A Camels Analysis on the Turkish Banking Sector: Rating of the 2004-2011 Period In Terms of Capital Ownership and Scale

Bu bildiri 9’uncu EBES Conference – Rome, Faculty of Economics Sapienza University of Rome konferansında Ocak 2013’te 338-352 sayfa aralığında yayınlanmıştır.

Asst. Professor PhD. E. Savas Basci, Hitit University, Turkey*

PhD. Adalet Hazar, Banking Expert, Turkey

Asst. Professor PhD. Senol Babuscu, Baskent University, Turkey

M. Oguz Koksal, Director of State-owned Bank, Turkey

Abstract

It is essential that banks serve efficiently and productively to increase their financial performance. For this reason, a number of methods are used to evaluate the financial performance of the banks. One of the most commonly used methods is the CAMELS method where C stands for capital adequacy, A stands for asset quality, M stands for management adequacy, E stands for earnings and S stands for sensitivity to market risk.

The main aim of this study is to determine if CAMELS values underwent any differentiation in line with the changes that the Turkish banking sector faced, especially after the 2001 financial crisis. For this purpose, this study briefly examines the practices in different countries and makes econometric analysis using the data gathered from the financial tables of the banks. The information will then be applied to evaluate the existence and nature of any possible differentiation.

In this paper, we analysed a CAMELS method in Turkish banking sector from 2004 to 2011. According to analyse we determined differences between banks which are 3 banks as state-owned banks, 10 banks as privately-owned banks and 8 banks as foreign capital banks. As a result of analyses we obtained a financial performance of separate banks for two dimensions. One dimension represents a successful banking, and others unsuccessful. In this analyse ,we have used Post Hoc Analysis and especially Scheffé’s method.

Key words: early warning, CAMELS

JEL Classifications: C14,E50,G21

1. Introduction

For an internal-level observation and assessment of the general condition and also for the purpose of an external, off-site surveillance of the commercial banks during risk-based supervision and auditing, the CAMELS method, incepted and developed by the regulatory and supervisory authorities in the U.S., is classified and designated as a rating system (Kaya: 2001:1).

Referred to shortly by its acronym of CAMELS, and employed both for measuring the aggregate performance of the banking industry and as a tool for early warning, this methodology was tackled in innumerous studies from a wide range of perspectives. This research paper also targets implicitly to generate converging results through the implementation of the similar methodology.

Taking into account six components, the initial letters of the CAMELS modeling denote: Capital adequacy, Asset quality, Management adequacy, Earnings, Liquidity and Sensitivity to market risk.

At the outset, research studies undertaken in various countries regarding the CAMELS technique are presented in a summarized format, followed by econometric analyses structured on data obtained from the banks’ financial tables, enabling us to verify and ascertain the existence and/or prevalence of divergence, dissimilarity and differentiation, if any, in terms of capital structure and scale, as well as gauging the direction thereof, in case such existence and prevalence are substantiated with plausible evidence.

2. Literature Review

The relevant literature features a plethora of studies accomplished in this subject matter. Compact information on selected academic works are provided here-below.

A research investigation achieved by R. Alton Gilbert, Andrew P. Meyer and Mark D. Vaughan (2000) rated the banks on the basis of the CAMEL modeling and the SEER risk ranking and tiering technique. As per the approach applied to the CAMEL method, the criteria utilized for the banks under coverage encompassed the following key indicators; return-on-assets, the ratio of commercial loans to assets, the ratio of shareholders’’ equity to assets, the ratio of immovable property and real estate to assets, the ratio of outstanding loans overdue by 30-89 days to assets, the ratio of outstanding loans overdue in excess of 90 days to assets, the ratio of unaccrued loans to assets, the ratio of securities to assets, the ratio of deposits of over US$100 million to assets and the ratio of housing loans to assets, while the natural logarithm of the total assets were adopted and a regression test was performed for the underlined aims of the research.

The study conducted by Harsh Vineet Kaur (2010), on the other hand, ranked and tiered 28 public, 26 private and 28 foreign banks by means of the CAMEL analysis technique. Rating and the correlated evaluation were implemented by selecting two ratios each for the capital adequacy, asset quality, management quality, earnings status and liquidity groupings – all forming the CAMEL constituents.

The departure-point of another exploratory study penned by Gary Whalen (2010) essentially endeavored to deduce answers to two basic questions. The foremost inquiry topic entailed whether there was a discernible decline in the propensity of the early warning models to engender accurate results and supply precise findings recently, in congruity with the acceleration of volatility in the banks. The second theme pertained to whether there is pressuring necessity to revisit and review the sculpts employed in the prediction and forecasting of bank risks more frequently and re-structure new estimations. For the stated purpose, the CAMELS model was applied, pillared on the annual year-end data covering the 1997-2002 period. A supplementary forecasting and prediction effort was also attempted for the ensuing and subsequent five-year term. At the end of the research, we were able to infer the conclusion that there surfaced no distinctly-manifested decrease in the forecasting and prediction accuracy and also there was no requirement to test the model more often than currently enforced.

Furthermore, Uyen Dang (2011) focuses on the adoption of the CAMEL system as relevant to the banking supervision and audit. In the model, scores are ascribed to the specified five bundles within the framework of the pre-designated ratios, enabling the procurement of a general score by means of taking the average of the compounded results of the points for the sub-groups. An overall rating is obtained by overlapping the scores assigned to the banks with their corresponding, pre-designated definitions.

