Janne Peltoniemi is Head of Degree Programme in Business Management and Principal Lecturer at Centria University of Applied Sciences, Kokkola, Finland, and is affiliated as a post-doc researcher at the University of Oulu, Department of Accounting and Finance, Oulu, Finland.
Personal Guarantees, Loan Pricing, and Lending Structure in Finnish Small Business Loans†
Article first published online: 18 MAR 2013
© 2013 International Council for Small Business
Journal of Small Business Management
Volume 51, Issue 2, pages 235–255, April 2013
How to Cite
Peltoniemi, J. and Vieru, M. (2013), Personal Guarantees, Loan Pricing, and Lending Structure in Finnish Small Business Loans. Journal of Small Business Management, 51: 235–255. doi: 10.1111/jsbm.12015
The authors wish to thank Professor Hannu Schadéwitz and the two anonymous reviewers for their insightful comments and extensive inputs to this article. This research is a part of FINNON research project financially supported by the Academy of Finland (Decision No. 116740). The bank is gratefully acknowledged for providing access to the confidential data used in this study.
- Issue published online: 18 MAR 2013
- Article first published online: 18 MAR 2013
This study analyzes the role of personal guarantees and collateral in the context of two different lending structures: one transaction and the other relationship based. The Finnish bank data, which were uniquely accessible for the study, enabled an exploration of credit files with specific details of the characteristics of the lending relationship during the period 1995–2001. According to the empirical results, the use of personal guarantees is an indication of transaction-based lending. Personal guarantees seem to increase the loan premium in transaction-based loans more than in relationship-based loans. Close ties between a bank and a firm seem to be a desirable basis for small and medium-sized enterprise bank lending.
Small business finance is an important part of the economy and its growth as a major part of small firms' external finances comes from bank loans (Berger and Udell 1998). Current global financial challenges come with high demands for small business bank loans and create financial pressures where every attempt to mitigate problems arising from asymmetric information should be important. Many studies show that relationship banking has been able to reduce some of these problems, particularly those concerning loan availability, cost of credit, or financial distress (D'Auria, Foglia, and Reedtz 1999; Hoshi, Kashyap, and Scharfstein 1990; Longhofer and Santos 2000; Petersen and Rajan 1995). Accordingly, increasing efforts have been made to understand the mechanisms of bank lending in small business finance and the characteristics of relationship banking especially. Although a great deal of the literature studies the benefits and costs of relationship banking, the role of personal guarantee in the context of relationship banking, however, is still ambiguous, in spite of the common practice of the use of personal guarantees in bank loan lending (Berger and Udell ; Bodenhorn 2003). For small firms in particular, collateralization and personal guarantees, as a part of collateral, are important factors that reveal key information on their borrower quality and creditworthiness for the bank. Furthermore, properly arranged collateralization tools can mean even less credit losses for banks. When customer firms are transparent in their information flow, the bank is increasingly able to assess the price and risk of the requested loan, a fact that should generate less credit losses than in bank–firm relationships with asymmetric information.
In a close bank–firm relationship, there are fewer problems with adverse selection, moral hazard, and costly verification, signs that are all harmful indications of the existence of asymmetric information (Binks and Ennew 1997; Boot and Thakor 1994; Foglia, Laviola, and Marullo Reedtz 1998; Thakor 1995). Consecutively, the more asymmetric information exists, the higher the costs are for the society in terms of credit losses and inefficient market conditions, which are then followed by mispriced loans and incorrect risk assessments. Therefore, by using collateralization correctly, including personal guarantees, banks and firms should be more efficient in providing financing. In order to achieve this financial environment, a better understanding of the role of collateralization, and especially the use and role of personal guarantees, is needed.
The purpose of this study is to emphasize small business bank lending characteristics by analyzing personal guarantees as part of the collateralization in two different lending types: in transaction-based and relationship-based structures. The characteristics of transactional and relationship lending is analyzed in the empirical setting. Previous research addresses the role of personal guarantees and collateral in bank–firm relationships from various aspects. The association between personal guarantees or collateral and different factors of lending (firm risk, information asymmetries, loan pricing, loan size, loan maturity, duration, and scope or breadth of relationship) are dispersed (Berger and Udell 2003; Degryse and van Cayseele 2000).1 In addition, it is not very clear why personal guarantees and collateral are sometimes used as a risk-controlling tool (Ortiz-Molina and Penas 2008; Scheelings 2006) and other times as a relationship-enhancing tool (Calcagnini, Farabullini, and Giombini 2007; Ono and Uesugi 2005). Furthermore, fewer studies provide hypotheses or expectations of how decisions on personal guarantees should be handled or modeled.
