Factors affecting performance of listed commercial banks: evidence on Vietnamese securities market

642 | Policies and Sustainable Economic Development Factors Affecting Performance of Listed Commercial Banks: Evidence on Vietnamese Securities Market TRAN QUOC THINH Banking University of Ho Chi Minh city - thinhtq@buh.eud.vn HOANG YEN NHI BNP PARIPAS Bank, Branch of Ho Chi Minh City - hoangnhi94.buh@gmail.com Abstract In the trend of integration with the region and the world, Vietnam has joined economic organizations such as the ASEAN Economic Community (AEC) or Trans-Pacific Partnership (TPP) This contributes to creating conditions for the development of Vietnam, especially domestic commercial banks (CBs), the central coordinator for economic capital However, the competitive challenges also posed many problems to be solved The authors uses quantitative methods to determine the factors affecting the performance of the listed commercial banks On this basis, the authors proposed a number of policies related to cost control, credit quality, and development strategies in line with the orientation of the sector, increasing investment in machinery and application technology This should lead to enhanced operational efficiency for listed commercial banks to make their presence felt along with sustainable development Keywords: performance; commercial banks; economic integration Policies and Sustainable Economic Development | 643 Introduction Bank birth and development are associated with the development of commodity economy to address the needs of capital distribution, billing, production and business expansion of economic organizations and individuals In the context of globalization, Vietnam joined the ASEAN Economic Community (AEC) in 2015; the issue of financial liberalization is the inevitable trend of the country Moreover, the Trans-Pacific Partnership (TPP) signed in 2016 brought many economic benefits to Vietnam together with favorable chances of development in many economic aspects However, the integration process requires commercial banks to improve competitiveness as well as administrative capacity and to ensure safe operation and efficiency Thus, bank performance becomes an important criterion for assessing its existence and development in such a competitive international environment Overview of research and methodology 2.1 The concept of performance There are many existing views about perceived performance, depending on the field of study A British economist, Adam Smith (1737-1790), stated that the performance is the result achieved in economic activity, the consumption of goods turnover In this view, Adam Smith equated the effectiveness and results; the different cost levels yielded the same result and the same effectiveness (Fry, 2005) Some other views for that performance are determined by the ratio of the results achieved and the cost of money to get results One typical example is Kuhn (1999), who suggested that the effectiveness is determined by taking the results calculated by dividing the unit value for business expenses, as agreed by many economists and business executives given operational efficiency However, correlations among quantity and quality and cost outcomes have not been supported Woele (1990) asserted that efficiency ratio is the relationship between output per unit in kind and the amount of input factors (labor hours, day labor, equipment units, materials, etc ), called the performance of technical nature According to this view the performance reflected rises in the productivity of inputs and outputs Doring (2000) suggested that the operational efficiency ratio is the relationship between the business expenses paid out in the most favorable conditions for business and the actual cost to spend, considered effective in terms of value Efficiency is determined by identifying the lowest business costs in the most favorable conditions bringing actual costs incurred versus planned costs (Weber, 2009) Rivard and Thomas (1997), one of the pioneers to suggest performance efficiency measurement which was recognized by many researchers, maintained that the performance is measured through several indicators such as earnings per share (EPS), rate of return on assets (ROA), and return on equity (ROE) Within the scope of the study of all this article, the authors focus on ROE as this is one of the indicators used in the field of popular economy 644 | Policies and Sustainable Economic Development 2.2 Previous studies Staikouras (1999) used quantitative analysis method to study the effectiveness of the 685 European banks during 1994-1998 Many factors found to positively affect performance include the stock market size, total assets, total loans to total assets, risk provisions, and loans to total loan amount, which also affects bank performance but in the opposite direction Naceur (2003) performed non-parametric analysis of the impacts of these factors on the performance of 10 banks in Tunisia for the period 1980-2000 The results showed that the economic indicators such as inflation and macroeconomic growth have no impact on net profit Badola and Verma (2006) used multivariate regression models to study the factors affecting the profitability of 27 listed banks in India for the period 1998-2004 They detected effects of total revenue, expenditures