Non-performing loans (NPLs) and non-performance: evidence from South Asian banks

This paper examines the consequences of banks’ performance on bank risk. The paper forms a theoretical model and delivers empirical evidence to identify that banks suffer in performance as the loans become bad. Using panel data from a sample of five (05) South-Asian emerging economies from 2011 to 2019, we have found that the banks are highly influenced by the development of non-performing loans (NPLs). We have primarily used Return on Asset (ROA) followed by Return on Equity (ROE) as a substitution to the performance of the banks and NPL as the proxy of bank risk. Simultaneous regression applying 3sls finds that Non-Performing Loan (NPL) hinders banks' growth, negatively affecting their profitability. © 2024 by the authors. Licensee SSBFNET, Istanbul, Turkey. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).


Introduction
In the last decade, the South-Asian banking sector has evolved significantly.At the same time, compliance loan disbursements have deteriorated, which directly affects banks' performance.There are several determinants of a bank's viability.These are either external or internal factors.Internal factors derive from bank accounts and are represented as bank-related profitability elements.External elements are not linked to bank administration and are placed exogenously on banks.
Macroeconomic indicators and strategies, the regulatory climate, technical advances, the international economy & and politics are among these exogenous variables.The activity and efficiency of financial institutions in an open economy would be influenced by all these conditions (Li, 1999).
The critical performance of banks, which work on the notion of money creation, is credit development; consequently, the likelihood of banks failing if the money they lend is not payback by borrowers is substantial.As a result, banks place a high value on bank credit risk supervision, primarily focused on decreasing the likelihood of borrowers defaulting on the loan payoff, particularly to producing non-performing loans (NPLs).NPLs are loans that have not generated returns for long, i.e., the principal and interest on these loan disbursements have not been paid for a minimum period of 90 days.Nkusu (2011) shows that NPLs are a central causative element in credit market frictions and macro-financial fragility when looking at the link between NPLs and their macroeconomic repercussions.Increases in NPLs, according to the report, are a prelude to debilitating macroeconomic activities.Credit risk administration is crucial to the effective operation of banks and the financial structure as a whole; this is especially true when the proportion of owners' equity to total assets is too poor that non-performing loans (NPLs) may consequently lead a bank to be unsuccessful.DeYoung and Rice (2004) recognized the interest spread of banks as a vital way of bank income.If NPLs are not controlled, the bank will collapse because interest revenue is reduced, decreasing the net interest margin.
On the other hand, this logic contradicts one of finance's most basic tenets: conventional asset pricing, which states that when a company's risk exposure is more, its likely return should also rise.In this instance, greater credit risk, as measured by NPLs, may result in high returns.Findings like those of Jesus and Gabriel (2006), found strong indication that riskier borrowers take large bank loans in upturns while collateral-backed loans decrease.Sufian (2012) showed a significant positive connection between credit risk and banks' income.As a result, while riskier lending may end in higher NPLs and lower returns, it may also elevate profitability.
This suggests that NPLs and bank profitability have a positive connection.As a result, the effect of NPLs on company profitability remains unknown, and further research is required.There is proof to assist the notion that in an economy with a well-functioning commercial banking system, significant levels of economic progress are recorded, according to Zhang et al. (2012).A poorly functioning financial system, on the other hand, may lead to increased poverty and sluggish economic progress.Banking institutions might be economically debilitating and calamitous if they fall under the weight of large levels of non-performing loans.This claim is not far from the truth since Reinhart and Rogoff (2010) found that NPLs can be used to forecast the beginning of a financial crunch, one of the most common reasons for economic distress.Many of the previous bank crises have been triggered by higher levels of non-performing loans (NPLs) within the industry, according to Fofack (2005).The global financial crisis of 2007/2008, caused by borrowers' nonpayment of sub-prime mortgages/credits in the United States, exhibits the negative effect that NPLs may have on economic fortunes.

