Diversification strategies, bank risk and performance empirical evidence from vietnam

MINISTRY OF EDUCATION AND TRAINING UNIVERSITY OF ECONOMICS HO CHI MINH CITY - PHAM KHANH DUY DIVERSIFICATION STRATEGIES, BANK RISK AND PERFORMANCE: EMPIRICAL EVIDENCE FROM VIETNAM Major: Finance and Banking Code: 9340201 DOCTORAL DISSERTATION ACADEMIC ADVISOR: Assoc Professor Dr TRUONG THI HONG Ho Chi Minh City - 2021 ii STATEMENT OF AUTHENTICATION I solemnly declare that this dissertation, “Diversification strategies, bank risk and performance: Empirical evidence from Vietnam”, is my own research Except for the references cited in this dissertation, I hereby guarantee that the whole or any part of this dissertation has never been published or used to obtain a degree elsewhere Any products/studies of other authors that have been used in this dissertation were properly cited This dissertation has never been submitted to any university or institution Ho Chi Minh City, 2021 ACKNOWLEDGEMENT In completing this dissertation, I would like to express my gratitude to the University of Economics Ho Chi Minh City (UEH), which provides an ideal environment and financial support for my research journey At UEH, I have gained valuable academic knowledge and research skills as a lecturer and as a Ph.D candidate My utmost appreciation is to my academic supervisor, Assoc Prof Dr Truong Thi Hong, for dedicated, whole-heartedly coaching and guidance throughout the completion of my thesis My gratitude also goes to my dearest colleagues at the School of Banking and the School of UEH Graduate for their continuously dedicated support, contributions, and guidance, for me in writing this dissertation I want to thank Prof Vo Xuan Vinh for the preliminary research ideas and suggestions My special thanks to Dr Nguyen Huu Huan, Dr Ngo Minh Vu, Dr Phan Chung Thuy, Dr Hoang Hai Yen, and Dr Ngo Minh Hai for their best efforts in helping me with various challenges during my Ph.D candidature Saying thank you is just simply not enough for what they have offered me Finally and most importantly, my beloved family and best friends have always been by my sides, with unconditional love, encouragement, and endless support Without them, this dissertation would have never been written Thank you all for the most profound appreciation Ho Chi Minh City, October 2021 Pham Khanh Duy TABLE OF CONTENTS STATEMENT OF AUTHENTICATION i ACKNOWLEDGEMENT ii TABLE OF CONTENTS iii ABBREVIATION LIST vii LIST OF TABLES viii LIST OF FIGURES ix ABSTRACT x TÓM TẮT xi CHAPTER INTRODUCTION 1.1 Research motivation 1.2 Research background 1.3 Research gap identification 11 1.4 Research objectives 12 1.5 Research questions 14 1.6 The scope of research 15 1.7 Research procedure and methodology 15 1.8 Research contributions 17 1.9 Structure of the dissertation 19 CHAPTER LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT 22 2.1 Bank diversification definition 22 2.2 Classification of bank diversification strategies 24 2.2.1 Asset diversification in the banking sector 28 2.2.2 Income diversification in the banking sector 30 2.2.3 Funding diversification in the banking sector 32 2.3 Theories of bank diversification 35 2.3.1 Theories of diversification 35 2.3.2 Theories of bank diversification 40 2.4 Concept and measurement of bank risk and performance 43 2.4.1 Concept and measurement