In its investigative pursuit, the study furthered by Wirnkar Alphonsius Dzeawuni and Dr. Muhammad Tanko (2008) processed the data, facts and figures of 11 commercial banks located in Nigeria, recorded over the 1997-2007 period. The study used the purposive sampling technique. Efficiency Measurement System (ES) 1.30 Software of Holger School and the independent T-test equality were put to use in the analysis and testing of the data. This venture produced findings on the impact of each factor in the CAMEL on the aggregate, total bank performance. Moreover, this outcome also enabled the identification of the most significant dynamics and paradigms that influenced the CAMEL.

Likewise, dwelling on a similar realm in their joint research work, Suvita Jha and Xiaofeng Hui (2012) appraised the financial performances of 18 commercial banks in Nepal with ownership portraying diverse and different types, stretching over a period between the years 2005-2010. Doubtless to mention, the CAMEL model was attached priority in the rating venture, cultivating a cluster of determinants. More to the point, the multivariate regression analysis was brought into the foreground to facilitate the forecast and estimation of the impact of primary yardsticks such as the capital adequacy ratio, the non-performing loans ratio, the ratio of the interest expenditures to the loans, the net interest margin ratio and the ratio illustrating the conversion of deposits into loans on the return-on-assets and the return-on-shareholders’ equity. The findings of the analysis demonstrated clearly that the levels of efficiency registered by the privately-owned local banks and the foreign banking entities were in close proximity to each other, while the performances of the public banks emerged at a lower par. What is more noteworthy, it became irrefutably apparent that the return-on-assets was predominantly influenced by the capital adequacy ratio, the ratio of the interest expenditures to loans and the net interest margin, while the capital adequacy ratio was most effective on the return-on-shareholders’ equity.

Shifting to a different domain of the industry in their scientific work, Rehana Kouser and Irum Saba (2012) provided a comparison of the commercial banks, mixed and hybrid banks and the Islamic-oriented banking institutions operating in Pakistan, again relying on the CAMEL method. Initially, the CAMEL-relevant ratios were defined in a bid to expose the divergences, dissimilarities and differences, as ANOVA was picked as an instrument for probing purposes. Data analysis was tested through resort to the SPSS. Among the prominent inferences of the study was the re-affirmation that the Islamic banks were in a more viable standing and sound status among the peer group under consideration, in pivotal areas such capital adequacy and asset quality.
3. Significance of Performance Rating in Banking and Applicable Methods

The operations of the banking sector with an accelerated efficiency is crucially important and vital for it to fulfill its underlying mission of performing an effective intermediary function for supporting the conducive environment and fundamental infrastructure warranted for a sustainable growth pattern. While competition will eventually assist a deceleration in the intermediation costs against a backdrop defined by an effectively-operating banking industry, the upshot will invariably enhance transparency and secure the steady fluidity of information (Mercan: 2008:103).

As already accentuated above, a variety of analyses are generated, incepted and designed to rate and assess the performance records of the banks and capture the early-warning signals emitted from this critical industry. Such researches are defined as the ratio analysis and the parametric and non-parametric approaches.
3.1 Ratio Analysis

The elementary objective of the ratio analysis is to conceptualize and envision the status of a bank in the future through utilization and processing of a bank’s current and past financial information. Elucidated in plain terminology, the purpose of the ratio analysis is to measure and rate a bank’s capital adequacy, liquidity, asset quality and profitability by means of constructing a meaningful and significant relationship among the relevant columns of the balance-sheet and the income statement (Gökmen: 2007:51).
3.2 Parametric Methods

While rating the efficiency magnitude by means of the parametric approaches, forecasting and prediction are conducted through the regression techniques, and, glancing at the definition of the production function, several inputs are correlated with only one output.

In the parametric methods, there are commonly a set of observations, and assuming that the best performance among this cluster takes place on the regression line (effective boundary), observations that do not denote deviation and divergence from this linear pattern is considered as effective, while those that fail insofar as the observations are concerned are deemed as lacking effectivity (Yardımcı: 2006:7).

Regression analysis is counted among the most frequently-employed measurements methods retrieved for the rating and measurement of the effectivity, providing also the setting for the methodological architecture sculpted to facilitate the determination of the causal structure between the dependent and independent variables, known to harbor a cause-and-effect inter-connectedness. The existence of a causal relationship between the independent (discloser) and the dependent (disclosed) variables, regarded from a theoretical framework, as well as the prior knowledge of the functional structure of the relationship between the variables portray significant importance and consequential value for the regression testing. Performance rating is conducted through regression analysis, patterned on the regression line. Judged units hovering above the regression line are accepted as relatively productive, while the decision-units submerging below the line are dubbed as unproductive. Through the residual obtained from the regression outputs, the targeted intention is to reflect the relative technical productivity. Negative residuals denote the unproductive decision-units, while the positive residuals signify the unproductivity angle (Beycan: 2007:79).
3.3 Non-Parametric Methods

Drawing on techniques tracing their origins to the linear-programming (i.e. optimization under constraints), non-parametric methods attempt to rate and assess the distance and proximity to the boundary of effectivity. Since these approaches are not compelled to engage in behavioral assumptions pertaining to the structure of the production unit, as also witnessed in the parametric models, they are equipped with more advantages, certainly purely in relative terms. Besides, the mentioned methods harbor superiority epitomized in their resilience to employ more than one explanatory and explained variable. By contrast, since they do not incorporate any random error term, they are inclined to transmit the data and measurement flaws, and any and all other defects and shortcomings precipitated by chance or other detrimental exogenous factors directly to the model and thus may also misrepresent and misplace the boundary of effectivity (Demirbaş, Sezgin, 2010, p. 146).

