Comprehensive Evaluation of the Operating Performance for Commercial Banks in China Based on Improved TOPSIS

Transcription

1 Internatonal Conference on Global Economy Commerce and Servce Scence (GECSS 04) Comprehensve Evaluaton of the Operatng Performance for Commercal Banks n Chna Based on Improved TOPSIS Chao L School of Busness enmn Unversty of Chna Beng Chna sxh_b@6.com Caqn Ye School of Economcs and Management North Chna Electrc Power Unversty Beng Chna Yecaqn89@6.com Abstract Wth the purpose of presentng a scentfc reasonable and feasble method to evaluate the performance of lsted commercal banks ths thess takes the current 6 lsted commercal banks n Chna as study obects selects ther data up to 0 and studes the performance evaluaton system for lsted commercal banks. Frst of all based on prncples of comprehensveness and representatveness ths study bulds a set of ndex system usng cluster analyss and multple correlaton coeffcent method. Secondly analytc herarchy process (AHP) and entropy method are used to dentfy the weght coeffcent. Last ths paper ranks and assesses the operatng performance for these commercal banks usng the mproved TOPSIS method to calculate the comprehensve scores for each bank. Keywords-commercal banks; operatng performance; comprehensve evaluaton; mproved TOPSIS method; cluster analyss I. INTODUCTION Wth the entry of foregn banks and ntegraton of nternatonal fnance Chnese commercal banks are facng unprecedented challenges ntensfyng the competton between banks. To some extent the core compettveness les n the operaton level and therefore mprovng operatng performance s the key factor for commercal banks to survve n the ferce competton. Puttng more research n ssues regardng to commercal banks operatng performance s not only necessary to mprove ther own self-management but also prerequste to enhance the operaton and management mechansm for the entre bankng ndustry and boost the stable and coordnated development of economy[][]. Ths paper takes the current 6 lsted commercal banks n Chna as study obects selects ther data up to 0 and studes the performance evaluaton system for lsted commercal banks. Double-classfcaton s used to buld the ndex system that s both comprehensve and representatve. As for the determnaton of ndex weght combnaton weghtng of analytc herarchy process (AHP) and entropy method are appled n order to make ndex system both ndcatve and practcal. In terms of comprehensve evaluaton ths thess uses the mproved TOPSIS method to calculate the comprehensve scores for each bank and assessed and ranked ther operatng performance. II. ESTABLISHMENT OF INDEX SYSTEM FO EVALUATION OF OPEATING PEFOMANCE A. Intal Evaluaton Index System There are a large number of ndcators to reflect commercal banks operatng performance. In order to buld a comprehensve and scentfc ndex system that can more accurately present and assess banks performance ths thess adopts lterature summarzng method consults predecessors studes and exstng regulatons and then ntally selects ndexes that are able to reflect status of fnancal proftablty asset lqudty asset securty and development ablty for these lsted commercal banks[3][4][5]. Ths ntal evaluaton ndex system s shown n table. Ths paper takes the current 6 lsted commercal banks n Chna as research obects studes ther data provded by SSE (Shangha Stock Exchange) and 0 annual reports of these banks and selects the above ndex data up to 0 as sample. Orgnal data s shown n table. B. Screenng of the Evaluaton Index System In practce t s not the case that the more ndexes the better and therefore we should use a few key ndcators n practcal evaluaton process. Based on predecessors work and the ntal ndex system [6] ths paper uses cluster analyss and multple correlaton coeffcent to screen these ndexes to fnalze them. Ths knd of method both subectvely consders the economc sgnfcance of the evaluaton ndexes and obectvely pcks the representatve ones. Frst of all these ndexes wll be dvdedly nto 4 categores usng cluster analyss. Ths s acheved by enterng the sample data nto SPSS.0 software and then applyng cluster analyzng the 4 categores of ndexes: fnancal proftablty ndexes asset lqudty ndexes asset securty ndexes and development ablty ndexes. The cluster analyss result s shown on fgure to fgure 4. Secondly the representatve ndex from each category wll be selected by calculatng the multple correlaton coeffcents. If there s only one ndex n a partcular category then t wll automatcally be selected. If the number of ndexes n a 04. The authors - Publshed by Atlants Press 7

