Factors impacting on abnormal investment of listed firms in ho chi minh stock exchange luận văn thạc sĩ tài chính ngân hàng

MINISTRY OF EDUCATION AND TRAINING STATE BANK OF VIETNAM BANKING UNIVERSITY OF HO CHI MINH CITY KIEU CONG BAO TRAN FACTORS IMPACTING ON ABNORMAL INVESTMENT OF LISTED FIRMS IN HO CHI MINH STOCK EXCHANGE MASTER'S THESIS Ho Chi Minh City - 2020 MINISTRY OF EDUCATION AND TRAINING STATE BANK OF VIETNAM BANKING UNIVERSITY OF HO CHI MINH CITY KIEU CONG BAO TRAN FACTORS IMPACTING ON ABNORMAL INVESTMENT OF LISTED FIRMS IN HO CHI MINH STOCK EXCHANGE Major: Finance and Banking Code : 8340201 MASTER'S THESIS ACADEMIC ADVISOR: PhD TRAN ANH TUAN Ho Chi Minh City – 2020 i DECLARATION I hereby declare that this master thesis entitled “FACTORS IMPACTING ON ABNORMAL INVESTMENT OF LISTED FIRMS IN HO CHI MINH STOCK EXCHANGE” is the result of my own original research work under the guidance of Mr Tuan Anh Tran, my thesis advisor This thesis has never been submitted for a master's degree at any other universities This thesis is the author's own research and the results of the research are trustworthy The thesis does not consist of any previously published content or content made by others except for citations which are fully cited in the thesis Ho Chi Minh City, September , 2020 KIEU CONG BAO TRAN ii ACKNOWLEDGEMENTS Apart from the efforts of myself, the completion of this thesis depends largely on the help, encouragement and guidelines of many others I shall take this opportunity to show my gratitude to the people who have been instrumental in this accomplishment of the thesis I would like to express my special appreciation and thanks to my academic advisor Mr Tuan Anh Tran for his tremendous support and help as well as for his patience, enthusiasm, and immense knowledge Without his guidance, this project would not have materialized Furthermore, I would like to give my sincere thanks to Banking University of Ho Chi Minh City for giving me the opportunity to this research and to all the lecturers there who have imparted the useful knowledge and practical experiences over the last one year For the greatest thing, a special thanks to my dearest family is indispensable Words cannot express how grateful I am to my mother, my father, my sister and brother for all of the sacrifices that you have made on my behalf Your prayers for me were what sustained me thus far I could not have done it without you Also, I would like to thank all of my friends who supported me in writing and incented me to strive towards my goal Thank you KIEU CONG BAO TRAN iii ABSTRACT Title: FACTORS IMPACTING ON ABNORMAL INVESTMENT OF LISTED FIRMS IN HO CHI MINH STOCK EXCHANGE Abstract: The purpose of the thesis is to study factors impacting on abnormal investment in Vietnam, namely that factors are free cash flow and dividends To achieve the main purpose, the study aims to first classify abnormal investment into two categories, over- and under-investment, by using an accounting-based framework developed by Richardson (2006) and Guariglia and Yang (2016) Secondly, by defining overinvestment as investments in negative NPV projects exceeds firm needs and underinvestment as the act of passing positive NPV investments essential for firm growth, the study relates these two categories of abnormal investment to firm’s free cash flow corresponding to financial constraints and agency problems theory Finally, the study examines the influence of dividends on overinvestment The studying data consists of 306 non-financial listed firms in Ho Chi Minh Stock Exchange with 3,672 firm-year-observations over the period 2008–2019 The study employs a proper research process with system GMM and FEM with clustered standard errors as final methods for conclusions The findings documented strong evidence of investment inefficiency exists among Vietnamese listed firms, which can be explained by financing constraints and agency problems Specifically, the results showed that firms with free cash flow below (above) their optimal level tend to under-(over-) invest as a consequence of financial constraints (agency costs) Moreover, the results indicated that dividends could increase investment efficiency Taking these findings into account, both corporate governance practices and financial market need taking actions to improve investment efficiency in Vietnamese firms Also, the findings support existing literature and become a foundation for further researches in investment inefficiency Keywords: Abnormal investment, free cash flow, dividend, financial constraints, agency cost iv ABSTRACT (VIETNAMESE) Tiêu đề: CÁC NHÂN TỐ TÁC ĐỘNG ĐẾN ĐẦU TƯ BẤT THƯỜNG CỦA NHỮNG CÔNG TY NIÊM YẾT TRÊN SỞ GIAO DỊCH CHỨNG KHỐN THÀNH PHỐ HỒ CHÍ MINH Tóm tắt: Mục tiêu luận văn nghiên cứu nhân tố tác động đến đầu tư bất thường doanh nghiệp Việt Nam, cụ thể nhân tố dòng tiền tự cổ tức Để đạt mục tiêu, nghiên cứu phân đầu tư bất thường thành hai loại: đầu tư mức, cách sử dụng mơ hình phân tích Richardson (2006) Guariglia and Yang (2016) Thứ hai, việc định nghĩa đầu tư mức khoản đầu tư kể vào dự án có NPV âm vượt nhu cầu công ty đầu tư mức hành động phải cắt giảm dự án có NPV dương cần thiết cho tăng trưởng công ty, nghiên cứu liên kết hai loại đầu tư bất thường với dòng tiền tự công ty tương ứng theo lý thuyết hạn chế tài chi phí đại diện Cuối cùng, nghiên cứu xem xét ảnh hưởng cổ tức việc đầu tư mức Dữ liệu nghiên cứu bao gồm 306 doanh nghiệp phi tài niêm yết Sở Giao dịch Chứng khốn Thành phố Hồ Chí Minh với 3.672 quan sát giai đoạn 2008-2019 Nghiên cứu thực theo trình tự phù hợp với ước lượng system GMM mơ hình FEM with clustered standard errors làm phương pháp cuối để đưa kết luận Kết nghiên cứu cho thấy chứng cụ thể hiệu đầu tư tồn doanh nghiệp niêm yết Việt Nam, điều giải thích hạn chế tài hay việc gặp vấn đề chi phí đại diện Cụ thể, cơng ty có dịng tiền tự (trên) mức tối ưu có xu hướng đầu tư (quá) mức bị hạn chế tài (gặp vấn đề chi phí đại diện) Ngồi ra, kết cịn cho thấy cổ tức làm tăng hiệu đầu tư Khi xem xét kết này, cần có phương án cho việc áp dụng thực quy định quản trị cơng ty thị trường tài để nâng cao hiệu đầu tư doanh nghiệp Việt Nam Hơn nữa, kết hỗ trợ cho lý thuyết hữu trở thành tảng cho nghiên cứu sau vấn đề hiệu đầu tư Từ khóa: Đầu tư bất thường, dịng tiền tự do, cổ tức, hạn chế tài chính, chi phí đại diện v LIST OF ABBREVIATIONS Abbreviations Definition 3SLS Three-Stage Least Squares Regression FEM Fixed-effects model HOSE Ho Chi Minh City Stock Exchange NPV Net Present Value OLS regression Ordinary Least Squares Regression REM Random-effects model ROA Return on Assets System-GMM System-Generalized Method of Moments vi TABLE OF CONTENTS DECLARATION i ACKNOWLEDGEMENTS ii ABSTRACT iii ABSTRACT (VIETNAMESE) iv LIST OF ABBREVIATIONS v TABLE OF CONTENTS vi LIST OF TABLES ix LIST OF EQUATIONS AND FIGURES x CHAPTER 1: INTRODUCTION 1.1 Introduction and Background 1.2 Research gap identification and new contributions 