Bank Governance, Regulation, Supervision, and Risk Reporting:

Evidence from Operational Risk Disclosures in European Banks

Ahmed Barakata Khaled Hussaineyb

Abstract

This paper investigates the direct and joint effects of bank governance, regulation, and
supervision on the quality of risk reporting in the banking industry, as proxied for by operational
risk disclosure (ORD) quality in European banks. After controlling for the endogeneity between
bank stability and risk reporting quality, we find that banks having a higher proportion of outside
board directors, lower executive ownership, concentrated outside non-governmental ownership,
and more active audit committee, and operating under regulations promoting bank competition
(i.e., less stringent entry to banking requirements) provide ORD of higher quality. In addition, we
find that the contribution of bank supervisors to enhancing ORD quality depends on the
ownership structure of the bank. Specifically, powerful and independent bank supervisors
mitigate the incentives for entrenched bank executives to withhold voluntary ORD. Moreover,
bank supervisors and largest shareholders perform substitutive roles in monitoring the bank
management’s compliance with mandatory ORD requirements. For the sake of enhancing risk
reporting quality in banks, our findings recommend sustaining board independence, enhancing
audit committee activity, easing entry to banking requirements, and promoting a more proactive
role for bank supervisors.

Keywords: Bank Regulation and Supervision, Basel II (Pillar 3), Corporate Governance,
European Banks, Operational Risk Disclosure.

JEL Classification: G21, G28, G32, G34, G38.

_____________________________
a
Faculty of Economics and Business Administration, Goethe University Frankfurt, Germany. Email:
barakat@finance.uni-frankfurt.de. Corresponding author.
b
School of Management (Plymouth Business School), Plymouth University, the UK. Email:
khaled.hussainey@plymouth.ac.uk.
1. Introduction

Disclosure of financial and risk information represents an important mechanism for improving
market efficiency through various channels. First, it serves as an outside mechanism for
monitoring the behavior of senior management (Eng & Mak, 2003). Second, it lowers investors’
uncertainty about firm’s expected future cash flows and enables public firms to access external
finance at a reasonable cost of capital (Botosan, 1997; Botosan & Plumlee, 2002; Easley &
O'
Hara, 2004; Kothari et al., 2009; Nier & Baumann, 2004). Third, it supports the firm’s
legitimacy and reputation thus maintaining the trust of various stakeholders (Lindblom, 1994;
Oliveira et al., 2011b).

In the banking industry, one main stream of disclosure is risk reporting. The Basel Committee
on Banking Supervision (BCBS), in the Basel II Capital Accord (Pillar 3), emphasizes the
importance of informative risk disclosures in banks for enhancing market discipline (BCBS,
2006b). Literature also showed that banks disclosing more comprehensive risk information
choose a higher capital buffer and lower default risk (Boot & Schmeits, 2000; Cordella & Yeyati,
1998; Nier & Baumann, 2006). Moreover, banks are risk-oriented institutions whose disclosures
should be studied independently of those of non-financial firms (Bessis, 2002; Linsley & Shrives,
2005, 2006). Furthermore, following the Global Financial Crisis (GFC), risk disclosures in banks
have been emphasized as an effective tool for avoiding banking crises (Financial Stability Board,
2012).

Previous research has mainly focused on the mechanical effects of a bank’s contextual factors
such as size, leverage, riskiness, profitability, and capital adequacy on risk reporting quality
(Bischof, 2009; Ford et al., 2009; Helbok & Wagner, 2006; Linsley et al., 2006; Oliveira et al.,
2011a, 2011b; Woods et al., 2009; Yong et al., 2005). However, to the best of our knowledge, no
previous research has examined the direct and joint effects of bank governance, regulation, and
supervision on risk reporting quality. Agency, management entrenchment, legitimacy, resource
dependence and stakeholder theories suggest that the heavy regulation and strict supervision of
the banking sector as well as bank level governance structures could shape the discretionary
decision of bank management to report risk information. This paper contributes to the governance
and disclosure literature by theoretically justifying and empirically examining the implications of
these theories with respect to risk reporting quality in the heavily regulated and strictly

1
supervised banking sector. Practically, this contribution is important because our findings inform
banking stakeholders, regulators, supervisors, and other relevant policy makers on the most
important country level and bank level governance mechanisms that could be employed to
enhance risk disclosures in banks thus boosting market efficiency and discipline.

The Basel II Capital Accord identifies three major risk types against exposure to which banks
are required to reserve sufficient capital resources (i.e., regulatory capital). These are credit risk,
market risk, and operational risk (BCBS, 2006b). The Basel II Capital Accord defines operational
risk as ‘… the risk of loss resulting from inadequate or failed internal processes, people and
systems or from external events. This definition includes legal risk, but excludes strategic and
reputational risk’ (BCBS, 2006b, p.144). Moreover, operational risk per se is a major source of
risk and financial distress in banks.1 Therefore, operational risk management and disclosure
practices in financial institutions have recently attracted increased attention from academics,
professionals, and regulators (BCBS, 1998b, 2001; Ford et al., 2009; Helbok & Wagner, 2006).

In this paper, we argue that disclosures on operational risk exposure and management, unlike
disclosures on credit and market risks, present a unique opportunity to consistently evaluate the
discretionary decision of bank management to provide risk disclosures of a certain quality in the
annual reports and risk reports. Our argument is based on the following grounds. First,
operational risk represents a purely idiosyncratic risk that is free from any contagion effects
(Daníelsson et al., 2001; Perry & de Fontnouvelle, 2005). Hence, disclosures on operational risk
exposure and management represent information that is unique to the disclosing bank. Second,
annual reports and risk reports are the major media through which banks communicate their
regular operational risk information to various stakeholders. While regular information on other
major risk types (i.e., credit risk and market risk) can be extracted from other sources such as
credit ratings announced by specialized rating agencies (e.g., Fitch Inc., Moody'
s, and Standard &
Poor’ s) as well as currency, commodity and stock exchanges, no regular information on

1
Many huge operational losses have hit large financial institutions all over the world leading to severe financial
disturbance or even the collapse of these institutions. For example, Allied Irish Banks (AIB), the second largest bank
in Ireland, was hit by one of the worst scandals in banking history in 2002, losing over $691 million when a currency
trader in their Baltimore office invested unsuccessfully in Japanese yen and kept the bank’ s losses a secret for about
five years. More recently, there was a rogue trading loss of €4.9 billion that was discovered by Société Générale in
2008. This operational risk event had been running undetected for about one year. In terms of the financial
consequences of operational losses, one of the worst examples might have been the unauthorized speculative trading
loss caused by Nick Leeson to Barings Bank during the period 1992-1995 which accumulated a loss of £827 million
(approximately $1.3 billion) leading the United Kingdom' s oldest investment bank to totally collapse.

2
operational risk is disclosed in public sources. Third, while comprehensive disclosures on credit
and market risks are imposed by the requirements of International Financial Reporting Standard
No. 7: Financial Instruments: Disclosures (IFRS 7), no accounting standard has yet been
proposed to regulate operational risk disclosures (ORD). Hence, unlike ORD, the quality of
disclosures on credit and market risks could be contaminated by contagion effects, diluted by the
availability of such disclosures from other public sources of information, or driven by the
mandatory requirements of accounting standards.

The remainder of this paper is organized as follows. Section 2 presents a background on

operational risk disclosures in the banking industry and its implementation in European banks.
Section 3 reviews prior research on risk disclosures in banks and develops our research
hypotheses. Section 4 introduces our measure for operational risk disclosure quality and provides
details on the data and sample selection, variable definitions, and empirical model design. Section
5 presents and discusses the empirical results. Section 6 concludes.

2. Overview of Operational Risk Disclosures in the Banking Industry

The Basel II Capital Accord calls on banks to reserve sufficient capital resources to protect
against operational risk exposure (i.e., regulatory capital for operational risk). Regarding
operational risk in banks, the Basel II Capital Accord uses a “three pillars” concept. Pillar 1
capital represents the minimum of their own funds that banks have to hold against their ex post
operational risk exposure from a regulatory point of view. Pillar 2 capital is determined according
to the Internal Capital Adequacy Assessment Process (ICAAP) by which additional operational
risk capital could be computed using the bank’ s internal models to reflect the bank’ s own
assessment of its ex post operational risk exposure. Pillar 2 capital is sometimes called economic
capital or operational Value-at-Risk (operational VaR). Pillar 2 capital is subject to review by the
relevant banking supervisory authorities if it differs from the Pillar 1 capital. Pillar 3 requires
certain operational risk disclosures which could encourage market discipline. Although
considered as an important step toward enhancing operational risk disclosures in banks, many
professionals and academics still see these requirements as being very general and qualitative in
nature (Ford et al., 2009).

3
 In 2006, the European Parliament (EP) issued the Capital Requirements Directive (CRD) to
implement the requirements of the three pillars of the Basel II Capital Accord in banks within the
European Union (EU) (EP, 2006). Most EU banks started implementing the CRD requirements in
the fiscal year 2008 (Deloitte, 2009). However, as is the case for Pillar 3 of the Basel II Capital
Accord, the disclosure requirements in the CRD are very general in nature and do not impose a
clear-cut framework for disclosure on operational risk exposure and management. Thus, bank
managers in the EU still have a great deal of discretion regarding the ORD quality to be provided
in their annual reports and risk reports. Such a high level of discretion is evident from the
outcomes of two recent surveys conducted by the Committee of European Banking Supervisors
(CEBS) and Deloitte in 2009 (CEBS, 2009; Deloitte, 2009). CEBS, in its assessment of Pillar 3
disclosures in 25 European banks for the fiscal year 2008, documented the fact that only 32% of
these banks had disclosed detailed information on their operational risk exposure and
management (CEBS, 2009). Moreover, CEBS (2009) criticized the lack of information disclosed
on the measurement methodology used to quantify the regulatory capital for operational risk
especially in banks applying Advanced Measurement Approaches (AMA). Furthermore, Deloitte
(2009) documented the fact that 29% of the 47 banks they surveyed had disclosed no information
on the measurement methodology used to quantify their regulatory capital for operational risk.
Both surveys clearly show that there are still remarkable variations in the ORD quality in EU
banks even after the promulgation and implementation of the CRD.

3. Prior Research and Hypothesis Development

3.1 Prior Research on Risk Reporting Quality in the Banking Industry

The determinants of the quality of risk reporting in financial institutions are under-researched.
For example, using the BCBS and International Organization of Securities Commissions
(IOSCO) joint recommendations as a benchmark, Yong et al. (2005) documented significant
variations in the level of risk management disclosures (not including operational risk) by Asia
Pacific banks across regions and levels of economic development. Helbok and Wagner (2006)
inspected the attributes and determinants of ORD in the annual reports of 59 banks in North
America, Asia, and Europe in the period 1998-2001. They documented an increase in ORD in
terms of extent (measured by word and page counts) and content (measured by an ORD index).

4
Moreover, they found that banks with a lower equity to assets ratio or profitability ratio are more
likely to disclose detailed information about operational risk in their annual reports. Linsley et al.
(2006) analyzed the contents of the annual reports of 9 UK and 9 Canadian banks in 2001 and
found that the level of risk disclosure is positively associated with bank size yet unrelated to
profitability and risk exposure. Using the requirements of the Basel II Capital Accord (Pillar 3) as
a benchmark for ORD quality in 65 internationally active financial institutions in the period
2004-2006, Ford et al. (2009) found that almost half of their sample firms were not meeting the
minimum disclosure requirements of the Basel II Capital Accord (Pillar 3). Furthermore, they
documented a lack of consistency in ORD. Surprisingly; they found that their sample firms were
reducing the extent of their ORD over time. They interpreted this latest result by arguing that
financial institutions generally prefer to disclose less when they utilize less complicated capital
measurement and allocation techniques. Several studies showed that the adoption of IAS/IFRS
has led to an increase in the level of risk disclosures in banks (Bischof, 2009; Oliveira et al.,
2011a; Woods et al., 2009). Using a sample of Portuguese banks, Oliveira et al. (2011b) showed
that stakeholder monitoring and corporate reputation are the main drivers of voluntary risk
reporting. However, Oliveira et al. (2011b) did not consider the impact of time series
fluctuations, country level regulations, or bank level governance mechanisms on the quality of
risk disclosures in the banking industry.

3.2 Hypothesis Development

Our paper contributes to the extant risk disclosure literature by utilizing a number of theories
(i.e., agency theory, management entrenchment theory, and organization-society theories) to
explain how bank governance, regulation, and supervision could affect risk reporting quality in
the banking industry. In the following paragraphs, we briefly explain the definitions and unique
aspects of organization-society theories (i.e., legitimacy theory, resource dependence theory, and
stakeholder theory) and how they may apply to the banking industry.

Legitimacy theory (Dowling & Pfeffer, 1975; Lindblom, 1994; Suchman, 1995) explains how
organizations show their adherence to the values of society and meet social expectations.
According to this theory, banks might: a) comply with mandatory risk disclosures to confirm
their adherence to the already established institutional values of society (i.e., institutional

5
legitimacy), and b) employ voluntary risk disclosures as a tool of legitimation2 (i.e.,
organizational legitimacy). However, Chen and Roberts (2010) asserted that, although legitimacy
theory is important, it has abstract premises which can be further operationalized using resource
dependence and stakeholder theories.

Resource dependence theory (Pfeffer, 1972; Pfeffer & Salancik, 1978, 2003) posits that the
firm depends in its survival and growth on the finite important resources available in its external
environment and that it competes with other firms to control and benefit from these resources.
These resources can take various forms such as experienced labor, capital flows, loyal customers,
and reputation. According to this theory, banks might employ risk disclosures as an effective tool
to attract cheaper capital, widen customer base, and maintain reputation.

Stakeholder theory (Freeman, 1984) differs from resource dependence theory in that it
recognizes that stakeholders are different in their expectations, powers, and influence on the
organization. According to this theory, banks might disclose risk information to more effectively
interact and better communicate with influential stakeholders (e.g., bank supervisors, largest
shareholders, and governmental blockholders).

