Corruption's impact on non-performing loans of banks in emerging markets: Empirical insights

Rehman, Abdul; Mehmood, Waqas; Al-smady, Ahnaf Ali; Arshian Sharif Article Corruption's impact on non-performing loans of banks in emerging markets: Empirical insights Research in Globalization Provided in Cooperation with: Elsevier Suggested Citation: Rehman, Abdul; Mehmood, Waqas; Al-smady, Ahnaf Ali; Arshian Sharif (2024) : Corruption's impact on non-performing loans of banks in emerging markets: Empirical insights, Research in Globalization, ISSN 2590-051X, Elsevier, Amsterdam, Vol. 9, pp. 1-9, https://doi.org/10.1016/j.resglo.2024.100241 This Version is available at: https://hdl.handle.net/10419/331167 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc/4.0/ Corruption’s impact on non-performing loans of banks in emerging markets: Empirical insights Abdul Rehman a , * , Waqas Mehmood b , Ahnaf Ali Alsmady c , Arshian Sharif d , e , f a Department of Management Sciences, National College of Business Administration and Economics, Rahim Yar Khan, Pakistan b Research Institute of Humanities &Social Sciences, University of Sharjah, United Arab Emirates c Accounting Department, Faculty of Business Administration, University of Tabuk, Tabuk, Saudi Arabia d Department of Economics and Finance, Sunway Business School, Sunway University, Subang Jaya, Malaysia e University of Economics and Human Sciences in Warsaw, Warsaw, Poland f College of International Studies, Korea University, Seoul, South Korea ARTICLE INFO Keywords: Control of Corruption Corruption Perception Index Non-performing Loans Banks Pakistan India and Bangladesh ABSTRACT Purpose: This study investigates the impact of control of corruption (COC) on nonperforming loans (NPLs) in South Asian countries. Methodology: The present study selected three South Asian countries (Pakistan, India, and Bangladesh) from 2000 to 2019. Further, the current research employed the fixed effect model (FEM) analysis on a sample of 81 conventional banks based in South Asian countries. Findings: The result reveals that control of corruption has a significant and negative association with NPLs, indicating that weak control of corruption would lead to an increase in NPLs in the sample countries. Implications: The findings provide insights to the policymakers in making strategic decisions about NPLs regarding control of corruption impact. Originality: The significant increase in NPLs is worrying as it may undermine banks’ability to grant credit and support a country’s economic recovery. The increasing trend of NPL rates and limited studies in related literature assessing NPLs in the South Asian region has motivated this study, which examines the external and internal factors that may influence NPLs. Introduction Financial institutions, especially banks, have various risk exposures such as liquidity, operational, market, and credit risks. Managing credit risk is among the essential practices of the financial industry because the value and exposure of credit risk are significant, given that it is directly linked to the failure of banks (Ghenimi, Chaibi, &Omri, 2017). For bank stability, careful credit portfolio management is essential because most earnings come from lending activities (Adzobu, Agbloyor, &Aboagye, 2017). Nonperforming loans (NPLs), 1 which are related to credit risk, arise when borrowers fail to repay loans to the bank. Banks’other common indicators of credit risks are loan loss provision (LLP) 2 and total loan (TL). The nonperforming loans to total loan (NPLs/TL) ratio is an important indicator as it influences the capability of grand new loans, level of capital provision and bank profitability (Pop, Cepoi, &Anghel, 2018). In addition, they also indicate that a higher level of NPLs might delay monetary policy transmission and reduce the efficiency of the financial system. The global financial crisis (GFC) of 2007–2008 triggered a rise in the value of credit risk in terms of NPLs for the banks in the global context (Ghosh, 2017b). With the significant increase in credit risk, many banks failed during the financial crisis period (Imbierowicz &Rauch, 2014). Such emerging issues in the banking industry have provided a pathway for discussing key indicators of NPLs (Ghosh, 2017b). In addition, Makri, Tsagkanos, and Bellas (2014) indicate that banks in developed economies have also faced the issue of increasing NPLs. Such an issue has triggered discussion in the current body of literature for European banks and other countries. In this respect, credit risk in terms of NPLs has been * Corresponding author. E-mail addresses: [email protected] (A. Rehman), [email protected] (W. Mehmood), [email protected] (A. Ali Alsmady), arshian. [email protected] (A. Sharif). 1 A loan is categorized as NPLs if principal payment and/or interest is 90 days or overdue. 2 Estimation for probable loan losses. Contents lists available at ScienceDirect Research in Globalization journal homepage: www.sciencedirect.com/journal/research-in-globalization https://doi.org/10.1016/j.resglo.2024.100241 Received 1 May 2024; Received in revised form 20 June 2024; Accepted 27 July 2024 Research in Globalization 9 (2024) 100241 Available online 30 July 2024 2590-051X/© 2024 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). examined by various studies (Acharya, Drechsler, &Schnabl, 2014; Imbierowicz &Rauch, 2014). Global trend of NPLs Globally, NPL trends indicate mixed ratios among the regions where the world as an indicator, on the whole, shows a downward trend from 2014 to 2017, as presented in Table 1. Regionally, the United States and Australia led the best performance since their NPL ratios declined consecutively from 2011 to 2017. The ratios were observed to be the lowest in Australia among all regions. Furthermore, the European Union, which had higher NPL ratios than Australia and the U.S., only experienced a downward trend from 2012 to 2017. On the other hand, South Asia showed an upward trend from 2011 to 2016, where its highest ratio in the region was 9.19 % in 2016 compared to the world’s highest of 4.16 % in 2014. Thus, Table 1, shows consistently high NPL ratios, peaking at 9.19 % in 2016, indicating significant financial stress, though there’s a slight improvement to 6.06 % by 2021. NPLs in South Asia Table 2 shows the NPL ratios of South Asian countries (Pakistan, India and Bangladesh) from 2007 to 2019. These three countries represent almost 70 % of the entire banking industry in South Asia. After the financial crisis, Pakistan faced a severe problem of increasing NPLs. In 2007, the NPL ratio of Pakistan was 7.60 % and showed an upward trend from 2007 to 2011, where its highest ratio was observed at 16.21 % in 2011. Similarly, India also faced the NPL crisis, wherein in 2009, the NPL ratio was only 2.20 %; however, in 2017, it jumped to 9.98 %. In Bangladesh, the NPL ratio was more than 10 % in 2017. Comparing NPLs in all three countries, Bangladesh had the highest NPL ratio in 2017, 10.41 %. On average, the NPL ratio of Pakistan, India and Bangladesh