1.3 Research objectives 1.4 Research questions 1.5 The scope of the study 1.6 Research data and Methodology 1.6.1 Research data 1.6.2 Methodology 1.7 Research structure CHAPTER 2: LITERATURE REVIEW 2.1 Theoretical Framework 2.1.1 Pecking - order theory 2.1.2 Financial constraints theory 10 2.1.3 Agency problems theory – Free cash flow theory 11 vii 2.2 Free cash flow 12 2.3 Dividends 13 2.4 Abnormal investment and measurement framework 13 2.4.1 Financial constraints with free cash flow-underinvestment relationship 16 2.4.2 Agency problems with free cash flow-overinvestment relationship 18 2.4.3 The relationship between dividends and overinvestment 20 2.5 The previous researches 21 2.6 Hypotheses for models 26 2.6.1 Hypotheses for research question 26 2.6.2 Hypothesis for research question 27 CHAPTER 3: RESEARCH METHODOLOGIES 29 3.1 Data, Sample and Variables 29 3.2 Research Models 32 3.2.1 Expectation model for firm investment expenditure decision level 32 3.2.2 The relationship between free cash flow and abnormal investment level 36 3.2.3 The relationship between dividends and overinvestment 37 3.3 Research methods 40 3.3.1 Descriptive statistics 40 3.3.2 Estimation methods 41 3.3.3 Research process 43 CHAPTER 4: EMPIRICAL RESULTS 45 4.1 Descriptive statistics 45 4.2 Correlation matrix 49 4.3 Regression analysis by REM and FEM 50 4.3.1 Expectation model for firm investment expenditure decision level 50 viii 4.3.2 The relationship between abnormal investment level and free cash flow 52 4.3.3 The relationship between cash dividend payout and overinvestment 53 4.4 Test of consistent and unbiased 55 4.5 FEM with clustered standard errors, System-GMM results and discussion 57 4.5.1 System-GMM test for model 57 4.5.2 FEM with clustered standard errors for model 2.1 and 2.2 60 4.5.3 System-GMM test for model 61 4.6 Empirical evidence of over–and under–investing Vietnamese firms 63 CHAPTER 5: CONCLUSIONS AND POLICY IMPLICATIONS 65 5.1 Findings 65 5.2 Implications 66 5.2.1 Implications to companies 66 5.2.2 Implications to policy makers 69 5.3 Limitations 70 CONCLUSION 71 REFERENCES i APPENDIX xv xxviii Model xtset firm_ind year panel variable: time variable: delta: firm_ind (unbalanced) year, 2008 to 2019, but with gaps unit xtreg iu dpr tobinq leverage fcfo fcfu age size ms, re Random-effects GLS regression Group variable: firm_ind Number of obs Number of groups R-sq: Obs per group: within = 0.0295 between = 0.1150 overall = 0.0464 corr(u_i, X) = = 1,414 297 = avg = max = 4.8 12 = = 56.82 0.0000 Wald chi2(8) Prob > chi2 = (assumed) iu Coef Std Err z dpr tobinq leverage fcfo fcfu age size ms _cons -.0000317 01453 0356984 1113236 -.0008911 -.0016959 0004659 0405223 0310795 0001319 004448 0171604 0211951 0023822 0005894 0013379 0216394 0347023 sigma_u sigma_e rho 07190547 07056261 50942483 (fraction of variance due to u_i) -0.24 3.27 2.08 5.25 -0.37 -2.88 0.35 1.87 0.90 P>|z| 0.810 0.001 0.038 0.000 0.708 0.004 0.728 0.061 0.370 [95% Conf Interval] -.0002903 005812 0020645 0697821 -.0055602 -.0028511 -.0021563 -.0018902 -.0369359 0002269 023248 0693322 1528652 003778 -.0005408 0030881 0829348 0990948 Appendix 16 Fixed-effects model (FEM) regression for models Model xtset firm_ind year panel variable: time variable: delta: xtreg firm_ind (strongly balanced) year, 2008 to 2019 unit inew linew lcash ltobinq lsize age lroa lleverage, fe Fixed-effects (within) regression Group variable: firm_ind Number of obs Number of groups R-sq: Obs per group: within = 0.0606 between = 0.0452 overall = 0.0544 corr(u_i, Xb) Coef linew 1796517 3,672 306 = avg = max = 12 12.0 12 = = 30.94 0.0000 F(7,3359) Prob > F = -0.1124 inew = = Std Err .0184465 t 9.74 P>|t| [95% Conf Interval] 0.000 1434842 2158192 between = 0.0452 overall = 0.0544 corr(u_i, Xb) = -0.1124 12.0 12 = = 30.94 0.0000 F(7,3359) Prob > F xxix inew Coef linew lcash ltobinq lsize age lroa lleverage _cons 1796517 0442633 -.0044984 0009948 -.003716 0930942 -.0610563 0220647 0184465 0227353 003626 0003507 0006082 0286596 014436 0030808 sigma_u sigma_e rho 03567283 08618106 14627481 (fraction of variance due to u_i) Std Err avg = max = t 9.74 1.95 -1.24 2.84 -6.11 3.25 -4.23 7.16 P>|t| 0.000 0.052 0.215 0.005 0.000 0.001 0.000 0.000 [95% Conf Interval] 1434842 -.0003131 -.0116078 0003073 -.0049085 0369022 -.0893605 0160241 F test that all u_i=0: F(305, 3359) = 1.53 2158192 0888396 002611 0016824 -.0025235 1492863 -.032752 0281052 Prob > F = 0.0000 Model 2.1 gen FCFU = fcf*fc gen FCFO = fcf*ag xtset firm_code year panel variable: time variable: delta: firm_code (unbalanced) year, 2008 to 2019, but with gaps unit xtreg iu FCFO FCFU, fe Fixed-effects (within) regression Group variable: firm_code Number of obs Number of groups R-sq: Obs per group: within = 0.5949 between = 0.3942 overall = 0.5646 corr(u_i, Xb) = = 2,258 303 = avg = max = 7.5 12 = = 1434.22 0.0000 F(2,1953) Prob > F = -0.0087 iu Coef Std Err t FCFO FCFU _cons -.0899538 6636196 -.0079641 0154816 0124377 0012414 sigma_u sigma_e rho 02995689 0464137 29407569 (fraction of variance due to u_i) -5.81 53.36 -6.42 F test that all u_i=0: F(302, 1953) = 2.30 P>|t| 0.000 0.000 0.000 [95% Conf Interval] -.120316 6392271 -.0103987 -.0595917 6880121 -.0055295 Prob > F = 0.0000 xxx Model 2.2 xtset firm_code year panel variable: time variable: delta: firm_code (unbalanced) year, 2008 to 2019, but with gaps unit xtreg iu FCFU FCFO ,fe Fixed-effects (within) regression Group variable: firm_code Number of obs Number of groups R-sq: Obs per group: within = 0.0156 between = 0.1101 overall = 0.0224 corr(u_i, Xb) = = 1,414 297 = avg = max = 4.8 12 = = 8.86 0.0002 F(2,1115) Prob > F = 0.0491 iu Coef FCFU FCFO _cons -.0013595 0937151 0533756 0024552 0223469 0024482 sigma_u sigma_e rho 08654998 07095777 59803255 (fraction of variance due to u_i) Std Err t -0.55 4.19 21.80 P>|t| 0.580 0.000 0.000 [95% Conf Interval] -.0061768 0498683 0485721 F test that all u_i=0: F(296, 1115) = 3.07 0034578 1375618 0581791 Prob > F = 0.0000 Model xtset firm_ind year panel variable: time variable: delta: firm_ind (unbalanced) year, 2008 to 2019, but with gaps unit xtreg iu dpr tobinq leverage fcfo fcfu age size ms, fe Fixed-effects (within) regression Group variable: firm_ind Number of obs Number of groups R-sq: Obs per group: within = 0.0318 between = 0.0631 overall = 0.0279 corr(u_i, Xb) Coef dpr tobinq leverage fcfo fcfu age -.0000204 0091131 058738 09424 -.0012386 -.0016351 1,414 297 = avg = max = 4.8 12 = = 4.56 0.0000 F(8,1109) Prob > F = -0.0313 iu = = Std Err .0001355 0053127 0224941 0225418 0024422 0006457 t -0.15 1.72 2.61 4.18 -0.51 -2.53 P>|t| 0.880 0.087 0.009 0.000 0.612 0.011 [95% Conf Interval] -.0002864 -.0013109 0146022 0500108 -.0060305 -.0029021 0002455 0195372 1028738 1384693 0035533 -.0003681 between = 0.0631 overall = 0.0279 corr(u_i, Xb) = -0.0313 4.8 12 = = 4.56 0.0000 F(8,1109) Prob > F xxxi iu Coef dpr tobinq leverage fcfo fcfu age size ms _cons -.0000204 