The Role of Collateral in the Credit Acquisition Process: Evidence from SME Lending

• DOI: http://doi.org/10.1111/jbfa.12207
• Published date: 01 May 2016
• Date: 01 May 2016
• Author: Fábio Duarte,José Paulo Esperança,Ana Paula Matias Gama

Journal of Business Finance & Accounting

Journal of Business Finance & Accounting, 43(5) & (6), 693–728, May/June 2016, 0306-686X

doi: 10.1111/jbfa.12207

The Role of Collateral in the Credit

Acquisition Process: Evidence from SME

Lending

F´

ABIO DUARTE,ANA PAULA MATIAS GAMA AND JOS ´

EPAULO ESPERANC¸A

∗

Abstract: This study tests the simultaneous impact of observed characteristics and private

information on debt term contracts in a multi-period setting, using a dataset of 12,666 credit

approvals by one major Portuguese commercial bank during 2007–2010. The main results show

that borrowers with good credit scores that know they have a high probability of success and

are unlikely to default are more willing to pledge collateral in return for a lower interest rate

premium (IRP). Furthermore, lenders tailor the specific terms of the contract, increasing both

collateral requirements and the IRP from observed risk, for borrowers operating in riskier

industries and with less credit availability. The results are robust to controls for joint debt terms

negotiation and the degree of collateralization offered by the borrower.

Keywords: SME, entrepreneurship signaling, private information, collateral

1. INTRODUCTION

A central characteristic of small and medium-sized enterprises (SMEs) is their depen-

dence on bank credit for external financing (Degryse and Van Cayseele, 2000). Since

SMEs tend to provide lower financial reporting quality, asymmetric information and

agency costs underlie their inadequate financing; therefore, banks issue restrictive

loan term contracts to reduce information risk and their default exposure (Garc´

ıa-

Teruel et al., 2014). Thus, in the design of loan term contracts, collateral assumes a

key role as a risk management instrument, especially for SME lending (Bonfim, 2005).

To analyze the role of collateral in debt term contracts, Han et al. (2009) propose

a “sorting by signaling and self-selection” (SBSS) model that emphasizes two unique

∗The first author is from Research Center in Business Sciences (NECE)/University of Beira Interior

(UBI). The second author is from University of Beira Interior (UBI)/Management and Economics

Department. The third author is from ISCTE Business School/Businsess Reseach Unit/ISCTE – Instituto

Universit´

ario de Lisboa. The authors appreciate financial support from the Research Unit in Business

Sciences (NECE), financed by the Portuguese Foundation for Science and Technology (FCT), pluriannual

programme for R&D units. Sandra Pinto provided database access. The authors also gratefully acknowledge

helpful comments from Liang Han and anonymous referees and conference participants at the Research

in Entrepreneurship and Small Business conference (RENT), Vilnius, 21–22 November 2013, and the

International Network of Business and Management Journals Conference (INBAM), Barcelona, Spain, 24–

27 June 2013. (Paper received December 2014, revised revision accepted April 2016).

Address for correspondence: Ana Paula Matias Gama, University of Beira Interior (UBI), Management and

Economics Department (DGE), Portugal.

e-mail: amatias@ubi.pt

C

2016 John Wiley & Sons Ltd 693

694 DUARTE, GAMA AND ESPERANC¸A

views on how collateral can mitigate credit risk better than sorting by observed risk

(SBOR) (e.g., Berger and Udell, 1990) or sorting by private information (SBPI) (e.g.,

Bester, 1985). In the SBOR model, borrowers’ risk types are observable, so lenders

require collateral and higher interest rate premiums (IRP) on the basis of borrowers’

observed characteristics (demand-side argument). In the SBPI model, good borrowers

instead offer collateral as a reliable signal of their low risk, and in return, they expect

a loan contract with lower IRP (supply-side argument). Combining features of these

existing paradigms, the SBSS model introduces borrowers’ observable characteristics

(i.e., SBOR) to design an incentive-compatible menu of loan contracts that works as a

self-regulating mechanism (i.e., SBPI). With this model, both demand- and supply-side

factors guide the design of loan contracts negotiated with banks.

In testing the simultaneous impacts of observed characteristics and private infor-

mation on debt term contracts, we extend Han et al.’s (2009) model in three ways.

First, because loan term contracts (i.e., collateral and IRP) are jointly determined

(Brick and Palia, 2007), we test the signaling value of collateral and consider the

determinants of IRP, with the assumption that they can be determined endogenously,

as predicted by the SBSS model (Godlewski and Weill, 2011). Second, our analysis

spans a multi-period setting, which allows us to control for the survivor bias effect by

including not only surviving firms but also default firms. Third, unlike previous studies,

our unique dataset permits us to conduct the analysis with a continuous variable for

collateral. Thus, we test the SBSS model by controlling not just for whether collateral

is provided but also for the amount pledged. These extensions in turn lead to three

main contributions to extant literature.

First, though some empirical studies examine the role of collateral in mitigating

informational asymmetries in loan contracts (e.g., Jimenez et al., 2006; Menkhoff

et al., 2006; Voordeckers and Steijvers, 2006; Berger et al., 2011), few explore the

interdependencies of collateral and IRP. Those that do generally use US data (e.g.,

Brick and Palia, 2007; Han et al., 2009), which reflect market-based financial systems

(La Porta et al., 1998). The current study instead analyzes the role of collateral in

loans granted to SMEs in Portugal, a country characterized by a bank-based system

and an economy dominated by SMEs. This issue is particularly relevant considering

the structural differences in the size and importance of US and EU banking sectors

relative to their respective overall financial sectors (Schildbach and Wenzel, 2013).

