Institutions and credit risk in banking system: the case of emerging economies

In fact, there are numerous works focusing on the determinants of credit risk in banking system including microeconomic factors such as bank liquidity, bank capital, bank size, bank comp

Institutions and Credit Risk in Banking System:

The Case of Emerging Economies

SU DINH THANH

University of Economics HCMC - dinhthanh@ueh.edu.vn

NGUYEN PHUC CANH

University of Economics HCMC - canhnguyen@ueh.edu.vn

Abstract

The extant literature has documented the determinants of bank credit risk, but does not pay much attention

to institutions This study fills this gap by investigating the effects of institutions on bank credit risk Our data covers 33 emerging economies over the period of 2002 - 2013 Applying system GMM estimator for unbalanced panel data, the study finds intriguing findings Given credit supply, the effects of bank liquidity and bank profitability are negative, while the effects of bank capital and bank size and bank concentration are significantly positive On the other hand, in terms of credit demand the effect of real GDP growth rate is negative Our main finding is that institutions have negative effects on bank credit risk Particularly, in the present of institutions, the interaction effects of bank liquidity, bank capital and bank size, bank profitability, and real GDP growth rate are strengthened These results imply that improvements in institutions are significant for control over bank credit risk in emerging countries

Keywords: institution; banking system; credit risk; emerging economies

1 Introduction

There has been a growing literature on determinants of bank credit risk in the recent years, especially since the 2008 global financial crisis Several studies blame the crisis on excessive risk taking of banking system (Agnello & Sousa, 2012; Hoque, Andriosopoulos, Andriosopoulos, & Douady, 2015) A deep understanding of credit risk determinants is crucial importance for the proper evaluation of banking system risk and has a direct link with the development of suitable regulation and prudential tools

In fact, there are numerous works focusing on the determinants of credit risk in banking system including microeconomic factors such as bank liquidity, bank capital, bank size, bank competition, credit derivatives, internal rating systems, collateral, relationship between lender and customer (e.g., Imbierowicz & Rauch, 2014; Agarwal et al., 2016; De Lis et al., 2001; Mandala et al., 2012); and macroeconomic factors such as inflation, unemployment, house price, credit cycles, business cycle, and economic growth (e.g., Hoque et al., 2015; Castro, 2013; Chen, 2007; Jiménez et al., 2005; Tajik

et al., 2015; Marcucci & Quagliariello, 2009; Jiménez et al., 2005)

The credit activities of banking system are the function of the asymmetric information problem (Lindset et al., 2014; Miller, 2015; Neyer, 2004) This problem may be reduced if the institution of a country is improved The quality of institution and financial regulation are definitely determinants

of bank credit risk since the improvements in institution and financial regulation reduce the asymmetric information problems, thus induce banking system to provide bank loans with better quality On the other hand, financial insurance regulation may stimulate banks in taking more risks Indeed, the role of public institutions or public governance in reducing the overall risk in the economy is explored in several studies (Cohen et al., 1983; Dal Bó & Rossi, 2007; Dutta et al., 2013;

Ho & Michaely, 1988; Williamson, 1981) However, not much is known about public institutions in determining credit risk in banking system in emerging countries In addition, the associations between public institutions and other determinants of credit risk such as bank capital, bank liquidity, bank size, bank competition, and economic growth are ignored

Therefore, this study goes to fill this gap by investigating the effects of public institutions and the associations between them with bank characteristics including bank capital, bank liquidity, bank size, and bank competition, which reflect the determinants of bank credit supply, on bank credit risk This paper is not the first to explore the determinants of credit risk in banking system, but it is innovative

in several ways First, this is the first study using three dimensions of public institutions including

government effectiveness, regulatory quality, and rule of law to investigate the effects of institutional

quality on credit risk in banking system in emerging countries Second, we use interaction terms

between each institutional indicator and bank credit supply factors in order to examine the effect of credit supply’s determinants on bank credit risk when institutions are present In addition, we use interaction terms between each institutional indicator and real economic growth in order to examine the effect of bank credit demand factors when institutions are present These methodologies ensure

that our model catches all the main drivers of credit activities relating to credit risk in banking system