0091131 058738 09424 -.0012386 -.0016351 0003207 0265584 0182119 0001355 0053127 0224941 0225418 0024422 0006457 0014722 0238477 0369837 sigma_u sigma_e rho 08627923 07056261 59921059 (fraction of variance due to u_i) Std Err avg = max = t -0.15 1.72 2.61 4.18 -0.51 -2.53 0.22 1.11 0.49 P>|t| 0.880 0.087 0.009 0.000 0.612 0.011 0.828 0.266 0.623 [95% Conf Interval] -.0002864 -.0013109 0146022 0500108 -.0060305 -.0029021 -.0025679 -.0202334 -.054354 F test that all u_i=0: F(296, 1109) = 2.96 0002455 0195372 1028738 1384693 0035533 -.0003681 0032094 0733501 0907778 Prob > F = 0.0000 Appendix 17 Hausman test for models Model hausman fe re, sigmamore Coefficients (b) (B) fe re linew lcash ltobinq lsize lroa age lleverage 1796517 0442633 -.0044984 0009948 0930942 -.003716 -.0610563 3129872 0228286 -.0008667 0002847 0474752 -.0012181 -.0170998 (b-B) Difference -.1333355 0214346 -.0036317 0007101 045619 -.0024979 -.0439565 sqrt(diag(V_b-V_B)) S.E .0072279 015222 0022047 0002302 0152192 0003967 0111105 b = consistent under Ho and Ha; obtained from xtreg B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic chi2(7) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 401.44 Prob>chi2 = 0.0000 Model 2.1 hausman fe re, sigmamore Coefficients (b) (B) fe re FCFU FCFO 6636196 -.0899538 6624191 -.1038316 (b-B) Difference 0012005 0138777 sqrt(diag(V_b-V_B)) S.E .0034242 0050696 b = consistent under Ho and Ha; obtained from xtreg B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic (b) fe FCFU FCFO 6636196 -.0899538 (B) re 6624191 xxxii -.1038316 (b-B) Difference sqrt(diag(V_b-V_B)) S.E .0012005 0138777 0034242 0050696 b = consistent under Ho and Ha; obtained from xtreg B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic chi2(2) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 8.25 Prob>chi2 = 0.0161 Model 2.2 hausman fe re, sigmamore Coefficients (b) (B) fe re FCFU FCFO -.0013595 0937151 -.001059 1154429 (b-B) Difference sqrt(diag(V_b-V_B)) S.E -.0003005 -.0217278 0003584 0070351 b = consistent under Ho and Ha; obtained from xtreg B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic chi2(2) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 10.47 Prob>chi2 = 0.0053 Model hausman fe re, sigmamore Coefficients (b) (B) fe re dpr tobinq leverage FCFO FCFU age size ms -.0000204 0091131 058738 09424 -.0012386 -.0016351 0003207 0265584 -.0000317 01453 0356984 1113236 -.0008911 -.0016959 0004659 0405223 (b-B) Difference 0000113 -.0054168 0230397 -.0170836 -.0003475 0000608 -.0001452 -.0139639 sqrt(diag(V_b-V_B)) S.E .0000215 0027707 014065 0067203 000358 0002415 0005648 0092235 b = consistent under Ho and Ha; obtained from xtreg B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 16.47 Prob>chi2 = 0.0362 xxxiii Appendix 18 Heteroskedasticity test for models Model xttest3 Modified Wald test for groupwise heteroskedasticity in fixed effect regression model H0: sigma(i)^2 = sigma^2 for all i chi2 (306) = Prob>chi2 = 1.3e+07 0.0000 Model 2.1 xttest3 Modified Wald test for groupwise heteroskedasticity in fixed effect regression model H0: sigma(i)^2 = sigma^2 for all i chi2 (303) = Prob>chi2 = 1.8e+31 0.0000 Model 2.2 xttest3 Modified Wald test for groupwise heteroskedasticity in fixed effect regression model H0: sigma(i)^2 = sigma^2 for all i chi2 (297) = Prob>chi2 = Model xttest3 1.3e+35 0.0000 Modified Wald test for groupwise heteroskedasticity in fixed effect regression model H0: sigma(i)^2 = sigma^2 for all i chi2 (297) = Prob>chi2 = 1.6e+36 0.0000 Appendix 19 Autocorrelation test for models Model xtserial inew linew lcash ltobinq lsize age lroa lleverage Wooldridge test for autocorrelation in panel data H0: no first-order autocorrelation F( 1, 305) = 118.115 Prob > F = 0.0000 xxxiv Model 2.1 xtserial iu FCFO FCFU Wooldridge test for autocorrelation in panel data H0: no first-order autocorrelation F( 1, 234) = 1.600 Prob > F = 0.2072 Model 2.2 xtserial iu FCFO FCFU Wooldridge test for autocorrelation in panel data H0: no first-order autocorrelation F( 1, 143) = 2.498 Prob > F = 0.1162 Model xtserial iu dpr tobinq leverage FCFO FCFU age size ms Wooldridge test for autocorrelation in panel data H0: no first-order autocorrelation F( 1, 143) = 1.934 Prob > F = 0.1665 Appendix 20 Endogeneity test for models Model xtivreg2 inew lroa linew lsize ltobinq age y1 y2 y3 y4 y5 y6 y7 y8 y9 y10 y11 y12 ( > y10 cluster( abc) endog( lleverage y5 y6 y7 y8 y9 y11 y12 ( lleverage lcash = lcash) dleverage dcash droa startyear ), fe xtivreg2 inew lroa linew lsize ltobinq age y1 y2 y3 y4 y5 y6 y7 y8 y9 y10 y11 y12 ( lleve Warning - collinearities detected > cluster( abc) endog( lleverage lcash) Vars dropped: y12 Warning - collinearities detected Vars dropped: y12 FIXED EFFECTS ESTIMATION FIXED EFFECTS ESTIMATION Number of groups = Number12of groups = group: = avg = max = 306 12.0 Warning 12 - collinearities detected Warning - collinearities detected Vars dropped: y12 Vars dropped: Obs per group: = Obs per group: = avg = 12 avg = max = 12.0 max = 12 12.0 12 12 y12 IV (2SLS) estimation IV (2SLS) estimation Estimates efficient for homoskedasticity only Estimates efficient for homoskedasticity only Statistics robust to heteroskedasticity and clustering on abc Statistics robust to heteroskedasticity and clustering on abc n abc mber of obs 18, 305) ob > F ntered R2 centered R2 ot MSE 306 Number of of clusters clusters(abc) (abc) Number = = = = = = = = 306306 3672 5.88 Total (centered) (centered)SSSS 26.55645026 Total = = 26.55645026 0.0000 Total (uncentered) (uncentered)SSSS = =26.55645026 26.55645026 Total -0.1263 Residual = = 29.91092674 Residual SS SS 29.91092674 -0.1263 09427 Robust Robust Number of = obs =3672 3672 Number of obs F( 18, 305) =5.88 5.88 F( 18, 305) = Prob > F = 0.0000 Prob > F = 0.0000 Centered R2 -0.1263 = -0.1263 Centered R2 = Uncentered = -0.1263 Uncentered R2 = R2 -0.1263 Root Root MSE MSE = 09427 = 09427 Estimates efficient for homoskedasticity only Statistics robust to heteroskedasticity and clustering on abc Number of clusters (abc) = 306 Number of obs F( 18, 305) Prob > F Centered R2 Uncentered R2 Root MSE xxxv Total (centered) SS Total (uncentered) SS Residual SS = = = inew Coef lleverage lcash lroa linew lsize ltobinq age y1 y2 y3 y4 y5 y6 y7 y8 y9 y10 y11 y12 -.4347834 1203848 -.1454159 2241894 0079852 -.0109007 -.0062596 -.0172413 -.0383802 -.0250488 -.0219975 -.0229844 -.0166015 -.0194985 -.0029813 -.0026189 -.0269311 -.004704 26.55645026 26.55645026 29.91092674 Robust Std Err .0847125 0607972 068669 044323 0016449 0055007 0019444 0169494 0168446 0158398 0152946 0139381 0131938 0117763 010808 00951 0090789 0095595 (omitted) z -5.13 1.98 -2.12 5.06 4.85 -1.98 -3.22 -1.02 -2.28 -1.58 -1.44 -1.65 -1.26 -1.66 -0.28 -0.28 -2.97 -0.49 P>|z| 0.000 0.048 0.034 0.000 0.000 0.048 0.001 0.309 0.023 0.114 0.150 0.099 0.208 0.098 0.783 0.783 0.003 0.623 = = = = = = 3672 5.88 0.0000 -0.1263 -0.1263 09427 [95% Conf Interval] -.6008168 0012244 -.2800047 1373178 0047612 -.0216819 -.0100705 -.0504616 -.071395 -.0560942 -.0519743 -.0503026 -.0424609 -.0425797 -.0241646 -.0212581 -.0447254 -.0234402 Underidentification test (Kleibergen-Paap rk LM statistic): Chi-sq(3) P-val = -.2687499 2395452 -.010827 3110609 0112091 -.0001195 -.0024487 0159789 -.0053654 0059965 0079793 0043339 0092579 0035827 018202 0160204 -.0091368 0140322 126.797 0.0000 Weak identification test (Cragg-Donald Wald F statistic): 65.245 (Kleibergen-Paap rk Wald F statistic): 66.488 Stock-Yogo weak ID test critical values: 5% maximal IV relative bias 11.04 10% maximal IV relative bias 7.56 20% maximal IV relative bias 5.57 30% maximal IV relative bias 4.73 10% maximal IV size 16.87 15% maximal IV size 9.93 20% maximal IV size 7.54 25% maximal IV size 6.28 Source: Stock-Yogo (2005) Reproduced by permission NB: Critical values are for Cragg-Donald F statistic and i.i.d errors Hansen J statistic (overidentification test of all instruments): Chi-sq(2) P-val = -endog- option: Endogeneity test of endogenous regressors: Chi-sq(2) P-val = Regressors tested: lleverage lcash 1.592 0.4512 22.620 0.0000 Instrumented: lleverage lcash Included instruments: lroa linew lsize ltobinq age y1 y2 y3 y4 y5 y6 y7 y8 y9 y10 y11 Excluded instruments: dleverage dcash droa startyear Dropped collinear: y12 xxxvi Model 2.1 xtivreg2 iu ( FCFU FCFO = L.cash L.leverage L.roa ), fe cluster( firm_code ) endog( FCFO FCFU) Warning - singleton groups detected 26 observation(s) not used FIXED EFFECTS ESTIMATION Number of groups = 253 Obs per group: = avg = max = 5.7 11 IV (2SLS) estimation Estimates efficient for homoskedasticity only Statistics robust to heteroskedasticity and clustering on firm_code Number of clusters (firm_code) = Total (centered) SS Total (uncentered) SS Residual SS = = = iu Coef FCFU FCFO 9389955 0895644 253 Number of obs F( 2, 252) Prob > F Centered R2 Uncentered R2 Root MSE 4.379943178 4.379943178 2.763990155 Robust Std Err .178239 1802288 z 5.27 0.50 P>|z| 0.000 0.619 = = = = = = 1448 19.22 0.0000 0.3689 0.3689 04809 [95% Conf Interval] 5896535 -.2636775 1.288338 4428063 Underidentification test (Kleibergen-Paap rk LM statistic): Chi-sq(2) P-val = 10.681 0.0048 Weak identification test (Cragg-Donald Wald F statistic): (Kleibergen-Paap rk Wald F statistic): Stock-Yogo weak ID test critical values: 10% maximal IV size 15% maximal IV size 20% maximal IV size 25% maximal IV size Source: Stock-Yogo (2005) Reproduced by permission NB: Critical values are for Cragg-Donald F statistic and i.i.d errors Hansen J statistic (overidentification test of all instruments): Chi-sq(1) P-val = -endog- option: Endogeneity test of endogenous regressors: Chi-sq(2) P-val = Regressors tested: FCFO FCFU 6.368 7.833 13.43 8.18 6.40 5.45 0.215 0.6428 2.083 0.3530 Instrumented: FCFU FCFO Excluded instruments: L.cash L.leverage L.roa Model 2.2 xtivreg2 iu ( FCFU FCFO = ms dpr cfo roa L.size ), fe cluster( firm_code ) endog( FCFO FCFU) Warning - singleton groups detected 49 observation(s) not used FIXED EFFECTS ESTIMATION Number of groups = 161 Obs per group: = avg = max = 4.2 11 IV (2SLS) estimation Estimates efficient for homoskedasticity only Statistics robust to heteroskedasticity and clustering on firm_code Number of clusters (firm_code) = Total (centered) SS Total (uncentered) SS Residual SS = = = 161 1.310408326 1.310408326 1.382116738 Number of obs F( 2, 160) Prob > F Centered R2 Uncentered R2 Root MSE = = = = = = 678 0.88 0.4147 -0.0547 -0.0547 0517 IV (2SLS) estimation Estimates efficient for homoskedasticity only Statistics robust to heteroskedasticity and clustering on firm_code xxxvii Number of clusters (firm_code) = Total (centered) SS Total (uncentered) SS Residual SS = = = iu Coef FCFU FCFO -.0022412 1717854 161 Number of obs F( 2, 160) Prob > F Centered R2 Uncentered R2 Root MSE 1.310408326 1.310408326 1.382116738 Robust Std Err .0017139 1809742 z P>|z| -1.31 0.95 0.191 0.343 = = = = = = 678 0.88 0.4147 -0.0547 -0.0547 0517 [95% Conf Interval] -.0056003 -.1829175 0011179 5264882 Underidentification test (Kleibergen-Paap rk LM statistic): Chi-sq(4) P-val = 18.933 0.0008 Weak identification test (Cragg-Donald Wald F statistic): (Kleibergen-Paap rk Wald F statistic): Stock-Yogo weak ID test critical values: 5% maximal IV relative bias 10% maximal IV relative bias 20% maximal IV relative bias 30% maximal IV relative bias 10% maximal IV size 15% maximal IV size 20% maximal IV size 25% maximal IV size Source: Stock-Yogo (2005) Reproduced by permission NB: Critical values are for Cragg-Donald F statistic and i.i.d errors Hansen J statistic (overidentification test of all instruments): Chi-sq(3) P-val = -endog- option: Endogeneity test of endogenous regressors: Chi-sq(2) P-val = Regressors tested: FCFO FCFU 2.537 4.160 13.97 8.78 5.91 4.79 19.45 11.22 8.38 6.89 3.759 0.2887 4.155 0.1252 Instrumented: FCFU FCFO Excluded instruments: ms dpr cfo roa L.size Model xtivreg2 iu dpr tobinq fcfu fcfo age size ms ( leverage = d.cash d.leverage ), fe cluster( firm_code Warning - singleton groups detected 49 observation(s) not used cash d.leverage ), fe cluster( firm_code) endog( leverage ) FIXED EFFECTS d xtivreg2 iuESTIMATION dpr tobinq fcfu fcfo age size ms ( leverage = avg = max = Warning - singleton groups detected Number of groups = 4.2 11 Obs per group: = avg = max = FIXED EFFECTS ESTIMATION Number of groups = IV (2SLS) estimation 161 Number of clusters (firm_code) = 678 5.52 0.0000 -0.1266 -0.1266 05344 4.2 11 Obs per group: = avg = max = 4.2 11 161 Number of obs = 678 F( 8, 160) = 5.52 Estimates efficient for homoskedasticity only Prob > F = 0.0000 Statistics robust to heteroskedasticity and clustering on firm_code Total (centered) SS = 1.310408336 Centered R2 = -0.1266 Total (uncentered) SS = 1.310408336 Uncentered R2 = -0.1266 Number of clusters (firm_code) = 161 Number of obs = Residual SS = 1.476287067 Root MSE = 05344 Total (centered) SS = Robust 1.310408336 Total (uncentered) SS = Std 1.310408336 iu Coef Err z Residual SS = 1.476287067 Conf Interval] 345 161 Estimates efficient for homoskedasticity only Statistics robust to heteroskedasticity and clustering on firm_code IV (2SLS) estimation code obs = 160) = = R2 = d R2 = = = d.cash d.leverage ), fe cluster( 49 observation(s) not used .4233658 leverage dpr tobinq 2575501 9.94e-06 0075249 0846014 0000658 0123427 Robust 3.04 0.15 0.61 P>|z| 0.002 0.880 0.542 678 F( 8, 160) = 5.52 Prob > F = 0.0000 Centered R2 = -0.1266 Uncentered R2 = -0.1266 [95% Conf Interval] Root MSE = 05344 0917345 -.000119 -.0166663 4233658 0001389 0317162 Estimates efficient for homoskedasticity only Statistics robust to heteroskedasticity and clustering on firm_code Number of clusters (firm_code) = 161 Number of obs F( 8, 160) Prob > F Centered R2 Uncentered R2 Root MSE xxxviii Total (centered) SS Total (uncentered) SS Residual SS = = = iu Coef leverage dpr tobinq fcfu fcfo age size ms 2575501 9.94e-06 0075249 -.0005993 