As obvious from the discussion above, organization-society theories suggest that bank
management might use risk disclosures as a tool to sustain the bank’ s legitimacy and reputation
thus maintaining resource flow with various stakeholders and ultimately supporting the bank’ s
survival and growth (Oliveira et al., 2011b).

Restrictions on Non-traditional Banking Activities

Bank regulations which impose restrictions on non-traditional banking activities (i.e.,

securities market activities, insurance activities, real estate activities, and the ownership of
nonfinancial firms) normally weaken the current and expected diversification of the bank’ s
investment portfolio (Barth et al., 2001). Consequently, bank management has less need to access
external finance because it anticipates lower future investment potentials. Furthermore, banks
with restricted activities typically expect no great future expansion in their customer base. Hence,
according to organization-society theories, under more activity restrictions, bank management

2
According to Maurer (1971, p. 361), “legitimation is the process whereby an organization justifies to a peer or
superordinate system its right to exist.”

6
might rely less on risk disclosures as a tool to facilitate access to capital markets and gain better
communication with current and potential customers. Therefore, we formulate our first
hypothesis as follows:

H1: There is a negative association between the restrictions imposed on non-traditional

banking activities and the quality of risk reporting in banks.

Entry to Banking Requirements

Imposing stringent requirements for obtaining a banking license might prevent new domestic
and foreign banks from entering the national banking industry thus mitigating current and
expected bank competition (Barth et al., 2001). Hence, weaker competition could make it less
costly for banks to access external finance (Corvoisier & Gropp, 2002; Hannan & Berger, 1991;
Maudos & Fernández de Guevara, 2004) thus mitigating the incentives for bank management to
enhance the quality of risk disclosures. Furthermore, if we apply organization-society theories to
the banking industry, lower competition intensity might further weaken the incentives for bank
management to employ risk disclosures as a tool to support its reputation from the perspectives of
customers and potential investors. Therefore, we formulate our second hypothesis as follows:
H2: There is a negative association between entry to banking requirements and the quality
of risk reporting in banks.

Official Powers and Independence of Banking Supervisory Authorities

Banking Supervisory authorities3 are crucial to maintaining the integrity, efficiency, and
transparency of the banking industry for the benefit of the national economy (Barth et al., 2001;
BCBS, 1998a, 2006a, 2006b). Nier and Baumann (2004) showed that improved bank disclosure
could be beneficial for bank supervisors since it allegedly reduces equity return volatility and
hence enhances their view of the relative performance and risk of the bank. Moreover,
organization-society theories suggest that bank supervisors who promptly take corrective actions
and are independent of political influence and the banking industry itself could be seen as
effective outside monitors and influential outside stakeholders that are better able to motivate

3
We use the terms ‘bank supervisors’ and ‘banking supervisory authorities’ interchangeably.

7
bank management to provide risk disclosures of higher quality. Therefore, we formulate our third
hypothesis as follows:

H3: There is a positive association between the official powers and independence of
banking supervisory authorities and the quality of risk reporting in banks.

Board Composition

Resource dependence theory suggests that outside directors have expertise, prestige, and
contacts that enable them to provide firms with links to the external environment (Boyd, 1990;
Carpenter & Westphal, 2001; Dalton et al., 1999; Hillman et al., 2000; Pfeffer, 1972). Moreover,
agency theory suggests that outside directors play a vital role in monitoring managers’
performance and limiting their opportunism (Fama, 1980; Fama & Jensen, 1983; Walsh &
Seward, 1990). Hence, a higher proportion of outside directors sitting on a board is more likely to
result in effective monitoring and a higher level of corporate transparency (Frankel et al., 2011).
Furthermore, in its guidance on enhancing corporate governance for banking organizations,
BCBS says that ‘…Independence and objectivity can be enhanced by including qualified non-
executive directors on the board…’ (BCBS, 2006a, p. 7). Therefore, we formulate our fourth
hypothesis as follows:

H4: There is a positive association between the proportion of outside directors sitting on
the board and the quality of risk reporting in banks.

Outside Ownership Concentration

Motivated by the direct linkage of their wealth to the performance and market valuation of the
disclosing firm, larger outside shareholders have stronger incentives to monitor the integrity and
efficiency of management thus alleviating agency conflicts (Admati et al., 1994; Huddart, 1993;
Maug, 1998; Noe, 2002; Shleifer & Vishny, 1986). In the banking industry, the first shareholder4
holding strong voting rights is easily able to affect the strategic decisions taken and investment
projects chosen by bank managers (Laeven & Levine, 2009). Hence, if incumbent bank managers
did not effectively perform their fiduciary duties, then the first shareholder might activate his

4
The first (largest) shareholder is the non-managing, non-governmental shareholder holding the largest voting rights
in the bank.

8
influential voting rights and change that inefficient bank management. In this sense, a first
shareholder with strong voting rights represents an effective outside monitor that might be better
able to press bank management to provide risk disclosures of higher quality. Therefore, we
formulate our fifth hypothesis as follows:

H5: There is a positive association between the proportion of voting rights held by the first
shareholder and the quality of risk reporting in banks.

Governmental Blockholding

In its guidance on enhancing corporate governance for banking organizations, BCBS states
that ‘Supervisors, governments and depositors are among the stakeholders due to the unique role
of banks in national and local economies and financial systems, and the associated implicit or
explicit deposit guarantees’ (BCBS, 2006a, p. 4). Motivated by their tendency to protect the
rights of their own people, governments might be more careful to continuously follow up and
assess the legitimacy of the banks they materially own. Hence, according to organization-society
theories, bank management might rely more on risk disclosures as a tool to support the bank’ s
legitimacy from the perspective of the governmental blockholder.5 Therefore, we formulate our
sixth hypothesis as follows:

H6: There is a positive association between governmental blockholding and the quality of
risk reporting in banks.

Executive Ownership

In the literature, two competing theories explain the relationship between executive ownership
and disclosure quality. These theories are agency theory and management entrenchment theory.
Agency theory suggests a positive association between executive ownership and disclosure
quality since the extent of executive ownership could serve to align the interests of managers and
shareholders (Jensen & Meckling, 1976). By contrast, management entrenchment theory posits
that concentrated executive ownership could be counter-productive to the firm’ s long-term value,
as management could effectively wield outside monitoring. Hence, managers are more likely to
maximize their private benefits from control by reducing disclosure levels (Eng & Mark, 2003;

5
The governmental blockholder is the domestic government holding at least 5% of the voting rights in the bank.

9
Gelb, 2000; Li et al., 2008; Nagar et al., 2003; Ruland et al., 1990). However, given the known
complexity and high information asymmetry between insiders and outsiders that characterizes the
banking industry, we expect management entrenchment theory to outweigh agency theory in
explaining the risk disclosure decisions of bank executives. Therefore, we formulate our seventh
hypothesis as follows:

H7: There is a negative association between the proportion of voting rights held by
executive managers and the quality of risk reporting in banks.

Audit Committee Activity

Audit committees are crucial to improving financial reporting quality (Anderson et al., 2003),
combating financial statement fraud (Beasley et al., 2000; Klein, 2002), and enhancing voluntary
disclosure quality in public firms (Kelton & Yang, 2008; Li et al., 2008). Moreover, active audit
committees are able to more effectively discharge their governance and monitoring
responsibilities (Abbott et al., 2000; Collier & Gregory, 1999; DeZoort et al., 2002).
Furthermore, BCBS (2006a) emphasizes the vital role audit committees are expected to play
regarding the quality of disclosures in banks. Therefore, we formulate our eighth hypothesis as
follows:

H8: There is a positive association between the frequency of audit committee meetings and
the quality of risk reporting in banks.

Interactions between Bank Governance, Regulation, and Supervision

We realize that bank regulation and supervision schemes are institutional settings over which
banks have no control. However, one of the main objectives of our paper is to test how these
institutional settings might explain cross-country variations in risk reporting quality in banks. To
the extent that bank regulators and supervisors might change these institutional settings in the
medium term or long term, our findings on their direct effects are informative regarding how to
enhance risk reporting quality in banks. Typically, bank regulation and supervision schemes do
not change much at the country level in the short term. Therefore, in addition to testing their
direct effects, we also aim to examine whether and how the impact of bank regulation and
supervision on risk reporting quality within a given country could differ across various bank level

10
governance structures. In developing our hypotheses regarding these joint effects, we do not test
the interactions of bank regulation and supervision with outside board directorships and activities
of audit committees because both tend to be independent monitors and advisors to bank
management.6 Hence, their contribution to enhancing risk reporting quality is most likely to be
unaffected by changes in country level regulations.7

Stronger bank level outside monitors (i.e., largest shareholders and governmental
blockholders) are more likely to reduce the adverse impact of disclosure-mitigating regulations
(i.e., activity restrictions and entry to banking requirements) on the disclosure incentives of bank
management. By contrast, weaker disclosure incentives of entrenched bank executives could be
further amplified when banks operate under more restrictive and stringent regulations. Therefore,
we formulate our ninth hypothesis as follows:

H9: The negative association between disclosure-mitigating bank regulations and the
quality of risk reporting is weaker for banks having concentrated outside ownership
and governmental blockholding, yet stronger for banks having concentrated executive
ownership.

According to organization-society theories, bank supervisors, largest shareholders, and

governmental blockholders could all be seen as outside monitors of the disclosure decisions of
bank management. Therefore, they are most likely to perform substitutive roles in enhancing risk
disclosure quality in banks. Furthermore, bank supervisors as country level outside monitors and
influential stakeholders are better able to mitigate the incentives for entrenched bank executives
to reduce risk disclosure levels. Therefore, we formulate our tenth hypothesis as follows:

H10: The positive association between bank supervision and the quality of risk reporting is
weaker for banks having concentrated outside ownership and governmental
blockholding, yet stronger for banks having concentrated executive ownership.

6
Note that audit committees in almost all our sample banks are fully composed of independent directors.
7
Empirically, in unreported robustness checks, when we include interactions of bank regulation and supervision with
outside board directorships and activities of audit committees in our multivariate regressions, we indeed find
insignificant coefficients for all these interaction terms.

11
4. Research Methodology

4.1 Our Operational Risk Disclosure Quality Index

We use a self-constructed disclosure index to measure ORD quality for two reasons.8 First, we
find it impossible to determine a pre-defined set of words that can properly be employed to
consistently identify sentences containing information on operational risk in banks. Second, by
using the disclosure index we try to alleviate the possible bias in the disclosure score that might
result if the disclosing bank focuses on providing an extensive amount of information on certain
aspects of operational risk while totally failing to disclose on other crucial aspects. Therefore, we
believe that the disclosure scores which are determined using our self-constructed disclosure
index are more likely to accurately capture the relative weights of various operational risk
disclosures.

We utilize different sources to construct our disclosure index. First, we use the ORD
requirements in Annex XII: Technical Criteria on Disclosure in the Capital Requirements
Directive (CRD) issued by the European Parliament (EP) in 2006 (EP, 2006). Second, we employ
the most important ORD items used in two closely related studies (Ford et al., 2009; Helbok &
Wagner, 2006).9 Hence, our ORD quality index is composed of 14 main items comprising 56
sub-items (i.e., 42 sub-items for textual disclosures of qualitative, quantitative, and forward-
looking information plus 14 sub-items for graphical illustration or tabular presentation providing
information not included in textual disclosures). Each sub-item is binary-coded (i.e., 1 or 0) thus
making the maximum total ORD score possible amount to 56 points. As explained in Section
4.3.1, there are 6 mandatory ORD sub-items thus making the maximum voluntary ORD score
possible amount to 50 points. Appendix A presents the scoring criteria and composition of our
ORD quality index.

4.2 Data and Sample Selection

We aim to empirically examine the direct and joint effects of bank governance, regulation, and
supervision on the quality of risk reporting in the banking industry, as proxied for by ORD
quality in the annual reports and risk reports of EU banks in the fiscal years 2008, 2009 and
8
Disclosure indexes have been heavily used in prior research (for a comprehensive review, see Berger, 2011; Beyer
et al., 2010).
9
The most important ORD items are those disclosed in the annual reports and risk reports in our sample.

12
2010.10 We begin our sample period in 2008 because it is the year in which most EU banks
started implementing the CRD, thus unifying the mandatory requirements to publicly disclose
operational risk information.11 We restrict our sample to the EU in order to exclude any
unobserved effects due to differences in the official adoption dates of the Basel II Capital Accord
(Pillar 3) requirements.12 We choose our sample from the main banking specializations
(according to the classification used in the Bankscope database): Bank holdings and holding
companies, commercial banks, investment banks, savings banks and cooperative banks with at
least €1,000 million in total assets. We use this size threshold to include in our sample large and
medium-sized banks whose managers tend to have stronger incentives to provide informative
disclosures on operational risk exposure and management.

We apply these initial criteria to the Bankscope database and get an original sample of 137
banks in all EU member countries. Then, we exclude 38 banks which are subsidiaries of other
banking or insurance groups. Additionally, we exclude 14 banks for which we cannot find
complete financial or corporate governance data. Thus, we end up with 85 banks from 20 EU
member countries which are Austria, Belgium, Bulgaria, Cyprus, Denmark, Finland, France,
Germany, Greece, Hungary, Ireland, Italy, Malta, Netherlands, Poland, Portugal, Spain, Sweden,
UK, and Lithuania. We believe that our sample is a fair representative of the whole EU banking
sector since the total assets of our sample banks (€24,199 billion) amount to approximately 74%
of the total EU banking assets (€32,744 billion) in the beginning of the fiscal year 2008. Our
sample takes an unbalanced panel data structure since complete data are available for the fiscal
years 2008 through 2010 for 73 banks while available for the fiscal years 2009 and 2010 for 12
banks. Hence, our final sample is composed of 243 bank-year observations. Appendix B presents
the composition of our final sample by country and specialization.

