PERFORMANCE EVALUATION OF TURKISH TYPE A MUTUAL FUNDS AND PENSION STOCK FUNDS BY USING TOPSIS METHOD

PERFORMANCE EVALUATION OF TURKISH TYPE A MUTUAL FUNDS AND PENSION STOCK FUNDS BY USING TOPSIS METHOD Nesrn ALPTEKIN Anadolu Unversty Faculty of Economcs and Admnstratve Scences Department of Busness Admnstraton 26470 Eskşehr, TURKEY E-mal: nesrnesen@anadolu.edu.tr Abstract In ths paper, t s evaluated performance of Turksh Type A mutual funds and penson stock funds by usng TOPSIS method whch s a multcrtera decson makng approach. Both of these funds compose of stocks n ther portfolos, so t can be enabled to compare each other. Generally, mutual or penson funds are evaluated accordng to ther rsk and return. At ths pont, t s used tradtonal performance measurement technques of funds lke Sharpe rato, Sortno rato, Treynor ndex and Jensen s alpha. TOPSIS method takes nto consderaton all of these fund performance measurement technques and provdes more reasonable performance measurement. Key Words: Penson Funds, Performance Evaluaton, Topss Method, Type A Fund. JEL Classfcaton: C02, G20, G23 1. INTRODUCTION The performance evaluaton of penson and mutual funds has been a very nterestng research topc not only for researchers, but also for managers of fnancal, bankng and nvestment nsttutons and ndvdual nvestors. Mutual funds are popular nvestment vehcles whch make t easy for small nvestors to nvest ther money n a dversfed pool of securtes. Penson funds are a pool of assets formng an ndependent legal entty that are bought wth the contrbutons to a penson plan for the exclusve purpose of fnancng penson plan benefts. Mutual funds are dsplayed actvty for long years, but penson funds are farly a new nvestment vehcle n Turksh Captal Market. Before 1960, portfolo managers evaluated portfolo performance usually on the rate of return, although they knew that rsk was a very mportant varable n determnng. The lack of knowledge how to measure and quantfy of rsk s the reason for omttng t (Gürsoy and Erzurumlu, 2001:44). Performance evaluaton of funds has an mportance for nvestors and portfolo management companes. After the development of portfolo theory n early 60s, and CAPM n subsequent years, studes on portfolo performance evaluaton have pcked up speed. There are lots of papers on portfolo performance evaluaton n the lterature, but some fundamental papers are as follows. Treynor (1965) was the frst researcher developng a composte measure of portfolo performance and he measured portfolo rsk wth beta and calculated portfolo s market rsk premum to ts beta. 11

Sharpe (1966) developed a composte ndex whch s smlar to the Treynor measure but the only dfference s that Sharpe used the standard devaton, nstead of beta, to measure the portfolo rsk. Jensen (1968) evaluated 115 mutual funds performances usng alpha whch s an ndcator of the fund managers forecastng ablty. He found that fund managers dd not have superor performance. 