INTRODUCTION:

Non-Performing Assets (NPAs) are frequently associated with the lending procedures of banks, as they can create a financial burden for banks and impact their overall stability. One of the most significant steps towards recognizing and addressing the issue of NPA in the Indian banking sector was by Narasimham Committee in its report on ‘Financial System Reforms’ in 1991.

According to the Reserve Bank of India (RBI), ‘terms loan on which interest or instalment of principal remain overdue for a period of more than 90 days from the end of a particular quarter is called a Non-performing Asset’ (Reserve Bank of India, 2004)

The Reserve Bank of India (RBI) issued a circular titled 'Classification of Non-Performing Assets' in 2004, which provided guidelines for banks to classify such assets based on their performance. The circular stated that assets which have been non-performing for a period of 12 months or less are considered substandard, while those which have remained substandard for 12 months or more are categorized as doubtful. Assets that are deemed uncollectible and have little value as a bankable asset are known as loss assets, though there may be some potential for recovery.

The problem of NPAs in India began to surface in the early 1990s, when economic reforms were introduced. The opening up of the financial sector provided an opportunity for private entities to venture into the banking industry. The significant policy changes resulted in increased competition and growth opportunities for banks, but it also brought about a problem of reckless lending and high rates of loan defaults, leading to the accumulation of NPAs.

The Indian government has taken several measures to tackle the problem of NPAs. The Securitization and Reconstruction of Financial Assets and Enforcement of Security Interest (SARFAESI) Act in 2002 streamlined the recovery process. It allowed banks and financial institutions to auction off secured assets of defaulting borrowers without court intervention. The SARFAESI Act allowed the establishment of specialised financial institution called Asset Reconstruction Company (ARC). ARCs buy NPAs from banks and financial institutions at a reduced price and try to recover the loan through restructuring or selling the asset. Additionally, the RBI implemented the Debt Recovery Tribunals (DRTs) in 2005 to accelerate the resolution of bad debts. In 2014, the Central Repository of Information on Large Credit (CRILC) was established to regulate large credit exposures of banks and financial institutions. In 2015, the RBI introduced the Asset Quality Review (AQR), which required banks to classify all stressed assets as NPAs and make adequate provisions for them. In 2016, the government introduced the Insolvency and Bankruptcy Code (IBC), which offers a framework for the resolution of NPAs. The IBC permits the liquidation or restructuring of insolvent companies, which aids in the recovery of unpaid loans. The RBI and the Indian government implemented several steps in 2020 to alleviate the impact of the COVID-19 pandemic on the economy and the banking sector, resulting in a reduction in NPAs.

The Reserve Bank of India's Financial Stability Report shows that the gross non-performing assets ratio was 14.6% in March 2018, 8.2% in March 2020, 7.3% in 2021, and decreased further to 5% in September 2022. This reduction in NPAs indicates that the measures taken by the RBI to address the issue of NPAs have been effective

It is crucial for the Indian banking industry to identify effective remedies to improve the financial health of banks and guarantee the long-term sustainability of the sector.

The existing literature has examined several variables in relation to Non-Performing Assets. These include various bank-specific^1-5, industry-specific¹and macroeconomic variables^6,7. Some researchers have utilized CAMEL parameters to identify the bank specific variables^8,9. The management of NPAs in public sector banks has been explored in terms of size, growth, and interbank differences¹⁰. There are variations in the factors that affect NPAs across the three types of bank ownership (public, private, and foreign)¹¹. In a similar strand of literature, the comparison of performance of public sector banks with that of private sector banks in terms of NPAs has also been done^2,12.

Most authors have used static and/or dynamic panel regression on a sample that varied from 27 to 45 scheduled commercial banks^5-15. In another study, several variants of the Generalized Method of Moments technique in dynamic models have been used to analyze the data³.The relative effectiveness of several bank groups in managing their NPAs has also been studied using the exponential growth equation⁴.

The research work spans over different time periods like 1995-2011³, 1997-2009¹⁵, 1997-2017¹³, 2003-2013⁵, 2004-2018¹, 2005-2014¹¹,2010-2019^8,16and 2015-2019¹⁴

The macroeconomic variables that have been found to be prime drivers of the growth of NPAs in Indian public sector banksare GDP growth rate^10,17, interest rate^10,17, external debt, FDI inflows, Index of Industrial Production, Consumer Price Index, Inflation, Policy Repo Rate, and Exchange Rate were significant in explaining NPAs^11,13,16,18

The bank-specific variables like revenue efficiency¹⁰, capital adequacy^6,11, return on assets, and return on equity have been found significant in explaining NPAs¹³.Other Bank-specific variables, such as gross loans and advances, provisions and contingencies, income on investments, and the sector of the bank (public versus private), were also found to be significant determinants of NPAs^5,16.Some findings suggested that the extension of credit to government-specified sectors, known as Priority Sector Lending, did not result in a significant increase in the bad loan portfolio of banks^1,15. Additionally, a study revealed that unique bank-specific factors such as size and the ratio of total loans of the bank to total loans in the banking industry also played a significant role in private banks¹¹. Other important factors influencing NPAs in Indian banks were determined to be asset quality, management quality^8,9 and earning quality⁸. Another study also revealed that the relative quantum of sensitive sector advances, consisting of commercial real estate, commodity, and capital market, had a positive and significant relationship with NPA ratio⁶. It has been found that net interest margin and capital adequacy ratio had a negative and significant impact on gross non-performing advances⁶.

