CC BY
9Original article
Operational bank risk Issues of Risk Analysis, Vol. 17, 2020, No. 2
JEL Classification codes: GZ10 https://doi.org/10.3Z686/181Z-5ZZ0-Z0Z0-17-Z-10Z-119
ISSN 181Z-5ZZ0
© Проблемы анализа риска, Z0Z0
Повышение эффективности управления операционными рисками в российских банках
Абу-Алроп Д. Х.,
Казанский федеральный университет,
420008, Россия, Республика Татарстан, г. Казань, ул. Кремлевская, д. 18
Аннотация
В настоящем исследовании рассматривается эффективность управления операционными рисками 85 российских коммерческих банков за период 2008—2017 гг. В этом исследовании используется ориентированная на ввод модель анализа оболочки данных (DEA) с финансовыми коэффициентами для оценки эффективности управления операционным риском. В исследовании используется базовый подход к измерению операционных рисков. Кроме того, в исследовании используется чистая процентная маржа (NIM), доходность активов (ROA) и доходность собственного капитала (ROE) для измерения эффективности банков. Исследование показало, что малые банки наиболее эффективны в управлении операционным риском, в то время как крупные банки более эффективны, чем средние.
Ключевые слова: операционный риск, эффективность, анализ конвертов данных (DEA), производительность, российские банки.
Для цитирования: Абу-Алроп Д.Х. Повышение эффективности управления операционными рисками в российских банках // Проблемы анализа риска. Т. 17. 2020. № 2. С. 102—119, https://doi. org/10.32686/1812-5220-2020-17-2-102-119
Автор заявляет об отсутствии конфликта интересов.
Jalal H. Abu-Alrop
Assecs the Efficiency of Operational Risk Management in Russian Banks
Assecs the Efficiency of Operational Risk Management in Russian Banks
Jalal H. Abu-Alrop,
Kazan Federal University, 420008, Russia, Republic of Tatarstan, Kazan, Cremlyovskaya str., 18
Keywords: operational Risk, Efficiency, Data Envelopment Analysis (DEA), Performance, Russian Banks.
For citation: Abu-Alrop Jalal H. Assecs the Efficiency of Operational Risk Management in Russian Banks // Issues of Risk Analysis. Vol. 17. 2020. No. 2. P. 102—119, https://doi.org/10.32686/1812-5220-2020-17-2-102-119
The author declare no conflict of interest.
Содержание
Introduction
1. Literature Review
2. Methodology [Data Envelopment Analysis (DEA) Application to Measure Banks Efficiency]
3. Empirical Analysis Conclusion References
Annotation
This study examines the efficiency of operational risk management of 85 Russian commercial banks during the period 2008—2017. This study uses data envelopment analysis (DEA) with financial ratios to assess the efficiency of operational risk management. The study adopts the basic indicator approach (BIA) to measuring operational risk. Also, the study adopts net interest margin (NIM), return on assets (ROA), and return on equity (ROE) for measuring banks performance. The study found that the small banks were the most effective in managing operational risk, while large banks were more efficient than medium banks.
Introduction
Banking performance is a wide concept that includes many issues, such as competition, concentration, efficiency, productivity and profitability (Bikker & Bos 2008, Heffernan 2005). There is a lot of studies that dealt with the subject of banking performance, but there is no approval among researchers on the most appropriate way to measure the efficiency and performance of banks. Most banking performance studies focus on performance and ignore the impact of risk. The study of bank performance and its relationship to risk is very important because of the long-term impact of risk factors. When looking at profitability, the risks associated with profitability indicators should also be analyzed. Research on the impact of risk on banks' performance is expanding rapidly because of its practical importance. The issue of banking risk assessment has become very important, so the study of risk preferences and their impact on bank efficiency is rapidly evolving and has become a magnet for researchers (Begumhan & Cenktan 2008).
The purpose of this study is to measure banks' performance with respect to operational risk preferences, and to assess whether operational risks are reasonably priced using the data envelopment analysis (DEA) approach which is a mathematical programming technique for measuring the performance of organizations in comparison with the boundaries in the sample. Comparing the efficiency of the Bank's operational risk management with its competitors
Original article
Operational bank risk Issues of Risk Analysis, Vol. 17, 2020, No. 2
may provide additional insights to regulatory and supervisory authorities as well as management of the Bank.
1. Literature Review
1.1. Overview of Operational Risk Concept
Operational risk is one of three major risks faced by banks, credit risk is believed to be the biggest risk to the bank, a senior risk officer in large German bank said: "more than 80% of our credit risk is really just operational risk" (a. s. khan, 2006: p. 7). ironically, over the last few years, the focus has been on developing models for measuring and managing credit risk and market risk, but most of the major losses in financial institutions were due to mismanagement of operational risks — more specifically, the behavior of individual individuals or small groups of individuals, operational risks have therefore gained more attention recently.
"Operational risk is the risk associated with business strategy, internal systems, processes, technology and mismanagement" (li sun, 2011, p. 55). in January 2001, the Basel committee on banking supervision (BCBS) convention defined operational risk as "the risk of direct or indirect loss resulting from inadequate or failed internal processes, people and systems or from external events" (BCBS, 2001, p. 2). However, the committee believes that banks should not rely on this general definition, but each bank should have a unique definition of operational risk in accordance with the size, nature and complexity of its activities. Basel committee believes that shortage in understanding and managing operational risk — which almost exists in all bank activities and transactions — to a large extent might decrease the possibility of identifying and controlling some of the risks, thus, the operational loss mainly has three exposure classes namely: people, processes and systems:
1. People: people's risk determines the human error, lack of experience and fraud, including non-compliance with existing procedures and policies.
2. Processes: process risk scope includes insufficient procedures and controls for reporting, monitoring and decision-making, add also insufficient procedures for processing information, such as errors in booking transactions and failure to audit legal documents, organizational deficiencies risk surveillance and excess limits, management deficiencies in risk monitoring, such as not providing the right incentives to report risks, or not abiding by the procedures and policies in force and errors in the recording process of transactions.
3. Systems: technical: technical risks relate to model errors and implementation and lack of sufficient one's instruments for measuring of risk, information technology risks relate to deficiencies in the information system and system failure.
Operational risk management has become important for banks for the higher level of automation in rendering banking and financial services, and increase in global financial inter-linkages scope of operational risk is very wide, the most common operational risks are:
1. Transaction risk: risks arising from fraud, internal or external, failed business processes, inability to maintain business continuity, and information management.
2. Compliance risk: the risk of legal or regulatory sanction, financial loss or loss of reputation that the bank may suffer as a result of its failure to comply with any applicable laws, regulations, codes of conduct and standards of good practice, it is also called the risk of integrity because the bank's reputation is closely linked to its commitment to the principles of integrity and fair dealing.
Since operational risk is measured in terms of a total loss, there are two operational risk components: frequency and severity, effective operational risk management requires a framework designed to convert primary operational risk data into information that supports management decision making, this is much more difficult than market risk or credit risk.
1.2. Overview of Operational Risk Measurement Methods
1. The Basic Indicator Approach (BIA): the basic indicator approach (BIA) is the simplest method and can be applied by all banks. In the basic indicator approach (BIA), operating capital for operating risks is calculated as a fixed percentage of the annual positive gross income average of the financial institution for three years.
2. Standardized Approach (TSA) or Alternative Standardized Approach (ASA): In contrast to the basic indicator approach (BIA), a negative gross income can be used in a single line of action to offset the positive gross income in other lines, resulting in a lower total capital charge, however, the total cost of capital cannot be negative and therefore cannot be used as a deduction from the level of capital or market risk in the financial institution, a financial institution that uses standardized approach (TSA) must map its overall gross lines to eight business lines, which were previously determined by BCBS. for details, please refer to (BCBS, 2006).
The 2007 crisis highlighted shortcomings in the Basel II framework. The main concern was the simpler methods the basic indicator approach (BIA), standardized approach (TSA), alternative standardized approach (ASA), which reflecting lower operational risk exposure despite the highest observed losses during the crisis. These methods are based on the bank's total income as a medium for exposure to operational risk. These methods also assume the linear re-
Jalal H. Abu-Alrop Assecs the Efficiency of Operational Risk Management in Russian Banks
lationship between total income and exposure to risk, but this becomes more complex with increasing size in large organizations, making this relationship nonlinear.
3. Advanced Measurement Approach (AMA): This approach is based on the development of financial institutions for their own methodologies based on internal losses and risk measurement systems. According to BCBS (2001), the goal is to provide the opportunity for development and innovation, but this made comparisons between financial institutions difficult, and the problem of lack of transparency and lack of clarity emerged.
2011 Basel Committee revised its principles for the sound management of operational risks (BCBS-195) and issued supervisory guidelines for the AMA Approach (BCBS-196). However, in 2014, BCBS concluded that banks had made insufficient progress in applying the BCBS-292 principles, which meant that many organizations had not considered operational risks and dealt with them seriously despite losses since 2003.