10
To maximize our final sample, we use the annual reports and risk reports of the fiscal years ending between
01/08/2008 and 31/07/2011.
11
There have been no changes in regulations governing operational risk management and disclosure practices in the
EU after the Global Financial Crisis (GFC) which coincides with our sample period.
12
In the U.S., the Basel II Capital Accord will be implemented only in the largest, internationally active ‘core’
banks. Moreover, its Pillar 3 requirements have not yet been officially approved by U.S. bank regulators for any of
these banks (Dugan & Xi, 2011).

13
4.3 Variable Definitions

4.3.1 Dependent Variables

We have two multivariate regression models. The dependent variable in our first model is the
Total Operational Risk Disclosure Quality Score for each bank-year in our final sample
determined using the disclosure index and scoring criteria explained in Appendix A. In the
second model which serves as a robustness check, we examine the extent to which the association
of bank governance, regulation, and supervision with ORD quality is driven by the discretionary
disclosure decision of bank management not just by its compliance with mandatory ORD
requirements in the CRD (EP, 2006). Hence, we derive a disclosure score for each bank-year that
captures only the voluntary items of the ORD quality index. To do that, we subtract the points
scored, if any, for the disclosure sub-items labeled as “mandatory” under disclosure items 1-6 of
the ORD quality index (see Appendix A) which represent the ORD requirements in the CRD.
Meanwhile, we do not subtract any points scored for disclosure items 7-14 of the ORD quality
index (see Appendix A) because no ORD under these items is required by the CRD. By applying
these refining criteria, the resulting score is the Voluntary Operational Risk Disclosure Quality
Score which is the dependent variable in our second model.13

4.3.2 Explanatory Variables

Country Level Bank Regulation and Supervision

We consider in our analysis only those aspects for which there are variations between the EU
member countries in bank regulation and supervision. Therefore, we exclude all constituents of
the original indexes (Barth et al., 2001, 2004, 2006, 2008; World Bank, 2008) for which all EU
member countries have identical values. We collect data on bank regulation and supervision from
a global database available from the website of the World Bank (World Bank, 2008) and acquire
the description and composition of the indexes from previous research by Barth et al. (2001,
2004, 2006, 2008).14 We use three proxies to examine the impact of bank regulation and
supervision on ORD quality in EU banks. First, we use the restrictions on securities market

13
Data on total and voluntary operational risk disclosure quality scores in our sample (distributed by country,
specialization, and fiscal year) are available upon request from the corresponding author.
14
Appendix D provides data on each index of bank regulation and supervision in the EU (2008-2010). For the sake
of brevity, compositions of the indexes are excluded (available upon request from the corresponding author).

14
activities, insurance activities, real estate activities, and the ownership of nonfinancial firms.
Second, we use the official requirements for obtaining a banking license. Third, we employ the
official powers and independence of banking supervisory authorities.

Bank Level Governance

We use the proportion of outside board directors as a proxy for board composition.15
Regarding the ownership structure of the bank, we use three proxies. First, we proxy for outside
ownership concentration by the proportion of voting rights held by the largest non-managing
non-governmental shareholder. Second, we use a dummy variable that equals 1 if the domestic
government holds at least 5% of the voting rights in the bank. Third, we measure executive
ownership as the proportion of voting rights held by bank executives. Finally, we proxy for audit
committee activity by the frequency of audit committee meetings held during the fiscal year.16
Data on these variables are collected from the annual reports and corporate governance reports of
individual banks. Additional data on largest shareholders and governmental blockholders are
collected from the Bankscope and Datastream databases.

4.3.3 Endogenous Variables

We use the degree of bank stability as an endogenous variable in our multivariate regressions.
Concern about endogeneity arises since as much as instable banks tend to enhance their risk
disclosures in order to signal a stable condition to various stakeholders (Helbok and Wagner,
2006), enhanced risk disclosures might also reduce uncertainty about the bank’ s expected future
cash flows and facilitate its access to external finance at a lower cost (Botosan, 1997; Botosan &
Plumlee, 2002; Kothari et al., 2009) which enhances its capital base and profitability thus

15
It is noteworthy that in some countries, such as Germany and Austria, there is a two-tier board system in which
there are two boards governing the bank; namely the management board (sometimes called the managing or
executive board) and the supervisory board. The supervisory board is elected by shareholders and includes employee
representatives. The supervisory board appoints the management board and supervises its actions. Since this paper
focuses on the monitoring rather than the executive role of the board of directors, we therefore consider the
proportion of outside directors on the supervisory board not the management board. Furthermore, in the two-tier
board structure, we consider the employee representatives to be inside directors.
16
Almost all our sample banks (94%) have separate CEO and board chairman positions. Therefore, we do not find
enough cross-sectional variation in this variable to test its effect on our dependent variables. Furthermore, audit
committee independence (whether proxied for by its ratio or a dummy variable for its full independence) is highly
correlated with the board independence ratio, suggesting that it is sufficient to use only one of them to proxy for
board composition. Moreover, all banks in our sample are audited by BIG-4 audit firms.

15
increasing bank stability. Hence, it is likely that the causality direction of bank stability and its
risk reporting quality can go either way. Therefore, we use two standard measures of bank
stability in the literature, namely Z-score and stock return volatility, as endogenous variables.
Additionally, we instrument for Z-score and stock return volatility using a group of bank level
and country level variables which tend to drive bank stability yet are not expected to directly
affect a bank’ s risk disclosures.

The first endogenous variable is Z-score which equals (CAR + ROA)/ (ROA) where CAR is
the capital-assets ratio, ROA is the return on assets, and (ROA) is the standard deviation of
ROA. The intuition behind Z-score is that banks with weaker capital base, lower profitability,
and higher volatility in their earnings tend to be more prone to insolvency and less financially
stable (Laeven & Levine, 2009). Consequently, the senior management of financially unstable
banks might tend to provide various stakeholders with enhanced risk disclosures in order to
reduce the systematic estimation risk premium incorporated in their expected cost of equity
capital thus making it less costly for them to raise new capital and engage in profitable
investments (Helbok & Wagner, 2006). Hence, we expect a negative impact of Z-score on ORD
quality. Data on Z-score are collected from Bankscope.

Since our paper is mainly focused on ORD quality, we devise a measure of Z-score that
captures the operating stability of the disclosing bank by employing the net operating income
after adding back loan loss provision (LLP) when computing ROA. We believe that this modified
ROA measure is intuitively matched to the objectives of our empirical analysis for a number of
reasons. First, net operating income typically excludes extraordinary items, tax effects, and
discontinued operations thus making it more efficiently capture recurring and ordinary sources of
the bank’ s operating costs and revenues. Second, adding LLP back to the net operating income
excludes expected credit risk exposure from ROA and makes it a more accurate measure of the
current operating efficiency of the bank. Third, previous research has showed that bank
management might smooth income by manipulating LLP (Abou El Sood, 2012; Kanagaretnam et
al., 2004; Shrieves & Dahl, 2003). Hence, excluding LLP should produce a modified ROA
measure that is free of hidden misreporting on the part of bank management.17

17
The majority of our sample banks do not disclose the exact gross amount of their historical operational losses thus
limiting the scope for us to devise a direct measure of the current operational risk exposure to use as an endogenous
variable in our empirical analysis.

16
 The second endogenous variable is stock return volatility which is the annualized standard
deviation of bank’ s daily stock returns. For less stable banks, there is greater uncertainty about
their expected future cash flows thus leading to more severe fluctuations in their stock prices
(Esty, 1998; Laeven & Levine, 2009; Saunders et al., 1990). One important advantage of using
this measure is that it is a market-oriented proxy for bank stability. Banks with a higher stock
return volatility might attempt to enhance their risk disclosures in order to reduce investors’
uncertainty about their expected future cash flows thus enhancing bank stability from the equity
market’ s perspective. Hence, we expect a positive impact of stock return volatility on ORD
quality. Data on stock returns are collected from Datastream.

4.3.4 Instrumental Variables

For the endogenous variables (i.e., Z-score and stock return volatility), we choose instrumental
variables which have a common feature: they affect ORD quality only through their impact on
bank stability. At the country level, we use the gross domestic product (GDP) per capita, inflation
rate, and degree of political stability. We choose these variables because they are the main
indicators of the macroeconomic and political conditions which materially affect the stability of
the banking sector in any given country (Laeven & Levine, 2009; Uhde & Heimeshoff, 2009).
Meanwhile, there is no clear intuition or previous evidence in the banking literature that could
argue for a direct impact of these macro factors on risk reporting quality in banks.18 Formally, we
expect a positive association of GDP per capita and political stability with bank stability. Since
higher inflation rates might indicate more opportunities for profitable investments while also
incurring higher operating costs, we do not expect a specific association with bank stability. Data
on GDP per capita and inflation rate are collected from the World Bank’ s World Development
Indicators Database while data on political stability are collected from the World Bank’ s
Worldwide Governance Indicators Database.

At the bank level, we use cost-income ratio, and cash and short-term investments to total
deposits ratio (henceforth, cash-deposits ratio). Typically, cost-income ratio is inversely related to
a bank’ s operating efficiency (Uhde & Heimeshoff, 2009). Therefore, we expect a negative

18
Note that all our sample banks use IFRS and that the mandatory disclosures in our ORD index are required by the
CRD (EU, 2006) which excludes any possible impact of macroeconomic and political variables on the quality of
accounting standards and prudential regulations.

17
association between cost-income ratio and bank stability. On one hand, a lower cash-deposits
ratio might indicate higher liquidity risk19 and less bank stability. On the other hand, a higher
cash-deposits ratio might indicate an overloaded cash portfolio and more foregone successful
investments thus leading to less bank stability. Therefore, we do not expect a specific association
of cash-deposits ratio with bank stability. Moreover, there is no clear argument or previous
evidence which could support a direct impact of these bank level measures on risk reporting
quality beyond their influence on bank stability. Data on cost-income ratio and cash-deposits
ratio are collected from Bankscope.

Empirically, we lag all instrumental variables by one fiscal year to eliminate any look-ahead
bias. In an unreported robustness check, we measure all instrumental variables at the end of the
fiscal year 2007 (i.e., before our sample period began and, most importantly, before the Global
Financial Crisis (GFC) erupted) and still get qualitatively similar results.

4.3.5 Control Variables

To address the issues related to the omitted correlated variables, our models have a
comprehensive list of control variables. First, we control for the size of the bank using the natural
logarithm of total assets. Data on total assets are collected from Bankscope. Second, we control
for the effects of cross-listing when banks are registered with the U.S. Securities and Exchange
Commission (SEC). Data on cross-listing are collected from SEC filings. Third, we control for
the measurement approach used to quantify regulatory capital for operational risk. Data on
measurement approaches are collected from the annual reports and risk reports of individual
banks.20

Fourth, we control for the strength of shareholder rights using the anti-director rights index
developed by Djankov et al. (2008). Fifth, we proxy for country level legal setting by the
strictness of the legal system using the rule of law index available from the World Bank’ s

19
Regarding liquidity risk, BCBS says that ‘Liquidity is the ability of a bank to fund increases in assets and meet
obligations as they come due, without incurring unacceptable losses. The fundamental role of banks in the maturity
transformation of short-term deposits into long-term loans makes banks inherently vulnerable to liquidity risk, both
of an institution-specific nature and that which affects markets as a whole.’ (BCBS, 2008, p. 1).
20
We thought of including analyst coverage as a control variable but complete data are not available for almost 20%
of our sample banks which makes it impossible to include this variable in our empirical analysis. Moreover,
including analyst coverage might cause a further endogeneity problem because some banks might attempt to enhance
risk disclosures to attract more analysts.

18
Worldwide Governance Indicators Database. Sixth, we control for the degree of national bank
concentration using the proportion of total assets of the five largest banks to total national
banking assets. Higher bank concentration might indicate stronger monopolistic power on the
part of the largest banks and hence weaker bank competition. It is important to use this control
variable in order to isolate the impact of competition arising from the national banking structure
when assessing the association between anti-competition regulations (i.e., entry to banking
requirements) and risk reporting quality.21 Data on bank concentration are collected from the
website of the European Central Bank (ECB). Finally, we control for time fixed-effects using
year dummies (the fiscal year 2008 is used as the reference category) and type fixed-effects using
bank specialization dummies (the “bank holdings and holding companies” specialization is used
as the reference category).

4.4 Econometric Model

To account for the endogeneity, auto correlation, and unobserved heterogeneity inherent in the
panel structure of our sample, we use a two stage random-effects linear panel-data specification
with generalized least squares (GLS). The model is typically composed of two stages. In the first-
stage regression, our endogenous variables (i.e., Z-score and stock return volatility) are the
dependent variables regressed on both their instrumental variables (i.e., excluded instruments)
and exogenous variables of the main regression (i.e., included instruments). In the second-stage
regression, the estimated Z-score and stock return volatility from the first-stage regression are
used as explanatory variables. Formally, our aim is to estimate the following two-stage model for
each bank i of type j in a country c for the fiscal year t:

21
In an unreported robustness check, we use the Herfindahl index, which equals the sum of the squares of the market
shares of all banks in a given country, as a measure of national bank concentration and get qualitatively similar
results.

19
 First-stage regression:

Second-stage regression:

where and are the estimated Z-score and stock return volatility
from the first-stage regression, yeart captures time fixed-effects, typej captures type fixed-
effects22, regulation is a set of variables capturing the country level bank regulation and
supervision (i.e., activity_restrictions, entry_requirements, supervisor_pwrind), governance is a
set of variables capturing bank level governance (i.e., brd_outside, largest_own,
government_own, executive_own, ac_meet), and and are error terms. The interaction
regulationc * governanceijct captures the joint effects of bank governance, regulation, and
supervision on ORD quality. Appendix C presents the definitions and data sources of all
variables used in our empirical analysis.

