NGÂN HÀNG NHÀ NƢỚC VIỆT NAM BỘ GIÁO DỤC VÀ ĐÀO TẠO TRƢỜNG ĐẠI HỌC NGÂN HÀNG TP HỒ CHÍ MINH NGUYỄN THỊ THÙY LINH CÁC YẾU TỐ ẢNH HƢỞNG ĐẾN LỢI NHUẬN CỦA CÁC NGÂN HÀNG THƢƠNG MẠI CỔ PHẦN VIỆT NAM KHÓA LU[.]
NGÂN HÀNG NHÀ NƢỚC VIỆT NAM BỘ GIÁO DỤC VÀ ĐÀO TẠO TRƢỜNG ĐẠI HỌC NGÂN HÀNG TP HỒ CHÍ MINH - NGUYỄN THỊ THÙY LINH CÁC YẾU TỐ ẢNH HƢỞNG ĐẾN LỢI NHUẬN CỦA CÁC NGÂN HÀNG THƢƠNG MẠI CỔ PHẦN VIỆT NAM KHÓA LUẬN TỐT NGHIỆP Chuyên ngành: Tài – Ngân hàng Mã ngành: 7340201 NGƢỜI HƢỚNG DẪN KHOA HỌC ThS NGUYỄN THỊ MỸ HẠNH TP Hồ Chí Minh, tháng 01 - 2020 Tai ngay!!! Ban co the xoa dong chu nay!!! NGÂN HÀNG NHÀ NƢỚC VIỆT NAM BỘ GIÁO DỤC VÀ ĐÀO TẠO TRƢỜNG ĐẠI HỌC NGÂN HÀNG TP HỒ CHÍ MINH - NGUYỄN THỊ THÙY LINH CÁC YẾU TỐ ẢNH HƢỞNG ĐẾN LỢI NHUẬN CỦA CÁC NGÂN HÀNG THƢƠNG MẠI CỔ PHẦN VIỆT NAM KHÓA LUẬN TỐT NGHIỆP Chuyên ngành: Tài – Ngân hàng Mã ngành: 7340201 NGƢỜI HƢỚNG DẪN KHOA HỌC ThS NGUYỄN THỊ MỸ HẠNH TP Hồ Chí Minh, tháng 01 - 2020 i TĨM TẮT Bài nghiên cứu phân tích yếu tố ảnh hƣởng đến lợi nhuận ngân hàng thƣơng mại cổ phần Việt Nam giai đoạn từ năm 2012 đến năm 2018 số liệu dùng đƣợc lấy từ báo cáo tài 30 ngân hàng thƣơng mại cổ phần hoạt động Tác giả phân tích yếu tố thơng qua việc xây dựng ba mơ hình hồi quy cho liệu theo phƣơng pháp bình phƣơng nhỏ (Ordinary Least Squares – OLS), mơ hình tác động cố định (Fixed Effects Model – FEM), mơ hình tác động ngẫu nhiên (Random Effects Model – REM), sau thực kiểm định nhƣ F – test Hausman để chọn đƣợc mơ hình phù hợp nhất, đồng thời tiến hành số kiểm định khác để kiểm tra xem mơ hình mắc phải khuyết tật đƣa phƣơng pháp khắc phục hồn thiện mơ hình Sau sử dụng phƣơng pháp bình phƣơng nhỏ tổng quát (generalized least squares – GLS) để khắc phục tƣợng phƣơng sai thay đổi đa cộng tuyến mơ hình kết nghiên cứu thu đƣợc yếu tố có yếu tố tác động chiều đến lợi nhuận quy mô ngân hàng, tỷ lệ an toàn vốn, tỷ lệ cho vay, rủi ro khoản, tăng trƣởng tín dụng, lạm phát tốc độ tăng trƣởng, hai biến tỷ lệ tiền gửi rủi ro tín dụng tác động ngƣợc chiều đến lợi nhuận Với kết đạt đƣợc phần giúp ngân hàng quan liên quan khác việc đánh giá, quản lý hoạt động hiệu hơn, nhƣ đóng góp thêm nghiên cứu thực nghiệm cho tƣơng lai ii ABSTRACT The banking system is the most important section of the economy Banks play a very important role in providing capital, demand deposits, making loans and providing other services for the public The health of the country‟s economy is closely related to the soundness of banking system Therefore, profit maximization is an important target that always concerned by all companies in general and joint stock commercial banks in Viet Nam in particular The profitability of a bank is affected by numerous factors, including internal elements of financial institutions and several external factors of the economy Therefore, the author has chosen the research topic „Factors affecting Vietnamese commercial banks‟ profitability‟ in the hope of making some recommendations for bankers in Vietnam The author used two theories in this study: the Efficiency Structure theory (ES) and the Market Power theory (MP) The ES hypothesis, which written by Smirlock (1985), demonstrates a direct relationship between