of bank risk 43 2.4.2 Concept and measurement of bank performance .46 2.5 Theoretical overview of banking crisis 47 2.6 The impact of diversification on bank risk and performance 50 Although the concern on the impact of diversification on banks' risks and performance is discussed in various articles, the conclusion of this topic is still unclear 50 2.6.1 Asset diversification, bank risk, and performance 52 2.6.2 Income diversification, bank risk, and performance 56 2.6.3 Funding diversification, bank risk, and performance .60 2.6.4 Combinations of diversification strategies 61 2.6.5 Bank diversification, risk and performance in the financial crisis 62 2.6.6 Role of ownership structure in the nexus between diversification, risk and performance 64 2.7 Summary 67 CHAPTER DATA AND METHODOLOGY 68 3.1 Data sample 68 3.2 Construction of variables 70 3.2.1 Dependent variables – Bank risk and performance 70 3.2.2 Key explanatory variables - Bank diversification measures .72 3.2.3 Control variables 74 3.2.5 The role of financial distress and bank ownership 76 3.3 Econometric models 79 3.4 Robustness check 83 3.5 Conclusion 86 CHAPTER EMPIRICAL RESULTS AND DISCUSSION 87 4.1 Descriptive statistics 87 4.2 Correlation matrix 89 4.3 Empirical results 91 4.3.1 Diversification strategies, bank’s risk and performance 91 4.3.2 Diversification strategy combination, bank risk and performance 102 4.3.3 Diversification, bank risk and performance during the crisis 105 4.3.4 The role of bank ownership structure 109 4.4 Robustness check results 117 4.5 Conclusion 117 CHAPTER CONCLUSION 118 5.1 Summary of research findings 118 5.2 Contributions 119 5.3 Policy implications 120 5.4 Limitations 122 LIST OF PUBLICATION i REFERENCES ii APPENDIX xv ABBREVIATION LIST ASEAN Association of Southeast Asian Nations BSI The banking stability index CPI Consumer Price Index DEA Data envelopment analysis FOB Foreign-owned bank GDP Gross domestic product GMM Generalized Method of Moments HHI Herfindahl-Hirschman Index JPoD Joint Probability of Distress JSB Joint-stock bank NIM Net Interest Margin NPL Non-performing Loan PB Policy Bank POLS Pooled Ordinary Least Squared RBV Resources Based View SBV State Bank of Vietnam SGMM System Generalized Method of Moments SOB State-owned Bank US The United States of America WTO World Trade Organisation LIST OF TABLES Table 1.1: Descriptive number of banks in Vietnam, 2005-2019 Table 2.1: Summary of hypotheses associated with research questions 66 Table 3.1: Definition of Variables 78 Table 3.2: The expected sign of variables 79 Table 3.3: Testing steps for the hypotheses associated with the research question 85 Table 4.1: Summary statistics of variables 88 Table 4.2: Cross-Correlation Matrix of Variables 90 Table 4.3: Empirical results – Using Pooled OLS estimation Method Basic model of the impact of diversification on banks performance and risks 94 Table 4.4: Empirical results – Using Random effects estimation Method Basic model of the impact of diversification on banks performance and risks 96 Table 4.5: Empirical results – Using Fixed effects estimation approach Basic model of the impact of diversification on banks performance and risks 98 Table 4.6: Empirical results – Using SGMM estimation approach The baseline model of the effectiveness of diversification strategies on banks performance and risks 