A study on Yes Bank revealed that the long-term impact of profitability and growth of loans and advances significantly affected its NPAs. Additionally, in the short-term causality, a unidirectional relationship was found between NPA-led bank rate and loans and advances-led NPA¹⁹.

Comparative studies between private and public banks in managing NPAs suggest that per capita income has a significant impact on NPAs in both sectors¹¹. While some studies indicate higher NPA growth in public sector banks compared to private sector banks², others suggest that nationalized, private, and foreign banks are less effective at controlling NPAs than public sector banks⁴. Public sector banks have shown improvement in managing bad debts, while private banks have maintained stability due to better risk management procedures and technology^15,33.

It has also been found that NPAs in the Indian banking system had a significant time persistence. Furthermore, they found that larger banks were more likely to default than smaller banks³.

Multiple studies have explored the determinants of non-performing assets (NPAs) in the banking sector, but their findings have varied due to differences in sample selection and time periods analyzed. The current research employs the CAMELS model to investigate NPAs within a different time frame. However, this study differs in its approach by utilizing statistical techniques across multiple years to assess the ongoing significance of these variables. By conducting multiple regressions for each year, the specific impact of the selected variables on NPAs can be examined and it can be assessed whether their significance holds consistently over time. It allows for a more detailed understanding of the dynamics and variations in the determinants of NPAs across different years.

RESEARCH METHODOLOGY:

The present research builds upon previous studies that have utilized the CAMELS model to investigate the determinants of non-performing assets in the banking sector. This model considers six parameters, including Capital Adequacy, Asset Quality, Management, Earning Capacity, Liquidity, and Sensitivity to risk, which are widely used by regulatory bodies and credit rating agencies to evaluate bank performance.

The objective of this study is to examine the relationship between non-performing assets and bank-specific variables selected from each of the CAMELS parameters. The study employed multiple linear regression to analyze the financial data of 32 banks, covering the period from 2017 to 2022, obtained from the Report on Trend and Progress of Banking in India, which is published by the Reserve Bank of India²⁰.

By analyzing the same set of variables across multiple years, the study aimed to identify whether these variables remained significant over time. The variables considered in this study are explained in the following paragraphs.

Capital Adequacy Ratio (CAR) is a crucial metric used to assess a bank's solvency and ability to manage risks. It is obtained by dividing tier 1 and tier 2 capital by risk-weighted assets, as per the Basel accord. Past research has shown conflicting results regarding the relationship between CAR and NPAs, with some studies indicating a negative relationship^6,8while others suggest a positive relationship²¹. Banks with a higher CAR that is mostly comprised of owned capital tend to be more conservative and implement more stringent measures to manage risks, which can lead to reduced levels of NPAs. Conversely, banks with a higher CAR due to a higher loan loss provision in Tier-2 of its capital may experience higher levels of defaulted loans.

Priority sector lending (PRSEC) and Secured advances (SECAD) are two indicators used to assess the quality of assets in the banking sector. Priority sector lending refers to a policy that mandates banks to allocate a certain percentage of their total loans to sectors that have historically been neglected, such as agriculture and small industries. These sectors are perceived to carry a higher risk of default, but there is no strong empirical evidence to support the idea that higher priority sector lending leads to higher non-performing assets²¹. Secured advances are loans that are backed by collateral, which reduces the risk of default for the bank. Studies have shown that a higher proportion of secured loans to total loans is associated with lower levels of non-performing assets^15,22.This is because the presence of collateral reduces the risk of default, providing the bank with an additional layer of protection against credit losses.

Business per employee (BPE) and Profit per employee (PPE) are two measures that indicate the efficiency of bank management in terms of the contribution of an employee towards bringing in business and appraising and monitoring loan accounts. The effectiveness of employees in maintaining the asset quality of the bank is crucial. Banks with higher BPE and PPE ratios are generally linked to lower levels of non-performing assets⁶.

Net Interest Margin (NIM) is an important measure of a bank's earnings and is calculated by subtracting the interest paid on deposits from the interest earned on loans and investments, and then dividing it by the bank's total assets. While some studies suggest that there is a positive relationship between NIM and NPA²³ others have found a negative relationship⁶. One possible reason for the positive relationship between NIM and NPA is that banks may increase their interest margins to compensate for anticipated loan defaults. Therefore, a higher NIM could indicate a higher level of NPAs. Conversely, a negative relationship between NIM and NPA could indicate that banks with higher NIMs are better equipped to manage risks and have stronger practices in place for identifying and managing problematic loans before they become non-performing.

