The Effect of Risk Type on ERM Effectiveness and Bank
Performance: An Empirical Analysis Of European Banks
Sari Minjari Damayanti
Faculty of Economics and Business, University of Indonesia, Jl. Prof. Dr. Sumitro Djojohadikusumo, Depok, Indonesia
Keywords: Banking Sector, Contingency Theory, COSO Framework, Corporate Governance, Credit Risk, Enterprise
Risk Management, Type of Risks.
Abstract: Ths study aims to provide empirical evidence on the significance of risk types on the effectiveness of ERM
and to the relationship between effective risk management and entities’ performance by using logistic and
residual analysis in European banking sectors in the period 2013-2016. This study provides empirical
evidence on the significance of credit risk and off-balance sheet exposure on the effectiveness of enterprise
risk management. The significance of credit risk may arise due to the banks close supervision on their credit
risks by implementing processes to monitor key risks to ensure they stay within the approved risk appetite
and mitigating efforts. Additionally, the significance of off-balance-sheet items may due to the consideration
that off-balance-sheet risk is the integral part of banks’ risk profile that need to be assessed carefully. While
this study does not provide support to contingency theory proposed by Gordon et. al. (2009), it provides
support for Kaplan and Mikes (2014) conception that risk management will be most effective when it matches
the intrinsic nature and controllability of the different types of risk the organization faces. As this study only
focuses on the banking sector, some standard measurement as suggested by previous studies cannot be fully
measured. It is possible that these results may not be generalizable to a broader range of risk and risk
management research. This study provides the empirical evidence of the significance of different type of risks
on the effectiveness of ERM in the European banks.
1 INTRODUCTION
Enterprise risk management (ERM) have been listed
by Harvard Business Review as one of their
breakthrough ideas (Buchanan, 2004). Most
legislative bodies, professional associations, rating
agencies, regulators, and stock exchange hold the
view that an ERM is an important tool to manage all
the risks an organization faces and have actively
advised firms to adopt ERM. Nevertheless, the
financial crisis of 2008 has cast doubts upon the
efficacy of ERM. Serious failures across financial
institutions during financial crisis, many ERM best
practice firms faced bankruptcy in the 2008 financial
crisis (Bromiley et al., 2015), have been link to risk
management flaws and low transparency in managing
risks (Stulz, 2008). Power (2009) argues that the
benefits of ERM are limited to certain states of the
world and that ERM is not well equipped to address
the complex realities of interconnectedness. Further,
Fraser, Schoening-Thiessen, and Simkins (2008)
confirmed that many practitioners recognize the lack
of information on management of ERM.
Literature on risk management after the financial
crisis shows that organisations can improve their
performance by implementing an enterprise risk
management, a holistic approach to risk management
(Gordon, Loeb, & Tseng, 2009; McShane, Nair, &
Rustambekov, 2011; Bromiley, McShane, Nair, &
Rustambekov, 2015). The efficiency in managing
risks can be achieved through a deeper understanding
of risk management across the institutions. Gordon et
al. (2009) show that the relation between a firm’s
ERM and its performance is dependent on the proper
match between a firm’s ERM and the contextual
variables surrounding firms. However, they
acknowledged that there is a limitation in their study
that contingency variables selection is only based on
the way the authors’ interpretation of the extant
literature and there is no theoretical model that could
adequately explain which contingency variables
438
Damayanti, S.
The Effect of Risk Type on ERM Effectiveness and Bank Performance: An Empirical Analysis of European Banks.
In Proceedings of the Journal of Contemporary Accounting and Economics Symposium 2018 on Special Session for Indonesian Study (JCAE 2018) - Contemporary Accounting Studies in
Indonesia, pages 438-446
ISBN: 978-989-758-339-1
Copyright © 2018 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved

should be considered in ERM studies. Nevertheless,
this limitation also presents opportunities to find more
fitting contingency variables in ERM studies. A
recent article written by Kaplan and Mikes (2014)
concludes that the effective risk management depends
on the organization’s context and circumstances.
Further, they have proposed that risk management
will be most effective when it matches the inherent
nature and controllability of the different types of risk
the organization faces. The primary aim of this paper
is to provide empirical evidence for the proposal that
the different types of risk is one of the contingency
variables that influence the relationship between a
firm’s ERM and its performance. Additionally, this
paper assesses the different risk types in banking
sector and their significance on the effectiveness of
risk management implementation.
2 LITERATURE REVIEW
Many academic scholars, standards setting
organizations, industry publications, industry
associations, consulting firms, and rating agencies
has offered their ERM definitions and descriptions.
The most recognised definition has been proposed by
COSO framework (2004) which define ERM as a
process that is designed to identify potential events
and manage risk in relation to the achievement of
entity’s objective. It indicates that enterprise risk
management addresses internal control need and a
fuller risk management process. The framework focus
on strategy, which is the key for implementing the
right ERM direction. Gordon et al. (2009) show that
enterprise risk management assess risks that
encompasses all functions and levels in an
organization. Additionally, research by Kaplan and
Mikes (2014) show that enterprise risk management
is an important component of corporate governance
reforms in the entities.
Nevertheless, findings by Paape and Speklé
(2012) shows that ERM implementation is influenced
by the regulatory environment. However, they did not
find any support that application of the COSO
framework and mechanistic view on risk appetite and
tolerance improves risk management effectiveness. In
response, COSO (2016) provides clarification on a
few misconceptions about its original framework
since it was introduced in 2004 that may alter research
findings. Further, it offers a more concise definition
of enterprise risk management as the culture,
capabilities, and practices, integrated with strategy
and execution, that organizations rely on to manage
risk in creating, preserving, and realizing value.
2.1 Theoretical Paradigm
Gordon, Loeb, & Tseng (2009) show that risk
management effectiveness and performance relation
is dependent upon the proper match between a firm’s
enterprise risk management (ERM) and its contextual
variables. Additionally, Kaplan and Mikes (2014)
indicates that the effective risk management depends
on the organization’s context and circumstances.
Further, they have indicated that risk management
will be most effective when it matches the inherent
nature and controllability of the different types of risk
the organization faces.
The banking sector is chosen because signalling
theory suggests that firms within the same sector try
to adopt the same level of disclosure to keep pace with
their peers and to avoid being perceived as firms that
hiding bad news (Craven & Marston, 1999).
Additionally, the firms may use internet disclosure to
signal high effectiveness disclosures that provide
signal to investors that the firm is profitable and keep
up with the latest technology (Oyelere et al., 2003).
Further, the banking sector has its own unique
characteristics and always attempted to diversify its
risk to prevent unexpected default from sinking the
entire bank.
2.2 Regulatory Context
This study focus on in the banking sector. The
banking industry is a heavily regulated industry.
Harnay and Scialom (2016) stated that there is a
paradigmatic change in the conception of regulatory
instruments of banking authorities, in which the
regulations have shifted from public interest theory
regulation to private interest theory regulation for the
substitution of micro-prudential for macro-prudential
regulations. They show that micro- prudential
regulations have failed to take the global features and
caused the 2007-2008 financial crisis. Kaminski and
Robu (2016) said that bank managers are often left to
their own ways to figure out what specific controls
are required to address regulatory requirements which
lead to uncertain effectiveness in control activities.
Further, they stated that tighter compliance
regulations have challenged financial institutions in a
variety of ways. In spite of that, those who adapt best
may enjoy a distinct competitive advantage and make
them more robust and sustainable over time.
Banking industry need more practical guidance
that could provide structural answers in detail
manner. Basel Committee on Banking Supervision
(BCBS) has developed Basel III which is aim to
strengthens micro-prudential regulation and
The Effect of Risk Type on ERM Effectiveness and Bank Performance: An Empirical Analysis of European Banks
439