market concentration, competition and efficiency This hypothesis that efficient firms can reduce production costs and thus gain higher profits and greater market share There are two distinct approaches: the X – efficiency and Scale – efficiency theory According to the X – efficiency approach, more efficient companies are more profitable because they have lower costs According to the Scale – efficiency approach, the relation between market share and efficient firm will be explained by scale Larger companies can obtain lower cost and higher profit than others With MP theory, there are also two distinct approaches: the Struture – conduct – performance (SCP) and the Relative market power theory (RMP) SCP theory indicates that performance is determined by the characteristics of a market structure because market structure influences a company's behavior which will affect the results achieved in the market RMP theory that companies with a large market share combined with differences in products can make a profit without too much competition iii In this thesis, the author will analyse factors affecting the profitability of Vietnamese joint stock commercial banks of in the period from 2012 to 2018, with collect statistical data from 30 banks By building regression models with three methods: the pooled regression model (Ordinary Least Squares - OLS), Fixed Effects models (FEM) and Random effects models (REM), the author also carry out the inspection as F - test and Hausman to choose the best model and perform some testing to check defects of the model to take remedial measures to help improve the research results th The research model is built based on financial theories, domestic and international studies In this study, author will base on models of Sehrish Gul, Faiza Irshad, Khalid Zaman (2011), “Factors affecting bank profitability in Pakistan” Model as follows: However, the author only chooses two indicators, ROA and ROE as dependent variables, which the most commonly used factors and change some independent variables in models research: ROA = X0 + X1*Sizei,t + X2*Capitali,t + X3*Loani,t + X4*Depositi,t + X5*LQi,t + X6*CRi,t + X7*LGi,t + X8*GDPi,t + X9*INFi,t + ui,t ROE = X0 + X1*Sizei,t + X2*Capitali,t + X3*Loani,t + X4*Depositi,t + X5*LQi,t + X6*CRi,t + X7*LGi,t + X8*GDPi,t + X9*INFi,t + ui,t ROA is a ratio calculated by dividing the net income on total assets, which used in most of the studies about profitability of banks A higher ROA shows that the company is more efficient in using its resources iv ROE is a measure of financial performance calculatied by dividing net income by shareholders‟ equity The higher ROE indicates that the higher the profit generated and the better the capital management SIZE: Bank size is calculated by log (total assets) Larger banks will generate more profits than small banks A positive relationship is expected between size and bank‟s profit CAPITAL is a ratio of equity capital to total assets, which is expected to have a positive relationship with profitability of joint stock commercial banks LOAN is the main source of income for banks and is expected to have a positive effect on banks performance