Securitization and credit quality in the European market

• Author: Alper Kara,David Marques‐Ibanez,Steven Ongena
• Date: 01 March 2019
• Published date: 01 March 2019
• DOI: http://doi.org/10.1111/eufm.12168

407Eur Financ Manag. 2019;25:407–434. wileyonlinelibrary.com/journal/eufm © 2018 John Wiley & Sons, Ltd.

DOI: 10.1111/eufm.12168

ORIGINAL ARTICLE

Securitization and credit quality in the European

market

Alper Kara

1

|

David Marques-Ibanez

2

|

Steven Ongena

3,4,5,6

1

Business School, University of

Huddersfield, Huddersfield, UK

Email: a.kara@hud.ac.uk

2

Financial Research Division, European

Central Bank, Frankfurt am Main,

Germany

Email: David.Marques@ecb.int

3

Department of Banking and Finance,

University of Zürich, Plattenstr, Zurich

4

SFI

5

KU Leuven

6

CEPR

Email: steven.ongena@bf.uzh.ch

Funding information

Ongena acknowledges financial support of

ERC ADG 2016–GA 740272 lending.

Abstract

We assess the effect of securitization activity on relative credit

quality employing a uniquely detailed dataset from the euro-

denominated syndicated loan market. We find that at issuance,

based on observable characteristics, banks do not seem to select

and securitize loans of lower credit quality. Following

securitization, the credit quality of borrowers whose loans are

securitizeddeteriorates more than those in the control group. We

find that poorer performance by borrowers of securitized loans

seems to be connected to banks’reducedmonitoring incentives.

Our results are supported by two additional methodologies and

robust to controlling for predetermined borrower–lender

matching.

KEYWORDS

credit risk, European market, securitization

JEL CLASSIFICATION

G21, G28

We are very grateful to John A. Doukas, two anonymous referees as well as to Peter Lindner, Giovanni Dell’Ariccia,

Alfonso Del Giudice, Soledad Martinez Peria, and the colleagues at the IMF's Macro Financial Research division for their

kind hospitality and fruitful discussions. We also thank Rajeev H. Dehejia, Jose Luis Peydro, Alex Popov, John Rogers,

and Joao Santos for helpful comments or discussions. Our thanks also to participants at seminars held at the European

Central Bank (ECB), World Bank, Loughborough University, University of Hull, the 6th IFABS 2014 Lisbon conference

on ‘Alternative Futures for Global Banking: Competition, Regulation and Reform,’the X Seminar on ‘Risk, Financial

Stability and Banking’organized by the Banco Central of Brazil, Wolpertinger 2016 conference, and 2016 Portsmouth–

Fordham conference on ‘Banking and Finance’for their useful comments and discussions. We are also most grateful to

Raffaele Passaro and Luiz Paulo Fichtner for their help with the initial data and for providing us with technical suggestions.

We would also like to thank Oliver Goß and Priti Thanki from Standard and Poor's and, in particular, Jean-Paul Genot for

their invaluable help finding data on securitized syndication credit from all major securitization trustees in Europe. This

work was completed while David Marques-Ibanez was at the IMF's Macro Financial Division of the Research Department.

The opinions expressed in this paper are those of the authors only and do not necessarily reflect those of the IMF or the

ECB. Ongena acknowledges financial support of ERC ADG 2016 –GA 740272 lending.

2

|

KARA ET AL.

K

408

1

|

INTRODUCTION

Banks generate proprietary information and tend to have superior knowledge on the credit quality of

the loans they originate. As a result, banks might have an incentive to securitize loans of lower credit

quality to unsuspecting investors (Gorton & Pennacchi, 1995). Largely for this reason, securitization

has been perceived as a major contributing factor to the 2007–2009 financial crisis (Financial Crisis

Inquiry Commission, 2011).

Most of the recent empirical ev idence on the impact of securitization is US-base d and focuses on

the mortgage markets. It sug gests that prior to the financ ial crisis banks securitiz ed riskier

mortgages (Elul, 2015; Kraine r & Laderman, 2014), and that s ecuritization led to poorer qu ality

mortgages (Keys, Mukherje e, Seru, & Vig, 2011; Purnanand am, 2011). Evidence on the effec t of

securitization on the corpor ate loan market is more limited and it s findings are contradicto ry. A

number of recent studies find t hat the credit quality of secu ritized corporate loans is no different

than non-securitized loan s (Benmelech, Dlugosz, & Ivas hina, 2012; Shivdasani & Wa ng, 2011;

Wang & Xia, 2015). In contrast, B ord and Santos (2015) find that securitization led to poorer quality

corporate loans.

We contribute to this latter litera ture by providing the first evide nce on the European corporate

loan market. The euro-denom inated collateralized lo an obligations (CLO) marke t, is a good

laboratory to assess the imp act of securitization on cred it quality on an alternative institution al

setting as there are significa nt differences between the US a nd European securitizatio n markets.

Firstly, the growth of European secu ritization has been relatively r ecent and swift (Kara, Marques-

Ibanez, & Ongena, 2016). It started tim idly in the late 1990s and developed signi ficantly only from

2004 to 2007. On the contrary, in the Un ited States the development of secur itization markets has

been much more continuous ov er time since the late 1960s. The sudden appearance of sec uritization

in Europe allows for a clearer ass essment of its effects on the bank ing behavior and the financial

system. Secondly, unlike in the U nited States the developm ent of the securitization marke t in

Europe has not been driven by go vernment-sponsored in stitutions such as Fannie M ae and Freddie

Mac. Therefore, our results ca nnot be ascribed to the impact of gove rnment intervention via s uch

institutions.

We also contribute to the literature using a novel dataset obtained directly from securitization

trustees operating in the European Union (EU). We construct our dataset by getting access to the

portfolios of the majority of euro-denominated CLOs so we are able to form a representative picture of

the market. Our detailed loan level dataset allows us to distinguish among all syndicated loans, those

that were eventually securitized.

In practical terms, we contrast the credit performance of non-financial companies whose loans are

securitized versus non-financial companies whose loans are not securitized. Loans extended to these

companies, headquartered in the euro area, are granted by euro-area banks between 2005 and 2007. We

track changes in borrowers' credit quality as measured by expected default frequencies (EDF) after

issuance. We then examine the link between future changes in EDF and loan securitization controlling

for a set of loan, borrower, and lender characteristics. As a complementary methodology, we also use

propensity score matching to compare the credit risk of corporates whose loans which, prior to

securitization, were very similar.

We find that at loan origination, based on observable characteristics, banks do not select and

securitize corporate loans of lower credit quality. Following securitization, the credit quality of

borrowers whose loans are securitized deteriorates by more than those in the control group.

Distinguishing between collateralized and uncollateralized loans, we find that poorer performance by

companies with securitized loans is linked to banks’reduced monitoring incentives. Our findings are

r

m

f

u

c

2

T

i

m

N

s

2

2

s

s

L

a

f

a

a

S

m

t

b

c

c

e

t

(

o

n

l

s

q

I

1

V

2

2