100 Table 4.7: Empirical Results Using SGMM Estimation Approach 103 Table 4.8: Empirical results Using SGMM Estimation Approach 107 Table 4.9: Regression Results Using SGMM Estimation Approach .111 Table 4.10: Regression Results Using SGMM Estimation Approach Diversification, risk and performance of Foreign banks 113 Table 4.11: Robustness test results 115 LIST OF FIGURES Figure 1.1: Diversification indexes of the Vietnam banks from 2005 to 2019 Figure 1.2: Research procedures 17 Figure 2.1: Research conceptual model 67 clxxv APPENDIX 17 Empirical results – Using SGMM estimation approach The baseline model of the effectiveness of diversification strategies on banks performance and risks Dynamic panel-data estimation, two-step system GMM Group variable: id Number of obs = 365 Time variable : yr Number of groups = 33 Number of instruments = 21 Obs per group: = F(6, 32) = 42.36 avg = 11.06 Prob > F = 0.000 max = 15 roe | Coef Std Err t P>|t| [95% Conf Interval] + astdiv | 0838574 0481959 1.74 0.091 -.0143144 1820292 eta | -.0050462 0454867 -0.11 0.912 -.0976995 0876071 lnasset | 0270458 0033139 8.16 0.000 0202957 0337959 llp | -3.344463 7147818 -4.68 0.000 -4.800426 -1.8885 gdp | 7882147 6687761 1.18 0.247 -.5740376 2.150467 inf | 3291028 0397735 8.27 0.000 2480869 4101187 _cons | -.4626077 0758065 -6.10 0.000 -.6170205 -.3081949 Instruments for orthogonal deviations equation Standard FOD.(eta lnasset llp gdp inf) GMM-type (missing=0, separate instruments for each period unless collapsed) L(1/14).astdiv collapsed Arellano-Bond test for AR(1) in first differences: z = -1.33 Pr > z = 0.183 Arellano-Bond test for AR(2) in first differences: z = Sargan test of overid restrictions: chi2(14) -1.99 Pr > z = 0.146 = 35.65 Prob > chi2 = 0.001 (Not robust, but not weakened by many instruments.) Hansen test of overid restrictions: chi2(14) (Robust, but weakened by many instruments.) = 21.77 Prob > chi2 = 0.183 Dynamic panel-data estimation, two-step system GMM Group variable: id Number of obs = 393 Time variable : yr Number of groups = 34 Number of instruments = 21 Obs per group: = F(6, 33) = 22.42 avg = 11.56 Prob > F = 0.000 max = 15 roe | Coef Std Err t P>|t| [95% Conf Interval] + incdiv | 2297641 0558655 4.11 0.000 116105 3434233 eta | -.0764665 0442824 -1.73 0.094 -.1665597 0136267 lnasset | 0171849 0036531 4.70 0.000 0097527 0246172 llp | -3.477695 6557074 -5.30 0.000 -4.811741 -2.143648 gdp | 7947485 6194294 1.28 0.208 -.46549 2.054987 inf | 3880792 0698039 5.56 0.000 2460621 5300964 _cons | -.2796313 0828746 -3.37 0.002 -.4482409 -.1110216 Instruments for orthogonal deviations equation Standard FOD.(eta lnasset llp gdp inf) GMM-type (missing=0, separate instruments for each period unless collapsed) L(1/14).incdiv collapsed Arellano-Bond test for AR(1) in first differences: z = -1.46 Pr > z = 0.145 Arellano-Bond test for AR(2) in first differences: z = Sargan test of overid restrictions: chi2(14) -2.60 Pr > z = 0.109 = 20.49 Prob > chi2 = 0.115 (Not robust, but not weakened by many instruments.) Hansen test of overid restrictions: chi2(14) = 25.24 Prob > chi2 = 0.132 (Robust, but weakened by many instruments.) Difference-in-Hansen tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(13) Difference (null H = exogenous): chi2(1) = 22.98 Prob > chi2 = 0.042 = 2.25 Prob > chi2 = 0.133 = 11.35 Prob > chi2 = 