Return on Assets (ROA) is a financial ratio that indicates the profitability of a bank relative to its total assets. If a bank is unable to generate sufficient profits, it may be more vulnerable to credit risk, resulting in a higher level of NPAs⁶. However, ROA had to be excluded from the analysis as an independent variable due to high multicollinearity with other variables. Despite this exclusion, other measures of profitability such as ROE and NIM were included to capture crucial aspects of bank performance and potentially offset the loss of information from the omission of ROA.

Return on Equity (ROE) is a financial ratio that indicates the profitability of a bank relative to its equity investment. Higher ROE indicates higher profitability, while a lower ROE suggests lower profitability Higher profitability, as measured by ROE, can help a bank to manage its credit risk and reduce the likelihood of NPAs¹⁸.

Credit Deposit Ratio (CDR) and Investment Deposit Ratio (IDR) are measures of a bank's ability to manage its liquidity. CDR is calculated by dividing the total loans disbursed by the bank by the total deposits received, while IDR is calculated by dividing the total investments made by the bank by the total deposits received. A high CDR suggests that the bank is using its deposits aggressively to extend credit, while a high IDR indicates that the bank is investing its funds in government or government-approved securities beyond the regulatory requirements. If a bank disburses credit without appropriate screening, it may lead to a rise in impaired loans²⁴. Conversely, if a bank diverts its funds to secure investments, it can help to lower NPAs²⁵.

The ownership of a bank is another independent variable included in the study, where a value of '0' is assigned to public sector banks and '1' is assigned to private sector banks. If the analysis shows a positive significant relationship between ownership of the bank and NPA, it would suggest that private sector banks have a lower level of NPA compared to public sector banks. This highlights that non-performing assets are more prevalent in public sector banks²⁶. A negative significant relationship between ownership of the bank and NPA would suggest that public sector banks have a lower level of NPA compared to private sector banks, which implies that public sector banks are more effective at controlling NPAs than nationalised, private, and foreign banks⁴.

In the current study, the net non-performing asset to net advances ratio has been utilized as the dependent variable. This ratio provides a more accurate indication of a bank's actual non-performing assets as it excludes the impact of provisions made by the bank for potential loan losses. It helps to identify the true level of risk in a bank's loan portfolio, which in turn provides a clearer picture of the bank's financial health.

To begin with, the study conducted preliminary investigations that involved selecting and transforming the explanatory variables, as well as analyzing the correlation and descriptive statistics to understand the characteristics of the dataset. The study also tested the assumptions of linear regression and made appropriate data transformations. Thereafter, multiple linear regression was used to analyze the data for each of the six years.

The study formulated hypotheses based on the existing literature to investigate the relationship between bank-specific factors and non-performing assets. These hypotheses were tested using statistical analysis.

The aim of this study is to examine the relationship between bank-specific factors and non-performing assets of 32 banks from 2017 to 2022. To achieve this, the following hypotheses have been formulated based on existing literature:

H1: Capital adequacy has no significant relationship with non-performing asset ratio of banks.

H2: Asset quality has no significant relationship with non-performing asset ratio of banks.

H3: Management quality has no significant relationship with non-performing asset ratio of banks.

H4: Profitability has no significant relationship with non-performing asset ratio of banks.

H5: Liquidity has no significant relationship with non-performing asset ratio of banks.

H6: Ownership has no significant relationship with non-performing asset ratio of banks.

These hypotheses were tested using statistical analysis after the selection and transformation of explanatory variables, as well as testing for the assumptions of linear regression. Descriptive statistics and correlation statistics were also used to understand the characteristics of the dataset.

The primary aim of this study is to determine the significance of various financial parameters on non-performing assets by using multiple regression analysis. The multiple regression model includes the net non-performing assets ratio as the dependent variable and nine predictors. The model is expressed using the following equation:

Y= β₀ + β_i X_i +ε_i

where Y represents the net NPA for a particular bank "i" as the dependent variable, X_i represents the independent or explanatory variables, β₀ is the Y intercept, β_i represents the slope of Y with respect to X_i while