5. Empirical Results

5.1 Descriptive Statistics

Table 1 presents our descriptive statistics. There is a high variation in total ORD quality score
which ranges from a 3-point minimum to a 28-point maximum. Given the 6-point maximum

22
Petersen (2009) recommends that when both a firm effect and a time effect exist in panel data (i.e., residuals are
correlated across years and firms), standard errors are unbiased if one effect is isolated parametrically (i.e., by
including firm or year dummies) and standard errors are clustered on the other dimension. Following this
recommendation, we do two things: a) parametrically, we isolate the time effect by including year dummies (i.e.,
time fixed-effects) and the time-invariant part of the firm effect by including specialization dummies (i.e., type fixed-
effects); b) we cluster standard errors by firm in order to isolate the remaining part of the firm effect. Even when we
eliminate both year dummies and specialization dummies and just cluster standard errors by both firm and year, we
still get qualitatively similar results (unreported).

20
mandatory ORD score possible and 56-point maximum total ORD score possible, the statistics
mentioned show that: a) approximately 15% of our sample banks do not fully comply with
mandatory ORD requirements, most probably due to their laxity and the absence of the relevant
legal enforcement in individual EU countries, and b) total ORD quality provided by EU banks is
still moderate, thus leaving much potential for future enhancements. The considerable variation
in total ORD quality score is also reflected in a coefficient of variation of 0.46 thus implying
much inconsistency in ORD quality in our sample and revealing a remarkable level of discretion
in the bank management’ s decision regarding ORD quality in the annual reports and risk
reports.23

Similar findings hold for voluntary ORD quality score which ranges from a 0-point minimum
to a 24-point maximum. Given the 50-point maximum voluntary ORD score possible, these
statistics show that: a) approximately 7.50% of our sample banks do not provide voluntary ORD,
and b) voluntary ORD quality provided by EU banks tends to be low thus calling for more future
improvements. It is also noteworthy that the mean (median) voluntary ORD quality score is 9.19
(9) points amounting to approximately 70% of the mean (median) total ORD quality score, which
is 13.25 (13) points, thus indicating that the majority of ORD in our sample is voluntary in
nature.

Finally, all (unreported) normality tests (i.e., Shapiro-Wilk test (Shapiro and Wilk, 1965) and
Shapiro-Francia test (Shapiro and Francia, 1972)) and skewness/kurtosis test (D’ Agostino et al.,
1990; Royston, 1991) are unable to reject the null hypothesis stating the normality of the
distributions of both total and voluntary ORD scores on a year-by-year basis.

5.2 Univariate Analysis

Table 2 presents the Pearson’ s correlation matrix. The correlation coefficients between ORD
quality scores and all control variables coincide with extant literature and our expectations, thus
supporting the validity of our disclosure index. As expected, ORD quality scores are positively
and significantly correlated with the variables total_assets, cross_listing, ama_dum, ruleoflaw,

23
The coefficient of variation is the standard deviation of the variable divided by its mean. It is considered a
normalized measure of the dispersion of probability distribution. The higher this coefficient, the more dispersed the
probability distribution of the variable is.

21
and antidirector (insignificant), and negatively and significantly correlated with the variable
concentration.

As posited by our hypotheses, ORD quality scores are negatively correlated with the variables
activity_restrictions, entry_requirements (significant), and executive_own, and positively
correlated with the variables brd_outside, government_own, and ac_meet (significant). Contrary
to our hypotheses, the variables supervisor_pwrind (significant) and largest_own (significant) are
both negatively correlated with ORD quality scores. However, all univariate results should be
treated as suggestive because they do not account for the effects of the interrelationships between
control and explanatory variables which are explicitly controlled for in our multivariate analysis.

5.3 Multivariate Analysis

5.3.1 Baseline Regressions

Table 3 presents the results of our baseline regressions estimating the direct effects of bank
governance, regulation, and supervision on ORD quality in European banks according to
Equations (1) and (2). Panel A presents the results of the first-stage regression where Z-score and
stock return volatility are the dependent variables, and Panel B presents the results of the second-
stage regression where total/voluntary ORD quality score is the dependent variable and both Z-
score and stock return volatility are the endogenous variables (i.e., the estimated Z-score and
stock return volatility from the first-stage regression are used as explanatory variables in the
second-stage regression).

First-Stage Regression

For the first-stage regression (Table 3 - Panel A), we focus on testing: a) the endogeneity of Z-
score and stock return volatility (i.e., whether the endogenous Z-score and stock return volatility
have no effect on ORD quality), b) the validity (strict exogeneity) of our instrumental variables
(i.e., whether our instrumental variables affect ORD quality only through their impact on Z-score
and stock return volatility), and c) the relevance (strength) of our instrumental variables (i.e.,
whether our instrumental variables adequately explain variations in Z-score and stock return
volatility).

To examine the endogeneity of Z-score and stock return volatility, we employ the Stock-
Wright test (henceforth, endogeneity test) introduced by Stock and Wright (2000). As clear from
22
Table 3 (Panel A), low Stock-Wright statistics conforming to p-values of higher than 0.10 cannot
reject the null hypothesis of the zero joint effect of the endogenous variables in the main equation
thus supporting our selection of Z-score and stock return volatility as endogenous variables.

To inspect the validity (strict exogeneity) of our instrumental variables, we use the Sargan test
(henceforth, overidentification test) introduced by Sargan (1958). As obvious from Table 3
(Panel A), small Sargan statistics corresponding to p-values of higher than 0.10 cannot reject the
null hypothesis of no correlation between the first-stage residuals and exogenous variables in the
main equation thus confirming the validity (strict exogeneity) of our instrumental variables.

To scrutinize the relevance (strength) of our instrumental variables, we use the F-test and
partial R2 of excluded instruments. In Table 3 (Panel A), the statistics for the F-test of excluded
instruments with almost null p-values strongly reject the null hypothesis that our instrumental
variables can be omitted from the first-stage regression. Moreover, the partial R2 of excluded
instruments ranges from 0.1142 to 0.1500 thus showing that our instrumental variables explain
considerable variations in Z-score and stock return volatility. Hence, these statistics confirm the
absence of the problem of weak instruments.24

24
When we re-ran the first-stage regression using each instrumental variable separately, the p-value of the F-test of
excluded instruments never came out lower than 10%, indicating that each instrumental variable on its own cannot
be excluded from the first-stage regression.

23
Second-Stage Regression

We proceed to discuss the results of the second-stage regression (Table 3 - Panel B) in order to
empirically test our hypotheses. As expected, Z-score enters negatively, supporting our view that
more stable banks disclose lower risk information. However, stock return volatility enters
unexpectedly negative yet insignificant. This might imply that, in making its risk disclosure
choice, bank management considers capital adequacy and profitability to be more powerful than
stock price fluctuations as an indicator of bank stability.

We begin by discussing the results for total ORD quality (Model 1). As posited in our
hypothesis H2, more stringent entry to banking requirements negatively impact ORD quality. In
Table 3 (Panel B – Model 1), a one-point increase in the variable entry_requirements would
move a bank with a median total ORD quality score to the 46th percentile of the total disclosure
score. However, our results do not provide evidence that activity restrictions or bank supervisors
produce any direct impact on ORD quality. Hence, our hypotheses H1 and H3 are not empirically
supported.

As expected in our hypotheses H4 and H8, outside board directors and active audit committees
enhance ORD quality. In Table 3 (Panel B – Model 1), one standard deviation increase in the
variable brd_outside (i.e., 18.37%) would move a bank with a median total ORD quality score to
the 58th percentile of the total disclosure score. In Table 3 (Panel B – Model 1), one standard
deviation increase in the variable ac_meet (i.e., 5 meetings) would move a bank with a median
total ORD quality score to the 59th percentile of the total disclosure score. However, our results
do not show that ownership structure has any direct effect on ORD quality. Hence, our
hypotheses H5, H6, and H7 are not empirically supported.

As is obvious from Table 3 (Panel B - Model 2), the results for voluntary ORD quality are
qualitatively similar to those for total ORD quality. This finding is expected because, as our
descriptive statistics reveal, the majority of ORD in our sample is voluntary in nature and hence
results seem to be driven by the impact of explanatory variables on the discretionary decision of
bank management to provide voluntary ORD of a certain quality. Moreover, in unreported
baseline regressions, mandatory ORD scores exhibit no clear association with explanatory

24
variables, thus initially implying that mandatory ORD quality is merely determined by the degree
of bank management’ s compliance with ORD requirements in the CRD (EP, 2006).

For the variables that turn out to be insignificant, we expect that the interaction regressions
would reveal more information as we anticipate that the impact of sticky country level bank
regulation and supervision schemes on the quality of risk disclosures could differ according to the
ownership structures of individual banks.

5.3.2 Interactions between Bank Governance, Regulation, and Supervision

Table 4 presents the results of interaction regressions estimating the joint effects of bank
governance, regulation, and supervision on ORD quality in European banks according to
Equations (1) and (2). For the sake of brevity, we report only the coefficients of the explanatory,
endogenous, and instrumental variables and the interaction terms. The coefficients of control
variables are qualitatively similar to those presented in Table 3.

For the first-stage regression (Table 4 - Panel A), similar to the baseline regressions, the
statistics of the endogeneity test, overidentification test, and F-test and Partial R2 of excluded
instruments confirm the presence of endogeneity, and the validity and relevance of our
instrumental variables.

For the second-stage regression (Table 4 - Panel B), we begin by discussing the results for
total ORD quality (Model 1). First, the variable largest_own enters significantly positive only if
total ORD quality score is the dependent variable. This result suggests that largest shareholders
only monitor the compliance of bank management with mandatory ORD requirements but not its
choice regarding voluntary ORD. This finding is intuitive; since what should matter most for
largest shareholders is that their own wealth is not adversely affected if their bank is sanctioned
by regulatory bodies or supervisory authorities for violating obligatory risk disclosures. Hence,
our hypothesis H5 is empirically supported only for mandatory risk disclosures. Second,
coinciding with agency theory and organization-society theories, the negative coefficient of the
interaction term supervisor_pwrind * largest_own shows that bank supervisors and largest

25
shareholders perform substitutive roles as outside monitors of bank management’ s compliance
with mandatory ORD requirements. 25

Regarding voluntary ORD quality (Table 4 – Panel B – Model 2), unlike largest shareholders,
the variable executive_own enters significantly negative only if voluntary ORD quality score is
the dependent variable. This result shows that the entrenchment of bank executives, due to
stronger voting rights, could mitigate their incentives to voluntarily provide ORD but not their
compliance with mandatory ORD requirements. This finding is also intuitive; since, regardless of
their own voting rights, bank executives should be willing to adhere to obligatory risk disclosures
to protect their own reputation from being harmed if their bank is sanctioned by regulatory bodies
or supervisory authorities for violating obligatory risk disclosures. Hence, our hypothesis H7 is
empirically supported only for voluntary risk disclosures. In accordance with management
entrenchment theory and organization-society theories, the positive coefficient of the interaction
term supervisor_pwrind * executive_own shows that bank supervisors, as country level
influential stakeholders, mitigate the incentives for entrenched bank executives to withhold
voluntary ORD from the public.

To summarize, the results of interaction regressions reveal that: a) bank supervisors, as

country level outside monitors, substitute largest shareholders, as bank level outside monitors, in
fostering bank management’ s compliance with mandatory ORD requirements, and b) bank
supervisors, as country level influential stakeholders, counter the adverse impact of bank
executives’ entrenchment on voluntary ORD quality. Hence, our hypothesis H10 is empirically
supported. Practically, the policy implication inferred from these findings is that, to enhance
ORD quality, bank supervisors should focus their advisory and monitoring efforts on banks
having higher executive ownership (i.e., lower voluntary ORD quality) and lower outside
ownership concentration (i.e., less compliance with mandatory ORD requirements).

Finally, our results provide no evidence that banking regulations (i.e., activity restrictions and
entry to banking requirements) affect ORD quality differently across various ownership
structures. Hence, our hypothesis H9 is not empirically supported. Moreover, our hypotheses H1

25
Note that, in unreported interaction regressions of mandatory ORD score, only the variables supervisor_pwrind
and largest_own enter significantly positive and their interaction term supervisor_pwrind * largest_own enters
significantly negative, while other explanatory variables and the whole model enter insignificant. These results
explicitly support our interpretation of the results for interaction regressions of total ORD score.

26
(activity restrictions) and H6 (governmental blockholders) are still not empirically supported.
This might indicate that: a) the diversification of investment portfolio and customer base which
are affected by restrictive regulations do not influence the bank management’ s choice of risk
disclosures, and b) governmental blockholders follow a free-riding strategy thus informally
delegating largest non-governmental shareholders to monitor the risk disclosure decisions taken
by bank management. Furthermore, year dummies and bank specialization dummies are
insignificant indicating that, ceteris paribus, ORD quality does not change over time or across
bank types.26

6. Conclusions

We aimed to theoretically justify and empirically examine the impact of bank governance,
regulation, and supervision on the discretionary decision of bank management to provide risk
disclosures of a certain quality, as proxied for by operational risk disclosure (ORD) quality in the
annual reports and risk reports of European banks. After controlling for the endogeneity between
bank stability and risk reporting quality, we found evidence that banks having a higher proportion
of outside board directors, lower executive ownership, concentrated outside non-governmental
ownership, and more active audit committee, and operating under regulations promoting bank
competition (i.e., less stringent entry to banking requirements) provide their stakeholders with
higher ORD quality.

Most importantly, we found that it would be misleading to consider the determinants of risk
reporting quality in banks without explicitly considering how bank governance, regulation, and
supervision interact in this context. Specifically, our results show that the contribution of bank
supervisors to enhancing ORD quality is positively associated with executive ownership and
negatively associated with concentrated outside non-governmental ownership of the individual
bank. Coinciding with organization-society theories, this finding proves that powerful and
independent bank supervisors can serve as effective outside monitors and influential stakeholders

26
In an unreported robustness check, we include a dummy for international banks (i.e., banks that have majority-
owned or wholly-owned branches or subsidiaries in at least one foreign country) yet it enters insignificantly, thus
implying that ORD quality does not differ between local and international banks. Moreover, we rerun our baseline
and interaction regressions for a sub-sample of 66 international banks and still get results that are qualitatively
similar to those for the main sample, thus indicating that national bank regulations and supervisors constitute the
most important factors in shaping risk disclosure policy in banks.