4. The Operational Risk Capital-At-Risk Approach (OP-CAR): This approach provides the foundation for the new approach, the standardized measurement approach (SMA), it only aims to replace the simpler approach (i.e., non-AMA).
5. The Standardized Measurement Approach (SMA): In 2016, the name SMA (BCBS-d355) appeared and the approach was expanded to replace the advanced measurement approach's (AMA). The calculation method is the same but the details have been reviewed. The new version was published by the Basel Committee as part of the final Basel III provisions in December 2017. Advanced flexible measurement method (AMA) and also the simple methods currently available Will be replaced to suit the new the standardized measurement approach (SMA) as of 2022.
2. Methodology [Data Envelopment Analysis (DEA) Application to Measure Banks Efficiency]
Researchers used different techniques to evaluate the efficiency of banks, three important surveys included bank efficiency studies:
1. The first (Berger and Humphrey 1997) in their review of 130 studies on the efficiency of banks found that 69 of them used data envelopment analysis (DEA).
2. The second (Fathi and Basoras 2010) in their review of 196 studies found that 151 of them used techniques similar to data envelopment analysis (DEA).
3. The third (Paradi and Zhu 2013) reported that there are 275 applications of Data Envelopment Analysis (DEA) in banking efficiency studies.
"Data envelopment analysis (DEA) is a mathematical programming technique for the development of production frontiers and the measurement of efficiency relative to these frontiers (Charnes et al, 1978). Data envelopment analysis (DEA), a non-parametric technique, for the estimation of production frontiers for given inputs and outputs of a set of decision-making units (DMUs). Introduced by Farrell (1957) and developed by Charnes, Cooper and Rhodes (1978), data envelopment analysis (DEA) assumes that if a unit can produce a certain level of output utilizing specific input levels, another unit of equal scale should be capable of doing the same. The most efficient producers can form a «composite producer», allowing the computation of an efficient solution for every level of input or output as a «Virtual producer» and to make comparisons." (Saha et al. 2015, p: 29). "The efficiency rate of a unit can be expressed as:
Woighted sum of outputs Xsi=1 uoyk
Weighted sum of inputs Ztl=1 v.y.
(1)
)=1 j'jq
yiq: is the quantity of output i produced by firm q. vj is the weight of input i. Xjq: is the quantity of input j consumed by firm q. ui: is the weight of output i. s: is the number of outputs. m: is the number of inputs. n: is the number of firms to be evaluated. To estimate the efficiency rate in Formula (1) above, this is based on an estimate of the input and output weights. This requires specifying the weights Vj and Ui in advance, meaning that the decision maker must determine the relative importance of the inputs and outputs in the analysis, thus modules can be rated from worst to best performing. Data envelopment analysis (DEA) models derive input and output weights by optimizing. Accordingly, units can be classified as efficient and inefficient. The data envelopment analysis (DEA) analysis can determine the target values for inputs and outputs that may lead to efficiencies." (Kristina, 2005, p. 25)
"Data envelopment analysis (DEA) helps to identify efficient companies to build efficient production frontier. Data envelopment analysis (DEA) models measure the relative efficiency that is the efficiency of each company relative to similar companies in the sample, thus applying data envelopment analysis (DEA) in evaluating the performance of a set of companies, it is possible to form two groups: companies that comprise an efficient frontier and inefficient companies lying below the frontier. In applying the data envelopment analysis (DEA) model, the efficiency score is estimated as the ratio of weighted outputs to weighted inputs (Charnes et al.1978). Weights are selected for each variable
Original article Operational bank risk Issues of Risk Analysis, Vol. 17, 2020, No. 2
of every analyzed unit in order to maximize its efficiency score. The efficiency rate for each unit of DMU is evaluated relative to the other set members (Charnes et al. 1978). The maximal efficiency score is equal to 1, and the lower values indicate relative inefficiency of the analyzed DMU." (Jelena et al. 2014, p. 742—743). "However, there are two conditions, the first is that the efficiency of any other units in the population should not be more than 1, the second condition is that weights of all inputs and outputs must be greater than zero. Such a model is defined as a linear divisive programming model:" (Kristina, 2005, p. 25).
Maximize: -
1 u.y.
= 1 i_ iq
Zm, v.y.
i=i vv
Subject to:
v vq
Zs , u.y.
r=1 , < 1 k = 1, 2 ..., n (2)
vjXjq ()
where: ui > e i= 1, 2, ..., s. vj < e i= 1, 2, ..., m. yiq: is the quantity of output i produced by firm q. v.: is the weight of input i. x.q: is the quantity of input j consumed by firm q. ui: is the weight of output i. s: is the number of outputs. m:is the number of inputs. n:is the number of firms to be evaluated.
"This model can be converted into a linear programming model1 and transformed into a matrix:
Maximize: z = uTY
q
Subject to: vTXq = 1
(3)
uTy - vTX Where: u > e, v < e.
Model (3) is often called the primary CCR (Charnes, Cooper, Rhodes) model. The dual model to this can be stated as follows:" (Kristina, 2005, p. 25).
Maximize: f = 9 - e (eTs++ (eTs-) (4)
Subject to: YX - s+ = Yq
XX + s- = 0Xq Where: X, s+, s- > 0.
X = (Xj, X2, Xn), Xj > 0, is a vector assigned to individual productive units.
s+ and s-, are vectors of addition input and output variables.
eT= (1, 1, ..., 1) and e, is a constantl greater than zero, which is normally pitched at 10-6 or 10-8.
1 The term linear programming consists of two words explaining the substance of this particular branch of operational research. Programming is a synonym for predicting future development. The word linear means that all equations and inequalities used in the model are linear (Jablonsky, 2002. cited in Ing, 2005, p. 25)
"In evaluating the efficiency of unit DMUq, model (4) seeks a virtual unit characterized by inputs XX and outputs YX, which are a linear combination of inputs and outputs of other units of the population and which are better that the inputs and outputs of unit DMUq which is being evaluated. For inputs of the virtual unit XX < Xq and for outputs YX > Yq. Unit DMUq is rated efficient if no virtual unit with requested traits exists or if the virtual unit is identical with the unit evaluated, i.e. XX = X and YX = Y . If unit DMU is
CCR efficient, then:
• The value of variable 9 is zero.
• The values of all additional variables s+ and equal zero." (Kristina, 2005, p. 25).
"Consequently, unit DMUq is the primary CCR (Charnes, Cooper, Rhodes) model efficient if the optimum value of the model (4) objective function equals one. Otherwise, the unit is inefficient. The optimum value of the objective function/* marks the efficiency rate of the unit concerned. The lower the rate, the less efficient the unit is compared to the rest of the population. In inefficient units 9 is less than one. This value shows the need for a proportional reduction of inputs for unit DMU q to become efficient. The advantage of the data envelopment analysis (DEA) model is that it advises how the unit evaluated should mend its behavior to reach efficiency. Models (3) and (4) are input-oriented — they try to find out how to improve the input characteristics of the unit concerned for it to become efficient" (Kristina, 2005, p. 25).
On the other hand, there are output-oriented models, but we will not address that because our study uses an input-oriented model. In data envelopment analysis (DEA) models, the input-oriented models are the most used to measure bank efficiency (Arshinova 2011; Asror 2010; Yang 2009; Zreika, Ekanj 2011). This is because bank managers have more control over inputs rather than outputs (Fethi,Pasiouras 2010).
"Later, the model was modified to the model Banker, Charnes, & Cooper (BCC) in 1984, which used the variable returns to scale technology (VRS) assumption. The variable returns to scale technology (VRS) assumption suggests that equiproportionate increases in factor inputs yield a greater (or less) than equiproportionate increase in output (Heffer-nan, S. 2005). Experts point to the fact that constant returns to scale (CRS) can only be applied for the companies which operate optimally (Coelli et al. 2005). However, in many industries (including banking sector) such factors, as imperfect competition or government regulations, may cause the deviation from an optimal scale (Coelli et al. 2005; Beccalli et al. 2006; Singh et al. 2008). In addition, the variable returns to scale technology (VRS) is considered to be a more
Jalal H. Abu-Alrop Assecs the Efficiency of Operational Risk Management in Russian Banks
appropriate assumption for measuring efficiency in developed banking sector (McAllister & McMaus 1993; Whee-lock & Wilson 1995)." (Jelena et al, 2014, p. 743—744). So, our study will use variable returns to scale technology (The variable returns to scale technology (VRS) model).
