SMEs’ soft information and credit rationing in France

Abstract

The decision-making processes seen in business enterprises and other social organizations, often entail soft data and information. Such is a situation regarding the banking industry in France, where soft information plays a vital role in the decision-making processes that grant credit to small and medium enterprises (SMEs). This article assesses how soft information reduces SMEs credit rationing. Based on a questionnaire sent to SMEs’ in France, the study’s sample consists of 296 companies that had applied for credit. The results show that the integration of soft information into the bank’s decision-making process can significantly reduce SMEs credit rationing. Scores tend to replace bankers and their analysis. Nevertheless, this article points out the importance of information resulting from relationships not considered in ratings. More generally, this work motivates organizations to use computerisation as complementary tools and not as substitute to human experience.

Keywords

Credit rationing SMEs bank-SME relationship soft information

Aymen Smondel, is an assistant professor of Finance at the “Université Côte d’Azur”, IAE, Nice received his doctoral degree in 2011 from the university of Paris Dauphine. His research activities cover a wide range of issues concerning corporate finance and especially on SME’s financing, loans application and rationing. He has been regularly participating in evaluating undergraduate and graduate degrees for the HCERES as an expert. In addition, he has served as a reviewer for several scientific journals and international conferences.

1 Introduction

Information asymmetry can lead firms to make biased decisions. A reliable information is then crucial to make good choices. However collecting the most complete set of information is not easy, because information is coming from several sources and many of these sources lie outside the organization. This lack of reliable information is common to all economic actors, but it is particularly vital in the banking industry. Indeed banks need complete information to respond accurately to the credit applications.

“The rate of partial or complete failure in the search for a bank loan rose sharply between 2007 and 2010: It went from just over 5% to nearly 17% (with a complete failure in 7% of cases) and reached 23% in 2010 for high-growth firms”. (Survey on access to finance for SMEs employing at least 10 people, INSEE results N ° 53, Economy - August 2011).

These results published in 2011 show the increased rigidity in access to credit for small and medium-sized enterprises (SMEs). They are considered to be small as well as fragile companies, and therefore vulnerable to financial crises. Furthermore, the recommendations of the Basel Committee incites banks to allocate credits based on scores from accounting and financial data. The use of these scores should reduce the flexibility of bankers in order to meet the demand for credit [54], and more particularly those of SMEs. Indeed, the latter, characterized by their opacity, must overcome the problem of information asymmetry. Berger and Udell [8] assume that the establishment of bank-SME relationship is one of the best solutions. These credit relationships enabled bankers to collect private information beyond accounting and financial information. SMEs require more flexible and personalized treatment, based on a closer relationship.

The intention of this article is to empirically demonstrate that soft information, including non-accounting data, private order and offering a better visibility [42, 49] contributes to the reduction of SMEs credit rationing. SMEs are rationed when they obtain less than they requested. Other variables affecting credit rationing will also be considered within this study: SMEs’ characteristics, bank-SME relationship indicators and the credit terms.

This article is organized into four sections. The first section will present the literature review and the different variables affecting the decision-making process. In the second section, a description of data and variables will highlight the different measures and expected signs. The third section will present the methodology and the main results of the study, while the last section is devoted to conclusions.

2 Literature review

2.1 Credit rationing and SMEs

SMEs plays a significant role in the French economic landscape as key generators of employment and income, and drivers of innovation and growth. In 2011, France had just over 3 million SMEs (99.8% of companies) accounting for 48.7% of salaried employment (in full-time equivalents). They generated 35.7% of turnover and 43.9% of added value excluding taxes. (INSEE: The Preparation of Annual Business Statistics System (Esane), Local knowledge of the productive system (Clap), The Liaisons Financiers Information System (Lifi), 2011). Although SMEs represent a huge part in French economy, they are considered vulnerable, fragile and uncertain when taken alone. They are sensitive to their environment.

In times of crisis, they are the first to be affected. Moreover, they are characterized by their opacity and the lack, the absence or the unreliability of facts available on them. This represents an additional obstacle to their financing. SMEs find it more difficult than large companies to access external funding, and it is difficult for them to obtain all of requested credits [18]. As loans access is a key factor in a SME development, they are particularly concerned by credit rationing. However, in addition to their fragility, their opacity can further limit their access to bank financing. In order to overcome this problem of information asymmetry, [8] consider that strengthening the relation with their bank is one of the best solutions. These credit relationships allow bankers to collect private information, also known as soft information, beyond accounting and financial data. This soft information help bankers to establish a clearer vision of the company and its future. Thus, SMEs need more flexible and personalized processes of their credit applications. A firm is considered rationed if one or more of its credit applications have been totally or partially rejected [15, 35]. Banks can only provide part of requested funds [17, 19].

2.2 Credit rationing and soft information

Soft information is qualitative data that is limited to the judgement and opinion of the person who collects the information [49]. It is generally collected and used by agents directly in contact with SMEs. The information is collected during the entire period of the relationship and is derived from several bank-SMEs interactions [31, 39].

The soft information represents the personal analyses and interpretations of the loan officer based on the general context. As soft information is collected over a long period, it can for example help the loan officer to differentiate the company’s inherent failures from the consequences of economic and financial crises [8]. Consequently, it can provide more visibility to bankers to reduce credit rationing.

Credit rationing has generally been studied in a macroeconomic context and most studies have highlighted its relationship with monetary crises and periods of under-investment. However, this work considers credit rationing as a response to high levels of opacity or risk.

Credit rationing may result from increased competition in the credit market, from a financial crisis or an economic recession [5 , 37]. Especially for SMEs, it may also be a consequence of the recommendations of the Basel Committee on the use of credit scores for decision-making purposes. Indeed, the intentions of these recommendations is to reduce the risk on the credit market, although their implementation has changed the functioning of the credit market by causing credit rationing [46]. The unification of borrower risk assessment methods leads to a generalization of calculated credit scores, leading to greater competition in the credit market. Hence, what matters here is the informational character of the borrower alone.