0.253 iv(eta lnasset llp gdp inf) Hansen test excluding group: chi2(9) Dynamic panel-data estimation, two-step system GMM Group variable: id Number of obs = 393 Time variable : yr Number of groups = 34 Number of instruments = 21 Obs per group: = F(6, 33) = 37.30 avg = 11.56 Prob > F = 0.000 max = 15 roe | Coef Std Err t P>|t| [95% Conf Interval] + fundiv | 3406604 0550289 6.19 0.000 2287032 4526175 eta | 0381629 0349344 1.09 0.283 -.0329116 1092375 lnasset | 0307798 0044489 6.92 0.000 0217285 0398311 llp | -1.574691 8261627 -1.91 0.065 -3.255531 1061498 gdp | 1.772954 5667137 3.13 0.004 6199664 2.925942 inf | 16953 074406 2.28 0.029 0181499 32091 _cons | -.7356854 1129203 -6.52 0.000 -.9654234 -.5059474 Warning: Uncorrected two-step standard errors are unreliable Instruments for orthogonal deviations equation Standard FOD.(eta lnasset llp gdp inf) GMM-type (missing=0, separate instruments for each period unless collapsed) L(1/14).incdiv collapsed Arellano-Bond test for AR(1) in first differences: z = -1.50 Pr > z = 0.135 Arellano-Bond test for AR(2) in first differences: z = Sargan test of overid restrictions: chi2(14) -1.57 Pr > z = 0.117 = 29.16 Prob > chi2 = 0.010 (Not robust, but not weakened by many instruments.) Hansen test of overid restrictions: chi2(14) (Robust, but weakened by many instruments.) = 18.21 Prob > chi2 = 0.197 Dynamic panel-data estimation, two-step system GMM Group variable: id Number of obs = 365 Time variable : yr Number of groups = 33 Number of instruments = 21 Obs per group: = F(6, 32) = 16.89 avg = 11.06 Prob > F = 0.000 max = 15 roa | Coef Std Err t P>|t| [95% Conf Interval] + astdiv | 0098162 00643 1.53 0.137 -.0032813 0229136 eta | 0405653 0193694 2.09 0.044 0011112 0800194 lnasset | 001539 0004291 3.59 0.001 000665 002413 llp | -.3252691 0570422 -5.70 0.000 -.4414604 -.2090779 gdp | 030479 0570327 0.53 0.597 -.0856929 1466509 inf | 0291978 0052171 5.60 0.000 018571 0398246 _cons | -.0274353 0125714 -2.18 0.037 -.0530423 -.0018283 Instruments for orthogonal deviations equation Standard FOD.(eta lnasset llp gdp inf) GMM-type (missing=0, separate instruments for each period unless collapsed) L(1/14).astdiv collapsed Arellano-Bond test for AR(1) in first differences: z = -1.60 Pr > z = 0.109 Arellano-Bond test for AR(2) in first differences: z = Sargan test of overid restrictions: chi2(14) -1.61 Pr > z = 0.108 = 36.55 Prob > chi2 = 0.001 (Not robust, but not weakened by many instruments.) Hansen test of overid restrictions: chi2(14) (Robust, but weakened by many instruments.) = 23.28 Prob > chi2 = 0.156 Dynamic panel-data estimation, two-step system GMM Group variable: id Number of obs = 393 Time variable : yr Number of groups = 34 Number of instruments = 21 Obs per group: = F(6, 33) = 53.14 avg = 11.56 Prob > F = 0.000 max = 15 roa | Coef Std Err t P>|t| [95% Conf Interval] + incdiv | 0352057 0049126 7.17 0.000 0252111 0452004 eta | 0387064 0070851 5.46 0.000 0242917 0531211 lnasset | 0012329 0003418 3.61 0.001 0005375 0019282 llp | -.4018236 0742126 -5.41 0.000 -.5528103 -.2508369 gdp | -.0078893 0659161 -0.12 0.905 -.1419966 126218 inf | 0407192 0085868 4.74 0.000 0232492 0581891 _cons | -.020237 0082336 -2.46 0.019 -.0369883 -.0034856 Instruments for orthogonal deviations equation Standard FOD.