0423634 0004921 -.0065029 0060527 1.310408336 1.310408336 1.476287067 Robust Std Err .0846014 0000658 0123427 0001906 0225732 0011243 0025442 0226375 z 3.04 0.15 0.61 -3.14 1.88 0.44 -2.56 0.27 P>|z| 0.002 0.880 0.542 0.002 0.061 0.662 0.011 0.789 = = = = = = 678 5.52 0.0000 -0.1266 -0.1266 05344 [95% Conf Interval] 0917345 -.000119 -.0166663 -.0009729 -.0018793 -.0017116 -.0114895 -.038316 Underidentification test (Kleibergen-Paap rk LM statistic): Chi-sq(2) P-val = 4233658 0001389 0317162 -.0002257 0866061 0026957 -.0015164 0504213 30.820 0.0000 Weak identification test (Cragg-Donald Wald F statistic): 103.327 (Kleibergen-Paap rk Wald F statistic): 125.241 Stock-Yogo weak ID test critical values: 10% maximal IV size 19.93 15% maximal IV size 11.59 20% maximal IV size 8.75 25% maximal IV size 7.25 Source: Stock-Yogo (2005) Reproduced by permission NB: Critical values are for Cragg-Donald F statistic and i.i.d errors Hansen J statistic (overidentification test of all instruments): Chi-sq(1) P-val = -endog- option: Endogeneity test of endogenous regressors: Chi-sq(1) P-val = Regressors tested: leverage 0.565 0.4524 9.718 0.0018 Instrumented: leverage Included instruments: dpr tobinq fcfu fcfo age size ms Excluded instruments: D.cash D.leverage Appendix 21 System-GMM method for models Model xtset firm_ind year xtset firm_ind year xtset firm_ind year panel variable: firm_ind (strongly balanced) panel variable: firm_ind (strongly balanced) panel variable: firm_ind (strong m_ind year time variable: year, 2008 to 2019 xtset2008 firm_ind year time variable: year, to 2019 ed) time variable: year, 2008 to 20 l variable: firm_ind (strongly balanced) delta: unit delta: unit panel variable: firm_ind (strongly balanced) delta: unit e variable: year, 2008 to 2019 time variable: year, 2008 to 2019 delta: unit xtabond2 inew linew lcash ltobinq lsize lroa age lleverage ap* y*,gmm( linew, lag(2 3) collapse) gmm( L3.lsi delta: unit xtabond2 inew linew lcash ltobinq lsize lroa 1age lleverage ap* y*,gmm( linew,inew lag(2 3) collapse) gmm(lsiz L3 xtabond2 linew lcash ltobinq > obinq , lag (1 3) collapse) gmm(L4.lroa, collapse) gmm( L.leverage L.lcash, collapse)iv( lage l2dlinew dlca > L.lcash, obinq , lag (1 (2 3) 6)collapse) collapse) gmm(L4.lro e lleverage ap* y*,gmm( linew, lag(2 3) collapse) collapse) ap* gmm( L3.lsize, lag lag(2 (2 6)collapse) gmm(L.lt > obinq , lag (1 age 3) gmm(L4.lroa, collapse) gmm( L.leverage collapse)iv( lage l2dlinew inew linew lcash ltobinq lsize lroa lleverage y*,gmm( linew, 3) collapse) gmm( L3.lsize, lag gmm(L.lt > l2dtobinq ap* y*) nodiffsargan robust orthogonal xtabond2 inew linew lcash small ltobinq lsize lroa > agel2dtobinq lleverage ap* ap* y*) y*,gmm( linew, lag(2 nodiffsargan robust se)(1 gmm( L.leverage collapse)iv( lage l2dlinew dlcash l2dsize l2dlroa ldleverage > L.lcash, l2dtobinq ap* y*)speed nodiffsargan robust orthogonal ag 3) collapse) collapse) gmm( L.lcash, collapse)iv( lage l2dlinew dlcash l2dsize ldleverage Favoring space over ToL.leverage switch, click on small mata: mata set matafavor speed, perm.l2dlroa se) gmm( L3.lsize, gmm(L4.lroa, lag (2 6)collapse) gmm(L.lt > obinq , lag type (1 3)or collapse) gmm(L4.lroa, collapse) gmm( L.leverage L.lcash, collapse space over perm speed To switch, ty alap* small Favoring space speed matafavor speed, q y*) nodiffsargan robust orthogonal small To switch, type or click on mata: mata setFavoring dropped due over to collinearity e l2dlinew dlcash ap1 l2dsize l2dlroa ldleverage > l2dtobinq ap* y*) nodiffsargan robust orthogonal small due to collinearity ap1 dropped ck on mata: mataTo set matafavor speed, perm ap1 dropped due to collinearity ace over speed switch, type or click on mata: mata set matafavor speed, perm ap3 dropped due to collinearity Favoring space over speed To switch, type or click on mata: mata matafavor speed, ap3 dropped due to set collinearity ap3 dropped due to collinearity year dropped due due to to collinearity collinearity ap1 dropped due to collinearity year dropped due to collinearity yeardropped droppeddue duetotocollinearity collinearity due to collinearity y11 ap3 dropped due to collinearity y11 dropped due to collinearity y11 dropped due toestimated collinearity d due to collinearity Warning: Two-step covariance matrix of moments is singular year dropped due to collinearity Warning: Two-step estimated covariance m due to collinearity Warning: estimated covariance matrix of weighting moments is singular Using a Two-step generalized inverse to calculate robust matrix for Hansen test y11 dropped due to collinearity Using a generalized inverse to calcula momentsestimated is singular o-step covariance matrix of moments is singular Using a generalized inverse toTwo-step calculate robust weighting Hansen is test Warning: estimated covariance matrix matrix for of moments singular weighting matrix for Hansen test eneralized inverse to calculate robust estimation, weighting matrix for system Hanseninverse test to calculate robust weighting matrix for Hansen test Dynamic panel-data one-step GMM Using a generalized Dynamic panel-data estimation, one-step Dynamic panel-data estimation, one-step system GMM M el-data estimation, Group one-step system GMM variable: firm_ind Dynamic panel-data estimation, Number of obs = one-step system GMM 3672 Group variable: firm_ind Time : year Number of of groups = = 306 Groupvariable variable: firm_ind Number obs 3672 Time variable : year umber of obs = 3672 panel variable: time variable: delta: firm_ind (strongly balanced) year, 2008 to 2019 unit xtabond2 inew linew lcash ltobinq lsize lroa age lleverage ap* y*,gmm( linew, lag(2 3) collapse) gmm( L3.l > obinq , lag (1 3) collapse) gmm(L4.lroa, collapse) gmm( L.leverage L.lcash, collapse)iv( lage l2dlinew dl > l2dtobinq ap* y*) nodiffsargan robust orthogonal small Favoring space over speed To switch, type or click on mata: mata set matafavor speed, perm ap1 dropped due to collinearity ap3 dropped due to collinearity year dropped due to collinearity y11 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular Using a generalized inverse to calculate robust weighting matrix for Hansen test xxxix Dynamic panel-data estimation, one-step system GMM Group variable: firm_ind Time variable : year Number of instruments = 62 F(20, 305) = 6.03 Prob > F = 0.000 inew Coef linew lcash ltobinq lsize lroa age lleverage aprov ap2 y1 y2 y3 y4 y5 y6 y7 y8 y9 y10 y12 _cons 5769351 113861 024805 0017586 -.3197191 -.0016941 -.1014689 0024036 0024269 0151172 -.0084969 0022965 0033445 -.0013818 0019826 -.0063837 0097196 0012857 -.0238294 0033764 0065816 Number of obs Number of groups Obs per group: avg max Robust Std Err .1788398 0549445 0141494 0009624 213501 0006043 0477861 0018814 0070586 0154052 0165748 0144352 0134921 0132587 0127327 0120729 0108375 0118317 0165775 013154 0156168 t 3.23 2.07 1.75 1.83 -1.50 -2.80 -2.12 1.28 0.34 0.98 -0.51 0.16 0.25 -0.10 0.16 -0.53 0.90 0.11 -1.44 0.26 0.42 P>|t| 0.001 0.039 0.081 0.069 0.135 0.005 0.035 0.202 0.731 0.327 0.609 0.874 0.804 0.917 0.876 0.597 0.371 0.914 0.152 0.798 0.674 = = = = = 3672 306 12 12.00 12 [95% Conf Interval] 2250191 0057427 -.0030377 -.0001353 -.7398405 -.0028833 -.195501 -.0012987 -.0114628 -.0151968 -.0411124 -.0261087 -.0232049 -.0274719 -.0230725 -.0301404 -.0116061 -.0219964 -.0564501 -.0225077 -.0241487 9288512 2219793 0526478 0036524 1004024 -.0005049 -.0074368 0061058 0163166 0454312 0241185 0307017 0298938 0247083 0270377 017373 0310454 0245677 0087914 0292606 0373118 Instruments for orthogonal deviations equation Standard FOD.