27
to mitigate the incentives for entrenched bank executives to withhold voluntary ORD from the
public. Moreover, it also confirms the substitutive roles of bank supervisors and largest
shareholders as outside monitors of bank management’ s compliance with mandatory ORD
requirements.

Practically, our findings recommend that risk disclosures in banks can be enhanced by
supporting the official powers and independence of bank supervisors (for example, by
establishing independent specialized national committees or task forces to monitor and advise
Pillar 3 disclosures in banks). In this regard, our findings lend support for the recommendations
of the Financial Policy Committee of the Bank of England (BoE) that banks should collaborate
more effectively with their national supervisors in enhancing the quality of their Pillar 3
disclosures (BoE, 2012). Furthermore, our findings recommend that bank supervisors should
focus their efforts on monitoring and promoting risk disclosures in banks having stronger
executive voting rights and dispersed outside non-governmental voting rights because these
banks tend to provide the lowest-quality risk disclosures.

As another practical implication, our findings show that banks can enhance their own risk
disclosures by hiring outside board directors and improving the activities of audit committees as
effective bank level advisors and monitors of risk disclosure decisions taken by bank
management. In this regard, our findings support the view of the Basel Committee on Banking
Supervision (BCBS) positing these traditional governance mechanisms as crucial drivers of
transparency in banks (BCBS, 2006a). Moreover, our findings call for less stringent banking
entry policies in order to enhance both competition and transparency in the banking sector.

Finally, our work points to avenues of future research which could examine the direct and
joint effects of bank governance, regulation, and supervision on the level of bank management’ s
compliance with mandatory disclosure requirements in IFRS 7 and CRD (EP, 2006). Moreover, a
closer investigation might be needed to compare the contributions of outside directors as board
members or audit committee members in enhancing the quality of risk reporting in banks.
Furthermore, it might be interesting to investigate the effects of ORD quality on the cost of
equity and debt capitals in banks.

28
References
Abbott, L. J., Park, Y., & Parker, S. (2000). The effects of audit committee activity and independence on
corporate fraud. Managerial Finance, 26(11), 55-68.
Abou El Sood, H. (2012). Loan loss provisioning and income smoothing in US banks pre and post the
financial crisis. International Review of Financial Analysis, 25, 64-72.
Admati, A., Pfleiderer, P., & Zechner, J. (1994). Large shareholder activism, risk sharing, and financial
market equilibrium. Journal of Political Economy, 102(6), 1097-1130.
Anderson, K. L., Deli, D. N., & Gillan, S. L. (2003). Boards of directors, audit committees, and the
information content of earnings. Working Paper Weinberg Center for Corporate Governance.
Bank of England (2012). Record of the interim Financial Policy Committee meeting, 22nd June. Available
from: http://www.bankofengland.co.uk/publications/Documents/records/fpc/pdf/2012/record1207.pdf.
Barth, J. R., Caprio, G., & Levine, R. (2001). The regulation and supervision of banks around the world a
new database. Working Paper The World Bank.
Barth, J. R., Caprio, G., & Levine, R. (2004). Bank supervision and regulation: What works best? Journal
of Financial Intermediation, 13, 205-248.
Barth, J. R., Caprio, G., & Levine, R. (2006). Rethinking Bank Regulation: Till Angels Govern.
Cambridge University Press, New York.
Barth, J. R., Caprio, G., & Levine, R. (2008). Bank regulations are changing: For better or worse?
Working Paper The World Bank.
Basel Committee on Banking Supervision (1998a). Enhancing Bank Transparency. Bank for International
Settlements, Basel, Switzerland.
Basel Committee on Banking Supervision (1998b). Operational Risk Management. Bank for International
Settlements, Basel, Switzerland.
Basel Committee on Banking Supervision (2001).Operational Risk. Bank for International Settlements,
Basel, Switzerland.
Basel Committee on Banking Supervision (2006a). Enhancing Corporate Governance for Banking
Organisations. Bank for International Settlements, Basel, Switzerland.
Basel Committee on Banking Supervision (2006b). International Convergence of Capital Measurement
and Capital Standards. Bank for International Settlements, Basel, Switzerland.
Basel Committee on Banking Supervision (2008). Principles for Sound Liquidity Risk Management and
Supervision. Bank for International Settlements, Basel, Switzerland.
Beasley, M. S., Carcello, J. V., Hermanson, D. R., & Lapides, P. D. (2000). Fraudulent financial
reporting: Consideration of industry traits and corporate governance mechanisms. Accounting
Horizons, 14(4), 441-454.
Berger, P. G. (2011). Challenges and opportunities in disclosure research: A discussion of the financial
reporting environment: Review of the recent literature. Journal of Accounting and Economics, 51(1-2),
204-218.
Bessis J. (2002). Risk management in banking. Wiley, Chichester.
Beyer, A., Cohen, D. A., Lys, T. Z., & Walther, B. R. (2010). The financial reporting environment:
Review of the recent literature. Journal of Accounting and Economics, 50, 296-343.
Bischof, J. (2009). The effects of IFRS 7 adoption on bank disclosure in Europe. Accounting in Europe,
6(2), 167-194.
Boot, A. W., & Schmeits, A. (2000). Market discipline and incentive problems in conglomerate firms with
applications to banking. Journal of Financial Intermediation, 9, 240–273.
Botosan, C. A. (1997). Disclosure level and the cost of equity capital. The Accounting Review, 72(3), 323-
349.
Botosan, C. A., & Plumlee, M. A. (2002). A re-examination of disclosure level and the expected cost of
equity capital. Journal of Accounting Research, 40(1), 21-40.

29
Boyd, B. (1990). Corporate linkages and organizational environment: A test of the resource dependence
model. Strategic Management Journal, 11(6), 419-430.
Carpenter, M., & Westphal, J. (2001). The strategic context of external network ties: Examining the
impact of board appointments on board involvement in strategic decision making. Academy of
Management Journal, 44, 639-660.
Chen, J. C. & Roberts, R. W. (2010). Toward a more coherent understanding of the organization–society
relationship: A theoretical consideration for social and environmental accounting research. Journal of
Business Ethics, 97, 651-665.
Collier, P. & Gregory, A. (1999). Audit committee activity and agency costs. Journal of Accounting and
Public Policy, 18(4-5), 311-332.
Committee of European Banking Supervisors (2009). Assessment of Banks’ Pillar 3 Disclosures.
Cordella, T., & Yeyati, E. L. (1998). Public disclosure and bank failures. Staff Papers - International
Monetary Fund, 45(1), 110-131.
Corvoisier, S., & Gropp, R. (2002). Bank concentration and retail interest rates. Journal of Banking and
Finance, 26, 2155-2189.
D’ Agostino, R. B., Belanger A., & D’ Agostino, R. B., Jr. (1990). A suggestion for using powerful and
informative tests of normality. American Statistician, 44(4), 316- 321.
Dalton D. R., Daily C. M., Johnson J. L., & Ellstrand, A. E. (1999). Number of directors and financial
performance: A meta-analysis. Academy of Management Journal, 42, 674-684.
Daníelsson, J., Embrechts, P., Goodhart, C., Keating, C., Muennich, F., Renault, O., & Shin, H. S. (2001).
An academic response to Basel II. Special Paper No. 130, LSE Financial Markets Group.
Deloitte (2009). Basel II Pillar 3 Benchmarking Survey 2009.
DeZoort, F. T., Hermanson, D. R., Archambeault, D. S., & Reed, S. A. (2002). Audit committee
effectiveness: A synthesis of the empirical audit committee literature. Journal of Accounting
Literature, 21, 38-75.
Djankov, S., La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. (2008). The law and economics of self-
dealing. Journal of Financial Economics, 88, 430-465.
Dowling, J. & Pfeffer, J. (1975). Organizational legitimacy: Social values and organizational behavior.
Pacific Sociological Review, 18, 122-136.
Dugan, J. C., & Xi, J. (2011). U.S. implementation of Basel II: Final rules issued but no supervisory
approvals to date. Briefing Note, Policy Department, European Parliament.
Easley, D., & O' Hara, M. (2004). Information and the cost of capital. Journal of Finance, 59, 1553-1583.
Eng, L., & Mak, Y. (2003). Corporate governance and voluntary disclosure. Journal of Accounting and
Public Policy, 22(4), 325-345.
Esty, B. (1998). The impact of contingent liability on commercial bank risk taking. Journal of Financial
Economics, 47, 189-218.
European Parliament (2006). Directive 2006/48/EC of the European Parliament and the Council of the
European Union (Capital Requirements Directive).
Fama, E. (1980). Agency problems and the theory of the firm. Journal of Political Economy, 88(2), 288-
307.
Fama, E., & Jensen, M. (1983). Separation of ownership and control. Journal of Law and Economics,
26(2), 301-325.
Financial Stability Board (2012). Enhancing the risk disclosures of banks, Report of the Enhanced
Disclosure Task Force, Basel, Switzerland.
Ford G., Sundmacher M., Finch N., & Carlin T. M. (2009). Operational risk disclosure in financial
services firms, in: Gregoriou, G. N. (Ed.), Operational risk toward Basel III: Best practices and issues
in modeling, management and regulation. Hoboken, John Wiley & Sons, Inc, NJ, pp. 381-395.
Frankel, R., McVay, S., & Soliman, M. (2011). Non-GAAP earnings and board independence. Review of
Accounting Studies, 16, 719-744.
Freeman, R. E. (1984). Strategic management: A stakeholder approach. Pitman, Boston, Massachusetts.

30
Gelb, D. S. (2000). Managerial ownership and accounting disclosure: An empirical study. Review of
Quantitative Finance and Accounting, 15(2), 169-185.
Hannan, T. H., & Berger, A. N. (1991). The rigidity of prices: Evidence from the banking industry. The
American Economic Review, 81(4), 938-945.
Helbok, G., & Wagner, C. (2006). Determinants of operational risk reporting in the banking industry. The
Journal of Risk, 9(1), 49-74.
Hillman A., Cannella A., & Paetzold, R. (2000). The resource dependence role of corporate directors:
Strategic adaptation of board composition in response to environmental change. Journal of
Management Studies, 37, 213-255.
Huddart, S. (1993). The effect of a large shareholder on corporate value. Management Science, 39(4),
1407-1421.
Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Management behaviour, agency costs and
ownership structure. Journal of Financial Economics, 3(3), 305-360.
Kanagaretnam, K., Lobo, G., & Yang, D. (2004). Joint tests of signaling and income smoothing through
bank loan loss provisions. Contemporary Accounting Research, 21(4), 843-844.
Kelton, A. S., & Yang, Y. (2008). The impact of corporate governance on Internet financial reporting.
Journal of Accounting and Public Policy, 27(1), 62-87.
Klein, A. (2002). Audit committee, board of director characteristics, and earnings management. Journal of
Accounting and Economics, 33(3), 375-400.
Kothari, S., Li, X., & Short, J. (2009). The effect of disclosure by management, analysts, and business
press on cost of capital, return volatility, and analysts forecast: Study using content analysis. The
Accounting Review, 84(5), 1639-1670.
Laeven, L., & Levine, R. (2009). Bank governance, regulation and risk taking. Journal of Financial
Economics, 93, 259-275.
Li, J., Pike, R., & Haniffa, R. (2008). Intellectual capital disclosure and corporate governance structure in
UK firm. Accounting and Business Research, 38(2), 137-159.
Lindblom, C. K. (1994). The implications of organizational legitimacy for corporate social performance
and disclosure. Working Paper, Critical Perspectives on Accounting Conference, New York.
Linsley, P. M., & Shrives, P. J. (2005). Examining risk reporting in UK public companies. Journal of Risk
Finance, 6(4), 292-305.
Linsley, P. M., & Shrives, P. J. (2006). Risk reporting: A study of risk disclosures in the annual reports of
UK companies. The British Accounting Review, 38(1), 387-404.
Linsley, P. M., Shrives, P. J., & Crumpton, M. (2006). Risk disclosure: An exploratory study of UK and
Canadian banks. Journal of Banking Regulation, 7(3-4), 268-282.
Maudos, J., & Fernández de Guevara, J. (2004). Factors explaining the interest margin in the banking
sectors of the European Union. Journal of Banking and Finance, 28, 2259-2281.
Maug, E. (1998). Large shareholders as monitors: Is there a trade-off between liquidity and control?
Journal of Finance, 53(1), 65-98.
Maurer, J. G. (1971). Readings in organization theory: Open system approach. Random House, New
York.
Nagar, V., Nanda, D., & Wysocki, P. (2003). Discretionary disclosure and stock-based incentives. Journal
of Accounting and Economics, 34(1-3), 283-309.
Nier, E., & Baumann, U. (2004). Disclosure, volatility, and transparency: An empirical investigation into
the value of bank disclosure. Federal Reserve Bank of New York. Economic Policy Review, 10(2), 31-
45.
Nier, E., & Baumann, U. (2006). Market discipline, disclosure and moral hazard in banking. Journal of
Financial Intermediation, 15, 332–361.
Noe, T. (2002). Institutional activism and financial market structure. Review of Financial Studies, 15, 289-
319.
Oliveira, J., Rodrigues, L. L., & Craig, R. (2011a). Risk-related disclosure practices in the annual reports
of Portuguese credit institutions: An exploratory study. Journal of Banking Regulation, 12(2), 100-118.

