Descriptive analysis

4.2 The Second Stage of Efficiency Analysis

4.2.1 Descriptive analysis

The same as in the stage 1, the author also carry the data out (see Appendix A.2) by running CCR input-oriented model. The descriptive statistics on Input and Output data are shown in the following table.

Table 4.6: The second stage analysis - Descriptive statistics on Input and Output data

Max Min Average SD

Total deposits 113.4804 1.1089 20.4487 25.0659

Total loans 74.9831 0.8015 16.3988 19.7156

Interest income 3.7077 0.1052 0.8176 0.6914

Profit before tax 0.7774 0.0033 0.1773 0.1908

EPS 0.1215 0.0023 0.0495 0.0299

Revenue per employee 367.5631 68.7813 141.1558 57.3592 Total loans/bad debt 729.1849 11.3636 190.3091 197.5128

In the risk-management analysis, the risk in credit sector is the most important. Table 4.7 indicates Taishin International Bank (TW_TIB) has the lowest NPL (Non-performing loans) ratio. Besides, Saigon Hanoi Bank (VN_SHB) has the largest NPL ratio. It shows that this bank is inefficient in the bad debt recovery and the systemic risk is increasing. Other important element is bank’s benefit. It relates to Interest income, Profit before tax, EPS and Revenue per employee of the banks. In the business activities, due to the global financial crisis, the banks also have tighter controlled in the loan activities.

43

4.2.2 Correlation

Table 4.7: The second stage analysis – Correlation of Inputs and Outputs

M1 M2 M3 M4 Y1 Y2 Y3

M1 1

M2 0.9833 1

M3 0.5095 0.5553 1

M4 0.7080 0.7325 0.5762 1

Y1 0.1682 0.1991 0.5469 0.5882 1

Y2 0.5896 0.5586 0.2091 0.5167 0.1003 1

Y3 0.5060 0.4866 0.0199 0.5784 0.1629 0.4409 1

Notes: M is immediate variable (input); Y is output in stage 2. M1: Total deposits. M2: Total loans. M₃: Interest income. M₄: Profit before tax. Y₁: EPS. Y₂: Revenue per employee. Y₃: Total loans/bad debt.

The table 4.7 explains the correlations of input and output variables. These correlations as the whole are not too high or too low except for the correlation between Total loans and Total deposits that was mentioned in the previous section.

4.2.3 Efficiency analysis

Table 4.8: The second stage efficiency analysis

DMU TE PTE SE RTS

Efficient Projected

VN_VCB 0.2743 0.3506 0.7824 DRS

VN_CTG 0.2548 1 0.2548 DRS

VN_BIDV 0.1768 0.1779 0.9939 CRS

VN_VBARD 0.1631 0.4383 0.3722 DRS

VN_MSB 0.6447 0.6449 0.9997 CRS

VN_STB 0.3411 0.3667 0.9301 DRS

VN_EAB 0.8413 0.9133 0.9211 DRS

VN_EIB 0.6575 1 0.6575 DRS

VN_ACB 0.3096 0.3682 0.8409 DRS

VN_VPB 0.6953 0.7612 0.9133 DRS

44

Table 4.8 (cont.): The second stage efficiency analysis

DMU TE PTE SE RTS

45

In the second stage, the average overall TE of the banks is 0.5746, in which the average TE of Vietnamese banks is 0.6720 and Taiwanese banks is 0.4772. It shows that Vietnamese banks have more effectively risk-management system than Taiwanese banks. The number of efficient DMUs is 12 (8 DMUs from Vietnam and 4 DMUs from Taiwan), the number of inefficient DMUs is 32 (14 DMUs from Vietnam and 18 DMUs from Taiwan). Moreover, Southeast Asia Bank (VN_SEAB), Ocean Commercial Bank (VN_OJB), Mekong Housing Bank (VN_MHB), En Tie Commercial Bank (TW_ETB) and Cota Bank (TW_CTB) are the most efficient DMUs in both stages (overall TE equal to 1). From the result of DEA-CCR model, the main reason for inefficient DMUs can be defined. In the inputs utilization, most of inefficient banks make too many loans to clients and because of unstable global financial situation as well as customers’ situation leading to inability to pay their debts. Therefore, these banks are loss in business, NPLs increasing.