(lage l2dlinew dlcash l2dsize l2dlroa ldleverage l2dtobinq aprov ap1 ap2 ap3 year y1 y2 y3 y4 y5 y6 y7 y8 y9 y10 y11 y12) GMM-type (missing=0, separate instruments for each period unless collapsed) L(1/11).(L.leverage L.lcash) collapsed L(1/11).L4.lroa collapsed L(1/3).L.ltobinq collapsed L(2/6).L3.lsize collapsed L(2/3).linew collapsed Instruments for levels equation Standard lage l2dlinew dlcash l2dsize l2dlroa ldleverage l2dtobinq aprov ap1 ap2 ap3 year y1 y2 y3 y4 y5 y6 y7 y8 y9 y10 y11 y12 _cons GMM-type (missing=0, separate instruments for each period unless collapsed) D.(L.leverage L.lcash) collapsed D.L4.lroa collapsed D.L.ltobinq collapsed DL.L3.lsize collapsed DL.linew collapsed Arellano-Bond test for AR(1) in first differences: z = Arellano-Bond test for AR(2) in first differences: z = Sargan test of (Not robust, Hansen test of (Robust, but overid restrictions: chi2(41) = 108.83 but not weakened by many instruments.) overid restrictions: chi2(41) = 47.53 weakened by many instruments.) -3.09 1.21 Pr > z = Pr > z = 0.002 0.228 Prob > chi2 = 0.000 Prob > chi2 = 0.224 GMM-type (missing=0, separate instruments for each period unless collapsed) D.(L.leverage L.lcash) collapsed D.L4.lroa collapsed D.L.ltobinq collapsed DL.L3.lsize collapsed xl DL.linew collapsed Arellano-Bond test for AR(1) in first differences: z = Arellano-Bond test for AR(2) in first differences: z = Sargan test of (Not robust, Hansen test of (Robust, but overid restrictions: chi2(41) = 108.83 but not weakened by many instruments.) overid restrictions: chi2(41) = 47.53 weakened by many instruments.) -3.09 1.21 Pr > z = Pr > z = 0.002 0.228 Prob > chi2 = 0.000 Prob > chi2 = 0.224 Model xtset xtset firm_ind firm_ind year year xtset firm_ind year panel panel variable: variable: firm_ind firm_ind(unbalanced) (unbalanced) panel variable: firm_ind (unbalanced) time variable: year, 2008 to 2019, but with gaps variable: gapsvariable: year, 2008 to 2019, but with gaps time xtset time firm_ind year year, 2008 to 2019, but with m_ind (unbalanced) xtset firm_ind year delta: delta: 11 unit unit delta: unit panel variable: firm_ind (unbalanced) r, 2008 to 2019, but with gaps panel variable: firm_ind (unbalanced) timetime variable: year, 2008 to 2019, but with gaps year, fcfo 2008 fcfu to 2019, size but with gaps nit xtabond2 iu dpr variable: tobinq leverage ms ap* y*, gmm( L2.leverage, lag (4 0) collapse) gmm( tobinq,la xtabond2 iu dprdelta: tobinq leverage fcfo fcfuage size ms gmm( leverage L2.leverage, (4 age 0) collapse) gmm( tobin age xtabond2 iuap* dpry*, tobinq fcfolag fcfu size ms ap* y*, gmm unit delta: 12).fcfu, unit > 2.size, collapse) gmm( L(0 collapse) gmm( L.dpr, lag (0 2) collapse) gmm( L(1 2).fcfo , lag (4 6)collapse > 2.size, collapse) (2 gmm( L(0 2).fcfu,gmm( collapse) gmm( L.dpr, lag gmm( (0 2)L(0 collapse) gmm( L(1 2).fcfo ,L.dpr, lag (4 lag 6)coll > 2.size, collapse) 2).fcfu, collapse) gmm( (0 lag (4 0) collapse) gmm( tobinq,lag 1) collapse) L > size binq ldleverage l2dsize lddpr lage dms ms(4ap* y*, eq(level)) nodiffsargan robust orthogonal small L verage fcfo fcfu age ms ap* y*, gmm( L2.leverage, lag 0) collapse) gmm( tobinq,lag (2 1) collapse) gmm( > binq ldleverage l2dsize lddpr lage dms ms ap* y*, eq(level)) nodiffsargan robust orthogonal small > binq ldleverage l2dsize lddpr lage dms ms ap* y*, eq(level)) nodi xtabond2 iu dpr tobinq leverage fcfo fcfu age size ms ap* y*, gmm( L2.leverage, lag (4 0) collapse) gmm( tobinq, m( L(1 2).fcfo , lag (4 6)collapse) iv( dfcfu l2dfcfo l2dto xtabond2 iu dpr tobinq leverage fcfo fcfu age size ms ap* y*, gmm( L2.leverage, lag (4 0) collapse) gmm( Favoring space lag over (0 speed To switch,gmm( typeL(1 or 2).fcfo click on ,mata: mata set matafavor perm 2).fcfu, collapse) gmm( L.dpr, 2) collapse) lag (4 6)collapse) iv(speed, dfcfu l2dfcfo l2dto Favoring space collapse) over speed ToL(0 switch, type or clickgmm( on mata: mata set speed, perm ag (2 1) collapse) gmm( > L2.size, gmm( 2).fcfu, collapse) L.dpr, lag (0 2) matafavor collapse) gmm( L(1 ,onlag 6)collap Favoring space over speed To switch, type or2).fcfo click mata: mata orthogonal > y*, 2.size, collapse) gmm( L(0 2).fcfu, collapse) gmm( L.dpr, lag (0 2) collapse) gmm( L(1 2).fcfo ,(4 lag (4 s ap1 dropped due to nodiffsargan collinearity ddpr lage dmssmall ms ap* eq(level)) robust orthogonal small ap1 dropped due to collinearity > binq ldleverage l2dsize lddpr lage dms ms ap* y*, eq(level)) nodiffsargan robust orthogonal small e) iv( dfcfu l2dfcfo l2dto ap1 dropped due to collinearity ap3 dropped due mata to collinearity > binq ldleverage l2dsize lddpr lage dms perm ms ap* y*, eq(level)) nodiffsargan robust orthogonal small eed, perm type or click To switch, on mata: set matafavor speed, ap3 dropped to collinearity Favoringdue space speed To switch, type ordropped click ondue mata: mata set matafavor speed, perm ap3or to collinearity year dropped due to over collinearity Favoring space over speed To switch, type click on mata: mata set matafavor speed, perm rity year ap1 dropped due due to collinearity dropped to collinearity y10 due year dropped due to collinearity ap1 dropped dropped dueto tocollinearity collinearity rity ap3 dropped due to collinearity y10 dropped due to collinearity Warning: Two-step estimated covariance matrixy10 of dropped moments is duesingular to collinearity ap3 dropped due to collinearity arity year dropped due to collinearity Warning: estimated matrix of weighting moments is singular Using a Two-step generalized inverse covariance to calculate robust matrix for Hansen test Warning: Two-step estimated covariance matrix of moments is singula year dropped duedue to to collinearity y10 collinearity rity Using adropped generalized inverse to calculate robust weighting matrix for Hansen test Using a generalized inverse to calculate robust weighting matrix Warning: Two-step estimated covariance matrix of moments is singular y10moments droppedisdue to collinearity covariance matrix of singular Dynamic panel-data estimation, one-step system GMM Using a matrix generalized inverse to calculate robust weightingismatrix for Hansen test Warning: Two-step estimated covariance matrix of moments singular se to calculate robust weighting for Hansen test Dynamic panel-data estimation, one-step system GMM Dynamic panel-data estimation, one-step