A STUDY ON RECOVERY CHANNELS OF NPAS: AN EMPIRICAL EVALUATION

Banks are the backbone of a sound financial system as they play a significant role in the growth and development of a country by advancing loans to various sectors of an economy. For sustainable growth of a developing country like India, it is very imperative that its banking sector should be healthy enough to meet capital requirements of different sections. But, from last few years, high magnitude of non-performing assets (NPAs) causes momentous effect on banks’ earnings as they have to make provisions on NPAs. Hence, it is very crucial that banks should take efficient resolution techniques for recovering massive amount of NPAs in order to improve their financial performance. The present study investigated the various parameters of NPAs in scheduled commercial banks (SCBs) as well as examined the performance of three major legal recovery channels for NPAs that is, Lok Adalats, SARFAESI and DRTs. The study employed ANOVA and Kruskal-Wallis test for analyzing the differences among recovery warning signals, management related signals, special mention category accounts, Lok Adalats, DRTs, and SARFAESI Act. The authors found that DRTs are the most efficient resolution channel followed by SARFAESI and concluded that banks should conduct proper credit monitoring of the project, carry out follow-up procedure, and adopt risk management mechanism for recovering enormous amount of NPAs. In another study by Chatterjee, Mukherjee, and Das (2012), the authors reported that most of the NPAs have been recovered through DRTs in 2010 and through SARFAESI in 2011. The study concluded that NPAs have severe negative impact on the overall banking operations and consequently on whole economy. Vaidyanathan (2013) elucidated the significance of credit risk management via different modes such as recovery through normal operations, corporate debt restructuring, recovery through compromise settlement, credit restructuring, Lok Adalats, DRTs, SARFAESI Act, arbitration, monitoring cash flow, and repossession of assets. The study suggested that banks should provide performance-based incentives, initiate legal measures at early stage, impart credit skills to personnel, follow disclosure procedure of NPAs, and carry out regulatory guidelines for improving their credit risk management. Another study on management of asset quality of banks conducted by Gandhi (2015) explicated various preventive measures, restructure/ rehabilitation measures, legal measures, and ARCs for efficient credit management. Further, the study discussed the new initiatives for NPA management such as bankruptcy law reforms, setting additional technical forums for evaluating projects, new secondary markets for security receipts, and independency in appraising credit proposals. Rajan (2016) suggested measures for resolving financial distress in PSBs. The author advised banks to conduct proper evaluation of projects, cautious post-lending monitoring process, ensure assets backing promoter guarantees, speeding up SARFAESI and DRTs proceedings, reforms in judiciary process, and infusing bank capital for recovering bad loans. Patel (2017) emphasized that time-bound resolution of NPA is very important for improving bank balance sheet position, for efficient allocation of capital, and for financial resilience of banks. The study suggested that banks, government, and RBI should work coordinately for addressing NPA challenge by

NPAs are classified into following three categories (a) Substandard Assets-It is an asset which has remained NPA for a period of less than or equal to 12 months; (b) Doubtful Assets-It is an asset which has remained in the substandard category for a period of 12 months; and (c) Loss Assets-It is an asset where loss has been identified by the bank or internal or external auditors or the RBI inspection but the amount has not been written off wholly (RBI, 2015a).
NPAs are categorized into Gross NPAs (GNPAs) and Net NPAs (NNPAs). GNPAs are the principal dues of NPAs plus Funded Interest Term Loan. Its ratio is GNPAs/Gross Advances. Gross Advances mean all outstanding loans and advances including advances for which refinance have been received but excluding rediscounted bills, and advances written off at Head Office level. On the other hand, NNPAs are the real NPAs which arrive after subtracting deductions from GNPAs. Its ratio is: Where deductions include provisions held in the case of NPA Accounts, deposit insurance/export credit guarantee corporation claims received and held pending adjustment, part payment received and kept in suspense/other account, balance in sundries account in respect of NPA accounts, floating provisions, provisions in lieu of diminution in the fair value of restructured accounts classified as NPAs and standard assets (RBI, 2015a).