With these strategies, we believe that our study has significant contribution to both scholar and practice In fact, our empirical results show that institutional quality has significant negative effect

on credit risk in banking system, indicating that improvements in institutions (such as government effectiveness, rule of law, and regulatory quality) have good effects on banking system stability and reducing systematic risk This result has strong contribution to the literature of institution, where the institutional quality impacts not only on economic growth (e.g., Acemoglu & Robinson, 2008; Dollar & Kraay, 2003; Fatás & Mihov, 2005), but also on the stabilization of financial market Most notably, we find that improvements in government effectiveness and rule of law in line with the higher of liquidity, capital, size, competition, and lower of profitability in banking system increase bank credit risk The interaction terms between regulatory quality with bank credit supply characteristics have same effects as government effectiveness and rule of law excluding the profitability and competition in banking system, which have opposite impacts on bank credit risk More precisely, we find that the interaction terms between improvements in institutions with economic growth increase bank credit risk This results show the risk-taking activities of banking system with higher liquidity, capital, size, competition, and lower profitability in emerging economies

in the case of better government effectiveness and rule of law Better regulations also simulate banks taking more risks excluding banking system with lower profitability and higher competition, which means that better regulations help banking system more stabilization and more safety

The rest of the paper proceeds as following manner Section 2 reviews related bank credit risk determinants through the drivers of credit demand and credit supply Section 3 introduces methodology and data Section 4 presents our empirical evidences from three dimensions of institutions including government effectiveness, rule of law, and regulatory quality Final section concludes

2 Literature review

Many previous empirical studies analyze determinants of bank credit risk or nonperforming loans, including macroeconomic factors and specific banking sector characteristics (Castro, 2013) The bank credit risk is generally defined as the risk that a loan is not being paid by borrowers to bank partially or totally The analysis of bank credit risk determinants in banking system is essential for policy makers since it provides early alarms for macroeconomic management in front of shocks and preventing financial system from a possible crisis (Agnello et al., 2011; Agnello & Sousa, 2012; Beltratti & Stulz, 2012)

In the literature, bank credit risk is divided into systematic credit risk and unsystematic credit risk (e.g., Ahmad & Ariff, 2007; Aver, 2008; Frye et al., 2000; Dietsch & Petey, 2002) The drivers of

systematic credit risk include: (i) macroeconomic factors such as unemployment, economic growth,

inflation, and exchange rate; (ii) changes in economic policies such as monetary policy, fiscal policy, economic legislation changes, or trade policy, (iii) and political changes These factors may influence

the probability of borrowers paying their loans For instance, Aver (2008) finds that employment, long-term interest rates, and the value of stock market have important impacts on credit risk of the Slovenian banking loan portfolio Kattai (2010) and Fainstein and Novikov (2011) show that banking systems highlight the importance of economic growth in Estonia, Latvia, and Lithuania Salas and Saurina (2002) and Jakubík (2007) also point out that GDP growth and interest rates are the main macroeconomic factors affecting bank credit risk for the Spanish and Czech, respectively

The divers of unsystematic credit risk includes: (i) the factors related to borrowers such as personality, financial solvency, and capital of individuals; (ii) management, financial position,

sources of funds, and financial reporting, and specific factors of the industry sector of firms (Castro, 2013) Zribi and Boujelbène (2011) studies ten commercial banks in Tunisia over the period 1995-