Godlewski [25] pointed out that the concentration of hard information (accounting and financial data) used in the decision-making process can skew the decision on granting loans, particularly in the case of SMEs. This rationing stems from the lack of existing information regarding the risk of potential borrowers default.

However, banks do not necessarily have all the information - beyond financial and accounting information - necessary to make the right decision. The establishment of a lasting relationship that is the source of private information is considered as a good solution to this lack of visibility [1 , 48] as well as for granting loan [12 , 43].

This credit relationship is very complicated and involves several actors, however in simpler terms, it is assumed that it has only two players: banks and SMEs. Collection, use and interpretation of soft information arising from this relationship is costly. It requires the deployment of significant financial and human resources, as it is collected after several interactions with enterprises [31]. For this reason, banks may favour limited to hard information in their decision-making process. However, the importance of soft information within the credit decision-making process has repeatedly been demonstrated [8 , 53]. The purpose of this study is therefore to empirically demonstrate that soft information reduces credit rationing, leading to the following hypothesis:

H1: Soft information reduces credit rationing.

The credit decision-making process is taken following two steps: granting the credit and fixing the share of the granted credit [17, 19]. The fact that soft information reduces the rationing measured by its actual rate leads to the second hypothesis:

H2: Soft information reduces the actual rate of credit rationing.

2.3 Other determinants of credit rationing

2.3.1 Characteristics of the bank-SME relationship

The strength of the bank-SME relationship is central in order to collecting private and reliable soft information as well as increasing credit allocation [15 , 41]. Maintaining several relationships with several banks can fragilize the relationship with main creditor and favour rationing. Multi-banking can make negotiating credit terms more difficult [13]. The borrower is in a more secure situation because he has other sources of funding and can get the money elsewhere. Putting the main bank in competition and negotiating with it the terms of the contract may harm it and reduce the bank’s profitability [15 , 55]. In fact, a firm that has negotiated with several creditors proves the existence of a weak bank-SMEs relationship [48]. Thus, it may be considered that putting several banks into competition on the same credit can reduce rationing, by offering more sources of finance and by taking advantage of competition in the credit market [15, 31].

However, bank-SME relationship can be altered by different credit history. With bad experiences, the bank can limit credits to SMEs. At this level, it is wise to distinguish the case of a credit without incidents, which strengthens the relationship, as well as the case of a default credit, which forces the bank not to venture with the firm in question. In fact, [51] assume that a firm can be rationed if it has been defaulting on a previous credit. In addition, switching from one credit to several credits increases financing costs and reduces their availability [41].

The strength of bank-SME relationship is generally measured by its duration. A longer duration allows to accumulate more collected information and consequently reduce rationing [15 , 41]. Steijvers and Voordeckers [48] argue that the strength of relationship is a substitute for the guarantee. By accepting restrictive conditions, the borrower gives a positive signal to the lender, showing a strong trust in their company and project [9, 47]. The borrower may also find himself obliged to renegotiate the terms of the credit each time their financial situation changes [9, 47], which implies increased control from the lender [9].

2.3.2 Characteristics of SMEs

Credit rationing for a SME is fundamentally linked to its financial and accounting characteristics. The purpose of this article is to prove that soft information reduces credit rationing without casting doubt on the role of all hard information. Soft information is collected by agents in direct contact with SMEs, resulting from a more personalized relationship. The size of the firm can determine the type, reliability and importance of collected information, and therefore may influence the credit decision [23]. Credit rejection is negatively related to firm size [35] and this can be explained by a higher cost of access to financing for small firms [14, 22], or by increased competition with large companies [18]. On the other hand, soft information incorporates opinions and interpretations of agents who have collected such and are in direct contact with the company, therefore shrinking those grey areas and giving more visibility to the lender. Thus, it is less likely to ration the smallest firms, based on the quality of the information collected [26, 44].

It may be difficult at times to collect reliable private information while making the credit decision and therefore reducing rationing. This is the case for innovative companies that prefer to limit the disclosure of private information. Innovative companies present a double risk: the first comes from its opacity and the second depends on the success of its projects. In addition, these innovative firms engage in riskier projects for greater gain [26]. Innovative firms find it difficult to obtain external financing to support their innovative projects. Indeed, [26] showed that the sensitivity of R&D to cash flows is generally due to the credit rationing of innovative firms. Hall [27] used a large panel of industrial firms in the United States and found a strong effect of cash flows in R&D spending. He interpreted this result by the evidence that innovative firms are financially thwarted.

2.3.3 Credit characteristics

The negotiations surrounding credit terms essentially focus on the guarantee and the interest rate. The combination of these two elements allows the bank to obtain a signal concerning the confidence level of the company within its project. The guarantee and the interest rate mutually replace one another [10, 11]. In fact, when it comes to financing a risky project, the company favours a loan with a high rate and a low guarantee. If successful, the project gains cover the high costs and in the event of failure - the company does not lose part of its assets [48]. As a result, simultaneous guarantee and interest rate fixing may reduce credit rationing [10, 11]. Moreover, credit agreements that use collateral as a means of self-selection can eliminate adverse selection and create a better balance. However, SMEs’ wealth must not be overlooked, because it may be difficult to present enough guarantees even with willingness [10, 51]. The non-risky borrower is ready to present more collateral [28, 38].

