Managing non-performing loans in banks: a review of the evidence of the impact of insolvency and creditor rights regimes Introduction Scope I have been asked to prepare a review of the literature on the impact of insolvency and creditor rights regimes (broadly construed) on the incidence/occurrence and management/resolution of non-performing loans in banks. It was agreed that whilst initiatives developed within the context of a crisis for the resolution of bank NPLs (e.g. the establishment of asset management corporations to which NPLs could be transferred) were within scope, the focus of the review would be on the resolution of non-performing bank loans rather than the resolution of distressed banks more generally, such that literature on the effects of special resolution regimes was not to be targeted. It was agreed that whilst quantitative evidence of impact was most desirable, relevant qualitative evidence would be included where apparently reliable. It was also agreed that literature that reliably showed that some change in insolvency or creditor rights regimes did not have an impact on the incidence or management/resolution of NPLs in banks, or had some adverse impact, would also be included. Whilst the focus of the review is on the impact of insolvency and creditor rights regimes, it was agreed that reliable evidence of the impact of the legal environment at a more general level (for example, court quality, strength of the rule of law, etc) on NPL levels would also be included. By “reliable” evidence, I mean evidence that has been obtained in a methodologically sound way. Weaknesses in research design might include, for example, a choice of variable that does not seem well suited to the inquiry at hand (such that there is good reason to be sceptical about whether the results shed any light on the question being asked or the relationship being explored), an unrepresentative sample, or a failure to acknowledge and deal with (e.g. by the use of appropriate robustness tests) the challenges that are inherent in the choice of methodological tool. A paper might also appear unsound because there are obvious errors in the way that the chosen methodology has been deployed, e.g. there are clear errors in the coding of data used to generate the results. Methodology My research assistant, Clara Natividade Martins Pereira (also at the University of Oxford), surveyed the literature and produced a very helpful summary document, which I attach as Annex A. The papers surveyed below are almost all drawn from this review. When deciding which papers from Ms Martins Pereira’s review to include, I focused on (i) those which appeared most methodologically sound (see immediately above), (ii) those which were most relevant to our inquiry, (iii) those which offered some new evidence of the impact of the legal framework that was more than merely anecdotal. I rejected some papers because of significant weaknesses in research design or obvious errors in the application of the chosen methodology; some because whilst the paper did touch on one or more aspects of our inquiry, it did so only tangentially, such that the paper did not appear the best source of evidence for our inquiry; and some because they merely restated evidence already uncovered in other literature (such that no new evidence was being offered), or merely reported anecdotal evidence. The methodology Ms Martins Pereira deployed to identify relevant papers is detailed at the beginning of Annex A. Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized
25
Embed
Managing non-performing loans in banks: a review of the …documents1.worldbank.org/curated/zh/389811561742153005/... · 2019. 6. 28. · Managing non-performing loans in banks: a
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Managing non-performing loans in banks: a review of the evidence of the impact of
insolvency and creditor rights regimes
Introduction
Scope
I have been asked to prepare a review of the literature on the impact of insolvency and
creditor rights regimes (broadly construed) on the incidence/occurrence and
management/resolution of non-performing loans in banks. It was agreed that whilst
initiatives developed within the context of a crisis for the resolution of bank NPLs (e.g. the
establishment of asset management corporations to which NPLs could be transferred) were
within scope, the focus of the review would be on the resolution of non-performing bank
loans rather than the resolution of distressed banks more generally, such that literature on
the effects of special resolution regimes was not to be targeted.
It was agreed that whilst quantitative evidence of impact was most desirable, relevant
qualitative evidence would be included where apparently reliable. It was also agreed that
literature that reliably showed that some change in insolvency or creditor rights regimes
did not have an impact on the incidence or management/resolution of NPLs in banks, or
had some adverse impact, would also be included.
Whilst the focus of the review is on the impact of insolvency and creditor rights regimes,
it was agreed that reliable evidence of the impact of the legal environment at a more general
level (for example, court quality, strength of the rule of law, etc) on NPL levels would also
be included.
By “reliable” evidence, I mean evidence that has been obtained in a methodologically
sound way. Weaknesses in research design might include, for example, a choice of variable
that does not seem well suited to the inquiry at hand (such that there is good reason to be
sceptical about whether the results shed any light on the question being asked or the
relationship being explored), an unrepresentative sample, or a failure to acknowledge and
deal with (e.g. by the use of appropriate robustness tests) the challenges that are inherent
in the choice of methodological tool. A paper might also appear unsound because there are
obvious errors in the way that the chosen methodology has been deployed, e.g. there are
clear errors in the coding of data used to generate the results.