2. PERFORMANCE EVALUATION TECHNIQUES Performance evaluaton of funds s an mportant ssue for fund management and s an mportant part of the nvestment actvtes. Attractng and keepng nvestors depend on performance of a fund or a portfolo manager. It s wdely accepted that performance evaluaton should consst of two components; rsk and return (Moy, 2002:226). In the lterature, there are several performance evaluaton technques that take nto consderaton rsk n dfferent ways. Some of these technques are based on standard devaton (total rsk) whch s a representaton of the rsk and some of them predcates on systematc rsk(beta). In ths paper, Sharpe rato, Sortno rato, Treynor ndex and Jensen s alpha are used n performance evaluaton of funds. Sharpe and Sortno ratos are the performance evaluaton technques whch based on total rsk; the others are based on systematc rsk. 2.1. Sharpe rato The Sharpe rato s a rsk-adjusted measure of performance, whch s often used to evaluate the performance of a portfolo and ts manager. The rato compares the return of the portfolo to the rsk-free rate as well as the rsk generated by the portfolo. The focus of ths rato s on the return generated by the portfolo n comparson to the amount of rsk taken. The more rsk taken, the hgher the return should be to compensate for the rsk. Sharpe rato s formulated as follows: Sharpe Rato r a r f σ =. a In ths formula; r a s the average return of the portfolo, r f s rsk-free rate of return and σ a s the standard devaton of the portfolo. The hgher the Sharpe rato, the better the performance of the portfolo s consdered to be. Arsng return dfferental or a fallng Standard devaton are both good events and they leads to a rse n the Sharpe rato; conversely, a fallng return dfferental or a rsng standard devaton are both bad events and they leads to a fall n the Sharpe rato. Hence, a hgher Sharpe rato s good, and a lower one s bad. When choosng between two alternatves, the Sharpe rato crteron s therefore to choose the one wth the hgher Sharpe rato (Dowd, 2000:211-212). 2.2. Sortno rato The Sortno rato s smlar to the Sharpe rato, whch measures the rsk-adjusted return of nvestments or portfolos. Unlke the Sharpe rato, the Sortno uses downsde-volatlty (sometmes 12

referred to as sem-volatlty) as the denomnator nstead of standard devaton. The use of downsde-volatlty allows the Sortno rato to measure the return of negatve volatlty. Downsde devaton dfferentates postve volatlty from negatve volatlty, unlke standard devaton. Standard devaton s the square root of volatlty. However, usng standard devaton as a measure of rsk may not be completely accurate. For example, assume nvestment A has a return of 10% n year one and -10% n year two. Investment B has a 0% return n year one and a 20% return n year two. The total varance n these nvestments s the same, 20%. However, nvestment B s obvously more favorable. Because the Sharpe rato measures rsk usng standard devaton, the Sharpe rato does not dfferentate between postve and negatve volatlty. Sortno rato s calculated by followng formula r MAR Sortno Rato =, DV where r a s the return of a asset or portfolo, MAR s the mnmum acceptable return and DV s the downsde-volatlty. The Sortno Rato dfferentates between ths postve and negatve volatlty by replacng standard devaton wth downsde-volatlty. Downsde-volatlty s the volatlty of returns below a mnmal acceptable return (MAR). Dstrbuton of returns s analyzed below the MAR. The denomnator of the Sortno rato s calculated only wth data from perods where performance was below the set MAR. Ths dfferentates the postve and negatve volatlty. Large Sortno Ratos ndcate a low rsk of large losses occurrng and should be consdered more by rsk conscous nvestors. 