DEPOSIT is taken as the ratio of total deposits to total assets, which are the main source of bank funding A positive relationship is expected between deposit and profitability of banks LQ: Liquidity risk is a ratio calculated by cash and cash equivalent on total assets and expected to have a positive relationship with bank‟s profit CR: Credit risk is a ratio calculated by loan loss reserves to gross loans and assumed have a negative effect on profitability LG: Loan growth is calculated by dividing total loans (previous loans) by previous loans LG will affect nagetive to banks‟s profit GDP: Real gross domestic product is the most frequently used macroeconomic indicators When GDP growth, demand for credits or loans increased as well as the quality of asset, so bank can make higher profit GDP growth has a positive impact on profitability of banks INF: Inflation is the rate at which the general level of prices for goods and services is increasing in economy overtime If banks can adjust interest rate timely when inflation rate rises, the revenues of banks will increase Therefore, the author expected to INF has a positive relationship with bank profitability v The next step is to carry out descriptive statistics for all the variables in this study The author will compare to average value, mean value, standard deviation, minimum and maximum values between 210 observations collected from period 2012 to 2018 After that, the pearson‟s correlation matrix analysis will help to make an examination of correlation of dependent variables and independent variables The author conducted a regression analysis for all independent variables of the study by VIF The results of the test are that all VIF value of these variables less than 10, so the research models not have multicollinearity Firstly, using F – test to choose a model between OLS and FEM, both ROA and ROE models, p value equals 0.0000 so FEM model will be chosen Secondly, Hausman test will be carried out to compare FEM and REM Because p value equals 0.0000 less than 1%, so FEM is the best suit model for all ROA and ROE Thirdly, the author will check heteroscedasticity by Wald test and autocorrelation by Wooldrige test The results of test, p values were 0.0000 lesst than 1% significance level, indicating that both ROA and ROE models are existed heteroscedasticity and autocorrelation, so the author used to GLS method to make good these flaws In conclusion, the analysis showed that there are seven factors, including size, capital ratio, loan ratio, liquidity risk, loan growth, GDP and INF, which had a benefit impact on bank‟s profit, while only two factors, deposit ratio and credit risk had a negative impact Based on the above findings, the study recommends that bankers would encourage revenue diversifications, reduce the cost of operation, banking system will have to follow stricter rules in crediting process to minimize credit risk and make them more accountable for their decisions This thesis would be useful for many parties, like government, regulators, bankers and other researchers, to understand determinants that affect banks‟ profitability performance so that they could make