Literature Review:-
NPAs pose a serious threat to the banking sector and therefore it is always an issue of concern for banks and policymakers. There are numerous studies on NPAs which highlighted the factors responsible for NPAs, identified its impact on bank operations, and discussed legal and non-legal mechanisms for its recovery. Few important studies on resolution techniques of NPAs are mentioned as follows: Saha (2006) inspected the various resolution measures of NPAs such as DRTs, Corporate Debt Restructuring (CDR), BIFR, and SARFAESI Act. The author stated that NPAs have a multiplier effect on the value destruction of an economy. The study asserted that banks should transfer a bad asset at an earlier stage to Asset Reconstruction Companies (ARCs) in order to ensure effective and timely recovery. It concluded that management of distressed assets lies in the speed of recycling those funds and their realization into cash. Karunakar, Vasuki, and Saravanan (2008) discussed the impact of NPAs on banks overall functioning as well as provided measures to restrain them efficiently. The study examined the trends of NPAs in public sector banks (PSBs) during 1992 to 2006, analyzed the capital to risk weighted asset ratio and recovery procedure. It concluded that banks should follow proper credit assessment mechanism before granting loans. Mehta (2009) found out that bad banking decisions and financial crises are the primary reasons for massive amount of NPAs. The study suggested that banks take assistance from ARCs, DRTs, and initiate winding up proceedings against defaulting borrowers for timely recovery of bad assets. Prasad and Veena (2011) explained the various preventive and management techniques to overcome the problem of NPAs, namely, early warning signals, financial 79 warning signals, management related signals, special mention category accounts, Lok Adalats, DRTs, and SARFAESI Act. The authors found that DRTs are the most efficient resolution channel followed by SARFAESI and concluded that banks should conduct proper credit monitoring of the project, carry out follow-up procedure, and adopt risk management mechanism for recovering enormous amount of NPAs. In another study by Chatterjee, Mukherjee, and Das (2012), the authors reported that most of the NPAs have been recovered through DRTs in 2010 and through SARFAESI in 2011. The study concluded that NPAs have severe negative impact on the overall banking operations and consequently on whole economy. Vaidyanathan (2013) elucidated the significance of credit risk management via different modes such as recovery through normal operations, corporate debt restructuring, recovery through compromise settlement, credit restructuring, Lok Adalats, DRTs, SARFAESI Act, arbitration, monitoring cash flow, and repossession of assets. The study suggested that banks should provide performancebased incentives, initiate legal measures at early stage, impart credit skills to personnel, follow disclosure procedure of NPAs, and carry out regulatory guidelines for improving their credit risk management. Another study on management of asset quality of banks conducted by Gandhi (2015) explicated various preventive measures, restructure/ rehabilitation measures, legal measures, and ARCs for efficient credit management. Further, the study discussed the new initiatives for NPA management such as bankruptcy law reforms, setting additional technical forums for evaluating projects, new secondary markets for security receipts, and independency in appraising credit proposals. Rajan (2016) suggested measures for resolving financial distress in PSBs. The author advised banks to conduct proper evaluation of projects, cautious post-lending monitoring process, ensure assets backing promoter guarantees, speeding up SARFAESI and DRTs proceedings, reforms in judiciary process, and infusing bank capital for recovering bad loans. Patel (2017) emphasized that time-bound resolution of NPA is very important for improving bank balance sheet position, for efficient allocation of capital, and for financial resilience of banks. The study suggested that banks, government, and RBI should work coordinately for addressing NPA challenge by adopting multi-pronged approach, which includes strengthening of legal, regulatory, supervisory, institutional, and operational framework. Dey (2018) scrutinized the recovery mechanisms of NPAs via DRTs, SARFAESI, and Lok Adalats as well as assessed its impact on NPAs during the period 2003-04 to 2016-17. The study found that DRTs are the most effective recovery channel followed by SARFAESI and Lok Adalats. In recent study by Kumar (2018), the trend of NPAs in SCBs for a period of 17 years ranging from 2000 to 2017 and the performance of various recovery channels of NPAs have been analyzed. The study found that the level of NPAs decreased significantly till 2005 due to establishment of Asset Reconstruction Corporation of India (ARCIL). It also revealed that SARFAESI and ARCs are the most effective channels for reducing stressed assets. In another recent study by Sowmya (2019), the reasons responsible for momentous amount of NPAs have been elucidated as well as measures were suggested for reducing them effectively. The author compared the NPAs of public and private sector banks (PvSBs) for the period covering from 2010-11 to 2016-17. The study found that PSBs contained more NPAs due to poor management policies and techniques. It suggested that banks should revise their recovery mechanisms and follow proper credit appraisal while lending loans.