Methodology
My research assistant, Clara Natividade Martins Pereira (also at the University of Oxford),
surveyed the literature and produced a very helpful summary document, which I attach as
Annex A. The papers surveyed below are almost all drawn from this review. When deciding
which papers from Ms Martins Pereira’s review to include, I focused on (i) those which
appeared most methodologically sound (see immediately above), (ii) those which were
most relevant to our inquiry, (iii) those which offered some new evidence of the impact of
the legal framework that was more than merely anecdotal. I rejected some papers because
of significant weaknesses in research design or obvious errors in the application of the
chosen methodology; some because whilst the paper did touch on one or more aspects of
our inquiry, it did so only tangentially, such that the paper did not appear the best source
of evidence for our inquiry; and some because they merely restated evidence already
uncovered in other literature (such that no new evidence was being offered), or merely
reported anecdotal evidence.
The methodology Ms Martins Pereira deployed to identify relevant papers is detailed at
the beginning of Annex A.
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Overview
The literature is limited: There is not a large amount of literature credibly assessing the
impact of insolvency and creditor rights regimes on NPL incidence or resolution.
This is not because impact is not expected: it is clear from the existing academic and
policy literature that insolvency and creditor rights regimes are expected to have an impact
on the incidence and resolution of NPLs. On the resolution side, for example, there is a
large literature reporting concerns that inefficiencies in insolvency and debt enforcement
processes have impeded NPL resolution in the recent post-crisis period.1 This link is fairly
intuitive2: non-performing loans can be dealt with by being sold (directly to the market, or
to an AMC), collected (enforced) or restructured 3 or written-off – technically only
collection relies on recourse to formal debt enforcement or insolvency procedures, but as
others have emphasised the price of any sale and the parties’ incentives to engage in any
restructuring will be informed by collection prospects and the tax treatment received after
collection is completed.4 It is also clear from the existing literature that the management of
NPLs is economically important: there is a very large body of literature on the adverse
macroeconomic consequences of high NPL levels.5
The relative paucity of the literature seems to me to be most likely a reflection of
methodological challenges, rather than of the unimportance of the subject matter. At the
cross-country level, attempts to draw links between the legal environment and NPL levels
are complicated by the fact that the relationship between creditor rights and the incidence
of NPLs is ambiguous: one might expect stronger creditor rights to reduce the likelihood
of default, but on the other hand such rights might expand the types of borrowers to which
banks are prepared to lend, such that the aggregate effect is an increased risk of default.6
The kind of coarse variables that tend to be used in cross-country work (to enable
comparability) do not seem very well suited to exploring this complexity.7 Given that there
1 See, e.g. Council of the European Union, Report of the FSC Subgroup on Non-Performing Loans, EF
113 ECOFIN 481, 9854/17, Brussels 31 May 2017, Ch.4; N. Jassaud and K.H. Kang, “A Strategy for
Developing a Market for Nonperforming loans in Italy”, IMF Working Papers, February 6 2015 Pt III;
A. Krueger and A. Tornell, “The Role of Bank Restructuring in Recovering from Crises: Mexico 1995-
98”, 24; see also the qualitative literature detailed below. 2 The expected link between creditor rights and the incidence or occurrence of NPLs is more complex.
As explained below, text to n 6, the overall expected effect of creditor rights on the incidence of NPLs
appears ambiguous, given that stronger creditor rights might at once reduce the likelihood of default
but also expand the range of borrowers to whom banks are prepared to lend to encompass ‘riskier’
borrowers. 3 J. Garrido, “Insolvency and Enforcement Reforms in Italy”, IMF Working Paper 16/134, 4. 4 ibid 6. On the impact of recovery times on the price at which non-performing loans can be sold in the
secondary market, see also L. G. Ciavoliello et al, “What’s the value of NPLs?”, Bank of Italy Notes
on Financial Stability and Supervision, No.3, April 2016. 5 See e.g. IMF, “Policy Options for Tackling Non-Performing Loans in the Euro Area”, July 10 2015,
Section B, and European Banking Coordination “Vienna Initiative”, Working Group on NPLs in
Central, Eastern and Southeastern Europe, March 2012, Section 2.5, reviewing some of the literature;
see also N. Klein, “Non-Performing Loans in CESEE: Determinants and Impact on Macroeconomic
Performance”, IMF Working Paper WP/13/72; M. Balgova, M. Nies and A. Plekhanov, “The economic
impact of reducing non-performing loans” EBRD Working Paper No. 193, Section 2; G. Impavido,
C.A. Klingen, and Y. Sun, “Non Performing Loans and the Macroeconomy”, Ch 2 of Working Group
on NPLs in Central, Eastern and Southeastern Europe, March 2012. 6 T. Jappelli, M. Pagano, and M. Bianco, “Courts and Banks: Effects of Judicial Enforcement on Credit
Markets” (2005) 37(2) Journal of Money, Credit, and Banking 223; see also on this complexity Kruger
fn 1 at 24, and J.F. Houston, C. Lin, P. Lin and Y. Ma, “Creditor rights, information sharing, and bank
risk taking” (2010) 96 Journal of Financial Economics 485. 7 See Armour et al in Dahan (ed) Research Handbook on Secured Financing in Commercial
Transactions (Elgar 2015).