31
Oliveira, J., Rodrigues, L. L., & Craig, R. (2011b). Voluntary risk reporting to enhance institutional and
organizational legitimacy: Evidence from Portuguese banks. Journal of Financial Regulation and
Compliance, 19(3), 271-289.
Perry, J., & de Fontnouvelle, P. (2005). Measuring reputational risk: The market reaction to operational
loss announcements. Working Paper Federal Reserve Bank of Boston.
Petersen, M. A. (2009). Estimating standard errors in finance panel data sets: Comparing approaches,
Review of Financial Studies, 22(1), 435-480.
Pfeffer, J. (1972). Size and composition of corporate boards of director: The organization and its
environment. Administrative Science Quarterly, 17, 218-229.
Pfeffer, J. & Salancik, G. R. (1978). The external control of organizations: A resource dependence
perspective. Harper and Row, New York.
Pfeffer, J. & Salancik, G. R. (2003). The external control of organizations: A resource dependence
perspective. Stanford University Press, Stanford, California.
Royston, J. P. (1991). Comment on sg3.4 and an improved D’ Agostino test. Stata Technical Bulletin, 3,
13-24.
Ruland, W., Tung, S., & George, N. (1990). Factors associated with the disclosure of managers’ forecasts.
The Accounting Review, 65(3), 710-721.
Sargan, J. D. (1958). The estimation of economic relationships using instrumental variables.
Econometrica, 26(3), 393-415.
Saunders, A., Strock, E., & Travlos, N. G. (1990). Ownership structure, deregulation, and bank risk
taking. Journal of Finance, 45, 643-654.
Shapiro, S. S. & Francia, R. S. (1972). An approximate analysis of variance test for normality. Journal of
the American Statistical Association, 67, 215-216.
Shapiro, S. S. & Wilk, M. B. (1965). An analysis of variance test for normality (complete samples).
Biometrika, 52 (3-4), 591-611.
Shleifer, A., & Vishny, R. W. (1986). Large shareholders and corporate control. Journal of Political
Economy, 94(3), 461-488.
Shrieves, R., & Dahl, D. (2003). Discretionary accounting and the behavior of Japanese banks under
financial duress. Journal of Banking and Finance, 27, 1219-1243.
Stock, J. H. & Wright, J. H. (2000). GMM with weak identification. Econometrica, 68(5), 1055-1096.
Suchman, M. C. (1995). Managing legitimacy: Strategic and institutional approaches. Academy of
Management Review, 20(3), 571-610.
Uhde, A. & Heimeshoff, U. (2009). Consolidation in banking and financial stability in Europe: Empirical
evidence. Journal of Banking and Finance, 33, 1299–1311.
Walsh J., & Seward, J. (1990). On the efficiency of internal and external corporate control mechanisms.
Academy of Management Review, 15, 421-458.
Woods, M., Dowd, K., & Humphrey, C. (2009). Market risk reporting by the world’ s top banks: Evidence
on the diversity of reporting practice and the implications for international accounting harmonization.
Spanish Accounting Review, 11(2), 9-42.
World Bank (2008). Database on Bank Regulation and Supervision around the World. Available from:
http://econ.worldbank.org/WBSITE/EXTERNAL/EXTDEC/EXTRESEARCH/0,,contentMDK:203450
37~pagePK:64214825~piPK:64214943~theSitePK:469382,00.html.
Yong, H. H. A., Chalmers, K., & Faff, R. (2005). Asia Pacific banks'derivative and risk management
disclosures. Asian Review of Accounting, 13(1), 15-44.

32
Table 1: Descriptive Statistics

This table presents the descriptive statistics of tested variables. The median, mean, and standard deviation (SD) are computed after
winsorizing the relevant variables. The composition of the final sample is presented in Appendix B. Variable definitions are presented in
Appendix C.

Num.
Variable Min 1p 5p 25p Median Mean SD 75p 95p 99p Max
Obs.
A) Bank Level Variables
ord_score_total 243 3.00 3.00 4.00 8.00 13.00 13.25 6.10 18.00 23.00 27.00 28.00
ord_score_voluntary 243 0.00 0.00 0.00 5.00 9.00 9.19 5.71 14.00 19.00 22.00 24.00
brd_outside 243 0.00 28.57 46.67 62.50 75.00 74.01 18.37 85.71 100.00 100.00 100.00
largest_own 243 0.00 0.01 1.20 6.32 14.55 25.22 23.22 41.70 71.50 92.00 98.62
government_own 243 0.00 0.00 0.00 0.00 0.00 0.14 0.35 0.00 1.00 1.00 1.00
executive_own 243 0.00 0.00 0.00 0.00 0.01 0.79 2.84 0.12 3.49 42.75 60.28
ac_meet 243 0.00 0.00 2.00 4.00 7.00 7.89 5.08 10.00 18.00 31.00 46.00
z_score 243 1.30 1.68 2.36 3.61 4.34 4.44 1.39 5.12 6.70 9.35 9.56
return_volatility 243 0.14 0.30 1.21 2.05 2.86 3.13 1.54 3.71 5.74 7.86 10.47
cost_income 243 0.62 29.31 42.54 51.84 61.54 62.25 18.31 69.50 84.96 138.59 185.39
cash_deposits 243 5.17 11.77 20.46 35.87 59.00 86.81 88.03 91.44 261.99 467.21 648.39
total_assets 243 6.91 6.97 7.29 9.31 10.88 10.99 2.07 12.49 14.32 14.65 14.74
cross_listing 243 0.00 0.00 0.00 0.00 0.00 0.14 0.34 0.00 1.00 1.00 1.00
ama_dum 243 0.00 0.00 0.00 0.00 0.00 0.17 0.38 0.00 1.00 1.00 1.00
B) Country Level Variables
activity_restrictions 20 4.00 4.00 5.00 7.00 9.00 8.75 2.11 10.50 12.00 12.00 12.00
entry_requirements 20 2.00 2.00 2.00 3.00 4.00 3.40 0.74 4.00 4.00 4.00 4.00
supervisor_pwrind 20 5.00 5.00 5.75 7.75 9.75 10.23 3.00 13.00 14.75 15.00 15.00
gdp_capita 20 9.02 9.02 9.33 10.22 10.75 10.56 0.60 10.99 11.20 11.31 11.31
inflation 20 -4.48 -4.48 -0.66 1.20 2.34 2.50 2.51 3.94 5.47 12.35 12.35
political_stability 20 -0.33 -0.33 -0.17 0.47 0.79 0.73 0.42 1.04 1.36 1.45 1.45
antidirector 20 2.00 2.00 2.00 2.50 3.50 3.33 1.00 4.00 5.00 5.00 5.00
ruleoflaw 20 -0.19 -0.19 0.15 0.78 1.43 1.27 0.59 1.77 1.95 1.97 1.97
concentration 20 22.70 22.70 31.10 43.65 58.55 58.52 17.06 71.00 84.10 86.80 86.80

33
Table 2: Correlation Matrix
This table presents the Pearson’ s correlation matrix of tested variables. Correlation coefficients are reported in bold. p-values are reported in parentheses. ***, **,
and * denote significance of the correlation coefficients at the 1%, 5%, and 10% levels, respectively. The composition of the final sample is presented in Appendix
B. Variable definitions are presented in Appendix C.

Panel A:
ord_score_ ord_score_ activity_ entry_ supervisor_ government_
Variable brd_outside largest_own executive_own ac_meet z_score
total voluntary restrictions requirements pwrind own
ord_score_total 1.0000

ord_score_voluntary 0.9917 1.0000

(0.0000)***
activity_restrictions -0.0678 -0.0802 1.0000
(0.2922) (0.2128)
entry_requirements -0.2764 -0.2599 0.0149 1.0000
(0.0000)*** (0.0000)*** (0.8177)
supervisor_pwrind -0.2727 -0.2953 0.0708 0.1646 1.0000
(0.0000)*** (0.0000)*** (0.2715) (0.0102)**
brd_outside 0.0319 0.0150 0.0187 0.0959 0.1071 1.0000
(0.6211) (0.8166) (0.7721) (0.136) (0.0956)*
largest_own -0.2080 -0.2094 0.0735 -0.0430 -0.1273 -0.0369 1.0000
(0.0011)*** (0.001)*** (0.2538) (0.5051) (0.0475)** (0.5674)
government_own 0.0070 -0.0051 -0.2157 0.1117 0.0375 0.0840 -0.0820 1.0000
(0.9133) (0.9373) (0.0007)*** (0.0822)* (0.5612) (0.1919) (0.2026)
executive_own -0.0472 -0.0600 0.0033 -0.0170 0.2796 -0.2589 -0.0321 -0.1109 1.0000
(0.4641) (0.3516) (0.9592) (0.7915) (0.0000)*** (0.0000)*** (0.6184) (0.0845)*
ac_meet 0.3860 0.3811 0.2964 0.0432 -0.0830 0.0854 -0.2065 -0.0520 -0.0361 1.0000
(0.0000)*** (0.0000)*** (0.0000)*** (0.5031) (0.1972) (0.1844) (0.0012)*** (0.4195) (0.5759)
z_score 0.1210 0.1258 0.1794 -0.1238 -0.0801 -0.1485 -0.0534 -0.1291 -0.0200 0.1411 1.0000
(0.0596)* (0.0502)* (0.005)*** (0.054)* (0.2137) (0.0206)** (0.4076) (0.0444)** (0.7566) (0.0279)**
return_volatility 0.1873 0.1922 -0.2600 -0.0906 -0.0321 0.0489 -0.1712 0.1703 -0.0614 0.1083 -0.1120
(0.0034)*** (0.0026)*** (0.0000)*** (0.1589) (0.6187) (0.4483) (0.0075)*** (0.0078)*** (0.3406) (0.0920)* (0.0814)*
cost_income 0.2676 0.2612 -0.0726 -0.1401 -0.2619 -0.0040 0.2486 0.0266 -0.0508 0.0668 -0.1959
(0.0000)*** (0.0000)*** (0.2593) (0.029)** (0.0000)*** (0.9505) (0.0001)*** (0.6795) (0.4304) (0.2995) (0.0022)***
cash_deposits 0.2592 0.2641 -0.1813 -0.1532 -0.2910 -0.1700 0.1455 -0.0416 -0.0759 0.1484 0.0631
(0.0000)*** (0.0000)*** (0.0046)*** (0.0168)** (0.0000)*** (0.0079)*** (0.0233)** (0.5188) (0.2387) (0.0206)** (0.3276)
gdp_capita 0.2430 0.2896 -0.3138 -0.0605 -0.4530 -0.3154 -0.1170 -0.0740 -0.0943 -0.0135 0.1283
(0.0001)*** (0.0000)*** (0.0000)*** (0.3475) (0.0000)*** (0.0000)*** (0.0686)* (0.2506) (0.1426) (0.8341) (0.0458)**
inflation -0.2432 -0.2527 -0.0432 0.1877 0.1551 0.1566 0.0095 0.0149 -0.0618 -0.1499 -0.1465
(0.0001)*** (0.0001)*** (0.5024) (0.0033)*** (0.0155)** (0.0145)** (0.8834) (0.8174) (0.3377) (0.0194)** (0.0223)**
political_stability -0.1212 -0.1026 -0.0228 -0.0894 -0.1037 -0.1388 0.1242 0.1585 -0.1608 -0.3662 0.0038
(0.0592)* (0.1105) (0.7234) (0.1648) (0.1067) (0.0305)** (0.0531)* (0.0134)** (0.0121)** (0.0000)*** (0.9536)
total_assets 0.7073 0.7097 -0.2349 -0.2173 -0.2575 -0.1307 -0.3524 0.1178 -0.0851 0.3740 0.1707
(0.0000)*** (0.0000)*** (0.0002)*** (0.0006)*** (0.0000)*** (0.0418)** (0.0000)*** (0.0667)* (0.1862) (0.0000)*** (0.0077)***
cross_listing 0.2698 0.2819 -0.4409 0.0398 0.0573 0.0221 -0.3332 0.1864 -0.0591 0.1296 -0.1377
(0.0000)*** (0.0000)*** (0.0000)*** (0.5371) (0.3739) (0.7316) (0.0000)*** (0.0035)*** (0.3593) (0.0436)** (0.0318)**
ama_dum 0.4174 0.4183 -0.0783 -0.0993 -0.2779 0.0268 -0.1764 0.0039 -0.1223 0.1604 0.0730
(0.0000)*** (0.0000)*** (0.2241) (0.1228) (0.0000)*** (0.6777) (0.0058)*** (0.9521) (0.0570)* (0.0123)** (0.2573)
antidirector 0.0687 0.0855 -0.5915 -0.1006 0.2262 -0.1037 -0.2108 0.0601 0.1123 -0.1453 -0.0071
(0.2858) (0.1839) (0.0000)*** (0.1176) (0.0004)*** (0.1067) (0.0009)*** (0.3507) (0.0807)* (0.0235)** (0.9120)
ruleoflaw 0.1306 0.1555 -0.5938 -0.2558 -0.0755 -0.1957 -0.0585 0.1010 -0.0208 -0.3884 0.0221
(0.042)** (0.0153)** (0.0000)*** (0.0001)*** (0.2409) (0.0022)*** (0.3639) (0.1164) (0.7475) (0.0000)*** (0.7318)
concentration -0.1422 -0.1567 0.0595 -0.0879 0.4013 0.2169 -0.0505 0.0270 0.0331 -0.2318 -0.2268
(0.0267)** (0.0145)** (0.3561) (0.1721) (0.0000)*** (0.0007)*** (0.4333) (0.6751) (0.6071) (0.0003)*** (0.0004)***