system GMM Using a generalized matrix for Group variable: firm_ind inverse to calculate robust Number weighting of obs = 1414Hansen test Dynamic panel-data estimation, one-step system GMM on, one-step system Time GMM variable Group variable: firm_ind Number obs 1414 : year Number of of groups = = 297 Group variable: firm_ind Number of obs Dynamic one-step system GMM Time variable : yearestimation, Number of groups Number ofpanel-data instruments = 73 Obs per group: = = 1297 Group variable: firm_ind Number of obs = 1414 Time variable : year Number of groups Number of obs = 1414 Number of instruments = 73 Obs per group: F(21, 296) = 3.14 avgmin = = 4.76 Time variable : year Number of groups = 297 Number of instruments = 73 Obs per group: Prob >Number F of of ==instruments 0.000 = = = = 12 F(21, 296) 3.14 avg 4.7611414 Number groups = Group variable: firm_ind Number of max obs = 73 297 Obs per group: F(21, 296) = 3.14 avg Prob > F = 0.000 max = 12 Obs per group: = Time F(21, variable Number of groups 296) : year = 3.14 avg = = 4.76 297 Prob > F = 0.000 max avg = =0.000 4.76 Prob F = = = 12 Number of >instruments 73 Robust Obs per group:max iu Coef Std Err t P>|t| [95% Conf Interval] Robust 12 F(21, 296) = max =3.14 avg = 4.76 Robust Coef Err t P>|t| [95% Conf = 0.000 Std Robust maxInterval] = 12 dpr -.0004851Coef .0002656 -1.83 t0.069 0000375 iu-.0010077 Std Err t P>|t| [95% Conf iu Std Err P>|t| [95% Coef Conf Interval] tobinq -.0187607 0143427 -1.31 0.192 -.0469872 0094658 dpr -.0004851 0002656 -1.83 0.069 -.0010077 0000375 t P>|t| [95% Conf Interval] Robust -1.74 leverage -.0912357 0523362 0.082 0117625 -.0004851 0002656 -1.31 -1.83 0.069 -.0010077 0000375 dpr-.1942339 -.0004851 0002656 -1.83 0.069 -.0010077 tobinq dpr -.0187607 0143427 0.192 -.0469872 0094658 iu Coef Std Err 2.14-1.31 t 0.0330.192 P>|t| 0222185 [95% Conf Interval] fcfo 2816652 131832 541112 tobinq -.0187607 0143427 -.0469872 0094658 tobinq -.0187607 0143427 -1.31 0.192 -.0469872 leverage-.0010077 -.0912357.0000375 0523362 -1.74 0.082 -.1942339 0117625 0002656 -1.83 0.069 fcfu -.0015553 0010448 0.138 0005009 leverage -.0912357 0523362-1.49-1.74 0.082-.0036115 -.1942339 0117625 leverage -.0912357 0523362 -1.74 0.082 -.1942339 fcfo 2816652 131832 2.14 0.033 0222185 541112 0143427 -1.31 0.192 age -.0469872 0094658 dpr -.0004851 0002656 -1.83 0.069-.0072715 -.0010077 0000375 -.0042433 0015387 0.006 -.0012152 fcfo 2816652 131832-2.76 2.14 0.033 0222185 541112 fcfo -.0036115 2816652 0005009 131832 2.14 0.033 0222185 fcfu-.1942339 -.0015553.0117625 0010448 -1.49 0.138 0523362 -1.74 0.082tobinq size fcfu 0087271 0049486 0.079 018466 -.0015553 0010448 1.76 -1.49 0.138 -.0036115 0005009 -.0187607 0143427 -1.31 0.192-.0010118 -.0469872 0094658 -.0042433 541112 0015387 -2.76 0.006 fcfu -.0072715 -.0015553 -.0012152 0010448 -1.49 0.138 -.0036115 131832 2.14 0.033 age 0222185 -.0042433 0015387 0.72 -2.76 0.006 -.0072715 -.0012152 ms age 0180052 0251519 0.475 0675044 leverage -.0912357 0523362 -1.74 0.082 -.031494 -.1942339 0117625 size-.0036115 0087271.0005009 0049486 1.76 0.079 age -.0010118 -.0042433 018466 0015387 -2.76 0.006 -.0072715 0010448 -1.49 0.138aprov size 0087271 0049486-0.012.14 1.76 0.079 -.0010118 018466 -.0000346 0036661 0.992 -.0072494 0071802 fcfo 2816652 131832 0.033 0222185 541112 ms-.0072715 0180052 0251519 0.72 0.475 -.031494 0675044 size 0087271 0049486 1.76 0.079 -.0010118 0015387 -2.76 0.006 ap2 -.0012152 ms 0243263 0180052 0251519 0.72 0.475 -.031494 0675044 0.067 0504113 fcfu -.0015553 0132545 0010448 1.84 -1.49 0.138-.0017586 -.0036115 0005009 0036661 -0.01 0.992 -.0072494 0071802 ms-.0086772 0180052 0251519 0.72 0.475 -.031494 aprov-.0000346 -.0000346 0036661 1.52 -0.01 0.992 -.0072494 0071802 0049486 1.76 0.079aprov 018466 y1-.0010118 0294788 0193881 0.129 0676347 age -.0042433 0015387 -2.76 0.067 0.006 -.0017586 -.0072715 0504113 -.0012152 -0.01 ap2 0243263 0132545 1.84 aprov -.0000346 0036661 0.992 -.0072494 ap2 0243263 0132545 1.84 0.067 -.0017586 0504113 y2 -.031494 -.025188 0675044 0159374 -1.58 0.115 -.056553 0061771 0251519 0.72 0.475 size 0087271 0049486 1.76 0.079 -.0010118 018466 y1 0294788 0193881 1.52 0.129 -.0086772 0676347 y1 0294788 0193881-1.39 1.52 0.129 -.0086772 0676347 ap2-.0430284 0243263 0132545 1.84 0.067 -.0017586 y3 -.0178066 0128159 0.166 0074153 0036661 -0.01 0.992 -.0072494 0071802 ms 0180052 0251519 0.72 0.475 -.031494 0675044 y2 -.025188 0159374 -1.58 0.115 -.056553 0061771 y2 -.025188 0159374 -1.58 0.115 -.056553 0061771 0294788 0193881 1.52 0.129 -.0086772 y4-.0017586 -.0060606 0504113 0132238 -0.46 0.647 y1-.0320852 0199641 0132545 1.84 0.067 aprov -.0000346 0036661 -0.01 0.992 -.0072494 0071802 -1.58 y3-.0178066 -.0178066 0128159-2.04 -1.39 0.166 -.0430284 0074153 y3 0128159 -1.39 0.166 -.0430284 0074153 y5 -.023426 0114776 0.042 -.000838 y2 -.046014 -.025188 0159374 0.115 -.056553 0193881 1.52 0.129 -.0086772 0676347 y4 -.0060606 0132238-0.24 -0.46 0.647 -.0320852 0199641 ap2 0243263 0132545 1.84 0.067 -.0017586 0504113 -1.39 y4 -.0060606 0132238 -0.46 0.647 -.0320852 0199641 y6 -.0029901 0127009 0.814 0220054 y3-.0279855 -.0178066 0128159 0.166 -.0430284 0159374 -1.58 0.115 -.056553 0061771 -.023426 0114776-2.27 -2.04 0.042 -.046014 -.000838 y5 -.023426 0114776 -2.04 0.042 -.046014 -.000838 y1 y5-.0194681 0294788 0193881 1.52 0.129 -.0086772 0676347 -0.46 y7 0085914 0.024 -.0025602 y4 -.036376 -.0060606 0132238 0.647 -.0320852 0128159 -1.39 0.166 -.0430284 0074153 y6 -.0029901 0127009 -0.24 0.814 -.0279855 0220054 y6 -.0029901 0127009 -0.24 0.814 -.0279855 0220054 y8 -.0043063 01047 0.6810.115 -.0249113 0162987 y2 -.025188 0159374 -0.41 -1.58 -.056553 0061771 -2.04 y5 -.023426 0114776 0.042 -.046014 0132238 -0.46 0.647 -.0320852 0199641 y7 -.0194681 0085914 -2.27 0.024 -.036376 -.0025602 y7 -.0194681 0085914 -2.27 0.024 -.036376 -.0025602 y9 0003879 0.963 -.0162781 017054 y3 -.0178066 0084685 0128159 0.05 -1.39 0.166 -.0430284 0074153 -0.24 y6 -.0029901 0127009 0.814 -.0279855 0114776 -2.04 0.042 -.046014 -.000838 y8 -.0043063 01047 -0.41 0.681 -.0249113 0162987 y11 008381 009616 0.87 0.384 -.0105434 0273053 y8 -.0043063 01047 -0.41 0.681 -.0249113 0162987 y4 -.0060606 0132238 -0.46 0.647 -.0320852 0199641 -2.27 y7 -.0194681 0085914 0.024 -.036376 0127009 -0.24 0.814 -.0279855 0220054 y9 0003879 0084685 0.05 0.963 -.0162781 017054 y12 0149477 0108045 1.38 0.168 -.0063157 0362111 y9 0003879 0084685 0.05 0.963 -.0162781 017054 y5 -.023426 0114776 -2.04 0.042 -.046014 -.000838 y8 -.0043063 01047 -0.41 0.681 -.0249113 y11 008381 009616 0.87 0.384 -.0105434 0273053 0085914 -2.27 0.024_cons -.036376 -.0025602 -.1159191 -0.90 0.368 -.3688392 137001 y11 008381 1285156 009616 0.87 0.384 -.0105434 0273053 y6 0127009 -0.24 0.814 -.0279855 0220054 y9 0003879 0362111 0084685 0.05 0.963 -.0162781 y12 -.0029901 0149477 0108045 1.38 0.168 -.0063157 01047 -0.41 0.681 -.0249113 0162987 y12 0149477 0108045 1.38 0.168 -.0063157 0362111 y7 -.0194681 0085914 -2.27 0.024 -.036376 -.0025602 _cons -.1159191 1285156 -0.90 0.368 -.3688392 137001 y11 008381 009616 0.87 0.384 -.0105434 0084685 0.05 Instruments 0.963 -.0162781 017054 _consfor orthogonal -.1159191 deviations 1285156 equation -0.90 0.368 -.3688392 137001 y8 -.0043063 01047 -0.41 0.681 -.0249113 0162987 y12 0149477 0108045 1.38 0.168 -.0063157 009616 0.87 0.384 -.0105434 0273053 GMM-type (missing=0, separate instruments for each period unless collapsed) Instruments for orthogonal.0084685 deviations equation y9 0003879 0.05 0.963 -.0162781 017054 _cons -.1159191 1285156 -0.90 0.368 -.3688392 0108045 1.38 Instruments 0.168 -.0063157 0362111 L(4/6).