Table 1: Descriptive Statistics

2016-17	CAR	CDR	IDR	PRSEC	SECAD	NIM	ROA	ROE	BPE	PPE
Mean	13.676	74.302	34.202	36.811	86.538	2.845	0.555	3.904	1345.692	3.701
Std. Deviation	3.156	14.769	17.258	14.437	9.826	1.429	1.205	13.043	422.660	10.913
Minimum	10.490	46.816	21.545	14.477	59.156	1.560	-2.040	-26.979	170.000	-28.000
Maximum	26.360	122.865	125.526	97.580	99.411	9.615	4.460	28.581	2345.000	32.000
Skewness	2.330	0.912	5.052	2.432	-0.765	3.596	0.607	-0.908	-0.158	-0.467
Kurtosis	7.565	2.763	27.422	9.611	0.376	16.578	2.695	0.393	1.469	2.301
2017-18	CAR	CDR	IDR	PRSEC	SECAD	NIM	ROA	ROE	BPE	PPE
Mean	14.061	76.400	33.529	37.584	85.637	2.827	0.182	-2.673	1433.876	-0.454
Std. Deviation	4.231	14.211	17.648	14.833	11.217	1.225	1.408	17.815	422.778	14.305
Minimum	9.040	45.881	23.251	12.303	52.915	1.397	-2.460	-46.631	232.000	-47.000
Maximum	31.480	108.230	126.979	94.283	99.845	8.135	4.030	19.461	2277.000	23.000
Skewness	2.276	0.191	5.069	1.741	-0.834	2.697	0.246	-0.896	-0.536	-1.081
Kurtosis	8.411	-0.143	27.491	6.003	0.583	10.851	0.561	-0.127	0.970	2.237
2018-19	CAR	CDR	IDR	PRSEC	SECAD	NIM	ROA	ROE	BPE	PPE
Mean	14.262	77.402	31.271	38.526	82.727	2.939	0.017	-3.646	1538.585	-2.805
Std. Deviation	3.581	16.150	11.191	14.689	17.226	1.267	1.601	19.253	433.497	19.897
Minimum	9.610	48.252	20.059	18.172	10.775	1.762	-4.680	-61.006	258.000	-88.000
Maximum	29.200	122.451	82.969	95.084	99.683	8.924	4.250	18.962	2327.400	23.000
Skewness	2.284	0.411	3.382	1.976	-2.579	3.567	-0.442	-1.438	-0.787	-2.761
Kurtosis	9.015	0.909	14.767	6.137	9.298	16.317	2.401	1.865	1.229	10.448
2019-20	CAR	CDR	IDR	PRSEC	SECAD	NIM	ROA	ROE	BPE	PPE
Mean	14.831	79.097	31.856	39.832	82.458	3.046	-0.086	-3.360	1514.325	-1.759
Std. Deviation	3.576	23.710	9.393	14.404	14.575	1.224	1.759	19.259	452.048	16.969
Minimum	8.460	48.158	21.827	22.650	34.606	1.850	-5.390	-67.522	322.000	-73.400
Maximum	27.430	162.715	69.737	90.774	99.231	8.537	4.180	22.910	2461.840	24.000
Skewness	1.588	1.773	2.354	1.571	-1.339	3.047	-0.939	-1.807	-0.369	-2.641
kurtosis	4.204	4.366	7.758	3.823	2.641	12.769	3.030	3.807	0.455	10.008
2020-21	CAR	CDR	IDR	PRSEC	SECAD	NIM	ROA	ROE	BPE	PPE
Mean	16.518	72.667	32.307	43.373	83.264	3.145	0.481	4.527	1592.222	4.619
Std. Deviation	2.903	15.811	6.773	14.735	14.385	1.049	0.883	9.787	476.705	9.406
Minimum	12.200	47.452	21.187	22.728	40.885	2.064	-2.550	-39.155	321.570	-30.940
Maximum	23.470	113.375	51.204	88.194	99.631	7.317	2.130	16.608	2403.000	26.000
Skewness	0.651	0.686	0.861	1.009	-1.158	2.216	-1.160	-3.083	-0.683	-1.615
Kurtosis	-0.339	0.391	0.982	1.450	1.553	7.079	4.146	12.826	0.573	6.729
2021-22	CAR	CDR	IDR	PRSEC	SECAD	NIM	ROA	ROE	BPE	PPE
Mean	17.282	73.391	31.229	42.017	81.380	3.170	0.774	8.256	1705.979	9.084
Std. Deviation	2.982	13.951	5.732	11.290	14.227	1.047	0.602	4.825	551.747	6.674
Minimum	13.230	49.074	21.563	24.088	39.276	1.926	-0.070	-0.591	325.100	-1.000
Maximum	25.900	111.571	43.684	72.638	99.336	6.865	2.130	18.979	2649.000	28.000
Skewness	0.872	0.543	0.667	0.639	-0.969	1.814	0.935	0.174	-0.512	0.865
Kurtosis	0.806	0.398	-0.170	0.377	1.122	4.218	-0.012	-0.196	0.315	0.895

Source: Based on SPSS output

holding all other variables constant, and ε_i represents the random error in Y for observation "i."

The regression analysis aims to investigate the direction and nature of the relationship between non-performing assets and independent variables in line with the research hypothesis.

RESULTS AND DISCUSSIONS:

The descriptive statistics that summarize and describe the key features of the data have been presented in Table- 1.

The data suggests that the banks' capital adequacy has improved over time and dispersion of CAR values around the mean is consistent over the years. The mean CDR has gradually increased over time, indicating that banks have been lending more credit relative to the deposits they hold. The standard deviation of CDR has also increased over time, indicating greater variability in the ratio. The mean IDR has gradually decreased over time, indicating that banks or financial institutions have been investing less relative to the deposits they hold. The standard deviation of IDR has also decreased over time, indicating less variability in the ratio. The mean PRSEC has gradually increased over time, indicating that banks have been lending more to the priority sectors as mandated by the RBI. The standard deviation of PRSEC has remained relatively stable over time. The average Investment Deposit Ratio (IDR) has shown a consistent trend over the years, with only a slight decrease in the year 2018-19. The standard deviation, however, shows significant variations, indicating changes in the loan portfolio mix over time. NIM has generally increased over the years, with some fluctuations. ROA and ROE have been volatile over the years and it can be seen that banks have not been generating profits consistently. The BPE values provided for each year show an increasing trend over time, indicating an improvement in the company's efficiency in generating revenue per employee. However, fluctuating PPE could mean increased business generated by the employees is not necessarily translated into higher profits.