4. Research Methodology

Within the context of the ratio analysis chosen for rating and measuring the banking sector’s performance, the CAMEL technique was conceived purposefully to reflect the banks’ financial standing, the extent of their compliance to the legislations and regulatory framework, their management quality and the status of their internal audit and supervisory architecture. Adopted generally for on-site supervision, this system emerged among the most crucial objectives in the U.S. for off-site supervision and auditing (Kosova, 2005:49).

The CAMEL approach has been sanctioned and acknowledged as the most influential internal supervisory implement employed extensively for rating the sensitivity of the financial institutions in the U.S. (Dang, 2011:16).

The five CAMEL factors — capital adequacy, asset quality, management, earnings and profitability, and liquidity — are increasingly gaining importance in stymieing and aborting bank failures and insolvencies as well as functioning as an asset of inspiration and guidance for their self-development and further sophistication. The method embodies the concept that each of the five groups actually signifies a crucial aspect of the banks’ financial tables (Kouser, Rehana, Saba, Irum, 2012:72).

Incepted and moulded as a technique to analyze and examine the financial structures of the banks within the context of specific criteria and principles, the CAMEL is an acronym, derived from the English-language capital letters of the five norms that possess exclusive importance (Babuşçu, 1997:81).

C Capital Adequacy

A Asset Quality

M Management

E Earnings

L Liquidity

Originally, the five basic criteria furnished here-above were accepted as key indicators of the banks’ financial performance, overall financial standing and condition, their safety and soundness in terms of operations and compliance with the regulatory and supervisory arrangements and stipulations. The technique began to be designated as CAMELS with the addition of the “Sensitivity to market risk” in 1996 as the sixth criterion. In application, the CAMELS technique offers rating and evaluation of the six criteria pertaining to the bank subject to examination and supervision on the scale of one (the best) to five (the worst) grading scores in ascending order (Mercan: 2008:116).

According to this, the ranking scales allude to the following:

“1” safe and sound banks in all aspects (scores for each of the bank’s CAMEL components are surmised to be 1 or 2);

“2” safe and sound banks, in a general satisfactory sense (scores for each of the components are surmised not to be worse than 3);

“3” flaws exist in connection with the bank’s performance and insufficiently resilient against exogenous shocks and vulnerable to risks;

“4” banks generally confronted with serious problems and encumbrances, displaying poor scaffolding or underperformance and afflicted by financial/managerial distortion and deterioration, and

“5” banks exposed to excessive risk, and distressed by grave financial/managerial predicaments (Kaya, 2001:1).

4.1 Capital Adequacy

From the perspective of a banking enterprise, the first and foremost function of the core capital is to cover any and all expenditures warranted for incorporation and launch of operations and also it constitutes a pool of funds for the intangible, fixed assets (Gökmen: 2007:60).

Capital adequacy is the fundamental determinant to ensure and rate the banks’ soundness and overall health. This yardstick also illustrates the capability of a bank’s shareholders’ equity to hedge and protect itself against shocks (Jha, Suvita and Hui Xiaofeng, 2012:7603).

When rating the capital adequacy of the banks, ratios are utilized to assess the banks’ equity resources in regard to quantity and quality. The factors considered by the supervisory and regulatory authorities during on-site supervision of the banks’ capital adequacy in the developed countries are provided below (Beycan: 2007:90):

– Rating of the core capital level and its quality by taking into account the bank’s overall financial standing as well as the bank’s size,

– Status of the resource and liquidity availability when and if confronted with the necessity for an emergency and supplementary capital injection,

– Status of the troubled and non-performing assets and whether adequate provisions have been set aside,

– Segregation and unbundling of the asset structure of the balance-sheet to incorporate also the risks,

– Rating of the risks created by the off-balance sheet columns,

– Level of profitability,

– Bank’s growth strategy and forward-looking plans and projections,

– Level of undistributed and retained profit, and

– Opportunities of access to the capital markets and other capital resources.

4.2 Asset Quality

Aspects such as the types of the bank’s assets, whether they are revenue-generating and, if indeed generating revenues, the ‘power of revenue-generation’ and the prevalence and persistence of revenue-generation capability are factors that bear significance for the rating and evaluation of the asset quality (Gökmen: 2007:76).

The most important asset group is the loan portfolio, as the largest risk posed to the banks is the non-performance of loans extended (Dang, 2011:19).

The factors considered by the supervisory and regulatory authorities during on-site supervision of the banks’ asset quality in the developed countries are provided below:

– Effectivity/eligibility/suitability of the entire lending process, the terms and conditions thereof, whether risk assessment is performed and the receipt of appropriate collateral and guarantees,

– Determination and monitoring of the non-performing, re-structured, reorganized and suspended loans subject to due administrative/legal action for collection and the level of success accomplished in such strides,

– Sufficient provisioning to cover loans, past-overdue credits and non-accruals, and provisioning against exposure to contingent asset losses, damages and claims,

– Exposure to and level of credit risk, and rating of the collaterals and guarantees, as well as the derivative transactions and margin-trading limits,

– Analysis of asset concentration,

– Level of success in the collection of non-performing assets and receivables, and

– Status of internal supervision and audit and the mainframe information systems (Beycan: 2007:91).

4.3 Management

Success of the executive board of directors and the upper management of the bank to identify, assess and control the perceived risks posed to the operations, pursuance and fulfillment of prudent overall strategies and business plans, their adaptation to the developments, new products and services in the sector, adequacy of the internal supervision and auditing system, level of determining the appropriate policies, the structure of the bank’s management, and information and risk management systems and a sound architecture and level of proficiency in regard to operations performed all contribute to, solidify and enhance the management quality and add value to operational performance (Ak: 2006:15).