Literature Review
Empirically and scientifically, the factors of bank efficiency have been thoroughly investigated.Financial businesses that run more effectively naturally generate more revenue and make more money.(Ramadan, Kilani, & Kaddumi, 2011) explored the connotation between bank profitability and the factors of internal and external influences in Jordan.The findings of his analysis indicate that the profitability of Jordanian banks varies considerably, including in terms of well-capitalization, high loan disbursement practices, low credit risk, and product management, combined with high profitability.The findings also indicate that the scale does not significantly affect the profitability of Jordanian banks.Bofondi and Ropele (2011) showed that NPLs were favorably connected to lending rates and unemployment but adversely related to Italy's GDP (gross domestic product) rate between 1990 and 2010.Salas and Saurina (2002) found that real GDP growth rate and fast credit expansion explained variance in NPLs when they studied the causes of bad loans in Spanish commercial and saving banks.GDP growth is also one of the primary drivers of variance in NPLs, according to Beck et al. (2013).They show that economic downturns lead to a rise in NPLs, whereas improved economic performance decreases NPLs.
Similarly, De Bock and Demyanets (2012) found that lending quality is affected by economic development.They argue NPL ratios are counter-cyclical, dropping during upturns in the business cycle and climbing during downturns.Espinoza and Prasad (2010), estimated a dynamic panel on eighty banks in the Gulf Cooperation Council from 1995-2008 to find that NPL increases with poor economic growth and higher rates of interest.Jesus and Gabriel (2006), investigated the Spanish banks (1984 to 2003) and showed that GDP growth rate, high real interest rate, and flexible loan terms affect NPLs.Messai and Jouini (2013) also found data identifying the negative connections between GDP growth and NPLs.Khemraj and Pasha (2009) revealed a significant inverse relationship between GDP and NPLs; inflation in the Guyanese banking system has been shown to have no effect on NPLs.Hoggarth et al. (2005) found that NPLs link interest rates and inflation positively.Rajan and Dhal (2003) identified that favourable macroeconomic conditions (as expressed by GDP growth rate) and financial features, for instance maturity, cost and terms of the loan, bank size, and credit orientation, substantially affect commercial banks' NPLs in India.
Currency rate depreciation, according to Beck et al. (2013), encourages a rise in non-performing loans in countries where there is a significant amount of lending in foreign currencies to unhedged borrowers.As stated by De Bock and Demyanets (2012), exchange rate depreciation diminishes private lending and worsens the quality of loans.According to Beck et al. (2013), A bearish market may considerably impact bank asset quality in economies with big stock markets compared to GDP.In terms of the stock exchange's element in influencing NPLs, De Bock and Demyanets (2012) discover that, despite common belief, when a capital market is doing good, it can safeguard pledgers from unpredicted shocks by enabling access to loans and help with servicing debt, real equity yields have no statistical association with NPLs.(Havrylchyk & Jurzyk, 2011) have noticed a substantial direct connection between bank earnings and capital assets.Since they can handle their assets and liabilities more efficiently, more competent banks are bound to generate more money.(Chirwa, 2003) uses a time series data from 1970 up to 1994 to examine the relationship between business dynamics and profitability of Malawi banks.He determines a long-term partnership of profitability capital asset ratio, loan-to-asset ratio, and demand-to-deposit ratio.(Staikouras & Wood, 2004) indicate that factors related to their successful management and macroeconomic climate affect European banks' profitability.(Zaman, 2011) applied the POLS approach to examine the effect of savings, loans, equity, and deposits on assets (ROA).Their empirical findings indicate these factors have a powerful impact on bank performance.The results further suggest that greater total assets cannot essentially initiate higher wages because of the economy's scale.Greater loans add to profitability but do not significantly affect it, whereas equity and savings influence profitability substantially.