are many determinants of NPL levels (macroeconomic, bank-level, institutional),8 the usual
problem of omitted variables also seems particularly acute.9 (There are also more basic
problems, including that NPLs are measured differently across countries, that data on NPL
levels may to varying degrees be unreliable,10 and that a change in NPL levels does not
necessarily imply any change in default levels or the resolution of problem loans, since
NPL levels (which are typically reported as the ratio of non-performing to total loans) can
be reduced simply by the issue of new loans11). Within-country studies, which can exploit
much more granular data, appear more promising – particularly where focused on the effect
of a change in the legal environment12 – but creditor rights are often reformed in the context
of peaks in NPL levels, and where this occurs the legal change will likely be one of a battery
of reforms, such that the effect of the legal change may be difficult to isolate.
Perhaps for these reasons, there are relatively few quantitative studies, both at the
cross-country and within country level, though there is a considerable amount of
qualitative evidence of the impact of the legal environment on NPL resolution.
Nevertheless, there is some relevant quantitative evidence of impact. These are the key
findings (each of which are discussed and analysed in the detailed review that follows this
summary):
• Countries that adopt policies for the active reduction of NPLs (including by facilitating
transfers to asset management companies, or reforming insolvency or debt enforcement
regimes, or changing rules on write-offs) did no better, in terms of growth outcomes,
than countries who responded to high NPL levels by encouraging growth in new loans,
but both groups did better than countries that took no action to address a high NPL ratio
(Balgova et al)*;
• In a cross-country study of banks in the MENA region, stronger legal rights, as
measured by a “legal rights index” that appears to have been derived from an earlier
version of the Doing Business “Getting Credit” report, were found to be associated with
lower NPL levels (Boudriga et al);
• Slower contract enforcement, as measured by the WB Doing Business “Enforcing
Contracts”, is associated with an increase in NPL levels in European banks (Chiorazzo
et al);
• European banks in jurisdictions with higher insolvency process effectiveness, as
measured by the recovery rate reported in the Doing Business Resolving Insolvency
indicator, experienced lower increases in NPLs in the post-crisis period (Cucinelli et
al);
• The better the “pre-insolvency efficiency” of an EU Member States’ laws, i.e. the extent
to which the law facilitates restructuring outside a manager-displacing insolvency
procedure, the faster the rate of NPL adjustment, i.e. the speed at which NPLs returned
to normal levels, in the post-crisis period (Carcea et al)*;
• Increasing judicial efficiency, measured by length and backlog of civil suits, increases
mortgage default rates (Jappelli et al)* - this is however found to be in line with the
authors’ theoretical model, which demonstrates that where judicial efficiency increases,
8 See Klein above fn 5; M. Nkusu, “Nonperforming Loans and Macrofinancial Vulnerabilities in
Advanced Economies”, IMF Working Paper 11/161; R. Beck, P. Jakubik and A. Piloiu, “Non-performing
loans: What matters in addition to the economic cycle?”, European Central Bank Working Paper
1515/February 2013. 9 Consistently with this, see Report of the FSC Subgroup on Non-Performing Loans, fn 1 above at [34]. 10 “Vienna Initiative” fn 5 above at 13. 11 See Balgova et al above fn 5 at 3; “Vienna Initiative” above fn 5 at [28]. 12 John Y. Campbell, Tarun Ramadorai and Benjamin Ranish, “Impact of Regulation on Mortgage Risk:
Evidence from India”, working paper, September 2014 at 13.
such that credit rationing is relaxed, riskier borrowers will be permitted to borrow –
leading to an increase in interest rates and default rates;
• While an improvement in out-of-court enforcement rights for banks had some positive
impact on the cumulative losses associated with delinquent loans in India, a far larger
impact was caused by a change in the regulatory classification of loans as non-
performing, bringing forward the classification of NPLs from 180 days delinquent to
90 days delinquent (Campbell et al)*; and
• The establishment of Debt Recovery Tribunals designed to reduce delays associated
with loan enforcement in the Indian courts reduced loan delinquency by 3-11%
(Visaria)*.