for technology upgrades, credit risk reserves, and the difference between interest receivable and interest payable Athanasoglou et al (2006) adopted quantitative methods to study the factors affecting the performance of 132 banks in southeast Europe (1998-2002), indicating that credit risk, operating costs, and capital size negatively impact on performance and that high credit risks as ROE decrease the impact Tariq et al (2014) studied the determinants of the efficiency of commercial banks in Pakistan for the period 2004-2010 with a sample of 17 commercial banks They explored the capital resources of banks that make sense in the performance of the bank Recently, Duraj and Moci (2015) examined the factors affecting the banking operation efficiency by means of multivariate regression for the case of 16 banks in Albania in 1999-2014, and demonstrated that except for ratio of nonperforming loans to total loans, the remaining factors such as liquidity risk or inflation negatively influence ROE, while debt levels and economic growth may have similar effects on ROE In Vietnam, Nguyen (2008) applied a combination of qualitative and quantitative methods to their study of the factors affecting the performance of 32 commercial banks in Vietnam, 2001-2005 His empirical results showed significant factors affecting liquidity risk and proportion of loans As such, in order to improve efficiency, banks need to reduce liquidity risk, strengthening the capacity of managers, reducing the proportion of lending to those that cannot afford to finance Trinh and Nguyen (2013) also studied the factors affecting the performance of 39 Vietnamese commercial banks for the period from 2005 to 2012 and concluded that the rate of return, ratio of equity to total assets, and ratio of loans to total assets have significant impact levels Most recently, Nguyen (2015) analyzed the factors affecting the profitability of listed commercial banks in Vietnam stock market between 2009 and 2014, showing the five factors affecting margins, the operating cost factors that affect the most profitable banks, followed by loans, liquidity, inflation and the end user and credit risk 2.3 Research methods The authors use quantitative methods with the assistance of SPSS 20 software as well as Microsoft Excel to make the calculations, statistical description, data processing, and analysis of regression models Policies and Sustainable Economic Development | 645 Study design 3.1 Sample The research sample contains listed joint-stock banks in the stock market of Vietnam, whereas the research phase covers the 2010 – 2015 period 3.2 Description of the variables studied Table Description of the variables used in the regression model Variable Description Measurement Dependent variable: ROE (Y) margins on equity = (Net Profit / Average Equity) x100% Independent variables: TCTR (X1) The ratio of operating expenses to total revenues = (Total operating expenses / net sales) x100% Independent variables: logTA (X2) bank size = Log (total assets) Independent variables: LOANTA (X3) rate of loans to total assets = (Loans to customers / Total assets) x100% Independent variables: ETA (X4) Ratio of equity to total assets = (Equity / Total assets) x100% Independent variables: TK (X5) Liquidity Index = (assets with high liquidity / Total assets) x100% Independent variables: NCA (X6) The percentage value of the investment in machinery and equipment and computer software on assets = (Value of investment in machinery and equipment and computer software / Total assets) x100 % Independent variables: GDP (X7) Economic growth y = (dY/Y)x100% 3.3 Research model The authors employ a combination of approaches as suggested by Athanasoglou et al (2006) and Nguyen (2015) in addition to surveying a number of experts to compute the ratio of investments in machinery computer software and equipment to total assets to match the characteristics and conditions of the commercial banks in Vietnam Therefore, a multivariate regression model is designed as follows: Yi= β + β1*X1+ β2*X2+ β3*X3+ β4*X4+ β5*X5+β6*X6+β7*X7+ ε where: 646 | Policies and Sustainable Economic Development Yi: dependent variable Y: Rate of return on equity (ROE) Xi: independent variable X1: The ratio of operating expenses to total revenue (TCTR) X2: The size of assets (logTA) X3: The rate of loans to total assets (LOANTA) X4: The ratio of equity to total assets (ETA) X5: Liquidity index (TK) X6: Percentage value of investment in machinery and equipment and computer software on assets (NCA) X7: Economic growth (GDP) Regression coefficients: β1, β2, β3, β4, β5, β6, β7 Regression error: ε Results and discussion 4.1 Descriptive statistical analysis Table Results of descriptive statistics of variables Variable N Minium Maxium Mean Std Deviation Y (ROE) 54 0.07 27.49 12.19963 706.1349 X1 (TCTR) 54 35.62 103.71 60.1276 1594.71 X2 (log(TA)) 54 7.3014 8.9298 8.2857 40.2389 X3 (LOANTA) 54 36.23 71 56.142 969.09 X4 (ETA) 54 4.26 14.75 8.0191 210.97 X5(TK) 54 6.41 42.9 22.0089 842.77 X6 (NCA) 54 0.3 0.0924 6.05 X7 (GDP) 54 6.78 5.8763 99.58 Valid N (listwise) 54 Source: data analysis using SPSS 20 The average of ROE reaches 12.20%; however, the standard deviation of up to 706.13% shows significant differences in the uses of capital efficiency by the bank owner, or in other