As mentioned in the previous paragraph, the simultaneous fixing of the guarantee and the interest rate may reduce credit rationing. The volume of credit granted is significantly related to the rate policy adopted by the banks [2, 52]. In fact, according to [4], the rationing rate is positively linked to the central bank’s key rate. In other words, the high rates hinder the granting of credits. Banks prefer rationing credit rather than raising interest rates [48].

Since interest rates are capped and banks are rationing risky businesses, the total risk incurred by the bank depends on the credit size. The requested amount determines the losses that the bank can assume and it will be taken into account when making the credit decision. There are two opposing theories of this influence: one assumes that credit size favours rationing [32] and that small loans are more profitable [6, 9], and the other assumes that big loans allow to cover costs and thus reduce rationing [14, 45].

The requested amount determines repayment periods and bank-SME relationship. However, this relationship can be influenced by the company size. Credits can be classified into two categories according to their size: larger investment credits reimbursed over a longer period and operating credits reimbursed in shorter terms. Short-term credit is less rationed than long-term credit for several reasons. Short-term credit reduces moral hazard and thus increases the chance for the borrower to obtain credit [7]. Short-term credit can reduce problems related to information asymmetry by strengthening bank control as well as the reliability of collected information and, consequently, increasing the credit availability [40]. In addition, a short-term contract can be used as a signalling instrument for the bank and is considered as a firm evaluation test [3, 21].

3 Data and variables

The data was collected from a questionnaire distributed to SMEs’ managers in 2008. This period of crisis represents a strong framework for this analysis. A lack of confidence was noted during the last subprime financial crisis. Access to finance has become even more difficult and more complicated. Analyzing the information role is more crucial than ever in explaining access to credit. The sample consists of 296 French SMEs operating within several economic sectors.

The questionnaire was sent by e-mail to 6102 SMEs and realized in the form of a logical tree, allowing to target the asked questions according to previous answers. This modality reduces the questionnaire length, decreases response time and thus improves response quality. Moreover, the fact that answers are anonymous and that there is no way to cross-reference such answers with other data, are mentioned, in order to encourage respondents to provide more sincere responses. Seven hundred fifty seven (757) responses (12.4%) are received, of which 296 (4.9%) were exploitable. The responses that demonstrated that the managers did not request any credit in the study period (for example, if the first question is answered in the negative) are eliminated.

The questions found in the questionnaire are summarized in Annex 1. The first section of the questionnaire provides information on the credit application. The second section deals with the credit conditions and terms (interest rate and presented guarantee). The third section is based on the bank-SME relationship (number of banks working with the company, the relationship duration and the information nature used for the credit granting). The fourth section presents firm characteristics (size, research and development). Table 1 summarizes the variables and the expected signs of their influence on rationing.

Table 1
Variables, their measures and expected signs

Dependant variables

Variable name Code Measure Expected signs

Rationing Ration Ration = 1: If the bank has refused to grant all or part of the credit N/A

Ration = 0: otherwise

Rationing rate Ration_rate The ungranted credit share proportion. The value is between 0 and 1. N/A

Independent variables

The nature of the information used in the decision-making process Inf Inf = 1 if the banker used accounting data and soft information –

Inf = 0 if the banker only used accounting and financial data

Characteristics of the bank-SME relationship

The number of banks Nmr_bk The number of banks requested for the same credit –

The existence of another credit (previous or present) with the same bank Cdt _his Cdt _his = 1: If the borrower has or has had another credit –

Cdt _his = 0: otherwise

The duration of the relationship Dur _rel This is the number of years since the company was a bank customer. –

Company characteristics

The size of the company Size Natural logarithm of total company balance sheet –/+

Technological level and innovation Innov Innov = 1: if the company claims to have an R &D department +

Innov = 0: otherwise

the amount of credit requested Log_at_cdt Natural logarithm of the amount of credit requested –/+

Credit characteristics

Credit maturity Cdt _mat Cdt _mat = 1 if the credit is repaid in the long term –

Cdt _mat = 0 if the credit is repaid in the short term

The interest rate Int _rate The interest rate that the company is spending in order to pay the bank +

The guarantee Grt Grt = 1: if the borrower has presented a guarantee –

Grt = 0: otherwise

Dependant variables
Rationing	Ration	Ration = 1: If the bank has refused to grant all or part of the credit	N/A
		Ration = 0: otherwise
Rationing rate	Ration_rate	The ungranted credit share proportion. The value is between 0 and 1.	N/A
Independent variables
The nature of the information used in the decision-making process	Inf	Inf = 1 if the banker used accounting data and soft information	–
		Inf = 0 if the banker only used accounting and financial data
Characteristics of the bank-SME relationship
The number of banks	Nmr_bk	The number of banks requested for the same credit	–
The existence of another credit (previous or present) with the same bank	Cdt _his	Cdt _his = 1: If the borrower has or has had another credit	–
		Cdt _his = 0: otherwise
The duration of the relationship	Dur _rel	This is the number of years since the company was a bank customer.	–
Company characteristics
The size of the company	Size	Natural logarithm of total company balance sheet	–/+
Technological level and innovation	Innov	Innov = 1: if the company claims to have an R &D department	+
		Innov = 0: otherwise
the amount of credit requested	Log_at_cdt	Natural logarithm of the amount of credit requested	–/+
Credit characteristics
Credit maturity	Cdt _mat	Cdt _mat = 1 if the credit is repaid in the long term	–
		Cdt _mat = 0 if the credit is repaid in the short term
The interest rate	Int _rate	The interest rate that the company is spending in order to pay the bank	+
The guarantee	Grt	Grt = 1: if the borrower has presented a guarantee	–
		Grt = 0: otherwise

Table 2 shows the descriptive statistics of variables. In the sample, 31% of the SMEs were rationed. The average rationing rate is 21% of the requested amount. About 76% of respondents said that soft information was used in decision-making process. 79% of SMEs applying for funding already have another credit. Companies compete with several banks before obtaining a loan (2.56 banks on average). The average duration of the bank-SME relationship exceeds 15 years. 27% of companies have an R&D department. The average interest rate was 4.52% and 54% of the sample had to deposit a guarantee.