From BCC model analysis, the average PTE score of the banks (0.7404) is less than the average SE score (0.7602). It indicates that the inappropriate in banks’ scale factor is not the main cause leading to inefficiency of the banks. That is weakness in inputs utilization of bank’s managers. As mentioned above, because of the ineffective inputs utilization for example the capital mobilization and lending activities are unbalanced leading to poor liquidity, less competitiveness compare with the other banks. See Appendix B.2, the DEA results for stage 2 over period 2008-12 were shown to see which banks are the most efficient through the years.

46

4.2.4 Slack analysis

Table 4.9: The second stage analysis – The slack analysis of the banks

DMU Score

47

Table 4.9 (cont.): The second stage analysis – The slack analysis of the banks

DMU Score employee. Y(3): Total loans/bad debt.

The same like slack analysis in the first stage, the table 4.10 shows some suggestions for inefficient DMUs to improve the efficiency of risk-management. In this stage, this research gets 32 inefficient DMUs and selects a single one (Bank of Taiwan – TW_BOT) for example.

In the second stage, the overall TE score of TW_BOT is 0.1229. For the slack analysis, this bank does not need to change anything of the output variables. However, in the input utilization, to improve the efficient score or in other word to improve the quality of risk, this

48

bank should reduce $44.751 Billion US in amount of Total deposits, $28.846 Billion US in amount of Total loans and $0.228 Billion US in amount of Interest income. (See more at Appendix C.2).

4.2.5 Sensitivity analysis

Table 4.10: The second stage analysis – Sensitivity analysis for the banks

DMU Original

TE M1 M2 M3 M4 Y1 Y2 Y3

VN_VCB 0.2743 0.2743 0.2743 0.1853 0.2732 0.1463 0.2743 0.2743 VN_CTG 0.2548 0.2335 0.2548 0.2224 0.2548 0.1114 0.2548 0.2548 VN_BIDV 0.1768 0.1768 0.1768 0.0937 0.1642 0.1047 0.1768 0.1768 VN_VBARD 0.1631 0.1631 0.1631 0.1396 0.1544 0.0559 0.1631 0.1631 VN_MSB 0.6447 0.6447 0.6232 0.6278 0.6395 0.6348 0.3999 0.6390 VN_STB 0.3411 0.3411 0.3411 0.2961 0.3128 0.2102 0.3411 0.3408 VN_EAB 0.8413 0.8413 0.8413 0.6109 0.7508 0.4091 0.8413 0.8413 VN_EIB 0.6575 0.4802 0.6575 0.6575 0.6575 0.4797 0.6560 0.6305 VN_ACB 0.3096 0.3096 0.3096 0.3073 0.2557 0.2233 0.3013 0.3094 VN_VPB 0.6953 0.6953 0.6435 0.6953 0.6039 0.5105 0.6730 0.6953 VN_TCB 0.3768 0.3768 0.3576 0.3712 0.3233 0.3043 0.3327 0.3768 VN_MB 0.7431 0.7363 0.7431 0.6260 0.7431 0.3665 0.7431 0.7431 VN_VIB 0.8921 0.8517 0.8771 0.8921 0.8921 0.6126 0.8591 0.8921

VN_SEAB 1 1 1 1 1 1 0.5658 1

VN_HDB 1 1 1 0.8905 1 0.9722 0.7756 1

VN_PNB 1 1 1 1 0.7292 0.7281 0.8794 1

VN_SCB 1 1 1 1 0.6757 1 0.2977 1

VN_SHB 0.4133 0.4077 0.4133 0.2250 0.4133 0.4133 0.0408 0.4133

VN_OJB 1 1 1 0.9429 1 0.9973 0.8699 1

VN_LPB 1 1 1 1 1 0.9060 1 1

VN_ABB 1 1 1 1 1 0.7858 1 1

VN_MHB 1 1 1 1 1 0.7132 1 1

TW_BOT 0.1229 0.1229 0.1213 0.0528 0.1229 0.1223 0.0896 0.1229

49

Table 4.10 (cont.): The second stage analysis – Sensitivity analysis for the banks