2.3. Treynor ndex Treynor (1965) was the frst researcher developng a composte measure of portfolo performance. Treynor ndex provdes a measure of excess return per unt of systematc rsk (beta). The underlyng assumpton of the Treynor ndex s that a mult-asset portfolo dversfes unsystematc rsk away and the relevant rsk that remans s systematc rsk. Treynor ndex s calculated as: a r a r f Treynor Index =. β a In ths ndex, the numerator s dentcal to the Sharpe rato. Therefore, both Treynor and Sharpe measure excess returns for a gven level of rsk. 2.4. Jensen s alpha Alpha s a coeffcent that s proportonal to the excess return of a portfolo over ts requred return, or ts expected return, for ts expected rsk as measured by ts beta. Jensen s alpha, developed by Jensen(1968), assumes that the Captal Asset Prcng Model(CAPM) s emprcally vald. Jensen s alpha s computed by the followng regresson equaton: 13

r r = + a f α + β( rm rf ) et. In the above equaton, r a s the arthmetc average of the returns, r f s the arthmetc average of the rsk-free nterest rate returns, r m s the return of the benchmark portfolo, β s the fund systematc rsk and e t s the random error term of the fund at perod t. A postve value of alpha ndcates superor rsk-adjusted performance, whle a negatve value ndcates nferor rsk-adjusted performance (Cesar and Panetta, 2002:106). Jensen performance crteron does not evaluate the ablty of portfolo managers to dversfy, snce the rsk premums are calculated n terms of β (Gürsoy and Erzurumlu, 2001:45). 3. TOPSIS METHOD TOPSIS (Technque for Order Preference by Smlarty to Ideal Soluton) method s one of the useful mult-crtera decson makng technques and was frstly proposed Hwang and Yoon (1981). Accordng to ths technque, the chosen alternatve should have the shortest dstance from the postve deal soluton(pis) and the farthest from the negatve deal soluton(nis). The PIS s a soluton that maxmzes the beneft crtera and mnmzes the cost crtera, whereas the NIS maxmzes the cost crtera and mnmzes the beneft crtera(bentez, Martn and Roman, 2007:548). The TOPSIS method takes nto consderaton smultaneously the dstances to both the PIS and the NIS. The soluton whch s closest the PIS and farthest to NIS s the deal soluton. In ths paper, TOPSIS method s used for determnng the fnal rankng of Turksh Type A mutual funds and penson stock funds. In the followng the steps of TOPSIS method are gven: Step 1: Decson matrx s normalzed by usng followng equaton: r j = a m = 1 j a 2 j, = 1,...,m; j = 1,..., n. Step 2: Weghted normalzed decson matrx s formed: v j = w r, 1,...,m; j = 1,..., n j =. Step 3: PIS (postve deal soluton) and NIS (negatve deal soluton) are determned: * * * * ( v1,v,...,v j,...,vn ) ( v,v,...,v,..., v ) * A 2 = maxmum values, A = 1 2 j n mnmum values. Step 4: The dstance of each alternatve from PIS and NIS s calculated as: 14

n * * 2 d = ( vj v j ), = 1, 2,...,m. j= 1 n 2 d = ( vj v j ), = 1, 2,...,m. j= 1 Step 5: The closeness coeffcent of each alternatve (CC ) s calculated as: d CC =. * d + d Step 6: The rankng of alternatves s determned by comparng CC values. 4. PERFORMANCE EVALUATION OF TURKISH TYPE A AND PENSION STOCK FUNDS 4.1. Research data Data used n ths research ncludes monthly returns of 11 Type A stock mutual funds and 11 Penson stock mutual funds n January 2007- December 2008 analyss perod. Funds names and codes whch are used n the research are gven n Table 1 and Table 2. Table 1. Penson Stock Funds and Funds Codes Penson Stock Mutual Funds Aegon Stock Income PMF Avvasa Stock Growth PMF Avvasa Stock PMF Anadolu Hayat Stock Growth PMF Anadolu Hayat Stock Growth Group PMF Anadolu Hayat Stock Growth PMF -Beyaz Başak Stock Growth PMF Garant Stock Growth PMF Oyak Stock Growth PMF Vakıf Stock Growth PMF Yapı Kred Stock Growth PMF Fund Code ANS AEB AEH AH5 AG3 AHB BEH GEH OEH VEH YEH 15