(eta lnasset llp gdp inf) GMM-type (missing=0, separate instruments for each period unless collapsed) L(1/14).incdiv collapsed Arellano-Bond test for AR(1) in first differences: z = -1.95 Pr > z = 0.051 Arellano-Bond test for AR(2) in first differences: z = Sargan test of overid restrictions: chi2(14) = -2.07 12.16 Pr > z = 0.138 Prob > chi2 = 0.594 (Not robust, but not weakened by many instruments.) Hansen test of overid restrictions: chi2(14) = 20.92 Prob > chi2 = 0.104 chi2(9) = 11.01 Prob > chi2 = 0.275 Difference (null H = exogenous): chi2(5) = 9.91 Prob > chi2 = 0.078 (Robust, but weakened by many instruments.) Hansen test excluding group: Dynamic panel-data estimation, two-step system GMM Group variable: id Number of obs = 393 Time variable : yr Number of groups = 34 Number of instruments = 21 Obs per group: = F(6, 33) = 41.25 avg = 11.56 Prob > F = 0.000 max = 15 roa | Coef Std Err t P>|t| [95% Conf Interval] + fundiv | 0252324 0063909 3.95 0.000 0122301 0382347 eta | 0540545 0064462 8.39 0.000 0409397 0671694 lnasset | 0021385 0004221 5.07 0.000 0012797 0029972 llp | -.2084307 0578915 -3.60 0.001 -.3262119 -.0906495 gdp | 1422175 0466535 3.05 0.005 0473003 2371346 inf | 0121688 0109513 1.11 0.275 -.0101119 0344495 _cons | -.0546014 010713 -5.10 0.000 -.0763972 -.0328055 Instruments for orthogonal deviations equation Standard FOD.(eta lnasset llp gdp inf) GMM-type (missing=0, separate instruments for each period unless collapsed) L(1/14).incdiv collapsed Arellano-Bond test for AR(1) in first differences: z = -2.03 Pr > z = 0.043 Arellano-Bond test for AR(2) in first differences: z = Sargan test of overid restrictions: chi2(14) -1.46 Pr > z = 0.144 = 38.79 Prob > chi2 = 0.000 (Not robust, but not weakened by many instruments.) Hansen test of overid restrictions: chi2(14) (Robust, but weakened by many instruments.) = 18.64 Prob > chi2 = 0.179 Dynamic panel-data estimation, two-step system GMM Group variable: id Number of obs = 365 Time variable : yr Number of groups = 33 Number of instruments = 21 Obs per group: = F(6, 32) = 11.12 avg = 11.06 Prob > F = 0.000 max = 15 sdroe | Coef Std Err t P>|t| [95% Conf Interval] + astdiv | -.039855 0330261 -1.21 0.236 -.1071271 027417 eta | -.0561724 0167994 -3.34 0.002 -.0903917 -.021953 lnasset | -.0018435 001529 -1.21 0.237 -.0049579 001271 llp | 9026276 3145431 2.87 0.007 2619243 1.543331 gdp | -.4506282 1655674 -2.72 0.010 -.7878779 -.1133786 inf | 0683787 0187521 3.65 0.001 030182 1065755 _cons | 105782 0324976 3.26 0.003 0395865 1719775 Instruments for orthogonal deviations equation Standard FOD.(eta lnasset llp gdp inf) GMM-type (missing=0, separate instruments for each period unless collapsed) L(1/14).astdiv collapsed Arellano-Bond test for AR(1) in first differences: z = -3.04 Pr > z = 0.002 Arellano-Bond test for AR(2) in first differences: z = Sargan test of overid restrictions: chi2(14) -1.68 Pr > z = 0.193 = 12.32 Prob > chi2 = 0.581 (Not robust, but not weakened by many instruments.) Hansen test of overid restrictions: chi2(14) (Robust, but weakened by many instruments.) = 19.32 Prob > chi2 = 0.153 Dynamic panel-data estimation, two-step system GMM Group variable: id Number of obs = 393 Time variable : yr Number of groups = 34 Number of instruments = 21 Obs per group: = F(6, 33) = 9.62 avg = 11.56 Prob > F = 0.000 max = 15 sdroe | Coef Std Err t P>|t| [95% Conf Interval] + incdiv | 0877571 0223252 3.93 0.000 0423362 133178 eta | -.0576269 0253886 -2.27 0.030 -.1092804 -.0059733 lnasset | -.0031102 0013771 -2.26 0.031 -.0059119 -.0003085 llp | 998387 3411193 2.93 0.006 3043746 1.692399 gdp | -.2658198 2289117 -1.16 0.254 -.7315441 1999045 inf | 0901798 0216101 4.17 0.000 0462137 1341459 _cons | 0790316 0311193 2.54 0.016 015719 1423442 Instruments for orthogonal deviations equation Standard FOD.