2008 by estimating a panel model controlling for random effects find that ownership structure, prudential regulation of capital, profitability are main drivers of bank credit risk Furthermore, Ahmad and Ariff (2007) point out the importance of micro characteristics from commercial banks of some emerging and developed economies on the bank credit risk including regulatory capital and management quality Meanwhile, Jiménez and Saurina (2004) use data from several Spanish credit institutions to investigate the role of collateral, type of lender, bank-borrower relationship, and the effects of characteristics of the borrowers and of the loan on bank credit risk They find that collateralized loans have a higher probability of default, and that while loans granted by savings banks are riskier, a close bank-borrower relationship increases the willingness for risk-taking behavior of banks

Nevertheless, there is a multitude of empirical studies looking at main drivers of bank credit risk and highlighting that macroeconomic and microeconomic variables should be included into the analysis due to their considerable influence The vast majority of these empirical works consider the macroeconomic and microeconomic factors as the most important drivers in the determinants of bank credit risk based on a single country analysis Some provide a multi-country comparative analysis without concerning to the effects of institution and its associations with microeconomic factors and macroeconomic factors In this study, we inspect the determinants of credit risk in banking system under the aspects of credit supply and credit demand factors to better understanding the influence of institution on credit function and risk-taking behavior of overall banking system The institution of a country which is defined as the rules of the game in a society (North, 1990), includes three features: (i) “humanly devised” which contrasts with other economic fundamentals; (ii) “the rules of the game” to set “constraints” on human behavior; and (iii) their major effect will

be through incentives (see North, 1981; Acemoglu & Robinson, 2008) Several works have studies the effects of institution, which is named as the new institutional economics, but the effects of institution on credit risk in banking system are still ignored Therefore, we try to build our arguments

to explain for this issue by four directions

First, better institutional quality induces higher economic growth, and then reduces credit risk in

banking system In the literature of economic growth, institution is considered the differences in residuals of economic growth Previous studies agree that better institutional quality is positively and significantly correlated with economic growth (Young & Sheehan, 2014) For instance, Djankov et al (2002) finds that the lower cost of opening a medium-size business in the United States in comparison with Nigeria, Kenya, Ecuador, and Dominican Republic is highly correlated with economic growth Meanwhile, Beekman et al (2014) find that corruption reduces incentives of individuals in both voluntary contributions and investments in Liberia, and thus impacts on economic growth Therefore, better institutional quality will be in line with higher economic growth and hence improving the financial situation of borrowers and reducing credit risk in banking system

Second, better institutional quality reduces asymmetric information problems (Cohen et al., 1983;

Ho & Michaely, 1988; Williamson, 1981); hence, banks will be less probability of making wrong decisions in lending to bad borrowers

Third, better institutional quality reduces overall risk in the economy (Cohen et al., 1983; Ho &

Michaely, 1988; Williamson, 1981), thereby inducing a lower systematic risk and reducing credit risk

of bank credit portfolio In fact, better institutional quality helps economic agents more trustable in business transactions Thus, institutional quality positively impacts the efficiency of businesses (Dal

Bó & Rossi, 2007), and mitigates systematic risk in overall economy Dutta et al (2013) find that worse corruption situation leads to high inequality, poverty, and employment and thus undermines the effectiveness of economic growth in India

Four, better institutional quality reduces transaction cost in economic activities in general and

credit activities in particularly As a result, better institutional quality induces higher efficiency of credit activities and better control of credit risk in banking system

Although the views of systematic and unsystematic credit risk dominate in the literature of bank credit risk, we can examine the determinants of credit risk in banking system under the basic determinants on demand and supply sides of credit function The approach of the systematic and unsystematic credit risk is suitable for bank level studies due to defining factors relating to systematic

or unsystematic risk, but the approach of determining the drivers of credit risk from the views of supply and demand side in credit function is better at the banking system level or country level, since

it is easy to define the drivers of credit risk in the overall banking system

In the literature, the function of credit is impacted by supply and demand side factors On the demand side, income level and growth rate of GDP per capita are main determinants of credit demand (Backé & Wójcik, 2008) For instance, Duprey (2012) uses real GDP growth in explaining the different pattern of bank behaviors over macroeconomic fluctuations at 459 public banks in 93 countries Similarly, Elekdag and Han (2015) find that domestic factors such as economic growth and monetary policy shocks are more dominant than external factors in driving rapid credit growth in emerging Asia

