Introduction. This chapter has introduced the necessity as well as the research objectives, scope and subjects of the research and the implementation process.

1.4. Research object and scope

According to the content presented in Section 1.2, previous studies have not fully assessed the factors affecting bad debt and analyzed the impact of bad debt on the operations of commercial banks on the same research sample. Therefore, this study exploits data from 34 Vietnamese commercial banks in the period 2005-2015, including state-owned commercial banks and joint-stock commercial banks. The sample size is 34 out of the total 35 joint-stock commercial banks, so this sample is representative of the group of commercial banks in Vietnam. At the same time, the thesis focuses on the main groups of subjects: (i) Group of macro factors affecting bad debt of Vietnamese commercial banks; (ii) Group of specific factors affecting bad debt of Vietnamese commercial banks;

(iii) Bad debt and its impact on efficiency, capital and credit growth of Vietnamese commercial banks.

Maybe you are interested!

Due to the limitation of data on bad debt variables, the thesis collected data from 34 commercial banks in the period 2005-2015. The thesis chose the time frame from 2005 as the period when the Vietnamese commercial banking system experienced a period of growth with many noteworthy events. This period has important significance for the Vietnamese banking system. This is a period of hot growth in quantity, capital, credit and bad debt began to increase. 2005 was also the year when the State Bank issued Decision 493/2005/QD-NHNN on promulgating regulations on debt classification, provisioning and use of reserves to handle credit risks in banking activities of credit institutions (hereinafter, the thesis will refer to Decision 493/2005/QD-NHNN).

1.5. Research method

Based on the problem of asymmetric information and moral hazard, theories on the relationship between macroeconomic factors as well as specific factors and bad debt, the hypotheses of Berger and De Young (1997) and previous empirical studies, the thesis builds a research model of factors such as bank efficiency, scale, credit growth, capital adequacy, level of owner control and macroeconomic factors such as economic growth, inflation, exchange rate, real estate prices.

real estate to bad debt. In addition, to seek empirical evidence on the long-term relationship of factors to bad debt, the study will add differences of the above variables based on the model of Klein (2013). At the same time, to assess the impact of bad debt on efficiency, capital safety and credit growth, the thesis will build a research model based on previous studies of Louzis et al. (2012), Le (2016) and Salas and Saurina (2013).

To determine the signs and magnitudes of the regression coefficients, the quantitative methods used in the thesis are estimation methods for regression models with panel data. First, the thesis uses a two-step system GMM generalized dynamic moment panel data to analyze the relationship between bank-specific factors and bad debt. Specifically, efficiency, capital adequacy and credit growth with the assumption that any change in specific factors is the cause of endogenous adjustment of other factors. In the model of factors affecting bad debt, the studies of Athanasoglou et al. (2008); Berger and De Young (1997); Karim et al. (2010) and Le (2016) all show that this is a two-way reciprocal relationship. Therefore, the model may have endogenous phenomena due to the reciprocal relationship between bad debt and efficiency, capital adequacy and credit growth. In that case, if using static panel estimates such as fixed effects FEM and REM, the results will be biased. Therefore, the thesis uses the GMM method to handle the endogeneity problem and lagged variables. At the same time, using the Sargan-Hansen test shows that the use of instrumental variables satisfies the endogeneity restriction condition of the model (overidentifying restrictions). The thesis uses the Stata 11.0 software support tool to perform the tests and estimate the regression coefficients of the variables in the model.

Models involving bank-specific factors often include endogenous variables. Karim et al. (2010) show that the endogeneity of bad debt when measuring the bad debt variable is closely correlated with the cost efficiency of the bank. Similarly, Le (2016) also shows the correlation between specific factors and bad debt. This leads to the problem of simultaneously making

Traditional estimators are biased. GMM estimation is an instrumental variable estimation technique and has several advantages over traditional estimators. Traditional estimators are inaccurate in the presence of heteroscedasticity. GMM estimation using moment conditions allows to produce accurate estimates even in the presence of cross-unit inconsistencies (Hansen, 2000). To check the robustness of the estimates, the study uses a two-step system GMM method developed for linear dynamic panel models (Arellano and Bond 1991; Arellano and Bover, 1995).