Table 2. Type A Stock Mutual Funds and Funds Codes Type A Stock Mutual Funds Fund Code Akbank Type A Stock MF AK3 Denzbank Type A Stock MF DAH Eczacıbaşı Menkul Değerler Type A Stock MF ECH Fnansbak Type A Stock MF FAF Forts Yatırım Menkul Değerler Type A Stock MF FAS ING Bank Type A Stock MF IGH T. Garant Bankası Type A Stock MF GHS T. İş Bankası Type A Stock MF TI2 TEB Yatırım Type A Stock MF TYH Tekstlbank Type A Stock MF TAH Yapı Kred Type A Stock MF YHS 4.2. Returns rate of funds Monthly returns of funds are calculated by usng unt prces of them n operaton date. Data are obtaned from offcal webste of Captal Markets Board of Turkey (CMB, 2009). Monthly return of a fund s calculated by followng equaton: r = ln r + ln r. t 1 t In the equaton; r s monthly return of the fund, r t+1 s closng prce of the fund at (t + 1) th month and r t s closng prce of the fund at t th month. 4.3. Rsk-free return rate Varous rsk-free return rates, whch are approprate to structure of the funds, are used n the lterature. In ths research, monthly Domestc Government Bonds (DGS) Performance ndex s used. Data of DGS performance ndex s obtaned from Istanbul Stock Exchange (ISE) offcal webste (ISE, 2009). 4.4. Benchmark In ths paper, ISE 100 Natonal Index s used for benchmark to penson stock funds and Type A stock funds. It s supposed approprate for penson stock funds n Indvdual Penson System Progress Report 2008 whch s prepared by Penson Montorng Centre. To confrm the approprateness of ths benchmark for Type A stock funds, correlaton analyss s used. It s found that correlaton coeffcent of Type A stock funds returns between ISE 100 Natonal Index returns s average 0.90. Therefore, ISE 100 natonal Index s an approprate benchmark for Type A stock funds. Monthly closng prces of ISE 100 natonal Index are obtaned from Istanbul Stock Exchange offcal webste (ISE, 2009). Monthly returns of ISE 100 are calculated as follows: r = ln r + ln r. t 1 t In the above equaton; r s monthly return of the ndex, r t+1 s closng prce of the ndex at (t + 1) th month and r t s closng prce of the ndex at t th month. 16

4.5. Unt Root Test Results In ths paper, unt root test s appled to test whether the tme seres are statonary or not. Statonarty of a seres s an mportant phenomenon because t can nfluence ts behavor. For a statonary seres a shock wll gradually de away n tme and the seres wll turn back ts average value for long run. In lterature, Dckey-Fuller and Augmented Dckey-Fuller(ADF) tests are the best known ones for statonarty of the seres. The testng procedure for the ADF test can be formulated as follows Δ y α β γ δ Δ δ Δ + ε t = + t + yt 1 + 1 yt 1 + L + p yt p t, where α s constant, β the coeffcent on a tme trend and p the lag order of the autoregressve process. The unt root test s then carred out under the null hypothess, H 0 : γ = 0, aganst the alternatve hypothess of H 1 : γ < 0. The value of the test statstc, ˆ DF = γ τ SE( ˆ γ ) s computed and t can be compared to the relevant crtcal value for the ADF