supervision, which raise resilience of individual
banking institutions to periods of stress, and adds a
macro-prudential overlay, which target system wide
risks and pro-cyclical amplification overtime (BCBS,
2011). Basel III has three pillars that includes capital
buffers, risk coverage, containing leverage, risk
management supervision, and market discipline. In
relation to risk management, Basel III address firm-
wide risk management by capturing risk of off-
balance sheet exposures, managing risk
concentration, strengthening counterparty credit risk
framework by risk coverage, and comprising
common equity of 2.5 percent of risk-weighted assets.
By meeting the Basel III requirements, individual
bank can have greater resilience in the period of stress
and global financial institutions can reduce the risk of
system wide shocks.
2.3 Hypotheses Development
This study follows Gordon et al. (2009), a firm’s
choice of ERM system should be properly matched
with several key firm-related factors that includes one
additional factor, different risk types, proposed by
Kaplan and Mikes (2014). Thus, the relation between
a firm’s ERM and its performance is contingent on
the proper match between a firm’s ERM and the
following six firm-related variables: environmental
uncertainty, industry competition, firm size, firm
complexity, board of directors’ monitoring, and risk
types.
This study set out to offer a new model for
understanding the relationship between different risk
types disclosure on ERM implementation
effectiveness and its effect on the fitting level of
contingency variables in ERM studies. Thus, the first
hypothesis is formulated as follows:
H
1
: There is a positive association between the
different types of risk disclosure and ERM
implementation effectiveness.
Healy and Palepu (2001) shows that disclosure is
an important means for management to communicate
firm performance and governance to outside
investors. Previous studies of risk management
provide mixed evidence on the relationship between
ERM effectiveness and market performance. Banks
and insurers with a strong and independent risk
management function have better performance and
reduce risk exposure (Ellul & Yeramilli, 2013;
McShane, Nair, & Rustambekov, 2011). However,
research by Baxter and Vermeulen (2013) shows that
there is no relationship between ERM effectiveness
and market performance in banking and insurance
sector.
This study examines the relationship between the
determinants and effectiveness of ERM systems, and
the consequences of ERM systems effectiveness on
financial and market performance of the entities. The
different risk type disclosure in this model is
represented by the variable ERWA, which is the
incorporation of risk types and risk level of European
banks. Thus, this research also seeks to address the
following hypothesis.
H
2
. There is a negative association between the
absolute value of the residuals and performance.
3 RESEARCH METHODOLOGY
3.1 Sample and Data
The risk types data are collected from Orbis Bank
Focus database which consist of world banking
information source from banks in 28 European Union
countries in the year of 2013-2015. After excluding
companies with missing data, preliminary sample
with complete risk types data consists of 14 variables
and 125 observations.
To test for association between risk types and
ERM effectiveness, I gather different risk types data
(market, credit, operational, counterparty, and off-
balance-sheet risk) of European banks annually from
Orbis. ERM Advanced data is manually collective
from selected European banks’ annual reports are
publicly available on their websites and Federal
Deposit Insurance Corporation (FDIC) websites. The
financial information and corporate governance data
are collected from DataStream, Reuters, and
Financial Times websites for the year 2013-2016.
3.2 Research Method
The first model is regressed using a logistic
regression to predict association between risk types,
other variables proposed by Gordon, Loeb, & Tseng
(2009) and ERM effectiveness. Logistic Regression
Models relationship between set of variables or
covariates x
i
. The advantages of the logit are simple
transformation of P(y|x), linear relationship with x,
can be continuous (Logit between -∞ to +∞), and
known binomial distribution (P between 0 and 1). A
logistic regression was chosen since the dependent
variable of ERM effectiveness is a binary dependent
variable (Wooldridge J., 2012). A binary variable
takes on only two values, zero and one. The binary
variable in this model is ERMadvanced that takes
value of 1 if ERM score is equal to or higher than 4,
and 0 if otherwise.
JCAE Symposium 2018 – Journal of Contemporary Accounting and Economics Symposium 2018 on Special Session for Indonesian Study
440

The second model is regressed using ordinary
least squares that regressed the absolute residual
value of the first model with the dependent variable
of bank’s performance, return on average assets and
Tobin Q. Both models are regressed using STATA 14
software that supports many aspects of logistic
regression.
3.3 Research Models and Variables
The first hypothesis is tested using the model in Eq.
(1). The coefficients in Eq. (1) describes the proposed
best practice match between ERM and the bank-
related factors (variables) discussed above:

,
=

+


,
+


,
+


,
+



,
+


,
+


,
+



,
+


,
+


,
+
,
(1)
where,
ERMA ERM advanced is a dummy variable equal to
1 if ERM score is equal to or higher than 4,
and 0 otherwise, which is a measurement by
Florio and Leoni (2017), ERMA is a
comprehensive measure for ERM
implementation effectiveness, whereas ERM
score is the sum of the following variables,
chief risk officer, risk committee, risk
committee to board of directors, risk
assessment frequency, risk assessment level,
risk assessment method in bank  at year .
MR Market risk is the risk of losses in the bank's
trading book due to changes in equity prices,
interest rates, credit spreads, foreign-
exchange rates, commodity prices, and other
indicators whose values are set in a public
market, in bank  at year , in millions of
dollars.
CR Credit risk is the potential risk that a bank
borrower or counterparty will fail to meet its
obligations in accordance with agreed terms,
in
b
ank  at year , in millions of dollars.
OR Operational risk is the risk of loss resulting
from inadequate or failed internal processes,
people and systems or from external events, in
b
ank  at year , in millions of dollars.
OBS Off-balance-sheet items are assets or
liabilities that exist, but are not required by
IFRS to be included on financial statements,
in
b
ank  at year , in millions of dollars.
EU Environmental uncertainty that represent the
difficulties for organizations due to the
increasing unpredictability of the future
events affecting the organization, in bank  at
year .
CI Industry competition that represent an inter-
industry variable represent a possible
competitive pressure banks face from other
sectors which is represented by the
measurement of market capital to GDP, in
b
ank  at year .
BS Bank size is the bank’s average total assets, in
b
ank  at year .
BC Bank complexity that captured scope and
diversity in business lines of the subsidiaries
of an organization, in
b
ank  at year .
MBD Monitoring by firm’s board of directors which
represents the number of directors for each
firm divided by the natural logarithm of total
assets, in
b
ank  at year .
β

