       (0.131)
                                                                       38
Notes: This table reports summary statistics on explanatory variables for the sample banks for the pre-crisis period
(2003-2006), and for the crisis and “post” crisis period (2007-2010).
Source: Bankscope Database and authors’ calculations.


Table 5. Correlations in the pre-crisis period and in the crisis and “post” crisis period.
                                           ROA                                                    ROE
                        Pre-crisis period     Crisis and “post”               Pre-crisis period       Crisis and “post”
Variable                                        crisis period                                           crisis period
SIZE                         -0.0795               0.0517                          0.2966*                  0.0706
SIZE2                        -0.0880               0.0513                          0.2909*                 0.0728*
NSFR                          0.2067*              0.0834*                        -0.2096*                  0.0881*
E/TA                          0.6156*              0.1007*                        -0.1207*                  0.0441
LLP/TL                        0.2145*             -0.6006*                         0.0786                 -0.5801*
CIR                          -0.3666*             -0.2948*                        -0.3910*                -0.2666*
IE/ATD                       -0.1705*             -0.0184                          0.0129                 -0.1841*
GDPC                          0.1707*              0.2021*                         0.2850*                  0.1748*
INFC                          0.1918*              0.1735*                         0.0213                   0.1661*
CR3                          -0.1514*              0.0258                          0.0139                   0.0393
Notes: The pre-crisis period spans from 2003 to 2006, while the crisis and the “post” crisis period extends from 2007 to
2010.
The variables with no * are independent.



Table 6. Panel regressions.
                                                 ROA as dependent variable
Variable                      Whole period             Whole period and       Pre-crisis period         Crisis and “post”
                                                        dummy crisis                                      crisis period
SIZE                             -1.217                     -0.893                  0.070                      2.948
                                (0.742)                    (0.747)                (0.546)                    (1.916)
SIZE2                             0.041                     0.037                  -0.011                     -0.101
                                (0.030)                    (0.030)                (0.023)                    (0.077)
NSFR                             1.438***                   1.523***                0.168                     3.120***
                                (0.331)                    (0.331)                (0.231)                    (0.584)
E/TA                              0.103***                  0.106***                0.012                     0.205***
                                (0.020)                    (0.020)                (0.013)                    (0.035)
LLP/TL                          -0.345***                 -0.341***                 0.165                   -0.289***
                                (0.015)                    (0.014)                (0.059)                    (0.020)
CIR                             -0.010***                 -0.010***               -0.025***                  -0.008***
                                (0.001)                    (0.001)                (0.001)                    (0.001)
IE/ATD                            1.449                     1.581                   2.919                     4.148
                                (1.928)                    (1.921)                (2.423)                    (2.732)

GDPC                             0.027*                     0.019                  0.022                      0.017
                                 (0.011)                   (0.011)                (0.020)                    (0.014)
INFC                              0.006                     0.006                  0.144**                    0.039
                                 (0.025)                   (0.025)                (0.044)                    (0.033)
CR3                              -0.426                    -0.363                  0.506**                  -3.365**
                                 (0.355)                   (0.354)                (0.166)                    (1.278)
Dummy crisis                                               -0.209**
                                                           (0.072)

Number of observations            1166                      1166                    418                       748
Number of sample banks             224                      224                     166                       219
Adjusted R2                      0.6039                    0.6071                  0.8943                    0.5761



                                                            39
                                               ROE as dependent variable
Variable                      Whole period           Whole period and          Pre-crisis period       Crisis and “post”
                                                      dummy crisis                                       crisis period
SIZE                             11.871                   16.694                    -7.129                  84.670*
                                (14.993)                (15.134)                    (8.099)                (39.416)
SIZE2                             -0.616                  -0.673                     0.158                  -3.280*
                                 (0.618)                 (0.617)                    (0.351)                 (1.595)
NSFR                              49.579***              43.852***                   0.488                  60.305***
                                 (6.689)                 (6.702)                    (3.434)                (12.023)
E/TA                               0.304                  0.343                    -1.030***                 2.207**
                                 (0.422)                 (0.422)                    (0.206)                 (0.736)
LLP/TL                           -6.347***               -6.288***                  -0.671                 -5.513***
                                 (0.313)                 (0.317)                    (0.464)                 (0.430)
CIR                              -0.182***               -0.178***                 -0.294***                -0.113***
                                 (0.021)                 (0.021)                    (0.018)                 (0.028)

IE/ATD                           -1.720***                -1.710***                 0.592                  -1.558**
                                 (0.389)                   (0.388)                 (0.359)                  (0.561)
GDPC                              0.193                     0.075                 1.636***                   0.020
                                 (0.228)                   (0.234)                 (0.297)                  (0.301)
INFC                              0.749                     0.742                 2.564***                  1.374*
                                 (0.512)                   (0.511)                 (0.664)                  (0.679)
CR3                              -10.801                    -9.857                6.774**                  -58.466*
                                 (7.173)                   (7.173)                 (2.471)                 (26.292)
Dummy crisis                                                -3.122*
                                                           (1.463)