The internal factors of the bank play a critical part in measuring the bank's profitability.Internal aspects of banks include (1) bank's financial proficiency, (2) risk appetite (capital ratio), (3) viability, and standard of credit.They have an essential role to play in assessing the bank's profits.Bank capital has a substantial impact on bank earnings by (Abreu & Mendes, 2001), (Kosmidou, Tanna, & Pasiouras, 2005) and (Flamini & Schumacher).A well-capitalized bank is risk-averse, increasing public confidence, lowering bankruptcies, and leading to a sustainable benefit impact.In 18 European nations from 1986 to 1989, (Molyneux & Thornton, 1992) showed that the bank's return-to-equity (ROE) profitability was positively connected to the concentration of banks, ownership, and some macroeconomic variables.(Samad, 2015) also collaborated on this discovery.
(P. Athanasoglou, Delis, & Staikouras, 2006) observed that ROA had a positive effect on total assets and the equity to total assets ratio and hurt the ratio of provisions to aggregate loans and operational costs to total assets in their study, including details from Romania, Bosnia-Herzegovina, Bulgaria, Croatia, Albania, and Serbia-Montenegro for the duration 1998-2002.(Abreu & Mendes, 2001) Analyzed data from Germany, Portugal, France, and Spain from 1986 to 1999.They concluded that credit and equity to assets positively affected ROA, and the bank's market share and the equity to overall assets ratio positively impacted ROE.Moreover, inflation and unemployment have a detrimental effect on all profitability rates.
According to Perry (1992), the association between inflation and bank production relies on whether or not the bank's management expects inflation.Managers can raise revenue more rapidly than costs by correctly anticipating inflation and changing interest rates.Several observers also found an ineffective financial structure in higher-inflation nations.Inflation is expected to have a detrimental effect on credit.
Historically, the considerable growth in non-performing loans has often been connected to the deterioration of the financial system, leading to insolvency (Fofack & Fofack, 2005).Similarly, the inability to reduce non-performing loan rates would contribute to the failure of banks (Richard et al., 2008).(Samir & Kamra, 2013) finds that NPLs harm bank earnings by reducing interest income and eroding existing profits and the capital base by the provision.Non-performing loans accounted for nearly 75% of the combined debt portfolios of more than sixty banks that failed in Indonesia during the financial crisis (Caprio & Klingebiel, 1996).
Non-performing loans are thought to affect profitability.Due to the write-off of doubtful and bad debt, which usually has an impact on profitability and capital levels, high non-performing loan levels are detrimental to banks (Ombaba, 2013).Thus, empirical research (Kithinji, 2010;Ombaba, 2013) has revealed a potential relationship between low profitability and a large percentage of nonperforming loans in areas where non-performing loan levels are greater.
The size of the bank is one of the significant factors influencing the output of the banking sector.Bigger banks are considered to be profitable than smaller banks due to economies of scale, as stated by (Molyneux Thornton, 1992), (Bikker Hu, 2002), and (Goddard et al., 2004).In most financial literature, the scale of the bank is deemed an essential consideration.Currencies or large economies may impact the bank's costs and earnings.However, the analytical observations on this issue are not definitive.Everything is messed up.However, in a non-competitive climate where large banks hold large market shares, more significant revenues are required through higher loan rates and low deposit rates (Ramadan et al., 2011).Since massive banks go through more agency and bureaucratic challenges than smaller banks, a reverse association exists between the bank's scale and outcome (P.P. Athanasoglou et al., 2008).
Broad bank sizes can contribute to cost reductions on a scale or hit countries and lead to credit and commodity diversification, permitting access to markets that a minor bank cannot enter.The indication is not conclusive with those economies.Studies describe the economies of scale for major banks (Altunbaş et al., 2001;Berger & Humphrey, 1997).Small or large banks have defined their economies of scale (Pallage, 1991;Vennet, 1998).(Golin & Delhaise, 2013) Points out that "a bank needs to protect against liquidity risk carefullythe risk that it might not have sufficient current assets, such as cash and quickly sold securities, to fulfil its current obligations, e.g., those of depositors, particularly during times of economic stress" (p.273) A bank may collapse without the necessary liquidity and funding to meet its short-term liabilities.Hence, the greater value of this ratio makes the bank highly liquid and less susceptible to losses.