$14.4 Earnings

The bank’s earnings and profitability level and capability to generate earnings and profitability and their sustainability signify crucial importance insofar as the main pillars of the financial structure are concerned.

The criteria considered by the supervisory and regulatory authorities during on-site supervision of the banks’ asset quality in the developed countries are provided below:

– Rating of the earnings base and revenues in regard to their overall trends and their prevalence and persistence,

– Status of the undistributed and retained earnings and profit, and verification whether adequate capital resources are generated through this channel,

– Sources of the earnings and revenues and their quality,

– Analysis of the budgeting systems and the management information mainframe systems,

– Analysis of the policy for provisioning and evaluation, and

– Sensitivity of the earnings and profit to the market risk (Beycan: 2007:94).

4.5 Liquidity

Probably the most prominent among the potent downside risk-drives intimidating the banking sector is the liquidity risk. Liquidity connotes the bank’s power and capability of meeting its cash requirements and obligations, derived from both the asset and liabilities sides of the balance sheet (Tulgar: 1993:40). Most often, the banks are inclined to maintain and utilize their assets as instruments of funding, which could not be steadfastly disinvested in the market when and if necessitated. As a consequence, maturity mismatch and thereby a liquidity crunch risk are manifested. Misperceptions and adverse reputation that could potentially circulate in the market about the bank, attributable to whatsoever reason and pretext, usually prompt and trigger a depositor run on the bank. However, since the banks are tempted to park the funds that they have collected with a short-term maturity in long-term assets, they prove unable to satisfy all of the demand for such reclaimed funds from the deposit-holders. The level of inadequacy of the liquid and readily available assets coerce the banks to fail in responding to the withdrawal demands of the deposit-owners or precipitate a quandary under which they are able to offset such claims only through high-cost fresh funding or roll-over. Clearly, resort to such a vicious circle would unleash and inflict a momentous damage on the bank’s financial constitution and composure and gradually paves the way for failure or collapse. In situations where deposit-guarantee is not available, the perception relating to a bank’s insufficient liquidity, formed among the ranks of the deposit-owners, would also instigate a panic among the deposit-holders of other banks and prod them to immediately rush to their banks for the purpose of withdrawing their funds (Suadiye: 2006:15).

4.6 Sensitivity to Market Risk

Under this category, rating is applied on the bank’s exposure and vulnerability to the interest rate patterns, exchange-rate volatility and fluctuations in the stock prices. For this purpose, rating and measurement are implemented on the sensitivity of the bank’s shareholders’ equity and the capital base to any and all changes in the earnings or variables such as the market interest rates, depending on the occurrence of such changes (Sarker, Abdul Awwal, 2008:12).

The basic criteria taken into account when conducting an on-site supervision and audit are furnished below:

– Sensitivity of the bank’s earnings and the capital value to the adverse changes in the market environment and circumstances,

– Bank management’s success to perceive, assess and control the bank’s exposure and vulnerability to market risk,

– Characteristics of the bank’s exposure to interest-rate risk in non-commercial transactions, and

– Status of the bank’s exposure to market risk emanating from the commercial and foreign-exchange transactions (Sakarya: 2010:16).

5. CAMELS Analysis in Terms of Scale and Capital Structure

The study covers public, resident-private and foreign capital-owned deposit banks (a total of 21 banks) operating in the Turkish banking sector. The grouping breakdown for the 21 deposit-taking banks is three for the public-sector, 10 for the privately-held and eight for banks featuring foreign capital control. At the same time, the scale distribution is as follows: seven large-scale, six medium-scale and eight small-scale. Forming the foundation-stone of the analysis, CAMELS indicators are composed of four ratios for capital adequacy, three for asset quality, five for management quality, four for earnings status, four for liquidity and four for sensitivity to market risks. The data set used in the research were derived and compiled from the internet web-site of the Banks’ Association.

5.1. Data Set

Utilized predominantly for on-site supervision and audit of the banking sector, the CAMELS rating system has evolved into one of the most effective and significant instruments of off-site supervision, surveillance and monitoring in the U.S. Implemented as a functional tool of both on-site supervision and off-site surveillance and monitoring to ensure that the banks operate safely, soundly and regularly in compliance with the regulations and to ascertain that they are on the right track, this approach is also defined as the “composite performance rating and evaluation” (Kaya: 2001:i).

The analysis was performed for a sample of twenty-one banks operating in Turkey of which three were public sector banks, ten were private sector banks and eight were foreign-private sector banks from 2004 to 2011. Our study covered financial ratio obtained from annual balance sheets and income states. The variables used in our study were based on CAMELS framework as below:

Table 1: Symbols and Definitions Used in Analysis

Symbols
Definitions
C1
Shareholders’ Equity / (Amount Subject to Credit Risk + Market Risk + Operational Risk)
C2
Shareholders’ Equity / Total Assets
C3
(Shareholders’ Equity-Permanent Assets) / Total Assets
C4
N(on+off) Balance-sheet Position / Total Shareholders’ Equity
A1
Total Loans and Receivables / Total Assets
A2
Total Loans and Receivables / Total Deposits
A3
Loans under follow-up (gross) / Total Loans and Receivables
M1
Total Assets / No. of Branches
M2
Total Deposits / No. of Branches
M3
Total Loans and Receivables / No. of Branches
M4
Total Employees / No. of Branches (person)
M5
Net Income / No. of Branches
E1
Net Profit (Losses) / Total Assets
E2
Net Profit (Losses) / Total Shareholders’ Equity
E3
Interest Income / Interest Expense
E4
Non-Interest Income (Net) / Other Operating Expenses
L1
Liquid Assets / Total Assets
L2
Liquid Assets / Short-term Liabilities
L3
Liquid Assets / (Deposits + Non-Deposit Funds)
L4
Permanent Assets / Total Assets
S1
Financial Assets (Net) / Total Assets
S2
FC Assets / FC Liabilities
S3
Net Interest Income After Specific Provisions / Total Assets
S4
On Balance-sheet FC Position / Shareholders’ Equity

In order to calculate the CAMELS ratings for the banks, the ratios corresponding to each CAMELS factor were considered: Capital Adequacy, Asset Quality, Management Soundness, Earnings And Profitability, Liquidity, Sensitivity To Market Risk. Lastgroup of analysisis overall CAMELS Ratings. The CAMELS rating was obtained as the total of the individual variable ratings.