Indeed, economic growth increases the expected income and profit in line with better financial conditions of private sector, thus allowing for higher levels of indebtedness (Kiss, et al., 2006), while households may want to increase debt levels to smooth consumption in current time since they expect higher income in the future (Backé & Wójcik, 2008) In addition, higher economic growth increases the disposable income of people and thus stimulates them in consumption and investment Firms are also induced to expand their operations to catch up with increasing demand, thereby stimulating the increased credit demand Even though the profit of firms may be higher due to economic growth, thus stimulating them to rely more on internal funds instead of bank loans (Kiss

et al., 2006), this effect may be not strong as the positive effects of economic growth on credit demand

in developing or emerging economies since high economic growth needs a bundle of capital

Accordingly, fluctuations in economic growth definitely change credit demand and then, in turn, impact on credit risk in banking system, which explains the relation between credit risk and business cycle (Marcucci & Quagliariello, 2009) For example, using lagged percentage change in GDP to explain for the banking crisis in the Nordic countries in the period of 1980s-1990s, Pesola (2001) finds that the shortfalls of GDP growth below forecast contribute to their banking crises Salas and Saurina (2002) finds that the GDP growth rate in line with other microeconomic factors such as firms, and family indebtedness, rapid past credit or branch expansion, inefficiency, portfolio composition, size, net interest margin, capital ratio, and market power are drivers of credit risk in Spanish commercial and savings banks in the period 1985-1997 Meyer and Yeager (2001) and Gambera (2000) find that macroeconomic variables are good predictors for the non-performing loans in the US Marcucci and Quagliariello (2008) use data from Italian banks and find that the credit risk is increased in economic downturns Similarly, Hoggarth et al (2005) provide the same evidence for the case of the UK

There are some possible explanations to the cyclicality of credit risk such as disaster myopia, optimism, herd behaviors, and insufficient market disciplines (Marcucci & Quagliariello, 2008) Marcucci and Quagliariello (2008) verify the cyclicality of credit risk from the following perspectives:

over-(i) economic growth increases the profit of firm, which raises asset prices rise and customers’

expectations at the beginning of expansionary phase and then increases aggregate demand As a result, the increasing of aggregate demand induces a rapid growth in bank credit portfolio and in economy's indebtedness, where banks usually underestimate their risk exposure and relax their credit standards due to over-optimism in the increasing of credit demand, which causes the

deterioration of borrowers’ creditworthiness; and (ii) customers’ profitability will be worsened when

an exogenous shock occurs The over-optimism is likely to become over-pessimism that can trigger the pitfall of asset prices and worsens customers’ financial wealth depressing deeper, while the downturn in asset prices worsens the value of collaterals of banking system Therefore, non-performing assets emerge, while firms’ financial distress increases, causing losses in banks’ balance sheets, and thus both banks’ profitability and capital adequacy deteriorate consequently as cyclicality

In fact, the effect of economic growth on credit risk in banking system follows the explanations of over-optimism, herd behaviors, and insufficient market disciplines, which depend much on institutional environment of a country Since better institutional quality reduces asymmetric information problems, transaction cost, and risk, while improving the efficiency of asset allocations, and property right protection (Cohen et al., 1983; Ho & Michaely, 1988; Williamson, 1981), better institution in association with higher economic growth may stretch cyclicality of credit risk in banking system under the impact of risk-taking behavior of banks