3. Empirical Analysis
3.1. Data and Variables
This study aims to assess the operational risks efficiency and financial performance of Russian commercial banks according to the data envelopment analysis (DEA) relative efficiency measurement characteristic. This banks group should be as homogeneous as possible to be meaningful. Therefore, banks with the largest assets were selected. In this study, the data of the largest 85 Russian banks. The total assets of the largest 85 banks selected for the study constitute 87% of the total assets of the banking sector in Russia. We divided the banks into three equal groups based on the size of the assets. The first group consisted of 28 banks, it included the banks which have total assets between (270 billion Rubles to 23 trillion Rubles) were categorized as large banks, The second group consisted of 29 banks, and included the banks which have total assets of between (102—270 billion Rubles) were categorized as medium banks, and The third group consisted of 28 banks, and included the banks which have total assets of between(5 — 102 billion Rubles) were categorized as small banks. The sample panel data include the year-end data for the period 2008—2017. This study uses financial ratios, simple regression and data envelopment analysis (DEA) model to assess the efficiency of Russian banks in managing operational
risks. All study data were obtained from the official website of the Bank of Russia. The study period includes 10 years (2008—2017). Table 1 defines the study variables, their abbreviations and the method of calculation.
Table 2 and Figure 1 shows the average of operational risks in Russian banks according to the size of the banks calculated on the basis of the basic indicator approach (BIA).
The next step is to find the appropriate variables to be included in the data envelopment analysis (DEA) model as inputs and outputs. The discriminatory power of the data envelopment analysis (DEA) will be reduced when there are a large number of variables. Therefore, until this problem is overcome, the variables must be minimized using appropriate scientific methods. This issue has been widely discussed and there are many ways to choose variables (Jenkins, Anderson, (2003); Fanchon, (2003); Ruggiero, (2005); Adler, Yazhemsky (2010); Luo, Liang (2012); Xie et al. (2014); Niranjan et al. (2011), Hiroshi Morita et.al., (2009); Subramanyam T (2016)). Here in our study, we will select the variables by analyzing the multiple regression of the variables to find the effect of dependent variables (inputs) on the independent variables (outputs) and then we will choose the variables with statistical significance.
3.2. The Simple-Regression Model
A general linear model of simple — regression is outlined in equation (5) where Y indicates the dependent variables (outputs), a is the constant, P is the regression coefficient, X is the independent factor (input) and finally, s is a random factor.
Y = a + P1X1 + s. (5)
Table l. Variables Definition and Measurement Units
Variables Description Abbreviation Variables Proxy
Independent variables (Inputs) Operational Risk OPR (Gross Income ) / (Total Shareholder's Equity)
Dependent variables Net Interest Income NIM (Net Interest Income) / (Total Assets)
(Outputs) Return On Assets ROA (Income After Tax) / (Total Assets)
Return On Equity ROE (Income After Tax) / (Total Shareholders' Equity)
Source: Author Design.
Table 2. The Average of Operational Risk in Russian Banks Based on The Basic Indicator Approach (BIA) (2008—2017)
Years 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Mean
Large Banks O.65 O.48 O.64 O.64 O.59 O.59 O.59 -O.58 O.31 O.45 0.44
Medium Banks O.57 O.42 O.55 O.69 O.68 O.69 O.55 O.26 O.58 O.55 0.55
Small Banks O.6O O.55 O.44 O.5O O.5O O.45 O.45 O.4O O.39 O.54 0.48
Mean 0.61 0.49 0.54 0.61 0.59 0.58 0.53 0.03 0.43 0.51 1.47
Source: Design and Calculation by Author Using (Excel). Data Source: Bank of Russia Website.
Original article Operational bank risk Issues of Risk Analysis, Vol. 17, 2020, No. 2
Operational Risk
■Ü 75%
S3 60%
s 45% .......................................... /
/
.................
I 30% If
Ö 15% „
§ 0% \ B
? -30% .
1-45% \ °
¿a -60%
Years
= Large Banks — — —m Medium Banks .......Small Banks
Figure 1. The Average of Operational Risk in Russian Banks Based on The Basic Indicator Approach (BIA) (2008—2017) Source: Design and Calculation by Author Using (Excel And SPSS Software). Data Source: Bank of Russia Websit.
Net interest margin (MM), return on assets (ROA), and return on equity (ROE) are the factors of profitability and performance that are influenced by operational risks (OPR) factor. By putting the study variables in the above equation, three equations can be formed as follows: NIM = a + Pj [(Gross Income)/ (Total Shareholder's Equity)]. (6) ROA =a + Pj [(Gross Income)/(Total Shareholder's Equity)]. (7) ROE =a + P1 [(Gross Income) / (Total Shareholder's Equity)]. (8)
3.2.1. The Main Hypotheses
The main hypotheses can be formulated as follows:
Ho: Operational risks (OPR) don't affect financial performance (expressed by NIM, ROA and ROE) of the Russian commercial banks. H1: Operational risks (OPR) affect financial performance (expressed by NIM, ROA, and ROE) in Russian commercial banks. 3.2.1.1. The Subset Hypothesis 1 — NIM Model:
Ho: Operational risks (OPR) doesn't affect MM in Russian banks. Hi: Operational risks (OPR) affect MM in Russian banks.
2 — ROA Model:
Ho: Operational risks (OPR) doesn't affect ROA in Russian banks. H1: Operational risks (OPR) affect ROA in Russian banks.
3 — ROE Model:
Ho: Operational risks (OPR) doesn't affect ROE in Russian banks. H1: Operational risks (OPR) affect ROE in Russian banks.
Based on the main and Subset hypotheses above, three sub-hypotheses will be tested for each year of study years which are MM model, ROA model and ROE model, because the study years are 10 years, therefore 30 models will be tested, 3 models for each year.
3.2.2. Testing(F) For the Suitability of The Research Models
To examine the suitability of the multiple regression models for analysis, by using the distribution (F-statistic) test, one of the following hypotheses will be rejected:
Ho: The model is unsuitable; when the independent variables don't affect the dependent variables.
H1. The model is suitable; when the independent variables do affect the dependent variables. The decision rule as follows: Accept Ho If p-value (Sig. F) > 0.05 Accept Hi vp-value (Sig. F) < 0.05 From the analysis output in Table 3, the results as follow: The Models (1), (2), (3), (4), (5), (6), (8), (9), (10), (11), (12), (13), (14), (15), (16) (17), (18), (19), (24), (25), (26), (27) and (28): values of p-value (Sig. F) < 0.05, So we shall refuse the null hypothesis Ho and accept the alternative hypothesis Hi, that means At the a = 0.05 level of significance, there is enough evidence to conclude that the predictor is useful for predicting the NIM or ROA or ROE ; therefore, the models are suitable.
The Models (7), (20), (21) ,(22), (23), (29) and (30): values of p-value (Sig. F) > 0.05 ,So we shall accept the null hypothesis Ho and refuse the alternative hypothesis Hi, that means At the a = 0.05 level of significance, there isn't enough evidence to conclude that the predictor is useful for predicting the MM or ROA or ROE ; therefore, the models are unsuitable (Table 3).
3.2.3. R-square for the Appropriate Models
Table 4 showing the variability percentage of independent variables. The (R square) demonstrates the relationship between dependent and independent variables whereas (R) represents the square root of (R). The value of (R) points out how independent variables are associated with MM, ROA and ROE.
Jalal H. Abu-Alrop
Assecs the Efficiency of Operational Risk Management in Russian Banks
Moreover, the (adjusted R square) mentions the statistical shrinkage of risks variables. Simply, (adjusted R square) refers to the compatibility of independent variables with dependent ones in order to validate the decisions based on the regression model (Table 4).
3.2.4. Testing (T) For the Appropriate Models
To examine the suitability of the multiple regression models for analysis, by using the distribution (T-statistic) test, one of the following hypotheses will be rejected:
Ho. The model is not suitable (when the independent variables don't affect the dependent variables).
Table 3. F-Test — ANOVA (2008—2017)
Years Model Name Model # F-Statistic Sig. F-Statistic The Decision Years Model Name Model # F-Statistic Sig. F-Statistic The Decision
200В NIM Model (1) 4.32B .041a Suitable 2013 NIM Model (16) 67.162 .000a Suitable
ROA Model (2) B.523 .005a Suitable ROA Model (17) 25.494 .000a Suitable
ROE Model (3) 26.400 .000a Suitable ROE Model (18) 50.934 .000a Suitable
2009 NIM Model (4) 6.077 .016a Suitable 2014 NIM Model (19) 22.077 .000a Suitable
ROA Model (5) 31.714 .000a Suitable ROA Model (20) 1.692 .197a Unsuitable
ROE Model (6) 50.204 .000a Suitable ROE Model (21) .202 .654a Unsuitable
2010 NIM Model (7) 1.979 .163a Unsuitable 2015 NIM Model (22) .327 .569a Unsuitable
ROA Model (B) 5.49B .021a Suitable ROA Model (23) .145 .705a Unsuitable
ROE Model (9) 19.559 .000a Suitable ROE Model (24) 43.900 .000a Suitable
2011 NIM Model (10) 44.921 .000a Suitable 2016 NIM Model (25) 15.518 .000a Suitable
ROA Model (11) 4.442 .038a Suitable ROA Model (26) 28.063 .000a Suitable
ROE Model (12) 31.693 .000a Suitable ROE Model (27) 49.373 .000a Suitable
2012 NIM Model (13) 4B.921 .000a Suitable 2017 NIM Model (28) 5.269 .024a Suitable
ROA Model (14) 12.966 .001a Suitable ROA Model (29) .362 .549a Unsuitable
ROE Model (15) 51.516 .000a Suitable ROE Model (30) 2.721 .103a Unsuitable
Source: Design and Calculation by Author Using (Excel And SPSS Software). Data Source: Bank of Russia Website.