Research and Methodology
We have used balanced panel data for five (05) Asian countries, e.g., India, Bangladesh, Sri Lanka, Pakistan, and Nepal.In order to explain the link between bank risk (NPL) and bank loan growth (LG) from 2011 to 2019, we first created an econometric model.Data availability is the rationale for the nine-year duration.Prior data are not accessible in BankFocus.It should be noted that several nations have missing data because of their weak banking sectors, limited economic bases, and restricted information access.
Countries, i.e., Afghanistan, Bhutan, and the Maldives, have been excluded from data development due to data unavailability.However, we have shown the graphs of different variables for all eight south-Asian countries.For a few missing data, we have used the exponential growth method.We have utilized the exponential growth rate formula to analyze the trend of each indicator and obtain some approximations of data for past periods.Put otherwise, Pn = Pn (1+r) n, where t is the time in equal intervals written as an integer, r is the rate of rise or decrease (exponential decay), and Pn is the beginning (missing) value.The amounts in this document are all in USD since BankFocus has already converted them.

Variable Definition
Return on assets (ROA): For this analysis, ROA is the dependent variable calculated by converting net profits after tax into total assets at the end of the financial year.Net income divided by total assets, or ROA, shows how the bank's management makes use of its actual resources to generate revenue.ROA is a calculation of profitability that measures how banks generate a profit relative to their investments, which implies how successful management uses company assets.On average, a higher ROA represents the efficient usage of the properties of a company to gain money.LG represents the percentage increase or decrease of total loans in year t relative to the base year t-1.Hence, the formula for loan growth is denoted by (Lit-Lit-1)/Lit-1, where L stands for loan disbursement.

Capital Efficiency (EQTA):
The equity / total asset ratio (EQTA) is the primary indicator of bank solvency.The equity ratio demonstrates that a bank may finance some form of unintended loss (due to lending or other activities).The bank's capital aversion is determined by EQTA, which represents the bank's aversion measure.The share capital of banks calculates the percentage of net assets.The traditional theory of risk returns suggests that bank capital and income are adversely interconnected.In other words, the greater the share of capital equity in general assets, the greater the risk aversion of a bank.Greater risk resistance implies low leverage and, thus, low sales.

Liquidity (LIQD):
LIQD comprises cash, balances with other financial institutions and, banks money at call and short notice, shortterm loans, and advances.Liquidity is the primary source of bank financing and impacts banks' profitability.If a bank keeps more cash deposits in its vault, it is more liquid and safer to settle depositors' claims.However, the sustainability of banks is affected.On the other hand, the sustainability of a bank is boosted as banks engage in long-term loans.If its capital is spent on long-term loans, the bank becomes vulnerable.The bank cannot entertain the huge demand for cash and fresh lending.Liquidity risk is the bank's failure to satisfy its customers' cash demand.The greater number of loans per dollar deposit raises bank liquidity risk.Liquid assets, however, are commonly correlated with lower return rates, so a negative relationship between this variable and profitability is typically predicted.

CIR (Cost to Income Ratio):
CIR is a significant financial measure essential for banks' appreciation.It indicates a company's costs in comparison to its earnings.Operating expenditures (administrative and fixed costs, such as payroll and property costs, but not bad loans written off) are divided into operational benefits to obtain the percentage.The ratio gives consumers a clear understanding of how the organization operates successfullythe bigger it is, the more efficient the bank will be.The ratio offers customers a good idea of how well the business operates-the smaller it is, the stronger the bank will be.The CIR ratio calculates the bank's overheads or charges, including personnel wages and pensions, occupancy, and other expenditures, such as office supply, as a portion of income.It is usually expressed as a measure of the capacity of management to reduce costs.Higher spending typically results in lower income and vice versa, so cost is anticipated to affect bank profits and margins adversely.

201
NPL (Non-Performing Loan): The NPL ratio, determined by dividing non-performing loans into total loans and advances, is used to measure credit risk.NPL is the crucial indicator of a Bank's financial performance.The greater the NPL ratio, the lower the quality of the credit and, thus, the greater the chance of charging more loan losses against profits.