Although the normality assumption refers to the distribution of the residuals in the statistical analysis, it is still important to check the distribution of the independent variables as well, especially if they are highly skewed or contain outliers. This is so because the presence of outliers or skewed data in independent variables can affect the regression coefficients and their interpretation. A skewness coefficient close to zero and kurtosis of 3 is considered reasonable. However, it is evident from Table 1 that certain variables have a high degree of skewness and kurtosis, suggesting that they do not follow a normal distribution. Therefore, data transformations were implemented to address these issues as presented in Table 2.

After transforming the data, normality assumptions of the independent variables were tested using statistical tests such as Shapiro-Wilk and Kolmogorov-Smirnov. A significance level of .05 was used to test the normality assumption. If the test statistic had a significance value greater than .05, it was concluded that the data followed a normal distribution. It was found that this assumption was met in all six years of the study.

One of the assumptions of the model is that the residuals follow a normal distribution. This assumption was checked by constructing a normal P-P plot, which showed that all points on the plot were roughly along a straight diagonal line. This assumption was found to hold true for all the years considered in the study.

As a number of ratios that have been considered represent different dimensions of the financial health of a bank, multicollinearity was expected. A correlation coefficient greater than 0.7 and a VIF greater than 5 are indicators of multicollinearity. Multicollinearity can reduce the statistical power of the analysis by reducing the precision of the estimates and inflating the standard errors. The problem of multicollinearity can be solved by reducing the number of collinear variables until there is only one remaining out of the correlated set. To address the issue of multicollinearity, the study excluded a few variables that were highly correlated with each other. For example, ROA was excluded from further analysis because it had a high VIF in all the years and was highly correlated with multiple other predictors. Other variables, such as PPE, CDR, IDR and ROE, were excluded in certain years where they exhibited high correlations with other variables. By reducing the number of collinear variables, the study was able to ensure that the remaining variables had reasonable tolerance values and VIFs in each year (Table 3)

Table 3: Tolerance and VIF of selected variables

	2016-17		2017-18		2018-19		2019-20		2020-21		2021-22
	Tolerance	VIF	Tolerance	VIF	Tolerance	VIF	Tolerance	VIF	Tolerance	VIF	Tolerance	VIF
CAR	0.292	3.427	0.347	2.883	0.254	3.938	0.465	2.148	0.283	3.529	0.269	3.718
PRSEC	0.256	3.912	0.231	4.321	0.285	3.511	0.559	1.790	0.552	1.813	0.456	2.192
SECAD	0.378	2.647	0.251	3.985	0.432	2.317	0.480	2.082	0.690	1.449	0.383	2.608
BPE	0.328	3.053	0.539	1.855	0.371	2.697	0.418	2.390	0.298	3.360	0.275	3.642
PPE	-	-	-	-	0.733	1.365	-	-	0.654	1.528	-	-
CDR	0.197	5.070	0.211	4.746	0.229	4.362	0.404	2.478	-	-	0.321	3.113
IDR	0.535	1.871			0.387	2.586	0.616	1.622	0.528	1.894	0.532	1.879
NIM	0.331	3.017	0.361	2.767	0.226	4.427	0.250	4.000	0.310	3.121	0.314	3.188
ROE	-	-	-	-	-	-	-	-	0.327	3.058	0.412	2.427

Source: Based on SPSS output

Table 4: Results of Multiple Linear Regression

Year	2016-17		2017-18		2018-19
R square	0.771		.768		.862
F(8,23)	9.674	F(8,23)	9.496	F(9,21)	14.575
Sig.	.000		.000	Sig.	.000
	ß value	Sig	ß value	Sig	ß value	Sig
RITCAR	130.402	0.011	118.760	0.004	48.139	0.089
SECAD	-0.034	0.607	-0.013	0.877	22.407	0.333
LOGPRSEC	7.205	0.182	4.904	0.410	0.884	0.776
BPE	-0.001	0.547	0.000	0.833	0.000	0.871
RITNIM	5.911	0.003	4.197	0.010	5.8340	0.007
CDR	0.095	0.126	0.048	0.507	-.066	0.044
LOGIDR	-3.833	0.380	-4.363	0.505	0.360	0.914
Pr/Pu	-1.858	0.104	-1.710	0.218	0.174	0.838
Year	2019-20		2020-21		2021-22
R square	.797		.799		72.7
F(8,23)	11.289	F(8,19)	9.414	F(9,22)	6.319
Sig.	.000		.000	Sig.	.000
	ß value	Sig	ß value	Sig	ß value	Sig
RITCAR	37.784	0.025	-0.069	0.489	-0.157	0.094
SECAD	0.010	0.598	0.001	0.921	-0.037	0.030
LOGPRSEC	1.790	0.329	0.033	0.027	0.009	0.645
BPE	-0.001	0.316	-0.001	0.208	-0.001	0.044
PPE	-	-	0.049	0.747	-	-
ROE	-	-	-1.009	0.004	-0.079	0.093
RITNIM	3.124	0.058	-.241	.902	2.444	0.050
CDR	-4.255	0.115	-	-	0.008	0.638
LOGIDR	0.490	0.830	-0.077	0.035	-0.020	0.550
Pr/Pu	-0.802	0.294	-1.419	0.066	-0.375	0.527