This paper contributes to the current literature in three main respects. First, this study is, to the best of the authors' knowledge, the first empirical examination to analyze personal guarantees and the cost of the credit in different lending structures. Second, the paper creates a factor that is new to the literature by examining how the lending structure is identified and controlled. This type of contextual lending structure component is very rare or even nonexistent in previous research. Third, we are able to use accurate measurements for the cost of credit and for the level of collateralization. The cost of credit is calculated as an effective rate, which includes the arrangement fees and/or provision for the bank in addition to the loan premium. This information gives a more accurate measurement of the cost of credit compared with the loan premium without the arrangement and/or the bank's provision. The previous literature often applies the loan premium without calculating the effective rate due to data restrictions. In addition, the level of collateralization is measured as a continuous measurement in this study, whereas in many other studies, it is available as a dichotomy variable (more restricted information content).
This study compares transaction-based lending (high information asymmetries) to relationship-lending (low information asymmetries). The use of personal guarantees in loan decisions is assumed to be determined as a tool to mitigate risk or to confirm the relationship. Whether personal guarantees are a mitigating risk or confirm the relationship may depend on the environment of information asymmetries. By comparing transaction-based loans and relationship-based loans, it is possible to analyze the role of personal guarantees and collateralization between these two structures. Eventually, we address two research questions: (1) Are personal guarantees associated with different lending structures?, and (2) do personal guarantees affect the cost of the credit?
The main results of this paper present new knowledge and understanding concerning how the personal guarantee affects the cost of credit when the lending structure is correctly identified and assessed. Our findings support that (1) transaction-based lending is associated with the personal guarantee, which refers to a higher asymmetric information level compared with relationship-based lending, and (2) when personal guarantees are used, the effect on the loan premium is higher in transaction-based loans than in relationship-based loans. Thus, transaction-based loans are more expensive than relationship loans when a personal guarantee is used. These results are considered robust after assessing the relationship, firm, loan, and collateral characteristics. Furthermore, the use of personal guarantees seems to be linked especially to unsecured loans as a supplementary part of the total pledged collateral. An analysis of relationship variables indicates that a longer duration of a relationship is associated with lower loan premiums. As an implication of the results, relationship-based loans refer to financial environments with lower asymmetric information, which may be a favorable condition for small business bank lending practices.
The rest of the paper is organized as follows. The next section reviews some of the prior research on issues of the role of personal guarantees and collateral in small business bank loans. The third section describes the data and methodology, whereas the fourth section reports the results, and the fifth concludes the paper.
Small business bank lending practices are very important in circumstances where informational asymmetries exist, that is, parties have unequal amounts of information of the same target (see Berger and Udell 1998, 2003). This is especially true in economic systems where information would hardly ever be symmetric. Informational asymmetries are costly, as their existence in the economy causes mispricing and inefficiencies in the markets (Thakor and Callaway 1983). If major parties in the economy, including firms, institutions, and individuals, would commit to taking actions according to more symmetric informational positions, it would maximize both shareholders' and stakeholders' welfare. Such positions would encompass long-term relationships, a high level of mutual trust, and ethically and morally ordered actions, for instance.