All of variables used in analysis were normalized using a formula given below (Dash and Das, 2009).

Where u represent the upper bound in descriptive statistics of its year and l is lower bound in descriptive statistics of its year. Result of this Formula ratings were assigned as follows:

Table 2: Camels Ratings

CAMELS Ratings
Range of Calculated Z
1
0.0 – 0.2
2
0.3 – 0.4
3
0.5 – 0.6
4
0.7 – 0.8
5
0.9 – 10.0

5.2. Findings and Results in Terms of Capital Ownership

According to our model, first of all we analyzed descriptive statistics as shown below.

Table 3: Descriptive Statistics

N
Minimum
Maximum
Mean
Std. Deviation
C
168
8
20
18,55
2,321
A
168
5
15
13,97
2,203
M
168
6
25
23,18
3,647
E
168
5
20
19,01
2,659
L
168
8
20
18,92
2,836
S
168
12
20
18,63
2,067
Valid N (listwise)
168

Each group has a 168 observation in model. Highest standart deviation were M group which are defined management skills of banking. As a ANOVA result of this study with compared group as shown ANOVA Table.

Table 4: Results of ANOVA in Terms of Capital Ownership

Sum of Squares
df
Mean Square
F
Sig.
C
Between Groups
4,463
2
2,231
,411
,663
Within Groups
895,055
165
5,425

Total
899,518
167

A
Between Groups
71,254
2
35,627
7,948
,001
Within Groups
739,597
165
4,482

Total
810,851
167

M
Between Groups
62,537
2
31,269
2,390
,095
Within Groups
2158,743
165
13,083

Total
2221,280
167

E
Between Groups
122,301
2
61,151
9,530
,000
Within Groups
1058,693
165
6,416

Total
1180,994
167

L
Between Groups
17,966
2
8,983
1,119
,329
Within Groups
1324,868
165
8,030

Total
1342,833
167

S
Between Groups
37,679
2
18,840
4,600
,011
Within Groups
675,696
165
4,095

Total
713,375
167

Result of ANOVA analyze, compare means between group has statistically significant and reject H0 hypothesis which are A, M, E, S. On the other way C and L were limited with legal rules by government. We expect that results because of accept H0 for C and L ANOVA Results. We know that C and L groups which were inside management and liquidity ratios for all banks has not a difference with other.

According to result of ANOVA, we need subtest for determining of some question. For purpose, which banks type are sufficient in CAMELS ratings in overall banking system. Data are nonparametric and banks counts insufficient. For detailed analyze for Nonparametric test in Scheffe test results are are shown below.

Table5: Multiple Comparisons Scheffe Test Results in Terms of Capital Ownership

Dependent Variable
(I) Type
(J) Type
Mean Difference
(I-J)
Std. Error
Sig.
C
Public Banks
Private Banks
,404
,542
,758

Foreign-owned Banks
,120
,557
,977

Private Banks
Public Banks
-,404
,542
,758

Foreign-owned Banks
-,284
,391
,768

Foreign-owned Banks
Public Banks
-,120
,557
,977

Private Banks
,284
,391
,768
A
Public Banks
Private Banks
-1,863(*)
,493
,001

Foreign-owned Banks
-1,859(*)
,507
,002

Private Banks
Public Banks
1,863(*)
,493
,001

Foreign-owned Banks
,003
,355
1,000

Foreign-owned Banks
Public Banks
1,859(*)
,507
,002

Private Banks
-,003
,355
1,000
M
Public Banks
Private Banks
1,033
,842
,472

Foreign-owned Banks
-,245
,866
,961

Private Banks
Public Banks
-1,033
,842
,472

Foreign-owned Banks
-1,278
,607
,112

Foreign-owned Banks
Public Banks
,245
,866
,961

Private Banks
1,278
,607
,112
E
Public Banks
Private Banks
,208
,590
,939

Foreign-owned Banks
1,911(*)
,606
,008

Private Banks
Public Banks
-,208
,590
,939

Foreign-owned Banks
1,703(*)
,425
,000

Foreign-owned Banks
Public Banks
-1,911(*)
,606
,008

Private Banks
-1,703(*)
,425
,000
L
Public Banks
Private Banks
-,158
,659
,972

Foreign-owned Banks
-,786
,678
,512

Private Banks
Public Banks
,158
,659
,972

Foreign-owned Banks
-,628
,475
,419

Foreign-owned Banks
Public Banks
,786
,678
,512

Private Banks
,628
,475
,419
S
Public Banks
Private Banks
-,258
,471
,860

Foreign-owned Banks
,760
,484
,294

Private Banks
Public Banks
,258
,471
,860

Foreign-owned Banks
1,019(*)
,339
,012

Foreign-owned Banks
Public Banks
-,760
,484
,294

Private Banks
-1,019(*)
,339
,012

* The mean difference is significant at the .05 level.

