Tóm tắt Luận án The effect of capital structure on sustainability and social performance of mfis in Vietnam

Harvey (2001), Flannery and Rangan (2006), Marsh (1982), Opler and Titman (1994), Taggart ( 1977), Hackbarth et al. (2007). On the contrary, Baker and Wurgler (2002, p. 3) argue against the theory of trade-offs and conclude "capital structure is the result of the effort over time in the stock market". In addition, Fama and French (1998), Dawar (2014), Daskalakis and Psillaki (2005), and Salim and Yadav (2012) also have counter-ideas to this theory. 14 1.5.3 Theory of pecking order theory 1.5.3.1 Theoretical content Contrary to the trade-off theory, the pecking order theory, which is introduced by Myers and Majluf (1984) denies the existence of a target debt ratio. However, this theory also agrees on building a priority selection for capital sources used by businesses. 1.5.3.2 Theoretical studies Supporters of this theory are researchers such as Remolona (1990), Baskin (1989), De Jong et al (2011). Also carried out research to check the suitability of pecking order theory, but Lery and Roberts (2010) found that this theory only explains about 20% of financial decisions of businesses, which is lower than the results given by De Jong et al (2011). Besides, a number of studies also show that the level of explanation of pecking order theory for the organizational capital structure is not high (Seifert and Gonenc, 2008; Helwege and Liang, 1996; Frank and Goyal, 2003; Graham and Harvey, 2001). 1.5.4. Capital structure characteristics of microfinance institution The difference in the mission and operation method has made the capital structure of microfinance institutions and enterprises clearly different. Firstly, microfinance institutions often receive more preferential capital sources than enterprises. Second, the capital structure of microfinance institutions always fluctuates based on changes in customer demand. Third, microfinance institutions have to comply with capital adequacy regulations, while businesses do not. 1.5.5 The mechanism of the impact of capital structure on social sustainability and efficiency Based on differences in capital structure characteristics of MFI and business‟ ones, the author finds that bankruptcy and agency costs do have little impact on the capital structure of microfinance institutions. In contrast, taxes and especially transaction costs have considerable impact on the relationship between capital structure, the sustainability and social performance of financial institutions. Therefore, the mechanism for impacting capital structure on the level of sustainability and social performance is built as follows: 15 Chart 1.3 Mechanism of capital structure impact on sustainability and social performance of microfinance institutions Source: Self-developed author The choice of different sources of funding determines the cost of capital that each microfinance institution must pay to maintain its operations. This cost will have a direct impact on the financial cost, and then on the sustainability of the organization. In addition, microfinance institutions have always been known to operate with the goal of helping to eradicate hunger and reduce poverty. Therefore, the use of capital sources with different costs is also intangible, creating a certain pressure on the managers and executives of the organization to ensure the balance between financial goals and social goals. As a result, the social performance of microfinance institutions will also be affected. Capital Structure Taxes, transaction costs Financial expenses Sustainability level Balancing between financial and social goals Social performance 16 CHAPTER 2. THE CURRENT SITUATION OF CAPITAL STRUCTURE, SUSTAINABILITY AND SOCIAL PERFORMANCE OF MICROFINANCE INSTITUTIONS IN VIETNAM 2.1. The process of establishment of microfinance institutions in Vietnam 2.2. The legal environment for microfinance activities in Vietnam 2.3. Microfinance institutions in Vietnam Microfinance providers in Vietnam are quite diverse. However, according to the definition in the concept of microfinance institutions, the thesis only focuses on analyzing the converted organizations licensed to operate under the credit institution law (including TYM, M&MFI, Thanh Hoa MFOs / MFIs, and CEP) and social funds, microfinance organizations/programs/projects that are registered and have annual statistics and reports to the State Bank. Also, due to being formed after microfinance institutions operating in other countries in Asia, MFIs in Vietnam are not only different in several service providers, but also, more importantly, there are distinct characteristics compared to MFIs in the peer group. First, the scale of Vietnamese MFIs is smaller than that of peers, and savings mobilization has not played a major role in microfinance institutions. By contrast, equity and debts make up the majority of the total capital of microfinance institutions in Vietnam. Besides, loans from private entities make up only a very small proportion. Second, the operating costs of MFIs in Vietnam are recognized as lower than MFIs in other countries, laying foundation for microfinance institutions in Vietnam to be easier to achieve profitability and maintain sustainability than microfinance institutions in other countries. Thirdly, the majority of MFIs Vietnam customers are women, showing that microfinance institutions in Vietnam are ensuring the pursuit of social performance. 