Objectives of the Study:-
1. To analyze the gross NPA and net NPA position of scheduled commercial banks in India. 2. To evaluate the performance of SARFAESI, DRTs, and Lok Adalats in recovering NPAs. 3. To suggest measures for improving the recovery of NPAs.

Hypothesis of the Study:
H 0 : The distribution of total amount recovered to total amount involved in NPAs is not significantly different across SARFAESI, DRTs, and Lok Adalats in Indian SCBs.
H a : The distribution of total amount recovered to total amount involved in NPAs is significantly different across SARFAESI, DRTs, and Lok Adalats in Indian SCBs.

Research Methodology of the Study:-
The study is descriptive and empirical in nature analyzing the trend of NPAs in SCBs for the period of 15 years covering from 2004-05 to 2018-19. The study is entirely based on secondary data which have been primarily taken from RBI website. The researcher has employed various statistical tools for analyzing the data which includes ratio analysis, average, standard deviation, one-way Analysis of Variance (ANOVA), and Kruskal-Wallis test (or H test). Kothari (2012) has defined ANOVA as a technique to examine the significant difference amongst more than two sample means at the same time. It can be described as: 80 F = Estimate of population variance based on between samples variance Estimate of population variance based on within sample variance There are certain assumptions which need to be satisfied for ANOVA, that is, each of the samples is drawn from a normal population; each of these populations has the same variance; and all the factors other than the one or more being tested are effectively controlled.
In addition to F-ratio, the researcher has employed nonparametric test, that is, independent samples Kruskal-Wallis test to re-verify the result as some of the aforesaid assumptions are violated. Though, ANOVA test is somewhat robust as its results remain trustworthy even after mild violations of assumptions (Analysis of Variance, n.d.). As per Kothari (2012), H test is used to test the null hypothesis that "k" independent random samples come from identical universes against the alternative hypothesis that the means of these universes are not equal. It is worked out as below:

Recovery Channels of NPAs:
In present context, NPAs are the core of financial problem of the banks. Higher level of NPAs implies higher financial instability in banks. A detailed explanation of the three most used recovery channels by banks are given below:

The Securitization and Reconstruction of Financial Assets and Enforcement of Security Interest (SARFAESI) Act, 2002 (54 of 2002):
The SARFAESI Act came into force on June 21, 2002 and it was re-promulgated on August 22, 2002. Under SARFAESI Act, 2002, a securitization company (SC)/reconstruction company (RC) carries on the business of securitization or asset reconstruction as provided in Section 10 of this Act (RBI, 2003). It is mandatory for every SC/RC to register itself with RBI before carrying the business of securitization and reconstruction of financial assets (FAs) (The Securitization and Reconstruction, 2016). This act is the outcome of suggestions recommended by Narasimham and Andhyarujina Committees. The minimum prescribed amount for carrying out the business of securitization or asset reconstruction should not be less than 15 percent of the total final assets acquired or to be acquired by the Company on an aggregate basis or Rs. 100 crore, whichever is lower and it should commence its business within 6 months from the date of obtaining certificate of registration. The period of realization of acquired assets extended from 5 years to 8 years and in case the SC/RC is unable in resolving the assets or redeeming the Security Receipt (SR) within this time frame, then the FAs will be considered as loss assets (RBI, 2013). Here company referred to SC/RC. The foremost provisions under this Act are given here below: Acquisition of rights or interest in financial assets: SC/RC can acquire the FAs of banks/ financial institutions (FIs) through issuing debentures, bonds or any other security whose amount must be fixed as per agreement. It can also enter into an agreement to transfer such FAs at an agreed terms and conditions.After acquisition, all rights or claims of banks/FIs regarding that FAs will be transferred to SC/RC (The Securitization and Reconstruction, 2016).