Test. If the test statstc s greater (n absolute value) than the crtcal value, then the null hypothess of γ = 0 s rejected and no unt root s present. The ADF tests results of the funds and the benchmark are gven n Table 3. Table 3: Unt Root Test Results Names of Funds and Benchmark ADF test Statstcs Aegon Stock Income PMF -5,54404* Akbank Type A Stock MF - 6,07976* Anadolu Hayat Stock Growth PMF -6,16713* Anadolu Hayat Stock Growth Group PMF -6,29392* Anadolu Hayat Stock Growth PMF Beyaz -5,62541* Avvasa Stock Growth PMF -6,07040* Avvasa Stock PMF -6,41663* Başak Stock Growth PMF -6,12941* Denzbank Type A Stock MF -5,70684* Eczacıbaşı Menkul Değerler Type A Stock MF -5,07937* Fnansbak Type A Stock MF -5,70935* Forts Yatırım Menkul Değerler Type A Stock MF -5,35576* Garant Stock Growth PMF -5,62230* ING Bank Type A Stock MF -5,35207* Oyak Stock Growth PMF -8,06000* T. Garant Bankası Type A Stock MF -5,39664* T. İş Bankası Type A Stock MF -5,28046* TEB Yatırım Type A Stock MF -5,80443* Tekstlbank Type A Stock MF -5,01947* Vakıf Stock Growth PMF -5,62811* Yapı Kred Stock Growth PMF -6,02109* Yapı Kred Type A Stock MF -5,73944* ISE 100 Natonal Index -5,60442* * ndcates that the seres s sgnfcant at 1% mportance level. McKnnon crtcal values at 1%, 5% and 10% mportance levels are -3, 76960, -3, 00487 and -2, 64224, respectvely. 17

4.6. Fndngs Table 4. Fund Performance Evaluaton wth Performance Measurement Technques Names of Funds Sharpe Sortno Treynor Jensen Aegon Stock Income PMF -0,2789-0,8244-0,0322 0,0000 Akbank Type A Stock MF -0,3009-0,6146-0,0345-0,0017 Anadolu Hayat Stock Growth PMF -0,2440-1,9112-0,0276 0,0042 Anadolu Hayat Stock Growth Group PMF -0,2624-0,9524-0,0296 0,0022 Anadolu Hayat Stock Growth PMF -0,2802-0,9983-0,0328-0,0005 Beyaz Avvasa Stock Growth PMF -0,3091-0,9499-0,0354-0,0028 Avvasa Stock PMF -0,3414-0,9649-0,0389-0,0389 Başak Stock Growth PMF -0,4114-1,1480-0,0487-0,0133 Denzbank Type A Stock MF -0,4577-0,6041-0,0545-0,0135 Eczacıbaşı Menkul Değerler Type A Stock -0,2457-0,1364-0,0280 0,0034 MF Fnansbak Type A Stock MF -0,2019-0,7518-0,0230 0,0077 Forts Yatırım Menkul Değerler Type A -0,2725-0,7277-0,0328-0,0004 Stock MF Garant Stock Growth PMF -0,2935-0,9523-0,0337-0,0013 ING Bank Type A Stock MF -0,3981-0,4873-0,0478-0,0090 Oyak Stock Growth PMF -0,2492-0,7544-0,0310 0,0012 T. Garant Bankası Type A Stock MF -0,2928-0,1551-0,0330-0,0005 T. İş Bankası Type A Stock MF -0,2600-0,3254-0,0302 0,0015 TEB Yatırım Type A Stock MF -0,3186-0,2446-0,0363-0,0033 Tekstlbank Type A Stock MF -0,3049-0,2929-0,0398-0,0048 Vakıf Stock Growth PMF -0,2811-0,9011-0,0320 0,0002 Yapı Kred Stock Growth PMF -0,2924-0,9070-0,0338-0,0014 Yapı Kred Type A Stock MF -0,2475-0,2327-0,0274 0,0037 Funds performances whch are evaluated based on Sharpe rato, Sortno rato, Treynor ndex and Jensen s alpha are gven n Table 4. Accordng to the Table 4, there s not a domnant alternatve. To determne the prortes of crtera, factor analyss s used for objectvty. In factor analyss, varables are clustered n three factors. These three factors explan almost overall varance,.e., 99,672%. Frst factor explans 63,394% of overall varance, second factor explans 25,443% of overall varance and thrd factor explans 10, 835% of overall varance. To fnd the prortes of crtera, factor rotaton s made usng varmax method and as a result of factor rotaton, factor matrx s obtaned n Table 5. Table 5. Factor Matrx Component 1 2 3 Sharpe,945,316 -,010 Sortno -,024,044,999 Treynor,954,286 -,032 Jensen,395,917,059. 18