words they not resemble each other TCTR average value is 60.13%, implying that the cost-to-income ratio reveals a very high proportion The range from 35.62% to 103.71%, which is a large gap, shows no similarities between commercial banks in the use of cost LogTA average value is 8.29%, according to which the standard deviation of up to 40.24% does not indicate the presence of similarities in terms of total assets among the Banks Policies and Sustainable Economic Development | 647 LOANTA average value is 56.14%, so the loan accounts for the majority of the total assets of the bank However, the range from the minimum value to the maximum value (36.23% - 71.00%) is relatively large, showing no similarities in the scale of customer loans among the banks ETA average value of 8.02% and standard deviation of 2.11% show the similarity of the ratio of equity to total assets of commercial banks listed on the stock Vietnam stock TK average value is 22.00%, while the minimum value is 6.41%, and the maximum value is 42.90%, so the very large range shows no similarities in liquidity among the banks NCA average value of 0.09% and standard deviation of 6.05% illustrate a very high level of similarity as seen in the increase in investments in the machinery and equipment and computer software by the joint-stock banks GDP average value of 5.88%, and a low range from 6.78% to 0.00% show the stability of economic growth between years 4.2 A correlation analysis Table Results of correlation analysis in the model Variable Y (ROE) Pearson correlation Y2 (ROE) X1 (TCTR) X2 (log(TA)) X3 (LOANTA) X4 (ETA) X5(TK) X6 (NCA) X7 (GDP) -.766** 434** -.127 -.431** 264 254 210 000 001 034 001 034 043 127 54 54 54 54 54 54 54 -.140 -.210 Sig N X1 (TCTR) Pearson correlation Sig N X2 (logTA) Pearson correlation Sig N X3 (LOANTA) ** -.766 000 ** 070 205 -.361 002 614 138 007 313 127 54 54 54 54 54 54 54 54 434** -.412** 489** -.609** -.092 119 -.055 001 002 000 000 506 390 690 54 54 54 54 54 54 54 137 -.078 54 ** -.127 070 489 Sig .360 614 000 54 54 54 Pearson correlation Sig N ** -.412 Pearson correlation N X4 (ETA) 54 -.431 ** ** ** -.135 -.738 329 000 322 573 54 54 54 54 54 -.014 043 -.106 918 760 446 54 54 54 205 -.609 -.135 001 138 000 329 54 54 54 54 54 648 | Policies and Sustainable Economic Development Y2 (ROE) X1 (TCTR) X2 (log(TA)) X3 (LOANTA) X4 (ETA) X5(TK) X6 (NCA) X7 (GDP) Pearson correlation 264 -.361** -.092 -.738** -.014 -.157 182 Sig .054 007 506 000 918 258 187 54 54 54 54 54 54 54 54 Pearson correlation 254 -.140 119 137 043 -.157 039 Sig .063 313 390 322 760 258 54 54 54 54 54 54 54 54 Pearson correlation 210 -.210 -.055 -.078 -.106 182 039 Sig .127 127 690 573 446 187 781 N 54 54 54 54 54 54 54 Variable X5(TK) N X6 (NCA) N X7 (GDP) 781 54 Source: data analysis using SPSS 20 Table shows the independent variables CI, logTA, LOANTA, ETA, TK, and NCA with Sig 5% 4.3 Assessing the suitability of the model R2 and adjusted R2 is used to evaluate the suitability of the model However, the adjusted R2, which is greater, indicates the better relevance of the model Table Evaluation of the relevance of the model Change Statistics R R Square Adjusted R Square Std Error of the Estimate R Square Change F Change df1 df2 Sig F Change DurbinWatson 857 734 693 3.9116178 734 18.103 46 000 1.920 Source: data analysis using SPSS 20 The results of analysis and evaluation of the relevance of the model show that the value of adjusted R2 is 69.3% 4.4 Testing for the suitability of the model The test is conducted with the following hypothesis of the relevance of the overall linear regression model: H0: βi = 0: Variables included in the model not affect the performance level H1: βi ≠ 0: Variables included in the model affect the degree of operational efficiency Policies and Sustainable Economic Development | 649 Table Results of model ANOVA analysis Model Sum of Squares df Mean Square F Sig 1938.886 276.984 18.103 000 703.835 46 15.301 2642.720 53 Regression Residual Total Source: data analysis using SPSS 20 According to ANOVA analysis, Sig = 0.000 implies that H0 should be rejected So, this model is suitable for analyzing the factors affecting ROE 4.5 The regression results Table Summary of the results of regression analysis Model Unstandardized Coefficients B Std Error (Constant) 32.978 21.752 X1 (TCTR) -30.059 4.312 Standardized Coefficients Collinearity Statistics t Sig Beta -.679 Tolerance VIF 1.516 136 -6.970 000 610 1.638 X2 (log(TA)) 2.383 2.509 136 950 347 283 3.530 X3 (LOANTA) -26.983 11.245 -.370 -2.399 021 243 4.114 X4 (ETA) -88.549 35.146 -.265 -2.520 015 525 1.904 X5(TK) -19.276 11.664 -.230 -1.653 105 299 3.347 X6 (NCA) 1953.206 933.045 167 2.093 042 906 1.104 X7 (GDP) 38.148 58.542 054 652 518 850 1.177 Source: data analysis using SPSS 20 Model results: ROE = 32,978 -0,679*TCTR -0,370*LOANTA -0,265*ETA+ 0,167*NCA 4.6 Discussion of research results Taken into account are the factors affecting the performance of listed banks drawn from the research results of the factors affecting ROE: the ratio of total operating expenses in total revenue, the ratio of loans to total assets, the ratio of equity to total assets inversely and proportionally impacting on ROE; however, the value of investment in machinery and equipment and computer software have a positive impact on ROE: β4 = -0.679