In small business finance, bank loans constitute the major part of the external financial needs of the firms, and informational asymmetries are at a very high level (Berger and Udell 2003). Many studies show that through a close and long-term bank–firm relationship, mutual benefits can be reached, and problems arising from asymmetric information, such as adverse selection, moral hazard, and costly ex-post verification, can be decreased (Berger and Hannan 1998; Berger, Demsetz, and Strahan 1999; Bhattacharya and Thakor 1993; Binks and Ennew 1997; Boot and Thakor 1994; Cyrnak and Hannan 1998; Fischer 1990; Greenbaum, Kanatas, and Venezia 1989; Peltoniemi 2004; Petersen and Rajan 1995; Sharpe 1990; Thakor 1995). Thus, loans will be more efficiently priced, and the availability of credit will increase (Thakor and Callaway 1983). This may lead to the conclusion that parties are more inclined toward morally and ethically correct behavior, which are assumed to be associated with a high level of small business loans. When firms have close ties with banks, they may be less probable to encounter loan defaults by banks' support during a period of financial distress of the firm (Edwards and Fischer 1994; Hoshi, Kashyap, and Scharfstein 1990; Longhofer and Santos 2000; Welch 1997). Also, close ties have positive impact and quality on bank–firm relationship (Binks and Ennew 1997; Ennew and Binks 1999). Firms may also more actively make debt payments and seek profits from the projects funded by the bank (not committing to moral hazard by executing projects outside the debt contract) because of the value of the relationship (Peltoniemi 2004). On the other hand, banks (in relationship banking) should be more willing to develop and maintain a reputation as a high-quality lender who obeys laws and regulations and in addition seeks the best outcomes for its own stakeholders.
The Effect of Relationship Banking on the Cost of Credit
Examining relationship banking and its effect on the cost of credit, many studies present a negative relationship between the strength of relationship and the cost of credit (Angelini, Di Salvo, and Ferri 1998; Berger and Udell 1995; Blackwell and Winters 1997; Boot and Thakor 1994; D'Auria, Foglia, and Reedtz 1999; Mörttinen 2000; Niskanen and Niskanen 1999; Peltoniemi 2007a). In contrast, however, numerous studies claim that the strength of the relationship and the cost of credit may be positively related (Caves and Uekusa 1976; Greenbaum, Kanatas, and Venezia 1989; Hanley, Ennew, and Binks 2006; Nakatani 1984; Weinstein and Yafeh 1998).
Petersen and Rajan (1994) studied relationship banking in a concentrated lending structure and found that lower cost of credit is associated with the main bank of the firm. These firms were long-established large firms and corporations. Berger and Udell (1995) also find that lower cost and lower collateral requirements are related to longer bank–firm relationships, especially with larger firms. As for small firms, many empirical studies show that duration and/or scope (or breadth) of relationship are important determinants for the strength of relationship, for instance Harhoff and Körting (1998), Mörttinen (2000) and Scott and Dunkelberg (1999).
These studies reflect that when the relationship between the bank and firm is tight and close, there are mutual economic benefits that the parties share. However, if either bank or firm starts behaving opportunistically, some negative consequences may follow, such as hold-up problems and informational capturing by the bank or moral hazard actions by the firm.
The analysis conducted in this study contributes to the previous studies in two ways: by using the effective loan rate and by examining the relation between the strength of the relationship and the cost of credit in the lending structure context.
The Role of Personal Guarantees and Collateral
The importance of personal guarantees and collateral in bank lending is evident (Berger and Udell 2003). However, the role of collateral in loan contracts is rather unclear despite the extensive theoretical and empirical research on the topic. For instance, according to conventional banking wisdom and certain academic studies, borrower risk and collateral requirements seem to be positively related (Berger and Udell ; Inderst and Müller 2006; Leeth and Scott 1989; Longhofer and Santos 2000; Orgler 1970; Welch 1997). Some academic research, however, indicate a negative relationship (Besanko and Thakor 1987; Bester 1985; Capra, Fernandes, and Ramirez 2001; Chan and Kanatas 1985; Chan and Thakor 1987; Machauer and Weber 2000; Peltoniemi 2007b). These implications and results can be found in both theoretical and empirical literature and highlight the need of additional understanding of the mechanisms that can better explain the large variation of the empirical and theoretical results on the association of the use of collateral and personal guarantees in bank loan contracts.