2.4 The capital structure of microfinance institutions in Vietnam 2.4.1. The capital structure of licensed MFIs in Vietnam Based on data provided by the Microfinance Working Group in Vietnam, the author found that the capital structure of licensed microfinance institutions in Vietnam is quite diverse. In particular, licensed microfinance institutions have an impressive growth in the ratio of capital mobilized from deposits to total capital, and also affirmed that savings are 17 gradually becoming an important source of capital. On the contrary, due to legal problems, the ratio of borrowing capital in total capital has declined over time. 2.4.2 Current status of the capital structure of registered MFIs in Vietnam Compared with licensed MFIs, registered institutions have a less diverse complementary capital structure. Particularly, equity accounts for nearly 50% of the total capital of registered MFIs. Loans and mobilized capital accounts for about 21% - 28%. Finally, other capital accounts for a relatively modest proportion of 1% - 4%. 2.5 Sustainability of microfinance institutions in Vietnam 2.5.1 Current Situation of Sustainability of Licensed MFIs in Vietnam The OSS ratios of institutions are all higher than 100% reflecting the operating income of licensed microfinance institutions can cover the full operating costs. More importantly, the OSS has not witnessed a big change over the years. However, based on international standard, currently only OSS of TYM and CEP fulfil the requirements, which is higher than 120%. More importantly, the OSS of TYM is not much different from the standard threshold, OSS of CEP is not really stable and can fluctuate in the following years. Therefore, maintaining and improving this index in the coming time is essential to ensure the sustainable operation of licensed MFIs. 2.5.2. Current Situation of Sustainability of Registered MFIs in Vietnam The average OSS of registered organizations over the years has been 120% higher, and particularly somewhat higher than that of licensed organizations. However, when focusing on the OSS analysis of each registered organization, it can be seen that there is a huge difference in the sustainability between organizations. Moreover, the number of registered organizations which have OSS higher than 120% is declining. 2.6. Social performance of microfinance institutions in Vietnam 2.6.1 Access level (A) A comparison of the average number of borrowing clients of licensed and registered microfinance institutions shows that the accessibility of licensed institutions is higher than that of registered institutions. 2.6.2 The ratio of average outstanding loans to average income (B) A comparison of the change in average outstanding loans/ GNI per capita over the years of the licensed and registered organization can be seen that this ratio has not changed much over time. 2.6.3 Cost per borrower (C) 18 The average cost per borrower of a licensed organization is relatively higher than that of a registered organization. 2.6.4 The proportion of female customers (D) The proportion of female customers to the total number of borrowing customers of licensed and registered microfinance institutions always accounts for over 88%. 2.6.5 Number of branches (E) The number of branches in this study is considered to have very little variation; and therefore, will be excluded when calculating the SPI value. 2.6.6 The portfolio at risk (G) and write-off ratio (H) The percentage of the gross loan portfolio for all open loans that is overdue by more than 30 day and the average write-off rate of licensed organizations are always lower than that of registered organizations. 