Number of observations            1166                     1166                     418                      748
Number of sample banks             224                     224                      166                      219
Adjusted R2                      0.5168                   0.5187                   0.8007                   0.5082

Notes: The dependent variables are ROA and ROE which measure the bank profitability. The explanatory variables are
bank-specific characteristics referring to bank size (SIZE and SIZE2), liquidity risk (NSFR), capital (E/TA), credit risk
(LLP/TL), efficiency (CIR), and funding costs (IE/ATD); macroeconomic factors referring to GDP growth (GDPC) and
INF growth (INFC); and industry-specific characteristic referring to bank market concentration (CR3). The dummy
crisis identifies the start of the crisis (from 2007).
The dependent variable and the independent variables are defined respectively in Section 3.
‘Whole period’ denotes the period from 2003 to 2010 (latest data available).
‘Pre-crisis period’ denotes the period from 2003 to 2006.
‘The crisis and “post” crisis period’ denotes the period from 2007 to 2010.
The small number of observations in the pre-crisis period depends both from having decided to focus only on large
banks and from a poor coverage of Bankscope database in the pre-IAS/IFRS period (or in the period prior to 2005).
Standard Errors of estimated coefficients are reported in parentheses. Adjusted R-squared derives from areg.
*** denotes coefficient statistically different from zero (1% level, two-tail test), ** 5% level, * 10% level.


Table 7. Panel regressions (robustness tests).
                                               ROA as dependent variable
Variable                      Whole period           Whole period and          Pre-crisis period       Crisis and “post”
                                                      dummy crisis                                       crisis period
SIZE                             -1.358                   -0.937                    -1.150                    2.742
                                 (0.734)                 (0.739)                   (0.554)                  (1.920)
SIZE2                             0.043                   0.038                      0.005                   -0.100
                                 (0.030)                 (0.030)                   (0.023)                  (0.077)
NSFR                              1.396***                1.510***                   0.392                   2.995***
                                 (0.331)                 (0.331)                   (0.224)                  (0.587)
E/TA                              0.102***                0.105***                  -0.011                   0.186***
                                 (0.020)                 (0.020)                   (0.014)                  (0.035)
LLP/TL                          -0.361***               -0.351***                  -0.078*                 -0.304***
                                (0.014))                 (0.014)                   (0.031)                  (0.019)
CIR                              -0.010***              -0.010***                  -0.025***                -0.009***
                                                           40
                                 (0.001)                   (0.001)                 (0.001)                 (0.001)
IE/ATD                            2.079                     2.012                   2.619                   3.847
                                 (1.861)                   (1.850)                 (2.431)                 (2.583)
Dummy crisis                                              -0.248***
                                                           (0.069)

Number of observations            1166                     1166                     418                      748
Number of sample banks             224                     224                      166                      219
Adjusted R2                      0.6015                   0.6064                   0.8895                   0.5696


                                               ROE as dependent variable
Variable                      Whole period           Whole period and          Pre-crisis period       Crisis and “post”
                                                      dummy crisis                                       crisis period
SIZE                             11.871                   13.033                   -11.176                  76.957*
                                (14.993)                (14.978)                   (8.690)                 (39.430)
SIZE2                             -0.616                  -0.509                     0.539                  -3.048*
                                 (0.618)                 (0.608)                   (0.370)                  (1.592)
NSFR                              42.579***             43.590***                    6.573                   57.806***
                                 (6.689)                 (6.708)                   (3.525)                 (12.068)
E/TA                               0.304                  0.212                    -1.100***                 1.770*
                                 (0.422)                 (0.415)                   (0.221)                  (0.725)
LLP/TL                            -6.347***             -6.475***                  -1.147*                 -5.870***
                                 (0.313)                 (0.295)                   (0.492)                  (0.405)
CIR                              -0.182***             -0.177***                  -0.308***                 -0.138***
                                 (0.021)                 (0.020)                   (0.020)                  (0.027)
IE/ATD                           -1.720***             -1.501***                    -0.621                 -1.455**
                                 (0.389)                 (0.374)                   (0.381)                  (0.530)
Dummy crisis                                             -3.591*
                                                         (1.413)