(eta lnasset llp gdp inf) GMM-type (missing=0, separate instruments for each period unless collapsed) L(1/14).incdiv collapsed Arellano-Bond test for AR(1) in first differences: z = -3.32 Pr > z = 0.001 Arellano-Bond test for AR(2) in first differences: z = Sargan test of overid restrictions: chi2(14) = -2.70 Pr > z = 0.107 6.31 Prob > chi2 = 0.958 (Not robust, but not weakened by many instruments.) Hansen test of overid restrictions: chi2(14) (Robust, but weakened by many instruments.) = 18.12 Prob > chi2 = 0.201 Dynamic panel-data estimation, two-step system GMM Group variable: id Number of obs = 393 Time variable : yr Number of groups = 34 Number of instruments = 21 Obs per group: = F(6, 33) = 13.17 avg = 11.56 Prob > F = 0.000 max = 15 sdroe | Coef Std Err t P>|t| [95% Conf Interval] + fundiv | -.0328864 0349334 -0.94 0.353 -.1039589 038186 eta | -.0110554 0168456 -0.66 0.516 -.045328 0232173 lnasset | -.0018667 0023704 -0.79 0.437 -.0066893 0029559 llp | 9913064 1965449 5.04 0.000 5914327 1.39118 gdp | -.3201636 2035245 -1.57 0.125 -.7342374 0939102 inf | 1024668 0277915 3.69 0.001 0459246 159009 _cons | 0849832 068332 1.24 0.222 -.0540392 2240056 Instruments for orthogonal deviations equation Standard FOD.(eta lnasset llp gdp inf) GMM-type (missing=0, separate instruments for each period unless collapsed) L(1/14).incdiv collapsed Arellano-Bond test for AR(1) in first differences: z = -3.85 Pr > z = 0.000 Arellano-Bond test for AR(2) in first differences: z = Sargan test of overid restrictions: chi2(14) -1.13 Pr > z = 0.260 = 20.82 Prob > chi2 = 0.106 (Not robust, but not weakened by many instruments.) Hansen test of overid restrictions: chi2(14) = 17.46 Prob > chi2 = 0.232 Dynamic panel-data estimation, two-step system GMM Group variable: id Number of obs = 365 Time variable : yr Number of groups = 33 Number of instruments = 21 Obs per group: = F(6, 32) = 24.78 avg = 11.06 Prob > F = 0.000 max = 15 sdroa | Coef Std Err t P>|t| [95% Conf Interval] 0.05 0.961 -.006646 0069788 + astdiv | 0001664 0033444 eta | 0147252 0031916 4.61 0.000 0082241 0212263 lnasset | -.0006763 0002373 -2.85 0.008 -.0011596 -.000193 llp | 0791803 0357368 2.22 0.034 0063869 1519737 gdp | -.0291879 0153491 -1.90 0.066 -.060453 0020772 inf | 0048582 0020795 2.34 0.026 0006225 009094 _cons | 0146057 004595 3.18 0.003 005246 0239654 Instruments for orthogonal deviations equation Standard FOD.(eta lnasset llp gdp inf) GMM-type (missing=0, separate instruments for each period unless collapsed) L(1/14).astdiv collapsed Arellano-Bond test for AR(1) in first differences: z = -1.46 Pr > z = 0.143 Arellano-Bond test for AR(2) in first differences: z = Sargan test of overid restrictions: chi2(14) = 0.53 Pr > z = 0.593 9.60 Prob > chi2 = 0.791 (Not robust, but not weakened by many instruments.) Hansen test of overid restrictions: chi2(14) (Robust, but weakened by many instruments.) = 18.19 Prob > chi2 = 0.198 Dynamic panel-data estimation, two-step system GMM