2.6.7 SPI index The Social Performance Index (SPI) of microfinance institutions in Vietnam in the period 2011 - 2017 tends to decrease over time. In particular, the comparison between SPI of registered microfinance institutions and SPI of licensed microfinance institutions show that the first one is better than the second one. 19 CHAPTER 3. ASSESSMENT OF THE IMPACT OF CAPITAL STRUCTURE ON SUSTAINABILITY AND SOCIAL PERFORMANCE OF MICROFINANCE INSTITUTIONS IN VIETNAM 3.1 Overview of econometric models to assess the impact of capital structure on the sustainability and social performance of microfinance institutions in Vietnam 3.1.1 The model for analysing the impact of capital structure on sustainability The first model: a model for data series in many countries. Figure 3.1: Models for data series in many countries The second model: the model for the data series in a country. Figure 3.2: Model for data series in a country The third model: the model only focuses on assessing the impact of capital structure on the sustainability of microfinance institutions. Figure 3.3: The model focuses on assessing the impact of capital structure 20 Through analyzing three models for measuring the impact of capital structure on the sustainability of MFIs, the author found that the second model - the model for the data series in a country is more suitable for this study. 3.1.2 The model for analyzing the impact of capital structure on social performance The first model: the model includes only the group of micro factors. Figure 3.4: The model includes only micro variables the second model: the model includes both based on both micro and macro variables. Figure 3.5: The model uses both micro and macro variables Through analyzing two models of measuring the impact of capital structure on the social performance of MFIs, the author found that the first model - an analysis model based on micro-factor groups is more appropriate because the study only focuses on microfinance institutions in a country is Vietnam. 3.2 Evaluation of the impact of capital structure on sustainability and social performance of microfinance institutions in Vietnam 3.2.1 Model description Based on the developed analytical framework, the impact of capital structure on the sustainability of microfinance institutions will be analyzed through the following model: OSS = β0 + β1* EAit + β2 *DPit +β3*BAit + β4*OAit + β5*Sizeit + β6*FBit + β7*OERit + β8*CPBit + β9*BPLOit + β10*PAR30it + β11*AGEit + β12*PSit + β13*RSit + β14*LMit + ε (1) Whereby: Β0: Intercept β1-14: estimated coefficients ε: error term 21 OSS: Sustainability level EA: The ratio of equity to total assets DP: The ratio of mobilized capital to total assets BA: The ratio of borrowings to total assets OA: The ratio of other debt to total assets Size: Gross loan portfolio of MFI FB: The ratio of female borrowers OER: The ratio of operating cost CPB: The ratio of cost per borrower BPLO: The ratio of borrowers per loan officer PAR 30: The ratio of portfolio at risk over 30 day AGE: The number of years in which organization operates PS: Profit status RS: Regulatory status LM: Lending method To answer research question 2, a multi-variable linear regression model was also used; in which, the dependent variable is SPI, the independent variable is equity over total assets (EA), capital raised from deposits on total assets (DP), borrowings to total assets (BA), other debt to total assets (OA), size (size), number of customers per credit officer (BC), profit stauts (PS), legal status (RS), lending method (LM), oerating years (AGE), lag SPI and lag PS. Specifically, the model has the following formula: SPI = β0 + β1* EAit + β2 *DPit +β3 *BAit + β4*OAit + β5*Sizeit + β6*BC it + β7*PSit + β8*RSit+ β9*LMit + β10*AGEit + β11*Lag SPIit + β12*Lag PSit € (2) β0: Intercept β1-12: Estimated coefficients €: error term 3.2.2 Description of the variables used in the model 3.2.2.1 Variables used in the model for analyzing the impact of capital structure on sustainability a) Dependent variable The dependent variable used in the model is OSS. b) Independent variables The sources forming the capital of MFIs (capital mobilized from deposits, borrowed capital, other debts, and equity) will be used as the proxy for capital structure. 22 Besides the group of capital structure variables, the group of variables on characteristics of microfinance institutions is also used in the model, including size; the percentage of female customers; cost per borrower; operating cost ratio; the number of customers per credit officer; the rate of risk balance over 30 days; operating period, the average yield on the total outstanding balance, legal status, lending method). Based on the theoretical basis of capital structure, results of previous studies, as well as the operating conditions of microfinance institutions in Vietnam, the author hypothesizes the impact of the variables on the dependent variable is as follows: Table 3.1. Hypothesis impacts of the independent variables on the dependent variable STT Impact variables Acronym Sign 1 The equity to total assets EA + 2 The mobilized capital to total assets DP + 3 The borrowings to total assets BA - 4 The other debt to total assets OA - 5 Gross loan portfolio Size + 6 The ratio of female borrowers FB + 7 Cost per borrower CPB - 8 The ratio of operating cost OER - 9 Borrowers per loan officer BPLO + 10 The portfolio at risk over 30 day PAR 30 - 11 Operating year AGE + 12 The average yield on the total outstanding balance PS + 13 Regulatory status RS + 14 Lending method LM + 3.2.2.2 Variables used in the model for analyzing the impact of capital structure on the social performance of microfinance institutions in Vietnam a) Dependent variable The dependent variable used in the model is the SPI which is formed by 7 variables named target of access (A), ratio of average outstanding loans to average income (B), cost per borrower (C), ratio of female clients (D), OSS (F), the ratio of portfolio at risk over 30 day (G) and write-off rate (H). 