Luận án Nghiên cứu tác động của dịch vụ ngân hàng quốc tế đến hiệu quả hoạt động của các ngân hàng thương mại Việt Nam

0,999 1 0,999 irs 26 0,992 1 0,992 irs 27 0,99 0,993 0,998 drs 28 1 1 1 - 29 1 1 1 - 30 0,882 0,893 0,988 irs 31 1 1 1 - 32 1 1 1 - 33 0,892 0,999 0,893 irs 34 0,97 1 0,97 irs 35 0,865 0,91 0,951 irs mean 0,946 0,966 0,979 PHỤ LỤC 6 HIỆU QUẢ KỸ THUẬT, HIỆU QUẢ KỸ THUẬT THUẦN, HIỆU QUẢ QUY MÔ CỦA CÁC NGÂN HÀNG NGHIÊN CỨU NĂM 2012 firm crste vrste scale 1 0,928 1 0,928 drs 2 1 1 1 - 3 0,978 1 0,978 drs 4 0,849 1 0,849 drs 161 5 0,87 0,94 0,926 drs 6 1 1 1 - 7 0,881 0,917 0,961 drs 8 0,984 1 0,984 drs 9 1 1 1 - 10 0,867 0,931 0,931 drs 11 0,913 0,916 0,997 drs 12 0,865 0,877 0,987 drs 13 0,882 0,896 0,985 drs 14 1 1 1 - 15 0,873 0,874 0,999 irs 16 0,993 0,993 1 - 17 0,9 0,904 0,995 irs 18 0,98 0,991 0,989 drs 19 0,927 0,928 0,999 drs 20 1 1 1 - 21 0,956 1 0,956 irs 22 1 1 1 - 23 1 1 1 - 24 0,882 1 0,882 irs 25 0,997 1 0,997 irs 26 1 1 1 - 27 1 1 1 - 28 0,979 1 0,979 irs 29 0,958 1 0,958 irs 30 0,948 0,985 0,963 irs 31 0,814 0,821 0,992 irs 32 1 1 1 - 162 33 1 1 1 - 34 0,937 1 0,937 irs 35 0,847 1 0,847 irs mean 0,943 0,971 0,972 PHỤ LỤC 7 HIỆU QUẢ KỸ THUẬT, HIỆU QUẢ KỸ THUẬT THUẦN, HIỆU QUẢ QUY MÔ CỦA CÁC NGÂN HÀNG NGHIÊN CỨU NĂM 2013 firm crste vrste scale 1 0,949 1 0,949 drs 2 1 1 1 - 3 1 1 1 - 4 1 1 1 - 5 0,917 0,936 0,98 drs 6 1 1 1 - 7 0,844 0,853 0,99 drs 8 0,862 0,873 0,988 drs 9 1 1 1 - 10 0,816 0,847 0,963 drs 11 0,841 0,857 0,981 drs 12 1 1 1 - 13 0,932 0,937 0,995 drs 14 0,863 0,877 0,985 drs 15 1 1 1 - 16 0,93 0,933 0,996 irs 17 0,952 0,959 0,993 irs 18 0,838 0,853 0,982 irs 19 0,944 0,946 0,998 irs 163 20 0,847 0,855 0,99 irs 21 1 1 1 - 22 1 1 1 - 23 1 1 1 - 24 1 1 1 - 25 0,992 1 0,992 irs 26 0,831 0,838 0,992 irs 27 0,975 1 0,975 irs 28 0,94 0,96 0,979 irs 29 0,817 0,955 0,855 irs 30 0,96 1 0,96 irs 31 0,817 1 0,817 irs 32 0,792 0,853 0,929 irs 33 1 1 1 - 34 1 1 1 - mean 0,931 0,951 0,979 PHỤ LỤC 8 HIỆU QUẢ KỸ THUẬT, HIỆU QUẢ KỸ THUẬT THUẦN, HIỆU QUẢ QUY MÔ CỦA CÁC NGÂN HÀNG NGHIÊN CỨU NĂM 2014 firm crste vrste scale 1 0,931 1 0,931 drs 2 1 1 1 - 3 1 1 1 - 4 0,9 0,927 0,971 drs 5 1 1 1 - 6 0,906 0,907 0,999 drs 7 0,848 0,849 0,999 drs 164 8 1 1 1 - 9 0,823 0,836 0,983 drs 10 0,981 0,981 0,999 drs 11 1 1 1 - 12 0,967 0,968 0,999 irs 13 0,85 0,874 0,973 drs 14 0,883 0,892 0,991 irs 15 1 1 1 - 16 0,921 0,932 0,989 drs 17 0,883 0,885 0,997 irs 18 0,866 0,899 0,963 irs 19 0,826 0,836 0,988 irs 20 0,897 0,92 0,974 irs 21 1 1 1 - 22 1 1 1 - 23 0,976 0,998 0,978 irs 24 0,889 1 0,889 irs 25 0,82 0,823 0,996 irs 26 1 1 1 - 27 0,872 0,878 0,993 irs 28 0,873 0,916 0,953 irs 29 0,926 0,992 0,933 irs 30 0,855 0,963 0,887 irs 31 0,841 0,854 0,985 irs mean 0,92 0,94 0,98 165 PHỤ LỤC 9 CHỈ SỐ MALMQUIST CỦA CÁC NHTMVN GIAI ĐOẠN 2008 – 2014 MALMQUIST INDEX SUMMARY OF ANNUAL MEANS year effch techch pech sech tfpch 2 1,067 0,892 1,047 1,019 0,952 3 0,988 1,034 0,985 1,002 1,022 4 1,049 0,92 1,025 1,023 0,965 5 1,006 0,906 1,004 1,003 0,911 6 0,964 1,01 0,975 0,989 0,974 7 1,01 0,998 1,004 1,006 1,008 mean 1,013 0,958 1,006 1,007 0,971 MALMQUIST INDEX SUMMARY OF FIRM MEANS firm effch techch pech sech tfpch 1 1,035 0,956 1 1,035 0,989 2 1,009 0,982 1 1,009 0,99 3 1,02 0,968 1 1,02 0,987 4 1,011 0,974 1,01 1,001 0,985 5 1,015 0,992 1 1,015 1,007 6 0,984 0,938 0,984 1 0,923 7 0,989 0,928 0,984 1,005 0,918 8 1,006 0,982 1 1,006 0,988 9 0,994 0,962 0,971 1,024 0,957 10 0,997 0,944 0,997 1 0,941 11 1,017 0,964 1,007 1,009 0,98 12 1,018 0,964 1,009 1,009 0,981 13 1,026 1,001 1,019 1,007 1,027 14 1,027 1,045 1,017 1,01 1,074 166 15 0,988 0,888 0,989 0,999 0,878 16 1,028 0,959 1,024 1,004 0,986 17 0,976 0,911 0,983 0,993 0,89 18 0,983 0,926 0,987 0,996 0,911 19 1 1,022 1 1 1,022 20 1,049 0,95 1,04 1,009 0,997 21 1,02 0,942 1,037 0,984 0,962 22 1,019 0,955 1,001 1,018 0,974 23 1,012 0,91 1,012 1 0,921 24 1,045 0,963 1,045 1 1,006 25 1,014 0,973 1,002 1,013 0,987 26 1,037 0,947 1,038 0,999 0,982 27 1,025 0,939 1 1,025 0,962 28 1,035 0,962 1,03 1,005 0,996 mean 1,013 0,958 1,006 1,007 0,971 PHỤ LỤC 10 MÔ HÌNH HỒI QUY VÀ PHƯƠNG PHÁP KIỂM ĐỊNH ĐỐI VỚI TOÀN BỘ CÁC NGÂN HÀNG NGHIÊN CỨU VỚI BIẾN PHỤ THUỘC HQKT MÔ HÌNH POOLED VỚI BIẾN PHỤ THUỘC HQKT _cons .6889828 .1307129 5.27 0.000 .4314526 .946513 lp -.1875544 .0911962 -2.06 0.041 -.3672289 -.0078799 ttkt -.8817011 1.285676 -0.69 0.494 -3.414738 1.651335 vhdcv .0020336 .0169176 0.12 0.904 -.0312973 .0353646 cv -.1373814 .0495655 -2.77 0.006 -.2350352 -.0397275 qmts .0179817 .0060824 2.96 0.003 .0059981 .0299652 vcsh .282574 .0809409 3.49 0.001 .1231044 .4420436 tsnnt .1184907 .055666 2.13 0.034 .0088175 .2281638 cvnt -.0109582 .0246905 -0.44 0.658 -.0596035 .0376871 hqkt Coef. Std. Err. t P>|t| [95% Conf. Interval] Total 1.5846411 241 .006575274 Root MSE = .07609 Adj R-squared = 0.1195 Residual 1.34892367 233 .005789372 R-squared = 0.1488 Model .235717432 8 .029464679 Prob > F = 0.0000 F( 8, 233) = 5.09 Source SS df MS Number of obs = 242 167 MÔ HÌNH FEM VỚI BIẾN PHỤ THUỘC HQKT MÔ HÌNH REM VỚI BIẾN PHỤ THUỘC HQKT F test that all u_i=0: F(37, 196) = 3.86 Prob > F = 0.0000 rho .58619073 (fraction of variance due to u_i) sigma_e .06308982 sigma_u .0750894 _cons .0827712 .224129 0.37 0.712 -.3592429 .5247853 lp .028427 .0957356 0.30 0.767 -.1603771 .2172311 ttkt -2.317437 1.157534 -2.00 0.047 -4.600257 -.0346168 vhdcv -.0034738 .0167118 -0.21 0.836 -.0364319 .0294843 cv -.0361059 .0640442 -0.56 0.574 -.1624102 .0901983 qmts .0530703 .0120027 4.42 0.000 .0293993 .0767414 vcsh .2180074 .1009236 2.16 0.032 .0189718 .417043 tsnnt .1705981 .0911971 1.87 0.063 -.0092555 .3504517 cvnt -.0051003 .0303196 -0.17 0.867 -.0648948 .0546943 hqkt Coef. Std. Err. t P>|t| [95% Conf. Interval] corr(u_i, Xb) = -0.6432 Prob > F = 0.0001 F(8,196) = 4.34 overall = 0.0424 max = 7 between = 0.0198 avg = 6.4 R-sq: within = 0.1505 Obs per group: min = 2 Group variable: bank Number of groups = 38 Fixed-effects (within) regression Number