In the above Table, varables (crtera) are lsted n the rows and factors are lsted n the columns. The values are placed n the Table are factor loadngs. Takng nto consderaton factor loadngs whch mean the relatonshps among each varable (crtera) and three factors, undetermned part of the overall varance by the each varable(1 0,99672 = 0,00328) s allocated to factors n ther percentage of explan. For example, the value 0,00328* 0,63394 = 0, 0020, s added to frst factor s percentage explan of overall varance. Therefore, adjusted percentage of explan all varables are 0,63594, 0,25523 and 0,10865,respectvely. These values also denote the mportance level of the factors. Then, the relatonshps are checked among the factors and the varables. In factor matrx, the frst factor s hghly related to the Sharpe rato and Treynor ndex, the second factor s hghly related to the Jensen s alpha and the thrd factor s hghly related to the Sortno rato. Importance level of frst factor (0,63594) s allocated 1 st and 3 rd varables n proporton wth ther factor loadngs. For example, the sum of frst factor loadngs 1 st and 3 rd varables s 0,945 + 0,954 = 1,899. Thus, mportance level of frst varable s 0,63594* 0,945 / 1,899 = 0, 3165. The mportance levels of the other varables are found as 0,1086, 0,3195 and 0,2552, respectvely. Table 6. Weghted Normalzed Decson Matrx of Funds Names of Funds Sharpe Sortno Treynor Jensen 0,3165 0,1088 0,3195 0,2552 Aegon Stock Income PMF -0,2789-0,8244-0,0322 0 Akbank Type A Stock MF -0,3009-0,6146-0,0345-0,0017 Anadolu Hayat Stock Growth PMF -0,244-1,9112-0,0276 0,0042 Anadolu Hayat Stock Growth Group PMF -0,2624-0,9524-0,0296 0,0022 Anadolu Hayat Stock Growth PMF Beyaz -0,2802-0,9983-0,0328-0,0005 Avvasa Stock Growth PMF -0,3091-0,9499-0,0354-0,0028 Avvasa Stock PMF -0,3414-0,9649-0,0389-0,0389 Başak Stock Growth PMF -0,4114-1,148-0,0487-0,0133 Denzbank Type A Stock MF -0,4577-0,6041-0,0545-0,0135 Eczacıbaşı Menkul Değerler Type A Stock MF -0,2457-0,1364-0,028 0,0034 Fnansbak Type A Stock MF -0,2019-0,7518-0,023 0,0077 Forts Yatırım Menkul Değerler Type A Stock MF -0,2725-0,7277-0,0328-0,0004 Garant Stock Growth PMF -0,2935-0,9523-0,0337-0,0013 ING Bank Type A Stock MF -0,3981-0,4873-0,0478-0,009 Oyak Stock Growth PMF -0,2492-0,7544-0,031 0,0012 T. Garant Bankası Type A Stock MF -0,2928-0,1551-0,033-0,0005 T. İş Bankası Type A Stock MF -0,2600-0,3254-0,0302 0,0015 TEB Yatırım Type A Stock MF -0,3186-0,2446-0,0363-0,0033 Tekstlbank Type A Stock MF -0,3049-0,2929-0,0398-0,0048 Vakıf Stock Growth PMF -0,2811-0,9011-0,032 0,0002 Yapı Kred Stock Growth PMF -0,2924-0,907-0,0338-0,0014 Yapı Kred Type A Stock MF -0,2475-0,2327-0,0274 0,0037 19

Then the dstance of each frm from PIS and NIS wth respect to each crteron are calculated. Then closeness coeffcent of each fund s calculated and the rankng of the funds are determned accordng to these values. The rankng of the funds are shown n Table 7. It s found that Anadolu Hayat Stock Growth PMF has the best performance and Yapı Kred Type A Stock MF has the worst performance among the 22 funds. In fnal rankngs of funds, t s seen that penson mutual stock funds have hgher performance than the Type A mutual stock funds. Table 7. Rankngs of Funds Accordng