has increased from 7.83 % in 2007 to 8.90 % in 2019. The NPL ratio is regarded as one of the indicators of a banking crisis. Demirgüç-Kunt and Detragiache (1998) suggest that if the NPLs ratio in the banking sector in a country exceeds 10 % and the cost of its recovery is 2 % of the GDP, it may cause a crisis. A crisis in the banking sector may result in governments taking over the banks and customers withdrawing their money, which may cause a bank run. NPLs are more critical among these South Asian countries because their average NPL ratio is very close to the international benchmark of the banking crisis. The national benchmark for the NPL ratio is set at 5 % by the banking and finance industry in Pakistan, India and Bangladesh. If the ratio exceeds 5 %, it may cause potential difficulties in banks that jeopardise their business continuity (RBI, 2015; SBP, 2015); (Banagladesh Bank, 2014). However, the issue of NPLs is a significant challenge for banks worldwide because it affects the source of their interest income, especially in developing countries and emerging markets. A high value of the NPL ratio will burden a bank’s balance sheets and profits (Balgova, Nies, &Plekhanov, 2016). In addition, NPL is an essential predictor of bank failure (Balgova et al., 2016). A high NPL ratio is not only an indication of poor banks’ asset quality but also an indisputable measure of a bank’s performance. This also affects banks’capacity to honour their customers’deposit obligations when demanded or due, causing liquidity crises in the banking industry and the whole economy (Ghosh, 2017a). This is because banks use depositors’money to grant loans to the borrowers. The GFC 2007–2008 witnessed an increase in the NPLs ratios of banks worldwide (Ghosh, 2017a). With the significant rise in NPLs, many banks failed during the 2008 financial crisis (Imbierowicz &Rauch, 2014). Hence, in the aftermath of the 2008 GFC, countries focused more on managing NPLs and developed strategies to reduce the problem. In 2017, the average NPLs ratio in the world was 3.45 %, but in South Asia, the ratio reached 8.43 %. NPL issues have triggered academic discussion, where existing studies on NPLs examine issues such as credit growth and NPLs (Peric & Konjusak, 2017), oil price movements and NPLs (Al-Khazali &Mirzaei, 2017), external deficit and NPLs (Kauko, 2012), carbon intensity loans (CIL) and NPLs (Guan, Zheng, Hu, Fang, &Ren, 2017), industry-specific, regional economic determinants and NPLs (Ghosh, 2015), credit cycle, business cycle and NPLs (Anastasiou, 2017), regulatory capital and NPLs (Osei-Assibey &Asenso, 2015), and bank efficiency and NPLs (Abd Karim, Sok, &Hassan, 2010). In general, the findings of these studies conclude that oil and gas price changes, lending interest rate, exchange rate, unemployment, share price, ROA, loan loss provisions, inefficiency, size, credit growth, budget deficit, credit and business cycle, the carbon intensity of loans, and leverage are all significant factors of banks’NPLs. However, the issue of corruption and NPL has received less attention from previous studies as corruption may distort the distribution of funds from good to poor projects and will affect NPL. Concerning this, moral hazard theory states that banks are difficult to identify between bad (high-risk) borrowers and good (low-risk) borrowers due to information asymmetry, resulting in the development of adverse selection problems. The higher levels of corruption in one country due to moral hazard may generate problems with non-performing bank loans. Empirical studies by Chen, Jeon, Wang, and Wu (2015) highlight that higher corruption levels will increase the banks’NPLs. In this respect, corruption can be defined as dishonest behaviour by those in positions of power, such as managers or government officials, that include giving or accepting bribes or inappropriate gifts, double-dealing, under-the-table transactions, diverting funds, and money laundering. In general, countries with high levels of corruption offer more opportunities to distort the credit assessment processes and may channel banks’loans to ineligible applicants. Concerning this, Pakistan has faced a significant issue of corruption for a long time. The types of corruption include bribes, nepotism, irregular payment in return for favourable judgements, and misuse of authority for personal use. Because of all these factors, Pakistan is ranked 140 out of 180 as the most corrupt country in the world (Transparency International, 2022). India is also facing a significant issue of corruption, like Pakistan, where the types of corruption in India include bureaucracy, embezzlement, fraud, and bribery. This has been reported that more than 62 per cent of Indians have paid a bribe to a public official in their lives (Transparency International, 2014). Bangladesh also has an issue of corruption, and the common types of corruption in Bangladesh are inappropriate use of government funds, excessive lobbying, pilferage, and irresponsible conduct from government officials. Due to this, Bangladesh is ranked 147 out of 180 as the most corrupt country in the world (Transparency International, 2022). Nevertheless, limited studies empirically examine corruption’s effect on NPLs in Pakistan, India, and Bangladesh. Although there is a study by Ahmad (2013) on corruption and NPL in Pakistan, the study, however, uses the “Corruption Perception Index”(CPI) for the measurement of corruption. The limitation of CPI is that this index measures corruption in the public sector only, which may not be accurate enough to reflect all data about corruption. Conversely, the current study uses the Control of Corruption (COC) data from the Worldwide Governance Indicator (WGI) covering corruption measurement for private and public sectors via interviews with experts and opinion polls. In addition, WGI is better because it measures the COC factor using 11 data sources from the CPI and 14 other data sources not used in the CPI (Rohwer, 2009). Additionally, other studies focus on Central and Eastern Europe (CEE) (Toader et al, 2017), Mexico, Indonesia, Nigeria, and Turkey (MINT) (Morakinyo &Sibanda, 2016) and Association of Southeast Asian Nations (ASEAN) countries (Ovi, Perera, &Colombage 2014). Therefore, the current study will focus on Pakistan, India and Bangladesh, countries in the South Asian region where the corruption level is alarming. Thus, the present study’s aim is •To investigate the impact of control of corruption on the NPLs of banks in South Asian countries. A. Rehman et al. Research in Globalization 9 (2024) 100241 2 The present research attempts to fill the prevailing gap in NPL literature by investigating the effect of corruption on NPLs of banks in South Asia, where these factors are not extensively investigated. The significance of the study can be observed with practical and theoretical contributions. The practical significance will benefit the regulatory authorities, financial analysts, policymakers, and practitioners by assisting them in creating a sustainable framework