of obs = 242 rho .34796779 (fraction of variance due to u_i) sigma_e .06308982 sigma_u .04608868 _cons .6118644 .1505694 4.06 0.000 .3167538 .906975 lp -.1293499 .0839068 -1.54 0.123 -.2938041 .0351043 ttkt -1.254605 1.114214 -1.13 0.260 -3.438425 .9292143 vhdcv -.0072557 .0159013 -0.46 0.648 -.0384217 .0239103 cv -.1321147 .0526093 -2.51 0.012 -.2352269 -.0290024 qmts .0245635 .0079107 3.11 0.002 .0090589 .0400681 vcsh .2074875 .0897825 2.31 0.021 .0315171 .3834579 tsnnt .0904065 .0690564 1.31 0.190 -.0449415 .2257546 cvnt -.0082521 .0266313 -0.31 0.757 -.0604484 .0439443 hqkt Coef. Std. Err. z P>|z| [95% Conf. Interval] corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0002 Random effects u_i ~ Gaussian Wald chi2(8) = 30.37 overall = 0.1258 max = 7 between = 0.1329 avg = 6.4 R-sq: within = 0.1144 Obs per group: min = 2 Group variable: bank Number of groups = 38 Random-effects GLS regression Number of obs = 242 168 KIỂM TRA ĐA CỘNG TUYẾN KIỂM ĐỊNH HAUSMAN KIỂM ĐỊNH BREUSCH-PAGAN LAGRANGIAN KIỂM ĐỊNH WOOLDRIDGE Mean VIF 1.74 cvnt 1.12 0.891150 ttkt 1.24 0.809133 tsnnt 1.26 0.791420 lp 1.32 0.755266 vcsh 2.02 0.494328 cv 2.22 0.449741 vhdcv 2.27 0.439852 qmts 2.47 0.405213 Variable VIF 1/VIF (V_b-V_B is not positive definite) Prob>chi2 = 0.0782 = 14.14 chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B) Test: Ho: difference in coefficients not systematic B = inconsistent under Ha, efficient under Ho; obtained from xtreg b = consistent under Ho and Ha; obtained from xtreg lp .028427 -.1293499 .1577769 .0460973 ttkt -2.317437 -1.254605 -1.062832 .3137062 vhdcv -.0034738 -.0072557 .0037819 .0051413 cv -.0361059 -.1321147 .0960087 .036523 qmts .0530703 .0245635 .0285069 .009027 vcsh .2180074 .2074875 .0105199 .0460942 tsnnt .1705981 .0904065 .0801915 .0595662 cvnt -.0051003 -.0082521 .0031518 .0144933 fe . Difference S.E. (b) (B) (b-B) sqrt(diag(V_b-V_B)) Coefficients Prob > chi2 = 0.0000 chi2(1) = 38.08 Test: Var(u) = 0 u .0021242 .0460887 e .0039803 .0630898 hqkt .0065753 .0810881 Var sd = sqrt(Var) Estimated results: hqkt[bank,t] = Xb + u[bank] + e[bank,t] Breusch and Pagan Lagrangian multiplier test for random effects Prob > F = 0.0016 F( 1, 36) = 11.597 H0: no first-order autocorrelation Wooldridge test for autocorrelation in panel data 169 MÔ HÌNH FGLS VỚI BIẾN PHỤ THUỘC HQKT PHỤ LỤC 11 MÔ HÌNH HỒI QUY VÀ PHƯƠNG PHÁP KIỂM ĐỊNH ĐỐI VỚI TOÀN BỘ CÁC NGÂN HÀNG NGHIÊN CỨU VỚI BIẾN PHỤ THUỘC HQKTT MÔ HÌNH POOLED VỚI BIẾN PHỤ THUỘC HQKTT _cons .738668 .1059039 6.97 0.000 .5311002 .9462358 lp -.0834843 .0629649 -1.33 0.185 -.2068933 .0399247 ttkt -1.735985 .8401119 -2.07 0.039 -3.382574 -.0893958 vhdcv -.0300667 .0133258 -2.26 0.024 -.0561847 -.0039487 cv -.1926906 .04281 -4.50 0.000 -.2765967 -.1087845 qmts .0222701 .0054582 4.08 0.000 .0115722 .0329679 vcsh .2398684 .0630296 3.81 0.000 .1163327 .3634041 tsnnt .1115416 .0498691 2.24 0.025 .0138 .2092832 cvnt -.0256816 .0208839 -1.23 0.219 -.0666132 .01525 hqkt Coef. Std. Err. z P>|z| [95% Conf. Interval] Prob > chi2 = 0.0000 Wald chi2(8) = 68.30 max = 7 avg = 6.368421 Estimated coefficients = 9 Obs per group: min = 2 Estimated autocorrelations = 1 Number of groups = 38 Estimated covariances = 38 Number of obs = 242 Correlation: common AR(1) coefficient for all panels (0.3785) Panels: heteroskedastic Coefficients: generalized least squares Cross-sectional time-series FGLS regression _cons .5679802 .1076613 5.28 0.000 .3558662 .7800942 lp -.009588 .0751135 -0.13 0.899 -.1575764 .1384003 ttkt -1.442498 1.058944 -1.36 0.174 -3.528826 .6438301 vhdcv .0100885 .0139341 0.72 0.470 -.0173645 .0375414 cv -.0528715 .0408245 -1.30 0.197 -.1333038 .0275608 qmts .0248042 .0050098 4.95 0.000 .014934 .0346744 vcsh .2900751 .0666668 4.35 0.000 .1587284 .4214218 tsnnt .0719499 .0458492 1.57 0.118 -.018382 .1622818 cvnt -.0124914 .0203363 -0.61 0.540 -.0525579 .0275752 hqktt Coef. Std. Err. t P>|t| [95% Conf. Interval] Total 1.08899436 241 .004518649 Root MSE = .06267 Adj R-squared = 0.1308 Residual .915102399 233 .003927478 R-squared = 0.1597 Model .17389196 8 .021736495 Prob > F = 0.0000 F( 8, 233) = 5.53 Source SS df MS Number of obs = 242 170 MÔ HÌNH FEM VỚI BIẾN PHỤ THUỘC HQKTT MÔ HÌNH REM VỚI BIẾN PHỤ THUỘC HQKTT F test that all u_i=0: F(37, 196) = 3.68 Prob > F = 0.0000 rho .47682051 (fraction of variance due to u_i) sigma_e .05248473 sigma_u .05010546 _cons .5488614 .1864541 2.94 0.004 .1811477 .9165751 lp .0068559 .0796429 0.09 0.931 -.1502112 .163923 ttkt -1.987262 .9629581 -2.06 0.040 -3.886351 -.0881724 vhdcv .0126838 .0139026 0.91 0.363 -.0147342 .0401018 cv -.0437124 .0532787 -0.82 0.413 -.1487855 .0613608 qmts .0270024 .0099851 2.70 0.007 .0073103 .0466944 vcsh .100462 .0839588 1.20 0.233 -.0651166 .2660407 tsnnt .2216493 .0758673 2.92 0.004 .0720282 .3712704 cvnt -.0002554 .025223 -0.01 0.992 -.0499988 .049488 hqktt Coef. Std. Err. t P>|t| [95% Conf. Interval] corr(u_i, Xb) = -0.4613 Prob > F = 0.0012 F(8,196) = 3.38 overall = 0.1014 max = 7 between = 0.0963 avg = 6.4 R-sq: within = 0.1212 Obs per group: min = 2 Group variable: bank Number of groups = 38 Fixed-effects (within) regression Number of obs = 242 rho .33542792 (fraction of variance due to u_i) sigma_e .05248473 sigma_u .03728735 _cons .6818037 .122974 5.54 0.000 .4407791 .9228284 lp -.0273097 .0689244 -0.40 0.692 -.162399 .1077796 ttkt -1.544875 .9169089 -1.68 0.092 -3.341983 .2522335 vhdcv .0068475 .013072 0.52 0.600 -.0187732 .0324681 cv -.0695559 .0431131 -1.61 0.107 -.154056 .0149442 qmts .0202788 .0064511 3.14 0.002 .0076348 .0329228 vcsh .1588311 .0735995 2.16 0.031 .0145787 .3030835 tsnnt .1150312 .0564203 2.04 0.041 .0044495 .2256129 cvnt -.0126792 .0218316 -0.58 0.561 -.0554683 .0301099 hqktt Coef. Std. Err. z P>|z| [95% Conf. Interval] corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0003 Random effects u_i ~ Gaussian Wald chi2(8) = 28.84 overall = 0.1394 max = 7 between = 0.1512 avg = 6.4 R-sq: within = 0.1055 Obs per