Group variable: id Number of obs = 393 Time variable : yr Number of groups = 34 Number of instruments = 21 Obs per group: = F(6, 33) = 151.27 avg = 11.56 Prob > F = 0.000 max = 15 sdroa | Coef Std Err t P>|t| [95% Conf Interval] 4.41 0.000 0056354 + incdiv | 0104691 0023758 0153028 eta | 0174833 004122 4.24 0.000 0090971 0258696 lnasset | -.0008753 0002749 -3.18 0.003 -.0014346 -.0003159 llp | 0741405 0354625 2.09 0.044 0019915 1462895 gdp | -.0309636 0266183 -1.16 0.253 -.0851188 0231917 inf | 0085212 0020395 4.18 0.000 0043719 0126706 _cons | 0165253 0059049 2.80 0.009 0045117 0285389 Instruments for orthogonal deviations equation Standard FOD.(eta lnasset llp gdp inf) GMM-type (missing=0, separate instruments for each period unless collapsed) L(1/14).incdiv collapsed Arellano-Bond test for AR(1) in first differences: z = -1.75 Pr > z = 0.079 Arellano-Bond test for AR(2) in first differences: z = Sargan test of overid restrictions: chi2(14) = 0.42 Pr > z = 0.676 9.61 Prob > chi2 = 0.790 (Not robust, but not weakened by many instruments.) Hansen test of overid restrictions: chi2(14) (Robust, but weakened by many instruments.) = 18.77 Prob > chi2 = 0.174 Dynamic panel-data estimation, two-step system GMM Group variable: id Number of obs = 393 Time variable : yr Number of groups = 34 Number of instruments = 21 Obs per group: = F(6, 33) = 209.46 avg = 11.56 Prob > F = 0.000 max = 15 sdroa | Coef Std Err t P>|t| [95% Conf Interval] + fundiv | 0022956 0024996 0.92 0.365 -.0027898 0073811 eta | 0209437 0036511 5.74 0.000 0135156 0283719 lnasset | -.0007705 0003173 -2.43 0.021 -.0014161 -.0001248 llp | 1061839 0356864 2.98 0.005 0335793 1787884 gdp | 0027413 0217793 0.13 0.901 -.041569 0470516 inf | 0029437 0023557 1.25 0.220 -.0018491 0077365 _cons | 0128031 0072349 1.77 0.086 -.0019165 0275227 Instruments for orthogonal deviations equation Standard FOD.(eta lnasset llp gdp inf) GMM-type (missing=0, separate instruments for each period unless collapsed) L(1/14).incdiv collapsed Arellano-Bond test for AR(1) in first differences: z = -1.61 Pr > z = 0.108 Arellano-Bond test for AR(2) in first differences: z = Sargan test of overid restrictions: chi2(14) 0.63 Pr > z = 0.526 = 19.77 Prob > chi2 = 0.138 (Not robust, but not weakened by many instruments.) Hansen test of overid restrictions: chi2(14) (Robust, but weakened by many instruments.) = 17.60 Prob > chi2 = 0.226 Dynamic panel-data estimation, two-step system GMM Group variable: id Number of obs = 365 Time variable : yr Number of groups = 33 Number of instruments = 21 Obs per group: = F(6, 32) = 42.24 avg = 11.06 Prob > F = 0.000 max = 15 zscore | Coef Std Err t P>|t| [95% Conf Interval] + astdiv | 4.011716 2.858082 1.40 0.170 -1.810007 9.833438 eta | 81.13133 11.84624 6.85 0.000 57.00132 105.2613 lnasset | 4110081 4300376 0.96 0.346 -.4649499 1.286966 llp | 28.04954 46.74314 0.60 0.553 -67.16313 123.2622 gdp | -35.14264 23.42719 -1.50 0.143 -82.86226 12.57698 inf | 13.373 2.61239 5.12 0.000 8.05173 18.69426 _cons | -.8199364 8.261271 -0.10 0.922 -17.6476 16.00772 Instruments for orthogonal deviations equation Standard FOD.