Transfer of pending applications to any one of DRTs in certain cases:
If banks/FIs have filed applications before two or more DRTs regarding FAs which is acquired by SC/RC, then the company may file an application to the Appellate Tribunals for resolution of such cases. The Appellate Tribunal may pass an order for transfer of the pending applications to any one of DRT which will execute all such pending cases under sub-section (2) as per the provision of Section 19(23) of Recovery of Debts Due to Banks and Financial Institutions (RDDBFI) Act, 1993.

Notice to obligor for discharge of obligation:
The bank/FI will issue a notice regarding acquisition of FA under sub-section (1) by SC/RC to the borrower/obligor and any other concerned person and to the concerned registering authority including Registrar of Companies in whose jurisdiction the mortgage, hypothecation, charge, assignment, or other interest created on the FA have been registered. On receiving of such notice, the borrower shall make payment and discharge all the obligations relating to the specified financial asset to the concerned SC/RC.

Securitization:
As per the directions provided in RBI (2003), a SC/RC follow the provisions as given in Sections 7 (1) and (2) under SARFAESI Act, 2002 and set up one or more trusts for issuing SRs. The trust will issue the SRs to the Qualified Institutional Buyers (QIBs) and after acquiring FAs by the company, the SC/RC subsequently issue SRs to QIBs for subscription as per the provisions of the Act and raise finances from QIBs through devising schemes for acquisition of FAs.

Measures for asset reconstruction:
As per the guidelines framed under SARFAESI Act, 2002 and following the rules laid down by RBI, the SC/RC may adopt one or more of the following measures for asset reconstruction: 1. to possess/take over the management of the business of the borrower; 2. to sale or lease of a part or whole of the business of the borrower; 3. to reschedule the payments of debts payable by the borrower; 4. to enforce the security interest as per the provisions of this Act; 5. to settle the dues which is payable by the borrower; and 6. to take possession/ownership of the secured assets as per the provisions of this Act.

Enforcement of security interest:
The SARFAESI Act 2002 granted the power to the secured creditor that is, banks/FIs (if 75% of the secured creditors agree) to enforce security interest for recovering its debts without the intervention of court or tribunal. Banks are required to send a notice to the borrower that if he/she will not discharge his/her liabilities within 60 days relating to loan which is classified as NPA, then they may enforce security interest and will exercise all or any of the rights which are conferred to them under sub-section (4) of the SARFAESI Act, 2002. If the borrower fails to do so, then banks/FIs may initiate any of the following measures to recover the debt: take possession of the secured asset of the borrower as well as the right to transfer that asset; take over the management of the business of the borrower; employ any person as the manger of the secured asset; and send a notice to the person/persons who has/have acquired any secured asset from the borrower, or to the person from whom any amount is due to the borrower and recover the amount . When secured creditors take over the ownership of secured assets, the next step they required is to obtain valuation of the assets by selling those assets to attain maximum value. They sell the assets via (a) inviting tenders from the public, (b) by private treaty, (c) by obtaining tenders from the persons who deals in such assets, and (d) by holding public auctions (The Securitization and Reconstruction, 2016).