Table 4. The Total Variation in The Dependent Variables (2008—2017)
Years Model Name Model # R2 Adjusted R2 Sig. R The Decision Years Model Name Model # R2 Adjusted R2 Sig. R The Decision
200В NIM Model (1) .050 .038 .223a Suitable 2013 NIM Model (16) .447 .441 .669a Suitable
ROA Model (2) .093 .082 .305a Suitable ROA Model (17) .235 .226 .485a Suitable
ROE Model (3) .241 .232 .491a Suitable ROE Model (18) .380 .373 .617a Suitable
2009 NIM Model (4) .068 .057 .261a Suitable 2014 NIM Model (19) .210 .201 .458a Suitable
ROA Model (5) .276 .268 .526a Suitable ROA Model (20) * * * Unsuitable
ROE Model (6) .377 .369 .614a Suitable ROE Model (21) * * * Unsuitable
2010 NIM Model (7) * * * Unsuitable 2015 NIM Model (22) * * * Unsuitable
ROA Model (B) .062 .051 .249a Suitable ROA Model (23) * * * Unsuitable
ROE Model (9) .191 .181 .437a Suitable ROE Model (24) .346 .338 .588a Suitable
2011 NIM Model (10) .351 .343 .593a Suitable 2016 NIM Model (25) .158 .147 .397a Suitable
ROA Model (11) .051 .039 .225a Suitable ROA Model (26) .253 .244 .503a Suitable
ROE Model (12) .276 .268 .526a Suitable ROE Model (27) .373 .365 .611a Suitable
2012 NIM Model (13) .371 .363 .609a Suitable 2017 NIM Model (28) .060 .048 .244a Suitable
ROA Model (14) .135 .125 .368a Suitable ROA Model (29) * * * Unsuitable
ROE Model (15) .383 .376 .619a Suitable ROE Model (30) * * * Unsuitable
* A Model Was Excluded Because It Failed to Pass An F-Test That Measures the Suitability of The Model for Prediction. Source: Design and Calculation by Author Using (Excel and SPSS software). Data Source: Bank of Russia Website.
Original article Operational bank risk Issues of Risk Analysis, Vol. 17, 2020, No. 2
Table 5. T-Test (2008—2017)
Years Out- Model # Inputs B T Sig. The Years Out- Model # Inputs B T Sta- Sig. The
puts Statis- Statis- Decision puts tistic Statis- Decision
tic tic tic
2008 NIM Model (1) constant .048 7.180 .000 Suitable 2013 NIM Model (16) constant .012 1.992 .050 Suitable
OPR .019 2.080 .041 Suitable OPR .071 8.195 .000 Suitable
ROA Model (2) constant .000 .054 .957 Unsuitable ROA Model (17) constant -.001 -.161 .873 Unsuitable
OPR .015 2.919 .005 Suitable OPR .024 5.049 .000 Suitable
ROE Model (3) constant -.254 -3.700 .000 Suitable ROE Model (18) constant -.001 -.067 .947 Unsuitable
OPR .476 5.138 .000 Suitable OPR .189 7.137 .000 Suitable
2009 NIM Model (4) constant .044 5.471 .000 Suitable 2014 NIM Model (19) constant .022 2.966 .004 Suitable
OPR .032 2.465 .016 Suitable OPR .054 4.699 .000 Suitable
ROA Model (5) constant -.023 -3.784 .000 Suitable ROA Model (20) constant * * * Unsuitable
OPR .054 5.632 .000 Suitable OPR * * * Unsuitable
ROE Model (6) constant -.151 -4.832 .000 Suitable ROE Model (21) constant * * * Unsuitable
OPR .355 7.086 .000 Suitable OPR * * * Unsuitable
2010 NIM Model (7) constant * * * Unsuitable 2015 NIM Model (22) constant * * * Unsuitable
OPR * * * Unsuitable OPR * * * Unsuitable
ROA Model (8) constant .004 .905 .368 Unsuitable ROA Model (23) constant * * * Unsuitable
OPR .013 2.345 .021 Suitable OPR * * * Unsuitable
ROE Model (9) constant -.028 -1.073 .286 Unsuitable ROE Model (24) constant -.334 -2.090 .040 Suitable
OPR .166 4.423 .000 Suitable OPR .336 6.626 .000 Suitable
2011 NIM Model (10) constant .017 2.711 .008 Suitable 2016 NIM Model (25) constant .035 7.675 .000 Suitable
OPR .058 6.702 .000 Suitable OPR .026 3.939 .000 Suitable
ROA Model (11) constant -.002 -.321 .749 Unsuitable ROA Model (26) constant -.014 -3.179 .002 Suitable
OPR .022 2.108 .038 Suitable OPR .034 5.297 .000 Suitable
ROE Model (12) constant -.016 -.604 .548 Unsuitable ROE Model (27) constant -.173 -4.896 .000 Suitable
OPR .206 5.630 .000 Suitable OPR .368 7.027 .000 Suitable
2012 NIM Model (13) constant .016 2.962 .004 Suitable 2017 NIM Model (28) constant .037 7.126 .000 Suitable
OPR .055 6.994 .000 Suitable OPR .017 2.295 .024 Suitable
ROA Model (14) constant -.004 -.612 .542 Unsuitable ROA Model (29) constant * * * Unsuitable
OPR .038 3.601 .001 Suitable OPR * * * Unsuitable
ROE Model (15) constant -.012 -.573 .568 Unsuitable ROE Model (30) constant * * * Unsuitable
OPR .226 7.177 .000 Suitable OPR * * * Unsuitable
A model was excluded because it failed to pass an f-test that measures the suitability of the model for prediction. Source: Design and Calculation by Author Using (Excel and SPSS software). Data Source: Bank of Russia Website.
Hr The model is suitable (when the independent variables affect the dependent variables).
The decision rule as follows: Accept Ho If p-value (Sig. T) > 0.05, Accept H If p-value (Sig. T) < 0.05 (Table 5). From the T-test analysis in Table 5, the results as follow: The Models (1), (3), (4), (5), (6), (10), (13), (16), (19), (24), (25), (26), (27) and (28): values of p-value (Sig. T) < 0.05 ,So we shall refuse the null hypothesis Ho and accept the alternative hypothesis H1, that means At the a = 0.05 level of significance, there exists enough evidence to conclude that the slope (B) of the variables mentioned above is not zero and, hence, that variables are useful for
predicting MM, ROA and ROE in Russian banks; therefore, the models are suitable.
The Models (2), (8), (9), (11), (12), (14), (15), (17) and (18): values of p-value (Sig. T) < 0.05 for (OPR), but for (constant) (Sig. T) > 0.05, So we shall refuse the null hypothesis Ho and accept the alternative hypothesis H1 with exclusion the constant of the regression equation , that means At the a = 0.05 level of significance, there exists enough evidence to conclude that the slope (B) of the variable (OPR) is not zero Thus, this variable is useful for predicting MM, ROA and ROE in Russian banks with exclusion the constant, therefore, the models are suitable with exclusion the constant.
Jalal H. Abu-Alrop Assecs the Efficiency of Operational Risk Management in Russian Banks
Table 6. Results of Multiple Regression Analysis of Study Models
Regression Analysis Results Models #
Accepted Models 1, 3, 4, 5, 6, 10, 13, 16, 19, 24, 25, 26, 27, 28
Accepted Models Provided the Constant is Excluded 2, 8, 9, 11, 12, 14, 15, 17, 18
Rejected Models 7, 20, 21, 22, 23, 29, 30
Source: Design and Calculation by Author Using (Excel and SPSS software). Data Source: Bank of Russia Website.
NIM (2008-2017) = 5.02% ROA (2008-2017) = 0.35% ROE (2008-2017) = -0.27%
□ OPR □ Other veriabeh Q OPR □ Other veriabels H OPR □ Other veriabels
Figure 2. The Ratios of The Contribution of Operational Risk Indicators in the Formation of Performance Indicators (2008—2017)
OPR: Operational Risk. NIM: Net Interest Income. ROA: Return on Assets. ROE: Return on Equity. Source: Design and Calculation by Author Using (Excel, Win4deap2 Software). Data Source: Bank of Russia Website.
The Models (7), (20), (21) ,(22), (23), (29) and (30): values of p-value (Sig. F) > 0.05 ,So we shall accept the null hypothesis Ho and refuse the alternative hypothesis Hi, that means At the a = 0.05 level of significance, there isn't enough evidence to conclude that the variable is useful for predicting the MM or ROA or ROE ; therefore, the models are unsuitable. Table 6 below summarizes the results of multiple regression analysis (Table 6).