Source: Based on SPSS output

The year wise regression results are summarised in the following paragraphs:

2016-17: A multiple regression analysis was conducted to investigate the relationship between net non-performing assets ratio and the independent variables - CAR, SECAD, PRSEC, BPE, NIM, CDR, IDR, and Ownership. The results showed that the independent variables significantly predicted the net NPA ratio, F(8,23)= 9.674, p<.05, with an R-square value of 0.771, indicating that the model can explain 77.1% of the variance in the dependent variable. However, further analysis revealed that only two independent variables, CAR and NIM, exhibited a significant positive relationship with the net NPA ratio.

2017-18: The net non-performing assets ratio was regressed on independent variables- CAR, SECAD, PRSEC, BPE, NIM, CDR, IDR and Ownership. The results showed that the independent variables significantly predicted the net NPA ratio, F(8,23)= 9.496, p<.05 with an R-square value of 0.768 indicating that the model can explain 76.8% of the variance in the dependent variable. However, further analysis revealed that only two independent variables, CAR and NIM, exhibited a significant positive relationship with the net NPA ratio.

2018-19: The net non-performing assets ratio was regressed on independent variables- CAR, SECAD, PRSEC, BPE, PPE, NIM, CDR, IDR and Ownership. The results showed that the independent variables significantly predicted the net NPA ratio, F(9,21)= 14.577, p<.05 with an R-square value of 0.862 which depicts that the model explains 86.2% variance in the dependent variable. However, further analysis revealed that only two independent variables, CDR and NIM, exhibited a significant relationship with the net NPA ratio. CDR was found to have a negative relationship with the net NPA ratio while NIM was found to have a positive relationship with the net NPA ratio.

2019-2020: The net non-performing assets ratio was regressed on independent variables- CAR, SECAD, PRSEC, BPE, NIM, CDR, IDR and Ownership. The results showed that the independent variables, taken together significantly predict the net NPA, F(8,23)= 11.289, p<.05 with an R-square value of 0.797 indicating that the model explains 79.7% of the variance in the dependent variable. Further examination of the individual relationships between the independent variables and the net NPA ratio revealed that only one variable, CAR, had a significant positive relationship with the net NPA ratio at a significance level of 5%. Another variable, NIM, was found to be significant at a 10% significance level.

2020-21: The net non-performing assets ratio was regressed on independent variables- CAR, SECAD, PRSEC, BPE, PPE, NIM, ROE, IDR and Ownership. The results showed that the independent variables, taken together significantly predict the net NPA, F(8,19)= 9.414, p<.05 with an R-square value of 0.799 which depicts that the model explains 79.9% variance in the dependent variable. Further examination of the individual relationships between the independent variables and the net NPA ratio revealed that IDR and ROE had a significant negative relationship and PRSEC had a significant positive relationship with the net NPA ratio at a significance level of 5%. Additionally, ownership was found to be significant at a 10% significance level.

2021-22: The net non-performing assets ratio was regressed on independent variables- CAR, SECAD, PRSEC, BPE, NIM, ROE, CDR, IDR and Ownership. The results showed that the independent variables, taken together significantly predict the net NPA, F(9,22)= 6.319, p<.05 with R-square value of 0.727 indicating that the model explains 72.7% variance in the dependent variable. Further examination of the individual relationships between the independent variables and the net NPA ratio revealed that SECAD and BPE had a significant negative relationship with the net NPA ratio at a significance level of 5%. Additionally, NIM was found to be significant at a 10% significance level.

The present study investigated the relationship between net non-performing assets ratio and various independent variables for 32 banks over a period of six years. The results suggest that the independent variables, taken together, significantly predict the net NPA ratio each year. However, only a few independent variables were found to have significant relationships with the net NPA ratio in each year.

In particular, the variables that consistently showed significant positive relationships with the net NPA ratio were CAR and NIM, while CDR showed a significant negative relationship in one year. Other variables, such as SECAD, PRSEC, BPE, PPE, ROE, and IDR, had varying degrees of significant relationships with the net NPA ratio across different years. Ownership was found to be significant at a 10% significances level in one year.

Overall, these findings suggest that banks should pay close attention to their CAR and NIM ratios, as they appear to have consistent and significant relationships with the net NPA ratio. It may also be important for banks to consider the other variables that showed significant relationships with the net NPA ratio in certain years, such as CDR, PRSEC, and BPE, to better manage their risk and improve their financial performance.