DMU Original

TE M1 M2 M3 M4 Y1 Y2 Y3

TW_LBOT 0.2220 0.2220 0.2220 0.0742 0.2177 0.1690 0.2048 0.2220 TW_TCB 0.1357 0.1357 0.1357 0.0652 0.1341 0.1186 0.1236 0.1357 TW_FCB 0.1926 0.1926 0.1926 0.0612 0.1891 0.1367 0.1813 0.1926 TW_CHB 0.1962 0.1962 0.1962 0.0815 0.1962 0.1554 0.1662 0.1926 TW_HNB 0.1574 0.1574 0.1574 0.0563 0.1574 0.1371 0.1215 0.1561 TW_TBB 0.2018 0.2006 0.2018 0.0762 0.2018 0.1963 0.1465 0.2018 TW_MCB 0.2712 0.2712 0.2712 0.1244 0.2712 0.1946 0.2503 0.2665 TW_ESB 0.2926 0.2926 0.2882 0.1855 0.2926 0.1887 0.2926 0.2708 TW_CCB 0.2000 0.1963 0.2000 0.0848 0.2000 0.1167 0.2000 0.1955 TW_TIB 0.3612 0.3578 0.3612 0.2628 0.3612 0.2417 0.3612 0.2835 TW_SPB 0.2710 0.2710 0.2710 0.1322 0.2710 0.2123 0.2413 0.2655

TW_ETB 1 1 1 0.3609 1 0.6422 1 1

TW_JSB 1 1 1 0.5809 1 0.7204 1 1

TW_TCCB 0.6240 0.6240 0.6240 0.2609 0.6240 0.4888 0.5288 0.6151 TW_UBOT 0.6861 0.6861 0.6421 0.5060 0.6861 0.4471 0.6861 0.6090 TW_FEIB 0.5715 0.5715 0.5163 0.2392 0.5715 0.5253 0.4629 0.5715

TW_BOK 1 1 1 0.5472 1 1 0.7437 1

TW_BTC 0.4627 0.4492 0.4627 0.4235 0.4627 0.4243 0.4627 0.4507 TW_YTB 0.5347 0.5347 0.5023 0.2591 0.5347 0.4995 0.4000 0.5347

TW_CTB 1 1 1 1 1 1 1 1

TW_BOP 0.9958 0.9958 0.9958 0.3668 0.9175 0.9958 0.2730 0.9935 Average 0.5746 0.5684 0.5690 0.4451 0.5513 0.4640 0.4859 0.5689 No. of eff.

DMUs 12 12 12 7 10 4 6 12

Notes: M is immediate variable (input); Y is output in stage 2. M(1): Total deposits. M(2):

Total loans. M(3): Interest income. M(4): Profit before tax. Y(1): EPS. Y(2): Revenue per employee. Y(3): Total loans/bad debt.

In the second stage, removing input/output variables are also applied. The results shown 7 new technical efficiency scores when removing each input and output variable. In this stage,

50

there is only 1 DMU achieving the absolute value (equal to 1) through all step of the removing variables. That is Cota Bank (TW_CTB). Besides, this bank gets the absolute score in the first stage too. It indicates that Cota Bank is not only good in bank’s business operation, but also efficient in risk-management activities.

Here, the lowest average TE score is 0.4451 that retrieved when removing Interest income factor. This element becomes the important factor when evaluating the risk-management activities of the banks because it affects all the banks. If the risk in the bank’s operation is at high level, the banks will have to face more difficulties in business especial in finance factor and then it is difficult for them to gain benefits.

4.2.6 Sorting

Applying the same ways like stage 1, Common compromise weights and Assurance region method are to sort efficient DMUs from the results of CCR model. Under CCW method, there is only one DMU get efficient score (TW_CTB). However, there are two DMUs with efficiency score 1 (VN_SEAB and TW_CTB) under AR method. (See Appendix D.2).

51

Figure 4.2: The second stage analysis - Comparison between CCR, CCW, AR method

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

VN_VCB VN_CTG VN_BIDV VN_VBARD VN_MSB VN_STB VN_EAB VN_EIB VN_ACB VN_VPB VN_TCB VN_MB VN_VIB VN_SEAB VN_HDB VN_PNB VN_SCB VN_SHB VN_OJB VN_LPB VN_ABB VN_MHB TW_BOT TW_LBOT TW_TCB TW_FCB TW_CHB TW_HNB TW_TBB TW_MCB TW_ESB TW_CCB TW_TIB TW_SPB TW_ETB TW_JSB TW_TCCB TW_UBOT TW_FEIB TW_BOK TW_BTC TW_YTB TW_CTB TW_BOP

CCR CCW AR2 AR5 AR10

52