When each group compared to the other groups one by one in terms of CAMELS components, details of the components that create significant difference is as follows:

In terms of asset quality: There is a significant separation between the state-owned banks and the other groups.There is no separation between private banks and foreign banks.

In terms of profitability: There is a significant separation between the foreign banks and the other groups. There is no separation between private banks and the state-owned banks.

In terms of sensitivity to market risk: There is a significant separation between the foreign banks and private banks. There is no separation between state-owned banks and other groups.

5.3. Findings and Results in Terms of Scale

Deposit banks in the scope of analysis divided into 3 groups including large, medium and small banks.

Table 6: Groups by Scala

Name of Banks
Scala
Türkiye Cumhuriyeti Ziraat Bankası A.Ş.
Large Bank
Türkiye Halk Bankası A.Ş.
Large Bank
Türkiye Vakıflar Bankası T.A.O.
Large Bank
Akbank T.A.Ş.
Large Bank
Türkiye Garanti Bankası A.Ş.
Large Bank
Türkiye İş Bankası A.Ş.
Large Bank
Yapı ve Kredi Bankası A.Ş.
Large Bank
Alternatif Bank A.Ş.
Small Bank
Anadolubank A.Ş.
Small Bank
Tekstil Bankası A.Ş.
Small Bank
Turkish Bank A.Ş.
Small Bank
Citibank A.Ş.
Small Bank
Eurobank Tekfen A.Ş.
Small Bank
Fibabanka A.Ş.
Small Bank
Turkland Bank A.Ş.
Small Bank
Şekerbank T.A.Ş.
Medium Bank
Türk Ekonomi Bankası A.Ş.
Medium Bank
Denizbank A.Ş.
Medium Bank
Finans Bank A.Ş.
Medium Bank
HSBC Bank A.Ş.
Medium Bank
ING Bank A.Ş.
Medium Bank

A result of ANOVA test about whether CAMELS components in terms of scale has a significant separation is as follows:

Table 7: Results in Terms of Scale

Sum of Squares
df
Mean Square
F
Sig.
C
Between Groups
3,673
2
1,836
,338
,714
Within Groups
895,845
165
5,429

Total
899,518
167

A
Between Groups
36,574
2
18,287
3,897
,022
Within Groups
774,277
165
4,693

Total
810,851
167

M
Between Groups
88,602
2
44,301
3,427
,035
Within Groups
2132,678
165
12,925

Total
2221,280
167

E
Between Groups
121,611
2
60,805
9,471
,000
Within Groups
1059,383
165
6,421

Total
1180,994
167

L
Between Groups
68,432
2
34,216
4,430
,013
Within Groups
1274,401
165
7,724

Total
1342,833
167

S
Between Groups
77,992
2
38,996
10,127
,000
Within Groups
635,383
165
3,851

Total
713,375
167

A, M, L groups are significant at 5% level, and E, S groups are significant at 1% level. C is not significant.

Secondly, significant seperation were tested each of the scale for each of CAMELS components separately. Test results are as follows:

Table 8: Scheffe Test in Terms of Scale

Dependent Variable
(I) Scala
(J) Scala
Mean Difference (I-J)
Std. Error
Sig.
C
Small Banks
Medium Banks
,365
,445
,715

Large Banks
,183
,426
,912

Medium Banks
Small Banks
-,365
,445
,715

Large Banks
-,182
,458
,925

Large Banks
Small Banks
-,183
,426
,912

Medium Banks
,182
,458
,925
A
Small Banks
Medium Banks
-1,125(*)
,414
,027

Large Banks
-,259
,396
,808

Medium Banks
Small Banks
1,125(*)
,414
,027

Large Banks
,866
,426
,130

Large Banks
Small Banks
,259
,396
,808

Medium Banks
-,866
,426
,130
M
Small Banks
Medium Banks
-,307
,686
,905

Large Banks
-1,650(*)
,658
,046

Medium Banks
Small Banks
,307
,686
,905

Large Banks
-1,342
,707
,168

Large Banks
Small Banks
1,650(*)
,658
,046

Medium Banks
1,342
,707
,168
E
Small Banks
Medium Banks
-1,786(*)
,484
,001

Large Banks
-1,721(*)
,464
,001

Medium Banks
Small Banks
1,786(*)
,484
,001

Large Banks
,065
,498
,991

Large Banks
Small Banks
1,721(*)
,464
,001

Medium Banks
-,065
,498
,991
L
Small Banks
Medium Banks
-1,495(*)
,531
,021

Large Banks
-,203
,509
,923

Medium Banks
Small Banks
1,495(*)
,531
,021

Large Banks
1,292
,547
,064

Large Banks
Small Banks
,203
,509
,923

Medium Banks
-1,292
,547
,064
S
Small Banks
Medium Banks
-1,620(*)
,375
,000

Large Banks
-1,096(*)
,359
,011

Medium Banks
Small Banks
1,620(*)
,375
,000

Large Banks
,524
,386
,400

Large Banks
Small Banks
1,096(*)
,359
,011

Medium Banks
-,524
,386
,400

* The mean difference is significant at the .05 level.

In terms of asset quality: There is a significant separation between small banks and medium banks.

In terms of management: : There is a significant separationbetween small banks and large banks.

In terms of profitability: There is a significant separation among all scales.

In terms of Liquidity: There is a significant separation between small banks and medium banks.

In terms of sensitivity to market risk: There is a significant separation among all scales.