The value of slope B in the above Table 5 represents the ratio of effect and the type of relationship between the independent variables and the dependent variable. In order to know the importance of operational risk indicator and its impact on performance indicators, it is necessary to determine its real value compared to other variables. Therefore, we multiply the value B by the mean of the dependent variable (OPR), this makes us know the value of its effect as compared to other variables. Figure 2 shows the contribution of the operational risk index to the formation of performance indicators during the study period (2008-2017). Operational risk has contributed to the formation of NIM, ROA and ROE performance indicators at 37%, 60% and 49% respectively, reflecting the impact of operational risk on the performance of Russian banks (Figure 2).
Based on the above, inputs and outputs will be adopted in the data envelopment analysis (DEA) analysis as shown in Table 7.
3.3. The Efficiency of Operational Risk [Data Envelopment Analysis (DEA)]
Tables 8A and 8B present the results of Data Envelopment Analysis (DEA). We use an input-oriented model [data envelopment analysis (DEA) — the variable returns to scale technology (VRS)] to assess the technical efficiency of operational risk management. The results showed that no
Table 7. The Inputs and Outputs Which Will Use in Data Envelopment Analysis (DEA)
Year Inputs Outputs
2008 OPR NIM,ROA,ROE
2009 OPR NIM,ROA,ROE
2010 OPR ROA,ROE
2011 OPR NIM,ROA,ROE
2012 OPR NIM,ROA,ROE
2013 OPR NIM,ROA,ROE
2014 OPR NIM
2015 OPR ROE
2016 OPR NIM,ROA,ROE
2017 OPR NIM
OPR: Operational Risk. NIM: Net Interest Income. ROA: Return on Assets. ROE: Return on Equity.
Source: Author Design.
Original article
Operational bank risk Issues of Risk Analysis, Vol. 17, 2020, No. 2
bank achieved full efficiency in operational risk management continuously in all ten years of study. In 2008 eight banks achieved the perfect efficiency score (1) namely, Banks # 32, 35, 42, 45, 49, 51, 71, and 76. while the worst bank in operational risk Management was namely, Bank # 24 with efficiency score 0.29.
In 2009 twenty-four banks achieved the perfect efficiency score (1) namely, Banks # 12, 24, 33, 39, 42, 46, 47, 48, 50, 51, 57, 59, 60, 61, 63, 64, 69, 70, 72, 73, 77, 78, 82 and 83. while the worst bank in operational risk Management was namely, Bank # 25 with efficiency score 0.12.
In 2010 three banks achieved the perfect efficiency score (1) namely, Banks # 35,41 and 69. while the worst banks in operational risk Management were namely, Banks #21 with efficiency score 0.29 (Table 8A).
In 2011 six banks achieved the perfect efficiency score 1.0, namely, Banks # 3, 12, 33, 39, 51 and 69. while the worst bank in operational risk Management was namely, Banks # 48 with efficiency score 0.42.
In 2012 three banks achieved the perfect efficiency score 1.0, namely, Banks # 23, 33 and 51. while the worst banks in operational risk Management were namely, Bank # 65 with efficiency score 0.07.
In 2013 three banks achieved the perfect efficiency score 1.0, namely, Banks # 29, 44 and 56. while the worst banks in operational risk Management were namely, Bank # 34 with efficiency score 0.29.
In 2014 nine banks achieved the perfect efficiency score 1.0, namely, Banks # 5, 21, 39, 44, 49, 61, 69, 73 and 80. while the worst banks in operational risk Management were namely, Bank # 77 with efficiency score 0.25.
In 2015 thirty two banks achieved the perfect efficiency score 1.0, namely, Banks 3, 4, 9, 11, 14, 18, 19, 20, 25, 28, 29, 31, 32, 37, 38, 40, 41, 43, 44, 45, 46, 49, 56, 59, 60, 62, 66, 79, 81, 82, 84 and 85. while the worst bank in operational risk Management was namely, Bank # 48 with efficiency score 0.01.
In 2016 eight banks achieved the perfect efficiency score 1.0, namely, Banks 3, 11, 14, 22, 36, 44, 64 and 66. while the worst bank in operational risk Management was namely, Bank # 48 with efficiency score 0.11.
In 2017 six banks achieved the perfect efficiency score 1.0, namely 22,23, 41, 44, 67 Banks and 83. while the worst bank in operational risk Management was namely, Bank #18 with efficiency score 0.08.
Table 8. The Technical Efficiency [(DEA) — Input Oriented — (VRS)] of Operational Risk Management in Russian banks (2008—2017)
Bank Bank # Years Mean
2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
Sberbank Of Russia 1 0.99 0.9В 0.9В 0.В4 0.39 0.В3 0.99 0.92 0.72 0.В2 0.85
VTB Bank 2 0.В1 0.96 0.В4 0.7В 0.52 1.00 0.33 0.99 0.97 0.90 0.81
Gazprombank 3 0.93 0.93 0.В7 1 0.59 0.94 0.В3 1 1 0.В1 0.89
Rosselkhozbank 4 0.63 0.В5 0.97 0.74 0.66 0.96 0.46 1 0.ВВ 0.25 0.74
Alfa-Bank 5 0.97 0.В6 0.90 0.96 0.36 0.В4 1 0.В5 0.99 0.95 0.87
Credit Bank Of Moscow 6 0.9В 0.97 0.93 0.92 0.3В 0.В6 0.99 0.99 0.96 0.92 0.89
Bank Otkritie Financial Corporation 7 0.9В 0.97 0.91 0.94 0.50 0.В9 0.90 0.99 0.97 0.29 0.83
Unicredit Bank В 0.95 0.99 0.94 0.99 0.57 0.ВВ 0.В7 0.99 0.99 0.97 0.91
B®N Bank 9 0.99 0.66 0.В1 0.94 0.72 0.92 0.54 1 0.72 0.29 0.76
Promsvyazbank 10 0.64 0.6В 0.91 0.97 0.43 0.В2 0.49 0.91 0.9В 0.40 0.72
Rosbank 11 0.94 0.74 0.В9 0.9В 0.59 0.91 0.95 1 1 0.51 0.85
Raiffeisenbank 12 0.ВВ 1 0.95 1 0.46 0.В2 0.99 0.93 0.69 0.В5 0.86
Sovcombank 13 0.70 0.9В 0.60 0.ВВ 0.31 0.73 0.72 0.43 0.70 0.64 0.67
Bank Saint-Petersburg 14 0.99 0.97 0.В4 0.95 0.70 0.В7 0.95 1 1 0.93 0.92
Bank Uralsib 15 0.9В 0.72 0.90 0.79 0.76 0.92 0.75 0.В6 0.7В 0.9В 0.84
Bank RRDB 16 0.99 0.99 0.В9 0.92 0.6В 0.96 0.В4 0.В7 0.90 0.ВВ 0.89
Jalal H. Abu-Alrop
Assecs the Efficiency of Operational Risk Management in Russian Banks
Bank Bank # Years Mean
2008 2003 2010 2011 2012 2013 2014 2015 201a 2017
Citibank 1У 0.54 0.96 0.B1 0.99 0.4B 0.B2 0.9B 0.5У 0.У1 0.У1 0.7a
Growth Bank 1B 0.B6 0.BB 0.У5 0.91 0.У2 0.95 0.64 1 0.11 0.0B 0.a3