Table 2

Descriptive statistics

	Mean	Std	Min	Median	Max
Ration	0.31	0.46	0.00	0.00	1.00
Ration_rate	0.21	0.36	0.00	0.00	1.00
Inf	0.76	0.43	0.00	1.00	1.00
Nmr_bK	2.56	1.27	1.00	2.00	10.00
Cdt _his	0.79	0.41	0.00	1.00	1.00
Dur _rel	15.53	12.28	0.00	15.00	85.00
Size	7.28	1.36	3.18	7.21	13.47
Innov	0.27	0.44	0.00	0.00	1.00
Log_at_cdt	4.77	1.25	0.69	4.79	8.52
Cdt _mat	0.80	0.40	0.00	1.00	1.00
Int _rate	4.52	1.44	0.00	4.50	10.00
Grt	0.54	0.50	0.00	1.00	1.00

Table 3 presents the correlation matrix for study variables. It shows that correlations between explanatory variables are small.

Table 3

Pearson correlation matrix

Variable		1	2	3	4	5	6	7	8	9	10	11	12
1	Ration		0.84	–0.15	–0.08	–0.07	–0.16	–0.14	–0.05	0.00	–0.20	0.19	0.20
2	Ration_ lv			–0.17	–0.08	–0.16	–0.11	–0.12	–0.02	–0.02	–0.26	0.15	0.13
3	Inf				0.08	0.16	0.08	0.03	0.10	0.07	0.02	–0.03	–0.13
4	Nmr_bK					0.26	–0.07	0.42	0.25	0.41	0.05	–0.08	0.07
5	Cdt _his						0.09	0.17	0.09	0.18	0.13	–0.11	–0.05
6	Dur _rel							0.13	–0.01	0.02	0.02	–0.08	–0.12
7	Size								0.21	0.63	0.10	–0.03	0.03
8	Innov									0.22	–0.06	0.03	0.01
9	Log_at_cdt										0.08	–0.03	0.07
10	Cdt _mat											–0.13	–0.03
11	Int _rate												0.14
12	Grt

4 Methodology and results

4.1 Methodology

Rationing may be total or partial. The object of hypothesis H1 is to show if the use of soft information reduces the probability for a SME to be rationed. Different models of logistic regression applied to the dummy variable (Ration) are used to respond to the research question. The variable (Ration) equals 1 when a firm is rationed and to 0 otherwise. The following control variables are also integrated: characteristics of the Bank-SME relationship, firm characteristics and characteristics of requested credits. Table 4 presents different logistic regression models used to confirm the first hypothesis.

Table 4
Logistic regression models

(L0) $Ln (\frac{P (Ration)}{1 - P (Ration)}) =$ α₀ + α₁Inf + ɛ

(L1) $Ln (\frac{P (Ration)}{1 - P (Ration)}) =$ α₀ + α₁Inf + α₂Nmr _ bk + α₃Cdt _ his + α₄Dur _ rel + ɛ

(L2) $Ln (\frac{P (Ration)}{1 - P (Ration)}) =$ α₀ + α₁Inf + α₅Size + α₆Innov + ɛ

(L3) $Ln (\frac{P (Ration)}{1 - P (Ration)}) =$ α₀ + α₁Inf + α₇Log _ at _ cdt + α₈Cdt _ mat + α₉Int _ rate + α₁₀Grt + ɛ

(Lfull) $Ln (\frac{P (Ration)}{1 - P (Ration)}) =$ α₀ + α₁Inf + α₂Nmr _ bk + α₃Cdt _ his + α₄ Dur _ re + α₅Size + α₆Innov + α₇Log _ at _ crt + α₈Cdt _ mat + α₉Int _ rate + α₁₀Grt + ɛ

The second hypothesis focuses on the impact of using soft information on the rationing rate. The endogenous variable is the actual rate of rationing (Ration_rate). This variable is left censored by 0 and right censored by 1. This double censoring explains the use of TOBIT censored regression models. The same control variables as above are considered. Table 5 presents different Tobit regression models used to confirm the second hypothesis.

Table 5

TOBIT regression models

Ration _ rate * = α₀ + α₁Inf + ɛ
Noting that	Ration _ rate * =	${\begin{matrix} Ration_R ate , & if 0 < Ration_r ate < 1 \\ 1 & if Ration_r ate \geq 1 \\ 0 & Ration_r ate * ⩽ 0 \end{matrix}$
We obtain the following models:
(T0)	Ration _ rate * =	α₀ + α₁Inf + ɛ
(T1)	Ration _ rate * =	α₀ + α₁Inf + α₂Nmr _ bk + α₃Cdt _ his + α₄Dur _ rel + ɛ
(T2)	Ration _ rate * =	α₀ + α₁Inf + α₅Size + α₆Innov + ɛ
(T3)	Ration _ rate * =	α₀ + α₁Inf + α₇Log _ at _ cdt + α₈Cdt _ mat + α₉Int _ rate + α₁₀Grt + ɛ
(TFull)	Ration _ rate * =	α₀ + α₁Inf + α₂Nmr _ bk + α₃Cdt _ his + α₄ Dur _ re + α₅Size + α₆Innov + α₇Log _ at _ crt + α₈Cdt _ mat + α₉Int _ rate + α₁₀Grt + ɛ

4.2 Results and discussions

The results of logistic regression models presented in Table 6 show a persistent significance of soft information role in reducing the SMEs rationing. This validates the first hypothesis H1. In fact, the first model is simple as it only integrates the variable representing the used information used. More complex models are then considered by adding separately characteristics of bank-SME relationship, of firms and of credits. One last model incorporates all previous variables. In every model, the results show that the use of soft information is negatively correlated with rationing, in line with the expectations. These results empirically prove the theoretical postulate affirming that soft information reduces rationing [7 , 53]. By integrating the characteristics of bank-SME relationship (model L1), the duration (considered to be a good indicator of the strength of the relationship [29]) reduces rationing by confirming theoretical postulates linking strength of relationship, its duration and credits availability. This argues that soft information, which is collected throughout the relationship by agents directly related to SMEs, contributes to reduce credit rationing [1 , 53].

