Management of Non-Performing Asset

Introduction

In today’s world, the banking industry across the world has become a significant factor in determining

the country’s economy and financial system. The stability of the banking system of the country is an

indicator of sustained economic growth, as it performs a vital role in the development of a country

through channelizing funds for profitable activities and supporting the financial and economic policies

of the government. Over the past few decades, India has become an emerging economic giant with

the world’s largest independent democracy. Post liberalization, Indian Banking Sector has undergone

multiple changes in terms of credit management and asset quality. Indian Banks have become very

cautious in sanctioning loans due to a decline in the quality of the Assets or increase in the Non-

Performing Assets.

Non-performing assets (NPA) are assets that cease to generate income through interest earned on

the principal loan amount and the repayment of the principal loan amount. Non-Performing Assets are

an outcome when the borrower intentionally defaults on the loan payment or is unable to repay the

loan due to poor economic conditions affecting his business. In either case, for a bank it means that

the loan asset may not be fully recovered or maybe only partly recovered. Non-performing assets are

a reflection of the bank’s overall efficiency while performing its business of converting deposits into

loans and recovering these loans. Non-recovery or partial recovery of loans has an impact on the

bank’s balance sheet and income statement items in the form of a reduction in interest earned on

loan assets, an increase in the provision on NPAs, increase in the capital requirement, and lower

profits. With the rise in NPA, the bank faces a scarcity of funds and becomes reluctant to lending out

loans which hampers the growth of the economy. It also soars up the interest rates which directly

impacting the investors and the individuals thereby affecting the overall demand in the economy.

NPA of a country is generally measured in the terms of Net NPA ratio or gross NPA ratio as the

absolute value of NPA does not tell about the true health of the banking sector of a country. In the
past few years, NPAs have become a major force in pulling down the bank’s profitability and

efficiency at the global level. As per the International Monetary Fund report 2019, the average NPA

ratio of the top five countries, with the highest NPA ratio, stands at 43.56% hereby indicating that

NPA has become a major hurdle for the banking sector across the globe. According to the same

report, India has the second worst NPA ratio among large economies. Macroeconomic factors

including GDP, Inflation, Interest Rate, Unemployment rate, Exchange rate, Repo rate, Country’s

external debt to GDP, Broad money, Government effectiveness, and Corruption level play a major

role in determining the NPAs. A good number of empirical studies show that bank specific variables

like banks’ efficiency, operational capability, capitalization, size, liquidity, solvency, and Net Advances

have also exhibited a significant influence on NPAs. Hence, determinants of factors which may create

NPAs should be identified prior to loans turning into NPAs.

Most academicians have examined NPA determinants and forecasted NPA, for a country or a set of

similar country. We extend the literature on NPAs by considering a mixture of developed and

undeveloped economies for determining the macroeconomic and bank specific variables affecting the

NPA and forecasting it specifically for the Indian banking sector.

2. Literature review

Numerous studies have been conducted by researchers and analysts to evaluate the factors affecting

non-performing assets of the banking sector in India and at the global level. The studies have

identified two sets of factors that determine the NPA of a country. A set of studies consider the

macro-economic variables which tend to influence the creditworthiness of the borrowers while the

other set associates the level of NPAs to a set of banks -specific factors emphasizing more on the

variability of NPAs across banks. Another strand of studies looks at both the sets of factors

simultaneously. Subsequently, after analyzing the factors affecting NPA some academicians have

also predicted the NPA of the country using various statistical models.
Among macro-economic variables GDP, construction expenditure, growth rate in per capita income,

foreign exchange reserves, stock market index, and volatility have a statistically significant inverse

relationship with NPA ratios whereas exchange rate, repo rate, and interest rate had a significant

positive impact on NPAs. (Prasanna et al, 2014)

Klein, (2013) observed that the macroeconomic factors have relatively higher explanatory power than

bank specific factors in CESEE countries. The results also confirmed that the level of NPLs tends to

increase when unemployment rises, the exchange rate depreciates, and Inflation is high. Beyond the

country-specific effects, factors, such as the euro area GDP growth and the global risk aversion, have

a direct impact on banks’ asset quality. He also found that the higher quality of the bank’s

management, as measured by the previous period’s profitability, leads to lower NPLs, while moral

hazard incentives, such as low equity, tend to worsen NPLs.