for bank loan quality portfolios. Furthermore, the findings of this study will also guide policymakers in formulating more sound and effective banking policies based on the stipulated policy implications of the study. In addition, it will help in identifying critical vulnerabilities and formulation of policies that will strengthen general financial stability. Literature review Corruption and nonperforming loans (NPLs) The misuse of authority for personal benefit is widely known as corruption. It is a global political, social, and economic phenomenon more prevalent in developing economies. Corruption is also established in conspiracy, nepotism, embezzlement, fraud, cronyism, deceit, the misuse of public resources, and other related practices besides extortion and bribery, which generally characterise corruption (Chen et al., 2015; Mohd-Rashid, Mehmood, Ooi, Che Man, &Ong, 2023). The complicated social concept of corruption has serious adverse effects on the development of society and the economy. Researchers of politics, sociology, and history have long focused on this topic (Dimant &Tosato, 2017). There are two contradictory opinions; one is grease the wheel, and the other is sand the wheel. The prior opinion reveals that an increase in corruption leads to the rise in investments in the banks and businesses through illegal money that increases the wealth of the ordinary person, as he wants to invest it for more profit, which reduces the NPLs and increases the economic growth (Friedrich, 1972; Huntington, 1968). Conversely, the other opinion reveals that improper implementation of rules and regulations causes corruption and bribery for legal work, which decreases the efficiency of institutional quality, ceases economic growth and increased NPLs (Barreto, 1996; Tanzi &Davoodi, 1997; Vito, 1998). Some authors found that corruption has a positive effect on NPL (Bougatef (2016)Chen et al. (2015); Goel and Hasan (2011), and some found negative results (Lee, Yahya, Habibullah, &Ashhari, 2019) Toader, Onofrei, Popescu, and Andrieș (2017) Ovi, Perera, and Colombage (2014), while some found insignificant effect on NPLs (Ahmad (2013), Ozili (2018). The starting argument for corruption and bank-lending can be viewed from the study of (LaPorta, Lopez-de-Silanes, Shleifer, &Vishny, 1998). They presented a theoretical assumption under the law and finance theory, explaining that financially strong institutions provide the loan facility to complete the loan contracts. Such enforcement of the legal contracts forces the banks to offer more loans, and meanwhile, they have to impose the screening against those who are not paying the loan amount on time. However, in those economies where corruption is a critical issue, banks are not sure to impose and fulfil the contracts regarding the recovery of the loans (Becker &Stigler, 1974; LaPorta et al., 1998). Such an uncertain situation leads to low loan recovery, more loan reduction and increasing credit risk in the banking sector. Various earlier studies considered the influence of corruption on NPLs. For instance, Goel and Hasan (2011) examined the impact of corruption and NPL using a sample of 60 countries in 2007. Seven institutional quality measures were used, including the CPI indicator of corruption, the autonomy of the central bank, European Union membership, and the development of the financial sector. The results showed that higher levels of NPLs were found in countries that are more corrupt. In addition, countries which are enjoying higher growth rates have lower bad loans. Park (2012) examined the influence of corruption on economic growth and banking industry soundness as calculated by nonperforming loans (NPLs) utilising data from more than 70 countries, covering the period from 2002 to 2004. The findings suggested that corruption worsens the issue of nonperforming loans (NPLs) in the banking industry. Furthermore, corruption distorts the distribution of fund from good to poor projects and increase NPLs, lowering the standard of private investment and, as a result, lowering economic growth. A study was conducted by Bougatef (2016) on how corruption affected loan portfolio (NPL) quality in 22 developing nations between 2008 and 2012. The study’s findings showed that corruption worsens the issue of NPLs. Furthermore, corruption can hinder economic development because of the misallocation of funds in the economy. When analysing the subsample, the entire sample was subdivided to create two groups based on the median involving highly corrupt and less corrupt nations. High corrupt group countries include Egypt, Indonesia, India, Mexico, Philippines, Russia, Thailand, Colombia, Peru, China, Brazil and Greece. Subsample analysis of corruption shows that the impact of corruption on NPL is more in highly corrupt group countries. Similar to Bougatef (2016) and Chen et al. (2015) examined the influence of corruption on risk-taking behavior as calculated by the z-score and bank NPL in 35 developing countries. The period of the study was from 2000 to 2012. The findings showed that when the level of corruption increases, the stability of banks declines. Additionally, a bank experiences greater vulnerability if they become embroiled in high-risk schemes in economies where corruption is widespread. Furthermore, high levels of corruption cause banks to adopt risk-taking behavior concerning NPLs rather than adopt a “sanding the wheel”stance in the nexus of corruption-development. Corruption increases the volatility in banks’returns and raises the probability of borrowers’default. In the context of Central and Eastern European (CEE) countries, Toader et al. (2017) examined how the stability of banks (z-score, NPLs, and credit growth) was influenced by corruption between 2005 and 2012. Their findings showed that the strength of banks was positively impacted by COC, resulting in a reduction in credit losses from NPLs and credit growing moderately. Furthermore, these results emphasised the importance of the nation’s attributes and the bank’s status in shaping corruption’s asymmetrical effect on the bank’s stability. Son, Liem and Khuong (2020) evaluate the influence of corruption on the banking sector and economic development using aggregate data encompassing 120 countries from 2004 to 2017. The findings of 3SLS regressions demonstrate that the association between corruption and NPL was positive, degrading the banking sector’s soundness. Therefore, the financial system becomes weak, which confirms the “sand the wheels” argument. Jenkins, Alshareef, and Mohamad (2021) study the effects of nationwide corruption on the credit risk of commercial banks with varying degrees of credit risk. They employ the quantile regression (QR) estimate approach for panel data of 191 commercial banks, from 18 MENAP nations, from 2011 to 2018. The study result reveals that corruption considerably magnifies the issue of bad loans from banks. For economies like Mexico, Indonesia, Nigeria and Turkey (MINT), Morakinyo and Sibanda (2016) examined the determinants of NPL for the banks