34
Table 2: Correlation Matrix (cont.)

Panel B:
return_ cash_ political_
Variable cost_income gdp_capita inflation total_assets cross_listing ama_dum antidirector ruleoflaw concentration
volatility deposits stability
return_volatility 1.0000

cost_income 0.2157 1.0000

(0.0007)***
cash_deposits 0.2172 0.4661 1.0000
(0.0007)*** (0.0000)***
gdp_capita 0.1413 0.0860 0.2519 1.0000
(0.0277)** (0.1817) (0.0001)***
inflation -0.1228 0.0522 -0.0764 -0.3799 1.0000
(0.0559)* (0.4178) (0.2355) (0.0000)***
political_stability 0.0405 0.0427 0.0818 0.3458 -0.1443 1.0000
(0.5301) (0.5080) (0.2038) (0.0000)*** (0.0245)**
total_assets 0.4169 0.1571 0.4225 0.3294 -0.1880 -0.1450 1.0000
(0.0000)*** (0.0142)** (0.0000)*** (0.0000)*** (0.0033)*** (0.0238)**
cross_listing 0.4134 -0.0036 0.1734 0.1670 -0.0407 -0.1373 0.4808 1.0000
(0.0000)*** (0.9556) (0.0067)*** (0.0091)*** (0.5273) (0.0324)** (0.0000)***
ama_dum 0.2161 0.1218 0.2412 0.1484 -0.0724 -0.0755 0.5714 0.2953 1.0000
(0.0007)*** (0.0580)* (0.0001)*** (0.0207)** (0.2607) (0.2409) (0.0000)*** (0.0000)***
antidirector 0.2467 -0.0861 0.1823 0.3039 -0.0373 -0.0650 0.2117 0.3910 0.0265 1.0000
(0.0001)*** (0.1808) (0.0044)*** (0.0000)*** (0.5624) (0.3128) (0.0009)*** (0.0000)*** (0.6808)
ruleoflaw 0.1541 0.0368 0.2541 0.6663 -0.2220 0.6052 0.1988 0.1781 0.0758 0.5925 1.0000
(0.0162)** (0.5680) (0.0001)*** (0.0000)*** (0.0005)*** (0.0000)*** (0.0018)*** (0.0054)*** (0.2390) (0.0000)***
concentration 0.0399 -0.1226 -0.1828 -0.0955 0.0857 0.2480 -0.2012 -0.0482 -0.2133 -0.0048 0.1919 1.0000
(0.5357) (0.0563)* (0.0042)*** (0.1377) (0.1828) (0.0001)*** (0.0016)*** (0.4545) (0.0008)*** (0.9410) (0.0027)***

35
Table 3: Determinants of Operational Risk Disclosure Quality in European Banks
(Baseline Regressions)

This table presents the results of the baseline regressions estimating the direct effects of bank governance, regulation, and supervision on
operational risk disclosure quality in European banks using a two-stage random-effects linear panel-data specification with generalized
least squares (GLS). z_score and return_volatility are the endogenous variables. z-statistics (in parentheses) are based on robust standard
errors clustered by firm. ***, **, and * denote significance of the regression coefficients at the 1%, 5%, and 10% levels, respectively. For
directional hypotheses (+/-), we use one-tailed tests. If the coefficient’ s sign is opposite from the one hypothesized, then significance
levels according to two-tailed tests are reported. The composition of the final sample is presented in Appendix B. Variable definitions are
presented in Appendix C.

Panel A: First-Stage Regression (z_score and return_volatility are the dependent variables. Results for instrumental variables are
reported in bold.)

(1) (2)
Total Operational Risk Voluntary Operational Risk
Variable Disclosure Quality Score Disclosure Quality Score
Exp. Exp. Exp. Exp.
z_score return_volatility z_score return_volatility
Sign Sign Sign Sign
cost_income - -0.0320 + 0.0103 - -0.0321 + 0.0104
(-3.51)*** (1.25) (-3.47)*** (1.24)
cash_deposits ? -0.0031 ? 0.0014 ? -0.0034 ? 0.0015
(-1.20) (0.40) (-1.27) (0.40)
gdp_capita + 0.0551 - 0.2741 + 0.0978 - 0.1750
(0.06) (0.38) (0.09) (0.22)
inflation ? 0.1900 ? -0.2537 ? 0.1906 ? -0.2517
(3.09)*** (-3.34)*** (3.06)*** (-3.29)***
political_stability + -0.9490 - 2.3515 + -1.0090 - 2.3091
(-0.67) (2.11)** (-0.66) (1.93)*
total_assets + 0.1117 - 0.4197 + 0.0687 - 0.4622
(0.73) (2.85)*** (0.38) (2.69)***
cross_listing ? -0.1534 ? 0.4396 ? -0.0527 ? 0.4270
(-0.28) (0.84) (-0.09) (0.78)
ama_dum ? -0.0884 ? -0.2228 ? 0.0083 ? -0.3405
(-0.16) (-0.51) (0.01) (-0.70)
antidirector + -0.2559 - 0.6291 + -0.2628 - 0.5981
(-0.73) (1.94)* (-0.70) (1.74)*
ruleoflaw + 1.7595 - -2.4071 + 1.8008 - -2.2616
(0.94) (-1.67)** (0.89) (-1.46)*
concentration - -0.0328 + 0.0110 - -0.0337 + 0.0093
(-1.86)** (0.66) (-1.66)** (0.50)
activity_restrictions + 0.2768 - -0.2096 + 0.2823 - -0.1967
(1.71)** (-1.67)** (1.61)* (-1.45)*
entry_requirements - -0.2499 + 0.0110 - -0.2753 + 0.0528
(-0.80) (0.05) (-0.79) (0.21)
supervisor_pwrind + 0.0050 - 0.0434 + 0.0036 - 0.0423
(0.04) (0.50) (0.03) (0.45)
brd_outside ? 0.0021 ? 0.0098 ? 0.0027 ? 0.0106
(0.25) (0.93) (0.30) (0.81)
largest_own ? 0.0082 ? 0.0015 ? 0.0083 ? 0.0020
(1.35) (0.40) (1.25) (0.49)
government_own ? 0.1220 ? -0.6846 ? 0.1268 ? -0.7313
(0.37) (-1.80)* (0.37) (-1.87)*
executive_own ? -0.0042 ? 0.0248 ? -0.0027 ? 0.0333
(-0.11) (0.38) (-0.06) (0.39)
ac_meet ? 0.0135 ? -0.0411 ? 0.0128 ? -0.0430
(0.29) (-1.04) (0.27) (-1.05)
constant 3.2758 -4.2984 3.3763 -3.9876
(0.39) (-0.57) (0.37) (-0.50)
F-test of excluded instruments 6.55 5.17 6.33 5.02
(Prob > F) (0.0000)*** (0.0004)*** (0.0000)*** (0.0004)***
Partial R2 of excluded instruments 0.1495 0.1165 0.1500 0.1142
Endogeneity (Stock-Wright) test 8.22 5.34
(p-value) (0.1447) (0.3762)
Overidentification (Sargan) test 0.276 3.501
(p-value) (0.9644) (0.3207)

36
Table 3: Determinants of Operational Risk Disclosure Quality in European Banks
(Baseline Regressions) (cont.)

Panel B: Second-Stage Regression (z_score and return_volatility are the endogenous variables)

(1) (2)
Exp. Total Operational Risk Voluntary Operational Risk
Variable
Sign Disclosure Quality Score Disclosure Quality Score
ord_score_total ord_score_voluntary
z_score - -0.6169 -0.3937
(-2.16)** (-1.63)*
return_volatility + -0.6288 -0.3315
(-1.38) (-1.08)
total_assets + 1.7765 1.4552
(5.07)*** (4.55)***
cross_listing + 0.2705 0.2904
(0.18) (0.21)
ama_dum + 0.3297 0.3803
(0.24) (0.29)
antidirector + 0.0072 0.1908
(0.01) (0.33)
ruleoflaw + 1.2621 0.9804
(1.33)* (1.11)
concentration - -0.0224 -0.0274
(-0.96) (-1.37)*
activity_restrictions - 0.1806 0.1661
(0.60) (0.57)
entry_requirements - -1.4152 -1.1236
(-2.19)** (-1.80)**
supervisor_pwrind + -0.1731 -0.2472
(-0.65) (-0.99)
brd_outside + 0.0538 0.0412
(2.17)** (1.81)**
largest_own + 0.0033 0.0013
(0.24) (0.11)
government_own + -0.3861 -0.3929
(-0.92) (-0.92)
executive_own - 0.1546 0.1150
(0.81) (0.64)
ac_meet + 0.2597 0.2675
(4.60)*** (5.27)***
constant -3.3411 -4.9241
(-0.54) (-0.78)
Year Dummies Included/Non-significant Included/Non-significant
Type Dummies Included/Non-significant Included/Non-significant
Num. Obs. 243 243
F-Statistic 60.81 35.09
(Prob > F) (0.0000)*** (0.0000)***
R2 0.7876 0.7081

37
Table 4: Determinants of Operational Risk Disclosure Quality in European Banks
(Interaction Regressions)

This table presents the results of the interaction regressions estimating the direct and joint effects of bank governance, regulation, and supervision on
operational risk disclosure quality in European banks using a two-stage random-effects linear panel-data specification with generalized least squares
(GLS). z_score and return_volatility are the endogenous variables. z-statistics (in parentheses) are based on robust standard errors clustered by firm. ***,
**, and * denote significance of the regression coefficients at the 1%, 5%, and 10% levels, respectively. For directional hypotheses (+/-), we use one-
tailed tests. If the coefficient’ s sign is opposite from the one hypothesized, then significance levels according to two-tailed tests are reported. The
composition of the final sample is presented in Appendix B. Variable definitions are presented in Appendix C.

Panel A: First-Stage Regression (z_score and return_volatility are the dependent variables. Results for instrumental variables are reported in bold. For
the sake of brevity, results for control variables are excluded)

(1) (2)
Total Operational Risk Voluntary Operational Risk
Variable Disclosure Quality Score Disclosure Quality Score
Exp. Exp. Exp. Exp.
z_score return_volatility z_score return_volatility
Sign Sign Sign Sign
cost_income - -0.0311 + 0.0109 - -0.0311 + 0.0108
(-3.11)*** (1.30)* (-3.10)*** (1.28)
cash_deposits ? -0.0030 ? 0.0016 ? -0.0032 ? 0.0017
(-1.09) (0.45) (-1.13) (0.45)
gdp_capita + -0.0501 - 0.2746 + -0.0421 - 0.1405
(-0.04) (0.34) (-0.03) (0.17)
inflation ? 0.1702 ? -0.2536 ? 0.1699 ? -0.2512
(2.66)*** (-3.17)*** (2.61)*** (-3.13)***
political_stability + -1.0381 - 2.4534 + -1.1214 - 2.4346
(-0.69) (1.97)** (-0.71) (1.86)*
activity_restrictions + 0.4474 - -0.2095 + 0.4882 - -0.1940
(1.54)* (-0.85) (1.53)* (-0.70)
entry_requirements - -0.2069 + 0.3061 - -0.2146 + 0.4011
(-0.48) (0.79) (-0.46) (0.92)
supervisor_pwrind + 0.0458 - 0.0406 + 0.0385 - 0.0303
(0.31) (0.31) (0.24) (0.21)
brd_outside ? 0.0020 ? 0.0001 ? 0.0008 ? -0.0028
(0.17) (0.01) (0.06) (-0.21)
largest_own ? 0.0802 ? 0.0039 ? 0.0858 ? 0.0028
(1.50) (0.12) (1.48) (0.08)
government_own ? 1.0927 ? -2.3610 ? 1.1790 ? -2.5276
(0.85) (-2.41)** (0.92) (-2.57)**
executive_own ? -0.3084 ? 2.7996 ? -0.0136 ? 3.6563
(-0.30) (1.34) (-0.01) (1.46)
ac_meet ? 0.0124 ? -0.0411 ? 0.0127 ? -0.0420
(0.25) (-0.99) (0.25) (-0.98)
activity_restrictions * largest_own ? -0.0082 ? 0.0020 ? -0.0095 ? 0.0019
(-0.97) (0.25) (-0.99) (0.22)
activity_restrictions * government_own ? 0.0473 ? 0.1200 ? 0.0491 ? 0.1393
(0.29) (1.02) (0.31) (1.14)
activity_restrictions * executive_own ? 0.0053 ? -0.0068 ? 0.0054 ? -0.0075
(0.18) (-0.23) (0.16) (-0.21)
entry_requirements * largest_own ? 0.0095 ? -0.0110 ? 0.0105 ? -0.0117
(0.50) (-0.67) (0.48) (-0.62)
entry_requirements * government_own ? 0.0565 ? 1.2754 ? 0.0801 ? 1.2706
(0.13) (2.75)*** (0.18) (2.61)***
entry_requirements * executive_own ? 0.0042 ? -0.5206 ? -0.0499 ? -0.6827
(0.03) (-1.31) (-0.27) (-1.45)
supervisor_pwrind * largest_own ? -0.0044 ? 0.0013 ? -0.0042 ? 0.0017
(-0.70) (0.44) (-0.62) (0.54)
supervisor_pwrind * government_own ? -0.1767 ? -0.3748 ? -0.1967 ? -0.3732
(-1.10) (-2.46)** (-1.23) (-2.45)**
supervisor_pwrind * executive_own ? 0.0202 ? -0.0848 ? 0.0102 ? -0.1115
(0.35) (-1.18) (0.16) (-1.26)
constant 2.9870 -3.5593 3.3744 -2.4498
(0.29) (-0.42) (0.31) (-0.27)
F-test of excluded instruments 4.44 4.18 4.29 4.11
(Prob > F) (0.0012)*** (0.0019)*** (0.0016)*** (0.0022)***
Partial R2 of excluded instruments 0.1290 0.1138 0.1289 0.1113
Endogeneity (Stock-Wright) test 6.15 4.98
(p-value) (0.2921) (0.4184)
Overidentification (Sargan) test 0.046 2.897
(p-value) (0.9974) (0.4078)

38
Table 4: Determinants of Operational Risk Disclosure Quality in European Banks
(Interaction Regressions) (cont.)

Panel B: Second-Stage Regression (z_score and return_volatility are the endogenous variables. For the sake of
brevity, results for control variables are excluded)

(1) (2)
Exp. Total Operational Risk Voluntary Operational Risk
Variable
Sign Disclosure Quality Score Disclosure Quality Score
ord_score_total ord_score_voluntary
z_score - -0.5955 -0.3971
(-1.93)** (-1.49)*
return_volatility + -0.5258 -0.2729
(-1.20) (-0.85)
activity_restrictions - -0.0695 -0.0641
(-0.18) (-0.17)
entry_requirements - -1.0776 -0.8261
(-1.29)* (-0.97)
supervisor_pwrind + -0.1214 -0.2455
(-0.36) (-0.73)
brd_outside + 0.0674 0.0525
(2.81)*** (2.24)**
largest_own + 0.1729 0.1515
(1.56)* (1.02)
government_own + 1.2371 -1.3934
(0.42) (-0.55)
executive_own - -2.0929 -2.5042
(-1.26) (-1.67)**
ac_meet + 0.2590 0.2659
(4.50)*** (5.16)***
activity_restrictions * largest_own + 0.0039 0.0037
(0.42) (0.38)
activity_restrictions * government_own + -0.3344 -0.0651
(-1.00) (-0.22)
activity_restrictions * executive_own - 0.0167 0.0076
(0.28) (0.14)
entry_requirements * largest_own + -0.0065 -0.0058
(-0.29) (-0.24)
entry_requirements * government_own + 0.9410 1.2712
(0.91) (1.18)
entry_requirements * executive_own - 0.0583 0.0865
(0.12) (0.18)
supervisor_pwrind * largest_own - -0.0114 -0.0087
(-1.68)** (-1.45)*
supervisor_pwrind * government_own - -0.2375 -0.2866
(-0.59) (-0.85)
supervisor_pwrind * executive_own + 0.1905 0.1952
(1.71)** (1.86)**
constant -2.8952 -4.5094
(-0.42) (-0.65)
Year Dummies Included/Non-significant Included/Non-significant
Type Dummies Included/Non-significant Included/Non-significant
Num. Obs. 243 243
F-Statistic 64.88 35.34
(Prob > F) (0.0000)*** (0.0000)***
R2 0.7905 0.7129

39
Appendix A: Operational Risk Disclosure Quality Index

Scoring Criteria:
• The original score for each disclosure item is 0.
• One additional point is given for each disclosure sub-item.
• One additional point is given for graphs or tables only if they provide information not included in textual disclosures.