In addition, to measure the cost efficiency of commercial banks as a factor affecting the bad debt of commercial banks, the non-parametric data envelopment method DEA is used in the study. The DEA data envelopment method is a linear programming technique to examine how a bank performs compared to other banks in the sample. This technique creates a frontier set by efficient banks and compared to inefficient banks. The efficiency of banks ranges from 0 to 1, with a perfectly efficient bank having a result of 1. After collecting the data, the author uses the DEAP 2.1 software written by Coelli (1996) to estimate the cost efficiency.

The thesis collects secondary data on the situation of bad debt, specific factors and macro factors and market competition related to banking activities to assess the relationship between bad debt and related factors. The data used in the study is mainly secondary data, including: Consolidated financial statements and audited annual reports, reports on the situation of operations through documents of shareholders' meetings of 34 commercial banks and official data of the State Bank, General Statistics Office. The source of macro data collection is from statistics of international financial institutions such as IMF, WB and other official data sources.

1.6. New results and contributions of the research

Empirical research on Vietnamese commercial banks has produced the following outstanding results: (i) Improving bank efficiency, increasing capitalization, credit growth, economic growth, and controlling the level of market competition in the industry will help reduce bad debt; (ii) Reducing risk provisions, bank size, level of owner control, inflation, interest rates, and housing prices will reduce bad debt; (iii) The above-mentioned factors also have an impact on bad debt in the long term; (iv) Increasing bad debt negatively impacts bank efficiency, capital safety, and credit growth.

Compared with previous studies on the same topic that the thesis has referred to, the thesis has the following new contributions:

Firstly , the thesis for the first time tests the relationship between bad debt and cost efficiency of Vietnamese commercial banks. This inverse relationship shows that ineffective cost control is one of the important causes of bad debt of Vietnamese commercial banks. To improve their operational efficiency, commercial banks need to cut input costs, which will help to control loans more closely and reduce bad debt. At the same time, the thesis analyzes the causes of bad debt of Vietnamese commercial banks by quantitative methods from many perspectives: cost efficiency, profitability, capital, ownership level, industry competition, macroeconomic factors... in which the impact of bad debt lag is taken into account, estimated through the generalized dynamic moment panel data estimation model GMM.

Second , the thesis for the first time deeply studies the impact of bad debt on banking operations on a sample of Vietnamese commercial banks and shows how important bad debt is to business performance, cost efficiency, capital safety or credit growth. The important policy implication from this research result is that to increase banking efficiency, managers should strengthen supervision and monitoring of debt risks.

1.7. Research process and structure of the thesis

The research process is carried out as shown in Figure 1.1. Step one is to identify the research problem, step two is to review the theoretical framework and empirical results to identify research gaps, step three is the research methodology, step four is to collect data and quantitative analysis and finally discuss the results and make policy recommendations.

The content of the thesis is presented in 5 chapters, specifically as follows:

Chapter 1. Introduction. This chapter introduces the necessity as well as the research objectives, scope and objects of the research and the research implementation process.

Chapter 2. Theoretical framework and overview of previous studies on bad debt of commercial banks. In this chapter, the thesis presents a theoretical framework explaining the causes of bad debt and the impact of bad debt from previous documents. The main theoretical framework that the study relies on to explain the relationship between specific factors and bad debt is the theory of financial cash flow, bank lending channel, transmission channel of monetary policy and hypotheses such as moral hazard, poor management and scale effect. In addition, the thesis also reviews previous empirical studies to identify quantitative factors to build an empirical model of factors affecting bad debt as well as the impact of bad debt on banking operations in Vietnam.

Chapter 3. Research model, research method and research data. Based on the theoretical framework in Chapter 2, inheriting the empirical models of the above studies, this chapter will build the empirical models of the thesis, including: models of factors affecting bad debt and the impact of bad debt. The highlight of this chapter is the detailed presentation of the steps and estimation methods to find evidence for the research objectives of the thesis. Finally, the measurement of variables and data sources are also presented in detail in this chapter.