Findings and Discussions
Table 1 illustrates the inflation factor test for variation.This test is used to confirm the validity of the regression effects as well as to identify multicollinearity, or the variance of the inflation component, in the model.In VIF, it is not advised to use multicollinearity if the result is less than 10 and has nearly zero tolerance.3 shows the association matrix.The correlation study reveals that ROA has a positive connection with LG, EQTA, LIQD, and INF, whereas NPL and GDP have a adverse relationship with ROA as a dependent variable.Since they are exposed to more prospects relative to small banks, big banks are argued to be in a stronger position in terms of their profitability.We primarily executed different diagnostic checks (white test for heteroskedasticity, VIF test for autocorrelations) and found no multicollinearity among the variables used.Also, the initial OLS assessment presented heteroskedasticity and first-order serial correlation.Therefore, BLUE coefficients could not be provided by the OLS.Next, to select a fixed-effect model over a randomeffect model, we ran the Hausman specification test.The outcomes from the fixed effect model are discussed in table-04.Table 4 displays the regression test results used to determine the theory's scientific validity on the factors contributing to bank profitability.Through the Hausman test, we have found that a fixed effect exists.LIQD have a negative relationship with non-performing loans with statistical significance.The significances of the results are that in the South-Asian economy, because of a high level of NPLs, the performance (ROA) of the banks deteriorate at the same time, excess liquidity (LIQD) is not handled correctly by the banking sector, and hence excess liquidity ends in bad lending consequently poor performance by the banks.In addition, excessive cost (CIR) tells upon profitability as well.
Important to Liquidity (LIQD) also has a negative association with Profitability (ROA).NPL ratio of South Asian banks had been overgrowing as the bank tends to lend irrationally with high liquidity, the chance of bad loan increases, and profitability decreases.
As empirical studies show, one of the fundamental ratios for capital strength is also considered to be the equity to total assets (EQTA) ratio (Golin, 2001).The higher the percentage, the lower the need for external funding and, thus, the higher the bank's profitability is expected.Well-capitalized banks, by contrast, face lower failure rates, thus reducing their borrowing costs.
In both models, the R 2 s are satisfactory.The R 2 s for variables are 0.73 and 0.96, with ROA being the dependent variable and 0.52 and 0.95 with ROE as the dependent variable, respectively, meaning the models fit well.
Among macroeconomic variables, GDP has significant negative relationships with banks' performance, whereas inflation (INF) positively affects banks' performance.The countries we have considered in our studies are emerging economies and have sufficient banks in the economy.Hence it is clear from the results that banks of the South-Asian region are not efficient enough.Therefore, with the growth in the economy, the banks' lending has increased, but the recovery is not satisfactory, and hence loans are getting non-performing.

Robustness check of regression results
To check the robustness of the study, we have replaced ROA with ROE in table 06.The results are almost similar to the regression results found with ROA, except ROE has a negative relationship with the solvency of banks (EQTA).Finally, we check the results in FGLS regression, whereas well we have found similar results.

Figure 1 :
Figure 1: Return on Assets (ROA) in billion USD; Source: BankFocus Database Loan Growth (LG):LG represents the percentage increase or decrease of total loans in year t relative to the base year t-1.Hence, the formula for loan growth is denoted by (Lit-Lit-1)/Lit-1, where L stands for loan disbursement.

Table 1 :
Variance Inflation FactorTable2delivers descriptive analysis and summary statistics for the factors included in the analysis.Every variable has a positive mean value except Loan Growth and ROA.Loan decreased by around 28.21% in Merchant Bank Of Sri Lanka & Finance PLC in 2013, and in 2012 the ROA went down by 5.20% in BASIC Bank Ltd-Bangladesh.The ROA has a 1.17 average and a 1.72 standard deviation.As the table indicates, relative to other variables, variables of Cost to Income Ratio (CIR) presents a greater standard deviation.It indicates that banks' CIR has more important differences than other variables.

Table 4 :
Pooled regression, Fixed EffectWe then executed simultaneous equations in our 3SLS regression analysis.The results are enumerated in table-05.
*** The coefficients at a 1 percent level are significant; ** The coefficients at a 5 percent level are significant; * The coefficients at a 10 percent level are significant 203

Table 5 :
Regression results for 3SLS; dependent variable ROA LG) has a significant positive effect on banks' profitability, which is measured as Return on Asset (ROA) in this current study, meaning that as the loan multiplies, profitability increases as well.Results for 3 stage least square regression suggest that ROA has a strong negative relationship with NPLs with 1% statistical significance.Other bank-level variables e.g., CIR and