This table presents the composition of our operational risk disclosure quality index.
Disclosure Item Disclosure Sub-items Reference(s)

1. Amount of Regulatory Capital for Operational Risk (Pillar 1 Capital)
1.1 The amount of regulatory capital for operational risk (mandatory). EP, 2006, annex xii, part 2,
1.2 The categorization of regulatory capital for operational risk by division, business line, subsidiary or point 4, p. 187
country.
1.3 The reasons for the change in regulatory capital for operational risk from previous years or quarters.
1.4 Graphical illustration or tabular presentation.

2. Measurement Approach of Regulatory Capital for Operational Risk
2.1 Definition and qualitative explanation of the measurement approach used to quantify the regulatory capital EP, 2006, annex xii, part 2,
for operational risk (mandatory). point 4, p. 187
2.2 Quantitative explanation of the measurement approach.
2.3 Prior or subsequent change in the measurement approach.
2.4 Graphical illustration or tabular presentation.

3. Strategies and Processes of Operational Risk Management 3.1 Qualitative Information (mandatory). EP, 2006, annex xii, part 2,
3.2 Quantitative information. point 1.a, p. 186
3.3 Forward-looking information.
3.4 Graphical illustration or tabular presentation.

4. Structure and Organization of the Operational Risk Management Function 4.1 Qualitative Information (mandatory). EP, 2006, annex xii, part 2,
4.2 Quantitative information. point 1.b, p. 186
4.3 Forward-looking information.
4.4 Graphical illustration or tabular presentation.

5. Scope and Nature of the Operational Risk Reporting System 5.1 Qualitative Information (mandatory). EP, 2006, annex xii, part 2,
5.2 Quantitative information. point 1.c, p. 186
5.3 Forward-looking information.
5.4 Graphical illustration or tabular presentation.

6. Operational Risk Transfer/Mitigation/Hedging Techniques 6.1 Qualitative Information (mandatory). EP, 2006, annex xii, part 2,
6.2 Quantitative information. point 1.d, p. 186
6.3 Forward-looking information.
6.4 Graphical illustration or tabular presentation.

7. Operational Value-at-Risk (VaR / Economic Capital / Pillar 2 Capital) 7.1 Qualitative Information. Ford et al. (2009)
7.2 Quantitative information.
7.3 Forward-looking information.
7.4 Graphical illustration or tabular presentation.

40
Appendix A: Operational Risk Disclosure Quality Index (cont.)

Scoring Criteria:
• The original score for each disclosure item is 0.
• One additional point is given for each disclosure sub-item.
• One additional point is given for graphs or tables only if they provide information not included in textual disclosures.

This table presents the composition of our operational risk disclosure quality index.
Disclosure Item Disclosure Sub-items Reference(s)

8. Internal Audit Function / Internal Control System
8.1 Qualitative Information on their roles in the operational risk management function and reporting system. Helbok and Wagner (2006)
8.2 Quantitative information on their roles in the operational risk management function and reporting system.
8.3 Forward-looking information on their roles in the operational risk management function and reporting
system.
8.4 Graphical illustration or tabular presentation.

9. Key Risk Indicators (KRIs) / Early Warning Systems (EWSs) 9.1 Qualitative Information. Ford et al. (2009)
9.2 Quantitative information.
9.3 Forward-looking information.
9.4 Graphical illustration or tabular presentation.

10.1 Qualitative Information.

10. Self-assessment Techniques 10.2 Quantitative information. Ford et al. (2009)
10.3 Forward-looking information.
10.4 Graphical illustration or tabular presentation.

11. Scorecard Models / Scenario Analyses / Stress Tests 11.1 Qualitative Information. Ford et al. (2009)
11.2 Quantitative information.
11.3 Forward-looking information.
11.4 Graphical illustration or tabular presentation.

12. Operational Risk Event Databases (Internal / External) 12.1 Qualitative Information. Helbok and Wagner
12.2 Quantitative information. (2006),
12.3 Forward-looking information. Ford et al. (2009)
12.4 Graphical illustration or tabular presentation.

13. Legal Risks 13.1 Qualitative Information. Helbok and Wagner (2006)
13.2 Quantitative information.
13.3 Forward-looking information.
13.4 Graphical illustration or tabular presentation.

14. Additional Information on Operational Risk Exposure and Management (e.g., 14.1 Qualitative Information.
cumulative amounts of historical operational losses classified by event types 14.2 Quantitative information.
and business lines; corrective actions subsequent to specific operational risk 14.3 Forward-looking information.
events) 14.4 Graphical illustration or tabular presentation.

41
Appendix B: Composition of the Final Sample by Country and Specialization

This table presents the composition of the final sample which consists of 85 banks with at least
€1,000 million in total assets from 20 member countries of the European Union.

Bank
Commercial Cooperative Investment Savings
Country Holding Total
Banks Banks Banks Banks
Companies

Austria 3 1 0 0 1 5
Belgium 0 0 0 0 2 2
Bulgaria 2 0 0 0 0 2
Cyprus 2 0 1 0 0 3
Denmark 7 0 0 0 1 8
Finland 1 0 0 0 1 2
France 3 1 0 0 0 4
Germany 5 0 1 0 1 7
Greece 5 0 0 1 0 6
Hungary 1 0 0 0 0 1
Ireland 2 0 0 0 0 2
Italy 6 6 2 0 0 14
Malta 0 0 1 0 0 1
Netherlands 1 0 0 0 2 3
Poland 1 0 0 1 0 2
Portugal 2 0 0 0 2 4
Spain 5 1 0 0 0 6
Sweden 2 0 0 1 1 4
UK 0 0 2 0 6 8
Lithuania 1 0 0 0 0 1
Total 49 9 7 3 17 85

42
Appendix C: Variable Definitions
This table presents the definitions and data sources of tested variables.
Variable Definition Data Source(s)

1. Dependent Variables:
ord_score_total The total operational risk disclosure quality score determined Annual Reports,
according to the disclosure index and scoring criteria explained Risk Reports
in Appendix A.
ord_score_voluntary The voluntary operational risk disclosure quality score Annual Reports,
determined according to the disclosure index and scoring criteria Risk Reports
explained in Appendix A. Additionally, we subtract points
scored for mandatory disclosures required by the Capital
Requirements Directive (EP, 2006).
2. Explanatory Variables:
a) Country Level Bank Regulation and Supervision Variables
activity_restrictions An index of regulatory restrictions on the activities of banks. It Barth et al. (2001,
measures regulatory impediments to banks engaging in 2004, 2006, 2008),
securities market activities (e.g., underwriting, brokering, World Bank (2008)
dealing, and all aspects of the mutual fund industry), insurance
activities (e.g., insurance underwriting and selling), real estate
activities (e.g., real estate investment, development, and
management), and the ownership of nonfinancial firms. It is
measured at the country level. It ranges from 4 to 16. Higher
values indicate greater restrictiveness.
entry_requirements An index of the entry requirements to the banking industry. It Barth et al. (2001,
measures the various types of legal submissions required to 2004, 2006, 2008),
obtain a banking license. It is measured at the country level. It World Bank (2008)
ranges from 0 to 4. Higher values indicate more stringent
banking entry policy.
supervisor_pwrind An index of the official powers and independence of the bank Barth et al. (2001,
supervisor. It measures the degree to which the supervisor has 2004, 2006, 2008),
the authority to take specific actions to prevent and correct World Bank (2008)
problems and also the independence of the supervisor from the
political pressures and the banking industry. It is measured at the
country level. Higher values indicate more powerful and
independent banking supervisory authorities.
b) Bank Level Governance Variables
brd_outside Proportion of outside board directors (percentage). Annual Reports,
Corporate
Governance Reports
largest_own Proportion of voting rights held by the largest non-managing, Annual Reports,
non-governmental shareholder (percentage). Corporate
Governance Reports,
Bankscope
government_own 1 if the domestic government holds at least 5% of voting rights; Annual Reports,
0 otherwise. Corporate
Governance Reports,
Datastream

43
Appendix C: Variable Definitions (cont.)

Variable Definition Data Source(s)

2. Explanatory Variables: (cont.)

b) Bank Level Governance Variables
executive_own Proportion of voting rights held by bank executives Annual Reports,
(percentage). Corporate
Governance Reports

ac_meet Number of audit committee meetings held annually. Annual Reports,

Corporate
Governance Reports
3. Endogenous Variables:
z_score (Capital-assets ratio + Return on assets) / Standard deviation of Bankscope
return on assets in the last five years (percentage). Return on
assets is computed using the net operating income after adding
back the loan loss provision. It is transformed using its natural
logarithm.
return_volatility Annualized standard deviation of daily stock returns. Datastream
4. Instrumental Variables:
gdp_capita The natural logarithm of gross domestic product (GDP) per World Bank’ s World
capita (in €). It is measured annually at the country level. Development
Indicators Database
inflation The percentage change in the cost to the average consumer of World Bank’ s World
acquiring a basket of goods and services that may be fixed or Development
changed at specified intervals. It is measured annually at the Indicators Database
country level.
political_stability An index that reflects perceptions of the likelihood that the World Bank’ s
government will be destabilized or overthrown by Worldwide
unconstitutional or violent means, including politically- Governance
motivated violence and terrorism. It ranges from -2.50 (weak Indicators Database
governance) to 2.50 (strong governance). It is measured
annually at the country level.
cost_income Cost-income ratio = Overhead costs / Total revenues Bankscope
(percentage).
cash_deposits Cash-deposits ratio = (Cash and equivalents + Short-term Bankscope
investments) / Total deposits (percentage).
5. Control Variables:
total_assets The natural logarithm of total assets (in € million). Bankscope
cross_listing 1 if the bank is registered and reporting with the U.S. Securities SEC filings
and Exchange Commission (SEC); 0 otherwise.
ama_dum 1 if the bank is using the Advanced Measurement Approaches Annual Reports,
(AMA) to quantify regulatory capital for operational risk; 0 Risk Reports
otherwise.

44
Appendix C: Variable Definitions (cont.)

Variable Definition Data Source(s)

5. Control Variables: (cont.)

antidirector Aggregate index of shareholder (anti-director) rights. The index Djankov et al. (2008)
is formed by summing: (1) vote by mail; (2) shares not
deposited; (3) cumulative voting; (4) oppressed minority; (5)
pre-emptive rights; and (6) capital to call a meeting. The source
of the index is the commercial laws of the various countries.
Higher values indicate stronger shareholder (anti-director)
rights. It is measured at the country level.
ruleoflaw An index that reflects perceptions of the extent to which agents World Bank’ s
have confidence in and abide by the rules of society, and in Worldwide
particular the quality of contract enforcement, property rights, Governance Indicators
the police, and the courts, as well as the likelihood of crime and Database
violence. It ranges from -2.50 (weak governance) to 2.50 (strong
governance). It is measured annually at the country level.
concentration Total assets of the five largest banks / Total assets of the European Central
national banking sector (percentage). It is measured annually at Bank
the country level.
Year dummies Dummies for the fiscal years 2009 (year_2009) and 2010 Bankscope
(year_2010). The fiscal year 2008 is used as the reference
category.
Type dummies Dummies for the business models of banks: commercial for Bankscope
commercial banks, cooperative for cooperative banks,
investment for investment banks and savings for savings banks.
The category: bank holding and holding companies is used as
the reference category.

45
Appendix D: Data on Bank Regulation and Supervision in the European Union (2008-2010)

Country activity_restrictions entry_requirements supervisor_pwrind

Austria 7 4 10
Belgium 7 4 10.50
Bulgaria 10 4 12
Cyprus 11 2 14
Denmark 9 4 9.50
Finland 9 3 8.50
France 9 3 7
Germany 7 2 6.50
Greece 8 3 9
Hungary 11 4 15
Ireland 7 4 12
Italy 12 4 7
Malta 10 4 14
Netherlands 6 3 7
Poland 8 4 8.50
Portugal 12 3 14.50
Spain 7 3 11
Sweden 10 2 5
UK 4 4 9
Lithuania 11 4 14.50

Source: Barth et al. (2001, 2004, 2006, 2008); World Bank (2008)

46