various model parameters, i = 0 to 5
ε
residual or error term.
Variables market risk (MR), credit risk (CR),
operational risk (OR), and off-balance-sheet items
(OBS), represent different risk types in European
banks, proposed by Mike & Kaplan (2013). The other
independent variables are contingency variables
proposed by Gordon et al. (2009).
The second hypothesis is tested using the model
in Eq. (2). Eq. (2) is a residual analysis model. The
basis for using a residual analysis is a better test of the
holistic relation concerning the way contingency
factors interact with ERM in affecting bank
performance (Gordon, Loeb, & Tseng, 2009). The
Eq. (2) is written below and regressed by an OLS
regression.

,
=

+


,
+
,
(2)
where,
P Firm performance, measured by accounting
measures, ROAA, market measures, Tobin
Q, in bank  at year +1.
ARES absolute value of residuals from Equation
1) that represent “lack of fit”, in bank  at
year .
β

various model parameters, i = 0 to 5
ε
residual or error term.
In order to see whether the above argument is
right and whether ARES, the absolute value of
residuals in Eq. (1), is related to performance. The
ARES coefficient should show a significant negative
association with banks’ performance in Eq. (2). The
derived coefficients are based on ‘‘minimizing” the
sum of the squared deviations of the residual. The
negative significance of ARES coefficient in Eq. (2)
is critical in assessing the ‘‘lack of fit” in the match
between an ERM system and the sixth contingency
variables.
The Effect of Risk Type on ERM Effectiveness and Bank Performance: An Empirical Analysis of European Banks
441

4 RESULTS
4.1 Main Results
4.1.1 Descriptive Statistics
The table 1 presents descriptive statistics for the
dependent and independent variables. Sampled
companies present a high operating profitability on
average, as mean ROA is equal to 61.4%. Mean
Tobin's Q ratio (Q) is 0.11, signalling means that the
cost to replace a firm's assets is greater than the value
of its stock, which implies that the stock is
undervalued. Meanwhile, 87.2% of the sample shows
an advanced ERM system, having 4 or more ERM
components. The average credit risk is 107,063
million dollars, the highest among all risks measured.
Table 1: Descriptive Statistics.
Variable Mean Std. Dev Min Max
ROAA 0.614 1.719 -3.420 16.340
TQ 0.110 0.351 0.000 3.930
ERMA 0.872 0.335 0.000 1.000
CR 107063.000 173515.600 49.000 955300.000
MR 9474.480 19544.120 2.000 119731.000
OR 15656.230 27218.830 2.000 119200.000
OBS 81552.780 148014.600 2.000 752775.000
EU 5.158 1.136 2.187 7.016
CI 0.024 0.049 0.000 0.426
BS 10.993 2.586 3.995 14.798
BC 612.840 1185.485 2.000 6024.000
MBD 2.646 0.887 0.760 5.074
N 125
4.1.2 Logistic regression result
The table 2 below is the logistic regression outcome.
I use Stata’s predict to obtain the predicted
probabilities of the outcome, the value of the logit
index, and the standard error of the logit index.
From 125 observations, the model likelihood is -
31.7, where null model has a lower value (more
negative). The LR Chi
2
(9) indicates G-square for 9
degrees of freedom. The Prob > chi
2
or p-value of the
first model is 0.0002. The p < 0.05 indicates a
significantly better model. In other words, the
Prob > chi
2
= 0.0002 show that the model as a whole
is statistically significant (p < 0.0001). The Pseudo R
2
of 0.336 indicates that model explain 33.6% of
variation in the effectiveness of ERM. In other words,
the McFadden Pseudo R
2
= 0.336 indicates that an
approximate amount of variability explained by the
fitted model is 33.6 percent.
The odds ratio of credit risk is 0.99 (p = 0.022). It
indicates that the odds of having an effective ERM
are increased by a factor of 0.99 for having credit risk
rather than not having credit risk, controlling for other
variables in the model. In other words, each one-unit
increase in credit risk variable increases the odd ratio
of ERM advanced by 0.99, when the other
independent variables are held constant, and this
effect is statistically significant.
The odds ratio of off balance sheet items is 1.00
(p=0.09). It indicates that the odds of having an
advanced ERM are increased by a factor of 1.00 for
having off balance sheet item rather than not having
off balance sheet item, controlling for other variables
in the model. It means that each one-unit increase in
off balance sheet items variable increases the odd
ratio of ERM advanced by 1.00, when the other
independent variables are held constant, and this
effect is statistically significant.
The odds ratio of competition in inter-industry is
1.6e+201 (p = 0.024) show that the odds of having an
advanced ERM are increased by a factor of 1.6e+201
for having competition in inter-industry rather than
not having competition in inter-industry, controlling
for other variables in the model. It means that each
one-unit increase in operational risk variable
increases the odd ratio of ERM advanced by
1.6*10
201
, when the other independent variables are
held constant, and this effect is statistically
significant.
Table 2: Logistic regression outcome.
(1)
ERMA
(
odd rat.
)
ERMA
CR 0.999*
(
-2.29
)
MR 0.999
(
-0.05
)
OR 1.000
(
0.84
)
OBS 1.00**
(
1.69
)
EUI 1.030
(
0.06
)
CI 1.6e+201*
(
2.26
)
BS 1.518
(
1.43
)
BC 0.998
(-1.56)
MBD 1.828
(1.11)
_cons 0.010
(-1.74)
N 125
t statistics in parentheses
* p<0.05, ** p<0.01, *** p<0.001
JCAE Symposium 2018 – Journal of Contemporary Accounting and Economics Symposium 2018 on Special Session for Indonesian Study
442