informed decisions vi LỜI CAM ĐOAN Với tƣ cách ngƣời thực khố luận, tơi xin có lời cam đoan nhƣ sau: Tôi tên là: Nguyễn Thị Thùy Linh Sinh ngày: 19/03/1997 Hiện sinh viên lớp HQ3 – GE02, trực thuộc Khoa Ngân hàng, chƣơng trình đào tạo Chất lƣợng cao, trƣờng Đại học Ngân hàng Thành phố Hồ Chí Minh Đề tài khóa luận tốt nghiệp: Các yếu tố ảnh hƣởng đến lợi nhuận ngân hàng thƣơng mại cổ phần Việt Nam Giảng viên hƣớng dẫn: ThS Nguyễn Thị Mỹ Hạnh Bài luận văn hồn tồn tơi thực dƣới hƣớng dẫn ThS Nguyễn Thị Mỹ Hạnh, đúc kết từ tồn kiến thức tơi học đƣợc trƣờng kết hợp với việc tham khảo nghiên cứu thực nghiệm trƣớc Nội dung làm khơng chép đâu thông tin đƣợc trích dẫn rõ ràng, cụ thể Tơi xin hồn toàn chịu trách nhiệm cho lời cam đoan Tác giả Nguyễn Thị Thùy Linh vii LỜI CẢM ƠN Tôi xin gửi lời cảm ơn chân thành đến giảng viên hƣớng dẫn ThS Nguyễn Thị Mỹ Hạnh hết lòng hỗ trợ, hƣớng dẫn bảo cho suốt quãng thời gian làm khóa luận tốt nghiệp, nhờ giúp đỡ mà khố luận tơi hoàn thành cách hiệu Tuy nhiên, giới hạn thời gian thực nhƣ hạn chế kiến thức thân nên q trình làm đề tài khơng thể tránh khỏi sai sót, tơi mong nhận đƣợc báo nhƣ ý kiến đóng góp từ thầy/cơ để đề tài đƣợc hồn thiện phục vụ cho nghiên cứu sau Tôi xin chân thành cảm ơn Nguyễn Thị Thùy Linh 60 KẾT LUẬN CHƢƠNG Kết thúc nghiên cứu tác giả tìm đƣợc câu trả lời cho mục tiêu nghiên cứu đề tài nhƣ việc chấp nhận hay bác bỏ giả thuyết kỳ vọng đ t Đồng thời khuyến nghị giúp cho nhà lãnh đạo ngân hàng tham khảo tìm hƣớng phát triển phù hợp làm tăng lợi nhuận tƣơng lai Cuối cùng, thông qua làm giúp nhà nghiên cứu phát triển lĩnh vực hoàn thiện 61 TÀI LIỆU THAM KHẢO Danh mục tài liệu tham khảo tiếng Việt: Lê Tấn Phƣớc, Bùi Xuân Diễn 2016, „Các yếu tố tác động đến tỷ suất sinh lợi Ngân hàng thƣơng mại Việt Nam‟, Tạp chí khoa học – Đại học Đồng Nai, số Nguyễn Trần Thịnh 2013, Phân tích nhân tố ảnh hưởng đến lợi nhuận ngân hàng TMCP có niêm yết Việt Nam, Luận văn Thạc sĩ Kinh tế, Trƣờng đại học Kinh tế Thành phố Hồ Chí Minh Nguyễn Việt Hùng 2008, Phân tích nhân tố ảnh hưởng đến hiệu hoạt động ngân hàng thương mại Việt Nam, Luận án Tiến Sĩ Kinh tế, Đại học Kinh tế Quốc dân Phan Hiền Giang 2011, Phân tích hiệu hoạt động kinh doanh ngân hàng thương mại nhà nước ngân hàng thương mại cổ phần TP Cần Thơ, Luận văn thạc sỹ, Trƣờng đại học Cần thơ Danh mục tài liệu tham khảo tiếng Anh: Abdus Samad 2015, „Determinants Bank Profitability: Empirical Evidence from Bangladesh Commercial Banks‟, International Journal of Financial Research, vol 6, no Ahmad Aref Almazari 2014, „Impact of Internal Factors on Bank Profitability: Comparative Study between Saudi Arabia and Jordan‟, Journal of Applied Finance & Banking, vol 4, no 1, pp 125-140 Aremu Mukaila Ayanda, Imoh Christopher, Mustanpha Adeniyi Mudashiru 2013, „Determinants of Capital Structure in Nigerian Banking Sector‟, International Journal of Academic Research in Economics and Management Sciences July 2013, vol 2, no 62 Bourke, P., 1989, „Concentration and other determinants of bank protitability in Europe, North America and Australia‟, Journal of Banking and Finance 13, pp 65 – 79 Deger Alper Adem Anbar 2011, „Bank Specific and Macroeconomic