group: min = 2 Group variable: bank Number of groups = 38 Random-effects GLS regression Number of obs = 242 171 KIỂM ĐỊNH HAUSMAN KIỂM ĐỊNH BREUSCH-PAGAN LAGRANGIAN KIỂM ĐỊNH WOOLDRIDGE (V_b-V_B is not positive definite) Prob>chi2 = 0.2888 = 9.67 chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B) Test: Ho: difference in coefficients not systematic B = inconsistent under Ha, efficient under Ho; obtained from xtreg b = consistent under Ho and Ha; obtained from xtreg lp .0068559 -.0273097 .0341656 .0399052 ttkt -1.987262 -1.544875 -.4423868 .2942216 vhdcv .0126838 .0068475 .0058364 .0047335 cv -.0437124 -.0695559 .0258435 .0313031 qmts .0270024 .0202788 .0067236 .0076214 vcsh .100462 .1588311 -.0583691 .0404004 tsnnt .2216493 .1150312 .1066181 .0507208 cvnt -.0002554 -.0126792 .0124238 .0126326 fe . Difference S.E. (b) (B) (b-B) sqrt(diag(V_b-V_B)) Coefficients Prob > chi2 = 0.0000 chi2(1) = 40.35 Test: Var(u) = 0 u .0013903 .0372873 e .0027546 .0524847 hqktt .0045186 .0672209 Var sd = sqrt(Var) Estimated results: hqktt[bank,t] = Xb + u[bank] + e[bank,t] Breusch and Pagan Lagrangian multiplier test for random effects Prob > F = 0.0023 F( 1, 36) = 10.724 H0: no first-order autocorrelation Wooldridge test for autocorrelation in panel data 172 MÔ HÌNH FGLS VỚI BIẾN PHỤ THUỘC HQKTT PHỤ LỤC 12 MÔ HÌNH HỒI QUY VÀ PHƯƠNG PHÁP KIỂM ĐỊNH ĐỐI VỚI TOÀN BỘ CÁC NGÂN HÀNG NGHIÊN CỨU VỚI BIẾN PHỤ THUỘC HQQM MÔ HÌNH POOLED VỚI BIẾN PHỤ THUỘC HQQM _cons .66676 .0768905 8.67 0.000 .5160574 .8174627 lp -.0017012 .039563 -0.04 0.966 -.0792432 .0758409 ttkt -.8488412 .5445592 -1.56 0.119 -1.916158 .2184753 vhdcv .0059352 .0092168 0.64 0.520 -.0121293 .0239998 cv -.0320864 .0331425 -0.97 0.333 -.0970444 .0328717 qmts .0182934 .004281 4.27 0.000 .0099028 .0266839 vcsh .1634918 .0519448 3.15 0.002 .0616818 .2653018 tsnnt .0671865 .0327671 2.05 0.040 .0029642 .1314089 cvnt -.0004067 .0167505 -0.02 0.981 -.033237 .0324236 hqktt Coef. Std. Err. z P>|z| [95% Conf. Interval] Prob > chi2 = 0.0000 Wald chi2(8) = 39.09 max = 7 avg = 6.368421 Estimated coefficients = 9 Obs per group: min = 2 Estimated autocorrelations = 1 Number of groups = 38 Estimated covariances = 38 Number of obs = 242 Correlation: common AR(1) coefficient for all panels (0.4936) Panels: heteroskedastic Coefficients: generalized least squares Cross-sectional time-series FGLS regression _cons 1.095848 .0837498 13.08 0.000 .9308439 1.260851 lp -.1898781 .0584308 -3.25 0.001 -.3049984 -.0747579 ttkt .5987865 .8237527 0.73 0.468 -1.024169 2.221742 vhdcv -.0061593 .0108394 -0.57 0.570 -.027515 .0151963 cv -.0853608 .0317574 -2.69 0.008 -.1479291 -.0227925 qmts -.005817 .0038971 -1.49 0.137 -.013495 .0018611 vcsh .0048439 .0518601 0.09 0.926 -.0973307 .1070186 tsnnt .0520516 .0356661 1.46 0.146 -.0182177 .1223209 cvnt .002265 .0158196 0.14 0.886 -.0289027 .0334328 hqqm Coef. Std. Err. t P>|t| [95% Conf. Interval] Total .61096355 241 .002535118 Root MSE = .04875 Adj R-squared = 0.0625 Residual .553755128 233 .002376631 R-squared = 0.0936 Model .057208422 8 .007151053 Prob > F = 0.0031 F( 8, 233) = 3.01 Source SS df MS Number of obs = 242 173 MÔ HÌNH FEM VỚI BIẾN PHỤ THUỘC HQQM MÔ HÌNH REM VỚI BIẾN PHỤ THUỘC HQQM F test that all u_i=0: F(37, 196) = 4.31 Prob > F = 0.0000 rho .61692073 (fraction of variance due to u_i) sigma_e .03946075 sigma_u .0500767 _cons .4886467 .1401859 3.49 0.001 .2121804 .765113 lp .0165544 .0598797 0.28 0.782 -.1015368 .1346456 ttkt -.3513773 .724002 -0.49 0.628 -1.779211 1.076457 vhdcv -.0149486 .0104527 -1.43 0.154 -.0355629 .0056656 cv .0087397 .0400577 0.22 0.828 -.0702597 .0877392 qmts .028432 .0075073 3.79 0.000 .0136265 .0432375 vcsh .1273834 .0631246 2.02 0.045 .0028928 .2518741 tsnnt -.0435729 .057041 -0.76 0.446 -.1560658 .06892 cvnt -.0070343 .018964 -0.37 0.711 -.044434 .0303654 hqqm Coef. Std. Err. t P>|t| [95% Conf. Interval] corr(u_i, Xb) = -0.7022 Prob > F = 0.0030 F(8,196) = 3.04 overall = 0.0165 max = 7 between = 0.1345 avg = 6.4 R-sq: within = 0.1105 Obs per group: min = 2 Group variable: bank Number of groups = 38 Fixed-effects (within) regression Number of obs = 242 rho .29355267 (fraction of variance due to u_i) sigma_e .03946075 sigma_u .02543715 _cons .9301615 .0938172 9.91 0.000 .7462831 1.11404 lp -.1225975 .0536342 -2.29 0.022 -.2277186 -.0174764 ttkt .3552342 .7178766 0.49 0.621 -1.051778 1.762246 vhdcv -.0122304 .0101948 -1.20 0.230 -.0322117 .007751 cv -.0682715 .0332611 -2.05 0.040 -.133462 -.003081 qmts .0040047 .0048934 0.82 0.413 -.0055863 .0135956 vcsh .0535347 .0568113 0.94 0.346 -.0578134 .1648829 tsnnt -.0132859 .0430567 -0.31 0.758 -.0976756 .0711037 cvnt .0039679 .016857 0.24 0.814 -.0290711 .037007 hqqm Coef. Std. Err. z P>|z| [95% Conf. Interval] corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.1318 Random effects u_i ~ Gaussian Wald chi2(8) = 12.46 overall = 0.0509 max = 7 between = 0.0734 avg = 6.4 R-sq: within = 0.0490 Obs per group: min = 2 Group variable: bank Number of groups = 38 Random-effects GLS regression Number of obs = 242 174 KIỂM ĐỊNH HAUSMAN KIỂM ĐỊNH MODIFIED WALD KIỂM ĐỊNH WOOLDRIDGE (V_b-V_B is not positive definite) Prob>chi2 = 0.0002 = 29.96 chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B) Test: Ho: difference in coefficients not systematic B = inconsistent under Ha, efficient under Ho; obtained from xtreg b = consistent under Ho and Ha; obtained from xtreg lp .0165544 -.1225975 .1391519 .0266261 ttkt -.3513773 .3552342 -.7066115 .0939796 vhdcv -.0149486 -.0122304 -.0027182 .0023079 cv .0087397 -.0682715 .0770112 .0223231 qmts .028432 .0040047 .0244273 .0056933 vcsh .1273834 .0535347 .0738487 .0275171 tsnnt -.0435729 -.0132859 -.030287 .0374138 cvnt -.0070343 .0039679 -.0110022 .0086877 fe . Difference S.E. (b) (B) (b-B) sqrt(diag(V_b-V_B)) Coefficients Prob>chi2 = 0.0000 chi2 (38) = 4880.16 H0: sigma(i)^2 = sigma^2 for all i in fixed effect regression model Modified Wald test for groupwise heteroskedasticity