4.1.3 Residual analysis
The residual analysis outcome is shown in table 3
below. The residuals () is derived from equation (1),
where =−

using a postestimation command in
STATA. Variable () is the absolute value of the
residual () is obtained using an absolute syntax in
STATA. The number of observations in table (3) is
reduced to 117 observations, due to eight missing
values generated in predicting () and generating
().
The coefficients of ARES (0.579) for ROAA and
(0.117) for Tobin Q are positive and significant (at the
level of 0.05). In other words, ARES is positively
associated with firm performance. These results are
contrary to the expected negative sign from Gordon
et al., (2009). These results do not support the main
argument that the proper match between ERM and the
contingency variables is an important driver of firm
performance. The different result may due to several
factors. First, the different ERM effectiveness and
entities performance measurements used, I used
Florio and Leoni (2017) measurement instead of
Gordon et. Al. (2009) and different sectors observed,
instead of multi-sectors observation, this paper only
focus on single sector, banking. Secondly, the lack of
variable control for different set of rules that varies
across countries in Europe. The lack of countries’
controls due to rigid application of Gordon et.al.
model that contain only the contingency variables.
Table 3: Residual analysis outcome.
(1)
ROAA
(2)
Tobin Q
abs res 0.579*
(2.24)
0.117*
(2.21)
_cons 0.355
(1.78)
0.0573
(1.41)
N 117 117
t statistics in parentheses
* p<0.05, ** p<0.01, *** p<0.001
4.2 Post Estimation Results
The classification model test yields predicted p>.5 for
118 banks, 106 banks actually have an effective
ERM. Overall 88% of banks are correctly classified.
Out of all banks observation that have an effective
ERM 97.25% were correctly predicted to have an
effective ERM. Out of all banks observation that do
not have an effective ERM 25% were correctly
predicted.
The regression collinearity diagnostic procedures
(coldiag) were also performed in STATA follow
Belsley, Kuh, & Welsch (2005) that examine the
"conditioning" of the matrix of independent variables.
Coldiag computes the condition number of the
matrix. If this number is "large", Belsley et. al. (2005)
suggest 30 or higher, then there may be collinearity
problems. The condition number is the largest
singular value. All "large" singular may be worth
investigating. The all condition numbers (singular
values) are below 30 which most numbers are
relatively smaller than 30, that indicates that there
may not be collinearity problems in this model.
Model specification is tested by LR Diagnostics
using linktest in STATA. The insignificant _hatsq (p
= 0.858) indicates the link function is correctly
specified. In other words, it indicates that there is no
specification error. Additionally, insignificant _hatsq
means that there are no omitted relevant variables.
Moreover, it also indicates that the link function is
correctly specified.
A goodness of fit test shows how well the data fits
the model. Specifically, the Hosmer-Lemeshow test
(HL test) calculates if the observed event rates match
the expected event rates in population subgroups
(Hosmer, Lemeshow, & Sturdivant, 2013). The HL
test is a goodness of fit test for logistic regression,
especially for risk prediction models. The output
returns a chi-square value (a Hosmer-Lemeshow chi-
squared) and a p-value (e.g. Pr > Chi
2
) are the main
concerns in this test. Small p-values (usually under
5%) mean that the model is a poor fit. The large
insignificant p-value (0.8349), suggests that the
model fits the data reasonably well.
The AIC in this result show a smaller value of
0.66, that indicates the better fit of the model.
Meanwhile, the current model is preferred when BIC
is negative. The more negative the BIC, the better the
fit. The BIC in this result is large negative that show
a better fit of the model.
The marginal effect outcome for the first model
indicates the following, one-unit increase in credit
risk from the baseline mark of 107063 increases the
probability of ERMA improvement by -1.44e-09,
one-unit increase in the off balance sheet items from
the baseline (81552.8) increases the probability of
ERMA improvement by 8.08e-10, and one-unit
increase in the competition in inter-industry from the
baseline (0.023) increases the probability of ERMA
improvement by 0.012 or 1.2 percent. (Basel
Committee on Banking Supervision (BCBS), 1986).
The robust estimate of variance of the first model
estimates the standard errors that are robust to the fact
that the error term is not identically distributed. The
standard errors in the robust regression can be used to
make valid statistical inference on the coefficients,
The Effect of Risk Type on ERM Effectiveness and Bank Performance: An Empirical Analysis of European Banks
443

