Bank Quality, Judicial Efficiency, and Loan Repayment Delays in Italy

• Published date: 01 August 2020
• Author: FABIO SCHIANTARELLI,MASSIMILIANO STACCHINI,PHILIP E. STRAHAN
• Date: 01 August 2020
• DOI: http://doi.org/10.1111/jofi.12896

THE JOURNAL OF FINANCE •VOL. LXXV, NO. 4 •AUGUST 2020

Bank Quality, Judicial Efficiency, and Loan

Repayment Delays in Italy

FABIO SCHIANTARELLI, MASSIMILIANO STACCHINI,

and PHILIP E. STRAHAN∗

ABSTRACT

Italian firms delay payment to banks weakened by past loan losses. Exploiting Credit

Register data, we fully absorb borrower fundamentals with firm-quarter effects. Iden-

tification therefore reflects firm choices to delay payment to some banks, depending

on their health. This selective delay occurs more where legal enforcement of collateral

recovery is slow. Poor enforcement encourages borrowers not to pay when the value

of their bank relationship comes into doubt. Selective delays occur even by firms able

to pay all lenders. Credit losses in Italy have thus been worsened by the combination

of weak banks and weak legal enforcement.

THE LONG AND DEEP RECESSION after the financial and foreign debt crises in

Europe has left a legacy of nonperforming loans on Italian banks’ balance

sheets. In December 2015, bad loans equaled about 200 billion, approximately

11% of the total amount of loans outstanding (18% if we include other troubled

loans not written off). Unlike other recent banking problems, where losses were

concentrated in real estate or sovereign debt exposure, close to 80% of these

bad debts came from bank lending to nonfinancial businesses (Bank of Italy

(2016)).1

∗Fabio Schiantarelli is with Boston College and IZA, Massimiliano Stacchini is with the Bank

of Italy, and Philip E. Strahan is with Boston College and NBER. None of the authors received

financial support other than their normal salaries from their institutions. The article has been

reviewed by the Bank of Italy, but the views expressed in it are the authors’ alone and do not

necessarily represent those of the institutions with which they are affiliated. We are grateful to

Massimiliano Affinito; Giorgio Albareto; Alberto Alesina; Fabio Braggion; Francesco Columba;

Riccardo De Bonis; Emilia Bonaccorsi di Patti; Francesco Giavazzi; Luigi Guiso; Harry Huizinga;

Francesco Manaresi; Paola Sapienza; Alfonso Rosolia; Ricardo Serrano-Padial; Paolo Sestito; Al-

berto Zazzaro; and seminar participants at the Bank of Italy, Bocconi University, Boston College,

University of Chicago, the Federal Reserve Bank of New York, Georgia Tech, the University of

Illinois at Champaign-Urbana, Notre Dame, Nova School of Business, the Bank of Portugal, SAIF,

Tilburg University, Tsinghua University, the University of Western Ontario, Drexel University,

and the NBER Summer Institute for helpful comments. We also thank Marco Errico, Ana Lariau,

and Danilo Liberati for the helpful research assistance.

Correspondence: Philip Strahan, Finance Department, Boston College, 140 Commonwealth

Avenue, 324 Fulton Hall, Chestnut Hill MA 02467; email: philip.strahan@bc.edu.

1The stock of bad loans fell to 160 billion by 2017 but remains substantial.

DOI: 10.1111/jofi.12896

C2020 the American Finance Association

2139

2140 The Journal of Finance R

In this paper, we show that the combination of weak bank balance sheets

and inefficient legal enforcement leads borrowers to delay debt repayment.

Borrowers selectively delay payment to banks already weakened by past bad

loans while continuing to pay healthier banks. Ineffective legal enforcement

exacerbates this problem, as the magnitude of our estimates is larger in areas

of Italy where it takes longer to resolve disputes about the recovery of collateral,

while accumulated bad loans do not have a significant effect in areas where

legal enforcement is quick and efficient. For example, where legal inefficiency

is high (top quartile of its distribution), a one-standard-deviation increase in a

bank’s past bad loans increases payment delays by about 50%, relative to the

unconditional mean.2

Our data allow us to capture a firm’s decision to delay repayments at the

level of the bank-borrower. Obviously, there can be many reasons for a delay in

loan repayment, ranging from firm financial distress to strategic considerations

about how such behavior may affect the firm’s ongoing or future relationship

with lenders. With regard to the latter, a firm will be trading off the short-term

gain of retaining financial resources (i.e., by not paying now) against the poten-

tial future loss of impairing their relationships with their current lender(s) or