Table 6
Logistic regression results

Dependent variable Ration Model L0 Model L1 Model L2 Model L3 Model LFull

Inf –0.743* –0.657 –0.729 –0.657 –0.657

(–2.625) (–2.259) (–2.538) (–2.180) (–2.100)

Nmr_bk –0.156 –0.121

(–1.354) (–0.920)

Cdt _his –0.084 0.088

(–0.261) (0.256)

Dur_rel –0.035* –0.028

(–2.702) (–2.031)

Size –0.218 –0.308

(–2.184) (–2.230)

Innov –0.033 –0.231

(–0.109) (–0.704)

log_at_cdt 0.042 0.308

(0.391) (2.050)

Cdt _mat –0.957* –0.968*

(–3.009) (–2.930)

Int_rate 0.211 0.210

(2.246) (2.169)

Grt 0.784* 0.766*

(2.831) (2.692)

Constant –0.230 0.667 1.341* –1.177 0.493

(–0.954) (1.556) (1.802) (–1.548) (0.514)

pseudo R² 0.018 0.046 0.033 0.092 0.132

chi2 6.782 16.978 12.084 33.995 48.735

df_m 1 4 3 5 10

P 0.009 0.002 0.007 0.000 0.000

Dependent variable Ration	Model L0	Model L1	Model L2	Model L3	Model LFull
Inf	–0.743***	–0.657**	–0.729**	–0.657**	–0.657**
	(–2.625)	(–2.259)	(–2.538)	(–2.180)	(–2.100)
Nmr_bk		–0.156			–0.121
		(–1.354)			(–0.920)
Cdt _his		–0.084			0.088
		(–0.261)			(0.256)
Dur_rel		–0.035***			–0.028**
		(–2.702)			(–2.031)
Size			–0.218**		–0.308**
			(–2.184)		(–2.230)
Innov			–0.033		–0.231
			(–0.109)		(–0.704)
log_at_cdt				0.042	0.308**
				(0.391)	(2.050)
Cdt _mat				–0.957***	–0.968***
				(–3.009)	(–2.930)
Int_rate				0.211**	0.210**
				(2.246)	(2.169)
Grt				0.784***	0.766***
				(2.831)	(2.692)
Constant	–0.230	0.667	1.341*	–1.177	0.493
	(–0.954)	(1.556)	(1.802)	(–1.548)	(0.514)
pseudo R²	0.018	0.046	0.033	0.092	0.132
chi2	6.782	16.978	12.084	33.995	48.735
df_m	1	4	3	5	10
P	0.009	0.002	0.007	0.000	0.000

*p < 0.10, **p < 0.05, ***p < 0.01; t-statistics are in parenthesis.

Concerning firm characteristics, the results of the third model (L2) demonstrate a negative relationship between firm size and rationing. Small sized companies need a closer and more personalized relationship to obtain requested credit. Contrary to expectations and the results of [32], this result indicates that bigger companies have greater chances to obtain credits. This result confirms those of [35] and further shows that the firm size is a reassuring factor for the credit granting.

Fourth regression (L3) argues that loans maturity, interest rate and guarantee have a significant impact on rationing. As expected, highest interest rates favour rationing, in line with the assertion of a part of literature assuming that higher-risk borrowers pay higher interest rates [4] and prefer to pay these high rates in order to access the credit market. Furthermore, in the context of “separation equilibrium”, a negotiated high interest rate should indicate that the borrower is risky and that he does not want to present a guarantee, he may lose in the event of bankruptcy. The second trend of literature argues that banks ration credits when interest rates are considered too high [48, 50]. The equilibrium between interest rate, collateral and credit volume is demonstrated [18]. Guarantee should reassure banks and encourage credit availability, but the results of this study highlight a positive relationship between rationing and guarantee. This result may be explained by the fact that companies offer guarantees to convince banks because they have a total confidence in their project [16, 26]. Risky companies are adopting this strategy in order to provide a positive signal to their banks.

By introducing all variables, model Lfull shows a significant and positive relationship between the amount of requested credit and rationing [32]. Indeed, an amount too high has more chances of being rationed. The requested amount is an important variable to determine banks potential losses in the case of failure. Banks prefer to ration large loans in order to better diversify funds allocation.

The relationship between rationing and requested credit size points to the question of granting total requested amount to a limited number of borrowers or of financing more borrowers with restricted amounts. These alternatives lead to the second section of this study. Logistic models have already showed the significant role of soft information on the rationing reduction. The second part based on Tobit models seeks to confirm the role of soft information on actual rationing rate.

The results presented in Table 7 mostly confirm those found previously. They show that soft information has contributed to reduce the actual rate of rationing. This result validates the second hypothesis.