Recovery of Debts due to Banks and Financial Institutions Act (RDDBFI Act), 1993 (Act 51 of 1993):
The Government of India (GoI) has set up a committee under the chairmanship of Shri T.Tiwari in 1981 for examining the legal difficulties confronted by banks/FIs in recovering loan which suggested for establishment of Special Tribunals for speedy recovery of debts. Also the Narasimham Committee (1991) advocated the formation of Special Tribunals for fast recovery of loans. As a result of these recommendations, the RDDBFI Act, 1993 came into force on June 24, 1993 which was passed in parliament on August 27, 1993. Consequently, Debt Recovery Tribunals (DRTs) and Debt Recovery Appellate Tribunals (DRATs) were set up under RDDBFI Act, 1993 with the specific aim of "expeditious adjudication and recovery of debts due to banks and FIs" under the powers conferred to them by this Act. It deals with the cases where the amount of debt due to any banks/FIs or to a consortium of banks/FIs is more than 1 lakh rupees but less than 10 lakh rupees (Department of Financial Services [DFS], 2015a, 2015b).
The DRTs and DRATs can be established by GoI under sub-section (1) of Section 3 and Section 8 of this Act respectively. Before the establishment of these tribunals, the procedure of recovering suit filed loan was quite cumbersome and prolonged for banks/FIs due to lengthy legal proceedings. There were huge amount of funds stuck in NPAs which was unproductive and that cannot be recycle in profitable avenues, thereby hindering the growth of country. But after their establishment the number of cases came down and the situation was improved (DFS, 2015a; The Debt Recovery Appellate Tribunal, 1997; The Debt Recovery Tribunal, 1997).
Banks/FIs have to file an application to DRTs for recovering the debt from borrower. On receiving application, the tribunal will issue the summon to the defendant to explain the reason within 30 days for not paying the debt. Thereafter, DRTs will conduct the hearings between defendant and applicant (i.e., banks/FIs). In case, the defendant will not able to provide proper evidences in his favor regarding debt failure or fails to supply security for recovering debt, then the tribunal will issue an order for attachment of whole or part of that security for recovering the debt 82 amount claimed by the applicant to the extent of secured debt. If the defendant does not follow any order given by the Tribunal, then he will be detained in the civil prison for a term not exceeding 3 months, unless in the meanwhile the Tribunal directs his release (DFS, 2015b). The applicant can appeal to DRATs, if he is not satisfied by the order passed by the DRTs, but the appeal will not consider by the Appellate Tribunal unless he will deposit 75 per cent of the amount due from him (Report on Non-Performing Assets, n.d.   It is apparent from figure 1 that the GNPA ratio increased continuously from 2.4% in 2011 to 11.2% in 2018, reporting an absolute increase of 8.8% and relative increase of 366.67%. The GNPA ratio has decreased extensively from 5.2% in 2005 to 2.4% in 2011 which signifies that banks have adopt efficient measures for reducing NPAs. The highest growth rate in NNPA was 98.94% in 2016 followed by 56.37% in 2012 and 51.36% in 2013 respectively. It clearly shows that banks recovery procedure is slow and ineffective due to sluggish legal system, incompetent credit staff, faulty lending practices, and inadequate follow up measures.   Table 2 demonstrates that the amount of standard advances increased significantly but its ratio reduced significantly from 97.8% in 2008 to 90.9% in 2019 resulting in a decline of 6.9% in absolute terms which indicates failure of banks in managing NPAs efficiently. There has been momentous increase in the amount of doubtful advances from 2007 to 2018 which signify that banks were unable to prevent slippage of standard advances into doubtful advances. Source: Based on Data in Table 4.

Analysis and Interpretation:
Note: SAR: standard advances ratio; SSAR: sub-standard advances ratio; DAR: doubtful advances ratio; and LAR: loss advance ratio It is evident from figure 2 that the ratios of sub-standard, doubtful, and loss advances increased drastically during 2015 to 2019 which imply that banks were failed to recognize early warning signals and have not taken preventive measures timely in controlling stressed assets.  Adalats, that is, 6.9% because the recovery procedure in latter is quite cumbersome and lengthy. Hence, it can be inferred that DRTs are better recovery mechanism for NPAs due to its expeditious recovery of debts. On the other hand, the total numbers of cases referred to SARFAESI are 1897522 and on an average basis, the cases are 126502 86 which are lower than Lok Adalats but higher than DRTs. Although the percentage of amount recovered to total amount involved is highest in SARFAESI, that is, 29.4% on an average basis which implies that it is the most effective recovery channel for NPAs than the other two.

Results and Discussions:-
For running ANOVA, certain assumptions need to be satisfied. The results of the assumptions and the test are given below: The normality assumption of original data is violated. Hence, the data have been transformed into log values. Table  4 depicts that the significant values under Kolmogorov-Smirnov and Shapiro-Wilk in all the three groups is greater than 0.05 which indicates that the data is normal. It is apparent from table 5 that the mean percentage of amount recovered to amount involved is highest in SARFAESI, that is, 1.43 followed by DRTs (i.e., 1.25) and Lok Adalats (i.e., 0.76) respectively. The standard deviation is also lowest in SARFAESI which indicates that its values less deviate from mean as compared to DRTs and Lok Adalats.  Another assumption for ANOVA is homogeneity of variances which is checked by Levene's test statistic. Table 6 shows that the significant value is 0.010 which is less than 0.05 and hence the assumption of equality of error variances is violated.  Table 7 depicts that F (2, 42) =19.301 and p-value (=Sig.) is 0.000. Since p-value is less than 0.05, we shall reject null hypothesis. Hence, it can be inferred that there is significant difference in total amount recovered to total amount involved in NPAs across SARFAESI, DRTs, and Lok Adalats.