Ak Bars Bank 19 0.93 0.9У 0.51 0.7B 0.4B 0.69 0.39 1 0.У6 0.УУ 0.73
Bm-Bank 20 0.BB 0.93 0.93 0.У9 0.B1 0.B2 0.B2 1 0.65 0.45 0.81
NB Trust 21 0.5B 0.33 0.29 0.B9 0.95 0.B2 1 0.93 0.У2 0.15 0.aa
Mosobl bank 22 0.9B 0.30 0.4B 0.54 0.B4 0.У6 0.У0 0.94 1 1 0.75
Smp Bank 23 0.9У 0.96 0.56 0.90 1 0.B9 0.95 0.9B 0.B0 1 0.30
Russian Standard Bank 24 0.29 1 0.56 0.У1 0.41 0.B7 0.65 0.ЗУ 0.B6 0.7B 0.a5
Bank Dom.Rf 25 0.УЗ 0.12 0.42 0.53 0.69 0.90 0.42 1 0.23 0.35 0.54
Novikom bank 26 0.У6 0.9У 0.9У 0.9У 0.У2 0.BB 0.96 0.90 0.BB 0.B9 0.83
The Ural Bank For Reconstruction And Development 2У 0.BB 0.42 0.B7 0.6B 0.90 0.B1 0.96 0.91 0.93 0.25 0.7a
Moscow Industrial Bank 2B 0.9B 0.99 0.B5 0.У5 0.96 0.95 0.B2 1 0.49 0.25 0.80
Sviaz-Bank 29 0.44 0.66 0.B3 0.99 0.B3 1 0.У6 1 0.66 0.BB 0.80
HCF Bank 30 0.93 0.B1 0.62 0.4B 0.24 0.55 0.34 0.03 0.42 0.56 0.50
Absolut Bank 31 0.96 0.УУ 0.94 0.65 0.У4 0.94 0.B6 1 0.7B 0.ЗУ 0.80
Vozrozhdenie Bank 32 1 0.9B 0.B2 0.9У 0.6B 0.B3 0.9У 1 0.94 0.94 0.31
Post Bank 33 0.99 1 1 1 1 0.55 0.49 0.29 0.66 0.УУ 0.77
Tinkoff Bank 34 0.5У 0.31 0.53 0.43 0.1B 0.29 0.33 0.09 0.29 0.44 0.34
Orient Express Bank 35 1 0.91 1 0.У1 0.42 0.B7 0.45 0.04 0.6У 0.92 0.70
Surgutneftegas bank 36 0.У2 0.42 0.63 0.94 0.61 0.B5 0.99 0.95 1 0.9B 0.81
Bank Zenit ЗУ 0.9B 0.99 0.B3 0.90 0.9B 0.94 0.B0 1 0.ЗУ 0.31 0.81
Trans kapital bank 3B 0.9B 0.99 0.94 0.95 0.46 0.7B 0.9B 1 0.92 0.32 0.83
Rosevro bank 39 0.99 1 0.91 1 0.35 0.B1 1 0.66 0.У2 0.94 0.84
Nordea Bank 40 0.B7 0.95 0.9У 0.BB 0.6B 0.90 0.96 1 0.B0 0.9B 0.30
Cb Deltacredit 41 0.9У 0.91 1 0.94 0.4У 0.B7 0.9B 1 0.У0 1 0.88
Ing Bank (Eurasia) 42 1 1 0.9У 0.9У 0.61 0.53 0.B7 0.BB 0.B6 0.90 0.8a
Mts Bank 43 0.У0 0.50 0.49 0.46 0.У6 0.90 0.B4 1 0.9B 0.9У 0.7a
Avers 44 0.93 0.96 0.B6 0.B3 0.92 1 1 1 1 1 0.35
Renaissance Credit 5 4 1 0.92 0.УУ 0.УЗ 0.33 0.B7 0.У2 1 0.13 0.55 0.70
Invest trade bank 46 0.B9 1 0.У2 0.B7 0.9B 0.91 0.91 1 0.У2 0.1B 0.82
Cetelem Bank 4У 0.6B 1 0.BB 1.00 0.45 0.B9 0.9B 0.31 0.92 0.7B 0.73
Jsc"Otp Bank" 4B 0.У2 1 0.У6 0.42 0.41 0.B6 0.ЗУ 0.01 0.11 0.7B 0.54
Joint Stock West Siberian Commercial Bank 49 1 0.9У 0.B4 0.99 0.40 0.У5 1 1 0.9У 0.9B 0.83
Avangard Joint Stock Bank 50 0.99 1 0.B5 0.93 0.5У 0.У5 0.9B 0.B1 0.У9 0.91 0.8a
Original article
Operational bank risk Issues of Risk Analysis, Vol. 17, 2020, No. 2
Bank Bank # Years Mean
2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
Bank Finservice 51 1 1 0.95 1 1 0.93 0.93 0.77 0.81 0.77 0.92
Skb-Bank 52 0.98 0.97 0.86 0.83 0.40 0.82 0.62 0.81 0.72 0.98 0.80
Rgs Bank 53 0.84 0.89 0.87 0.94 0.78 0.86 0.95 0.99 0.93 0.31 0.83
Rusfinance Bank 54 0.99 0.92 0.71 0.60 0.49 0.85 0.70 0.17 0.59 0.77 0.68
Credit Europe Bank Ltd 55 0.72 0.98 0.98 0.86 0.50 0.83 0.79 0.27 0.69 0.99 0.76
Globexbank 56 0.69 0.99 0.79 0.88 0.74 1 0.33 1 0.35 0.83 0.76
Asian-Pacific Bank 57 0.92 1 0.72 0.84 0.34 0.64 0.88 0.99 0.80 0.64 0.78
Center-Invest Bank 58 0.97 0.95 0.91 1.00 0.45 0.82 0.99 0.93 0.94 0.93 0.89
Sme Bank 59 0.97 1 0.75 0.73 0.47 0.73 0.51 1 0.88 1.00 0.80
Eximbank Of Russia 60 0.83 1 1.00 0.82 0.76 0.94 0.38 1 0.81 0.64 0.82
Kuban Credit 61 0.78 1 0.93 0.98 0.71 0.83 1 0.95 0.88 0.96 0.90
Baltinvestbank 62 0.87 0.83 0.78 0.94 0.90 0.83 0.83 1 0.70 0.22 0.79
Locko-Bank 63 0.98 1 0.92 0.93 0.39 0.81 0.98 0.69 0.79 0.95 0.84
Hsbc Bank (Rr) 64 0.83 1 0.77 0.95 0.80 0.89 0.89 0.89 1 0.96 0.90
Rn Bank 65 0.75 0.87 0.90 0.94 0.07 0.89 0.65 0.78 0.68 0.71 0.72
Bank Soyuz 66 0.54 0.74 0.52 0.81 0.86 0.80 0.91 1 1 0.99 0.82
Deutsche Bank 67 0.95 0.92 0.91 0.63 0.41 0.89 0.94 0.89 0.89 1 0.84
Metallinvestbank 68 0.84 0.99 0.79 0.93 0.56 0.91 0.88 0.91 0.82 0.97 0.86
Centro Credit Bank 69 0.99 1 1 1 0.73 0.74 1 0.80 0.88 0.71 0.88
Expobank 70 0.65 1 0.83 0.77 0.79 0.76 0.98 0.60 0.75 0.99 0.81
Sdm-Bank 71 1 0.96 0.94 0.97 0.40 0.86 0.97 0.81 0.75 0.98 0.86
Bbr Bank 72 0.99 1 0.90 0.95 0.97 0.79 0.99 0.64 0.93 0.87 0.90
Toyota Bank 73 0.31 1 0.91 0.99 0.54 0.69 1 0.67 0.87 0.98 0.80
Banca Intesa 74 0.75 0.92 0.86 0.98 0.81 0.89 0.88 0.18 0.76 0.53 0.76
Primsotsbank 75 0.62 0.99 0.86 0.62 0.31 0.82 0.96 0.76 0.70 0.69 0.73
Bcs Bank 76 1 0.97 0.80 0.86 0.87 0.95 0.95 0.97 0.86 0.94 0.92
Bnp Paribas Bank 77 0.84 1 0.95 0.96 0.86 0.93 0.25 0.92 0.98 0.60 0.83
Levoberezhny 78 0.66 1 0.95 0.77 0.29 0.74 0.96 0.77 0.75 0.59 0.75
International Financial Club 79 0.95 0.70 0.77 0.95 0.97 0.83 0.55 1 0.91 0.67 0.83
Chelindbank 80 0.97 0.96 0.83 0.98 0.89 0.75 1 0.89 0.91 0.98 0.92
Credit Agricole Cib 81 0.80 0.99 0.76 0.93 0.92 0.93 0.85 1 0.86 0.51 0.85
Chelyabinvestbank 82 0.99 1 0.87 0.99 0.60 0.76 0.99 1 0.96 0.87 0.90
Commerzbank (Eurasija) 83 0.88 1 0.92 0.58 0.96 0.90 0.93 0.88 0.97 1 0.90
Sotsinvestbank 84 0.95 0.92 0.84 0.74 0.62 0.94 0.85 1 0.21 0.33 0.74
Mosuralbank 85 0.93 0.98 0.89 0.93 0.91 0.91 0.95 1 0.85 0.76 0.91
Mean 0.86 0.89 0.83 0.85 0.63 0.84 0.81 0.83 0.78 0.73 0.80
Source: Design and Calculation by Author Using (Excel, Win4deap2 Software). Data Source: Bank of Russia Website.
Jalal H. Abu-Alrop Assecs the Efficiency of Operational Risk Management in Russian Banks
The year 2009 was the best year in the efficiency of operational risk management during the study period, where the average efficiency of banks combined to score 89%, while in 2012 was the worst, the average efficiency of banks combined score was 63%. In 2008, 2010, 2011, 2013, 2014, 2015, 2016 and 2017 the measure of the efficiency of operational risk management for banks combined were score 86%, 83%, 85%, 84%, 81%, 83%, 78%, 73% respectively. The average operational risk efficiency of the combined banks from 2008-2017 indicates that Russian banks could have reduced their inputs (operational risk) by 14%, 11%, 17%, 15%, 37%, 16%, 19%, 17%, 23% and 27% % Respectively.