Chapter 4. Analysis of research results. Based on the empirical model and data collected from 34 Vietnamese commercial banks, the thesis uses Stata 11.0 software as a support tool to conduct tests and estimate regression coefficients of variables in the model. Then, discuss the empirical results based on the theoretical foundation of the research and compare them with previous studies to interpret the results logically. This result provides evidence to help answer the research questions of the thesis.

Chapter 5. Conclusion and solutions . This chapter summarizes the main empirical results associated with the research objectives of the thesis. From there, the thesis proposes a number of policy implications to control bad debt through influencing factors as well as limit the adverse impacts, if any, of bad debt on banking operations. These suggestions are expected to provide additional references for policy makers when implementing solutions to limit bad debt. At the same time, this chapter also recognizes a number of limitations that the thesis has not yet resolved. This is also the final chapter of the thesis.

CHAPTER 2

THEORETICAL FRAMEWORK AND OVERVIEW OF PREVIOUS STUDIES ON BAD DEBT OF COMMERCIAL BANKS

Introduce

The objective of the thesis is to evaluate the factors affecting bad debt in Vietnam and the impact of bad debt on banking operations. In particular, the thesis focuses on assessing the impact of specific factors and macro factors affecting bad debt, while also analyzing the impact of bad debt on banking operations, specifically efficiency, capital safety and credit growth.

Previous studies have shown that bad debt is a complex and inconsistent concept. The empirical results of these studies show that not only specific factors such as bank efficiency, scale, operational safety, financial capacity and credit growth, but also macro factors such as economic growth, inflation, exchange rates, real estate market and market competition have important impacts on credit quality. Therefore, Chapter 2 presents an overview of the conceptual framework of bad debt according to previous studies as well as organizations such as the International Monetary Fund (IMF), International Accounting Standards (IAS), Basel Committee, Vietnamese Accounting Standards (VAS ) . From there, the thesis also presents the conceptual framework and measurement of bad debt.

Thus, this chapter will present the theoretical framework that underpins this study, including the theory of financial cash flows, the theory of bank lending channels, the theory of balance sheet channels, the transmission channels of monetary policy and related hypotheses. In addition, empirical studies on factors affecting bad debt as well as the impact of bad debt on banking operations will be presented in detail in this chapter to serve as a foundation for building an empirical research model in the following chapter.

2.1. Theoretical framework

First of all, to understand the bad debt of commercial banks, it is necessary to understand the nature of the bank credit relationship and the issue of credit risk because bad debt is a category related to the credit relationship and credit risk. Credit is an economic category and a product of the commodity economy. The credit relationship is essentially an economic relationship associated with the process of creating and using credit funds, including goods and currencies, with the aim of satisfying temporary capital needs for production and life, according to the principle of repayment within a certain period of time. In particular, according to the Basel Committee on Banking Supervision (2006), credit risk is the risk of asset loss that may arise when a counterparty fails to fulfill its financial obligations or contractual obligations to a bank, including failure to make debt payments, whether principal or interest, when the debt is due. This risk is measured by the cost of obtaining replacement cash flows if the counterparty goes bankrupt (Jorion, 2009).

In Vietnam, “credit risk in banking activities is the potential loss of debt of credit institutions and foreign bank branches due to customers not performing or not being able to perform part or all of their obligations as committed”. This is the definition according to Circular No. 02/2013/TT-NHNN on classification of assets, provisioning levels, methods of setting up risk provisions and the use of provisions to handle risks in the operations of credit institutions and foreign bank branches.

2.1.1. Bad debt of commercial banks

According to previous studies by individuals and organizations, the concept and definition of bad debt are diverse and complex. It depends on the research objectives as well as the approach of the researcher. The term "bad debt" in English is bad debt, non-performing loan, doubtful debt, which are loans that begin to be classified as bad debt when the principal and interest are overdue for 90 days or more (Rose, 2009; Miskin, 2010). The world's definitions of bad debt are related to

Comment