Specifically, in the US, financial firms other than banks have grown more quickly in

recent decades, to the extent that the so-called shadow banking system overtook the

traditional banking system in size by the 1990s. In Europe, most financing still moves

through traditional credit institutions so, even though the economies are roughly

comparable in size, the EU’s banking sector has more than four times as many total

bank assets as the US sector, which has important repercussions for how SMEs find

external funding. European banks’ relatively limited capacity to provide credit to

private and public sectors thus determines the economy’s overall financial strength far

more in Europe than in the US, where actors other than banks are equally relevant.1

1 The differences between the US and EU banking markets grew after the financial crisis. Previously, both

US and European banks enjoyed profits; currently, only US banks do so. In addition, US lending is growing,

and loan loss provisions have returned to 2007 levels. In contrast, loan growth remains weak in Europe, and

loan loss provisions remain high. European banks’ greater need to raise their capital ratios, together with

their higher levels of deleveraging and shrinking of firm sizes, have put them at a competitive disadvantage

compared with their US counterparts (Schildbach and Wenzel, 2013). For example, whereas in the US the

C

2016 John Wiley & Sons Ltd

THE ROLE OF COLLATERAL IN THE CREDIT ACQUISITION PROCESS 695

In Portugal, this issue is critical: it is enduring the third worst credit conditions within

the EU (Gaspar, 2012).

In addition, the effect of creditor protection on bank lending to small businesses is

largely unexplored, even though small businesses constitute a crucial sector of virtually

all economies (Banco de Portugal, 2013). The differences in Portuguese and US

bankruptcy codes are best viewed from the perspective of super-priority financing,

automatic stays and the legislation goals in general. In the US, both liquidation

(Chapter 7) and reorganization (Chapter 11) are legal under the bankruptcy code

and, thus, the country is characterized as debtor friendly. In contrast, Portugal’s 2004

bankruptcy code not only moves away from debtor protection but also explicitly

emphasizes the liquidation of firms over their rescue and recovery. As part of this

protection of creditors’ rights, Portugal’s legislation does not allow for automatic stays

from creditors whereas, under US legislation, Chapter 11 allows firms in reorganiza-

tion to postpone all repayments of capital and interest until reorganization is complete

as a way of preserving the company as an operating concern. Finally, super-priority

financing under the Portuguese system only occurs if an insolvency plan expressly

provides for it, or if the firm has any collateral assets free to meet this potential claim.

In contrast, under the US system, firms can use super-priority financing without such

limitations.

Second, our sample covers 2007–2010, so this study provides empirical evidence of

the effect of macro-economic conditions on loan term contracts during credit crunch

and recession periods. We control for this effect by balancing the number and amount

of loans granted before and after the 2008 financial crisis and using a censored model

in addition to the SBSS model to predict the determinants of the amount of collateral

required (Lambrecht, 2009).

Third, to the best of our knowledge, this study is the first to focus on the

Portuguese loan market. According to a survey conducted by the European Central

Bank (ECB), SME access to finance in the euro area has remained broadly unchanged.

Substantial cross-country differences are evident, however, and, in general, financing

is more difficult for SMEs than for large companies (ECB, 2015).2Using 12,666 credit

approvals by one of the largest commercial banks in Portugal, we show that risky

borrowers must provide collateral and pay a higher IRP to obtain a loan, in line with

the SBOR (demand-side) model (i.e., the moral hazard [MH] effect). Sharing private

information between lenders and borrowers also exerts an impact, such that borrowers

ratio of bank non-performing loans to total gross loans decreased between 2009 and 2014, from 5% to

1.98% (i.e., the level of 2007), in the EU (and Portugal), the ratio increased annually from 4.7% (4.82%)

to 8.19% (11.19%), representing an increase in the value of the ratio greater than four times the level in

2007. Although the ratio for Portugal is higher than the EU average, it is lower than those recorded in other

EU economies, such as Greece, Spain, Italy and Ireland, which suffered the severest impacts of the financial

crisis on their national bank activity. Regarding lending activity to the private sector, World Bank statistics

show that US domestic credit to the private sector (as a percentage of GDP) has increased each year since

2011, rising from 182.35% to 194.75% in 2014. By contrast, lending activity in the EU has decreased since

2009. In 2014, the comparable ratio was 100.77% in the EU and 129.8% in Portugal (World Bank Data

Indicators).

2 According to the Survey on the Access to Finance in the Euro Area (SAFE, 2015), 42% of SMEs operating

in European countries in 2012 (47.2% in Portugal) suffer from difficulties when they try to access finance

(this indicator is not available prior to 2012). Among those SMEs that cited bank loans as (very) relevant in

their financing structure (i.e., 73.9%), less than half had actually negotiated a bank loan recently (within

the past six months). Similar to other euro area countries (cf. Germany), perceptions that access to finance

has deteriorated continued to increase in Portugal between 2009 (the first year this indicator was available)

and 2013.

C

2016 John Wiley & Sons Ltd