The literature of banking activities have concluded that there are many determinants of bank taking behavior such as capital requirement, bank size, bank leverage and financial liberalization (Ashraf et al., 2016; Bhagat et al., 2015; Blaško & Sinkey Jr., 2006; Borio & Zhu, 2012; Cubillas & González, 2014; Efing et al., 2015; Galloway et al., 1997; García-Marco & Robles-Fernández, 2008) Ashraf et al (2016) find strong evidence that bank risk-taking is significantly higher in countries, which have high individualism, low uncertainty avoidance, and low power distance cultural values Buch and DeLong (2008) examine cross-border bank mergers and find that the supervisory structures of the partners’ countries influence changes in post-merger total risk and strong host regulators limiting the risky activities of their local banks

risk-As a summary, the higher economic growth makes both people and banks more confidence and more optimistic, while better institutional quality boosts this confidence and optimistic higher due

to better quality of government, regulation, law system, property right protection, and also reduces corruption problem, therefore banks may take riskier activities in their credit portfolio Therefore,

we argue that better institutional quality in line with higher economic growth will simulate taking activities of banks, hence increasing the credit risk of overall banking system

risk-On the supply side, banks are assumed to make decisions on their credit portfolios depending on all available information and their internal financial conditions such as capital, liquidity, profitability, assets, therefore these internal characteristics of banks have strong impacts on credit risk The extent literature has examined the drivers of credit supply of banks (Auel & de Mendonça, 2011; Aysun, 2016; Bhaumik et al., 2011; Ciccarelli et al., 2015; Liu & Minford, 2014; Ramos-Tallada, 2015; Yagihashi, 2011) Marcucci and Quagliariello (2008), for instance, document that banks may react stronger to external shocks such as recession in economy by cutting their credit supply higher if they face to constraints such as capital in their balance sheet While, Bernanke and Gertler (1995) argue that banks can adjust their credit supply in responding with external shocks such as a contractionary

or expansionary monetary policy, where a contractionary monetary policy leads to a decreasing in available fund such as deposits or other funds, and thus banks have to reduce their credit portfolio Many empirical studies show that banks may not response to external shocks in the line of what the authorizers want in some cases For instance, banks may do risk-taking activities by increasing their credit supply despite contractionary monetary policy (Angeloni et al., 2015; Buch et al., 2014;

de Moraes et al., 2016; DellʼAriccia et al., 2014; García-Kuhnert et al., 2015; Montes & Peixoto, 2014)

Particularly, studies on credit channel argue that the actions of banks in their credit portfolio depend

on their internal conditions including capital, liquidity, size, profitability, and competition on banking sector (Aiyar et al., 2016; Aysun, 2016; Gunji & Yuan, 2010; Imbierowicz & Rauch, 2014; Khan et al., 2016; Leon, 2015; Yang & Shao, 2016)

There is usually an adequate requirement in the minimum capital ratio that every bank needs to meet Whenever, there is an external shock such as a downturn in economic growth or a contractionary monetary policy, banks with lower capital buffers above the minimum capital requirement will reduce their credit supply much more (Bliss & Kaufman, 2002) If firms cannot substitute bank loans by other sources, they may face to insufficient funding for their investment projects, and in turn, increasing the credit risk of banks Hence, banks with higher capital will be less credit risk But, banks with lower capital ratio have to raise additional capital before expanding credit

to bind risk-based capital requirement, thus they may optimally forgo profitable loans to reduce the risk of future capital inadequacy (Van den Heuvel, 2002) As a result, they may choose to reject bad projects for lending to reduce credit risk Konishi and Yasuda (2004) find that the implementation

of the capital adequacy requirement reduces risk taking at Japanese commercial banks, while banks with higher capital ratio may take riskier activities for higher profits, leading to increasing credit risk The liquidity is also very important for banking system due to the risk of bankruptcy (Diamond

& Rajan, 1999; Gatev et al., 2009; Imbierowicz & Rauch, 2014) In fact, banks have to consider their liquidity situation when they make credit decisions Banks with higher liquidity may be more flexible

in supplying their credit portfolio, thus they are better management their credit risk However, higher liquidity may simulate banks taking risk in their credit activities such as the studies of Nguyen and Boateng (2013) and Nguyen et al (2015)