even though the data are not identically distributed.
The model likelihood and Pseudo R
2
has the same
value with the standard logistic regression, -31.7 and
0.336 respectively. The Prob > chi
2
or p-value of the
robust model is slightly higher, 0.004, nevertheless, it
still below p < 0.05 that indicates a significant model.
The odds ratio of credit risk, off balance sheet items,
and competition in inter-industry have the same value
with the standard logistic regression. However, the
robust regression has the lower p-value that indicates
the higher odds of having an effective ERM are
increased by a factor of 0.99, 1.00, and 1.6e+201 for
having credit risk, off balance sheet items, and
competition in inter-industry rather than not having
those variables, when the other independent variables
are held constant, and this effect is statistically
significant.
5 CONCLUSIONS
5.1 Key Findings
This study provides empirical evidence on the
significance of credit risk, off-balance sheet exposure
on the relationship between effective risk
management using measurement by Florio and Leoni
(2017) and entities’ performance. The significance of
credit risk on enterprise risk management (ERM) may
due to that ERM takes a broader view of risk that
identify risks that could impact the institution’s
ability to achieve their goals. It implements processes
to monitor key risks to ensure they stay within the
approved risk appetite. Further, it seeks to identify all
aspects of credit risk that might be present throughout
the institution, regardless of where the risk occurs.
The credit risk needs to be identified, aggregated, and
managed that contributes to the effectiveness of ERM
(Hoover, 2016). Meanwhile, the significance of off
balance sheet items can be explained by the paper
conclusion of Basel Committee on Banking
Supervision (1986) that stated the individual types of
risk associated with most off-balance-sheet business
are in principle no different from those associated
with on-balance-sheet business. The off-balance-
sheet risks should not be analysed separately from the
risks arising from on-balance-sheet business, but
should be regarded as an integral part of banks’
overall risk profiles. Thus, the existence of off-
balance-sheet risks contributes to the effectiveness of
ERM, since it considered as integral part of banks’
risk profile that need to be assessed carefully.
Additionally, it also provides support for Kaplan
and Mikes (2014) findings that the effective risk
management depends on the organization’s context
and circumstances. However, this study does not
provide support to contingency theory proposed by
Gordon et. al. (2009).
5.2 Limitations and Further Research
There are limitations in this study. First, the study is
unable to encompass the different industries, since it
only focus on the banking sector. The measurement
as suggested by Gordon et al. (2009) cannot be fully
measure due to single industry study. It is possible
that these results may not be generalizable to a
broader range of risk and risk management study. In
other words, the generalisability of these findings is
limited to the banking sectors. Thus, the further
studies need to be carried out in a cross-industry study
involving different sectors to investigate the
association between risk types and ERM
effectiveness. A second limitation is that a theoretical
model with selected contingency variables is based on
subjective interpretation of the literature.
Schoonhoven (1981) suggests that the contingency
theory has several problems lack of clarity in its
theoretical statements to the embedding of
symmetrical and non-monotonic assumptions in the
theoretical arguments. Thus, it is recommended that
further research be undertaken from different
theoretical perspectives. A third limitation to this
study is that banks with complete risks data are more
likely to have more funding in risk management
which implies higher assets. It may hinder the
inclusion of banks sample with moderate or lower
assets. Further studies should assess different periods
beyond the year of 2018 to include more banks with
different assets range, as the upcoming 2018 Basel III
requirements will generate more complete risks data
in banking industry. The forth limitation is the ERM
components may not be a complete representation of
ERM effectiveness. Thus, more research is needed to
account for other potentials representation of ERM
effectiveness, e.g. risk committee experiences and
risk assessment complete disclosures. The last
limitation is the lack of variable control for different
set of rules that varies across countries in Europe.
Thus, the future research need to consider the
inclusion of controls for countries’ rules, such as a
rules index.
Hopefully, this study could offer some important
insights into the significance of credit risk, off-
balance sheet exposure, and competition in industry
in incentivizing banks to better manage their risk by
meeting the upcoming 2018 Basel III requirements
that are in line with the new 2016 COSO framework.
JCAE Symposium 2018 – Journal of Contemporary Accounting and Economics Symposium 2018 on Special Session for Indonesian Study
444