in these countries from 1998 to 2014. The results reported that corruption positively influences bank NPLs in these MINT countries. Increased corruption is the cause of rising NPLs in the banks. Therefore, it is suggested that institutions should be improved to eliminate corruption and promote transparency. Murdock, Ngo, and Richie (2023) examine the impact of governmental corruption on the performance and risk of financial enterprises based in the United States. They indicate that a corrupt environment is linked to decreased bank performance without a concurrent decrease in risk. Major financial institutions often misjudge the rise in credit risk. Small—and medium-size banks aim for higher profits in corrupt regions by decreasing their liquidity. The study conducted by Abuzayed, Ammar, Molyneux, and Al- Fayoumi (2024) analyses a dataset of 7235 banks from 160 countries spanning the years 2000 to 2016. They explored the relationship A. Rehman et al. Research in Globalization 9 (2024) 100241 3 between corruption, lending practices, and bank performance. The findings reveal that corruption has a positive effect on bank lending, but it negatively affects bank earnings and raises risks such as credit, solvency, and distance to default. Corruption at the level of banks significantly impacts the performance of banks in both developed and developing nations, whereas corruption at the level of the country has a comparatively less impact on lending in developing countries. The research further reveals that increased bank competition, market concentration, and enhanced regulatory regimes mitigate the impact of corruption on bank lending and performance. Jiang and Wang (2024) discovered that banks provide superior loan conditions to borrowers in nations characterised by higher levels of lending corruption. The strength of this relationship is enhanced when borrowers face limitations in obtaining funding but weakened in nations with more stringent oversight of foreign bank ownership or higher levels of religiosity. Additionally, we also found that banks located in nations with elevated levels of corruption in lending exhibit inferior loan quality and earnings performance, making them more vulnerable to difficulties in times of financial crises. Jungo, Madaleno, and Botelho (2024) investigate the impact of financial inclusion and institutional determinants, such as corruption and the rule of law, on the credit risk and stability of banks. The findings validate that enhancing financial inclusion and bolstering the rule of law may make a substantial contribution to mitigating credit risk and enhancing the financial stability of banks. Conversely, the authors observe that inadequate control of corruption exacerbates credit risk. Furthermore, they discovered that heightened rivalry within the banking industry leads to an increase in credit risk. Ali, Sohail, Khan, and Puah (2019) examine the influence of liquidity risk, credit risk, financing risk, and corruption on the stability of the banking sector in Pakistan. The results of this research indicate that the size of a bank, liquidity risk, financing risk, and corruption have a favourable influence on the stability of the bank. In addition, they found a negative correlation between credit risk and bank stability. Lisbinski and Burnquist (2024) examine the impact of institutional features on the level of financial development in both developed and developing nations. The findings from the comprehensive study indicate that welldefined and stable institutions have a crucial role in promoting financial growth. Sol Murta and Gama (2023) investigated how the way corruption is seen at the national level affects the lending activity of commercial banks in Europe. They specifically focused on the significance of loans and the quality of loan portfolios. The findings indicate that corruption has a detrimental effect on the significance of loans in bank assets and a favourable effect on the percentage of non-performing loans. Furthermore, trade openness amplifies the significance of loans and the prevalence of nonperforming loans. Bank lending activity is influenced by factors such as the size of the bank, its capital, and the level of risk. Tran (2022) examines the correlation between corruption and corporate risktaking in developing nations, where corruption is often seen as the foremost risk to public welfare. The findings reveal a detrimental impact of corruption on corporate risk-taking. In summary, empirical findings from previous studies indicate that corruption is an important indicator which creates the problem of NPLs. Corruption affects both the demand and supply sides of the banking and financial market. Corruption increases banks’risk as funds are misallocated from good to destructive projects, eventually reducing economic growth. In addition, previous studies showed that NPLs are high in corrupt countries. Thus, this motivates this study to investigate further whether corruption will significantly impact NPLs in South Asian countries. Although there have been extensive studies establishing a connection between corruption and non-performing loans worldwide, there is a significant lack of studies that explicitly examine South Asian nations. Prior studies have mostly investigated this correlation in more extensive settings or within certain areas such as Europe or emerging markets. Nevertheless, the unique economic, political, and institutional dynamics of South Asia have not been thoroughly investigated in regard to this issue. This gap highlights the need for focused research to comprehend the impact of corruption on non-performing loans in banking systems of South Asian countries, while considering the particular challenges and circumstances of the region. Research methodology For the dependent variable (NPLs) and bank-specific control variables, data were collected from the DataStream database, central bank websites and banks’annual reports. Data was obtained from the World Bank and Worldwide Governance Indicators (WGI) database for corruption. Meanwhile, data were collected from the World Bank database for macroeconomic variables. The present study used panel data from 2000 to 2019. This study utilises a sample of conventional banks from three countries in the South Asian region: namely, Pakistan, India and Bangladesh. This zone provides a unique situation to explore how NPLs were affected by corrupt practices, country governance, and concentrated loan portfolios because 2017 saw the average ratio of NPLs in the South Asian region reach 8.43 %. The severity of the NPLs in South Asia is worth investigating as it may expose banks vulnerability to financial crisis and could adversely affect the banks’financial stability, which hampers the good borrowers from banks (Bernanke, Lown, &Friedman, 1991; Kanas &Molyneux, 2017). Furthermore, in South Asian countries, corruption is more prevalent in Pakistan, India and Bangladesh (GCI, 2017). 