2 category exceeds two then compare calculate the multple correlaton coeffcent for each ndex and pck the one that has the largest multple correlaton coeffcent because the larger the multple correlaton coeffcent the more useful nformaton ths ndex s supposed to provde n ths category. Take the Fnancal proftablty ndex category as an example; the process to dentfy the representatve ndex s as follows. TABLE I. INITIAL INDEX SYSTEM FO OPEATING PEFOMANCE EVALUATION OF CHINA S LISTED COMMECIAL BANKS Obect Operatng Performance Evaluaton for lsted commercal banks n Chna Evaluaton Index Fnancal Proftablty Asset Lqudty Asset Securty Ablty Evaluaton Factors Code Unt Nature OAA X % Postve OE X % Postve EPS X3 yuan Postve Cost/evenue X4 % Inverse Per Capta Proft X5 K Postve eturn on Captal X6 % Postve Lqudty ato X % Moderate Loan to Depost ato X % Moderate Cash eserve ato X3 % Postve NPL X3 % Inverse Overdue Loans ato X3 % Inverse Captal Adequacy ato X33 % Postve Core Captal Adequacy ato X34 % Postve NPL Provson coverage X35 % Postve evenue Growth ate X4 % Postve Total Assets Growth ate X4 % Postve Depost Growth ate X43 % Postve Loan Growth ate X44 % Postve Non-nterest ncome growth rate X45 % Postve ate of captal accumulaton X46 % Postve TABLE II. OIGINAL DATA FO EVALUATION OF OPEATING PEFOMANCE FO 6 LISTED COMMECIAL BANKS IN CHINA BANK NAMES X X X3 (Yuan) X4 Shenzhen Bank of Nngbo SPD Bank Hua Xa Bank Chna Mnsheng Bank Chna Merchants Bank Bank of Nanng Industral Bank Bank of Beng ABC Bank of Communcatons ICBC Chna Everbrght Bank Chna Construct-on Bank Bank of Chna Chna CITIC Bank X5 (K) X6 X X X3 X3 Bank Names X3 X33 X34 X35 X4 X4 X43 X44 X45 X46 Shenzhen Bank of Nngbo SPD Bank Hua Xa Bank Chna Mnsheng Bank Chna Merchants Bank Bank of Nanng Industral Bank Bank of Beng ABC Bank of Communcatons ICBC Chna Everbrght Bank Chna Construct-on Bank Bank of Chna Chna CITIC Bank

3 In the same way we can select the representatve ndexes for the other 3 categores and form establsh the fnal comprehensve ndex system for evaluatng commercal banks operatng performance as shown n table 3. TABLE III. COMPEHENSIVE EVALUATION INDEX SYSTEM FO LISTED BANKS OPEATING PEFOMANCE FIGUE I. EXAMPLE OF A FIGUE CAPTION. (FIGUE CAPTION) Obect Operatng Performance Evaluaton for lsted commercal banks n Chna Evaluaton Index Fnancal Proftablty Indcators Asset Lqudty Indcators Asset Securty Indcators Ablty Indcators Evaluaton Factors Cod e Unt OE X % eturn on Captal X % Lqudty ato X3 % Loan to Depost ato X4 % Core Captal Adequacy ato X5 % NPL Provson coverage X6 % evenue Growth ate X7 % Depost Growth ate X8 % FIGUE II. EXAMPLE OF A FIGUE CAPTION. (FIGUE CAPTION) FIGUE III. EXAMPLE OF A FIGUE CAPTION. (FIGUE CAPTION) FIGUE IV. EXAMPLE OF A FIGUE CAPTION. (FIGUE CAPTION) Accordng to the cluster analyss result we can further dvde the 6 ndexes that reflect fnancal proftablty nto two groups wth X3 X5 and X6 beng group and X X X and X4 beng group. For each of the 3 ndexes n group we calculate ts multple correlaton coeffcent wth the other 5 and get the followng results: X 3 = X 5 = X 6 = Obvously X6 (return of equty) has the largest multple correlaton coeffcent and t s therefore selected. Smlarly we do the calculaton for each of the ndexes n group and get the results as follows: X = X = X 4 = and agan we pck X(weght average proft margn on net asset) snce t yelds the largest multple correlaton coeffcent. Therefore X6 and X are selected as the representatve ndexes to reflect banks fnancal proftablty. III. DETEMINATION OF THE INDEX WEIGHT In the process of mult-subect comprehensve evaluaton weghtng determnes the accuracy of the evaluaton result. In order to avod the arbtrarness of subectve weghtng method and mpractcablty of obectve weghtng method ths thess uses the combnaton of AHP method and entropy method to dentfy the