The role of personal guarantees and collateral may have different outcomes depending on the level of information asymmetries. Under asymmetric information, personal guarantees and collateral can be mitigating tools to control or decrease problems of adverse selection (Bester 1985, 1987) or moral hazard (Bester 1994; Boot, Thakor, and Udell 1991). On the other hand, personal guarantees and collateral may be associated with information content established by financial intermediaries (Boot 2000; Boot and Thakor 1994; Manove, Padilla, and Pagano 2000; Rajan and Winton 1995; Sharpe 1990). If the information environment were symmetric and perfect, there would be no cost on collateral, but risk aversion of the firm owners would still exist (Bester 1987). In that case, the firm owner would lose collateral in a project default and increase profitability in a successful project through lower interest rates.
The results of Ortiz-Molina and Penas (2008) as well as Hernández-Cánovas and Martínez-Solano (2006) support the idea that collateralization and/or personal guarantees are related to relationship banking rather than transaction-based lending. However, the results of Pozzolo (2002), Ono and Uesugi (2005), Calcagnini, Farabullini, and Giombini (2007), Pozzolo (2004), and Brick and Palia (2007) have indications for transaction-based lending (Table 1).
|Ortiz-Molina and Penas (2008)||Lending to small businesses: the role of loan maturity in addressing information problems||The association between collateral/personal collateral, borrower risk/informational asymmetries, and loan maturities.||Loan maturity shorter for high-risk, informationally opaque firms. Maturity and collateral are substitutes in decreasing agency problems. Personal collateral associated with longer maturities than business collateral.||Short maturities indicates transaction-based lending, collateralization enables longer maturities, especially the use of personal collateral in reducing the borrower risk. Personal guarantees compensate the risk involved in transaction-based lending.|
|Scheelings (2006)||Firm risk and collateralized asset choice in small business bank lending: theory and evidence||The association between collateral/personal collateral, firm risk, firm size, and loan size.||The use of personal asset as collateral may indicate either decreasing or increasing firm risk. No relation to firm size or loan size.||When transaction-based lending indicates higher risk than relationship-based lending, the use of personal assets may either increase (moral hazard) or decrease (compensation) the borrower risk.|
|Pozzolo (2002)||Secured lending and borrower's riskiness||The association between guarantee loan premium, firm risk, and firm size.||Theory: the use of guarantee is related to higher loan premiums. Empirical: guarantees required riskier and smaller firms (high asymmetric information)||The positively associated personal guarantees, firm riskiness, and loan premiums indicate that guarantees are tools to lower informational asymmetries, which are evident in transaction-based lending.|
|Hernández-Cánovas and Martínez-Solano (2006)||Bank relationships: effects on the debt terms of the small Spanish firms||The association between bank relationship, loan premium, and collateral/personal guarantees.||Less number of relationships with banks reduces cost of credit. The use of personal guarantees positively related with duration of relationship. Relationship loans include lower premiums but higher real collateral.||Relationship lending indicates more use of personal guarantees than transaction-based lending.|
|Ono and Uesugi (2005)||The role of collateral and personal guarantees in relationship lending: evidence from Japan's small business loan market||The association between collateral/personal guarantees, borrower risk, and duration of relationship.||Contrary with conventional theory, collateralized or (personal guarantees used) loans are more frequently monitored.||Collateral and personal guarantees are complementary to relationship lending as a tool to diminish the adverse effect of asymmetric information.|
|Calcagnini, Farabullini, and Giombini (2007)||Loans, interest rates, and guarantees: is there a link?||The association between collateral/personal guarantees and loan premiums in firms, producer households, and consumer households.||Collateral decreases loan premium as a solution to adverse selection, and personal guarantees increase loan premiums as a solution to moral hazard problem.||Bank use personal guarantees to decrease borrower risk and to prevent opportunistic behavior, which are evident in high-asymmetric information environment. Personal guarantees indicate transaction-based lending.|
|Pozzolo (2004)||The role of personal guarantees in bank lending||The association between collateral/personal guarantees, borrower risk, loan premiums, and secured/unsecured loans.||Collateral is used as an internal device for protection of loan default. Personal guarantees are used as an external device for protection against moral hazard.||Collateral and personal guarantees are used in order to decrease ex-ante credit risk. Ideal tool to control credit risk in transaction-based lending.|
|Brick and Palia (2007)||Evidence of jointness in the terms of relationship lending.||The association between collateral/personal guarantees, loan premium, and bank relationship||Contrary to the theory that collateral is endogenously chosen, results show collateral and loan premium are positively related as a protection against moral hazard.||The effect of personal guarantees on loan premium (+) is stronger than collateral. The use of personal guarantees related to higher risk, thus indicating transaction-based lending, where risk levels are higher due to higher asymmetric information.|