The size of bank is usually mentioned as the main driver of credit risk (Demsetz & Strahan, 1997; Kishan & Opiela, 2000) At bank level, larger banks have higher market power and thus they have higher advantages in managing their credit portfolio, while smaller banks may take risk due to lower market share However, larger banks may be impacted by the theory “too big to fail” (O'hara & Shaw, 1990; Papachristos, 2011); thus, they take more risk-taking activities than smaller ones and have higher credit risk (Bhagat et al., 2015; Chen et al., 2015; Hakenes & Schnabel, 2011; Stever, 2007) At country level, larger banking system measuring under the relative size of assets in banking system with GDP defines the higher domination of banks in financial system and the more importance of their role in the economy, which creates more incentives for banks in taking risk due to the higher probability of government in intervening whenever they face to financial distress to protect people from panic as theory of “too big to fail”

At the last characteristic mentioned in our arguments, the profitability of banks measured by return on asset or return on equity may have goodness on bank credit risk The higher profitability

of banks makes them more flexible in managing their credit portfolio, and also reduce pressures in

expanding credit too quick so that the probability of wrong credit decisions or risk-taking activities will be limited

The competition in banking sector has strong impacts on bank choices and thus it has impacts on bank credit risk through the effects on bank risk-taking behaviors A large amount of literature argues that higher competition makes banks take more risk to aim at getting higher market share (Agoraki et al., 2011; Boyd & De Nicolo, 2005; Jiménez et al., 2013) Hou et al (2014), for example, find that intense market competition compels Chinese commercial banks to develop advanced technical experience and skills, thus improving their technical efficiency and then the technical efficiency is positive associated with the risk taking

However, countries with a few of large banks dominated reduce competitive pressures between banks When large banks have more confidence in taking risk and thus authorizers have to make sure they are safe (Boyd & De Nicolo, 2005; Hakenes & Schnabel, 2010; Jiménez et al., 2013; Wagner, 2010) Matutes and Vives (2000) study the links between competition for deposits and risk taking incentives, and conclude that the welfare performance of the market and the appropriateness of alternative regulatory measures depend on the degree of rivalry and the deposit insurance regime

As a summary, bank’s internal conditions such as liquidity, capital, size, profitability, and competition within banking sector have strong impacts on credit risk through credit activities Credit activities truly depend on bank’s manager expectations and financial behaviors The expectations are built from the available information that is faced to the problem of asymmetric information, while the financial behaviors of banks managers are impacted by the macroeconomic institution such as regulations and law system Therefore, institution definitely has associations with these drivers of bank credit risk Chen et al (2015) use bank-level data from more than 1200 banks in 35 emerging economies during the period 2000-2012 and find that the impact of monetary policy on banks' risk-taking behavior is more pronounced with the increasing severity of corruption da Silva and Divino (2013) use data for the Brazilian economy in the period from 1995 to 2009 and find that credit risk

is pro-cyclical and default risk depends on structural features

In the first view of associations between institution with bank characteristics and banking sector competition, better institution in line with higher capital, higher liquidity, bigger size, lower profitability, and lower competition may stimulate banking system in taking more risk due to lower asymmetric information problems, transaction cost, and overall risk in the economy

However, the difference aspects of institution including government effectiveness, regulation quality, and rule of law may have difference impacts Since government effectiveness reflects perceptions of the quality of public services and civil service, and the degree of its independence from political pressures in policy formulation and implementation, and the credibility of the government's commitment to such policies, so better government effectiveness may stimulate risk-taking behavior

in banking system due to better public services and thus lower transaction cost and better conditions for the growth in private sector Regulatory quality reflects perceptions of the ability of the

government to formulate and implement sound policies and regulations that permit and promote private sector development, thus better regulatory quality may induce some risk-taking activities of better banks due to the less harmful of policies from political pressures so that the “too big to fail” may be less impacts on bank’s managers Meanwhile, rule of law reflects perceptions of the extent to which agents have confidence in and abide by the rules of society, and in particular the quality of contract enforcement, property rights, and the police, therefore banks have more commitment in contract agreement and the property rights protections thus they may take risk more