to CC Values Names of Funds d * d CC Aegon Stock Income PMF 0,8346 0,6310 0,4305 10 Akbank Type A Stock MF 0,6406 0,4359 0,4049 15 Anadolu Hayat Stock Growth PMF 1,9038 1,7062 0,4726 1 Anadolu Hayat Stock Growth Group PMF 0,9559 0,7538 0,4409 7 Anadolu Hayat Stock Growth PMF Beyaz 1,0045 0,8020 0,4440 3 Avvasa Stock Growth PMF 0,9638 0,7601 0,4410 6 Avvasa Stock PMF 0,9874 0,7832 0,4423 4 Başak Stock Growth PMF 1,1822 0,9777 0,4527 2 Denzbank Type A Stock MF 0,7031 0,5062 0,4186 14 Eczacıbaşı Menkul Değerler Type A Stock MF 0,2086 0,1243 0,3733 18 Fnansbak Type A Stock MF 0,7476 0,5470 0,4225 12 Forts Yatırım Menkul Değerler Type A Stock 0,7395 0,5354 0,4200 13 MF Garant Stock Growth PMF 0,9624 0,7592 0,4411 5 ING Bank Type A Stock MF 0,5726 0,3783 0,3979 16 Oyak Stock Growth PMF 0,7596 0,5565 0,4228 11 T. Garant Bankası Type A Stock MF 0,2583 0,1579 0,3793 17 T. İş Bankası Type A Stock MF 0,3610 0,1651 0,3138 21 TEB Yatırım Type A Stock MF 0,3344 0,1786 0,3481 19 Tekstlbank Type A Stock MF 0,3608 0,1826 0,3360 20 Vakıf Stock Growth PMF 0,9097 0,7066 0,4372 9 Yapı Kred Stock Growth PMF 0,9181 0,7147 0,4377 8 Yapı Kred Type A Stock MF 0,2771 0,1063 0,2772 22 No 5. CONCLUSION In ths paper, t s evaluated performance of Turksh penson stock mutual funds and Type A stock mutual funds n the perod January 2007-December 2008 by usng monthly returns of the funds. Frstly, t s analyzed that the returns seres are statonary or not. After the determnng the all of the seres are statonary, performance evaluatons of the seres are obtaned by usng performance measurement technques; Sharpe rato, Sortno rato, Treynor ndex and Jensen s alpha. Accordng to performance evaluaton results, t s not found a domnant alternatve. Therefore, to obtan the prortes of the performance evaluaton technques t s used the factor analyss. The varables are clustered n there factors. The fnal rankngs of the funds are obtaned by usng TOPSIS method whch s a multcrtera decson makng approach. The am of the TOPSIS 20

method s to choose the alternatve that should have the shortest dstance from the postve deal soluton and the farthest from the negatve deal soluton. It s found that Penson Stock Mutual Funds have superor performance than the Type A Stock Mutual funds. Anadolu Hayat Stock Growth PMF has the best performance and Yapı Kred Type A Stock MF has the worst performance among the 22 funds. Ths paper has the unque about the performance evaluaton of Turksh penson Stock Funds and Type A Stock Mutual Funds wth multcrtera decson approach. It can be appled all penson and mutual funds. Fndngs obtaned from ths paper have benefts for ndvdual nvestors, fund managers and researchers. BIBLIOGRAPHY Alper, Yusuf (2002), Sosyal Güvenlkte Yen Br Adım: Breysel Emekllk, Çmento İşveren Dergs, Vol. 16, No. 2, pp.11 32. Altıntaş, Kadr Murat (2008), Türk Özel Emekllk Fonlarının Rsk Odaklı Yönetm Performansı: 2004 2006 Dönemne İlşkn Br Analz, Anadolu Ünverstes Sosyal Blmler Dergs, Clt 8, Sayı 1,pp. 85 110. Arslan, Mehmet (2005), A Tp Yatırım Fonlarında Yönetclern Zamanlama Kablyet ve Performans İlşks Analz: 2002 2005 Dönem Br Uygulama, Tcaret ve Turzm Eğtm Fakültes Dergs,2, pp.1 23. Bentez, Juan Manuel, Martn, Juan-Carlos and Concepcόn Román (2007), Usng Fuzzy Number for Measurng Qualty of Servce n the Hotel Industry, Toursm Management, Vol. 28, No.2, pp.544-555. Blake, C.R., Elton, E.J. and Gruber, M.J.