Luận án Tác động của sở hữu gia đình và tài sản tình cảm xã hội đến cấu trúc vốn của doanh nghiệp

285607 -.0706419 liq -.0127316 .0007708 -16.52 0.000 -.0142424 -.0112209 mtb .0059744 .0019041 3.14 0.002 .0022425 .0097064 ndts -.2959344 .0697189 -4.24 0.000 -.432581 -.1592878 fam .0412103 .0172747 2.39 0.017 .0073525 .075068 blev Coef. Std. Err. z P>|z| [95% Conf. Interval] corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 Wald chi2(30) = 1000.76 overall = 0.2611 max = 11 between = 0.2812 avg = 10.0 within = 0.2014 min = 2 R-sq: Obs per group: Group variable: firm Number of groups = 390 Random-effects GLS regression Number of obs = 3,905 23 1.2. Mô hình 1.2: FAM và CEO là các biến độc lập . rho .73081237 (fraction of variance due to u_i) sigma_e .09071412 sigma_u .14946872 _cons -.3890447 .0656666 -5.92 0.000 -.5177488 -.2603405 2020 -.0840174 .0080672 -10.41 0.000 -.0998287 -.068206 2019 -.0744682 .0078106 -9.53 0.000 -.0897767 -.0591597 2018 -.0556856 .0073785 -7.55 0.000 -.0701472 -.041224 2017 -.0485363 .0072827 -6.66 0.000 -.0628102 -.0342625 2016 -.042897 .0071875 -5.97 0.000 -.0569842 -.0288098 2015 -.027641 .0071267 -3.88 0.000 -.0416091 -.013673 2014 -.0200848 .0071923 -2.79 0.005 -.0341814 -.0059882 2013 -.0083102 .0072044 -1.15 0.249 -.0224305 .0058101 2012 -.0066254 .0071703 -0.92 0.355 -.0206789 .0074281 2011 .0039592 .0070997 0.56 0.577 -.009956 .0178744 year Tourism -.0128815 .0649067 -0.20 0.843 -.1400963 .1143333 Retail -.0560629 .0831096 -0.67 0.500 -.2189546 .1068288 Resources .0184953 .0546526 0.34 0.735 -.0886219 .1256125 Real_estate -.0579982 .0515594 -1.12 0.261 -.1590528 .0430565 Public_services -.0367365 .0572177 -0.64 0.521 -.1488813 .0754082 Medical -.1104012 .0605437 -1.82 0.068 -.2290647 .0082623 Industrial -.0283489 .0515375 -0.55 0.582 -.1293607 .0726628 IT -.0322567 .0676793 -0.48 0.634 -.1649058 .1003924 Foods -.0657421 .0519585 -1.27 0.206 -.1675788 .0360947 Consuming -.0152999 .0569366 -0.27 0.788 -.1268937 .0962939 Constructions .0684503 .0497406 1.38 0.169 -.0290394 .1659401 Communications -.0911741 .0790443 -1.15 0.249 -.2460981 .06375 Chemical -.068243 .0565181 -1.21 0.227 -.1790164 .0425304 ind tang .0094095 .0163733 0.57 0.566 -.0226816 .0415006 size .0706251 .0033806 20.89 0.000 .0639992 .077251 prof -.0995615 .0147754 -6.74 0.000 -.1285209 -.0706022 liq -.0127642 .0007715 -16.54 0.000 -.0142764 -.011252 mtb .0059521 .0019042 3.13 0.002 .0022199 .0096844 ndts -.2955625 .0697196 -4.24 0.000 -.4322103 -.1589147 ceo .0323742 .0195427 1.66 0.098 -.0059287 .0706772 fam .0549254 .0223482 2.46 0.014 .0111239 .098727 blev Coef. Std. Err. z P>|z| [95% Conf. Interval] corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 Wald chi2(31) = 1001.67 overall = 0.2620 max = 11 between = 0.2822 avg = 10.0 within = 0.2016 min = 2 R-sq: Obs per group: Group variable: firm Number of groups = 390 Random-effects GLS regression Number of obs = 3,905 . xtreg blev fam ceo ndts mtb liq prof size tang i.ind i.year, re 24 1.3. Mô hình 1.3: FAM và GEN là các biến độc lập . rho .73165965 (fraction of variance due to u_i) sigma_e .09047759 sigma_u .14940068 _cons -.4078564 .0653256 -6.24 0.000 -.5358922 -.2798207 2020 -.0849092 .0080523 -10.54 0.000 -.1006915 -.069127 2019 -.075303 .0077966 -9.66 0.000 -.0905841 -.0600219 2018 -.0560546 .0073641 -7.61 0.000 -.0704879 -.0416212 2017 -.0488407 .0072687 -6.72 0.000 -.0630871 -.0345943 2016 -.0432399 .0071741 -6.03 0.000 -.0573009 -.029179 2015 -.0280747 .0071144 -3.95 0.000 -.0420188 -.0141306 2014 -.020505 .0071801 -2.86 0.004 -.0345778 -.0064322 2013 -.0086695 .0071922 -1.21 0.228 -.022766 .005427 2012 -.0068083 .0071576 -0.95 0.342 -.020837 .0072203 2011 .003813 .0070872 0.54 0.591 -.0100776 .0177037 year Tourism -.0063792 .0649272 -0.10 0.922 -.1336342 .1208758 Retail -.0579622 .0831314 -0.70 0.486 -.2208968 .1049724 Resources .0303781 .0546562 0.56 0.578 -.0767462 .1375023 Real_estate -.0497502 .0514606 -0.97 0.334 -.1506111 .0511106 Public_services -.0296939 .0571743 -0.52 0.604 -.1417535 .0823657 Medical -.0973336 .0605258 -1.61 0.108 -.215962 .0212949 Industrial -.0268673 .0515068 -0.52 0.602 -.1278187 .0740841 IT -.0126281 .0675637 -0.19 0.852 -.1450506 .1197943 Foods -.0562261 .0518824 -1.08 0.278 -.1579137 .0454615 Consuming -.0085983 .0569526 -0.15 0.880 -.1202234 .1030268 Constructions .0756713 .0496841 1.52 0.128 -.0217078 .1730503 Communications -.0864275 .0788821 -1.10 0.273 -.2410335 .0681785 Chemical -.0690043 .0565177 -1.22 0.222 -.1797769 .0417683 ind tang .0091513 .0163438 0.56 0.576 -.022882 .0411846 size .0715253 .0033661 21.25 0.000 .0649278 .0781228 prof -.1001364 .0147511 -6.79 0.000 -.1290479 -.0712248 liq -.0127932 .0007698 -16.62 0.000 -.0143021 -.0112844 mtb .0061563 .0019016 3.24 0.001 .0024292 .0098835 ndts -.2935805 .069603 -4.22 0.000 -.4299998 -.1571611 gen -.0086687 .0221709 -0.39 0.696 -.0521228 .0347855 fam .0774394 .0200106 3.87 0.000 .0382192 .1166595 blev Coef. Std. Err. z P>|z| [95% Conf. Interval] corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 Wald chi2(31) = 1016.60 overall = 0.2617 max = 11 between = 0.2805 avg = 10.0 within = 0.2046 min = 2 R-sq: Obs per group: Group variable: firm Number of groups = 390 Random-effects GLS regression Number of obs = 3,905 . xtreg blev fam gen ndts mtb liq prof size tang i.ind i.year, re 25 1.4. Mô hình 1.4: FAM và BOD là các biến độc lập rho .7316647 (fraction of variance due to u_i) sigma_e .09048568 sigma_u .14941596 _cons -.4106254 .0652714 -6.29 0.000 -.5385551 -.2826958 2020 -.0849913 .0080702 -10.53 0.000 -.1008086 -.069174 2019 -.0753715 .0078108 -9.65 0.000 -.0906805 -.0600626 2018 -.0561772 .0073841 -7.61 0.000 -.0706498 -.0417047 2017 -.0489845 .0072941 -6.72 0.000 -.0632806 -.0346884 2016 -.0433657 .0071947 -6.03 0.000 -.0574671 -.0292643 2015 -.0281264 .0071197 -3.95 0.000 -.0420808 -.014172 2014 -.0205213 .0071827 -2.86 0.004 -.0345992 -.0064435 2013 -.0086781 .0071935 -1.21 0.228 -.0227771 .0054209 2012 -.0068156 .0071583 -0.95 0.341 -.0208456 .0072143 2011 .0038284 .0070872 0.54 0.589 -.0100623 .0177191 year Tourism -.0070059 .0649103 -0.11 0.914 -.1342277 .1202159 Retail -.0553085 .0830275 -0.67 0.505 -.2180393 .1074223 Resources .0297808 .054631 0.55 0.586 -.0772939 .1368556 Real_estate -.0488654 .0515061 -0.95 0.343 -.1498155 .0520848 Public_services -.0267987 .0570586 -0.47 0.639 -.1386316 .0850341 Medical -.0972712 .0605322 -1.61 0.108 -.2159122 .0213698 Industrial -.0245236 .051422 -0.48 0.633 -.1253089 .0762616 IT -.0111978 .0675815 -0.17 0.868 -.1436551 .1212596 Foods -.0561997 .0518877 -1.08 0.279 -.1578977 .0454984 Consuming -.0084831 .0569641 -0.15 0.882 -.1201306 .1031644 Constructions .0778346 .0496524 1.57 0.117 -.0194823 .1751515 Communications -.0832981 .0788559 -1.06 0.291 -.2378529 .0712567 Chemical -.0680553 .0565406 -1.20 0.229 -.1788728 .0427623 ind tang .0092397 .0163423 0.57 0.572 -.0227905 .04127 size .071466 .0033633 21.25 0.000 .0648741 .078058 prof -.1000812 .0147504 -6.78 0.000 -.1289913 -.071171 liq -.0127855 .0007699 -16.61 0.000 -.0142944 -.0112766 mtb .0061611 .0019023 3.24 0.001 .0024327 .0098895 ndts -.2926647 .0696235 -4.20 0.000 -.4291242 -.1562051 bod .0054857 .025014 0.22 0.826 -.0435408 .0545123 fam .0788028 .0198 3.98 0.000 .0399956 .11761 blev Coef. Std. Err. z P>|z| [95% Conf. Interval] corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 Wald chi2(31) = 1016.46 overall = 0.2623 max = 11 between = 0.2819 avg = 10.0 within = 0.2044 min = 2 R-sq: Obs per group: Group variable: firm Number of groups = 390 Random-effects GLS regression Number of obs = 3,905 . xtreg blev fam bod ndts mtb liq prof size tang i.ind i.year, re 26 1.5. Mô hình 1.5: FAM và SEW là các biến độc lập . rho .73114658 (fraction of variance due to u_i) sigma_e .09068772 sigma_u .14955224 _cons -.400392 .0656643 -6.10 0.000 -.5290918 -.2716923 2020 -.0854879 .0080853 -10.57 0.000 -.1013349 -.0696409 2019 -.0757867 .0078249 -9.69 0.000 -.0911232 -.0604502 2018 -.0568752 .0073913 -7.69 0.000 -.0713619 -.0423885 2017 -.0497689 .007297 -6.82 0.000 -.0640708 -.035467 2016 -.043968 .0071976 -6.11 0.000 -.058075 -.0298609 2015 -.0282303 .0071268 -3.96 0.000 -.0421986 -.014262 2014 -.0206392 .0071909 -2.87 0.004 -.0347332 -.0065453 2013 -.0087283 .0072016 -1.21 0.226 -.0228431 .0053866 2012 -.0068807 .0071664 -0.96 0.337 -.0209266 .0071652 2011 .0039258 .0070952 0.55 0.580 -.0099805 .0178321 year Tourism -.0109213 .0649202 -0.17 0.866 -.1381627 .11632 Retail -.052622 .083068 -0.63 0.526 -.2154324 .1101883 Resources .0255263 .0547535 0.47 0.641 -.0817886 .1328413 Real_estate -.0494958 .05173 -0.96 0.339 -.1508848 .0518931 Public_services -.0291755 .0571963 -0.51 0.610 -.1412781 .0829271 Medical -.1051264 .0606005 -1.73 0.083 -.2239011 .0136483 Industrial -.0231084 .0515258 -0.45 0.654 -.1240972 .0778803 IT -.0232381 .0677399 -0.34 0.732 -.1560059 .1095297 Foods -.0607284 .0520219 -1.17 0.243 -.1626895 .0412327 Consuming -.0106502 .0569894 -0.19 0.852 -.1223474 .1010469 Constructions .0756892 .0497892 1.52 0.128 -.0218958 .1732743 Communications -.079428 .0792005 -1.00 0.316 -.2346581 .0758021 Chemical -.0606561 .0566383 -1.07 0.284 -.1716651 .050353 ind tang .0093388 .0163583 0.57 0.568 -.0227228 .0414004 size .0709495 .0033824 20.98 0.000 .0643202 .0775788 prof -.1001244 .0147661 -6.78 0.000 -.1290656 -.0711833 liq -.012751 .0007711 -16.54 0.000 -.0142623 -.0112398 mtb .0061328 .0019048 3.22 0.001 .0023995 .0098661 ndts -.2918213 .0696995 -4.19 0.000 -.4284299 -.1552127 sew .0169431 .0063831 2.65 0.008 .0044324 .0294539 fam .0285983 .0210946 1.36 0.175 -.0127463 .0699429 blev Coef. Std. Err. z P>|z| [95% Conf. Interval] corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 Wald chi2(31) = 1007.03 overall = 0.2613 max = 11 between = 0.2823 avg = 10.0 within = 0.2024 min = 2 R-sq: Obs per group: Group variable: firm Number of groups = 390 Random-effects GLS regression Number of obs = 3,905 . xtreg blev fam sew ndts mtb liq prof size tang i.ind i.year, re 27 2. Kết quả hồi quy tác động ngẫu nhiên các mô hình có biến tương tác 2.1. Mô hình 2.1: Biến FAM là biến độc lập corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 Wald chi2(32) = 1064.36 overall = 0.2789 max = 11 between = 0.2988 avg = 10.0 within = 0.2075 min = 2 R-sq: Obs per group: Group variable: firm Number of groups = 390 Random-effects GLS regression Number of obs = 3,905 . xtreg blev pba fam pba_fam ndts mtb liq prof size tang i.ind i.year, re rho .69437452 (fraction of variance due to u_i) sigma_e .08998412 sigma_u .13563382 _cons -.3663133 .061847 -5.92 0.000 -.4875313 -.2450954 2020 -.0810318 .008054 -10.06 0.000 -.0968173 -.0652462 2019 -.0721153 .0077937 -9.25 0.000 -.0873907 -.05684 2018 -.0535445 .0073664 -7.27 0.000 -.0679824 -.0391065 2017 -.0467448 .0072746 -6.43 0.000 -.0610027 -.0324869 2016 -.0413261 .0071967 -5.74 0.000 -.0554313 -.0272208 2015 -.0244032 .0071437 -3.42 0.001 -.0384046 -.0104018 2014 -.0210883 .0072042 -2.93 0.003 -.0352083 -.0069683 2013 -.0102339 .0072267 -1.42 0.157 -.0243979 .0039301 2012 -.0091567 .0072074 -1.27 0.204 -.023283 .0049695 2011 .0041435 .0071132 0.58 0.560 -.0097981 .0180851 year Tourism -.0151983 .0597689 -0.25 0.799 -.1323432 .1019466 Retail -.054778 .0763906 -0.72 0.473 -.2045007 .0949448 Resources .0233516 .0500076 0.47 0.641 -.0746614 .1213647 Real_estate -.052345 .0472333 -1.11 0.268 -.1449205 .0402304 Public_services -.0332232 .0525677 -0.63 0.527 -.136254 .0698077 Medical -.108376 .0556013 -1.95 0.051 -.2173525 .0006005 Industrial -.0292032 .0473642 -0.62 0.538 -.1220353 .0636289 IT -.0295302 .0619888 -0.48 0.634 -.151026 .0919656 Foods -.0626247 .0475956 -1.32 0.188 -.1559104 .030661 Consuming -.0154184 .0523484 -0.29 0.768 -.1180193 .0871826 Constructions .0688124 .0456291 1.51 0.132 -.020619 .1582438 Communications -.0907971 .0725741 -1.25 0.211 -.2330397 .0514455 Chemical -.0687148 .0519605 -1.32 0.186 -.1705556 .0331259 ind tang .0073785 .0163419 0.45 0.652 -.0246509 .039408 size .0680272 .003292 20.66 0.000 .061575 .0744795 prof -.0837409 .0150397 -5.57 0.000 -.1132182 -.0542637 liq -.0126155 .0007645 -16.50 0.000 -.0141139 -.0111171 mtb .005232 .0018944 2.76 0.006 .001519 .008945 ndts -.2718539 .0697917 -3.90 0.000 -.4086431 -.1350647 pba_fam -.0171378 .0075634 -2.27 0.023 -.0319618 -.0023139 fam .0461235 .0159735 2.89 0.004 .0148161 .077431 pba .0281989 .0043047 6.55 0.000 .0197618 .036636 blev Coef. Std. Err. z P>|z| [95% Conf. Interval] 28 2.2. Mô hình 2.2: Biến CEO là biến độc lập corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 Wald chi2(32) = 1052.73 overall = 0.2742 max = 11 between = 0.2911 avg = 10.0 within = 0.2063 min = 2 R-sq: Obs per group: Group variable: firm Number of groups = 390 Random-effects GLS regression Number of obs = 3,905 . xtreg blev pba ceo pba_ceo ndts mtb liq prof size tang i.ind i.year, re _cons .0927044 .0352667 2.63 0.009 .0235614 .1618474 2020 -.0423861 .0136197 -3.11 0.002 -.0690885 -.0156836 2019 -.0403504 .0133528 -3.02 0.003 -.0665297 -.0141712 2018 -.0248129 .0128015 -1.94 0.053 -.0499113 .0002854 2017 -.0223674 .0127585 -1.75 0.080 -.0473814 .0026466 2016 -.0224254 .0127998 -1.75 0.080 -.0475205 .0026696 2015 -.0038489 .0128531 -0.30 0.765 -.0290484 .0213506 2014 -.0143025 .013087 -1.09 0.275 -.0399605 .0113554 2013 -.0107366 .0131507 -0.82 0.414 -.0365195 .0150463 2012 -.0145544 .0131527 -1.11 0.269 -.0403413 .0112325 2011 .0045093 .0130509 0.35 0.730 -.0210779 .0300966 year Tourism -.048155 .0246912 -1.95 0.051 -.096564 .000254 Retail -.0842527 .0284993 -2.96 0.003 -.1401279 -.0283776 Resources .0236321 .019223 1.23 0.219 -.014056 .0613202 Real_estate -.0405852 .0183435 -2.21 0.027 -.076549 -.0046215 Public_services -.0352331 .0202541 -1.74 0.082 -.0749429 .0044766 Medical -.1450684 .0215016 -6.75 0.000 -.1872239 -.1029128 Industrial -.0648074 .0182411 -3.55 0.000 -.1005704 -.0290444 IT -.0635694 .0232871 -2.73 0.006 -.1092255 -.0179133 Foods -.0688077 .0183257 -3.75 0.000 -.1047367 -.0328787 Consuming -.0451948 .0201857 -2.24 0.025 -.0847705 -.0056191 Constructions .0413669 .0174953 2.36 0.018 .0070661 .0756677 Communications -.1489311 .0281208 -5.30 0.000 -.2040641 -.0937981 Chemical -.0969522 .0199655 -4.86 0.000 -.1360961 -.0578082 ind tang -.03788 .0169463 -2.24 0.025 -.0711045 -.0046555 size .0365606 .0022147 16.51 0.000 .0322186 .0409027 prof -.0639287 .022487 -2.84 0.004 -.1080163 -.0198412 liq -.0196519 .0008886 -22.12 0.000 -.0213941 -.0179097 mtb -.0050028 .0021203 -2.36 0.018 -.0091598 -.0008459 ndts -.1331287 .0958772 -1.39 0.165 -.3211033 .0548459 pba_ceo -.0442517 .0148074 -2.99 0.003 -.0732827 -.0152207 ceo .0199172 .0090907 2.19 0.029 .0020942 .0377403 pba .0822612 .0068531 12.00 0.000 .0688251 .0956972 blev Coef. Std. Err. t P>|t| [95% Conf. Interval] rho .62614386 (fraction of variance due to u_i) sigma_e .07281673 sigma_u .09423581 29 2.3. Mô hình 2.3: Biến GEN là biến độc lập corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 Wald chi2(32) = 1052.69 overall = 0.2701 max = 11 between = 0.2840 avg = 10.0 within = 0.2078 min = 2 R-sq: Obs per group: Group variable: firm Number of groups = 390 Random-effects GLS regression Number of obs = 3,905 . xtreg blev pba gen pba_gen ndts mtb liq prof size tang i.ind i.year, re _cons .0976668 .0348683 2.80 0.005 .0293048 .1660289 2020 -.0421319 .0136322 -3.09 0.002 -.0688589 -.015405 2019 -.0398123 .0133669 -2.98 0.003 -.0660191 -.0136055 2018 -.0248888 .0128078 -1.94 0.052 -.0499994 .0002218 2017 -.0216055 .0127677 -1.69 0.091 -.0466375 .0034266 2016 -.022055 .0128121 -1.72 0.085 -.047174 .003064 2015 -.004294 .0128573 -0.33 0.738 -.0295017 .0209138 2014 -.0145245 .0130976 -1.11 0.268 -.0402034 .0111543 2013 -.0109793 .0131632 -0.83 0.404 -.0367867 .0148281 2012 -.0149572 .0131666 -1.14 0.256 -.0407713 .0108568 2011 .0040314 .0130609 0.31 0.758 -.0215755 .0296383 year Tourism -.0457843 .0247134 -1.85 0.064 -.0942368 .0026682 Retail -.0831892 .0284896 -2.92 0.004 -.1390453 -.0273331 Resources .0243266 .0192466 1.26 0.206 -.0134079 .062061 Real_estate -.0403998 .0182714 -2.21 0.027 -.0762224 -.0045773 Public_services -.0347758 .0201365 -1.73 0.084 -.074255 .0047034 Medical -.1436484 .0214669 -6.69 0.000 -.185736 -.1015608 Industrial -.0640467 .0181748 -3.52 0.000 -.0996797 -.0284137 IT -.064284 .0230502 -2.79 0.005 -.1094757 -.0190923 Foods -.0679316 .018294 -3.71 0.000 -.1037983 -.0320648 Consuming -.0439556 .0201749 -2.18 0.029 -.0835099 -.0044012 Constructions .0415729 .0173442 2.40 0.017 .0075682 .0755775 Communications -.1496052 .0279524 -5.35 0.000 -.2044079 -.0948024 Chemical -.0954624 .0199684 -4.78 0.000 -.134612 -.0563129 ind tang -.0389257 .016979 -2.29 0.022 -.0722144 -.0056371 size .0364091 .0022233 16.38 0.000 .0320501 .0407681 prof -.065004 .0225341 -2.88 0.004 -.109184 -.0208241 liq -.0197806 .0008872 -22.29 0.000 -.0215201 -.0180412 mtb -.0053731 .002121 -2.53 0.011 -.0095315 -.0012146 ndts -.1398291 .0957254 -1.46 0.144 -.327506 .0478479 pba_gen -.0299968 .017201 -1.74 0.081 -.0637206 .0037271 gen .0110319 .010436 1.06 0.291 -.0094286 .0314925 pba .0775864 .0066077 11.74 0.000 .0646315 .0905413 blev Coef. Std. Err. t P>|t| [95% Conf. Interval] rho .69580079 (fraction of variance due to u_i) sigma_e .09004482 sigma_u .13618276 30 2.4. Mô hình 2.4: Biến BOD là biến độc lập corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 Wald chi2(32) = 1057.81 overall = 0.2761 max = 11 between = 0.2938 avg = 10.0 within = 0.2069 min = 2 R-sq: Obs per group: Group variable: firm Number of groups = 390 Random-effects GLS regression Number of obs = 3,905 . xtreg blev pba bod pba_bod ndts mtb liq prof size tang i.ind i.year, re rho .69608904 (fraction of variance due to u_i) sigma_e .09001662 sigma_u .13623288 _cons -.3595292 .0618034 -5.82 0.000 -.4806615 -.2383968 2020 -.0821024 .0080768 -10.17 0.000 -.0979326 -.0662722 2019 -.0732638 .0078097 -9.38 0.000 -.0885706 -.057957 2018 -.0546042 .0073859 -7.39 0.000 -.0690803 -.0401281 2017 -.0479176 .0072984 -6.57 0.000 -.0622222 -.033613 2016 -.0424993 .0072149 -5.89 0.000 -.0566402 -.0283584 2015 -.0250332 .0071494 -3.50 0.000 -.0390457 -.0110207 2014 -.0216924 .0072073 -3.01 0.003 -.0358184 -.0075664 2013 -.0107513 .0072276 -1.49 0.137 -.024917 .0034144 2012 -.0096945 .0072046 -1.35 0.178 -.0238152 .0044263 2011 .0038809 .0071116 0.55 0.585 -.0100575 .0178193 year Tourism -.0126147 .0599944 -0.21 0.833 -.1302014 .1049721 Retail -.0609036 .0765882 -0.80 0.426 -.2110138 .0892066 Resources .0262517 .050208 0.52 0.601 -.0721541 .1246575 Real_estate -.0510196 .0474628 -1.07 0.282 -.1440449 .0420057 Public_services -.0405895 .052564 -0.77 0.440 -.1436131 .0624341 Medical -.1087609 .0558242 -1.95 0.051 -.2181743 .0006526 Industrial -.0337635 .0474498 -0.71 0.477 -.1267633 .0592363 IT -.03413 .062161 -0.55 0.583 -.1559633 .0877034 Foods -.0621808 .0477896 -1.30 0.193 -.1558467 .0314851 Consuming -.0141431 .0525717 -0.27 0.788 -.1171817 .0888954 Constructions .0647163 .0457216 1.42 0.157 -.0248964 .1543289 Communications -.094651 .0728214 -1.30 0.194 -.2373784 .0480764 Chemical -.0671115 .052188 -1.29 0.198 -.1693982 .0351751 ind tang .0061915 .0163475 0.38 0.705 -.025849 .0382321 size .0682874 .0032959 20.72 0.000 .0618276 .0747472 prof -.0839936 .0150389 -5.59 0.000 -.1134693 -.0545179 liq -.0125966 .000765 -16.47 0.000 -.014096 -.0110972 mtb .0052999 .0018963 2.79 0.005 .0015832 .0090165 ndts -.2715818 .0698287 -3.89 0.000 -.4084434 -.1347201 pba_bod -.0288241 .0150031 -1.92 0.055 -.0582296 .0005814 bod .0548826 .0234342 2.34 0.019 .0089526 .1008127 pba .0268192 .0041482 6.47 0.000 .0186889 .0349496 blev Coef. Std. Err. z P>|z| [95% Conf. Interval] 31 2.5. Mô hình 2.5: Biến SEW là biến độc lập corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 Wald chi2(32) = 1066.57 overall = 0.2768 max = 11 between = 0.2970 avg = 10.0 within = 0.2080 min = 2 R-sq: Obs per group: Group variable: firm Number of groups = 390 Random-effects GLS regression Number of obs = 3,905 . xtreg blev pba sew pba_sew ndts mtb liq prof size tang i.ind i.year, re rho .69551841 (fraction of variance due to u_i) sigma_e .09001969 sigma_u .136054 _cons -.3735676 .0622324 -6.00 0.000 -.4955408 -.2515944 2020 -.082503 .0080715 -10.22 0.000 -.0983228 -.0666831 2019 -.0734998 .007807 -9.41 0.000 -.0888013 -.0581983 2018 -.0547836 .0073813 -7.42 0.000 -.0692506 -.0403165 2017 -.0481249 .0072912 -6.60 0.000 -.0624154 -.0338344 2016 -.0427375 .0072074 -5.93 0.000 -.0568637 -.0286112 2015 -.0251758 .0071467 -3.52 0.000 -.0391831 -.0111686 2014 -.0217673 .0072046 -3.02 0.003 -.0358881 -.0076465 2013 -.0107736 .0072258 -1.49 0.136 -.0249359 .0033887 2012 -.0097367 .0072023 -1.35 0.176 -.023853 .0043796 2011 .0039419 .0071108 0.55 0.579 -.0099951 .0178789 year Tourism -.0115389 .0599036 -0.19 0.847 -.1289479 .1058701 Retail -.0550964 .0765389 -0.72 0.472 -.2051098 .0949171 Resources .0311561 .0501804 0.62 0.535 -.0671957 .1295079 Real_estate -.044138 .0475335 -0.93 0.353 -.137302 .049026 Public_services -.0304028 .0527731 -0.58 0.565 -.1338361 .0730304 Medical -.1036699 .0558042 -1.86 0.063 -.2130442 .0057044 Industrial -.0270788 .0475209 -0.57 0.569 -.120218 .0660604 IT -.0235587 .062324 -0.38 0.705 -.1457115 .0985941 Foods -.0579056 .0477612 -1.21 0.225 -.151516 .0357047 Consuming -.0110386 .0525149 -0.21 0.834 -.1139658 .0918887 Constructions .0727795 .0458677 1.59 0.113 -.0171195 .1626784 Communications -.0828659 .0729997 -1.14 0.256 -.2259426 .0602108 Chemical -.0616127 .0521895 -1.18 0.238 -.1639023 .0406769 ind tang .0066756 .0163386 0.41 0.683 -.0253475 .0386987 size .0685431 .0032958 20.80 0.000 .0620835 .0750027 prof -.0838756 .0150336 -5.58 0.000 -.1133409 -.0544103 liq -.0125859 .0007647 -16.46 0.000 -.0140847 -.0110871 mtb .0053584 .0018962 2.83 0.005 .001642 .0090748 ndts -.2694303 .0698117 -3.86 0.000 -.4062587 -.132602 pba_sew -.0062679 .0032602 -1.92 0.055 -.0126579 .000122 sew .0180125 .0060366 2.98 0.003 .0061811 .029844 pba .0267181 .0041253 6.48 0.000 .0186327 .0348036 blev Coef. Std. Err. z P>|z| [95% Conf. Interval] 32 PHỤ LỤC 7: KẾT QUẢ HỒI QUY SGMM 1. Kết quả hồi quy SGMM các mô hình không có biến tương tác 1.1. Mô hình 1.1: FAM là biến độc lập Difference-in-Sargan/Hansen statistics may be negative. Using a generalized inverse to calculate optimal weighting matrix for two-step estimation. Warning: Two-step estimated covariance matrix of moments is singular. Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm. > ation(level)) iv(i.ind i.year) twostep > of size tang, lag (1 3) equation(diff)) gmm(l.blev fam ndts mtb liq prof size tang, lag (1 1) equ . xtabond2 blev l.blev fam ndts mtb liq prof size tang i.ind i.year, gmm(l.blev fam ndts mtb liq pr Prob > chi2 = 0.000 max = 10 Wald chi2(33) = 19015.69 avg = 8.93 Number of instruments = 282 Obs per group: min = 1 Time variable : year Number of groups = 390 Group variable: firm Number of obs = 3482 Dynamic panel-data estimation, two-step system GMM L1. .8692185 .0125266 69.39 0.000 .8446669 .8937701 blev blev Coef. Std. Err. z P>|z| [95% Conf. Interval] tang -.0633106 .0091404 -6.93 0.000 -.0812255 -.0453956 size .0079849 .001634 4.89 0.000 .0047823 .0111875 prof -.1020359 .0122142 -8.35 0.000 -.1259752 -.0780966 liq -.0033649 .0006424 -5.24 0.000 -.0046241 -.0021058 mtb .0018647 .0007662 2.43 0.015 .0003631 .0033664 ndts -.0889366 .0592458 -1.50 0.133 -.2050564 .0271831 fam .0358863 .0058134 6.17 0.000 .0244923 .0472803 DL.(L.blev fam ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) _cons 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year 1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for levels equation L(1/3).(L.blev fam ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year) 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year D.(1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for first differences equation (Robust, but weakened by many instruments.) Hansen test of overid. restrictions: chi2(223) = 238.46 Prob > chi2 = 0.227 (Not robust, but not weakened by many instruments.) Sargan test of overid. restrictions: chi2(223) = 289.83 Prob > chi2 = 0.002 Arellano-Bond test for AR(2) in first differences: z = -1.33 Pr > z = 0.184 Arellano-Bond test for AR(1) in first differences: z = -8.83 Pr > z = 0.000 33 1.2. Mô hình 1.2: FAM và CEO là các biến độc lập Difference-in-Sargan/Hansen statistics may be negative. Using a generalized inverse to calculate optimal weighting matrix for two-step estimation. Warning: Two-step estimated covariance matrix of moments is singular. Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm. > g (1 1) equation(level)) iv(i.ind i.year) twostep > liq prof size tang, lag (1 3) equation(diff)) gmm(l.blev fam ceo ndts mtb liq prof size tang, la . xtabond2 blev l.blev fam ceo ndts mtb liq prof size tang i.ind i.year, gmm(l.blev fam ceo ndts mtb Prob > chi2 = 0.000 max = 10 Wald chi2(34) = 142709.92 avg = 8.93 Number of instruments = 298 Obs per group: min = 1 Time variable : year Number of groups = 390 Group variable: firm Number of obs = 3482 Dynamic panel-data estimation, two-step system GMM L1. .8537922 .0116445 73.32 0.000 .8309694 .876615 blev blev Coef. Std. Err. z P>|z| [95% Conf. Interval] tang -.0553512 .0086176 -6.42 0.000 -.0722413 -.0384611 size .0089359 .0014892 6.00 0.000 .0060172 .0118546 prof -.1132514 .011315 -10.01 0.000 -.1354285 -.0910743 liq -.0034314 .0006402 -5.36 0.000 -.0046863 -.0021766 mtb .0017673 .0007524 2.35 0.019 .0002926 .003242 ndts -.1492171 .0439634 -3.39 0.001 -.2353837 -.0630505 ceo .0130222 .0097568 1.33 0.182 -.0061007 .0321451 fam .0250645 .0077575 3.23 0.001 .0098601 .0402689 DL.(L.blev fam ceo ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) _cons 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year 1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for levels equation L(1/3).(L.blev fam ceo ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year) 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year D.(1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for first differences equation (Robust, but weakened by many instruments.) Hansen test of overid. restrictions: chi2(237) = 251.12 Prob > chi2 = 0.253 (Not robust, but not weakened by many instruments.) Sargan test of overid. restrictions: chi2(237) = 313.18 Prob > chi2 = 0.001 Arellano-Bond test for AR(2) in first differences: z = -1.47 Pr > z = 0.141 Arellano-Bond test for AR(1) in first differences: z = -8.71 Pr > z = 0.000 34 1.3. Mô hình 1.3: Biến FAM và GEN là các biến độc lập Difference-in-Sargan/Hansen statistics may be negative. Using a generalized inverse to calculate optimal weighting matrix for two-step estimation. Warning: Two-step estimated covariance matrix of moments is singular. Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm. > (1 1) equation(level)) iv(i.ind i.year) twostep > liq prof size tang, lag (1 3) equation(diff)) gmm(l.blev fam gen ndts mtb liq prof size tang, lag . xtabond2 blev l.blev fam gen ndts mtb liq prof size tang i.ind i.year, gmm(l.blev fam gen ndts mtb Prob > chi2 = 0.000 max = 10 Wald chi2(34) = 24868.22 avg = 8.93 Number of instruments = 297 Obs per group: min = 1 Time variable : year Number of groups = 390 Group variable: firm Number of obs = 3482 Dynamic panel-data estimation, two-step system GMM L1. .8603576 .0118314 72.72 0.000 .8371685 .8835466 blev blev Coef. Std. Err. z P>|z| [95% Conf. Interval] tang -.0591546 .0086023 -6.88 0.000 -.0760148 -.0422944 size .0079316 .0016777 4.73 0.000 .0046434 .0112197 prof -.1023285 .0107062 -9.56 0.000 -.1233123 -.0813447 liq -.0032374 .0006229 -5.20 0.000 -.0044582 -.0020167 mtb .0013872 .0007168 1.94 0.053 -.0000176 .0027921 ndts -.113546 .0570795 -1.99 0.047 -.2254199 -.0016722 gen .0129942 .0116214 1.12 0.264 -.0097834 .0357718 fam .0308306 .0073416 4.20 0.000 .0164414 .0452199 DL.(L.blev fam gen ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) _cons 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year 1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for levels equation L(1/3).