We summarize the technical efficiency values for the year of 2005 calculated by Ratio Output-oriented DEA model which is described in Eq.(2) as shown in Table 5. Of the total results of the DMUs, the mean technical efficiency score is quite high (0.992). This implies that the gap in efficiency difference among the 24 commercial banks is not too large in 2005, since all banks have operated in a highly competed environment in Taiwan and have already improved their efficiency as the same conclusion in TSER-YIETH CHEN(2000). Among 24 commercial banks, five of them (First, Bank Of China, E.Sun, Taishin, Shin Kong) are categorized as technically efficient banks because their efficiency scores are equal to one.
Note that all of these technically efficient banks are the bank subsidiaries in Financial Holding Companies (FHCs). Thus we conclude that the 11 bank subsidiaries in the FHC do
outperform independent banks and also supported by three other findings: First, the mean technical efficiency scores of bank subsidiaries in FHC is 0.984 and independent banks is 0.922. Second, the six poorest efficiency rating banks in 2005 among 24 commercial banks include Union(0.928), Cosmos(0.921),Ta Chong(0.915), Chinese Bank(0.849),Chinfon(0.821), and Bowa(0.808); however, all of these six are independent banks. Third, none of the
independent banks defines as optimal-efficient bank. All of the findings described above have supported that bank subsidiaries in FHC are relative efficiencies than independent banks.
We further conduct a Wilcoxon-Mann-Whitney test which serves to test the hypothesis that the two groups include bank subsidiaries and independent banks are in the same
population. In fact, Wilcoxon-Mann-Whitney test, also called Rank-Sum-Test, is one of the non-parametric tests. Since T1= -3.2 < -1.96= -T0.025, we reject the null hypothesis at the
1
12 / ) 1 ( + +
+
= +
n m mn
1)/2 n m(m -T S
where m is the number of DMUs which are bank subsidiaries in Financial Holding Companies( Group one);
n is the number of DMUs which are independent banks (Group two). S is the sum of rankings for all DMUs in group one
In our case, m=11, n=13, C=｛1, 1, 1, 1, 1, 0.998 , 0.989, 0.988, 0.977, 0.975, 0.968, 0.966, 0.964, 0.963, 0.949, 0.948, 0.945, 0.938, 0.928, 0.921, 0.915, 0.849, 0.821, 0.808｝; R=﹛3, 3, 3, 3, 3, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24﹜; (the underlined numbers indicate they are group one);
There is another finding that should be addressed that the types of main businesses of FHC do have influences on the efficiency of its bank subsidiary. For a clear explanation, we first divide bank subsidiaries of FHCs into three groups by the main businesses of FHCs. The main businesses can be categorized into securities, insurance and banking. By comparing the mean technical efficiency scores, we can find an interesting result that is those main business whose categories are not banking perform poorer than those whose main business is banking.
As shown in Table 5, the mean technical efficiency scores are 0.996, 0.983, and 0.943 which correspond to the main businesses of banking, insurance and security, respectively. It’s easy to explain these results, because after all, when banks are not the main business in FHC, the big part of resources would not be concentrated in the banks. Therefore, the performance of bank subsidiaries in FHCs which the main business is banking outperforms other types. The other reason may be explained by strategies of the company, when banks are not the main business in the FHC, the FHC may sacrifice benefits from their unimportant subsidiaries to trade off more benefits in their subsidiary main business.
We further summarize the technical efficiency scores and Financial Ratios which we used as output in our DEA model as shown in Table 6 for comparison. We have calculated the specific maximum, 2nd maximum, minimum and 2nd minimum values of seven financial ratios.