|Berger, Cerqueiro, and Penas (2009 AFA)||Does debtor protection really protect debtors? Evidence from the small business credit market||The association between debtor protection, collateral, maturity, loan premium, and loan size.||Under debtor-friendly bankruptcy laws sole proprietorships and partnerships pledge collateral, loans are shorter in maturity, loan premiums higher and loans smaller. Limited liabilities pay higher loan premiums.||Unlimited liability small businesses face collateral requirements more often than liabilities (under debtor-friendly bankruptcy laws). The incidence of personal collateral is lower than business collateral.|
To conclude, the previous results point to three major implications: (1) the use of personal guarantee and collateral is significantly dependent on the informational distribution level between the firm and the bank. When asymmetric information is high (low), it is probable that the ex-ante credit risk and firm risk is high (low) and that the incidence of moral hazard or adverse selection is high (low). (2) Previous studies seem to be diverse in explaining the role of personal guarantee and collateral in different information environments and lending structures, that is, transaction-based and relationship-based lending; and (3) moreover, the effect on loan premiums in these lending structures is ambivalent and need further research.
The previous implications lead us to establish two hypotheses to be empirically analyzed in this study, focusing especially on the role of personal guarantees as follows:
- H1: The use of personal guarantees is associated with the lending structure.
- H2: The use of personal guarantees is associated with the cost of credit.
H1 is established on the basis of previous literature, where both lending types are supported. In H2, higher loan premiums refer to a “mitigation tool,” and lower loan premiums refer to a “confirmation tool.” Mitigation refers to risk management, where the use of personal guarantees is aimed at reducing the risk involved in the loan contract. Confirmation means emphasizing or strengthening the existing relationship by the use of personal guarantees in the loan contract.
Data and Methodology
The original data covered all corporate loans originating from a local corporate division of a major Finnish bank from 1995 to 2001.2 The data represent the banking environment in Finland reasonably well, as the sample was gathered from a full long-term credit decision period of a large bank. However, due to the highly concentrated banking industry, where three major banks share over 80 percent of the market, when making generalizations of the results, the bank-based market structure has to be taken into account (for a review of Finnish financial markets, see Bank of Finland 1996; Peltoniemi 2004). Similar strong bank-based financial systems are found in Germany, Japan, and Italy, for instance. It is even argued that countries with strong bank-based financial systems have shown higher economic growth than other countries and that close bank–firm relationships may become economically significant in efficiency through the availability and relatively low cost of credit. Because the structure of Finnish financial markets is very concentrated, bank lending in small business finance and understanding the mechanics of bank lending become important. Finnish studies find that a close and/or long-term relationship may be desirable for small firms due to the potential benefits (Mörttinen 2000; Niskanen and Niskanen 1999; Peltoniemi 2004, 2007a). Compared with banking data in previous studies, the data in this study do not contain information on the total amount of relationships a firm holds with banks in general.3 However, the data include some pieces of information that are not generally included in previous studies, such as continuous measurement of the level of collateralization, an effective loan rate (including arrangement fees), and internal risk rating of firms (Table 2).
|Effective rate||Negotiated marginal effective interest rate that includes the total loan arrangement fees and provisions for the lender.|
|Duration||Length of the relationship between bank and firm (measured in years).|
|Financial services||Number of the subject bank's financial services used by the firm.|
|Loans||Number of currently performing loans from the subject bank.|
|Multiple||Dummy variable indicating (= 1) whether the firm owner has multiple client–bank relationships with the subject bank or not (= 0). Client–bank relationships can include either firms or individuals.|
|Commitment||Ratio of firm's total liabilities to the subject bank per total assets of the firm.a (= firm's total liabilities/firm's total assets)|
|Structure||A joint variable, which is established from four variables: duration, financial services, loan size, and loan maturity by forming 10 groups from each variables. Groups are ranked and scaled between values 0 and 1. Value 0 indicates the orientation for transaction-based lending and value 1 indicates the orientation for relationship-based lending.|
|Pergua||Dummy variable indicating whether personal guarantee has been used (= 1) or not (= 0).|
Coverage of pledged total collateral over the total liabilities of the firm to the subject bank after the current loan (measured in percent).