(eta lnasset llp gdp inf) GMM-type (missing=0, separate instruments for each period unless collapsed) L(1/14).astdiv collapsed Arellano-Bond test for AR(1) in first differences: z = -1.47 Pr > z = 0.141 Arellano-Bond test for AR(2) in first differences: z = Sargan test of overid restrictions: chi2(14) -0.24 Pr > z = 0.813 = 66.90 Prob > chi2 = 0.000 (Not robust, but not weakened by many instruments.) Hansen test of overid restrictions: chi2(14) (Robust, but weakened by many instruments.) = 16.95 Prob > chi2 = 0.259 Dynamic panel-data estimation, two-step system GMM Group variable: id Number of obs = 393 Time variable : yr Number of groups = 34 Number of instruments = 21 Obs per group: = F(6, 33) = 35.45 avg = 11.56 Prob > F = 0.000 max = 15 zscore | Coef Std Err t P>|t| [95% Conf Interval] 4.62 0.000 3.816434 9.834245 + incdiv | 6.825339 1.47893 eta | 69.86351 12.09566 5.78 0.000 45.2547 94.47232 lnasset | -.449885 4031173 -1.12 0.272 -1.270033 3702633 llp | -11.83756 59.1028 -0.20 0.842 -132.0831 108.408 gdp | -46.40584 21.96854 -2.11 0.042 -91.10116 -1.710515 inf | 13.9932 3.573093 3.92 0.000 6.723683 21.26271 _cons | 18.46875 7.374461 2.50 0.017 3.465293 33.4722 Instruments for orthogonal deviations equation Standard FOD.(eta lnasset llp gdp inf) GMM-type (missing=0, separate instruments for each period unless collapsed) L(1/14).incdiv collapsed Arellano-Bond test for AR(1) in first differences: z = -0.65 Pr > z = 0.513 Arellano-Bond test for AR(2) in first differences: z = Sargan test of overid restrictions: chi2(14) -0.19 Pr > z = 0.848 = 38.57 Prob > chi2 = 0.000 (Not robust, but not weakened by many instruments.) Hansen test of overid restrictions: chi2(14) (Robust, but weakened by many instruments.) = 19.10 Prob > chi2 = 0.161 Dynamic panel-data estimation, two-step system GMM Group variable: id Number of obs = 393 Time variable : yr Number of groups = 34 Number of instruments = 21 Obs per group: = F(6, 33) = 35.49 avg = 11.56 Prob > F = 0.000 max = 15 zscore | Coef Std Err t P>|t| [95% Conf Interval] + fundiv | 13.50463 4.653364 2.90 0.007 4.037294 22.97197 eta | 83.37977 10.6052 7.86 0.000 61.80332 104.9562 lnasset | 213859 5386394 0.40 0.694 -.882011 1.309729 llp | 45.91352 37.97608 1.21 0.235 -31.34939 123.1764 gdp | 14.69974 23.97777 0.61 0.544 -34.0834 63.48288 inf | 5.903287 3.275631 1.80 0.081 -.7610338 12.56761 _cons | -4.447661 10.39861 -0.43 0.672 -25.6038 16.70848 Instruments for orthogonal deviations equation Standard FOD.(eta lnasset llp gdp inf) GMM-type (missing=0, separate instruments for each period unless collapsed) L(1/14).incdiv collapsed Arellano-Bond test for AR(1) in first differences: z = -0.98 Pr > z = 0.327 Arellano-Bond test for AR(2) in first differences: z = Sargan test of overid restrictions: chi2(14) -0.18 Pr > z = 0.854 = 37.92 Prob > chi2 = 0.001 (Not robust, but not weakened by many instruments.) Hansen test of overid restrictions: chi2(14) (Robust, but weakened by many instruments.) = 14.45 Prob > chi2 = 0.417 ... 50 2.6.1 Asset diversification, bank risk, and performance 52 2.6.2 Income diversification, bank risk, and performance 56 2.6.3 Funding diversification, bank risk, and performance .60... risk and performance 91 4.3.2 Diversification strategy combination, bank risk and performance 102 4.3.3 Diversification, bank risk and performance during the crisis 105 4.3.4 The role of bank. .. RQ1.2: Does income diversification impact bank risk and performance? RQ1.3: Does funding diversification impact bank risk and performance? RQ1.4: Can banks combine these three diversification strategies