A large number of studies have provided evidence of the anti-cyclical behavior of the NPAs. The

researchers argued that the higher growth in GDP reflects a favorable economic environment in the

country which is beneficial for both households and businesses. In such a condition, the incomes of

households and businesses grow, and their capability to pay back their debts increases. In a reverse

situation, when the economy passes through a slowdown, the level of NPAs rises as rate of

unemployment is enhanced and borrowers encounter greater difficulties in repaying their loans

(Louzis et al., 2012; Ranjan and Dhal, 2003; Fofack, 2005; Jimenez and Saurina, 2006; Das and

Ghosh, 2007; Das and Kabra, 2010; Curak et al., 2013 ; Roy, 2014; Beaton et al., 2017 and Mazreku

et al., 2018)

Based on a dynamic panel formed using 75 countries it was estimated that real GDP growth, share

prices, the exchange rate, and the lending interest rate significantly affect the NPA ratio. In the case

of exchange rates, the direction of the effect depends on the extent of foreign exchange lending to

unhedged borrowers which is particularly high in countries with pegged or managed exchange rates.
In the case of share prices, the impact is found to be larger in countries that have a large stock

market relative to GDP. (Beck et al, 2013)

The effects of business cycles on the performance of commercial bank loan portfolios across major

developing economies in the period 1996–2008 indicates that while economic growth is the major

driver of loan portfolio performance, interest rates have second-order effects. Furthermore, they found

under extreme economic stress the relationship between loan loss provisions and economic growth to

be highly non-linear. During this stress, GDP growth needs to decline by more than 6 percentage

points in order to generate an increase in loan loss provisions equivalent to median emerging market

bank profits. While a decline of more than 10 percentage points in growth implies significant capital

losses, of at least 20 percent, for the median emerging market bank. Also, it was identified that the

higher loan loss provisions are associated with private sector leverage, poor loan portfolio quality, and

lack of banking system penetration and capitalization. (Glen et al, 2011)

Subsequently, researchers have also explored the relationship between bank specific variables like

profitability, loan growth, activities undertaken and the rising NPA. Clair (1992), Messai and Jouini

(2013), Curak et al. (2013), and Keeton (1999). Berger and DeYoung, 1997 examined the existence

of causality among loan quality, cost efficiency, and bank capital of the US banks through a series of

hypotheses and suggested the intertemporal relationships between loan quality and cost efficiency

run in both directions. They observed that increase in nonperforming loans tend to be followed by a

decrease in measured cost efficiency, suggesting that high levels of problem loans cause banks to

increase spending on monitoring, working out, and/or selling off these loans, and possibly become

more diligent in administering the portion of their existing loan portfolio. Berger and DeYoung

observed evidence in support of the “bad management” hypothesis in the case of US banks.

Das and Kabra (2010) and Kjosevski et al. (2019) observed the negative influence of credit growth on

NPAs. They argued that during the period of credit expansion, banks must operate with increased
caution. In addition, to achieve a satisfactory rate of credit growth, banks should focus on improving

their credit portfolio by cleaning up the NPAs.

After analyzing the position of NPAs in selected public and private sector banks of India. It was found

that the problem of NPA was greater in the public sector banks as compared to private and foreign

banks in India. Similarly, the problem of NPAs was more in the non-priority sector than the priority

sector. Further, they observed that the MSME sector has the largest share in the total NPA among

banks of the priority sector. As a result, the financial health of banks had been adversely affected.

(Samir and Kamra, 2013; Das and Dey, 2018; Singh.V.R., 2016).

Bank specific variables like inefficiency ratio had a significant positive impact on the non-performing

assets. However, bank size and performance indicators had a significant negative impact on the

NPAs of the bank indicating efficient operational management at the bank-level helps to reduce

nonperforming advances. (Prasanna et al. 2014)

Another class of researchers has taken into account the combination of macro-economic and bank

specific variables to determine the factors affecting NPAs. Analysis of BRICS countries for NPA using

multiple regression models indicates that the Trade balance, Bank provision to loans, return on

assets, and unemployment rate have a positive effect on the NPAs. The paper also suggested that

for better growth and control, governments must make structural changes in the loan process across

industries. Each country needs to develop a prototype for the structural transformation which can be

measured against the developed nations. (Indrakanti, M.N. and Sridevi, V. 2019).

Frank Betz et al (2013) contributed to the literature by developing an early warning model for the

prediction of the vulnerabilities which lead to distress in European banks using bank and country-level

data. The key findings of the research were that the complementing bank specific vulnerabilities with

indicators for macro-financial imbalances and banking sector vulnerabilities improve model

performance and yield useful out-of-sample predictions of bank distress during the current financial

crisis.
Celik and Karatepe (2006 ) further extended the previous studies by building two neural network

models to forecast the values of the output neurons consisting of Non-performing Loans/Total loans,

Capital/Assets, Profits/Assets, and Equity/Assets ratios by using 25 input neurons consisting of

macroeconomic variables, the variables related to the external balanced financial system’s structure,

and time with very small errors. He concluded that the NPA/Total Assets values that the models

forecasted tended to increase before the banking crises in Turkey.