Prob > F = 0.0377 F( 1, 36) = 4.653 H0: no first-order autocorrelation Wooldridge test for autocorrelation in panel data 175 MÔ HÌNH FGLS VỚI BIẾN PHỤ THUỘC HQQM PHỤ LỤC 13 MÔ HÌNH HỒI QUY VÀ PHƯƠNG PHÁP KIỂM ĐỊNH ĐỐI VỚI CÁC NGÂN HÀNG NHÓM 1 VỚI BIẾN PHỤ THUỘC HQKT MÔ HÌNH POOLED VỚI BIẾN PHỤ THUỘC HQKT _cons .981067 .0646598 15.17 0.000 .8543361 1.107798 lp -.0696219 .032509 -2.14 0.032 -.1333385 -.0059054 ttkt -.1710881 .4381833 -0.39 0.696 -1.029912 .6877354 vhdcv -.0187937 .0071428 -2.63 0.009 -.0327933 -.004794 cv -.082662 .020907 -3.95 0.000 -.123639 -.041685 qmts .0048155 .0034368 1.40 0.161 -.0019205 .0115514 vcsh .028787 .0306177 0.94 0.347 -.0312225 .0887965 tsnnt -.0317596 .0279002 -1.14 0.255 -.086443 .0229237 cvnt -.0062644 .010491 -0.60 0.550 -.0268265 .0142976 hqqm Coef. Std. Err. z P>|z| [95% Conf. Interval] Prob > chi2 = 0.0013 Wald chi2(8) = 25.45 max = 7 avg = 6.368421 Estimated coefficients = 9 Obs per group: min = 2 Estimated autocorrelations = 1 Number of groups = 38 Estimated covariances = 38 Number of obs = 242 Correlation: common AR(1) coefficient for all panels (0.3870) Panels: heteroskedastic Coefficients: generalized least squares Cross-sectional time-series FGLS regression _cons .9323893 .3080593 3.03 0.004 .3175004 1.547278 lp -.2211092 .1445136 -1.53 0.131 -.5095595 .0673411 ttkt -.828911 2.05012 -0.40 0.687 -4.920968 3.263147 vhdcv -.0500903 .0443562 -1.13 0.263 -.1386255 .038445 cv -.2337826 .1360365 -1.72 0.090 -.5053126 .0377473 qmts .0142129 .0159639 0.89 0.376 -.0176511 .046077 vcsh .2103276 .3060629 0.69 0.494 -.4005765 .8212317 tsnnt .0816045 .0941882 0.87 0.389 -.106396 .269605 cvnt -.0989029 .0520511 -1.90 0.062 -.2027974 .0049916 hqkt Coef. Std. Err. t P>|t| [95% Conf. Interval] Total .380690676 75 .005075876 Root MSE = .06717 Adj R-squared = 0.1110 Residual .302322451 67 .004512275 R-squared = 0.2059 Model .078368224 8 .009796028 Prob > F = 0.0408 F( 8, 67) = 2.17 Source SS df MS Number of obs = 76 176 MÔ HÌNH FEM VỚI BIẾN PHỤ THUỘC HQKT MÔ HÌNH REM VỚI BIẾN PHỤ THUỘC HQKT F test that all u_i=0: F(10, 57) = 2.02 Prob > F = 0.0478 rho .41890269 (fraction of variance due to u_i) sigma_e .06257961 sigma_u .05313308 _cons .2218203 .4600037 0.48 0.632 -.6993212 1.142962 lp -.0454251 .1869112 -0.24 0.809 -.4197082 .3288581 ttkt -1.343945 2.001207 -0.67 0.505 -5.351294 2.663404 vhdcv -.0215857 .0657215 -0.33 0.744 -.1531908 .1100194 cv .0131097 .2324369 0.06 0.955 -.4523372 .4785566 qmts .0393995 .0217065 1.82 0.075 -.0040671 .0828661 vcsh .4966837 .386375 1.29 0.204 -.277019 1.270386 tsnnt .2122024 .1683688 1.26 0.213 -.1249503 .5493551 cvnt -.0564119 .0973935 -0.58 0.565 -.2514391 .1386153 hqkt Coef. Std. Err. t P>|t| [95% Conf. Interval] corr(u_i, Xb) = -0.5106 Prob > F = 0.3757 F(8,57) = 1.10 overall = 0.0398 max = 7 between = 0.0058 avg = 6.9 R-sq: within = 0.1339 Obs per group: min = 6 Group variable: bank Number of groups = 11 Fixed-effects (within) regression Number of obs = 76 rho .22720711 (fraction of variance due to u_i) sigma_e .06257961 sigma_u .03393222 _cons .747534 .327944 2.28 0.023 .1047756 1.390292 lp -.1913941 .1431033 -1.34 0.181 -.4718713 .0890832 ttkt -.7627531 1.912856 -0.40 0.690 -4.511881 2.986375 vhdcv -.0328833 .0516899 -0.64 0.525 -.1341936 .0684271 cv -.1716585 .1553412 -1.11 0.269 -.4761216 .1328047 qmts .0192601 .0167431 1.15 0.250 -.0135557 .052076 vcsh .3794709 .3222557 1.18 0.239 -.2521388 1.011081 tsnnt .0990759 .1126068 0.88 0.379 -.1216294 .3197811 cvnt -.0827722 .0652355 -1.27 0.205 -.2106315 .045087 hqkt Coef. Std. Err. z P>|z| [95% Conf. Interval] corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.1882 Random effects u_i ~ Gaussian Wald chi2(8) = 11.25 overall = 0.1938 max = 7 between = 0.4036 avg = 6.9 R-sq: within = 0.1044 Obs per group: min = 6 Group variable: bank Number of groups = 11 Random-effects GLS regression Number of obs = 76 177 KIỂM TRA ĐA CỘNG TUYẾN KIỂM ĐỊNH HAUSMAN KIỂM ĐỊNH BREUSCH-PAGAN LAGRANGIAN KIỂM ĐỊNH WOOLDRIDGE Mean VIF 2.20 ttkt 1.26 0.796614 lp 1.35 0.741087 cvnt 1.37 0.731533 tsnnt 1.54 0.649014 vcsh 1.64 0.610306 qmts 2.81 0.355443 vhdcv 3.05 0.328400 cv 4.57 0.218956 Variable VIF 1/VIF Prob>chi2 = 0.8323 = 4.27 chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B) Test: Ho: difference in coefficients not systematic B = inconsistent under Ha, efficient under Ho; obtained from xtreg b = consistent under Ho and Ha; obtained from xtreg lp -.0454251 -.1913941 .145969 .1202382 ttkt -1.343945 -.7627531 -.5811919 .5880593 vhdcv -.0215857 -.0328833 .0112976 .0405891 cv .0131097 -.1716585 .1847682 .1729047 qmts .0393995 .0192601 .0201394 .0138146 vcsh .4966837 .3794709 .1172128 .2131593 tsnnt .2122024 .0990759 .1131265 .1251709 cvnt -.0564119 -.0827722 .0263603 .0723176 fe . Difference S.E. (b) (B) (b-B) sqrt(diag(V_b-V_B)) Coefficients Prob > chi2 = 0.2140 chi2(1) = 1.54 Test: Var(u) = 0 u .0011514 .0339322 e .0039162 .0625796 hqkt .0050759 .0712452 Var sd = sqrt(Var) Estimated results: hqkt[bank,t] = Xb + u[bank] + e[bank,t] Breusch and Pagan Lagrangian multiplier test for random effects Prob > F = 0.0184 F( 1, 10) = 7.906 H0: no first-order autocorrelation Wooldridge test for autocorrelation in panel data 178 MÔ HÌNH FGLS VỚI BIẾN PHỤ THUỘC HQKT PHỤ LỤC 14 MÔ HÌNH HỒI QUY VÀ PHƯƠNG PHÁP KIỂM ĐỊNH ĐỐI VỚI CÁC NGÂN HÀNG NHÓM 1 VỚI BIẾN PHỤ THUỘC HQKTT MÔ HÌNH POOLED VỚI BIẾN PHỤ THUỘC HQKTT _cons .9456188 .3313764 2.85 0.004 .296133 1.595105 lp -.148855 .122328 -1.22 0.224 -.3886134 .0909034 ttkt -.2971461 1.699891 -0.17 0.861 -3.628871 3.034578 vhdcv -.0252933 .0468556 -0.54 0.589 -.1171285 .0665419 cv -.0732797 .1439377 -0.51 0.611 -.3553924 .2088331 qmts .0056908 .0171035 0.33 0.739 -.0278316 .0392131 vcsh .1258016 .3243871 0.39 0.698 -.5099855 .7615887 tsnnt .1280488 .1045611 1.22 0.221 -.0768872 .3329847 cvnt -.1214874 .0564355 -2.15 0.031 -.232099 -.0108758 hqkt Coef. Std. Err. z P>|z| [95% Conf. Interval] Prob > chi2 = 0.1161 Wald