We have used NPL as measurement of risk control, so that the smaller the NPL value is; the better the performance of banks is. The minimum and 2nd minimum values of NPL are 0.5 and 0.79, which performed by Bank of China and E. Sun., respectively and both are categorized as efficient banks because their technical efficiency values are one. ROE, ROA, Margin and P/L are the measurements of profitability; thus the more the values are, the better the performance of banks are. The maximum (2nd maximum) values in ROE, ROA, Margin and P/L are
19.95(18.22), 1.61(1.19), 29.77(26.43), and 3700.42(2200.26), respectively. We can easily find that three of the optimal-efficient banks include the First Commercial bank, Bank Of
S=3+3+3+3+3+3+6+7+8+12+16+18=82
China and Taishin have higher values in at least two of these profitability financial ratios. In this case, we can realize why those three banks are categorized as optimal-efficient.
We use Growth Rate of Deposits (hereafter, GDR) and Growth Rate of Loans (hereafter, GLR) to measure Growth ability of banks. Usually, the higger value means the higher growth potential the bank is. The maximum (2nd maximum) of GDR and GLR are 53.02(38.98) and 52.56(33.34) performed by Shin Kong and E. Sun which are technically efficient since their technical efficiency values are one. In summary, the optimal-efficient banks have good values in at least two of these seven ratios, and we can divide these efficient banks into three
categories. First, banks performed well in at least two of three abilities, such as the Bank of China (Risk Control, Profitability) and E. Sun( Risk Control and Growth). Second, banks performed well in Profitability Ability, like the First (ROE, Margin) and Taishin( ROE and ROA) banks. Third, banks performed exceptionally well in Growth Ability, like Shin Kong.
It’s worth noting that Shin Kong as a relatively new bank only performed well at Growth Ability, but worst in other two abilities. In the conventional DEA model which has not considered the Growth Ability, the performance of Shin Kong would not be categorized as efficient. However, as we have mentioned in Chapter Three, the indices which were used to measure performance should also consider the potential growth in the future. We can thus include the growth ability index into our model for encouraging those long-term growth policies. We can suggest that Shin Kong Bank would have great potential to be a good performer in the following year.
In order to find out the trend of efficiency for DMUs, we calculate a five year technical efficiency comparison by the method of Window Analysis as shown in Table 7. In Window Analysis, we treat every observation as an independent DMU, thus we have 120 DMUs evaluated. Since the data covers five years, we have a deflated P/L variable by CPI. The CPI indices for five years are shown in Appendix 7. Since the Legislative Yuan passed the Financial Holding Company (FHC) Act in November 2001, Hua Nan Commercial Bank joined into Hua Na FHC in 2001 and Taishin, E. Sun, Chinatrust, Sinopac, Bank of China,
FHC since 2003. Shin Kong and Macoto Bank were merged by Shin Kong FHC in Oct. 2004 and Oct. 2005, respectively and two affiliated banks merged in Dec. 2005. In order to confirm the results of bank subsidiaries of FHCs outperforming independent banks, we thus categorize the 120 DMUs into several groups: Group One, if banks are independent banks in 2005, and Group Two, if banks are subsidiaries of FHC in 2005. Furthermore, we also separate those DMUs in Group Two into three groups by the main businesses of FHCs. For instance, Cathay Bank is in Cathay FHC with its main business focused on Cathay Insurance, therefore we classify it into Group B which is under Group Two. Chinatrust Bank is in Chinatrust FHC in which its main business is Chinatrust Insurance Brokerage, thus we classify it into Group A under Group Two.
The five-year averages technical efficiencies are 0.8118 for Group One and 0.9104 for Group Two. The average technical efficiencies for Group One are 0.7328, 0.7448, 0.8107, 0.86 and 0.9108 from 2001 to 2005. The average technical efficiencies for Group Two in 2005 are 0.8194, 0.8719, 0.9287, 0.9563 and 0.9759 from 2001 to 2005. It seems to be an
increasing trend of technical efficiency, no matter for banks in Group One or Two. Obviously, the performance of banks in Group One outperform than Group Two. We thus conclude that the bank subsidiaries in FHCs do outperform independent banks and the successful banks will succeed again.