Table 7

Tobit regression results

Dependent variable Ration_rate	Model T0	Model T1	Model T2	Model T3	Model Tfull
Inf	–0.818***	–0.706***	–0.825***	–0.694***	–0.678***
	(–3.006)	(–2.652)	(–3.047)	(–2.691)	(–2.654)
Nmr_bk		–0.105			–0.080
		(–1.101)			(–0.776)
Cdt _his		–0.374			–0.241
		(–1.324)			(–0.900)
Dur_rel		–0.022**			–0.015
		(–2.043)			(–1.423)
Size			–0.188**		–0.230**
			(–2.221)		(–2.204)
Innov			0.004		–0.122
			(0.014)		(–0.478)
Log_at_cdt				0.027	0.236**
				(0.311)	(2.015)
Cdt_mat				–0.947***	–0.892***
				(–3.475)	(–3.327)
Int_rate				0.143*	0.133*
				(1.953)	(1.856)
Grt				0.494**	0.449*
				(2.122)	(1.957)
Constant	–0.176	0.643*	1.198*	–0.546	0.788
	(–0.770)	(1.764)	(1.887)	(–0.883)	(1.089)
pseudo R²	0.021	0.039	0.032	0.074	0.099
Chi²	10.012	18.850	15.431	35.881	48.361
Model df	1	4	3	5	10
p-value	0.002	0.001	0.001	0.000	0.000
LogLikelihood	–238.433	–234.014	–235.724	–225.498	–219.259
LL intercept only	–243.439	–243.439	–243.439	–243.439	–243.439

*p < 0.10, **p < 0.05, ***p < 0.01; t-statistics are in parenthesis.

The influential variables in the model are the same as before, except for duration that is no longer significant in the full model. Soft information even displays a higher level of significance than in previous models. Among the control variables, three of the relationships in terms of rationing were not significant. More specifically, the number of banks requested for credit demand, the history of the relationship and the innovative character of the company do not have a significant impact on rationing. However, the negative relationship between the number of banks and rationing confirms the idea that a larger number of banks places the main bank in a position of competition and reduces rationing [20, 48]. The same negative relationship is found for the history of the bank-SME relationship and rationing, confirming the fact that the existence of another credit reimbursement or previous credit reduces rationing.

5 Conclusion

SMEs opacity poses a real problem for banks when selecting less risky borrowers. Several studies have highlighted the role that bank-SME relationship plays to reduce information asymmetry. This relationship provides private information and facilitates credit granting. The recommendations of Basel Committee, encouraging banks to use scores standardizes credit decision-making processes and disfavours the exploitation of relationships and soft information.

The purpose of this work is highlight the positive role of soft information on the reduction of credit rationing. The analysis is based on the assumption that the decision to grant credit is made following two steps: If credit is granted, it is necessary to determine the share that will be granted.

The results show the relevance of soft information in the reduction of rationing and the actual rationing rate. Bank-SME relationship, the characteristics of SMEs and the amount requested show their importance for reducing rationing. The duration, which is considered as a good indicator of the strength of the bank-SME relationship, contributes to rationing reduction. The same effect is observed for firm size (characteristic of SMEs) and the requested credit maturity (credit characteristics). While guarantee is supposed, reduce rationing, the study shows the contrary: a greater guarantee increases the risk of being rationed.

This article demonstrates the importance of soft information in the credit decision-making process and shows its role to improve granting credit for SMEs.

The managerial implications are double. On the one hand, SMEs must pay attention to this information, resulting from close bank-SMEs’ relationships, to increase the probability of obtaining their credits. On the other hand, banks must find a solution to integrate this soft information into their rating models to use a more complete information.

Bank-SMEs relationships still stay important even if bank industry is evolving. Moreover, the entire economy is changing to become more impersonal. This new setting reduces human interactions and this could lead to a lesser human implication in economic transactions. The contributions of this article can be extended and then explain the role of soft information into any decision-making process in business enterprises and other societal organizations. Furthermore, studying the hardening and the integration of soft information in impersonal decision-making processes could represent an interesting research question.

Footnotes

Appendix 1

Questionnaire summary

Attached, one will find a questionnaire summary - divided according to subject:

Section 1: credit application

1. Did you apply for a credit in 2008? (Y/N)

2. The requested credit is to be repaid in the short-term (repayment period less than or equal to 2 years) or in the long term?

3. Can you specify the amount of requested credit?

4. What percentage of this credit has been financed by the bank?

5. Did you obtain credit from the first application? (Y/N)

6. How many applications did you submit to obtain the credit?

7. Did you request credit from the same bank? (Y/N)

8. From how many banks did you request credit?

Section 2: credit terms

9. What is the interest rate you are paying?

10. Did obtaining the credit require the submission of a guarantee? (Y/N)

11. What share of the credit is covered by the guarantee?

12. Can you specify whether or not you are involved in financing the credit generating project? (Y/N)

13. Can you specify the part of self-financing for this project?

Section 3: the bank-SME relationship

14. Can you provide us with the number of banks that are working with you?

15. Is the credit in question requested from the main bank? (Y/N)

16. For how many years have you been a customer of this bank?

17. In order to process your credit application, do you believe that the banker relied solely on the accounting and financial documents or, in addition, exploited the information that they were capable of collecting during the period of your relationship (opinions, judgments, private information)?

-Solely accounting documents

-Accounting documents+judgments and opinions of the banker

18. Did you have a previously refunded credit or a credit in repayment with the bank that granted or refused you the credit in question? (Y/N)

19. If yes, was it obtained by that bank or another?

Section 4: The Company

In this final section, you are asked for a brief description concerning the characteristics of your company.

Can you estimate the size of your company in 2008?

-Staff: (trainees and temporary workers are not included)

- Balance sheet total:

20. Can you specify whether or not there is a Research & Development department within your company? (Y/N)

References

Albareto

, Benvenuti

, Mocetti

, Pagnini

and Rossi

, The Organization of lending and the Use of Credit Scoring Techniques in Italian Banks, Journal of Financial Transformation 32 (2011), 143–168.