23 b) Independent variables The sources forming the capital of MFIs (capital mobilized from deposits, borrowed capital, other debts, and equity) will be used as the proxy for capital structure. Besides the group of capital structure variables, the group of variables about the characteristics of microfinance institutions is also used in the model including size, number of customers per credit officer, time of operation, the average yield on total outstanding, legal status and lending method. Based on the capital structure theory and the results of previous studies, the author hypothesizes the impact of the independent variables on the dependent variable as follows: Table 3.2. Hypothesis impacts of the independent variables on the dependent variable STT Inpact variables Acronym Sign 1 Equity to total assets EA + 2 Mobilized capital to total assets DP + 3 Borrowings to total assets BA - 4 Other debt to total assets OA - 5 Gross loan portfolio to total assets SIZE + 6 Borrowers per loan officer BPLO + 7 The average yield on the total outstanding balance PS - 8 Regulatory status RS - 9 Lending method LM + 10 Operating year AGE - 11 Lag PS LagPS + 12 Lag SPI LagSPI + 3.3 Data and research results 3.3.1 Descrition of the statistics 3.3.2 Research results of model for assessing the impact of capital structure on the sustainability of microfinance institutions in Vietnam 3.3.2.1 Regression results Table 3.5: Results of the REM regression model estimation 24 Variable Coefficient Std. Error t-Statistic Prob. C 472.1715*** 110.7141 4.264781 0.0000 EA 1.126835*** 0.244797 4.603134 0.0000 DP 1.127743*** 0.409753 2.752251 0.0066 OA 0.513946 1.152903 0.445784 0.6564 LOG_SIZE -7.324464 11.25622 -0.650704 0.5162 FB -3.126682*** 0.905044 -3.454729 0.0007 LOG_CPB -27.11789 25.25183 -1.073898 0.2845 OER -2.552263* 1.528775 -1.669483 0.0970 BPLO -0.005756 0.025502 -0.225721 0.8217 PAR_30 -7.101701 5.166675 -1.374520 0.1713 AGE 0.013726 1.313880 0.010447 0.9917 PS 1.460687*** 0.544646 2.681901 0.0081 RS -32.59415 29.70525 -1.097252 0.2742 LM -47.42429*** 18.01713 -2.632177 0.0093 Note: *,**,*** indicate significance at the 10%, 5% and 1% levels Source: Author's analysis from secondary data Table 3.6: Results of reduced model 1 Variable Coefficient Std. Error t-Statistic Prob. C 401.0163 79.47154 5.046036 0.0000 EA 1.207609*** 0.225602 5.352828 0.0000 DP 1.054746*** 0.304674 3.461884 0.0007 FB -3.312479*** 0.833216 -3.975535 0.0001 OER -3.087107*** 1.142183 -2.702811 0.0076 PS 1.181572** 0.460051 2.568351 0.0111 LM -31.30115** 15.07611 -2.076208 0.0394 Note: *,**,*** indicate significance at the 10%, 5% and 1% levels Source: Authors' analysis from secondary data The results of the regression model show that all variables have a significant impact on the sustainability of microfinance institutions in Vietnam at the significance level of 1% or 5%. In addition, the impact dimension of the variables is quite similar to the research hypothesis given in Section 3.2.1. Table 3.7: Comparison between hypothesis and model 1 research results Impact variables Hypothesis Research result Conclusion Equity to total assets (EA) + + Accept 25 Mobilized capital to total assets (DP) + + Accept Female borrowers (FB) + - Deny Operating expense ratio (OER) - - Accept Profit status (PS) + + Accept Lending method (LM) + - Deny Specifically, the variables of capital structure including equity over total assets and mobilized capital to total assets have a positive impact on the sustainability level, as well as consistent with the given hypothesis. In addition, average return on total outstanding loans also shows a positive impact on the sustainability of MFIs in Vietnam. On the contrary, the variable of the proportion of female customers and the lending method have a negative sign reflecting the inverse relationship with the level of sustainability. Therefore, in the next section, the explanations of the impact trends of the variables will be clarified. 