On the other hand, the ranking of performance within Group Two is Group A, Group B and then Group C in all five years (0.8486 ［Group A］>0.7830［Group B］>0.7535［Group C］>0.7328［Group One］in 2001). Therefore, we can conclude that the bank subsidiaries in FHCs in which their main business are banking outperform others which focus their main business in insurance or securities. It confirms that main business of FHC does have
influences on the performance of their bank subsidiaries. This conclusion is intuitive, because the big part of financial funds and all kinds of resources would support the main business in FHC. Thus bank subsidiaries in FHC whose main business is banking would perform better.
We also find those technical efficiency banks on the list of good performers which is evaluated by The Banker. For example, E. Sun Commercial Bank, which is categorized on the list of technically efficient banks in 2004 and 2005 in our analysis, won the awards of “Bank of the Year”, which reflect the best overall performance by banks in the calendar year 2004.
The awards, decided by The Banker’s global editorial team, is not only based on the
assessment of the latest results and performance data provided by the banks but also analyzes strategic developments and overall achievements. E. Sun is one of the 510 selective banks from a record of 138 countries announced by The Banker (Sep. 05 2005). In addition, E. Sun Commercial Bank established the strategic alliances with the UK’s Prudential, thus E. Sun Bank can extend its services in 2004. On the other hand, according The Banker, much of the IT establishment has been achieved on the risk management front in preparation for Basel II, and E.SUN Bank is deemed to continue to be the leader in asset quality compared to its peers.
Taishin Bank which was categorized as technically efficient in 2002 is also the focus for foreign investment. (The Banker, July 04, 2005, pp.154).
Efficiency Scores Efficiency Scores
Subsidiary of a FHC 0.984
Main business of FHC
Banking
0.996
First # 1.000
Bank Of China # 1.000 E.Sun # 1.000 Taishin # 1.000 Chinatrust # 0.998 Sinopac # 0.989 Hua Nan # 0.988