chi2(8) = 12.88 max = 7 avg = 6.909091 Estimated coefficients = 9 Obs per group: min = 6 Estimated autocorrelations = 1 Number of groups = 11 Estimated covariances = 1 Number of obs = 76 Correlation: common AR(1) coefficient for all panels (0.4041) Panels: homoskedastic Coefficients: generalized least squares Cross-sectional time-series FGLS regression _cons .6450415 .2168671 2.97 0.004 .212173 1.07791 lp .1640911 .1017344 1.61 0.111 -.0389716 .3671539 ttkt -.7535215 1.44324 -0.52 0.603 -3.634241 2.127199 vhdcv .0090185 .0312258 0.29 0.774 -.0533084 .0713454 cv .1102052 .0957668 1.15 0.254 -.080946 .3013564 qmts .015462 .0112382 1.38 0.173 -.0069696 .0378937 vcsh -.1156767 .2154617 -0.54 0.593 -.54574 .3143866 tsnnt .1540297 .0663065 2.32 0.023 .0216814 .286378 cvnt -.0726497 .0366429 -1.98 0.052 -.1457892 .0004898 hqktt Coef. Std. Err. t P>|t| [95% Conf. Interval] Total .201589945 75 .002687866 Root MSE = .04729 Adj R-squared = 0.1680 Residual .149826638 67 .002236218 R-squared = 0.2568 Model .051763307 8 .006470413 Prob > F = 0.0079 F( 8, 67) = 2.89 Source SS df MS Number of obs = 76 179 MÔ HÌNH FEM VỚI BIẾN PHỤ THUỘC HQKTT MÔ HÌNH REM VỚI BIẾN PHỤ THUỘC HQKTT F test that all u_i=0: F(10, 57) = 3.57 Prob > F = 0.0010 rho .5156108 (fraction of variance due to u_i) sigma_e .04021022 sigma_u .04148588 _cons .8733539 .2955731 2.95 0.005 .2814789 1.465229 lp -.0244686 .1200989 -0.20 0.839 -.2649624 .2160252 ttkt .5601288 1.285866 0.44 0.665 -2.014774 3.135031 vhdcv .0448697 .0422291 1.06 0.292 -.0396925 .129432 cv .1355143 .1493512 0.91 0.368 -.1635564 .434585 qmts -.003978 .0139474 -0.29 0.777 -.0319072 .0239513 vcsh -.0269137 .2482634 -0.11 0.914 -.5240526 .4702252 tsnnt .3023456 .1081845 2.79 0.007 .0857097 .5189814 cvnt -.09563 .0625797 -1.53 0.132 -.2209438 .0296837 hqktt Coef. Std. Err. t P>|t| [95% Conf. Interval] corr(u_i, Xb) = -0.4874 Prob > F = 0.0032 F(8,57) = 3.37 overall = 0.1043 max = 7 between = 0.0001 avg = 6.9 R-sq: within = 0.3208 Obs per group: min = 6 Group variable: bank Number of groups = 11 Fixed-effects (within) regression Number of obs = 76 rho .30099115 (fraction of variance due to u_i) sigma_e .04021022 sigma_u .02638591 _cons .6976899 .2248918 3.10 0.002 .2569102 1.13847 lp .096351 .0968797 0.99 0.320 -.0935298 .2862318 ttkt -.0699193 1.270445 -0.06 0.956 -2.559946 2.420107 vhdcv .0346993 .0356912 0.97 0.331 -.0352542 .1046528 cv .1996138 .1078 1.85 0.064 -.0116703 .4108979 qmts .0062579 .011392 0.55 0.583 -.0160701 .0285858 vcsh -.0874461 .218971 -0.40 0.690 -.5166214 .3417291 tsnnt .2441341 .0787878 3.10 0.002 .0897128 .3985553 cvnt -.0978926 .0458293 -2.14 0.033 -.1877165 -.0080688 hqktt Coef. Std. Err. z P>|z| [95% Conf. Interval] corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0023 Random effects u_i ~ Gaussian Wald chi2(8) = 23.94 overall = 0.2224 max = 7 between = 0.1782 avg = 6.9 R-sq: within = 0.2985 Obs per group: min = 6 Group variable: bank Number of groups = 11 Random-effects GLS regression Number of obs = 76 180 KIỂM ĐỊNH HAUSMAN KIỂM ĐỊNH BREUSCH-PAGAN LAGRANGIAN KIỂM ĐỊNH WOOLDRIDGE (V_b-V_B is not positive definite) Prob>chi2 = 0.0683 = 14.56 chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B) Test: Ho: difference in coefficients not systematic B = inconsistent under Ha, efficient under Ho; obtained from xtreg b = consistent under Ho and Ha; obtained from xtreg lp -.0244686 .096351 -.1208196 .0709793 ttkt .5601288 -.0699193 .6300481 .1985466 vhdcv .0448697 .0346993 .0101705 .0225706 cv .1355143 .1996138 -.0640995 .103368 qmts -.003978 .0062579 -.0102358 .0080469 vcsh -.0269137 -.0874461 .0605324 .116989 tsnnt .3023456 .2441341 .0582115 .0741376 cvnt -.09563 -.0978926 .0022626 .0426133 fe . Difference S.E. (b) (B) (b-B) sqrt(diag(V_b-V_B)) Coefficients Prob > chi2 = 0.0173 chi2(1) = 5.67 Test: Var(u) = 0 u .0006962 .0263859 e .0016169 .0402102 hqktt .0026879 .0518446 Var sd = sqrt(Var) Estimated results: hqktt[bank,t] = Xb + u[bank] + e[bank,t] Breusch and Pagan Lagrangian multiplier test for random effects Prob > F = 0.0555 F( 1, 10) = 4.695 H0: no first-order autocorrelation Wooldridge test for autocorrelation in panel data 181 MÔ HÌNH FGLS VỚI BIẾN PHỤ THUỘC HQKTT PHỤ LỤC 15 MÔ HÌNH HỒI QUY VÀ PHƯƠNG PHÁP KIỂM ĐỊNH ĐỐI VỚI CÁC NGÂN HÀNG NHÓM 1 VỚI BIẾN PHỤ THUỘC HQQM MÔ HÌNH POOLED VỚI BIẾN PHỤ THUỘC HQQM _cons .5903455 .1625765 3.63 0.000 .2717014 .9089896 lp .0901457 .0589896 1.53 0.126 -.0254719 .2057632 ttkt -.6161471 .7932508 -0.78 0.437 -2.17089 .9385959 vhdcv .0008527 .0292233 0.03 0.977 -.0564239 .0581293 cv .0685118 .0787681 0.87 0.384 -.0858708 .2228944 qmts .0199237 .0082772 2.41 0.016 .0037007 .0361468 vcsh -.0913024 .1805952 -0.51 0.613 -.4452625 .2626578 tsnnt .0847568 .0371928 2.28 0.023 .0118601 .1576534 cvnt -.0367651 .0285121 -1.29 0.197 -.0926478 .0191176 hqktt Coef. Std. Err. z P>|z| [95% Conf. Interval] Prob > chi2 = 0.0002 Wald chi2(8) = 30.29 max = 7 avg = 6.909091 Estimated coefficients = 9 Obs per group: min = 6 Estimated autocorrelations = 0 Number of groups = 11 Estimated covariances = 11 Number of obs = 76 Correlation: no autocorrelation Panels: heteroskedastic Coefficients: generalized least squares Cross-sectional time-series FGLS regression _cons 1.281242 .2435447 5.26 0.000 .7951248 1.767359 lp -.3872294 .1142492 -3.39 0.001 -.6152717 -.1591871 ttkt -.0767315 1.620779 -0.05 0.962 -3.311819 3.158356 vhdcv -.060267 .035067 -1.72 0.090 -.130261 .009727 cv -.3491417 .1075474 -3.25 0.002 -.5638072 -.1344763 qmts -.0006776 .0126207 -0.05 0.957 -.0258686 .0245134 vcsh .3245717 .2419664 1.34 0.184 -.1583952 .8075387 tsnnt -.0731784 .0744631 -0.98 0.329 -.2218074 .0754505 cvnt -.0276023 .0411504 -0.67 0.505 -.109739 .0545343 hqqm Coef. Std. Err. t P>|t| [95% Conf. Interval] Total .2952914 75 .003937219 Root MSE = .05311 Adj R-squared = 0.2837 Residual .188955311 67 .002820229 R-squared = 0.3601 Model .10633609 8 .013292011 Prob > F = 0.0001 F( 8, 67) = 4.71 Source SS df MS