6. Conclusions

The aggregated CAMELS components of groups obtained from grouping of banks in terms of capital ownership shows significant difference in terms of A, M, E, S.When each group analyzed for the each component,the only indication seperated all the groups was active quality. Test results are as follows:

Table 9: CAMELS ComponentsResult Matrix in Terms of Capital Ownership

C
Public Banks
Private Banks
Foreign-owned Banks
Public Banks

x
x
Private Banks
x

x
Foreign-owned Banks
x
x

A
Public Banks
Private Banks
Foreign-owned Banks
Public Banks



Private Banks



Foreign-owned Banks



M
Public Banks
Private Banks
Foreign-owned Banks
Public Banks

x
x
Private Banks
x

x
Foreign-owned Banks
x
x

E
Public Banks
Private Banks
Foreign-owned Banks
Public Banks

x

Private Banks
x

x
Foreign-owned Banks

x

L
Public Banks
Private Banks
Foreign-owned Banks
Public Banks

x
x
Private Banks
x

x
Foreign-owned Banks
x
x

S
Public Banks
Private Banks
Foreign-owned Banks
Public Banks

x
x
Private Banks
x


Foreign-owned Banks
x

X= No significant difference.

√= There is a significant difference.

Therefore, it is predictable that capital ownership is important for forming of the asset composition. When an evaluation is made in terms of scale, the aggregated CAMELS components of groups shows significant difference for A, M, E, L, S. When CAMELS components is analyzed one by one, the seperation points to in terms of profitability and sensitivity to market risk indicators of all groups.

Table 10: CAMELS Components Result Matrix in Terms of Scale

C
Small Banks
Medium Banks
Large Banks
Small Banks

x
x
Medium Banks
x

x
Large Banks
x
x

A
Small Banks
Medium Banks
Large Banks
Small Banks


x
Medium Banks


x
Large Banks
x
x

M
Small Banks
Medium Banks
Large Banks
Small Banks

x

Medium Banks
x

x
Large Banks

x

E
Small Banks
Medium Banks
Large Banks
Small Banks



Medium Banks



Large Banks



L
Small Banks
Medium Banks
Large Banks
Small Banks


x
Medium Banks


x
Large Banks
x
x

S
Small Banks
Medium Banks
Large Banks
Small Banks



Medium Banks



Large Banks


X= No significant difference.

√= There is a significant difference.

In terms of scale, the significant seperation points to sensitivity to profitability and market risk which is two closest indicator for all groups. When performance analysis or early warning analysis is made, it is referred that the detail analysis should be done in terms of indicators derived from significant seperation findings taking into account capital ownership and scales of banks

References

(English translations are provided solely for information, elaboration and explanatory purposes and should not be construed necessarily as an authentic replication or representation of the original versions in Turkish)

Ak, Sebla (2006), “Ticari Bankalarda Fon Yönetimi” (“Fund Management in Commercial Banks”), Gazi Üniversitesi Sosyal Bilimler Enstitüsü İktisat Ana Bilim Dalı Yüksek Lisans Tezi (“Gazi University, Social Sciences Institute, Economics Main Department, Master’s Degree Thesis”).

Babuşcu, Şenol (1997), “Bankacılıkta Risk Derecelendirmesi (Rating) ve Türk Bankacılık Sektörüne Uygulanması” (“Risk Rating in Banking and Its Application to the Turkish Banking Sector”), Sermaye Piyasası Kurulu Yayını (“Capital Market Board Publication”), No: 94, Ankara.

Beycan, Mehmet (2007), “Bankalarda Performans Değerlemesi ve Bir Uygulama” (“Performance Rating in Banks and An Application”), Dokuz Eylül Üniversitesi Sosyal Bilimler Enstitüsü İşletme Anabilim Dalı Yüksek Lisans Tezi, İzmir (“Dokuz Eylül University, Social Sciences Institute, Business Administration Main Department, Master’s Degree Thesis, Izmir”)

Dang, Uyen (2011),” “The Camel Rating System in Banking Supervision: A Case Study,” Arcada University of Applied Sciences, International Business, Degree Thesis.

Dash, Mihir and Das, Annyesha, A CAMELS Analysis of the Indian Banking Industry (July 14, 2009). Available at SSRN: http://ssrn.com/abstract=1666900 or http://dx.doi.org/10.2139/ssrn.1666900

Demirbaş, Mahmut & Sezgin, Funda H. (2010), “Likidite Krizi Sürecinde Amerika Birleşik Devletleri, Avrupa Birliği’ne Üye Ülkeler ve Türkiye’deki Bankacılık Sektörünün Karşılaştırmalı Etkinlik Analizi : 2006-2010 Dönemi” (“An Efficiency Analysis of the Banking Sectors in the United States of America, Member Countries of the European Union and Turkey During the Era of Liquidity Crisis: 2006-2010 Period”), Gazi University, Journal of Economic and Administrative Sciences Faculty 12/3, pp. 135-158.

Doğan, Barış (2008), Bankaların Gözetiminde Bir Araç Olarak Kümeleme Analizi: Türk Bankacılık Sektörü İçin Bir Uygulama (“Cluster Analysis As A Tool in Banking Surveillance: An Application for the Turkish Banking Sector”), Kadir Has Üniversitesi Sosyal Bilimler Enstitüsü Finans ve Bankacılık Bilim Dalı Doktora Tezi, İstanbul (“Kadir Has University, Social Sciences Institute, Finance and Banking Department, Doctoral Dissertation, Istanbul”).

Dzeawuni, Wirnkar Alphonsius, Tanko, Muhammad (2008), “CAMELs and Banks’ Performance Evaluation: The Way Forward,” Social Science Research Network, June 24.