Determinants of Commercial Bank Profitability: Empirical Evidence from Turkey‟, Business and Economics Research Journal, vol 2, no 2, pp 139 – 152 Fadzlan Sufian, Royfaizal Razali Chong 2008, „Determinants of bank profitability in a developing economy: empirical evidence from the Philipines‟, Asian Academy of Management Journal of Accounting and Finance, vol 4, no 2, pp 91 – 112 Haroon Jabbar 2014, „Determinants of Banks Profitability‟, IOSR Journal of Business and Management (IOSR-JBM), vol 16, issue 1, pp 109 – 113 Imad Z Ramadan, Qais A Kilani, Thair A Kaddumi 2011, „Determinants of Bank Profitability: Evidence from Jordan‟, International Journal of Academic research, vol.3, no 4, pp 180 – 191 Indra Satria, Edy Supriyadi Agus s Irfani, Achmad djami 2018, „The Most Important Factors Affecting Profitability of the Top 10 Commercial Banks in Asean‟, The International Journal of Social Sciences and Humanities Invention, vol 5, no Sehrish Gul, Faiza Ishad, Khailid Zaman 2011, „Factors affecting bank profitability in Pakistan‟, The Romanian Economic Journal, pp 61 – 87 Tobias Olweny, Themba Mamba Shipho 2011, „Effects of Banking Sectoral Factors on the Profitability of Commercial Banks in Kenya‟, Economics and Finance Review, vol 1(5), pp – 30 Vincent Okoth Ongore, Gemechu Berhamu Kusa 2013, „Determinants of Financial Performance of Commercial Banks in Kenya‟, International Journal of Economics and Financial Issues, vol 3, no 1, pp 237 – 252 63 PHỤ LỤC Thống kê mô tả sum roa roe size capital loan deposit lq cr lg inf gdp Variable roa roe size capital loan deposit lq cr lg inf gdp Obs 210 210 210 210 210 210 210 210 210 210 210 Mean 0.0061 0.0720 8.0431 0.0904 0.5503 0.7672 0.1358 0.0131 0.2140 0.0362 0.0631 Kiểm tra đa cộng tuyến vif Variable size capital loan deposit lq cr lg inf gdp Mean VIF VIF 2.67 2.75 1.88 1.49 1.24 1.28 1.10 1.86 2.01 1.81 1/VIF 0.3745 0.3636 0.5319 0.6711 0.8065 0.7813 0.9091 0.5376 0.4975 Std Dev 0.0050 0.0569 0.4866 0.0393 0.1184 0.0844 0.0719 0.0050 0.1855 0.0209 0.0082 Min 0.0001 0.0007 7.1233 0.032253 0.21621 0.50903 0.01521 0.00541 -0.2459 0.0060 0.0525 Max 0.0264 0.24441 9.1183 0.238381 0.77413 0.89597 0.47324 0.04752 1.0682 0.0681 0.0780 64 Ma trận tƣơng quan biến pwcorr roa roe size capital loan deposit lq cr lg inf gdp, sig roa roe size capital loan deposit lq roa 1.0000 roe 0.8448 0.0000 1.0000 size 0.1416 0.0404 0.4799 0.0000 1.0000 capital 0.2465 0.0003 -0.2086 0.0024 -0.6915 0.0000 1.0000 loan 0.1899 0.0058 0.2868 0.0000 0.3143 0.0000 -0.0987 0.1541 1.0000 deposit -0.2428 0.0004 -0.1065 0.1239 0.1207 0.0810 -0.3038 0.0000 0.3822 0.0000 1.0000 lq 0.0285 0.6814 -0.0186 0.7890 -0.2421 0.0004 0.1428 0.0386 -0.3773 0.0000 -0.0940 0.1747 1.0000 cr 0.0304 0.6618 0.0071 0.9191 0.0964 0.1641 0.0920 0.1843 -0.2320 0.0007 -0.2861 0.0000 -0.0237 0.7322 lg 0.0853 0.2184 0.1587 0.0214 -0.0415 0.5499 -0.1639 0.0175 -0.1226 0.0762 0.0113 0.8711 0.0702 0.3115 inf 0.1266 0.0670 -0.0029 0.9666 -0.1514 0.0283 0.2413 0.0004 -0.1654 0.0164 -0.3054 0.0000 0.1203 0.0821 gdp 0.0405 0.5594 0.1982 0.0039 0.2568 0.0002 -0.3111 0.0000 0.3344 0.0000 0.2694 0.0001 -0.0970 0.1612 cr lg inf gdp cr 1.0000 lg -0.0625 0.3674 