3 Methodology and data

3.1 Methodology

This section presents our methodology in estimating the effects of institution and its associations with bank characteristics and banking sector competition through three step procedures First, we use the percentage change in government effectiveness indicators from the World Governance indicators to proxy for the change in institutional quality as following formula:

Since institutional quality indicators are scaled from -2.5 to +2.5 for each indicator, thus we standardize these indicators by calculating the percentage change of each indicator for better measuring changes in institutional quality This proxy of changes in institutional quality is used to evaluate the impacts of institution on credit risk in banking system in 33 emerging countries3 from

2002 to 2013 through the following modification

(3)

in which Crerisk is the ratio of non-performance loans to total outstanding loans which is used to proxy for credit risk in banking system; Inqua is the changes in institutional quality, including government effectiveness (Goveff), regulatry quality (Reguqua) and rule of law (Rulelaw); X is a set

of control variables including bank liquidity (Bankliq), bank capital (Bankcap), Bank asset (Bankasset), bank profitability (Bankroa), competition in banking sector (Bankcon), and real GDP growth rate (Gdpg) All definitions and calculations of variables are presented in Table 1

3 List of emerging economies: Brazil, Bulgaria, Chile, Colombia, China, Czech Republic, Egypt Arab Rep., Estonia, Greece, Hungary, India, Indonesia, Latvia, Lithuania, Malaysia, Mexico, Morocco, Nigeria, Oman, Pakistan, Peru, Philippines, Poland, Romania, Russian Federation, Slovenia, South Africa, Korea Rep., Thailand, Turkey, Ukraine, Venezuela RB, and Vietnam

Table 1

Definitions and sources of data

Crerisk The ratio of non-performing loans to total outstanding loans (%) GFDD

Bankliq The ratio of liquidity assets to deposit and short-term funding (%) GFDD

Bankasset The ratio of deposit money bank’s assets to GDP (%) GFDD

Goveff Percentage change in Government effectiveness indicator (%) Calculation from WGI Rulelaw Percentage change in Rule of law indicator (%) Calculation from WGI Reguqua Percentage change in Regulatory quality indicator (%) Calculation from WGI

Based on the work of Yurdakul (2014), the real GDP growth is used as the explanatory variable to proxy for the demand side of credit functions Other explanatory variables are based on the works of Ramos-Tallada (2015), Altunbas et al (2012), and Altunbas et al (2010) All the explanatory are lagged to capture data characteristic, where the year-end data to bank characteristics and bank competition will have impacts on bank’s manager decisions in following year The real GDP growth rate of this year will impact on the expectation of people in the future, and thus impacts on the next year credit demand

After examining the impacts of changes in government effectiveness on credit risk in banking system, we use interaction terms between institutions and the factors of credit demand side and credit supply side to evaluate interaction effects on credit risk in banking system In which, we add each interaction term one by one into our estimations to test the consistence of our results, which is proved for the robustness of our findings This is the second step in our study

Then, in the third step, we use changes in rule of law (Rulelaw) and regulatory quality (Reguqua),

respectively to replace for the government effectiveness and replicate all test procedures as mentioned above With this strategy, we believe our study is robustness in testing the impacts of institutions and interactions terms between institutions and the drivers of demand and supply on credit risk We use the system GMM estimators following the study of Arellano and Bond (1991), Arellano and Bover (1995), extended by Blundell and Bond (1998) and Blundell and Bond (1998) for advantage of addressing the bias associated with the fixed effects in short panels and solving the problem of endogeneity in dynamic panel data, which is suitable for this study due to the endogeneity emerging from the relationship between credit risk with other bank characteristics such as bank liquidity, bank profitability, or bank capital