3 Concerning this, all countries in the sample share the same corruption and country governance issues: higher corruption and bad country governance; As for loan portfolio concentration, these countries heavily focus on giving out loans to sectors such as textile, industry, and commercial sectors and ignore other economic sectors. Pakistan, India and Bangladesh represent almost 70 % of the total banking industry in South Asia. These three countries were colonies of the British Empire before 1947, and these countries are homogeneous in nature because of their culture and same historical background, and this explains why these three countries are selected in this study. Currently, there are 34 banks under the supervision of State Bank of Pakistan, 86 banks under the supervision of Reserve Table 1 NPL ratios across the region (%). Region 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 World 2.66 3.05 3.51 3.47 4.05 4.24 4.34 4.04 3.75 3.55 2.94 South Asia 4.70 4.96 5.58 7.78 8.40 9.19 8.43 7.14 7.27 6.61 6.06 European Union 6.01 7.48 6.41 5.48 4.91 4.25 3.72 3.81 3.22 2.63 1.85 Australia 1.97 1.75 1.40 1.05 0.92 0.98 0.89 0.90 0.96 1.10 0.91 United States 3.78 3.32 2.45 1.85 1.47 1.32 1.13 0.91 0.86 1.07 0.81 Source: World Bank Database. 3 Global Competitiveness Index (GCI) 2017–2018, analyse 114 indicators of 12 pillars of the economy. Most problematic factor for doing business is corruption in Pakistan, India and Bangladesh. A. Rehman et al. Research in Globalization 9 (2024) 100241 4 Bank of India and 59 banks under the supervision of Bangladesh Bank. The final sample contains 81 banks; 18 in Pakistan, 40 in India and 23 in Bangladesh. Variable and measurement Dependent Variable: Nonperforming loans (NPLs) is measured by the ratio of nonperforming loans to total gross loans, which is in line with the existing studies such as Al-Khazali and Mirzaei (2017) and Guan et al. (2017). NPL is associated with the quality of individual assets and the likelihood of default by borrowers (Koch &MacDonald, 2010). Nonperforming loans, according to Berger and DeYoung (1997), have been generally agreed as the most acceptable proxy of problem loans. Independent variable: The control of corruption (COC) is the only independent variable in this study. Following past studies such as, Mehmood, Mohd-Rashid, Ahmad, &Tajuddin, (2023), Mehmood, Mohy Ul Din, Aman-Ullah, Khan, &Fareed, (2023), Mehmood, Mohd-Rashid, Khalid, Aman-Ullah, &Abbas, (2022), this study uses the country-wide corruption levels to measure COC. COC refers to “capturing perceptions of the extent to which public power is exercised for private gain, including both petty and grand forms of corruption, as well as capture of the state by elites and private interests.”(Kaufmann, Kraay, &Mastruzzi, 2009). The value of COC estimates ranges from −2.5 to 2.5, where −2.5 indicates the lowest level of COC (high levels of corruption), while 2.5 highlights the maximum level of COC (low levels of corruption). The previous studies suggest that higher COC values (lower corruption levels) will decrease bank NPLs. Control Variables: Present study used bank-specific and macroeconomics factors as control variables. These control variables are bank size, profitability, capitalisation, income diversification, inefficiency, GDP growth rate, inflation and interest rate. Bank size (SIZE): The bank size is determined by using the “natural logarithm of total assets”(Ben Saada, 2018; Chaibi &Ftiti, 2015; Tarchouna, Jarraya, &Bouri, 2017). Bank size is perceived to negatively influence nonperforming loans as it is believed that larger banks have better resources and capabilities to recover their loans. Larger banks are more likely to employ more skilled individuals and market power and to use economies of scale (Adhikary, 2006). Profitability (PROFIT): Peric and Konjusak (2017) utilised “return on assets”(ROA) to measure the bank’s performance; it is evaluated by “EBIT to total assets”. Furthermore, Peric and Konjusak (2017) indicate that ROA has a negatively significant impact on NPLs. The higher levels of the bank’s earnings may lead to lower NPL and vice versa. Contrary to this, good governance should lead to lower credit risk or lower NPL (Dimitrios, Helen, &Mike, 2016). This implies that ROA has a negative association with NPL. Capitalisation (CAP): Saif-Alyousfi, Saha, and Md-Rus (2018) use the capitalisation variable and measure by equity capital to total assets. Saif- Alyousfi et al. (2018) find that capitalisation negatively impacts the Banks’NPLs. The managerial staff of banks faces moral hazards with less capital (low capitalisation) to involve in risky financing (Keeton & Morris, 1987). This ‘moral hazard’hypothesis implies an inverse relationship between capitalisation and NPLs. Income Diversification (INCD): Several prior studies use the income diversification variable that is analysed by the ratio of non-interest income to total income (Tarchouna et al., 2017). This reflects the bank’s dependence on various types of income other than interest income. Tarchouna et al. (2017) find that INCD negatively effect on NPLs of banks. Inefficiency (INEFF): Saif-Alyousfi et al. (2018) use an inefficiency variable measured by operating expenses to total income. Saif-Alyousfi et al. (2018) find that inefficiency has a positively significant influence on bank NPLs and gives support to the “bad management”hypothesis. High-cost inefficiency (bad management) may result in the rise of the NPL. GDP Growth (GDP): The GDP growth variable is an economic condition indicator and is measured by the annual growth rate of GDP. Alhassan, Coleman, and Andoh (2014) suggest that the GDP growth rate negatively affects NPLs. NPLs of the banks decrease during times of high economic growth. Inflation (INF): Pop et al. (2018) use an inflation variable measured by the annual inflation rate. In addition, Dimitrios et al. (2016) express that inflation rate can determine the extent of country’s NPL because it improves loan repayment abilities of the borrowing customers and makes loans cheaper. Interest Rate (IR): The real interest rate calculated as the difference between the long-term and inflation rates has been used in previous studies (Castro, 2013; Chaibi & Ftiti, 2015). According to Castro (2013), interest rates affect NPLs positively. A rise in interest rates would weaken a borrower’s ability to repay debt service, leading to an increase in the value of NPLs. Empirical model The present study used static panel estimation to test the hypothesis. OLS, fixed effect and random effect are used. For selecting the best model, this study runs the Pool ability F-test for selecting the OLS or fixed effect models. If the p-value of the F-test <0.05, then reject H 0, and the Fixed effect estimator will be superior to OLS—Breusch-Pagan LM test run for selecting OLS or Random effect. This study runs the Hausman test to choose the random effect or fixed effect model. For robustness analysis, the present study used Corruption Perception Index (CPI) to capture the impact of corruption on bank NPL, which is the alternate measure of corruption (See Equations (1) and (2). NPLi,t=β0+β1COCt+β2SIZEi,t+β3PROFITi,t+β4CAPi,t+β5INCDi,t +β6INEFFi,t+β7GDPt+β8INFt+β9IRt+ui,t (1) NPLi,t=β0+β1CPIt+β2SIZEi,t+β3PROFITi,t+β4CAPi,t+β5INCDi,t +β6INEFFi,t+β7GDPt+β8INFt+β9IRt+ui,t (2) Data analysis and discussion Preliminary results Descriptive findings are presented in Table 3, covering the mean score, deviation from the mean and minimum–maximum value of the variables in the study. Table 3 illustrates that for the credit risk in the form of NPL, the mean score is 0.0629, explains that on average, the banking sector has a mean trend of above 6 per cent in the form of weak loan performance over the whole period of study where the highest level of NPL is recorded 0.5156. The dependent variable is nonperforming loan as measured by NPL=nonperforming loan/total loan; COC is the control of corruption; SIZE is the bank size measured by logarithm of total assets; PROFIT is Table 2 NPL ratios in Pakistan, India and Bangladesh. Country 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Pakistan 7.60 9.10 12.20 14.75 16.21 14.47 12.99 12.27 11.36 10.06 8.43 7.97 8.58 India 2.70 2.40 2.20 2.39 2.67 3.37 4.03 4.35 5.88 9.19 9.98 9.46 9.23 Bangladesh 13.20 10.80 9.20 7.30 6.10 10.00 8.90 10.00 8.80 10.10 10.41 9.88 8.90 Average 7.83 7.43 7.87 8.15 8.33 9.28 8.64 8.87 8.68 9.78 9.61 9.10 8.90 Source: World Bank Database. A. Rehman et al. Research in Globalization 9 (2024) 100241 5 profitability measured by return on assets; CAP is capitalisation measured by equity capital/total assets; INCD is income diversification measured by non-interest income/total income; INEFF is inefficiency measured by cost/income; GDP is gross domestic product measured by the annual growth rate of GDP; INF is inflation measured by annual consumer price index; IR is interest rate measured by annual real interest rate. In addition, the mean for COC is −0.7161, whereas it’s minimum and maximum value is −1.4965 and −0.2395, respectively. Overall, these three countries have weak COC showing negative signs of the min and max. The bank-specific control variables are represented by SIZE, PROFIT, CAP, INCD and INEFF. The mean value of SIZE is 15.3311, with a maximum value of 20.0866. Meanwhile, the mean value of PROFIT and CAP is 0.0143 and 0.0779, with a maximum value of 0.2350 and 0.5394, respectively. Furthermore, the mean value of INCD is 0.1840 and INEFF mean score is 0.5402. GDP, INF, and IR represent the macroeconomic control variables. The overall GDP mean value is 0.0426, with a maximum value of 0.0704. For INF, the mean value is 0.0674, with a maximum value of 0.2029. Whereas for the IR, the average mean score is 0.0471 with a maximum value of 0.0926. The correlation analysis is used to detect multicollinearity among independent variables, which may impact their relationship with the dependent variables in a regression analysis (Pallant, 2009). Severe multicollinearity is a significant issue because it raises the variance of regression coefficients and causes them to be unstable. However, based on the correlation matrix shown in Table 4, all these values are above 0.70, the benchmark to identify multicollinearity (Pallant, 2009). Hence, there is evidence of multicollinearity problems among variables in the model. Multiple regression analysis Table 5 presents the results of the multiple regression analysis. The regression equation can explain 31.9 % of the variation (reflected through R-square) in the dependent variable (i.e. NPL). At the same time, the F-statistics for the equation is 12.58, which is significant at 1 per cent. Additionally, the Wooldridge and Wald tests’results for autocorrelation and heteroscedasticity yield significant findings at a 1 % level (p-value =0.0000). Hence, to overcome the autocorrelation and heteroscedasticity problem, t-statistics are computed using robust standard errors. The results show that the coefficients for the main variable of interest, COC, are statistically significant at a 5 % level. In contrast, other control variables like PROFIT, GDP, and INF are found to be statistically significant at a 1 % level. Meanwhile, INEFF and IR are statistically significant at a 10 % level. The significant effect of COC implies that an increase in country-wide corruption, including in the banking sector, will lead to the inappropriate approval of loans and make the recovery of loans very difficult. This is not surprising in the case of South Asia, where almost all the countries are confronted with a high level of corruption (low control over corruption). This might be particularly true for South Asian countries, which are facing a high level of corruption (the mean of COC is −0.7166, which shows low control of corruption, that’s why NPL is high in South Asian counties). The regression result indicates that increasing the COC Index by 1 unit may decrease the banks’NPL of 0.025 units. The effect of the first control variable in the analysis, bank size (SIZE), is not statistically significant. This implies that, on average, this notion is not supported by clear evidence that SIZE determines the NPL for the selected banks under the present study. This would justify the argument that the SIZE of the banks has no role in determining the NPL. Conversely, the coefficient of PROFIT, which is assumed as the second control variable, is negatively significant (p-value =0.000). This would indicate that a higher level of the bank’s earnings may lead to a lower level of NPL and vice versa. As for the economic magnitude, an increase of 1 unit in ROA will decrease 1.746 units in bank non-performing loans in the sample countries. This result is similar to Peric and Konjusak (2017) and Vithessonthi (2016) findings. The higher level of banking sector earnings indicates a good performance dynamic, resulting in a lower amount of NPL. In addition, the impact of capitalisation (CAP) on NPL is found to be negatively insignificant, meaning that there is inadequate evidence to recommend that CAP determines the NPL for the banks in the sample studied. Similarly, income diversification (INCD) shows an insignificant impact on the NPLs. This implies that, on average, no proper evidence supports the notion that INCD determines the nonperforming loans for banks in the sample studied. However, the bank inefficiency coefficient (INEFF) measured by the cost-to-income ratio is positive and statistically significant (p-value =0.076), supporting the weak management hypothesis. The result implies that a 1 unit increase in INEFF will increase 0.050 units in NPL. This indicates that high-cost inefficiency (inefficient management) may increase the NPL. After analysing the effect of bank-specific dynamics on NPL, the relationship between macroeconomic variables and NPL is observed. This is followed that the impact of GDP growth on NPL is negatively significant at 1 per cent (p-value =0.008). This result indicates that an increase of one unit in GDP will decrease 0.254 units in NPL which suggests that in good economic growth will result in lower NPL and the elimination of bad loans from the economy. These results are consistent with Peric and