In the above list I have starred those studies that seem to me to be the most reliable of those
reviewed. Some of the un-starred studies are limited by some of the methodological
challenges I identify above. I have noted the limitations of each study after reporting its
findings.
The way forward: if the objective of the project is to produce a Viewpoint-style overview
document, a promising way to proceed may be to select some of the literature below, and
report this in a document alongside some of the evidence focused on in the previous
Viewpoint on the impact of insolvency and creditor rights regimes more generally. The
latter body of literature is obviously directly relevant to our NPL inquiry, given the
expected impact of the insolvency and creditor rights regime on NPL resolution. Thus, for
example, literature showing how some improvement in the design or operation of an
insolvency procedure increased returns to creditors might helpfully be discussed in the NPL
context.
In drawing on this literature, it may be helpful to make the following points:
• the general insolvency/creditor rights framework may not be equipped to handle
an enormous influx of cases in a crisis-context13, such that where NPLs have
risen rapidly the general framework may have to be complemented by other
measures (such as an asset management corporation that can take loans but delay
enforcement14);
• more generally, strengthening the rights of banks or any other creditor
constituency on borrower default may not lead to a value maximising outcome,
and the possibility of this will have ex ante effects that must be borne in mind
when evaluating the merits of a particular reform.15
Other literature
In the below I have focused on literature that offers some new evidence of the impact of
insolvency and creditor rights regimes, or the legal environment more generally, on bank
NPLs. I have not included literature that merely asserts a relationship between creditor
rights or the broader legal environment and NPLs, without producing evidence, nor have I
included literature that offers an overview of the literature on the determinants of NPLs or
13 L. Laeven and T. Laryea, “Principles of Household Debt Restructuring”, IMF Staff Position Note,
SPN/09/2015, June 26, 2009 3; see also as an illustration Jassaud and Kang above fn 1 at 15. 14 ibid, noting also the possible disadvantages with government intervention; on AMCs, see also C.
Bergstrom, P. Englund and P. Thorell, “Securum and the Way out of the Banking Crisis”, May 2003.
Of course, if the enforcement/creditor rights framework remains weak, mere transfers to AMCs will be
insufficient to achieve resolution: Vítor Constâncio, “Resolving Europe’s NPL burden: challenges and
benefits” 3 February 2017. 15 Vikrant Vig, “Access to Collateral and Corporate Debt Structure: Evidence from a Natural
Experiment” (2013) 68(3) Journal of Finance 881.
on the range of strategies that can be used to manage high NPL levels. This overview
literature is however very helpful as an introduction to the subject. I found the following
particularly helpful:
• S. Claessens, “Policy Approaches to Corporate Restructuring Around the World: What
Worked, What Failed?” in Corporate Restructuring: International Best Practices
(World Bank).
• T. Laryea, “Approaches to Corporate Debt Restructuring in the Wake of Financial
Crises”, IMF Staff Position Note, January 26, 2010.
• Council of the European Union, Report of the FSC Subgroup on Non-Performing
Loans, EF 113 ECOFIN 481, 9854/17, Brussels 31 May 2017.
• Bergthaler, “Tackling Small and Medium Sized Enterprise Problem Loans in Europe”,
IMF Staff Discussion Note, March 2015 15/04.
• L. Laeven and T. Laryea, “Principles of Household Debt Restructuring”, IMF
Staff Position Note, SPN/09/2015, June 26, 2009.
Kristin van Zwieten
Oxford, September 2017
EVIDENCE OF IMPACT
A. CROSS-COUNTRY EVIDENCE
Quantitative cross-country evidence
1. A. Boudriga, N.B. Taktak, and S. Jellouli, “Bank specific, business and institutional
environment determinants of banks nonperforming loans: Evidence from MENA
countries” (Economic Research Forum Working Paper, September 2010)
Available at: http://erf.org.eg/publications/bank-specific-business-institutional-