ACKNOWLEDGEMENTS
I would like to thank Professor Jacco L. Wielhouwer,
professor in economics of accounting and tax at the
VU University Amsterdam for his valuable
comments and suggestions to improve the quality of
the paper.
Additionally, I would like to convey my gratitude to
Professor Ferdinand Gul, Alfred Deakin Professor,
Faculty of Business and Law, BL Deakin Business
School, Melbourne, for his valuable insights during
JCAE 2018 Conference in Bali, Indonesia.
* Data not shown, available upon request.
REFERENCES
Basel Committee on Banking Supervision (BCBS). (1986).
Principles for the Management of Credit Risk.
Retrieved from Banks for International Settlements:
http://www.bis.org/publ/bcbs75.pdf
Basel Committee on Banking Supervision (BCBS). (2011,
June 29). Basel Committee on Banking Supervision
reforms - Basel III. Retrieved from Bank for
International Settlements:
http://www.bis.org/bcbs/basel3/b3summarytable.pdf
Basel Committee on Banking Supervision (BCBS). (2011).
Basel III: A global regulatory framework for more
resilient banks and banking systems. Basel: Bank for
International Settlements.
Baxter, D., & Vermeulen, P. (2013). Intelligent ERM:
Evolving risk management. Journal of Financial
Perspectives, 1(3), 19-27.
Beier, N., Harreis, H., Poppensieker, T., Sojka, D., &
Thaten, M. (2010, June). Getting to grips with
counterparty risk. McKinsey Working Papers on Risk,
20.
Belsley, D. A., Kuh, E., & Welsch, R. E. (2005). Regression
diagnostics: Identifying influential data and sources of
collinearity (Vol. 571). John Wiley & Sons.
Bikker, J. A., & Spierdijk, L. S. (2009). Measuring and
Explaining Competition in the Financial Sector.
Discussion Paper Series, 9(1), 36.
Bromiley, P., McShane, M., Nair, A., & Rustambekov, E.
(2015). Enterprise Risk Management: Review,
Critique, and Research Directions. Long Range
Planning, 48(4), 265-276.
Buchanan, L. (2004). Breakthrough ideas for 2004.
Harvard Business Review, 2(1).
Carmassi, J., & Herring, R. (2016). The Corporate
Complexity of Global Systemically Important Banks.
Journal of Financial Services Research, 49(2), 175-
201.
Cetorelli, N., & Goldberg, L. S. (2014, March). Measures
of Complexity of Global Banks. Economic Policy
Review, 20(2), 39.
Committee of Sponsoring Organizations of the Treadway
Commission (COSO). (2004, September). Enterprise
Risk Management - Integrated Framework. Retrieved
from COSO Enterprise Risk Management — Integrated
Framework Update:
https://www.coso.org/documents/Framework%20Refe
rence%20Secured.pdf
Committee of Sponsoring Organizations of the Treadway
Commission (COSO). (2016, June). Enterprise Risk
Management Aligning Risk with Strategy and
Performance. Retrieved from ERM Integrated
Framework Update:
https://www.coso.org/Documents/COSO-ERM-Exec-
Summary-draft-Post-Exposure-Version.pdf
Craven, B. M., & Marston, C. L. (1999). Financial reporting
on the Internet by leading UK companies. European
Accounting Review, 8(2), 321-333.
Cumming, C. M., & Hirtle, B. J. (2001). The Challenges of
Risk Management in Diversified Financial Companies.
Economic Policy Review Federal Reserve Bank of New
York, 7(1), 1-13.
Ellul, A., & Yerramilli, V. (2013). Stronger Risk Controls,
Lower Risk: Evidence from U.S. Bank Holding
Companies. The Journal of Finance, 68(5), 1757-1803.
Federal Reserve Board. (2006). Calculation of Risk-
Weighted Assets.
Florio, C., & Leoni, G. (2017, January). Enterprise risk
management and firm performance: The Italian case.
The British Accounting Review, 49(1), 56-74.
Fraser, J. R., Schoening-Thiessen, K., & Simkins, B. J.
(2008). Who Reads What Most Often? A Survey of
Enterprise Risk Management Literature Read by Risk
Executives. Journal of Applied Finance, 18(1), 73-91.
Gordon, L. A., Loeb, M. P., & Tseng, C. Y. (2009).
Enterprise Risk Management and Firm Performance: A
Contingency Perspective. Journal of Accounting and
Public Policy, 28(4), 301-327.
Harnay, S., & Scialom, L. (2016). The influence of the
economic approaches to regulation on banking
regulations: a short history of banking regulations.
Cambridge Journal of Economics, 40(2), 401-426.
Healy, P. M., & Palepu, K. G. (2001). Information
asymmetry, corporate disclosure, and the capital
markets: A review of the empirical disclosure literature
. Journal of Accounting and Economics, 31(1-3), 405-
440.
Hoover, T. (2016, April). Pulse. Retrieved from Linkedin:
https://www.linkedin.com/pulse/risk-management-vs-
enterprise-terry-hoover-cpa-cgma-cia-cisa
Hosmer, D. W., Lemeshow, S., & Sturdivant, R. (2013).
Applied logistic regression (3rd Edition ed., Vol. 398).
John Wiley & Sons.
Jorion, P., & Zhang, G. (2009). Credit Contagion from
Counterparty Risk. The Journal of Finance, 64(5),
2053-2087.
Kaminski, P., & Robu, K. (2016, January). A best-practice
model for bank compliance. Retrieved from Mckinsey
& Company: http://www.mckinsey.com/business-
functions/risk/our-insights/a-best-practice-model-for-
bank-compliance
The Effect of Risk Type on ERM Effectiveness and Bank Performance: An Empirical Analysis of European Banks
445