Efficiency of operational risk management also indicates that the profitability of banks is exactly in parallel with their operational risk — taking preferences in a bank for five years, three banks for four years, three banks for three years, twenty-one banks for two years and thirty four banks for a year. This means that these banks may have had good operational risk management in those years, It also means that these banks were working better than other banks in those years because their degrees of efficiency is equal to (1). On the other hand, there were twenty-three banks that have never achieved the full degree of efficiency (1) over the ten-year period. This means that the profitability of those banks did not reasonably match their operational risk levels as expected. Many banks could have achieved higher returns at the same operational risk levels or could have achieved the same returns at lower risk levels (Table 9).
Table 9 shows the average technical efficiency of operational risk management according to the size of the banks.
Table 9. The Average Technical Efficiency [(DEA) — Input Oriented — (VRS)] of operational Risk Management by Size of Russian Banks (2008 — 2017)
Years Large banks Medium banks Small banks Mean
20СВ 0Б5 0Б4 0Б6
2009 0Б2 059 0.95 0Б9
2010 0.79 0БЗ 0Б6 0БЗ
2011 0Б6 0БЗ 0Б5
2012 0.63 0.60 0.67 0.63
2013 0Б7 0£1 0Б4 0Б4
2014 0.7B 0.79 0Б6 0Б1
2015 0.90 0.73 0Б5 0БЗ
2016 0.B0 0.70 0БЗ 0.7B
2017 0.64 0.75 0.B0 0.73
Mean 0.79 0.7B 0Б4 0.B0
Source: design and calculation by Author using (Excel, Win4deap2 Software). Data Source: Bank of Russia website.
During the ten years, the banks achieved average efficiency of operational risk management as follows: large banks (79%), medium banks (78%) and small banks (84%), In other words, small banks were the most effective in operational risk managing, while large banks were more Table 9 shows the average technical efficiency of operational risk management according to the size of the banks. During the ten years, the banks achieved average efficiency of operational risk management as follows: large banks (79%),
&
nei
k
si
la n
о
e §
100% 95% 90% 85% 80% 75% 70% 65% 60% 55%
Operational Risk Efficiency (DEA)
8 0 0
<N
9 0 0
<N
0
<N
0
<N
0
<N
0
<N
0
<N
0
<N
0
<N
0
<N
Large banks
Years Medium banks
Small banks
Figure 3. The Average Technical Efficiency [(DEA) — Input Oriented — (VRS)] of operational Risk Management by Size of Russian Banks (2008—2017)
Source: design and calculation by Author using (Excel, Win4deap2 Software). Data Source: Bank of Russia website.
Original article Operational bank risk Issues of Risk Analysis, Vol. 17, 2020, No. 2
Table 10. The Ratios of The Contribution of Operational Risk Indicators in The Formation of Performance Indicators, Comparison Between Real Ratios and Ideal Target Ratios (2008—2017)
The Variables The Real Ratios Mean The Target Ratios Mean
Large Banks Medium Banks Small Banks Large Banks Medium Banks Small Banks
OPR 0.436 0.554 0.484 0.491 0.217 0.308 0.198 0.241
NIM 0.039 0.066 0.045 0.050 0.049 0.072 0.052 0.058
ROA -0.004 0.011 0.004 0.003 0.021 0.026 0.023 0.023
ROE -0.055 -0.001 0.048 -0.003 0.095 0.112 0.097 0.101
Source: Design and Calculation by Author Using (Excel, Win4deap2 Software). Data Source: Bank of Russia Website.
NIM (2008-2017) = 5.02%
Tager NIM (2008-2017) = 5.79%
Other veriabels 63%
OPR
37%
Other veriabels 86%
OPR 14%
ROA (2008-2017) = 0.35%
Other veriabels -40%
Tager ROA (2008-2017) = 2.33%
OPR
60%
ROE (2008-2017) = -0.27%
Other veriabels -51%
OPR 49%
Tager ROE (2008-2017) = 10.1%
Other veriabels 43%
OPR 57%
Figure 4. The Ratios of The Contribution of Operational Risk Indicators in The Formation of Performance Indicators, Comparison Between Real Ratios and Ideal Target Ratios (2008—2017)
OPR: Operational Risk., NIM: Net Interest Income. ROA: Return on Assets. ROE: Return on Equity. Source: Design and Calculation by Author Using (Excel, Win4deap2 Software).
Jalal H. Abu-Alrop
Assecs the Efficiency of Operational Risk Management in Russian Banks
medium banks (78%) and small banks (84%), In other words, small banks were the most effective in operational risk managing, while large banks were more efficient than medium banks. The medium banks were the least efficient than other banks in efficiency of operational risk management. Figure 3 also shows that.
Conclusion
This study examines the efficiency of operational risk management of 85 Russian commercial banks During the period 2008—17. This study uses the input-oriented model [data envelopment analysis (DEA) — the variable returns to scale technology (VRS)] with financial ratios to assess the efficiency of operational risk management, also This study uses simple regression analysis to select variables that will enter as inputs and outputs in data envelopment analysis (DEA) approach. The study adopts the basic indicator approach (BIA) approach to measuring operational risk as this approach relies on gross income as an indicator of operational risk. Also, the study adopts net interest margin (MM), return on assets (ROA), and return on equity (ROE) to measuring banks performance. The study divided the banks into three equal major groups based on the size of the assets:
1. Large banks (L): consisted of 28 banks, it included the banks which have total assets between (270 billion rubles to 23 trillion rubles).
2. Medium banks (M): consisted of 29 banks, and included the banks which have total assets of between (102 — 270 billion rubles).
3. Small banks (S): consisted of 28 banks, and included the banks which have total assets of between (5 — 102 billion rubles).
The study found that:
• The impact of operational risk was positive on the performance of Russian banks in most years of study except for 2011, 2012, 2014 and 2017, as it had no effect on some performance indicators. Operational risk contributed to the formation of MM, ROA and ROE performance indicators at 37%, 60% and 49% respectively, reflecting the impact of operational risk on the performance of Russian banks.
• During the study period, 2009 was the best year in the efficiency of operational risk management in Russian banks, where the average efficiency of banks was 89%, While in 2012 was the worst, where the average efficiency of banks in operational risk management was 63%. In 2008,2010,2011,2013, 2014, 2015, 2016 and 2017, the efficiency of operational risk management in Russian banks
was 86%, 83%, 85%, 84%, 81%, 83%, 78% and 73%, respectively. The average efficiency of operational risk management in Russian banks from 2008-2017 indicates that Russian banks could reduce their inputs (operational risk) by 14%, 11%, 17%, 15%, 37%, 16%, 19%, 17 23% and 27%, respectively.
• Operational risk efficiency indicates that banks' profitability is fully consistent with their operational risk preferences in one bank for five years, three banks for four years, three banks for three years and twenty one banks for two years and thirty four banks for one year, meaning that these banks may have risk management It also means that these banks have been working better than other banks in those years because their degree of efficiency is equal to (1). On the other hand, there were 23 banks that had never achieved full proficiency (1) over the study period. This means that the profitability of these banks did not reasonably match operational risk levels as expected. Many banks could have achieved higher returns at the same operational risk levels or could achieve the same returns at lower risk levels. The average technical efficiency of operational risk management by size of banks was as follows: Large banks (79%), medium banks (78%) and small banks (84%), the difference was clear between small banks and other banks. In other words, small banks were the most effective in managing operational risk, while large banks were more efficient than medium banks. Medium banks were less efficient than other banks in the efficiency of operational risk.
The study concluded that:
• Operational risk is an important risk affecting the performance of Russian banks.
• Russian banks could have reduced their inputs (operational risk) by 14%, 11%, 17%, 15%, 37%, 16%, 19%, 17%, 23% and 27% in 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016 and 2017 Respectively that means, Many Russian banks could have achieved higher returns at the same operational risk levels or could have achieved the same returns at lower operational risk levels.
• Small Russian banks were the most effective in operational risk managing, while large banks were more efficient than medium banks.
• The banks' performance is not necessarily parallel to their risk preferences, by comparing the Bank's risk effectiveness with its competitors, it is possible to determine whether the Bank's performance and profitability are reasonable compared to risk levels. Data envelopment analysis (DEA) is an effective measurement tool for such a comparison.
Original article
Operational bank risk Issues of Risk Analysis, Vol. 17, 2020, No. 2
• These results may provide an alert for the inefficient banks to detect and verify their efficiency and compare it with their peers and delve deeper into this subject.
• The banks management should investigate in low profitability compared to other banks because in the long term this may not be sustainable or may result in loss of market shares and damage to the bank's financial health. A high-risk bank should continually review its position either to increase its profitability or to reduce its risk level.
• The risk management approach in standard banks can be seen as an exemplary approach.
References
1. Kristína, Vincová. (2005). Using DEA Models to Measure Efficiency, Biatec, Volume Xiii, 8/2005. Grant Project Vega No.1/1266/04.