Erdoğmuş, Buket (2010), Bankalarda Mali Başarısızlıkların Önceden Tespitinde Erken Uyarı Sistemi ve Bir Uygulama (“Early Warning System in Detecting Financial Failures in Banks and An Application”), Ankara Üniversitesi Fen Bilimleri Enstitüsü Yüksek Lisans Tezi, Ankara (“Ankara University, Physical Sciences Institute, Master’s Degree Thesis, Ankara”).

Ertuğrul, Ahmet, Osman Zaim (1996) “Türk Bankacılığında Etkinlik: Tarihi Gelişim Kantitatif Analiz” (“Efficiency in Turkish Banking: Historical Progress-A Quantitative Analysis”), Bilkamat İşletme ve Finans Yayınları (“Bilkamat Business Administration and Finance Publications”), No.:3, pp: 3-70.

Gilbert, R. Alton, Meyer, Andrew P. and Vaughan, Mark D. (2000),”The Role of a CAMEL Downgrade Model in Bank Surveillance,” Federal Reserve Bank of St. Louis, Research Division, Working Paper.

Gökmen, Berrak (2007), Bankalarda Finansal Tablolar Analizi (“Analysis of Financial Tables in Banks”), İstanbul Üniversitesi Sosyal Bilimler Enstitüsü İktisat Anabilim Dalı İktisat Teorisi Bilim Dalı Yüksek Lisans Tezi, İstanbul (“University of Istanbul, Social Sciences Institute, Economics Main Department, Economic Theory Discipline, Master’s Degree Thesis, Istanbul”).

Harsh Vineet Kaur (2010), “Analysis of Banks in India—A CAMEL Approach,” Global Business Review, vol. 11 no. 2 257-280.

Jha, Suvita and Hui, Xiaofeng (2012), “A Comparison of Financial Performance of Commercial Banks: A Case Study of Nepal,” African Journal of Business Management, Vol. 6(25), pp. 7601-7611,

Karadaş, Sadık (2006), “Türk Bankacılık Sisteminin Verimlilik Açısından Değerlendirilmesi” (“Rating of the Turkish Banking System in Terms of Productivity”), Marmara Üniversitesi Bankacılık ve Sigortacılık Enstitüsü Bankacılık Ana Bilim Dalı Yüksek Lisans Tezi, İstanbul (“Marmara University, Banking and Insurance Institute Main Discipline, Master’s Degree Thesis, Istanbul”)

Kaya, Yasemin Türker (2001), “Türk Bankacılık Sektöründe Camels Analizi” (“Camels Analysis in the Turkish Banking Sector”) BDDK Çalışma Raporu: 2001/6 (“BRSA Working Paper: 2001/6”).

Kouser, Rehana, Saba, Irum (2012), “Gauging the Financial Performance of Banking Sector Using CAMEL Model: Comparison of Conventional, Mixed and Pure Islamic Banks in Pakistan,” International Research Journal of Finance and Economics, Issue 82.

Muhammet, Mercan (2008), “Türk Bankacılık Sisteminin Yeniden Yapılandırılması ve Performans Gelişimi (“Re-Structuring of the Turkish Banking System and Its performance Pattern”), Marmara Üniversitesi Bankacılık ve Sigortacılık Enstitüsü Doktora Tezi, İstanbul (“Marmara University, Banking and Insurance Institute, Doctoral Dissertation, Istanbul”).

Sakarya, Şakir (2010), “Camels Derecelendirme Sistemine Göre İMKB’deki, Yerli ve Yabancı Sermayeli Bankaların Karşılaştırmalı Analizi” (“A Comparative Analysis of the Local and Foreign Capital Banks on the ISE on the Basis of the Camels Rating System”), Akademik Araştırmalar ve Çalışmalar Dergisi (“Journal of Academic Researches and Studies”, Prof. Dr. Alaeddin Yavaşça Özel Sayısı (Prof. Dr. Alaeddin Yavaşça Special Issue”), pp: 7-21.

Sarker, Abdul Awwal(2008), “CAMELS Rating System in the Context of Islamic Banking: A Proposed ‘S’ for Shariah Framework”

Suadiye, Gülhan (2006), “Banka Başarısızlıkları ve Bankacılık Düzenlemeleri: İMKB’de İşlem Gören Türk Ticaret Bankalarının Başarısızlık Olasılığının Tahmini” (“Banking Failures and Banking Regulation: Forecasting the Probability of Failure of the Turkish Commercial Banks Traded on the ISE”), Doktora Tezi, Ankara (“Doctoral Dissertation, Ankara”).

Tulgar, Koray (1993), “1993 Ticari Bankalarda Aktif-Pasif Yönetimi” (“1993 Asset-Liabilities Management in Comme4cial Banks”), TBB Yayınları, Ankara (“TBA Publications, Ankara”).

Whalen, Gary (2010), “Are Early Warning Models Still Useful Tools for Bank Supervisors,” Social Science Research Network, March 2.

Yardımcı, Nilüfer (2006), “Bankacılık Sektöründe Etkinlik Analizi: Türk Bankacılık Sektörü ve Avrupa Birliği’ne Üye Bazı Ülkelerin Bankacılık Sektörlerinin Karşılaştırmalı Bir Analizi” (“Efficiency Analysis in the Banking Sector: A Comparative Analysis of the Turkish Banking Sector with the Banking Sectors of Some Member Countries of the European Union”), Erciyes Üniversitesi Sosyal Bilimler Enstitüsü Yüksek Lisans Tezi, Kayseri (“Erciyes University, Social Sciences Institute, Master’s Degree Thesis, Kayseri”).

* Corresponding Author, esavasbasci@hitit.edu.tr, Hitit University, Faculty of Economics and Administrative Sciences, Business Management Department,Corum, Turkey