Kaplan, R. S., & Mikes, A. (2014). Towards a Contingency
Theory of Enterprise Risk Management. Harvard
Business Review.
Kwan, S., & Eisenbeis, R. A. (1997). Bank Risk,
Capitalization, and Operating Efficiency. Journal of
Financial Services Research, 12(2), 117-131.
McShane, M., Nair, A., & Rustambekov, E. (2011). Does
Enterprise Risk Management Increase Firm Value?
Journal of Accounting, Auditing & Finance, 26(4), 641-
658.
Mehta, A., Neukirchen, M., Pfetsch, S., & Poppensieker, T.
(2012, May). Managing market risk: Today and
tomorrow. McKinsey Working Papers on Risk.
Retrieved from Mckinsey & Company.
Oyelere, P., Laswad, F., & Fisher, R. (2003). Determinants
of internet financial reporting by New Zealand
companies. . Journal of International Financial
Management & Accounting, 14(1), 26-63.
Paape, L., & Speklé, R. F. (2012). The Adoption and Design
of Enterprise Risk Management Practices: An
Empirical Study. European Accounting Review, 21(3),
533-564.
Power, M. (2009). The Risk Management of Nothing.
Accounting, Organizations and Society, 34(6-7), 849-
855.
Stulz, R. M. (2008). Risk Management Failures: What Are
They and When Do They Happen? Journal of Applied
Corporate Finance, 20(4), 39-48.
Williams, R. (2017). Scalar Measures of Fit: Pseudo R2
and Information Measures (AIC & BIC). Retrieved
from University of Notredame:
https://www3.nd.edu/~rwilliam/stats3/L05.pdf
Wooldridge, J. (2012). Introductory Econometrics A
Modern Approach (5th Edition ed.). Boston: Cengage
Learning.
JCAE Symposium 2018 – Journal of Contemporary Accounting and Economics Symposium 2018 on Special Session for Indonesian Study
446