Konjusak (2017) and Alhassan et al. (2014), who emphasise that a good economic situation in the country may raise the borrower’s creditworthiness, thus leading to lower NPL for the banking sector. Similar to GDP, the impact of the second macroeconomic variable, entitled inflation (INF), shows a significant and negative coefficient at a 1 per cent level (p =0.000). This result indicates that an increase of one unit in INF will result in a 0.267 unit decrease in NPL. This result is similar to the finding of Pop et al. (2018), where inflation is significantly decreasing the NPL value, and a consistent trend is observed (Castro, 2013). In addition, Dimitrios et al. (2016) also express that the inflation rate can determine the extent of the country’s NPL because it improves loan repayment abilities of the borrowing customers and makes loans cheaper. Additionally, Ghosh (2015) argues that, theoretically, inflation decreases the debt value and thus makes the servicing of debt easier. Meanwhile, the coefficient estimation of interest rate (IR) is Table 3 Descriptive statistics of all variables. Variable Mean Standard deviation Min Max NPL 0.0629 0.0618 0.0008 0.5156 COC −0.7161 0.3489 −1.4965 −0.2395 SIZE 15.3311 1.6619 2.0692 20.0866 PROFIT 0.0143 0.0167 −0.1037 0.2350 CAP 0.0779 0.0424 0.0158 0.5394 INCD 0.1840 0.0876 0.0131 0.4349 INEFF 0.5402 0.1639 0.0810 1.4050 GDP 0.0426 0.0203 −0.0060 0.0704 INF 0.0674 0.0324 0.0201 0.2029 IR 0.0471 0.0317 −0.0677 0.0926 A. Rehman et al. Research in Globalization 9 (2024) 100241 6 positively significant (p-value =0.087). This result would justify the claim that an increase of one unit in IR would increase to 0.087 teams in NPL and vice versa. This result is similar to the finding of Louzis, Vouldis, and Metaxas (2012), Chaibi and Ftiti (2015), and Tarchouna et al. (2017), who have claimed similar results where a significant and positive relationship exists between interest rate and NPL. Furthermore, these findings align with the theory that a higher interest rate would lead to a higher level of NPL, implying that a high-interest rate increases the costs of funds by increasing debt servicing costs with the lower payment capacity of the borrowers as well. Meanwhile, this promotes the culture of high-risk-taking behaviour in the banking sector, where loans are approved for high-risk borrowers at a very high-interest rate. Thus, these loans most probably transformed into bad loans, hence increase NPL in the banking sector. Additional analysis The robustness results are presented in Table 6 using the corruption perception index. Similar to COC, the result also shows that CPI has a negatively significant relationship with NPL. The result is similar to Toader et al. (2017), who claimed a significant and adverse association between the CPI and NPL in the banking sector. Additionally, the effect of bank size, profitability, GDP growth, and inflation is found to be negatively significant. This would justify the earlier findings where it is explained that higher earnings in the banking sector lead to lower NPL, along with the negative effect of economic dynamics like economic growth and inflation for lower NPL and vice versa. Conclusion To address the study’s objective, control of corruption (COC) was used to investigate the relationship with NPL. As an alternative measure of corruption, this study also used the corruption perception index (CPI) for the robustness test. The results show that control of corruption (COC) is negatively related to nonperforming loans and is significant for conventional banks in South Asian countries. The findings correspond to those of Toader et al. (2017), Park (2012) and Bougatef (2016), who found that if corrupt behaviour were poorly controlled (i.e., corruption levels were higher), more non-performing loans would be the outcome. The estimates for COC had a significant impact, indicating that if corruption is practised across the economy, unsuitable credit and insider credit will be created. At the same time, loans may be diverted away from schemes the credit committee had approved. Such projects may ultimately be deemed “bad loans”, resulting in a higher NPL ratio. Loans would also become more expensive, so a borrower would find it less easy to make repayments. This result seems to suggest that better control on corruption in South Asia has the ability to restrict the unethical behavior associated with corruption during the 2000 to 2019 study period, otherwise it supports the increase of levels of nonperforming loans in the conventional banks of South Asian countries. Limitations of the study This is important to note various limitations. First, this study only examined conventional banks as a sample. Therefore, these research findings cannot necessarily be applied to other forms of banks in South Asia, like those focusing on investment, cooperation, or development. Secondly, the study utilised data from the operations of traditional banks only because these types of banks follow a separate regulatory system from Islamic banks. Furthermore, Islamic banks were omitted because they are few in number, and most banks were established in recent years, where sufficient data is unavailable. Third, this research primarily concentrated on the South Asian banking sector. Hence, the findings are restricted to the NPLs of South Asian banks and do not represent the NPLs of banks in other developing and developed nations. The results are also limited to the study period of 2000 to 2019. Fourth, Table 4 Correlation matrix. Variables NPL COC SIZE PROFIT CAP INCD INEFF GDPP INF IR NPL 1 COC −0.207* 1 SIZE −0.144* 0.618* 1 PROFIT −0.425* −0.264* −0.105* 1 CAP 0.065* −0.151* −0.260* 0.171* 1 INCD −0.072* −0.426* −0.327* 0.355* 0.188* 1 EFF 0.324* 0.135* −0.031 −0.575* −0.025 −0.117* 1 GDP −0.380* 0.408* 0.287* 0.053 −0.131* 0.100* −0.068* 1 INF −0.006 −0.184* 0.018 −0.03 0.150* −0.121* −0.043 −0.121* 1 IR −0.081* 0.053 −0.027 0.076* −0.156* 0.118* 0.051 0.071* −0.641* 1 *shows significance at the 0.05 level. Table 5 Regression result for COC. Variable Coefficient Standard error t-value p-value COC −0.025 0.012 −2.02 0.046** SIZE −0.004 0.003 −1.46 0.148 PROFIT −1.746 0.342 −5.10 0.000*** CAP −0.012 0.083 −0.15 0.884 INCD 0.069 0.052 1.35 0.182 INEFF 0.050 0.028 1.80 0.076* GDP −0.254 0.093 −2.74 0.008*** INF −0.267 0.045 −5.88 0.000*** IR 0.087 0.050 1.73 0.087* Constant 0.119 0.050 2.39 0.019** R-squared 0.319 Number of obs 960 F-test 12.580 Prob >F 0.0000 LM Test (p-value) 0.0000 Wooldridge Test (Prob >F) 0.0000 Hausman Test (p-value) 0.0000 Wald Test (p-value) 0.0000 ***Significant at 1 % level, **Significant at 5 % level, *Significant at 10 % level. Notes: The dependent variable is the nonperforming loan (NPL). The t-statistics and p-values are based on robust standard errors. A. Rehman et al. Research in Globalization 9 (2024) 100241 7 missing values under some of the variables have provided that this study is based on the unbalanced panel, which is another limitation of this study. Also, some of the conventional banks in South Asian