(1993), The Performance of Bond Mutual Funds, The Journal of Busness, Vol.66, No.3, pp.371 403. Bolak, Mehmet (1991), Sermaye Pyasası, Menkul Kıymetler ve Portföy Analz, Frst Edton, İstanbul:Beta. CMB(2009), http://www.cmb.gov.tr/ndex.aspx [Accessed 20.03.2009]. Cesar, Rccardo and Fabo Panetta (2002), The Performance of Italan Equty Funds, Journal of Bankng&Fnance, Vol.26, No.1, pp.99-126. Dağlar, Hüseyn (2007), Kurumsal Yatırımcılar Olarak Emekllk Yatırım Fonları ve Performanslarının Değerlendrlmes, İstanbul: Türkye Bankalar Brlğ. Dahlqust, M., Engström, S. and Soderlnd, P.(2000). Performance and Characterstcs of Swedh Mutual Funds, The Journal of Fnancal and Quanttatve Analyss, Vol.35, No.3, pp. 409 423. Detzler, M.L.(1999), The Performance of Global Bond Mutual Funds, Journal of Bankng&Fnance, 23, pp.1195 1217. Doğanay, M. Mete (2002). Hsse Sened Fonlarının Çok Krterl Karar Yaklaşımı le Derecelendrlmes, Ankara Ünverstes SBF Dergs, Clt 57, Sayı, 3, pp.31-48. Dowd, Kevn (2000), Adjustng for Rsk: An Improved Sharpe Rato, Internatonal Revew of Economcs and Fnance, 9, pp.209-222. 21

Ertuğrul, İrfan and Karakaşoğlu, Nlsen (2009), Performance Evaluaton of Turksh Cement Frms wth Fuzzy Analytc Herarchy Process and TOPSIS Methods, Expert Systems wth Applcatons, 36, pp.702-715. Gürsoy, Cud Tuncer and Erzurumlu, Y. Ömer (2001), Evaluaton of Portfolo Performance of Turksh Investment Funds, Doğuş Unversty Journal, Issue 4, pp.43-58. Hwang, C.L. and Yoon, K.(1981), Multple Attrbutes Decson Makng Methods, Berln:Sprnger. ISE (2009). http://www.mkb.gov.tr/ver.htm [Accessed 15.01.2009]. Kaya, Yılmaz (2004), Çok Amaçlı Karar Verme Yöntemlernden TOPSIS ve ELECTRE Yöntemlernn Karşılaştırılması. http://www.hho.edu.tr/huten/20032004%20seminer%20 INTERNET /YILMAZ%20KAYA/YILMAZ%20KAYA%20%5BWORD%5D.pdf, [Accessed 1.15.2009]. Korkmaz, Turhan and Uygurtürk, Hasan (2007), Türkye dek Emekllk Fonlarının Performans Ölçümü ve Fon Yönetclernn Zamanlama Yeteneğ, Akdenz İ.İ.B.F. Dergs, Vol.7, No.4, pp. 66-93. McDonald, John G. (1973), French Mutual Fund Performance: Evaluaton of Internatonally-Dversfed Portfolos, The Journal of Fnance, Vol.28, No.5, pp.1161-1180. Moy, Ronald L. (2002), Portfolo Performance Illustratons From Mornngstar, Journal of Educaton for Busness, Vol.77, No.4, pp.226-229. Oprcovc, S. and Tzeng, Gwo-Hshung (2004), Compromse soluton by MCDM methods: A comparatve analyss of VIKOR and TOPSIS European Journal of Operatonal Research, 156, pp.445 455. Pendarak, Konstantna, Doumpos, Mchael and Constantn Zopounds (2003), Assessng Equty Mutual Fund s Performance Usng A Multcrtera Methodology: A Comparatve Analyss, South Eastern Europe Journal of Economcs, 1, pp.85-104. Seçme, Neşe Yalçın, Bayrakdaroğlu, Al and Cengz Kahraman (2009), Fuzzy Performance Evaluaton n Turksh Bankng Sector Usng Analytc Herrachy Process and TOPSIS (In Press). Sharpe, Wllam F. (1966). Mutual Fund Performance, Journal of Busness, Vol.39, No.1, pp.119 138. Shh, Hsu-Shh (2008), Incremental analyss for MCDM wth an applcaton to group TOPSIS, European Journal of Operatonal Research, 186, pp.720-734. Teker, S., Karakurum, E. ve Tav, O. (2008). Yatırım Fonlarının Rsk Odaklı Performans Değerlemes, Doğuş Ünverstes Dergs, Clt 9, Sayı 1, pp.89-105. Treynor, Jack L. (1965), How to Rate Management of Investment Funds, Harvard Busness Revew, Vol.43, No.1, pp.63-75. 22