1.0000 inf 0.2731 0.0001 -0.0170 0.8061 1.0000 gdp -0.2531 0.0002 -0.0590 0.3947 -0.6407 0.0000 1.0000 65 MÔ HÌNH ROA Mơ hình Pooled OLS reg roa size capital loan deposit lq cr lg inf gdp Source SS df Model 0.00195 Residual 0.00323 200 Total 0.00518 209 roa size capital loan deposit lq cr lg inf gdp _cons MS Number of obs = F(9, 200) = 0.000216735 Prob > F = 0.000016166 R-squared = Adj R-squared = 0.000024803 Root MSE = 210 13.41 0.0000 0.3763 0.3482 0.0402 Coef Std Err t P>t [95% Conf Interval] 0.00616 0.08603 0.00596 -0.011 0.00649 -0.0859 0.00657 0.04122 0.12972 -0.057 0.000934 0.011724 0.0032163 0.0040234 0.0043028 0.0627607 0.0015738 0.0181238 0.0482959 0.0096677 6.6 7.34 1.85 -2.73 1.51 -1.37 4.17 2.27 2.69 -5.89 0.000 0.000 0.065 0.007 0.133 0.173 0.000 0.024 0.008 0.000 0.0043192 0.0080027 0.0629065 0.1091437 -0.0003788 0.0123055 -0.0189297 -0.0030623 -0.0019904 0.0149791 -0.2096184 0.0378968 0.0034635 0.0096702 0.0054825 0.076959 0.0344833 0.2249524 -0.0760307 -0.0379033 66 Mô hình FEM xtreg roa size capital loan deposit lq cr lg inf gdp, fe Fixed-effects (within) regression Group variable: id01 Number of obs = 210 Number of groups = 30 R-sq: within = 0.4586 between = 0.1882 overall = 0.2615 Obs per group: = avg = max = corr(u_i, Xb) = -0.5206 F(9,171) Prob > F roa Coef Std Err t P>t size capital 0.012194 0.096354 0.0029012 0.013303 4.20 7.24 0.000 0.000 loan 0.005809 0.0041124 1.41 0.016 deposit lq -0.01578 0.011249 0.0040887 0.004532 -3.86 2.48 0.000 0.014 cr -0.1325 0.0575095 -2.30 0.022 lg inf 0.002341 0.029446 0.0014209 0.013594 1.65 2.17 0.101 0.032 gdp 0.038744 0.0540279 0.72 0.474 _cons -0.09564 0.0228846 -4.18 0.000 sigma_u sigma_e rho 0.004026 0.002891 0.659852 = = 7.0 16.09 0.0000 [95% Conf Interval] 0.0064671 0.0700944 0.0023091 0.0238523 0.0023033 0.2460228 0.0004635 0.0026125 0.0679039 0.1408086 0.0179207 0.1226131 0.0139262 -0.0077104 0.0201951 -0.0189828 0.0051459 0.0562797 0.1453911 -0.0504633 (fraction of variance due to u_i) F test that all u_i=0: F(29, 171) = 7.44 Prob > F = 0.0000 67 Mơ hình REM xtreg roa size capital loan deposit lq cr lg inf gdp, re Random-effects GLS regression Group variable: id01 Number of obs = 210 Number of groups = 30 R-sq: within = 0.4474 between = 0.2650 overall = 0.3490 Obs per group: = avg = max = corr(u_i, X)=0(assumed) Wald chi2(9) Prob > Chi2 roa size capital loan deposit lq cr lg inf gdp _cons sigma_u sigma_e rho Coef Std Err z P>z 007203 085892 005918 -.015539 010781 -.140959 002915 034614 105799 -.059155 001340 011646 003594 003845 004246 054765 001379 013610 040742 011826 5.37 7.38 1.65 -4.04 2.54 -2.57 2.11 2.54 2.60 -5.00 0.000 0.000 0.100 0.000 0.011 0.010 0.035 0.011 0.009 0.000 0.002722 0.002891 0.469855 (fraction of variance due to u_i) 7.0 = 146.03 = 0.0000 [95% Conf Interval] 004576 063066 -.001126 -.023076 002458 -.248298 000211 007938 025946 -.082335 009829 108717 012961 -.008004 019104 -.033621 005618 061289 185653 -.035976 68 Kiểm định Hausman hausman fem rem Coefficients -(b) (B) (b-B) fem rem Difference size capital loan deposit lq cr lg inf gdp 0121939 0963537 0058085 -.0157814 0112492 -.1325028 0023412 0294461 0387436 B= 0072026 0858915 