Number of observations            1166                     1166                     418                      748
Number of sample banks             224                     224                      166                      219
Adjusted R2                      0.5168                   0.5176                   0.7671                   0.5082
Notes: The dependent variables are ROA and ROE which measure the bank profitability. The explanatory variables are
bank-specific characteristics referring to bank size (SIZE and SIZE2), liquidity risk (NSFR), capital (E/TA), credit risk
(LLP/TL), efficiency (CIR), and funding costs (IE/ATD). The dummy crisis identifies the start of the crisis (from 2007).
The dependent variable and the independent variables are defined respectively in Section 3.
‘Whole period’ denotes the period from 2003 to 2010 (latest data available).
‘Pre-crisis period’ denotes the period from 2003 to 2006.
‘The crisis and “post” crisis period’ denotes the period from 2007 to 2010.
The small number of observations in the pre-crisis period depends both from having decided to focus only on large
banks and from a poor coverage of Bankscope database in the pre-IAS/IFRS period (or in the period prior to 2005).
Standard Errors of estimated coefficients are reported in parentheses. Adjusted R-squared derives from areg.
*** denotes coefficient statistically different from zero (1% level, two-tail test), ** 5% level, * 10% level.




                                                           41
Appendix


Table A. The NSFR construction.
                                                Available Stable Funding (ASF)

  ASF Factor      Basel 3: ASF categories (Liabilities)                 Corresponding definition from Bankscope database
    100%          Capital: Tier 1 & 2 Capital Instruments               Equity

     100%         Other preferred shares and capital instruments        -
                  in excess of Tier 2 allowable amount having
                  an effective maturity of one year or greater

     100%         Long-term debt                                        Long-term debt

     90%          Stable short-term deposits of retail and small        Stable deposits: Total customer deposits * 75% (1)
                  business customers

     80%          Less stable short-term deposits of retail and         Less stable deposits: Total customer deposits * 25% (1)
                  small business customers

     50%          Wholesale short-term funding from non-                -
                  financial corporate

      0%          All other liabilities and equity not included         All other liabilities and equity not included above
                  above

                                                Required Stable Funding (RSF)


  RSF Factor      Basel 3: RSF categories (Assets)                          Corresponding definition from Bankscope database
     0%           Cash and short-term securities                            Cash and short-term securities

      5%          Sovereign Debt                                            Government Debt

      5%          Off balance sheet exposures undrawn amount                -
                  of committed credit and liquidity facilities

     20%          Highly rated corporate and covered bonds                  -

     50%          Short-term loans to non-financial corporate               Short-term corporate loans: Corporate & Commercial loans
                  clients                                                   * 25% (2)

     65%          Unencumbered residential mortgages of any
                  maturity with Basel II standardized risk weights
                                                                            Simplified assumption:
                  of 35% or below
                                                                            total retail loans (short- and long-term) * RSF factor at 85%
                                                                            (in a prudential perspective)
     85%          Short-term loans to retail clients

     100%         Other assets                                              All other assets not included above
Notes: This table shows the construction of the NSFR used in this paper using Bankscope database (see third column)
and its comparison with the NSFR of Basel 3 (see second column). The NSFR is defined as the ratio of Available
Stable Funding (ASF) to Required Stable Funding (RSF). This ratio must be greater than 100%.
The columns “Basel 3: ASF categories (Liabilities)”, “ASF Factor”, “Basel 3: RSF categories (Assets)”, and “RSF
Factor” were constructed using data from the December 2010 Final Document (Basel 3).
Some detail of the NSFR has been omitted for simplicity. See BCBS (2010) for full details.
Short- and long-term is defined as less and more than one year respectively.
ASF and RSF balance sheet items not present in the Bankscope database are indicated in table with “-“.
 (1) The supervisors’ assumptions affirm that the 75 per cent of deposits are stable and 25 per cent are less stable. See
King (2010); (2) The supervisors’ assumptions affirm that the 25 per cent of corporate loans are less than one year in
maturity. See King (2010).
In the Bankscope database, retail loans are calculated as the sum of Residential Mortgage Loans, Other Mortgage
Loans, and Other Consumer/Retail Loans. For a detailed description of the simplified assumptions on retail loans used
                                                                   42
to calculate the NSFR, see Section 3.
Source: BCBS (2010) and Bankscope Database, authors’ calculations.



Table B. NSFR distribution by specialization.
                                   Pre-crisis period                 Crisis and “post” crisis period
                              Mean              Min.-Max.               Mean             Min.-Max.
Specialization             (Std. Dev.)                               (Std. Dev.)
Bank Holding &                0.830            0.219-1.167              0.834           0.309-1.323
Holding Companies            (0.215)                                   (0.214)

Commercial Banks             0.753            0.284-1.214              0.791           0.263-1.559
                            (0.189)                                   (0.219)

Cooperative Banks            0.904            0.513-1.246              0.964           0.370-1.211
                            (0.190)                                   (0.168)

Savings Banks                1.027            0.176-1.364              0.997           0.180-1.358
                            (0.201)                                   (0.185)

Real Estate &                1.177            0.944-1.309              1.030           0.706-1.384
Mortgage Banks              (0.122)                                   (0.152)
Source: Bankscope Database and authors’ calculations.




                                                        43

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:15
posted:9/25/2012
language:English
pages:43