3. Variables, Descriptive Statistics and Methodology

Model 1

NetnpaRi,t = β1 UnEmpRi,t + β2RIntRi,t + β3ExRi,t + β4Log(Net Inc)i,t + β5Log(TD)i,t+ β6Dummy Vari,t +

β7GdpGRi,t + β8Log(gdp)i,t + β9Log(Domescred)i,t + u i + vi,t

The above model has been used for determining the macro economic factors influencing the NPAs at

Global level. The dataset used in this model contains data of 10 different countries (5 developed and

5 undeveloped countries) for the past 9 years i.e. 2010 – 2018. The list of variables with their

meaning, expected influence on NPA, and descriptive statistics is mentioned in Table 1 and 4

(Appendix 1).

Data has been collected mainly from the secondary/published sources i.e. reports by RBI, the

Capitaline Corporate Database, India stats database, CEIC database, and World Development

Indicators as published by the World Bank.

 Table 1 Explanatory Variables and their expected Influence on the Net NPA ratio

Expected
Variable Meaning Previous Study
Influence
It is the share of the labor force that is jobless, expressed as a
Unemployment percentage. It is a lagging indicator, meaning that it generally rises Santosh and Sundarsan
Positive
Rate (UnEmpR) or falls in the wake of changing economic conditions, rather than (2019)
anticipating them.
It is defined as the rate of interest an investor, depositor, or lender
-Independent Variables

Real Interest Rate (RIntR) receives after allowing for inflation. It reflects the real cost of funds Positive Louzis et al. (2012); Fofack,
to the borrower and the real yield to the lender or an investor. (2005)

An exchange rate is a rate at which the domestic currency of one Klein (2013); Fofack, (2005)
Exchange Rate (ExR) Positive
country is exchanged with that of a foreign country. and Ghosh Roy (2014),

The net income of the banking sector of the country is an indicator

Log (Net Income of
of its profitability which in return is an indicator of the bank
banking sector of a country) Negative Sahoo and Mishra (2018)
efficiency. Log has been taken to scale down the values to the level
(Log(Net Inc))
of other variables.
Log (Total deposit held by
It refers to the deposit held by the bank or submitted by the
banking sector of a country) Negative Das et al (2014)
depositor/saver.
(Log(TD))
Dummy (Dev countries = 0,
It is considered that developed countries have better Net NPA ratio Klein (2013), Beck et al
Undeveloped Countries = 1) Positive
and efficient banking system. (2013) [11][22]
(Dummy Var)
It measures how fast the economy is growing. It compares the Louzis et al. (2012); Ranjan
GDP Growth Rate (GdpGR) quarter of the country's gross domestic product to the previous Negative and Dhal (2003); Fofack
quarter. (2005)
GDP of a country is the total market value of all final goods and
Control Variables

services produced in a country in a given year. As GDP decreases,

Log (GDP) Log(gdp) Negative Das and Dey (2018)
the economic conditions of the country plunges giving rise to Net
NPA ratio.
Log (Domestic It refers to lending or credit that a country or territory’s bank makes
Negative Sikdar and Makkad (2013)
Credit) Log(Domescred) available to borrowers within the same territory.

4. Analysis of Result and Discussions

4.1 Determination of the macroeconomic factors affecting the Net NPA ratio of a country

A panel data set of 10 countries over the 9 years (2010 – 2018) has been created including 7

independent variables, 2 control variables, and the Net NPA ratio as the dependent variables. The

independent variables are being selected by considering the variables that can possibly affect the

NPA and on the basis of already reported work. The hypothesis used for the model under study are:
Null Hypothesis – Net NPA ratio is independent of all the variables

Alternative Hypothesis – Net NPA ratio is a dependent variable influenced by numerous

macroeconomic factors.

The data was tested for multicollinearity using the Variance Inflation Factor (VIF). In Table 5

(Appendix 1), as the Mean VIF including all variables was less than 5 indicating multicollinearity does

not exist. Breusch Pagan Lagrange Multiplier and white tests were used to confirm the

heteroskedasticity in the data set. As, the p-value is less than 0.05 for both the test, as shown in

Table 6 and Table 7 (Appendix 1), we will reject the null hypothesis i.e. data is homoscedastic in

nature. From Table 8 (Appendix 1), we can observe that F-statistic is (22.72) at 0.05 level of

significance. Hence, the null hypothesis of the F – Test will be rejected. Rejection of the null

hypothesis of the F – Test, and BPLM test implies that both REM and FEM are significant. Therefore,

the Hausman test was used to compare the REM and FEM and the result in Table 9 (Appendix 1)

suggests that we should stick to REM. Data was also being investigated for autocorrelation using

Wooldridge Test. In Table 10 (Appendix 1), the p-value less than 0.05 at a 95% confidence interval

indicates the rejection of the Null Hypothesis. Thus, the data has the problem of autocorrelation.