0059177 -.0155399 0107809 -.1409593 002915 0346135 1057996 sqrt(diag(V_b-V_B)) S.E .0049913 0104623 -.0001092 -.0002414 0004683 0084565 -.0005738 -.0051674 -.067056 0025731 0064297 0019992 0013901 0015835 0175524 0003408 0354837 b = consistent under Ho and Ha; obtained from xtreg inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic chi2(9) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 36.8 Prob>chi2 = 0.0000 (V_b-V_B is not positive definite) Kiểm định Wald xttest3 Modified Wald test for groupwise heteroskedasticity in fixed effect regression model H0: sigma(i)^2 = sigma^2 for all i chi2 (30) = 2150.89 Prob>chi2 = 0.0000 69 Kiểm định Wooldridge xtserial roa size capital loan deposit lq cr lg inf gdp Wooldridge test for autocorrelation in panel data H0: no first order autocorrelation F( 1, 29) = 28.023 Prob > F = 0.0000 Mơ hình GLS xtgls roa size cap loan deposit lq cr lg inf gdp, panels (heteroskedastic) corr(ar1) Cross-sectional time-series FGLS regression Coefficients: generalized least squares Panels: heteroskedastic Correlation: common AR (1) coefficient for all panels (0.5363) Estimated covariances 30 Number of obs Estimated autocorrelations = Number of groups = 30 Estimated coefficients 10 Time periods = Wald chi2(9) = 109.58 Prob > chi2 = 0.0000 roa size capital loan deposit lq cr lg inf gdp _cons = = Coef Std Err z P>z 006428 091259 001647 -.004827 007189 -.121653 002763 026199 105829 -.059134 000897 011175 003244 003157 003257 051985 001088 010734 036205 008353 7.16 8.17 0.51 -1.53 2.21 -2.34 2.54 2.44 2.92 -7.08 0.000 0.000 0.612 0.126 0.027 0.019 0.011 0.003 0.015 0.000 = 210 [95% Conf Interval] 004669 069356 -.004711 -.011014 000807 -.223540 000631 034868 005162 -.075505 008187 113162 008006 001359 013573 -.019765 004894 047237 176789 -.042762 70 MƠ HÌNH ROE Mơ hình Pooled OLS reg roe size capital loan deposit lq cr lg inf gdp Source SS df Model 0.27697 Residual 0.39913 200 Total 0.67611 209 roe size capital loan deposit lq cr lg inf gdp _cons Coef .07087 31920 11239 -.14576 11771 -.53315 07709 35544 1.3399 -.59986 Std Err .01038 13026 03574 04470 04781 69732 01749 20137 53661 10742 MS Number of obs = F(9, 200) = 0.030774641 Prob > F = 0.001995669 R-squared = Adj R-squared = 0.003234955 Root MSE = t P>t 6.83 2.45 3.15 -3.26 2.46 -0.76 4.41 1.77 2.50 -5.58 0.000 0.015 0.002 0.001 0.015 0.445 0.000 0.079 0.013 0.000 210 15.42 0.0000 0.4097 0.3831 0.04467 [95% Conf Interval] 05041 06234 04193 -.23391 02344 -1.90819 04261 -.04164 28175 -.81167 09134 57607 18286 -.05761 21198 84189 11157 75252 2.39801 -.38805 71 Mơ hình FEM xtreg roe size capital loan deposit lq cr lg inf gdp, fe Fixed-effects (within) regression Group variable: id01 Number of obs = 210 Number of groups = 30 R-sq: within = 0.3652 between = 0.3748 overall = 0.3602 Obs per group: = avg = max = corr(u_i, Xb) = -0.3003 F(9,171) Prob > F roe size capital loan deposit lq cr lg inf gdp _cons sigma_u sigma_e rho Coef Std Err t P>t 10103 31707 09566 -.16480 13195 -1.73537 02163 33434 72541 -.75327 03096 14197 04389 04364 04837 61376 01516 14508 57659 24423 3.26 2.23 2.18 -3.78 2.73 -2.83 1.43 2.30 1.26 -3.08 0.001 0.027 0.031 0.000 0.007 0.005 0.156 