weghtng coeffcent for the comprehensve ndex system. Accordng to the followng way: ( β ) w = β w + w where w s the combned weght coeffcent w s the obectve weght coeffcent determned by entropy method w s the subectve weght coeffcent β s the subectve preference coeffcent β s the obectve preference coeffcent and β [ 0]. β s determned by decson makers based on the actual stuaton of commercal banks background of specfc assessed proects and ther preferences. In ths thess the subectve weght coeffcent s set as 0.4 the obectve weght coeffcent s set as 0.6 and the combned weght coeffcent s therefore dentfed as follows: w = [ ] IV. COMPEHENSIVE EVALUATION OF OPEATING PEFOMANCE FO LISTED COMMECIAL BANKS IN CHINA A. Improved TOPSIS Method The tradtonal TOPSIS method s a way to measure a subect s comprehensve benefts accordng to the relatve dstance from deal soluton. The concept of ths method s to fnd the postve deal pont and negatve deal pont n a decson-makng process and choose the alternatve that has the shortest geometrc dstance from the postve deal pont or the largest geometrc dstance from the negatve deal pont. However when the geometrc dstance from the postve deal pont s equal to that from the postve deal pont ths method s unable to evaluate each alternatve s advantages and 9

4 dsadvantages [9]. Therefore ths thess adopts the mproved TOPSIS whch uses vertcal proecton method to evaluate banks operatng performance. The specfc steps are lsted as follows: ) Set the number of evaluaton subect as n number of evaluaton ndex as m and the ntal decson matrx as X then X x x x m x x x x n x n x n m m = ) Unform all the factors n Matrx X. x assocated wth the crtera havng a postve mpact () y = - x assocated wth the crtera havng a negatve mpact 3) Unform all the factors n Matrx X. z y = n ( y ) = 4) Unform all the factors n Matrx X. u = w z (4) where () (3) ω s the weght coeffcent dentfed by the combnaton of AHP and entropy method. 5) Unform all the factors n Matrx X. = m u = ( u u u n ) u + = m a x { u } (5) u = ( u u u n ) u = m n { u } (6) 6) Calculate the relatve dstance d and the larger the d the better the evaluaton obect s comprehensve benefts. + emember u = u u u = u u then the relatve dstance d between defnton pont u and negatve deal pont u s: = n (7) d = u u u where d [0] u u s the nner product of u and u u s Eucldean Norm of u namely u = u u m + ( ) = B. Comprehensve Evaluaton Based on Improved TOPSIS Accordng to the relatve dstance calculated under equaton () to (8) the operatng performance for the 6 commercal banks s ranked n table 6. / (8) TABLE IV. COMPEHENSIVE ANKING IN 6 LISTED COMMECIAL BANKS IN CHINA UNDE THE IMPOVED TOPSIS METHOD Corps Score(Z) ank Corps Score(Z) ank Shenzhen Bank of Beng Bank of Nngbo ABC SPD Bank Bank Of Communcatons Hua Xa Bank ICBU Chna Mnsheng Bank Chna Everbrght Bank Chna Merchants Chna Constructon Bank Bank 0.47 Bank of Nanng Bank of Chna Industral Bank 0.03 Chna CITIC Bank C. Emprcal esults Accordng to commercal banks nature of busness we can dvde the 6 mentoned commercal banks nto 3 categores: State Owned Commercal Banks (SOCB) ncludng Agrcultural Bank of Chna Bank of Communcatons Industral and Commercal Bank of Chna Chnese Constructon Bank natonal ont-stock commercal banks ncludng Shenzhen Bank Shangha Pudong Bank Huaxa Bank Chna Mnsheng Bank Chna Merchants Bank Industral Bank Chna Everbrght Bank Chna CITIC Bank and cty commercal banks ncludng Bank of Beng Bank of Nanng and Bank of Nngbo). Comprehensve assessment results show two man characterstcs of commercal banks operatng performance consstent wth actual stuaton. ) Postonng fgures and tables natonal ont-stock commercal banks have better operatng performance than state owned commercal banks. As shown n the results n table 6 natonal ont-stock commercal banks are hghly ranked wth better performance especally the TOP Shenzhen Bank and Industral Bank. In contrast stated owned commercal banks are mostly lowest ranked wth the Bg Four banks and Communcaton banks all ranked below 0th. Wth better regulated bankng system postve bankng culture and comprehensve ncentve mechansm natonal ont-stock commercal banks have developed quckly under severe competton and mproved