2. AH Samad-Khan. (2006). Stress Testing Operational Risk. Opries' Advisory LLC, The International Monetary Fund, Paper presented at the Expert Forum on Advanced Techniques on Stress Testing: Applications for Supervisors, Washington, DC- May 2—3, 2006. www.opriskadvisory. com.
3. Arshinova, T. (2011). The Banking Efficiency Measurement Using the Frontier Analysis Techniques, Journal of Applied Mathematics, 4(3), 165—176.
4. Asror, Nigmonov. (2010). Bank Performance & Efficiency in Uzbekistan, Eurasian Journal of Business & Economics, 3 (5), 1—25.
5. Banker, R., Charnes, A. & Cooper, W. W. (1984). Some models for estimating technical and scale inefficiencies in data envelopment analysis, Management Science, (30): 1078—1092.
6. Basel Committee on Banking Supervision (BCBS). (2017). Basel III: Finalizing post-crisis reforms. Bank for International Settlements Press & Communications CH-4002 Basel, Switzerland. https: //www, bis.org/bcbs/publ/d424.pdf.
7. Basel Committee on Banking Supervision (BCBS). (2016). Consultative Document Standardized Measurement Approach for operational risk. Bank for International Settlements Press & Communications, CH-4002 Basel, Switzerland, https://www.bis.org/bcbs/publ/d355.pdf.
8. Basel Committee on Banking Supervision (BCBS). (2014). Review of the Principles for the Sound Management of Operational Risk. Bank for International Settlements Press & Communications, CH-4002 Basel, Switzerland. https:// www.bis.org/publ/bcbs292.pdf.
9. Basel Committee on Banking Supervision (BCBS). (2011). Principles for the Sound Management of Operational Risk.
Bank for International Settlements Press & Communications, CH-4002 Basel, Switzerland. https://www.bis.org/ publ/bcbsl95.pdf.
10. Basel Committee on Banking Supervision (BCBS). (2011). Operational Risk — Supervisory Guidelines for the Advanced Measurement Approaches. Bank for International Settlements Press & Communications, CH-4002 Basel, Switzerland. https://www.bis.org/publ/bcbsl96.pdf.
11. Basel Committee on Banking Supervision (BCBS). (2006). International Convergence of Capital Measurement and Capital Standards. Bank for International Settlements Press & Communications, CH-4002 Basel, Switzerland, https://www.bis. org/publ/bcbs 128.pdf.
12. Basel Committee on Banking Supervision (BCBS). (2001). Sound Practices for the Management and Supervision of Operational Risk. Bank for International Settlements Press & Communications, CH-4002 Basel, Switzerland. https:// www.bis.org/publ/bcbs86.pdf.
13. Beccalli, E.; Casu, B. & Girardone, C (2006). Efficiency and stock performance in European banking, Journal of Business Finance & Accounting, 33(1-2), 245—262.
14. Begumhan Ozdincer & Cenktan Ozyildirim (2008). The Effects of Diversification on Bank Performance from the Perspective of Risk Return and Cost Efficiency, SSRN Electronic Journal. DOI: 10.2139/ssrn. 1253223. https://www. researchgate.net/publication/228265417.
15. Berger, A, N. & Humphrey, D. B. (1997). Efficiency of Financial Institutions: International Survey & Directions for Future Research, European Journal of Operational Research, 98(2): 175—212.
16. Bikker, J.A. & Bos, J.W.B. (2008J. Bank Performance: A theoretical and empirical framework for the analysis of profitability, competition and efficiency, Routledge International Studies in Money and Banking, Routledge, London & New York, 176 pages.
17. Charnes, A.; Cooper, W. W. & Rhodes, E. (1978). Measuring the efficiency of decision-making units, European Journal of Operational Research, 2: 429—444.
18. Coelli ,T, J. Rao, D, S, P . Christopher, J. Battese, O, G E. (2005). An Introduction to Efficiency and Productivity Analysis. 2nd Ed, Springer. USA. https://www.springer. com/us/book/9780387242651.
19. DeYoung, R, E & J. P. Hughes & C, G, Moon. (2001). Efficient Risk-Taking and Regulatory Covenant Enforcement in a Deregulated Banking Industry. Journal of Economics and Business, 53 (2—3): 255—282. https://doi.org/10.1016/ S0148-6195(00)00044-8
Jalal H. Abu-Alrop
Assecs the Efficiency of Operational Risk Management in Russian Banks
20. Fanchon, P. (2003). Variable Selection for Dynamic Measures Efficiency in the Computer Industry, International Advances in Economic Research, 9(3): 175—188.
21. Farrell, M. J. (1957). The measurement of productive efficiency. Journal of the Royal Statistical Society (Series A), 120(3), 253—281.
22. Fethi, M. D. & Pasiouras, F. (2010). Assessing bank efficiency and performance with operational research and artificial intelligence techniques: a survey, European Journal of Operational Research, 204(2): 189—198.
23. Heffernan, S. 2005. Modern Banking. Chichester: John Wiley & Sons, Ltd. ISBN: 978-0-470-02004-3. 736 Pages. https:// www.wilev.com/en-us/Modern+Banking-p-9780470020043.
24. Hiroshi Morita, Necmi K. Avkiran. (2009). Selecting Inputs and Outputs in Data Envelopment Analysis by Designing Statistical Experiments, Journal of the Operations Research Society of Japan, 52(2), 163—173.
25. Ing, Kristina, Vincova. (2005). Using DEA Models to Measure Efficiency, Biatec, Volume Xiii, 8/2005. Grant Project Vega No. 1/1266/04.
26. Jelena, Titko; Jelena, Stankeviciene & Natalja, Lace. (2014). Measuring Bank Efficiency: DEA Application, Technological & Economic Development of Economy, 20(4), 739—757.
27. Jenkins, L. & Anderson, M. (2003). A multivariate statistical approach to reducing the number of variables in data envelopment analysis, European Journal of Operational Research, 147(1), 51—61.
28. Lei Sun, Tzu-Pu Chang. (2011). A comprehensive analysis of the effects of risk measures on bank efficiency: Evidence from emerging Asian countries. Journal of Banking & Finance, 35(7), 1727—1735.
29. Luo, Y., Bi, G., & Liang, L. (2012). Input/output indicator selection for DEA efficiency evaluation: An empirical study of Chinese commercial banks, Expert Systems with Applications, 39(1), 1118—1123.
30. McAllister, P. H. & McMaus, D. (1993). Resolving the scale efficiency puzzle in banking. Journal of Banking and Finance, 17: 389—405.
31. Nataraja, Niranjan R. & Johnson, Andrew L. (2011). Guidelines for using variable selection techniques in data
envelopment analysis, European Journal of Operational Research, Elsevier, 215(3), 662—669.
32. Paradi, J. С & Zhu, H. (2013). A Survey on Bank Branch Efficiency & Performance Research with Data Envelopment Analysis, Omega, (41)1: 61—79.
33. Qiwei, Xie. Qianzhi, Dai. Yongjun, Li & An Jiang. (2014). Increasing the Discriminatory Power of DEA Using Shannon's Entropy, Entropy, 16, 1571—1585.
34. Ruggiero, J. (2005). Impact Assessment of Input Omission on DEA, International Journal of Information Technology & Decision Making, 04(03): 359—368.
35. Saha, A., Ahmad.; N. H., & Dash, U. (2015). Drivers of Technical Efficiency Inmalaysian Banking: A New Empirical Insight. Asian-Pacific Economic literature, 29(1), 161—173.
36. Singh, G. Singh, P. & Munisamy, S. (2008). A cross country comparison of banking efficiency: Asia Pacific banks, International Review of Business Research Papers, 4(3): 73—95.
37. Subramanyam T. (2016). Selection of Input-Output Variables in Data Envelopment Analysis — Indian Commercial Banks. International Journal of Computer & Mathematical Sciences, 5(6), 51—57.
38. Wheelock, D. С & Wilson, P. (1995). Why do banks disappear: the determinants of bank failures and acquisitions, the Review of Economics and Statistics, 82: 127—138.
39. Yang, Z. (2009). Bank Branch Operating Efficiency: A DEA Approach, The International Multi Conference of Engineers & Computer Scientists (IMECS 2009), 18—20 March 2009, Hong Kong.
40. Zreika, M. & Elkanj, N. (2011). Banking Efficiency in Lebanon: An Empirical Investigation, Journal of Social Sciences, (7) 2, 199—208.
Сведения об авторе
Абу-Алроп Джалал Хэфет: Казанский федеральный университет Российский институт управления, экономики и финансов, Финансы, денежное обращение и кредит Контактная информация:
Адрес: 420008, Республика Татарстан, г. Казань, ул. Кремлевская, д. 18
E-mail: jalalabualrop@gmail.com
Дата поступления: 03.12.2019 Came to edition: 03.12.2019
Дата принятия к публикации: 17.03.2020 Date of acceptance to the publication: 17.03.2020
Дата публикации: 30.04.2020 Date of publication: 30.04.2020