Number of obs = 76 182 MÔ HÌNH FEM VỚI BIẾN PHỤ THUỘC HQQM MÔ HÌNH REM VỚI BIẾN PHỤ THUỘC HQQM F test that all u_i=0: F(10, 57) = 3.56 Prob > F = 0.0010 rho .70804565 (fraction of variance due to u_i) sigma_e .04518342 sigma_u .07036428 _cons .3529386 .3321296 1.06 0.292 -.3121394 1.018017 lp -.030438 .1349527 -0.23 0.822 -.300676 .2398001 ttkt -1.856164 1.444902 -1.28 0.204 -4.74953 1.037202 vhdcv -.0645584 .0474519 -1.36 0.179 -.1595793 .0304624 cv -.1267759 .167823 -0.76 0.453 -.4628355 .2092838 qmts .0430017 .0156724 2.74 0.008 .0116181 .0743852 vcsh .5209135 .2789686 1.87 0.067 -.0377114 1.079538 tsnnt -.0877707 .1215648 -0.72 0.473 -.3312 .1556586 cvnt .0387694 .0703196 0.55 0.584 -.1020432 .179582 hqqm Coef. Std. Err. t P>|t| [95% Conf. Interval] corr(u_i, Xb) = -0.6895 Prob > F = 0.0258 F(8,57) = 2.41 overall = 0.0228 max = 7 between = 0.5275 avg = 6.9 R-sq: within = 0.2528 Obs per group: min = 6 Group variable: bank Number of groups = 11 Fixed-effects (within) regression Number of obs = 76 rho .02033747 (fraction of variance due to u_i) sigma_e .04518342 sigma_u .00651012 _cons 1.251849 .2447428 5.11 0.000 .7721616 1.731536 lp -.3777144 .1135805 -3.33 0.001 -.6003281 -.1551008 ttkt -.1336482 1.602445 -0.08 0.934 -3.274382 3.007085 vhdcv -.0606122 .0357906 -1.69 0.090 -.1307605 .0095361 cv -.3542843 .1092018 -3.24 0.001 -.5683159 -.1402527 qmts .0010299 .0126876 0.08 0.935 -.0238372 .0258971 vcsh .3495524 .2433472 1.44 0.151 -.1273994 .8265043 tsnnt -.082754 .0760564 -1.09 0.277 -.2318218 .0663138 cvnt -.0229074 .0424005 -0.54 0.589 -.1060109 .0601961 hqqm Coef. Std. Err. z P>|z| [95% Conf. Interval] corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 Random effects u_i ~ Gaussian Wald chi2(8) = 34.38 overall = 0.3596 max = 7 between = 0.7384 avg = 6.9 R-sq: within = 0.0744 Obs per group: min = 6 Group variable: bank Number of groups = 11 Random-effects GLS regression Number of obs = 76 183 KIỂM ĐỊNH HAUSMAN KIỂM ĐỊNH BREUSCH-PAGAN LAGRANGIAN KIỂM ĐỊNH WOOLDRIDGE (V_b-V_B is not positive definite) Prob>chi2 = 0.2865 = 9.70 chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B) Test: Ho: difference in coefficients not systematic B = inconsistent under Ha, efficient under Ho; obtained from xtreg b = consistent under Ho and Ha; obtained from xtreg lp -.030438 -.3777144 .3472765 .0728814 ttkt -1.856164 -.1336482 -1.722516 . vhdcv -.0645584 -.0606122 -.0039462 .0311564 cv -.1267759 -.3542843 .2275084 .1274343 qmts .0430017 .0010299 .0419717 .0092006 vcsh .5209135 .3495524 .1713611 .1364023 tsnnt -.0877707 -.082754 -.0050167 .0948337 cvnt .0387694 -.0229074 .0616768 .0560985 fe . Difference S.E. (b) (B) (b-B) sqrt(diag(V_b-V_B)) Coefficients Prob > chi2 = 0.3333 chi2(1) = 0.94 Test: Var(u) = 0 u .0000424 .0065101 e .0020415 .0451834 hqqm .0039372 .0627473 Var sd = sqrt(Var) Estimated results: hqqm[bank,t] = Xb + u[bank] + e[bank,t] Breusch and Pagan Lagrangian multiplier test for random effects Prob > F = 0.0709 F( 1, 10) = 4.084 H0: no first-order autocorrelation Wooldridge test for autocorrelation in panel data 184 PHỤ LỤC 16 MÔ HÌNH HỒI QUY VÀ PHƯƠNG PHÁP KIỂM ĐỊNH ĐỐI VỚI CÁC NGÂN HÀNG NHÓM 2 VỚI BIẾN PHỤ THUỘC HQKT MÔ HÌNH POOLED VỚI BIẾN PHỤ THUỘC HQKT MÔ HÌNH FEM VỚI BIẾN PHỤ THUỘC HQKT _cons .2611162 .219746 1.19 0.237 -.1729237 .6951562 lp -.0753591 .1225861 -0.61 0.540 -.3174898 .1667716 ttkt -1.448569 1.645688 -0.88 0.380 -4.699113 1.801975 vhdcv .0077257 .0188011 0.41 0.682 -.0294101 .0448615 cv -.0780945 .058642 -1.33 0.185 -.1939235 .0377344 qmts .0409463 .011467 3.57 0.000 .0182969 .0635957 vcsh .401728 .09948 4.04 0.000 .2052363 .5982197 tsnnt .1018308 .0789949 1.29 0.199 -.0541992 .2578608 cvnt .0075779 .0287573 0.26 0.793 -.0492233 .064379 hqkt Coef. Std. Err. t P>|t| [95% Conf. Interval] Total 1.20102703 165 .007278952 Root MSE = .07921 Adj R-squared = 0.1380 Residual .985085803 157 .006274432 R-squared = 0.1798 Model .215941225 8 .026992653 Prob > F = 0.0001 F( 8, 157) = 4.30 Source SS df MS Number of obs = 166 F test that all u_i=0: F(26, 131) = 4.11 Prob > F = 0.0000 rho .47981474 (fraction of variance due to u_i) sigma_e .06437032 sigma_u .06182206 _cons -.0083122 .2642345 -0.03 0.975 -.531031 .5144067 lp .0718645 .1178646 0.61 0.543 -.1612998 .3050288 ttkt -2.618872 1.468328 -1.78 0.077 -5.523576 .2858311 vhdcv -.0056092 .0178845 -0.31 0.754 -.0409891 .0297707 cv -.0287132 .0702489 -0.41 0.683 -.1676822 .1102558 qmts .0611276 .0148774 4.11 0.000 .0316964 .0905587 vcsh .2179748 .1112491 1.96 0.052 -.0021024 .438052 tsnnt .1132769 .1200583 0.94 0.347 -.1242272 .3507809 cvnt -.0032652 .0332665 -0.10 0.922 -.0690743 .0625439 hqkt Coef. Std. Err. t P>|t| [95% Conf. Interval] corr(u_i, Xb) = -0.3136 Prob > F = 0.0009 F(8,131) = 3.55 overall = 0.0963 max = 7 between = 0.0762 avg = 6.1 R-sq: within = 0.1784 Obs per group: min = 2 Group variable: bank Number of groups = 27 Fixed-effects (within) regression Number of obs = 166 185 MÔ HÌNH REM VỚI BIẾN PHỤ THUỘC HQKT KIỂM TRA ĐA CỘNG TUYẾN KIỂM ĐỊNH HAUSMAN rho .41184216 (fraction of variance due to u_i) sigma_e .06437032 sigma_u .05386469 _cons .1845554 .2282526 0.81 0.419 -.2628114 .6319222 lp -.0012097 .109867 -0.01 0.991 -.216545 .2141256 ttkt -2.021809 1.403096 -1.44 0.150 -4.771826 .7282084 vhdcv -.0056723 .0169017 -0.34 0.737 -.038799 .0274544 cv -.0677006 .0613351 -1.10 0.270 -.1879152 .052514 qmts .0494749 .0126666 3.91 0.000 .0246487 .074301 vcsh .2630833 .1021426 2.58 0.010 .0628876 .4632791 tsnnt .0654087 .0950758 0.69 0.491 -.1209363 .2517538 cvnt -.0008519 .0296573 -0.03 0.977 -.0589791 .0572753 hqkt Coef. Std. Err. z P>|z| [95% Conf. Interval] corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0002 Random effects u_i ~ Gaussian Wald chi2(8) = 29.75 overall = 0.1414 max = 7 between = 0.1369 avg = 6.1 R-sq: within = 0.1674 Obs per group: min = 2 Group variable: bank Number of groups = 27 Random-effects GLS regression Number of obs = 166 Mean VIF 1.84 cvnt 1.07 