3.3.2.2 Discussion of the results Equity to Total Assets (EA): The variable EA is positively related to OSS. This result is completely consistent with results in studies on the sustainability of microfinance institutions both in Vietnam (Dao Lan Phuong, 2019; Nguyen Quynh Phuong, 2017) and in other countries (Mwizarubi et al., 2015; Nyamsogoro, 2010). Specifically, the equity capital of microfinance institutions, especially registered ones, comes mainly from contributed capital sources and sponsored capital with very low costs. Therefore, the use of this capital will certainly help microfinance institutions to stabilize operations and create more profits. However, equity financing becoming increasingly declining, especially since Vietnam entered the ranks of middle-income countries. Therefore, the use of equity will be affected more or less in the future. Deposit-to-Total Assets (DP): The value of the regression coefficient is 1.054 showing that mobilized capital has important implications for the sustainability of microfinance institutions in Vietnam. This result is completely similar to conclusions from many researchers in the microfinance field such as Mwizarubi et al (2015); Iezza (2010); Muriu (2011) and Hossain and Asam (2016). Particularly, mobilized capital is considered to have a relatively low cost of capital; therefore, the increase in the ratio of mobilized capital to the total capital will help microfinance institutions in Vietnam save costs, thereby increasing profitability and sustainability. Kinde (2012) also argues that the use of mobilized capital also indirectly helps microfinance institutions strengthen strong customer relationships because they can attract and provide credit to customers at a more attractive and preferential interest rate. 26 The proportion of female customers (FB): The regression coefficient of this ratio is at -3.31, reflecting the inverse relationship with the OSS sustainability level. This result is somewhat different from the conclusion given by many researchers such as Nyamsogoro (2010), Ali (2013), and Muriu (2011) as many researchers all argue that female customers are guests with a very high repayment commitment. However, the inverse relationship between FB and OSS has also been confirmed in a previous studíe on the sustainability of microfinance institutions in Bangladesh (Hossain and Asam, 2016). In addition, in the Vietnamese context, female customers can be those with high repayment commitments; but, their power in the family is low. Moreover, the limited business capacity of female clients also is the barrier that makes loans for female customers is not working. Specifically, Le Thi Thu Huong (2016) in the study on difficulties and challenges of rural female workers pointed out four points to note, including (i) limited access to vocational training; (ii)an increasing in migration; (iii) limitations in improving social understanding and (iv) difficulties in integrating with the strong changes of socio-economic life. Besides, the amount of housework of rural women is also relatively large as most of them have to take care of housework, raise children, and take care of the elderly; therefore, the efficiency in using capital is not high. Operating cost ratio (OER): The variable OER has a regression coefficient of -3.08 reflecting the inverse relationship with the OSS sustainability level. This result is completely similar to conclusions made by many researchers such as Nyamsogoro (2010) Hossain and Asam (2016), Tehulu (2013), and Iezza (2010). Specifically, the authors argue that an increase in operating costs will reduce profits and thus negatively impact the sustainability of microfinance institutions. Besides, when operating costs increase too high, a part of this cost will be transferred to borrowers; therefore, it will also indirectly put pressure on borrowers and can affect the efficiency of capital use of customers in the future. Average Return on Total Debt (PS): The variable PS has a regression coefficient of 1.18, indicating a positive relationship with the OSS sustainability. This result is completely similar to the conclusion made by Iezza (2010). Lending method (LM): The variable LM has a regression coefficient of -31.3, reflecting the inverse relationship with the OSS sustainability. This result, according to the author, is due to the differences in customer characteristics when using different lending methods. Specifically, the introduction of group lending is often associated with disadvantaged customers in society - people with low incomes, relatively limited education, and no assets. accumulation. Therefore, group lending was born with the desire to use social constraints, and mutual assistance between-group borrowers to improve the repayment capacity of this group of customers. In contrast, individual lending is often applied to customers with relatively better qualifications and incomes. As a result, lending to 27 individual might help MFIs can be profitable and higher degree of sustainability. 3.3.3 Research results of