(L.blev fam gen ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year) 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year D.(1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for first differences equation (Robust, but weakened by many instruments.) Hansen test of overid. restrictions: chi2(236) = 248.61 Prob > chi2 = 0.274 (Not robust, but not weakened by many instruments.) Sargan test of overid. restrictions: chi2(236) = 313.49 Prob > chi2 = 0.001 Arellano-Bond test for AR(2) in first differences: z = -1.35 Pr > z = 0.178 Arellano-Bond test for AR(1) in first differences: z = -8.84 Pr > z = 0.000 35 1.4. Mô hình 1.4: Biến FAM và BOD là các biến độc lập Difference-in-Sargan/Hansen statistics may be negative. Using a generalized inverse to calculate optimal weighting matrix for two-step estimation. Warning: Two-step estimated covariance matrix of moments is singular. Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm. > g (1 1) equation(level)) iv(i.ind i.year) twostep > liq prof size tang, lag (1 3) equation(diff)) gmm(l.blev fam bod ndts mtb liq prof size tang, la . xtabond2 blev l.blev fam bod ndts mtb liq prof size tang i.ind i.year, gmm(l.blev fam bod ndts mtb Prob > chi2 = 0.000 max = 10 Wald chi2(34) = 26663.77 avg = 8.93 Number of instruments = 317 Obs per group: min = 1 Time variable : year Number of groups = 390 Group variable: firm Number of obs = 3482 Dynamic panel-data estimation, two-step system GMM L1. .8711695 .0106584 81.74 0.000 .8502794 .8920596 blev blev Coef. Std. Err. z P>|z| [95% Conf. Interval] tang -.0558306 .0077678 -7.19 0.000 -.0710551 -.0406061 size .0101472 .0012985 7.81 0.000 .0076022 .0126922 prof -.1214753 .0092924 -13.07 0.000 -.1396879 -.1032626 liq -.0032645 .0006197 -5.27 0.000 -.004479 -.0020499 mtb .0009259 .0006878 1.35 0.178 -.0004221 .0022739 ndts -.1810248 .0384319 -4.71 0.000 -.2563499 -.1056998 bod .0200895 .0079339 2.53 0.011 .0045393 .0356396 fam .0171956 .0071926 2.39 0.017 .0030984 .0312927 DL.(L.blev fam bod ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) _cons 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year 1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for levels equation L(1/3).(L.blev fam bod ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year) 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year D.(1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for first differences equation (Robust, but weakened by many instruments.) Hansen test of overid. restrictions: chi2(255) = 249.51 Prob > chi2 = 0.585 (Not robust, but not weakened by many instruments.) Sargan test of overid. restrictions: chi2(255) = 330.39 Prob > chi2 = 0.001 Arellano-Bond test for AR(2) in first differences: z = -1.30 Pr > z = 0.195 Arellano-Bond test for AR(1) in first differences: z = -8.80 Pr > z = 0.000 36 1.5. Mô hình 1.5: Biến FAM và SEW là các biến độc lập Difference-in-Sargan/Hansen statistics may be negative. Using a generalized inverse to calculate optimal weighting matrix for two-step estimation. Warning: Two-step estimated covariance matrix of moments is singular. Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm. > g (1 1) equation(level)) iv(i.ind i.year) twostep > liq prof size tang, lag (1 3) equation(diff)) gmm(l.blev fam sew ndts mtb liq prof size tang, la . xtabond2 blev l.blev fam sew ndts mtb liq prof size tang i.ind i.year, gmm(l.blev fam sew ndts mtb Prob > chi2 = 0.000 max = 10 Wald chi2(34) = 26789.29 avg = 8.93 Number of instruments = 317 Obs per group: min = 1 Time variable : year Number of groups = 390 Group variable: firm Number of obs = 3482 Dynamic panel-data estimation, two-step system GMM L1. .8714316 .0107177 81.31 0.000 .8504252 .892438 blev blev Coef. Std. Err. z P>|z| [95% Conf. Interval] tang -.0556974 .0077531 -7.18 0.000 -.0708931 -.0405016 size .0109522 .0013273 8.25 0.000 .0083508 .0135537 prof -.122085 .0093916 -13.00 0.000 -.1404923 -.1036778 liq -.0030842 .0006112 -5.05 0.000 -.0042822 -.0018862 mtb .0007983 .0006802 1.17 0.241 -.0005349 .0021315 ndts -.1792125 .0387617 -4.62 0.000 -.2551841 -.1032409 sew .0052268 .001692 3.09 0.002 .0019106 .008543 fam .0149666 .0069934 2.14 0.032 .0012598 .0286734 DL.(L.blev fam sew ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) _cons 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year 1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for levels equation L(1/3).(L.blev fam sew ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year) 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year D.(1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for first differences equation (Robust, but weakened by many instruments.) Hansen test of overid. restrictions: chi2(255) = 250.90 Prob > chi2 = 0.561 (Not robust, but not weakened by many instruments.) Sargan test of overid. restrictions: chi2(255) = 330.34 Prob > chi2 = 0.001 Arellano-Bond test for AR(2) in first differences: z = -1.24 Pr > z = 0.215 Arellano-Bond test for AR(1) in first differences: z = -8.82 Pr > z = 0.000 37 2. Kết quả hồi quy SGMM các mô hình có biến tương tác 2.1. Mô hình 2.1: Biến FAM là biến độc lập Difference-in-Sargan/Hansen statistics may be negative. Using a generalized inverse to calculate optimal weighting matrix for two-step estimation. Warning: Two-step estimated covariance matrix of moments is singular. Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm. > liq prof size tang, lag (1 1) equation(level)) iv(i.ind i.year) twostep > pba_fam ndts mtb liq prof size tang, lag(1 3) equation(diff)) gmm(l.blev pba fam pba_fam ndts mtb . xtabond2 blev l.blev pba fam pba_fam ndts mtb liq prof size tang i.ind i.year, gmm(l.blev pba fam Prob > chi2 = 0.000 max = 10 Wald chi2(35) = 75892.66 avg = 8.93 Number of instruments = 352 Obs per group: min = 1 Time variable : year Number of groups = 390 Group variable: firm Number of obs = 3482 Dynamic panel-data estimation, two-step system GMM tang -.0705639 .0252084 -2.80 0.005 -.1199716 -.0211563 size .0121465 .0031366 3.87 0.000 .0059988 .0182941 prof -.131246 .0078813 -16.65 0.000 -.1466931 -.1157988 liq -.0149974 .001904 -7.88 0.000 -.0187291 -.0112657 mtb .002577 .0009569 2.69 0.007 .0007016 .0044525 ndts -.2773513 .1416688 -1.96 0.050 -.555017 .0003145 pba_fam -.0434827 .0109114 -3.99 0.000 -.0648686 -.0220967 fam .0131488 .0053438 2.46 0.014 .0026752 .0236225 pba .0164318 .0043749 3.76 0.000 .0078571 .0250065 L1. .7938374 .0138562 57.29 0.000 .7666798 .8209951 blev blev Coef. Std. Err. z P>|z| [95% Conf. Interval] DL.(L.blev pba fam pba_fam ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) _cons 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year 1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for levels equation L(1/3).(L.blev pba fam pba_fam ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year) 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year D.(1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for first differences equation (Robust, but weakened by many instruments.) Hansen test of overid. restrictions: chi2(170) = 176.82 Prob > chi2 = 0.344 (Not robust, but not weakened by many instruments.) Sargan test of overid. restrictions: chi2(170) = 279.63 Prob > chi2 = 0.000 Arellano-Bond test for AR(2) in first differences: z = -1.44 Pr > z = 0.149 Arellano-Bond test for AR(1) in first differences: z = -9.04 Pr > z = 0.000 38 2.2. Mô hình 2.2: Biến CEO là biến độc lập Difference-in-Sargan/Hansen statistics may be negative. Using a generalized inverse to calculate optimal weighting matrix for two-step estimation. Warning: Two-step estimated covariance matrix of moments is singular. Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm. > liq prof size tang, lag (1 1) equation(level)) iv(i.ind i.year) twostep > pba_ceo ndts mtb liq prof size tang, lag(1 3) equation(diff)) gmm(l.blev pba ceo pba_ceo ndts mtb . xtabond2 blev l.blev pba ceo pba_ceo ndts mtb liq prof size tang i.ind i.year, gmm(l.blev pba ceo Prob > chi2 = 0.000 max = 10 Wald chi2(35) = 433447.63 avg = 8.93 Number of instruments = 353 Obs per group: min = 1 Time variable : year Number of groups = 390 Group variable: firm Number of obs = 3482 Dynamic panel-data estimation, two-step system GMM tang -.0942695 .0222524 -4.24 0.000 -.1378833 -.0506557 size .0129918 .00281 4.62 0.000 .0074844 .0184993 prof -.1252348 .0071933 -17.41 0.000 -.1393335 -.1111362 liq -.0163994 .001707 -9.61 0.000 -.019745 -.0130538 mtb .0024216 .0008879 2.73 0.006 .0006814 .0041619 ndts -.3641976 .1157638 -3.15 0.002 -.5910904 -.1373047 pba_ceo -.0260756 .0091858 -2.84 0.005 -.0440794 -.0080718 ceo .0052866 .0064653 0.82 0.414 -.0073852 .0179585 pba .0072328 .0032767 2.21 0.027 .0008105 .0136551 L1. .7719882 .0128734 59.97 0.000 .7467568 .7972196 blev blev Coef. Std. Err. z P>|z| [95% Conf. Interval] DL.(L.blev pba ceo pba_ceo ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) _cons 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year 1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for levels equation L(1/3).(L.blev pba ceo pba_ceo ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year) 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year D.(1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for first differences equation (Robust, but weakened by many instruments.) Hansen test of overid. restrictions: chi2(210) = 199.48 Prob > chi2 = 0.688 (Not robust, but not weakened by many instruments.) Sargan test of overid. restrictions: chi2(210) = 342.29 Prob > chi2 = 0.000 Arellano-Bond test for AR(2) in first differences: z = -1.58 Pr > z = 0.115 Arellano-Bond test for AR(1) in first differences: z = -8.76 Pr > z = 0.000 39 2.3. Mô hình 2.3: Biến GEN là biến độc lập Difference-in-Sargan/Hansen statistics may be negative. Using a generalized inverse to calculate optimal weighting matrix for two-step estimation. Warning: Two-step estimated covariance matrix of moments is singular. Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm. > liq prof size tang, lag (1 1) equation(level)) iv(i.ind i.year) twostep > pba_gen ndts mtb liq prof size tang, lag(1 3) equation(diff)) gmm(l.blev pba gen pba_gen ndts mtb . xtabond2 blev l.blev pba gen pba_gen ndts mtb liq prof size tang i.ind i.year, gmm(l.blev pba gen Prob > chi2 = 0.000 max = 10 Wald chi2(35) = 80395.06 avg = 8.93 Number of instruments = 352 Obs per group: min = 1 Time variable : year Number of groups = 390 Group variable: firm Number of obs = 3482 Dynamic panel-data estimation, two-step system GMM tang -.0612577 .0237667 -2.58 0.010 -.1078395 -.0146759 size .0117556 .0032008 3.67 0.000 .0054822 .018029 prof -.1298917 .0077742 -16.71 0.000 -.1451289 -.1146545 liq -.0144499 .0018505 -7.81 0.000 -.0180767 -.010823 mtb .0026161 .0009458 2.77 0.006 .0007622 .0044699 ndts -.2598844 .1369725 -1.90 0.058 -.5283454 .0085767 pba_gen -.0226464 .0131079 -1.73 0.084 -.0483374 .0030447 gen .0056178 .0094825 0.59 0.554 -.0129676 .0242032 pba .0069907 .0033778 2.07 0.038 .0003703 .0136112 L1. .7982089 .0136036 58.68 0.000 .7715464 .8248714 blev blev Coef. Std. Err. z P>|z| [95% Conf. Interval] DL.