0.938586 tsnnt 1.18 0.848270 ttkt 1.29 0.777204 lp 1.51 0.662633 cv 2.13 0.470041 vhdcv 2.20 0.455169 vcsh 2.36 0.422991 qmts 2.98 0.335159 Variable VIF 1/VIF Prob>chi2 = 0.6646 = 5.85 chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B) Test: Ho: difference in coefficients not systematic B = inconsistent under Ha, efficient under Ho; obtained from xtreg b = consistent under Ho and Ha; obtained from xtreg lp .0718645 -.0012097 .0730742 .0426768 ttkt -2.618872 -2.021809 -.5970634 .4327928 vhdcv -.0056092 -.0056723 .0000631 .0058472 cv -.0287132 -.0677006 .0389874 .0342477 qmts .0611276 .0494749 .0116527 .0078035 vcsh .2179748 .2630833 -.0451086 .0440823 tsnnt .1132769 .0654087 .0478681 .0733117 cvnt -.0032652 -.0008519 -.0024133 .01507 fe . Difference S.E. (b) (B) (b-B) sqrt(diag(V_b-V_B)) Coefficients 186 KIỂM ĐỊNH BREUSCH-PAGAN LAGRANGIAN KIỂM ĐỊNH WOOLDRIDGE MÔ HÌNH FGLS VỚI BIẾN PHỤ THUỘC HQKT Prob > chi2 = 0.0000 chi2(1) = 32.11 Test: Var(u) = 0 u .0029014 .0538647 e .0041435 .0643703 hqkt .007279 .0853168 Var sd = sqrt(Var) Estimated results: hqkt[bank,t] = Xb + u[bank] + e[bank,t] Breusch and Pagan Lagrangian multiplier test for random effects Prob > F = 0.0041 F( 1, 25) = 9.990 H0: no first-order autocorrelation Wooldridge test for autocorrelation in panel data _cons .3921823 .1698437 2.31 0.021 .0592948 .7250698 lp -.0207302 .083462 -0.25 0.804 -.1843128 .1428524 ttkt -2.440381 1.068111 -2.28 0.022 -4.533841 -.3469215 vhdcv -.0200198 .0133082 -1.50 0.132 -.0461033 .0060638 cv -.1511831 .0475824 -3.18 0.001 -.2444428 -.0579233 qmts .0414814 .0088869 4.67 0.000 .0240634 .0588994 vcsh .3038906 .0789031 3.85 0.000 .1492435 .4585378 tsnnt .1230141 .0711734 1.73 0.084 -.0164832 .2625114 cvnt .0102921 .0236386 0.44 0.663 -.0360388 .0566229 hqkt Coef. Std. Err. z P>|z| [95% Conf. Interval] Prob > chi2 = 0.0000 Wald chi2(8) = 54.23 max = 7 avg = 6.148148 Estimated coefficients = 9 Obs per group: min = 2 Estimated autocorrelations = 1 Number of groups = 27 Estimated covariances = 27 Number of obs = 166 Correlation: common AR(1) coefficient for all panels (0.3824) Panels: heteroskedastic Coefficients: generalized least squares Cross-sectional time-series FGLS regression 187 PHỤ LỤC 17 MÔ HÌNH HỒI QUY VÀ PHƯƠNG PHÁP KIỂM ĐỊNH ĐỐI VỚI CÁC NGÂN HÀNG NHÓM 2 VỚI BIẾN PHỤ THUỘC HQKTT MÔ HÌNH POOLED VỚI BIẾN PHỤ THUỘC HQKTT MÔ HÌNH FEM VỚI BIẾN PHỤ THUỘC HQKTT _cons .7290852 .1876258 3.89 0.000 .3584888 1.099682 lp -.120405 .1046677 -1.15 0.252 -.3271435 .0863335 ttkt -1.550119 1.405138 -1.10 0.272 -4.325532 1.225294 vhdcv .0117523 .016053 0.73 0.465 -.0199554 .0434599 cv -.0840328 .0500703 -1.68 0.095 -.1829311 .0148654 qmts .0170754 .0097908 1.74 0.083 -.0022634 .0364141 vcsh .26818 .084939 3.16 0.002 .1004094 .4359506 tsnnt .0761245 .0674483 1.13 0.261 -.0570987 .2093476 cvnt -.00178 .0245539 -0.07 0.942 -.0502786 .0467185 hqktt Coef. Std. Err. t P>|t| [95% Conf. Interval] Total .839049457 165 .005085148 Root MSE = .06763 Adj R-squared = 0.1005 Residual .718153365 157 .004574225 R-squared = 0.1441 Model .120896092 8 .015112012 Prob > F = 0.0016 F( 8, 157) = 3.30 Source SS df MS Number of obs = 166 F test that all u_i=0: F(26, 131) = 3.87 Prob > F = 0.0000 rho .48450057 (fraction of variance due to u_i) sigma_e .05569628 sigma_u .05399571 _cons .3553797 .2286283 1.55 0.123 -.0969017 .807661 lp .0427291 .1019821 0.42 0.676 -.1590159 .244474 ttkt -2.939926 1.270468 -2.31 0.022 -5.453214 -.4266375 vhdcv .0061754 .0154746 0.40 0.690 -.024437 .0367877 cv -.0240399 .0607827 -0.40 0.693 -.1442825 .0962027 qmts .0423004 .0128727 3.29 0.001 .0168352 .0677656 vcsh .1419431 .096258 1.47 0.143 -.0484782 .3323645 tsnnt .0912243 .1038802 0.88 0.381 -.1142756 .2967241 cvnt -.0004984 .0287838 -0.02 0.986 -.0574396 .0564428 hqktt Coef. Std. Err. t P>|t| [95% Conf. Interval] corr(u_i, Xb) = -0.3855 Prob > F = 0.0051 F(8,131) = 2.91 overall = 0.0426 max = 7 between = 0.0074 avg = 6.1 R-sq: within = 0.1510 Obs per group: min = 2 Group variable: bank Number of groups = 27 Fixed-effects (within) regression Number of obs = 166 188 MÔ HÌNH REM VỚI BIẾN PHỤ THUỘC HQKTT KIỂM ĐỊNH HAUSMAN rho .3775976 (fraction of variance due to u_i) sigma_e .05569628 sigma_u .04338156 _cons .5904594 .1962403 3.01 0.003 .2058355 .9750833 lp -.0413342 .0950796 -0.43 0.664 -.2276868 .1450184 ttkt -2.200606 1.217085 -1.81 0.071 -4.58605 .1848375 vhdcv .0038971 .0146417 0.27 0.790 -.0248001 .0325943 cv -.0661893 .0527421 -1.25 0.209 -.1695618 .0371833 qmts .0282488 .0108694 2.60 0.009 .0069452 .0495524 vcsh .1783482 .0881107 2.02 0.043 .0056544 .3510421 tsnnt .0348507 .0811598 0.43 0.668 -.1242195 .193921 cvnt -.0027622 .0255385 -0.11 0.914 -.0528167 .0472922 hqktt Coef. Std. Err. z P>|z| [95% Conf. Interval] corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0059 Random effects u_i ~ Gaussian Wald chi2(8) = 21.52 overall = 0.1020 max = 7 between = 0.0715 avg = 6.1 R-sq: within = 0.1327 Obs per group: min = 2 Group variable: bank Number of groups = 27 Random-effects GLS regression Number of obs = 166 Prob>chi2 = 0.4242 = 8.10 chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B) Test: Ho: difference in coefficients not systematic B = inconsistent under Ha, efficient under Ho; obtained from xtreg b = consistent under Ho and Ha; obtained from xtreg lp .0427291 -.0413342 .0840633 .0368811 ttkt -2.939926 -2.200606 -.7393196 .3644053 vhdcv .0061754 .0038971 .0022783 .0050083 cv -.0240399 -.0661893 .0421494 .0302127 qmts .0423004 .0282488 .0140516 .0068965 vcsh .1419431 .1783482 -.0364051 .038757 tsnnt .0912243 .0348507 .0563735 .0648397 cvnt -.0004984 -.0027622 .0022638 .0132775 fe . Difference S.E. (b) (B) (b-B) sqrt(diag(V_b-V_B)) Coefficients 189 KIỂM ĐỊNH BREUSCH-PAGAN LAGRANGIAN KIỂM ĐỊNH WOOLDRIDGE MÔ HÌNH FGLS VỚI BIẾN PHỤ THUỘC HQKTT Prob > chi2 = 0.0000 chi2(1) = 28.30 Test: Var(u) = 0 u .001882 .0433816 e .0031021 .0556963 hqktt .0050851 .0713102 Var sd = sqrt(Var) Estimated results: hqktt[bank,t] = Xb + u[bank] + e[bank,t] Breusch and Pagan Lagrangian multiplier test for random effects Prob > F = 0.0038 F( 1, 25) = 10.179 H0: no first-order autocorrelation Wooldridge test for autocorrelation in panel data _cons .8495037 .1570778 5.41 0.000 .5416368 1.157371 lp -.0537214 .0720377 -0.75 0.456 -.1949128 .0874699 ttkt -2.085647 .9351658 -2.23 0.026 -3.918538 -.2527554 vhdcv .002055 .0095541 0.22 0.830 -.0166707 .0207808 cv -.0963768 .0394909 -2.44 0.015 -.1737776 -.018976 qmts .0138718 .0085421 1.62 0.104 -.0028703 .030614 vcsh .1759197 .0658513 2.67 0.008 .0468534 .304986 tsnnt .0776117 .0607484 1.28 0.201 -.0414531 .1966764 cvnt -.0016853 .0201198 -0.08 0.933 -.0411195 .0377489 hqktt Coef. Std. Err. z P>|z| [95% Conf. Interval] Prob > chi2 = 0.0006 Wald chi2(8) = 27.34 max = 7 avg = 6.148148 Estimated coefficients = 9 Obs per group: min = 2 Estimated autocorrelations = 1 Number of groups = 27 Estimated covariances = 27 Number of obs = 166 Correlation: common AR(1) coefficient for all panels (0.4173) Panels: heteroskedastic Coefficients: generalized least squares Cross-sectional time-series FGLS regression 190 PHỤ LỤC 18 MÔ HÌNH HỒI QUY VÀ PHƯƠNG PHÁP KIỂM ĐỊNH ĐỐI VỚI CÁC NGÂN HÀNG NHÓM 2 VỚI BIẾN PHỤ THUỘC HQQM MÔ HÌNH POOLED VỚI BIẾN PHỤ THUỘC HQQM MÔ HÌNH FEM VỚI BIẾN PHỤ THUỘC HQQM _cons .4965681 .109734 4.53 0.000 .2798228 .7133134 lp .0304117 .0612155 0.50 0.620 -.0905004 .1513238 ttkt .1377362 .8218024 0.17 0.867 -1.485479 1.760951 vhdcv -.0018776 .0093887 -0.20 0.842 -.020422 .0166668 cv .0059055 .0292839 0.20 0.840 -.0519357 .0637467 qmts .0253941 .0057262 4.43 0.000 .0140837 .0367045 vcsh .1509605 .049677 3.04 0.003 .052839 .2490821 tsnnt .0355709 .0394475 0.90 0.369 -.0423453 .1134872 cvnt .010094 .0143605 0.70 0.483 -.0182706 .0384587 hqqm Coef. Std. Err. t P>|t| [95% Conf. Interval] Total .291357835 165 .001765805 Root MSE = .03956 Adj R-squared = 0.1139 Residual .245648482 157 .00156464 R-squared = 0.1569 Model .045709353 8 .005713669 Prob > F = 0.0006 F( 8, 157) = 3.65 Source SS df MS Number of obs = 166 F test that all u_i=0: F(26, 131) = 2.20 Prob > F = 0.0020 rho .30041858 (fraction of variance due to u_i) sigma_e .03613652 sigma_u .02368048 _cons .5704304 .1483372 3.85 0.000 .276984 .8638768 lp .0259915 .0661674 0.39 0.695 -.1049034 .1568864 ttkt .258377 .824297 0.31 0.754 -1.372279 1.889033 vhdcv -.0107464 .0100401 -1.07 0.286 -.0306082 .0091153 cv -.0011626 .0394366 -0.03 0.977 -.0791777 .0768525 qmts .0225017 .008352 2.69 0.008 .0059795 .0390239 vcsh .0914793 .0624535 1.46 0.145 -.0320686 .2150273 tsnnt .0300795 .0673989 0.45 0.656 -.1032517 .1634106 cvnt -.0059199 .0186753 -0.32 0.752 -.0428641 .0310244 hqqm Coef. Std. Err. t P>|t| [95% Conf. Interval] corr(u_i, Xb) = 0.0178 Prob > F = 0.1420 F(8,131) = 1.56 overall = 0.1362 max = 7 between = 0.2399 avg = 6.1 R-sq: within = 0.0871 Obs per group: min = 2 Group variable: bank Number of groups = 27 Fixed-effects (within) regression Number of obs = 166 191 MÔ HÌNH REM VỚI BIẾN PHỤ THUỘC HQQM KIỂM ĐỊNH HAUSMAN rho .2049295 (fraction of variance due to u_i) sigma_e .03613652 sigma_u .01834618 _cons .5425528 .1173575 4.62 0.000 .3125364 .7725692 lp .0263917 .0592802 0.45 0.656 -.0897953 .1425787 ttkt .18228 .7702863 0.24 0.813 -1.327453 1.692013 vhdcv -.0063504 .0091702 -0.69 0.489 -.0243236 .0116228 cv -.0016857 .0315117 -0.05 0.957 -.0634474 .060076 qmts .0235767 .006405 3.68 0.000 .011023 .0361303 vcsh .116519 .0534304 2.18 0.029 .0117975 .2212406 tsnnt .037697 .0464264 0.81 0.417 -.0532971 .1286911 cvnt .0038658 .015371 0.25 0.801 -.0262608 .0339924 hqqm Coef. Std. Err. z P>|z| [95% Conf. Interval] corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0064 Random effects u_i ~ Gaussian Wald chi2(8) = 21.32 overall = 0.1522 max = 7 between = 0.2927 avg = 6.1 R-sq: within = 0.0827 Obs per group: min = 2 Group variable: bank Number of groups = 27 Random-effects GLS regression Number of obs = 166 Prob>chi2 = 0.8692 = 3.86 chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B) Test: Ho: difference in coefficients not systematic B = inconsistent under Ha, efficient under Ho; obtained from xtreg b = consistent under Ho and Ha; obtained from xtreg lp .0259915 .0263917 -.0004002 .0293936 ttkt .258377 .18228 .076097 .2934699 vhdcv -.0107464 -.0063504 -.004396 .004088 cv -.0011626 -.0016857 .0005231 .0237121 qmts .0225017 .0235767 -.001075 .0053601 vcsh .0914793 .116519 -.0250397 .0323364 tsnnt .0300795 .037697 -.0076176 .048859 cvnt -.0059199 .0038658 -.0097857 .0106066 fe . Difference S.E. (b) (B) (b-B) sqrt(diag(V_b-V_B)) Coefficients 192 KIỂM ĐỊNH BREUSCH-PAGAN LAGRANGIAN KIỂM ĐỊNH WOOLDRIDGE MÔ HÌNH FGLS VỚI BIẾN PHỤ THUỘC HQQM Prob > chi2 = 0.0046 chi2(1) = 8.03 Test: Var(u) = 0 u .0003366 .0183462 e .0013058 .0361365 hqqm .0017658 .0420215 Var sd = sqrt(Var) Estimated results: hqqm[bank,t] = Xb + u[bank] + e[bank,t] Breusch and Pagan Lagrangian multiplier test for random effects Prob > F = 0.1910 F( 1, 25) = 1.806 H0: no first-order autocorrelation Wooldridge test for autocorrelation in panel data _cons .7489874 .0763581 9.81 0.000 .5993283 .8986465 lp -.0359849 .0327955 -1.10 0.273 -.1002629 .0282932 ttkt .1790037 .3573899 0.50 0.616 -.5214677 .879475 vhdcv -.007626 .0053583 -1.42 0.155 -.0181282 .0028761 cv -.0326891 .0180833 -1.81 0.071 -.0681317 .0027535 qmts .0137401 .0036888 3.72 0.000 .0065101 .02097 vcsh .0803187 .0356561 2.25 0.024 .010434 .1502033 tsnnt .0124882 .0261436 0.48 0.633 -.0387524 .0637288 cvnt .0050155 .0094298 0.53 0.595 -.0134665 .0234976 hqqm Coef. Std. Err. z P>|z| [95% Conf. Interval] Prob > chi2 = 0.0000 Wald chi2(8) = 103.31 max = 7 avg = 6.148148 Estimated coefficients = 9 Obs per group: min = 2 Estimated autocorrelations = 0 Number of groups = 27 Estimated covariances = 27 Number of obs = 166 Correlation: no autocorrelation Panels: heteroskedastic Coefficients: generalized least squares Cross-sectional time-series FGLS regression