0.022 0.210 0.002 03821964 03085186 60546879 = = 7.0 10.93 0.0000 [95% Conf Interval] 03992 03682 00903 -.25094 03647 -2.94688 -.0083 04796 -.41276 -1.23537 16215 59731 18229 -.07867 22742 -.52386 05157 62071 1.86358 -.27118 (fraction of variance due to u_i) F test that all u_i=0: F(29, 171) = 8.56 Prob > F = 0.0000 72 Mơ hình REM xtreg roe size capital loan deposit lq cr lg inf gdp, re Random-effects GLS regression Group variable: id01 Number of obs = 210 Number of groups = 30 R-sq: within = 0.3606 between = 0.3953 overall = 0.3811 Obs per group: = avg = max = corr(u_i, X)=0(assumed) Wald chi2(9) Prob > Chi2 roe size capital loan deposit lq cr lg inf gdp _cons sigma_u sigma_e rho Coef Std Err z P>z 07641 27005 09793 -.16609 13417 -1.60617 03051 35189 1.08724 -.57853 01471 12497 03871 04119 04553 58574 01475 14514 43760 12895 5.20 2.16 2.53 -4.03 2.95 -2.74 2.07 2.42 2.48 -4.49 0.000 0.031 0.011 0.000 0.003 0.006 0.039 0.015 0.013 0000 03055229 03085186 49512153 (fraction of variance due to u_i) 7.0 = 115.38 = 0.0000 [95% Conf Interval] 04758 02511 02207 -.24684 04494 -2.75421 00160 06741 22955 -.83127 10523 51499 17379 -.08535 22340 -.45814 05941 63636 1.94493 -.32579 73 Kiểm định Hausman hausman fem1 rem1 Coefficients -(b) (B) (b-B) fem rem Difference size capital loan deposit lq cr lg inf gdp B= 1010345 3170677 0956634 -.1648031 1319454 -1.73537 0216326 334337 7254109 076405 2700537 0979327 -.1660932 1341694 -1.606172 0305077 3518857 1.08724 sqrt(diag(V_b-V_B)) S.E .0246295 047014 -.0022693 0012901 -.002224 -.1291982 -.0088751 -.0175487 -.3618292 0272478 0673633 0206898 0143867 0163298 1833095 0035261 3754588 b = consistent under Ho and Ha; obtained from xtreg inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic chi2(9) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 52.53 Prob>chi2 = 0.0000 (V_b-V_B is not positive definite) Kiểm định Wald xttest3 Modified Wald test for groupwise heteroskedasticity in fixed effect regression model H0: sigma(i)^2 = sigma^2 for all i chi2 (30) = 3696.77 Prob>chi2 = 0.0000 74 Kiểm định Wooldridge xtserial roe size capital loan deposit lq cr lg inf gdp Wooldridge test for autocorrelation in panel data H0: no first order autocorrelation F( 1, 29) = 78.431 Prob > F = 0.0000 Mơ hình GLS xtgls roe size cap loan deposit lq cr lg inf gdp, panels (heteroskedastic) corr(ar1) Cross-sectional time-series FGLS regression Coefficients: generalized least squares Panels: heteroskedastic Correlation: common AR (1) coefficient for all panels (0.6675) Estimated covariances 30 Number of obs Estimated autocorrelations = Number of groups = 30 Estimated coefficients 10 Time periods = Wald chi2(9) = 93.44 Prob > chi2 = 0.0000 roe size capital loan deposit lq cr lg inf gdp _cons = = Coef Std Err z P>z 07931 37947 06853 -.09229 09442 -1.45579 01859 14466 48653 -.60665 01141 11546 03451 03317 03456 55183 01126 10463 37612 10206 6.59 3.29 1.99 -2.78 2.73 -2.64 1.65 1.38 1.29 -5.94 0.000 0.001 0.047 0.005 0.006 0.008 0.099 0.167 0.196 0.000 = 210 [95% Conf Interval] 05695 15317 00089 -.15732 02668 -2.53736 -.00348 -.06040 -.25064 -.80668 10167 60578 13617 -.02728 16216 -.37424 04067 1.22371 34972 -.40661