Anari

, Kolari

, Pynnönen

and Suvanto

, Further evidence on the credit view: The case of Finland, Applied Economics 34(3) (2002), 267–278.

Barclay

M.J.

and Smith

C.W.

, The maturity structure of corporate debt, Journal of Finance 50(2) (1995), 609–631.

Becchetti

, Garcia

and Trovato

, Credit Rationing and Credit View: Empirical Evidence from Loan Data, Centre for Economic and International Studies 7(2), 2009.

Beck

and Demirguc-Kunt

, Small and medium-size enterprises: Access to finance as a growth constraint, Journal of Banking and Finance 30(11) (2006), 2931–2943.

Berger

A.N.

, Frame

W.S.

and Miller

N.H.

, credit scoring and the availability, price, and risk of small business credit, Journal of Money, Credit, and Banking 37(2) (2005), 191–222.

Berger

A.N.

and Udell

G.F.

, Some evidence on the empirical significance of credit rationing, Journal of Political Economy 100(5) (1992), 1047–1077.

Berger

A.N.

and Udell

G.F.

, Small firm credit availability and relationship lending: The importance of bank organisational structure, Economics Journal 112(477) (2002), 32–53.

Berger

A.N.

, Udell

G.F.

, Small business and debt finance, In Acs

Zoltan J.

and Audretsch

David B.

(eds.) Handbook of entrepreneurship research. (2005).

10.

Besanko

and Thakor

, Collateral and rationing: Sorting equilibria in monopolistic and competitive credit markets, International Economic Review 28(3) (1987a), 671–689.

11.

Bester

, Screening vs. rationing in credit markets with imperfect information, American Economic Review 75(4) (1985), 850–855.

12.

Bharath

S.T.

, Dahiya

, Saunders

and Srinivasan

, Lending relationships and loan contract terms, Review of Financial Studies 24(4) (2011), 1141–1203.

13.

Bolton

and Scharfstein

D.S.

, Optimal debt structure and the number of creditors, The Journal of Political Economy 104(1) (1996), 1–25.

14.

Bolton

, Freixas

, Gambacorta

and Mistrulli

P.E.

, Relationship and transaction lending in a crisis, Review of Financial Studies 29(10) (2016), 2643–2676.

15.

Cenni

, Monferrà

, Salotti

, Sangiorgi

and Torluccio

, Credit rationing and relationship lending. Does firm size matter? Journal of Banking & Finance 53(2015), 249–265.

16.

Chen

, Collateral, loan guarantees, and the lenders’ incentives to resolve financial distress, Quarterly Review of Economics and Finance 46(1) (2006), 1–15.

17.

Chen

, Guo

R.J.

and Huang

R.L.

, Two stages credit evaluation in bank loan appraisal, Economic Modelling 26(1) (2009), 63–70.

18.

Cowling

, The role of loan guarantee schemes in alleviating credit rationing in the UK, Journal of Financial Stability 6(1) (2010), 36–44.

19.

Craig

S.G.

and Hardee

, The impact of bank consolidation on small business credit availability, Journal of Banking & Finance 31(4) (2007), 1237–1263.

20.

Detragiache

, Garella

and Guiso

, Multiple versus single banking relationships: Theory & evidence, The Journal of Finance 55(3) (2000), 1133–1161.

21.

Flannery

, Asymmetric information and risky debt maturity choice, Journal of Finance 41(1) (1986), 9–37.

22.

Garvin

W.J.

, The small business capital gap: The special case of minority enterprise, Journal of Finance 26(2) (1971), 445–457.

23.

Gertler

and Gilchrist

, Monetary Policy, Business Cycles and the Behaviour of Small Manufacturing Firms, Quaterly Journal of Economics 109(2) (1994), 309–340.

24.

Gobbi

and Sette

, Do firms benefit from concentrating their borrowing? Evidence from the great recession, Review of Finance 18(2) (2014), 527–560.

25.

Godlewski

, Rôle de la nature de l’information dans l’intermédiation Bancaire, université Robert Schuman Strasbourg III, la RGE, (2004).

26.

Guiso

, High-tech firms and credit rationing, Journal of Economic Behavior and Organization 35(1) (1998), 39–59.

27.

Hall

, Investment in Research and Development at the Firm Level: Does the Source of Finance Matters?, National Bureau of Economic Research, Working Paper N°4096, 1992.

28.

Hernández-Cánovas

and Martínez-Solano

, Banking relationships: Effects on debt terms for small Spanish firms, Journal of Small Business Management 44(3) (2006), 315–333.

29.

Hoshi

, Kashyap

and Scharfstein

, Corporate structure, liquidity and investment: Evidence from japanese industrial groups, The Quarterly Journal of Economics 106(1) (1991), 33–60.

30.

Keeton

, Equilibrium Credit Rationing, New York, Garland Press, 1979.

31.

Kirschenmann

, Credit rationing in small firm-bank relationships, Journal of Financial Intermediation 26 (2016), 68–99.

32.

Kjenstad

E.C.

, Sub

and Zhang

, Credit rationing by loan size: A synthesized model, The Quarterly Review of Economics and Finance 55 (2015), 20–27.

33.

Kysucky

and Norden

, The benefit of relationship lending in a cross-country context: A meta-analysis, Management Science 62(1) (2015), 90–110.

34.

Laeven

, Does financial liberalisation reduce financing constraints? Financial Management 32(1) (2003), 5–34.

35.

Levenson

A.R.

and Willard

K.L.

, Do Firms Get the Financing They Want? Measuring Credit Rationing Experienced by Small Businesses in the U.S, Small Business Economics 14(2) (2000), 83–94.

36.

Lobez

, le rationnement du crédit: Une synthèse, Finance 9(2), 1988.

37.