Insurance 0.983
Shin Kong # 1.000 Cathay # 0.966
Security 0.943
Fuhwa # 0.948 Jih Sun # 0.938
Independent Banks 0.922
Sunny 0.977
Far Eastern 0.975
Chang Hwa 0.968
Hwatai 0.964
Bank Of Taipei 0.963 Overseas Chinese 0.949
Entie 0.945
Union 0.928
Cosmos 0.921
Ta Chong 0.915
Chinese Bank 0.849
Chinfon 0.821
Bowa 0.808
Table 6. Summary of technical efficiency scores and ratios
# bank Subsidiaries in FHC
*maximum value; **2nd maximum value
***minimum value; ****2nd minimum value
12 Ranked by technical efficiency value.
Year Banks Efficiency Score Average Year Banks Efficiency Score Average
2001 0.7328 2001 0.8194
Sunny 0.8590 Group A 0.8486
Bank Of Taipei 0.8340 Chinatrust 0.9160
Union 0.7850 Sinopac 0.8990
Far Eastern 0.7710 Taishin 0.8860
Cosmos 0.7590 Bank Of China 0.8840
Hwatai 0.7520 E.Sun 0.8660
Ta Chong 0.7330 Hua Nan 13#14 0.7540
Chang Hwa 0.7300 First 0.7350
Chinfon 0.7130 Group B 0.7830
Entie 0.7020 Cathay 0.8450
Chinese Bank 0.6640 Shin Kong 0.7210
Bowa 0.6210 Group C 0.7535
Overseas Chinese 0.6030 Fuhwa 0.7910
Jih Sun 0.7160
2002 0.7448 2002 0.8719
Bank Of Taipei 0.8670 Group A 0.9250
Far Eastern 0.8580 Taishin15 # 1.0000
Sunny 0.8560 E.Sun16 # 0.9510
Cosmos 0.8410 Chinatrust17 # 0.9460
Union 0.7990 Sinopac18 # 0.9300
Ta Chong 0.7920 Bank Of China19 # 0.9250
Hwatai 0.7760 Hua Nan # 0.8700
Chinese Bank 0.7020 First 0.8530
Overseas Chinese 0.6870 Group B 0.8075
Chang Hwa 0.6810 Cathay20 # 0.8760
Entie 0.6390 Shin Kong 0.7390
Bowa 0.6120 Group C 0.7505
Chinfon 0.5720 Jih Sun21 # 0.7540
Fuhwa22 # 0.7470
13 The sign of # characteristices of those banks which are bank subsidiaries in FHC.
14 Hua Nan Commercial Bank has joined into Hua Na FHC since Dec. 2001.
15 Taishin Banks has joined into Taishin FHC since Feb. 2002.
16 E. Sun Commercial Bank has joined into E. Sun FHC since Jan. 2002.
17 Cinatrust Commercial Bank has joined into Chinatrust FHC since May 2002.
18 Sinopac Bank has joined into Sinopac FHC since May 2002.
19 The International Commercial Bank of China has joined into Mega FHC since Dec. 2002.
20 Cathay United Bank has joined into Cathay FHC since 2002.
21 Jih Sun Bannks has joined into Jih Sun FHC since Fed. 2002.
22 Fuhwa Banks has joined into Fuwa FHC since Aug. 2002.
Table 7. Five year comparable technical efficiencies (Cont.)
Group One Group Two Year Banks Efficiency Score Average Year Banks Efficiency Score Average
2003 0.8107 2003 0.9287
Far Eastern 0.9520 Group A 0.9551
Bank Of Taipei 0.9340 E.Sun # 0.9770
Sunny 0.9120 Taishin # 0.9740
Ta Chong 0.8790 Sinopac # 0.9630
Cosmos 0.8780 Bank Of China # 0.9550
Union 0.8630 Chinatrust # 0.9530
Hwatai 0.8420 First 23 # 0.9460
Entie 0.8180 Hua Nan # 0.9180
Chang Hwa 0.7540 Group B 0.9010
Chinfon 0.7210 Cathay # 0.9590
Chinese Bank 0.7060 Shin Kong 0.8430
Bowa 0.6410 Group C 0.8640
Overseas Chinese 0.6390 Fuhwa # 0.9040
Jih Sun # 0.8240
2004 0.8600 2004 0.9563
Bank Of Taipei 0.9690 Group A 0.9791
Entie 0.9640 E.Sun # 1.0000
Far Eastern 0.9540 Taishin # 0.9960
Hwatai 0.9470 Sinopac # 0.9910
Sunny 0.9430 Bank Of China # 0.9860
Union 0.9140 Chinatrust # 0.9700
Cosmos 0.9130 First # 0.9600
Ta Chong 0.8930 Hua Nan # 0.9510
Overseas Chinese 0.8200 Group B 0.9225
Chang Hwa 0.8020 Cathay # 1.0000
Chinese Bank 0.7670 Jih Sun # 0.8450
Chinfon 0.7070 Group C 0.9100
Bowa 0.5870 Fuhwa # 0.9230
Shin Kong24 # 0.8970
23 The First Commercial Bank was acquired into the First FHC since 2003.
24 Shin Kong Commercial Bank was acquired into the Shin Kong FHC since Oct. 2004.
Year Banks Efficiency Score Average Year Banks Efficiency Score Average
2005 0.9108 2005 0.9759
Far Eastern 0.9690 Group A 0.9833
Chang Hwa 0.9680 Bank Of China # 1.0000
Hwatai 0.9640 E.Sun # 1.0000
Bank Of Taipei 0.9630 Sinopac # 0.9890
Sunny 0.9520 Taishin # 0.9860
Overseas Chinese 0.9480 Chinatrust # 0.9780
Entie 0.9450 First # 0.9700
Union 0.9280 Hua Nan # 0.9600
Cosmos 0.9200 Group B 0.9830
Ta Chong 0.9150 Shin Kong # 1.0000
Chinese Bank 0.8490 Cathay # 0.9660
Bowa 0.8080 Group C 0.9430
Chinfon 0.7110 Fuhwa # 0.9480
Jih Sun # 0.9380
5-year Average 0.8118 5-year Average 0.9104