The research methodology adopted in the present study did not allow for the inclusion of macroeconomic variables in the regression model, despite their known significant impact on non-performing assets as indicated by previous studies. Instead, the study focused on regressing the dependent variable of net NPA with independent variables related to financial parameters for each of the six years under study, in order to assess their significance. However, it is important to acknowledge that the exclusion of macroeconomic variables from the analysis may limit the overall understanding of the determinants of NPAs within the broader economic context.

CONCLUSION:Top of Form:

Non-performing assets are a major cause of concern for banks worldwide, and their impact on a bank's profitability and financial stability cannot be overstated. The factors that contribute to NPA can vary across banks. There are several bank-specific factors and macroeconomic factors whose relationship with non-performing assets has been investigated in previous studies. The present study empirically tested the relationship between variables representing different parameters of CAMELS approach and non performing asset for each of six years from 2017 to 2022. In the present study, the R-squared values for the regression models were found to be high, indicating that a significant portion of the variance in the net NPA ratio can be attributed to the independent variables considered in the analysis. It is important that banks monitor various financial indicators to manage and mitigate their credit risk. While government initiatives are important in addressing the issue of non-performing assets in the banking sector, it is also important to acknowledge that NPAs can be attributed to poor credit discipline. This includes inefficient underwriting criteria, inadequate due diligence or collateral and lack of proper monitoring of borrowers' creditworthiness. To effectively tackle the issue of NPAs, it is necessary to implement internal controls and responsible lending practices within the banking sector. The automation and digitisation of processes can improve the efficiency and accuracy of credit assessments, while also ensuring better risk management and reduction of NPAs.

CONFLICT OF INTEREST:

The authors have no conflicts of interest regarding this investigation.

REFERENCES:

1. Gaur D, Mohapatra DR. The nexus of economic growth, priority sector lending and non-performing assets: case of Indian banking sector. South Asian Journal of Business Studies. 2021; 10(1):70-90.

2. Mohanty BK, Krishnankutty R. Determinants of profitability in Indian banks in the changing scenario. International Journal of Economics and Financial Issues. 2018; 8(3):235.

3. Memdani L. Macroeconomic and bank specific determinants of non-performing loans (NPLs) in the Indian banking sector. Studies in Business and Economics. 2017; 12(2):125-135.

4. Siraj KK, Pillai PS. Efficiency of NPA management in Indian SCBs - A bank-group wise exploratory study. Journal of Applied Finance and Banking. 2013; 3(2):123.

5. Bamoriya P, Jain R. NPA and select key financial heads: An empirical study of commercial banks of India using multiple regression modeling. The Journal of Accounting and Management. 2013; 3(2).

6. Sengupta R, Vardhan H. Non-performing assets in Indian Banks: This time it is different. Economic and Political Weekly. 2017; 52(12): 85-95.

7. Mittal RK, Suneja D. The problem of rising non-performing assets in banking sector in India: comparative analysis of public and private sector banks. International Journal of Management, IT and Engineering. 2017; 7(7):384-398.

8. Kuknor S, Rastogi S. Determinants of profitability in Indian banks: a panel data analysis. International Journal of Modern Agriculture. 2021; 10(2):978-986.

9. Gupta A, Gupta SC, Afreen A. An empirical analysis of the causal relationship between NPA and its determinants in Yes Bank.

10. Goyal A, Joshi S. Non-performing assets of Indian banking sector: A CAMEL approach. Journal of Public Affairs. 2019; 19(1):e1727.

11. Bawa JK, Goyal V, Mitra SK, Basu S. An analysis of NPAs of Indian banks: Using a comprehensive framework of 31 financial ratios. IIMB Management Review. 2019; 31(1):51-62.

12. Javheri J, Gawali R. A study on NPAs of selected private & public sector banks in India. Indian Journal of Finance and Banking. 2022;9(1):129-139.

13. Rahaman Sk Mujibar. Assessing the Impact of Non-Performing Assets on Profitability of Public Sector Banks in India. International Journal of Reviews and Research in Social Sciences. 2022; 10(3):129-135.

14. Mishra AK, Jain S, Abid M. Non‐performing assets and its determinants in the Indian banking system: An empirical analysis using dynamic panel data models. International Journal of Finance & Economics. 2021; 26(4):5948-5962.

15. Dhar S, Bakshi A. Determinants of loan losses of Indian Banks: a panel study. Journal of Asia Business Studies. 2015; 9(1):17-32.

16. Kumar Gulshan, Kapoor Gaurav. Financial Performance Analysis of Banks – A Study of SBI. Asian Journal of Management. 2022; 13(4): 285-288.

17. Roy SG. Determinants of non-performing assets in India - panel regression. Eurasian Journal of Economics and Finance. 2014; 2(3): 69-78.

18. Mittal RK, Suneja D. The problem of rising non-performing assets in banking sector in India: comparative analysis of public and private sector banks. International Journal of Management, IT and Engineering. 2017; 7(7): 384-398.