(L.fcfo L2.fcfo) collapsed for (missing=0, orthogonal deviations equation for each period unless collapsed) GMM-type separate instruments y11 008381.137001 009616 0.87 0.384 -.0105434 0273053 1285156 -0.90 0.368 -.3688392 L(0/2).L.dpr collapsed GMM-type (missing=0, L2.fcfo) separate collapsed instruments for each period unless collapsed) L(4/6).(L.fcfo y12 0149477collapsed 0108045 1.38 0.168for orthogonal -.0063157 deviations 0362111 equation Instruments L(1/11).(fcfu L2.fcfu) L(4/6).(L.fcfo L2.fcfo) collapsed L(0/2).L.dpr collapsed _cons -.1159191 -0.90 0.368 -.3688392 137001 GMM-type (missing=0, separate instruments for each period unless L(1/11).L2.size collapsed 1285156 deviations equation L(1/11).(fcfu L2.fcfu) collapsed L(0/2).L.dpr collapsed L(1/2).tobinq collapsed L(4/6).(L.fcfo L2.fcfo) collapsed rate instruments for each period unless collapsed) L(1/11).L2.size collapsed L(1/11).(fcfu L2.fcfu) collapsed L(0/4).L2.leverage collapsed L(0/2).L.dpr collapsed Instruments for orthogonal deviations equation collapsed L(1/2).tobinq collapsed L(1/11).L2.size collapsed Instruments for levels equation L(1/11).(fcfu collapsed GMM-type (missing=0, separate instruments for each periodL2.fcfu) unless collapsed) L(0/4).L2.leverage collapsed L(1/2).tobinq collapsed Standard Instruments for L2.fcfo) levels equation L(1/11).L2.size collapsed L(4/6).(L.fcfo collapsed collapsed L(0/4).L2.leverage collapsed dfcfu l2dfcfo l2dtobinq ldleverage l2dsize lddpr lage dms ms aprov ap1 ap2 Standard L(1/2).tobinq collapsed L(0/2).L.dpr collapsed ed Instruments for levels equation ap3 year y1l2dfcfo y2 y3 y4 y5 y6 y7ldleverage y8 y9 y10 l2dsize y11 y12 lddpr lage dms ms aprov ap1 ap2 dfcfu l2dtobinq L(0/4).L2.leverage collapsed L(1/11).(fcfu L2.fcfu) collapsed Standard _cons ap3 year y1 y2 y3 y4 y5 y6 y7 y8 y9 y10 y11 y12 Instruments fordms levels equation apsed L(1/11).L2.size collapsed dfcfu l2dfcfo l2dtobinq ldleverage l2dsize lddpr lage ms collapsed) aprov ap1 ap2 GMM-type (missing=0, separate instruments for each period unless _cons Standard tion L(1/2).tobinq collapsed ap3 year y1 y2 y3 y4 y5 y6 y7 y8 y9 y10 y11 y12 DL3.(L.fcfo L2.fcfo) collapsed GMM-type (missing=0, separate instruments for each period unless collapsed) dfcfu l2dfcfo l2dtobinq ldleverage l2dsize lddpr lage dms ms ap _cons L(0/4).L2.leverage collapsed Prob > F iu Robust Std Err y8 y9 y11 y12 _cons -.0043063 0003879 008381 0149477 -.1159191 01047 0084685 009616 0108045 1285156 -0.41 0.05 0.87 1.38 -0.90 xli 0.681 0.963 0.384 0.168 0.368 -.0249113 -.0162781 -.0105434 -.0063157 -.3688392 0162987 017054 0273053 0362111 137001 Instruments for orthogonal deviations equation GMM-type (missing=0, separate instruments for each period unless collapsed) L(4/6).(L.fcfo L2.fcfo) collapsed L(0/2).L.dpr collapsed L(1/11).(fcfu L2.fcfu) collapsed L(1/11).L2.size collapsed L(1/2).tobinq collapsed L(0/4).L2.leverage collapsed Instruments for levels equation Standard dfcfu l2dfcfo l2dtobinq ldleverage l2dsize lddpr lage dms ms aprov ap1 ap2 ap3 year y1 y2 y3 y4 y5 y6 y7 y8 y9 y10 y11 y12 _cons GMM-type (missing=0, separate instruments for each period unless collapsed) DL3.(L.fcfo L2.fcfo) collapsed DL.L.dpr collapsed D.(fcfu L2.fcfu) collapsed D.L2.size collapsed D.tobinq collapsed DL.L2.leverage collapsed Arellano-Bond test for AR(1) in first differences: z = Arellano-Bond test for AR(2) in first differences: z = Sargan test of (Not robust, Hansen test of (Robust, but -3.87 -1.17 overid restrictions: chi2(51) = 35.33 but not weakened by many instruments.) overid restrictions: chi2(51) = 60.55 weakened by many instruments.) Pr > z = Pr > z = 0.000 0.241 Prob > chi2 = 0.953 Prob > chi2 = 0.169 Appendix 22 FEM with clustering standard errors method for models Model 2.1 xtset firm_code year panel variable: time variable: delta: firm_code (unbalanced) year, 2008 to 2019, but with gaps unit xtreg iu FCFU FCFO, fe cluster ( firm_code) Fixed-effects (within) regression Group variable: firm_code Number of obs Number of groups R-sq: Obs per group: within = 0.5949 between = 0.3942 overall = 0.5646 corr(u_i, Xb) = = 2,258 303 = avg = max = 7.5 12 = = 57.19 0.0000 F(2,302) Prob > F = -0.0087 (Std Err adjusted for 303 clusters in firm_code) Robust Std Err iu Coef t FCFU FCFO _cons 6636196 -.0899538 -.0079641 0627619 0225136 0029545 sigma_u sigma_e rho 02995689 0464137 29407569 (fraction of variance due to u_i) 10.57 -4.00 -2.70 P>|t| 0.000 0.000 0.007 [95% Conf Interval] 5401136 -.1342573 -.0137781 7871256 -.0456504 -.00215 corr(u_i, Xb) = -0.0087 Prob > F = 0.0000 (Std Err adjusted for 303 clusters in firm_code) Robust Std Err xlii iu Coef FCFU FCFO _cons 6636196 -.0899538 -.0079641 0627619 0225136 0029545 sigma_u sigma_e rho 02995689 0464137 29407569 (fraction of variance due to u_i) t 10.57 -4.00 -2.70 P>|t| 0.000 0.000 0.007 [95% Conf Interval] 5401136 -.1342573 -.0137781 7871256 -.0456504 -.00215 Model 2.2 xtset firm_code year panel variable: time variable: delta: firm_code (unbalanced) year, 2008 to 2019, but with gaps unit xtreg iu FCFU FCFO ,fe cluster ( firm_code) Fixed-effects (within) regression Group variable: firm_code Number of obs Number of groups R-sq: Obs per group: within = 0.0156 between = 0.1101 overall = 0.0224 corr(u_i, Xb) = = 1,414 297 = avg = max = 4.8 12 = = 14.03 0.0000 F(2,296) Prob > F = 0.0491 (Std Err adjusted for 297 clusters in firm_code) Robust Std Err iu Coef t FCFU FCFO _cons -.0013595 0937151 0533756 0002568 0355737 0024685 sigma_u sigma_e rho 08654998 07095777 59803255 (fraction of variance due to u_i) -5.29 2.63 21.62 P>|t| 0.000 0.009 0.000 [95% Conf Interval] -.0018648 0237056 0485175 -.0008542 1637245 0582336 ...MINISTRY OF EDUCATION AND TRAINING STATE BANK OF VIETNAM BANKING UNIVERSITY OF HO CHI MINH CITY KIEU CONG BAO TRAN FACTORS IMPACTING ON ABNORMAL INVESTMENT OF LISTED FIRMS IN HO CHI MINH STOCK. .. CONG BAO TRAN iii ABSTRACT Title: FACTORS IMPACTING ON ABNORMAL INVESTMENT OF LISTED FIRMS IN HO CHI MINH STOCK EXCHANGE Abstract: The purpose of the thesis is to study factors impacting on abnormal. .. financial constraints and agency problems theory Finally, the study examines the influence of dividends on overinvestment The studying data consists of 306 non-financial listed firms in Ho Chi Minh Stock