potential future lender(s). Which side of this trade-off dominates may depend

on the lender’s financial health and internal enforcement capacity, on the firm’s

ability to borrow elsewhere, and on the quality of the institutional environment

(which affects banks’ ex post ability to recover collateral or otherwise force re-

payment through the judicial process). Everything else equal, a firm ought to

be more likely to delay repayment to weaker banks because the expected value

of continuing the relationship is smaller. Bond and Rai (2009) show formally

that concern about a lender’s long-run viability can lead to a “borrower run” in

which repayment incentives for individual borrowers depend on the payments

of other borrowers. As the lender fails given enough defaults, this externality

can lead to an equilibrium in which borrowers default because they expect

other borrowers to default. The incentive to delay debt payment may also be

enhanced if weak banks are less able to enforce contracts.

Totest how bank health affects repayment behavior, we exploit a unique data

set, namely the Italian Credit Register, which contains detailed information on

all bank loans above 30,000 euros. The data include information on repayment

delays and the degree of impairment of loans, including those that fall short

of being formally classified as “bad” by the bank. The solvency of Italian firms

and the quality of loans have been strongly affected by the double-dip recession

following the global financial crisis of 2007 to 2009 and the sovereign debt crisis

of 2010 to 2011.3

We match these data to bank balance sheets reported to the Bank of Italy,

as well as to borrower balance sheet data collected by the Balance Sheet

2The Wall Street Journal reports that, “The snail’s pace of Italy’s courts throws sand into the

wheels of the economy in myriad ways. Banks struggle to resolve bad loans because bringing

deadbeat debtors to court takes by far the longest in Europe.” (Zempano (2014))

3In seven years, manufacturing firms lost 17% of their productive capacity and net job destruc-

tion reached almost one million.

Bank Quality, Judicial Efficiency, and Loan Repayment Delays in Italy 2141

Register. (These data have been provided by lenders for information-sharing

purposes since 1983.) The data can also be matched to measures of local judicial

(in)efficiency in recovering collateral that we calculate using information from

the Italian Ministry of Justice.4Although civil law and procedures are formally

the same across Italy, the real-world effectiveness of the court system varies

widely across local jurisdictions (Carmignani and Giacomelli (2009), Giacomelli

and Menon (2013)). We exploit this regional and subregional variation to test

how legal enforcement affects repayment behavior. Based on the discussion

above, we expect firms to be more willing to delay loan repayment where it is

harder for lenders to protect their interests through the courts.

As in other studies, we exploit the fact that many Italian firms borrow from

multiple banks. This feature allows us to include firm-specific, time-varying

effects that absorb fundamentals that may drive firm decisions to delay loan

repayment. Our identification thus comes solely from variation in bank char-

acteristics, characteristics of the bank-firm relationship, and, importantly, on

the efficiency of the court system. In other words, we test how the same firm

behaves with respect to different banks, depending on the strength of the bank’s

balance sheet, the local judicial environment, and the nature of the past bank-

firm relationship.

The results suggest that bank balance sheet strength (particularly past bad

loans) affects the probability of a delay in loan repayment. In our basic specifi-

cation, the stock of past bad loans increases the probability of borrower delays.

This effect increases as legal efficiency decreases. Thus, on average banks with

weaker balance sheets due to past (and noncollectible) bad loans experience

more future defaults (in the form of temporary delays in repayment, many of

which ultimately become permanently impaired). That is, we find that borrow-

ers withhold payment to weak banks. To allay concerns that our results reflect

reverse causality, whereby bank balance sheet health is reduced by borrower

payment delays, as well as concerns about omitted variables, we construct an

instrument for bank weakness that depends only on a bank’s 2007 lending

portfolio shares (across sectors and provinces), combined with losses based on

aggregate loan outcomes at the sector and province levels (excluding, for each

firm, loans in the sector-province to which the firm belongs). These results are

qualitatively similar to those from our baseline models. In addition, we verify

that late repayment harms lenders, as their profits decrease with past levels

of payment delays.

Are distressed borrowers merely selecting which banks to pay by allocating a

fixed but limited cash-flow budget across lenders? Or, are borrowers paying less

than they otherwise would because lenders are weak? We find that some of the

payment delays are truly strategic in that borrowers pay less than they other-

wise would because one or more of their lenders is distressed. Specifically, we

reestimate the original model stratified by borrower health, and we show that

even the safest firms still choose not to pay some of their banks because of the

4The data can be downloaded from the web page of the Italian Ministry of Justice

(https://reportistica.dgstat.giustizia.it/).