Love

, Financial development and financing constraints: International evidence from the structural investment model, Review of Financial Studies 16(3) (2003), 765–791.

38.

Menkhoff

, Neuberger

and Suwanaporn

, Collateral-based lending in emerging markets: Evidence from Thailand, Journal of Banking and Finance 30(1) (2006), 1–21.

39.

Norden

and Weber

, Credit line usage, checking account activity, and default risk of bank borrowers, The Revue of Financial Studies 23(10) (2010), 3665–3699.

40.

Ortiz-Molina

and Penas

, Lending to small businesses: The role of loan maturity in addressing information problems, Small Business Economics 30(4), 2008.

41.

Petersen

M.A.

and Rajan

, The benefits of lending relationships: Evidence from small business data, Journal of Finance 49(1) (1994), 1367–1400.

42.

Petersen

M.A.

, Information: Hard and soft, Mimeo, Kellogg School of Management, Northwerstern University, 2004.

43.

Puri

, Rocholl

and Steffen

, Global retail lending in the aftermath of the US financial crisis: Distinguishing between supply and demand effects, Journal of Financial Economics 100(3) (2011), 556–578.

44.

Rondi

, Sembenelli

and Zanetti

, Is excess sensitivity of investment to financial factors constant across firms? Journal of Empirical Finance 1(3) (1994), 365–383.

45.

Schreft

S.L.

and Villamil

A.P.

, Credit rationing by loan size in commercial loan markets, Economic Review 78 (1992), 3–8.

46.

Slijkerman

J.F.

, Smant

D.J.C.

and De Vries

C.G.

, Credit Rationing Effects of Credit Value-at-Risk, Tinbergen Institute Discussion Paper 32(2) (2004).

47.

Smith

J.K.

, Smith

R.L.

, Entrepreneurial finance, John Wiley & Sons Inc, New York, 2000.

48.

Steijvers

and Voordeckers

, Collateral and credit rationing: A review of recent studies as a guide for future research, Journal of Economic Surveys 23(5) (2009), 924–946.

49.

Stein

J.C.

, Information production and capital allocation: Decentralized versus hierarchical firms, The Journal of Finance 57(5) (2002), 1891–1921.

50.

Stiglitz

J.E.

and Weiss

, Credit rationing in markets with imperfect information, The American Economic Review 71(3) (1981), 393–410.

51.

Stiglitz

J.E.

and Weiss

, Incentive effects of terminations: Applications to the credit and labour markets, American Economic Review 73(5) (1983), 912–927.

52.

Suzuki

, Credit channel of monetary policy in Japan: Resolving the supply versus demand puzzle, Applied Economics 36(21) (2004), 2385–2396.

53.

Uchida

, Udell

G.F.

and Yamori

, Loan officers and relationship lending to SMEs, Journal of Financial Intermediation 21(1) (2012), 97–122.

54.

Udell

G.F.

, What’s in a relationship? The case of commercial lending, Business Horizons 51(2) (2008), 93–103.

55.

Von

E.L.

, Thadden, Long-term contracts, short-term investment and monitoring, Review of Economic Studies 62(4) (1995), 557–575.

Dependant variables
Variable name	Code	Measure	Expected signs
Rationing	Ration	Ration = 1: If the bank has refused to grant all or part of the credit	N/A
		Ration = 0: otherwise
Rationing rate	Ration_rate	The ungranted credit share proportion. The value is between 0 and 1.	N/A
Independent variables
The nature of the information used in the decision-making process	Inf	Inf = 1 if the banker used accounting data and soft information	–
		Inf = 0 if the banker only used accounting and financial data
Characteristics of the bank-SME relationship
The number of banks	Nmr_bk	The number of banks requested for the same credit	–
The existence of another credit (previous or present) with the same bank	Cdt _his	Cdt _his = 1: If the borrower has or has had another credit	–
		Cdt _his = 0: otherwise
The duration of the relationship	Dur _rel	This is the number of years since the company was a bank customer.	–
Company characteristics
The size of the company	Size	Natural logarithm of total company balance sheet	–/+
Technological level and innovation	Innov	Innov = 1: if the company claims to have an R &D department	+
		Innov = 0: otherwise
the amount of credit requested	Log_at_cdt	Natural logarithm of the amount of credit requested	–/+
Credit characteristics
Credit maturity	Cdt _mat	Cdt _mat = 1 if the credit is repaid in the long term	–
		Cdt _mat = 0 if the credit is repaid in the short term
The interest rate	Int _rate	The interest rate that the company is spending in order to pay the bank	+
The guarantee	Grt	Grt = 1: if the borrower has presented a guarantee	–
		Grt = 0: otherwise

(L0)	$Ln (\frac{P (Ration)}{1 - P (Ration)}) =$	α₀ + α₁Inf + ɛ
(L1)	$Ln (\frac{P (Ration)}{1 - P (Ration)}) =$	α₀ + α₁Inf + α₂Nmr _ bk + α₃Cdt _ his + α₄Dur _ rel + ɛ
(L2)	$Ln (\frac{P (Ration)}{1 - P (Ration)}) =$	α₀ + α₁Inf + α₅Size + α₆Innov + ɛ
(L3)	$Ln (\frac{P (Ration)}{1 - P (Ration)}) =$	α₀ + α₁Inf + α₇Log _ at _ cdt + α₈Cdt _ mat + α₉Int _ rate + α₁₀Grt + ɛ
(Lfull)	$Ln (\frac{P (Ration)}{1 - P (Ration)}) =$	α₀ + α₁Inf + α₂Nmr _ bk + α₃Cdt _ his + α₄ Dur _ re + α₅Size + α₆Innov + α₇Log _ at _ crt + α₈Cdt _ mat + α₉Int _ rate + α₁₀Grt + ɛ