87
As the assumption of homogeneity of variances is violated, therefore, the researcher has re-verified the result by employing nonparametric test, that is, Kruskal Wallis test. This test is run on original values. The result is as follows: It is evident from table 8 that the significant value is 0.000 which is less than 0.05, meaning thereby that the distribution of ratio (i.e., the percentage of total amount recovered to total amount involved) is not same across categories of group. Hence, the null hypothesis is rejected under Independent-Samples Kruskal-Wallis test also. Consequently, it implies that the results are same in both the analysis and the re-verification is correct.
It is clear from boxplot that there is normal distribution of ratio across SARFAESI, DRTs, and Lok Adalats. The table clearly reveals that the asymptotic significant value is 0.000 which is less than 0.05. Hence, null hypothesis is rejected.

Conclusion of the Study:-
NPAs have become the major hurdle in the today's scenario of banking institutions. Their efficient management is the key to increase profitability, liquidity, credibility, and soundness of banks. In this study, the researcher has made an attempt to analyze the position of NPAs in SCBs and assess the effectiveness of Lok Adalats, DRTs, and SARFAESI in recovering NPAs. The study found that there has been significant increase in the GNPA and NNPA amount from 2007 to 2018 which consequently implies increase in their ratios as well. This indicates banks defective lending policy, failure to recognize early warning signals, lack of assessment of credit worthiness of borrowers, and project appraisal deficiencies. However, in 2019, there has been considerable decrease in the GNPA and NNPA amount as compared to previous year which is Rs. 1025.78 billion and Rs. 1656.03 billion in absolute terms. This decrease in NPA amount simultaneously results in decrease of GNPA and NNPA ratio from 11.2% and 88 6% in 2018 to 9.08% and 3.7% in 2019 respectively. The result of ANOVA and Kruskal-Wallis test reject null hypothesis meaning thereby that there is significant difference among channels in recovering NPAs. The author also found that though the maximum number of cases referred to Lok Adalats but SARFAESI is the most efficient channel in recovering NPAs. The reason being, that SARFAESI grants the power to secured creditors to recover their debts without the intervention of court. The procedure for recovering debts under SARFAESI is also not lengthy and consequently it is the most preferred recovery mechanism.

Suggestions of the Study:-
RBI and banks have undertaken numerous measures to bring down the voluminous level of NPAs. Following are some suggestions to resolve NPAs effectively: The key to effective management of NPAs basically involves three steps: First, banks should prevent fresh accretion of NPAs; second, they should identify potential NPAs so that banks control them at initial stage and take corrective measures to prevent their slippage into bad accounts and third, banks/(FIs) should implement prompt legal as well as non-legal procedures for recovering existing amount of NPAs.
Banks should conduct proper credit assessment of the project before lending funds as well as carry out regular monitoring and follow-up to inspect the progress of the project. They should impart credit skills to credit officials and staff associated with the task of credit appraisal and monitoring as well as gives incentives to them for prompt repayment.
Banks/FIs should carry out internal checks and controls to keep an eye on project vulnerability as well as follow Asset-Liability Management (ALM) approach to monitor credit risk, liquidity risk, and interest risk. They should also enable a well build Management Information System (MIS) as it will help them to increase their productivity and efficiency and consequently improves their asset quality.
The legal measures for recovering NPAs is very sluggish, time taking, and sympathetic towards borrowers. The judicial system should be pro-active for timely disposition of cases and should not keep the matter linger. It should carry out its duties properly and pass the decree timely so that there will be fewer cases of loan defaults which ultimately bring down the gigantic amount of NPAs.
The functioning of DRTs, Lok Adalats, and SARFAESI should be more tighten to make them function more efficiently and judiciously. They should be linked with all India database so that they can access financial debt information effortlessly and pass orders promptly. Moreover, the Government should increase the number of DRTs as the number of cases referred to them is quite large but they are unable to dispose the cases timely.