countries started their operations later than in 2006. Suggestions for future research In the light of the present study, several future research directions are highlighted. Firstly, a key topic of debate arising from the GFC has been the way that traditional banks address non-performing loans. Conventional banks are essential as intermediation parties involved in default risk or nonperforming loans in their business. Nevertheless, finding any studies beyond highly limited empirical research is impossible. Therefore, the suggestion is for further studies to recognise how conventional banks determine the issue of NPLs. Corruption should not be the only theme; different relevant aspects should also be extensively investigated, such as official policies, ownership, Islamic standpoints, and legal and public elements. Second, future studies should analyse nonperforming loans in a disaggregating method, whereby nonperforming loans are subdivided into consumer and corporate loans. Thirdly, for NPLs of traditional banks to be examined more comprehensively, research could be conducted that compares states, or regions, such as (the South Asian Association for Regional Cooperation) or ASEAN (the Association of Southeast Asian Nations), which have very similar external environments. In addition, a comparison between South Asian banks and other banks in developing and developed countries is also essential. The fourth recommendation is that researchers need to conduct an analysis comparing NPLs in both standard and Islamic banking systems. CRediT authorship contribution statement Abdul Rehman: Writing –original draft, Conceptualization. Waqas Mehmood: Supervision, Methodology, Conceptualization. Ahnaf Ali Alsmady: Writing –review &editing, Software. Arshian Sharif: Writing –review &editing, Validation, Conceptualization. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. References Abd Karim, M. Z., Sok, G. C., & Hassan, S. (2010). Bank efficiency and non-performing loans: Evidence from Malaysia and Singapore. Prague Economic Papers. Abuzayed, B., Ammar, M. B., Molyneux, P., & Al-Fayoumi, N. (2024). Corruption, lending and bank performance. International Review of Economics &Finance, 89, 802–830. Acharya, V., Drechsler, I., & Schnabl, P. (2014). A pyrrhic victory? Bank bailouts and sovereign credit risk. The Journal of Finance, 69(6), 2689–2739. Adhikary, B. K. (2006). Nonperforming loans in the banking sector of Bangladesh: Realities and challenges. Bangladesh Institute of Bank Management, 4(26), 75–95. Adzobu, L. D., Agbloyor, E. K., & Aboagye, A. (2017). The effect of loan portfolio diversification on banks’risks and return: Evidence from an emerging market. Managerial Finance, 43(11), 1274–1291. Ahmad, F. (2013). Corruption and information sharing as determinants of nonperforming loans. Business Systems Research Journal: International Journal of the Society for Advancing Business &Information Technology (BIT), 4(1), 87–98. Al-Khazali, O. M., & Mirzaei, A. (2017). The impact of oil price movements on bank nonperforming loans: Global evidence from oil-exporting countries. Emerging Markets Review, 31, 193–208. Ali, M., Sohail, A., Khan, L., & Puah, C.-H. (2019). Exploring the role of risk and corruption on bank stability: Evidence from Pakistan. Journal of Money Laundering Control, 22(2), 270–288. Alhassan, A. L., Coleman, A. K., & Andoh, C. (2014). Asset quality in a crisis period: An empirical examination of Ghanaian banks. Review of Development Finance, 4(1), 50–62. Anastasiou, D. (2017). Is ex-post credit risk affected by the cycles? The case of Italian banks. Research in International Business and Finance, 42, 242–248. Balgova, M., Nies, M., &Plekhanov, A. (2016). The economic impact of reducing nonperforming loans. Banagladesh Bank. (2014). Prudential regulations for banks. Retrieved from https://www. bb.org.bd/aboutus/regulationguideline/prudregjan2014.pdf. Barreto, R. A. (1996). Endogenous corruption, inequality and growth. European Economic Review, 44(1), 35–60. Becker, G. S., & Stigler, G. J. (1974). Law enforcement, malfeasance, and compensation of enforcers. The Journal of Legal Studies, 3(1), 1–18. Ben Saada, M. (2018). The impact of control quality on the non-performing loans of Tunisian listed banks. Managerial Auditing Journal, 33(1), 2–15. Berger, A. N., & DeYoung, R. (1997). Problem loans and cost efficiency in commercial banks. Journal of Banking Finance, 21(6), 849–870. Bernanke, B. S., Lown, C. S., & Friedman, B. M. (1991). The credit crunch. Brookings papers on Economic Activity, 1991(2), 205–247. Bougatef, K. (2016). How corruption affects loan portfolio quality in emerging markets? Journal of Financial Crime, 23(4), 769–785. Castro, V. (2013). Macroeconomic determinants of the credit risk in the banking system: The case of the GIPSI. Economic Modelling, 31, 672–683. Chaibi, H., & Ftiti, Z. (2015). Credit risk determinants: Evidence from a cross-country study. Research in International Business and Finance, 33, 1–16. Chen, M., Jeon, B. N., Wang, R., & Wu, J. (2015). Corruption and bank risk-taking: Evidence from emerging economies. Emerging Markets Review, 24, 122–148. Demirgüç-Kunt, A., & Detragiache, E. (1998). The determinants of banking crises in developing and developed countries. Staff Papers, 45(1), 81–109. Dimant, E., & Tosato, G. (2017). Causes and effects of corruption: What has past decade’s empirical research taught us? A survey. Journal of Economic Surveys, 32(2), 335–356. Dimitrios, A., Helen, L., & Mike, T. (2016). Determinants of non-performing loans: Evidence from Euro-area countries. Finance Research Letters, 18, 116–119. Friedrich, C. J. (1972). Opposition, and government, by violence. Government Opposition, 7(1), 3–19. GCI. (2017). The Global Competitiveness Index Retrieved from https://www.weforum. org/reports/the-global-competitiveness-report-2017-2018. Ghenimi, A., Chaibi, H., & Omri, M. A. B. (2017). The effects of liquidity risk and credit risk on bank stability: Evidence from the MENA region. Borsa Istanbul Review, 17(4), 238–248. Ghosh, A. (2015). Banking-industry specific and regional economic determinants of nonperforming loans: Evidence from US states. Journal of financial stability, 20, 93–104. Table 6 Regression result for CPI. Variable Coefficient Standard error t-value p-value CPI −0.001 0.000 −4.57 0.000*** SIZE −0.007 0.003 −2.25 0.027** PROFIT −1.702 0.337 −5.05 0.000*** CAP −0.015 0.081 −0.18 0.855 INCD 0.070 0.048 1.45 0.151 INEFF 0.054 0.027 2.04 0.045** GDP −0.217 0.096 −2.26 0.027** INF −0.208 0.044 −4.71 0.000*** IR 0.096 0.050 1.94 0.056* Constant 0.201 0.053 3.77 0.000*** R-squared 0.339 Number of obs 974 F-test 12.846 Prob >F 0.0000 LM Test (p-value) 0.0000 Wooldridge Test (Prob >F) 0.0000 Hausman Test (p-value) 0.0000 Wald Test (p-value) 0.0000 ***Significant at 1 % level, **Significant at 5 % level, *Significant at 10 % level. Notes: The dependent variable is the nonperforming loan (NPL). The t-statistics and p-values are based on robust standard errors. A. Rehman et al. Research in Globalization 9 (2024) 100241 8