ther operatng performance. As for state owned commercal banks as they grow larger so does the nternal nerta whch prevents ther market compettveness and proftablty from mprovng and ncreasng. In spte of the better operatng performance for natonal ont-stock commercal banks there are vast dspartes wthn them. Of the 6 banks 5 natonal ont-stock commercal banks are hghly ranked ndcatng an above-theaverage operatng performance. However there are also 3 natonal ont-stock commercal banks that are ranked very low wth operatng performance below the average level. ) Cty commercal banks are developng fast wth robust overall strength. Among the 6 banks the 3 cty commercal are ranked 3rd 5th and 8th respectvely wth hgh scores. It can be seen that even though they can t compete wth state owned banks and natonal ont-stock commercal banks n assets qualty and sze cty commercal banks are developng rapdly 0

5 wth ncreasng compettveness due to ther flexble busness pattern sharp response market and effectve operaton mechansm. In addton the development of cty commercal banks reles heavly on the local economy. These 3 banks are all headquartered n ctes wth developed economy. Strong local fscal revenues large number of small and medum-szed prvate enterprses and hgh per capta natonal ncome all provde a sold base for cty banks development. V. CONCLUSIONS AND POSPECTS Ths thess takes the lsted commercal banks as samples sets the ndex system dentfes the ndex weght and comprehensve assessment method and then bulds a evaluaton model for commercal banks performance and systematcally analyzes how to effectvely evaluate the operatng performance for commercal banks n Chna. However due to lack of access to the fnancal data from some of these banks the selecton of samples and ndexes s lmted whch wll have some mpacts on the comprehensveness of the evaluaton results. Moreover n the screenng process for ndcators restrcted by the sample data some of the key ndcators such as non-performng loan ratos are deleted omttng ther nfluence on the effectveness the evaluaton. Future studes should make mprovement n the followng two aspects. Frst selected samples should be more comprehensve. Wth ncreasng number of commercal banks research n ther operatng performance wll become more dffcult and therefore proper selecton of samples wll enhance the qualty and accuracy of evaluaton results. Secondly n order to get a better evaluaton of lsted commercal banks operatng performance data studed should not be restrcted to fnancal ndcators. A combnaton of fnancal and nonfnancal ndcators wll help to assess banks operatng performance more comprehensvely and scentfcally. ACKNOWLEDGMENT I would lke to partcularly show my best grattude to my supervsor Dr. Song who has provde me wth valuable gudance n every stage of the wrtng of ths paper. EFEENCES [] Zhang H. W. Wang KY. Comprehensve evaluaton of operatng performance for ont-stock commercal banks n Chna (n Chnese). Hanan Fnance (0) 6-8 (009) [] Zhang Q. Pan H. X. The emprcal research on the effcency of commercal banks n Chna (n Chnese). Journal of egonal Fnancal esearch (09) 6-65 (0) [3] George E. H. Dmtros S. S. Effcency measurement of the Greek commercal banks wth the use of fnancal ratos: a data development analyss approach. Management Accountng esearch (5) 0-4 (004) [4] Zhao X. X. Zhang Q. W. Comprehensve evaluaton of operatng performance for lsted commercal banks n Chna (n Chnese). Busness Economy (7): [5] Hu Z. F. Wang T. T. Wang S. Y. Operatng performance of Chna s commercal banks---an analyss based on 4 lsted banks data n 009 (n Chnese). Fnance Forum (05) (0) [6] Zhang L. J. Luo Z. Evaluaton ndcator screenng method for lsted companes operatng performance (n Chnese). Statstcs and Decson (8) (008) [7] Jang L. M. Applcaton of AHP method n comprehensve evaluaton of commercal banks operatng performance (n Chnese). Statstcs & Informaton Trbune (0) 9-3 (00) [8] Xe C. Zhong Z. Entropy method and ts applcaton n comprehensve evaluaton of bank's performance (n Chnese). Chna Soft Scence (09) 08-0 (00) [9] Zeng W. Comprehensve evaluaton of bank s operatng strength based on entropy coeffcent and TOPSIS method (n Chnese). Statstcs and Decson (9) (007)