(L.blev pba gen pba_gen ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) _cons 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year 1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for levels equation L(1/3).(L.blev pba gen pba_gen ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year) 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year D.(1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for first differences equation (Robust, but weakened by many instruments.) Hansen test of overid. restrictions: chi2(170) = 177.89 Prob > chi2 = 0.324 (Not robust, but not weakened by many instruments.) Sargan test of overid. restrictions: chi2(170) = 283.40 Prob > chi2 = 0.000 Arellano-Bond test for AR(2) in first differences: z = -1.63 Pr > z = 0.102 Arellano-Bond test for AR(1) in first differences: z = -9.09 Pr > z = 0.000 40 2.4. Mô hình 2.4: Biến BOD là biến độc lập Difference-in-Sargan/Hansen statistics may be negative. Using a generalized inverse to calculate optimal weighting matrix for two-step estimation. Warning: Two-step estimated covariance matrix of moments is singular. Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm. > liq prof size tang, lag (1 1) equation(level)) iv(i.ind i.year) twostep > pba_bod ndts mtb liq prof size tang, lag(1 3) equation(diff)) gmm(l.blev pba bod pba_bod ndts mtb . xtabond2 blev l.blev pba bod pba_bod ndts mtb liq prof size tang i.ind i.year, gmm(l.blev pba bod Prob > chi2 = 0.000 max = 10 Wald chi2(35) = 1.08e+06 avg = 8.93 Number of instruments = 370 Obs per group: min = 1 Time variable : year Number of groups = 390 Group variable: firm Number of obs = 3482 Dynamic panel-data estimation, two-step system GMM tang -.0643794 .0241698 -2.66 0.008 -.1117513 -.0170075 size .0124616 .003071 4.06 0.000 .0064425 .0184807 prof -.1295627 .0077862 -16.64 0.000 -.1448234 -.1143019 liq -.0146763 .001896 -7.74 0.000 -.0183925 -.0109602 mtb .0023233 .0009383 2.48 0.013 .0004843 .0041624 ndts -.2539557 .1403092 -1.81 0.070 -.5289567 .0210453 pba_bod -.0831432 .0222937 -3.73 0.000 -.126838 -.0394484 bod .0287579 .0117565 2.45 0.014 .0057156 .0518003 pba .0144107 .0041917 3.44 0.001 .0061952 .0226263 L1. .7940152 .0137576 57.71 0.000 .7670508 .8209797 blev blev Coef. Std. Err. z P>|z| [95% Conf. Interval] DL.(L.blev pba bod pba_bod ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) _cons 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year 1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for levels equation L(1/3).(L.blev pba bod pba_bod ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year) 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year D.(1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for first differences equation (Robust, but weakened by many instruments.) Hansen test of overid. restrictions: chi2(170) = 180.70 Prob > chi2 = 0.273 (Not robust, but not weakened by many instruments.) Sargan test of overid. restrictions: chi2(170) = 279.50 Prob > chi2 = 0.000 Arellano-Bond test for AR(2) in first differences: z = -1.57 Pr > z = 0.118 Arellano-Bond test for AR(1) in first differences: z = -9.08 Pr > z = 0.000 41 2.5. Mô hình 2.5: Biến SEW là biến độc lập Difference-in-Sargan/Hansen statistics may be negative. Using a generalized inverse to calculate optimal weighting matrix for two-step estimation. Warning: Two-step estimated covariance matrix of moments is singular. Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm. > liq prof size tang, lag (1 1) equation(level)) iv(i.ind i.year) twostep > pba_sew ndts mtb liq prof size tang, lag(1 3) equation(diff)) gmm(l.blev pba sew pba_sew ndts mtb . xtabond2 blev l.blev pba sew pba_sew ndts mtb liq prof size tang i.ind i.year, gmm(l.blev pba sew Prob > chi2 = 0.000 max = 10 Wald chi2(35) = 155470.44 avg = 8.93 Number of instruments = 370 Obs per group: min = 1 Time variable : year Number of groups = 390 Group variable: firm Number of obs = 3482 Dynamic panel-data estimation, two-step system GMM tang -.0634566 .0242513 -2.62 0.009 -.1109884 -.0159249 size .0129718 .00315 4.12 0.000 .0067979 .0191457 prof -.1290612 .0079915 -16.15 0.000 -.1447243 -.1133982 liq -.0149411 .00193 -7.74 0.000 -.0187239 -.0111584 mtb .0023521 .0009467 2.48 0.013 .0004966 .0042076 ndts -.2707068 .1422544 -1.90 0.057 -.5495202 .0081067 pba_sew -.0225901 .0046617 -4.85 0.000 -.0317267 -.0134534 sew .0075113 .0024584 3.06 0.002 .0026928 .0123298 pba .0159033 .0040332 3.94 0.000 .0079983 .0238083 L1. .7923179 .0138835 57.07 0.000 .7651068 .8195291 blev blev Coef. Std. Err. z P>|z| [95% Conf. Interval] DL.(L.blev pba sew pba_sew ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) _cons 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year 1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for levels equation L(1/3).(L.blev pba sew pba_sew ndts mtb liq prof size tang) GMM-type (missing=0, separate instruments for each period unless collapsed) 2015.year 2016.year 2017.year 2018.year 2019.year 2020.year) 12.ind 13.ind 14.ind 2010b.year 2011.year 2012.year 2013.year 2014.year D.(1b.ind 2.ind 3.ind 4.ind 5.ind 6.ind 7.ind 8.ind 9.ind 10.ind 11.ind Standard Instruments for first differences equation (Robust, but weakened by many instruments.) Hansen test of overid. restrictions: chi2(170) = 182.35 Prob > chi2 = 0.245 (Not robust, but not weakened by many instruments.) Sargan test of overid. restrictions: chi2(170) = 274.84 Prob > chi2 = 0.000 Arellano-Bond test for AR(2) in first differences: z = -1.47 Pr > z = 0.142 Arellano-Bond test for AR(1) in first differences: z = -9.05 Pr > z = 0.000