Considering the problem of Heteroskedasticity, autocorrelation, and output of Hausman Test,

Feasible Generalized least squares method has been used to estimate REM. The estimated results

for the model using FGLS are presented in Table 11 of Appendix 1. The six independent variables

were found to be statistically significant in the computed model using FGLS. The major finding arising

from the estimated model and impact of these variables as shown in Table 11 can be stated as

follows:
  Unemployment Rate – It has a positive impact on the Net NPA ratio because as

unemployment rises more people will face poverty and deficit of cash. Hence, the number of

defaulters will increase.

 Interest Rate – It is positively related to the Net NPA ratio. As the interest rate will increase the

borrower would also have to pay an extra amount for each penny borrowed and the total amount to

be paid will increase. Therefore, fewer people will be able to pay back the total amount borrowed, and

hence the Net NPA ratio will increase.

 Log (Net Income of the Banking Sector) – An increase in the Net Income of the banking sector

of a country indicates an improvement in the overall economy of the country which further indicates

the improved financial condition of the individual. Therefore, when the Net Income of the banking

sector increases the Net NPA ratio decreases.

 Dummy Variable (Developed countries = 0, Undeveloped countries = 1) – The economic

condition of the country effects the Net NPA ratio of the banking sector of a country. As a country

becomes more developed GDP/per capita increases thereby decreasing the number of defaulters.
Appendix 1
Table 1: Explanatory Variables and their expected Influence on the Net NPA ratio
Table 2: List of explanatory variables used for the analysis of Public and Private Sector Bank (To be added
after the completion of 2nd objective)

Table 3: Model Selection Procedure (To be added after the completion of 2nd objective)

Table 4 Descriptive Statistics for Variables used in Model 1

Ob
Variable s Mean Std. Dev. Min Max
           
Unemployment 90 5.460 2.122 2.400 10.360
Interest Rate 90 3.494 4.618 -13.100 16.900
Exchange Rate 90 105.334 13.353 78.730 127.050
Log (Total Net Income) 90 4.263 1.406 3.070 8.340
Log (Total Deposits) 90 6.168 0.780 4.620 7.100
Net NPA Ratio 90 3.255 2.372 0.730 10.120
Dummy (Dev = 0, Undev =1) 90 0.500 0.503 0.000 1.000
GDP Growth Rate 90 3.334 2.430 -2.310 14.530
Log (GDP) 90 12.127 0.601 11.340 13.310
Log (Domestic Credit) 90 12.124 0.741 10.820 13.320

Table 5 Variance Inflation Factor Table 6 Breusch-Pagan / Cook-Weisberg test for

heteroskedasticity

Variable VIF 1/VIF

      H0: Constant variance
Log (Total Deposits) 2.550 0.392 H1: Not Constant Variance
Exchange Rate 1.620 0.618 Variables: fitted values of npa
GDP Growth Rate 1.610 0.621 chi2(1) 17.780
Interest Rate 1.600 0.626 Prob > chi2 0.000
Dummy (Dev = 0, Undev =1) 1.480 0.678
Log (Total Net Income) 1.360 0.735
Unemployment 1.320 0.757
Mean VIF 1.650  
Table 7 White's test
Table 8 F - Test Table 9 Hausman Test

H0: homoskedasticity
H1: Heteroskedasticity F (7,82) 22.72 Chi2 (6) 10.63
chi2(34) 56.900 Prob > F 0.000 Prob > Chi 2 0.101
Prob>Chi (2) 0.008
Table 10 Wooldridge test for autocorrelation in panel data
H0: no first-order autocorrelation
H1: Autocorrelation exist
F (1, 9) 210.444
Prob > F 0.000

Table 11 Estimated model for determining factors effecting NPA of a country

Coefficien
NPA t Standard Error Z-Statistic P- value
Unemployment 0.262 0.072909 3.6 0.000
Interest Rate 0.129 0.039 3.310 0.001
Exchange Rate -0.001 0.013 -0.110 0.909
Log (Total Net Income) -0.463 0.165 -2.810 0.005
Log (Total Deposits) 3.034 0.469 1.450 0.311
Dummy (Dev = 0, Undev =1) 2.307 0.432 5.340 0.000
GDP Growth Rate 0.161 0.070 1.310 0.071
Log (GDP) 4.191 1.178 3.560 0.000
Log (Domestic Credit) -4.389 0.890 -4.930 0.000
_cons -14.505 6.302 -2.300 0.021