19. Chawla S, Rani S. Investigating the determinants of non-performing assets: the case of the Indian banking sector. Journal of Commerce & Accounting Research. 2021; 10(4).

20. Reserve Bank of India. Report on trend and progress of banking in India. Retrieved from https://www.rbi.org.in/Scripts/AnnualPublications.aspx?head=Trend%20and%20Progress%20of%20Banking%20in%20India (2022).

21. Das A, Ghosh S. Financial deregulation and efficiency: An empirical analysis of Indian banks during the post reform period. Review of Financial Economics. 2006;15(3):193-221.

22. Bhowmick G, Banerjee B. Non-performing assets in the priority sector advance of public sector banks during 1994-95 to 2005-06: an empirical analysis. Indian Accounting Review. 2008; 17(2):75-82.

23. Kaur J, Kaur J. Determinants of Non-Performing Assets in Indian Banks: A CAMEL Analysis. Journal of Financial Economic Policy. 2020; 12(3): 527-547.

24. Sangeetha R, Chinu Moorarka. Macroeconomic variables and Commercial banks in India. Asian Journal of Management. 2019;10(1):25-28.

25. Jimenez G, Salas V, Saurina J. Determinants of collateral. Journal of Financial Economics. 2006; 81(2):255-281.

26. Kadanda D, Raj K. Non-performing assets (NPAs) and its determinants: a study of Indian public sector banks. Journal of Social and Economic Development. 2018; 20: 193-212.

27. Bardhan S, Mukherjee V. Bank-specific determinants of nonperforming assets of Indian banks. International Economics and Economic Policy. 2016; 13: 483-498.

28. Goyal Neha, Agrawal Rachna, Aggarwal Renu. Review of Priority Sector Lending Performance of Indian Commercial Banks. Asian J. Management. 2015; 6(4): 302-306.

29. Pal Subhabaha, Pal Satyabrata, Kaushik Bhattacharjee. A PCA Approach for Comparing Performances of Major Indian Banks during 2004-05 to 2013-14. Asian J. Management. 2017; 8(3): 442-450.

30. Saluja Manminder Singh, Sharma Sweta. Effect of NPAs on Financial Performance of Public Sector Banks in India: A Study of Changing Scenario. Asian J. Management. 2017; 8(3): 577-586.

31. Yadav JR. Credit Is Inevitable In Banking: A Case Study Of Premier Bank On Credit Appraisal and Assessment. Journal of Business Management and Social Sciences Research. 2013; 2(7): 20-29.

32. Saddy Neeraj Kumar, Kaur Harpreet. Recovery Management of Non-Performing Assets of Public and Private Sector Banks. Asian J. Management. 2013; 4(4): 293-296.

33. Swamy V. Determinants of bank asset quality and profitability-An empirical assessment. Available at SSRN 2126761. 2012.

34. Dheenadhayalan V, Rajaprabu D. Loan Assets in New Private Sector Banks in India. Asian J. Management. 2014; 5(3): 347-353.

35. T. Vara Lakshmi, Reddy Srinivasa. Imprint of Insolvencies on Commercial Bank's Priority Sector Lending Practices. Asian J. Management. 2017; 8(3): 642-644.

36. Reserve Bank of India. Master Circular - Prudential norms on Income Recognition, Asset Classification and Provisioning pertaining to Advances. Retrieved from https://www.rbi.org.in/scripts/BS_CircularIndexDisplay.aspx?Id=2657 (2004, June 28).

37. Thiagarajan S, Ayyappan S, Ramachandran A. Credit risk determinants of public and private sector banks in India. European Journal of Economics, Finance and Administrative Sciences. 2011; 34(34): 147-153.

38. Yadav Shailaja P. Performance evaluation of public sector banks with reference to ‘CAMEL model’ for the period of 2006 – 2013. Asian J. Management. 2014; 5(4): 374-379.

Received on 26.07.2023 Modified on 30.08.2023

Asian Journal of Management. 2023;14(4):283-292.

DOI: 10.52711/2321-5763.2023.00046

	2016-17	2017-18	2018-19	2019-20	2020-21	2021-22
Capital Adequacy ratio	Inverse transformation	Inverse transformation	Inverse transformation	Inverse transformation	No change	No change
Secured advances to total advances	No change	No change	Inverse transformation	No change	No change	No change
Priority sector advances to total advances	Log transformation	Log transformation	Log transformation	Log transformation	No change	No change
Business per employee	No change	No change	No change	No change	No change	No change
Profit per employee	Excluded	Excluded	Inverse transformation	Excluded	Inverse transformation	Excludd
Return on Assets	Excluded	Excluded	Excluded	Excluded		Excluded
Return on Equity	Excluded	Excluded	Excluded	Excluded	Excluded	No change
Net Interest Margin	Inverse transformation	Inverse transformation	Inverse transformation	Inverse transformation	Inverse transformation	Inverse transformation
Credit Deposit Ratio	No change	No change	No change	Log transformation	Excluded	No change
Investment Deposit Ratio	Log transformation	Log transformation	Log transformation	Log transformation	No change	No change