[= (total collateral/liabilities)*100].
|Nonbank guarantee||Dummy variable indicating (= 1) whether the loan is guaranteed by a Finnish nonbank financial institution or not (= 0).|
|Leverage||Debt per assets of the firm (= debt/assets).|
|Firm age||Age of the firm (measured in years).|
|Firm size||Total assets of the firm.|
|Legal form||A dummy variable indicating (= 1) whether the legal form of the firm is partnership/proprietorship. Otherwise limited liability (= 0).|
|R3-5||Bank's internal risk-rating classification. (lower risk = R1-2, higher risk = R3-5).b|
|Industry||Ten industry dummies.|
|Maturity||Maturity of the loan (measured in years).|
|Loan size||Amount of the loan (in thousand FIM [Finnish markka, currency in Finland until 2002]).|
|Year||A dummy variable for each year in the data.|
The data from 1995 to 1999 were collected in 2000, and data from the 2-year period of 2000–2001 were collected in 2002. Each credit file includes detailed information on the credit contracts between the borrower firm and the bank, covering relationship-, collateral-, firm-, and loan-specific characteristics. The whole sample consisted of 1,436 loan decisions for 936 small firms.4 We included only limited liability companies or partnerships in the analysis by excluding 112 nonprofit organizations with 125 loans.
Next, we excluded 108 loans that provided no information on the interest rate.5 Information for effective loan rate calculations was missing for 14 loans. Thus, our final sample consisted of 1,189 loans for 768 firms.6 Over 100 variables were gathered from 1,436 credit files from 1995 to 2001. We note that during the data period 1995–1999, the economic growth was especially strong in the technology industry. However, our data do not contain any more than a small minority of technology-oriented firms. For instance, the electronic industry represents less than 1 percent of the estimated sample.
In the estimation model, the endogenous variable is the effective interest rate of the loan (Effective rate), that is, the interest rate spread that includes the total amount of loan arrangement fees for the bank, or personal guarantee (pergua). The effective loan rate is the actual and true value of the loan that the lender and the borrower have negotiated in the contracting process. The following explanatory variables, which are defined later, were used in the analysis.
The use of a personal guarantee is a dichotomy variable indicating whether the loan is personally guaranteed or not. Personal guarantee is a given promise (in contrast to real asset) to personally pay the bank a certain amount in case of a loan default. The value of personal guarantee is determined as to a maximum sum of cash one person is generally able to pay, regardless of the income (limited guarantee).7 Empirical research results show that collateral (decreasing the adverse selection problem) and personal guarantees (decreasing the moral hazard problem) may have their unique roles in small business loans (Brick and Palia 2007; Calcagnini, Farabullini, and Giombini 2007; Pozzolo 2004). Collateral (real assets) effects are proxied by the value of a firm's total collateral over total liabilities to the bank (collateralization). Collateralization is a continuous variable showing how much of the firm's liabilities to the bank after the current loan is covered by the pledged level of the total collateral.8 We also check whether the loan is guaranteed by the major nonbank financial institution (nonbank guarantee) or not. The nonbank in the data is owned by the Finnish government. This nonbank guarantee often works closely with banks by giving external guarantees to loans. Typically, banks require this type of guarantee for loans that are requested by start-ups or high-risk firms.
The relationship variables are measured by variables that are related to the strength of bank–firm relationship, indicating the dimensions of duration of relationship and the breadth or scope of relationship. Duration of relationship is one of the basic and most used single measures of the bank relationship strength (Berger and Udell 1995; Blackwell and Winters 1997; Bodenhorn 2003; Degryse and van Cayseele 2000; Petersen and Rajan 1994). However, it does not tell much about the intensity of the relationship between the bank and the firm. Therefore, we add four scope-based variables to measure the breadth and intensity of the relationship. Some previous studies have examined the scope measures as well (Degryse and van Cayseele 2000; Elsas 2005; Mörttinen 2000). These are the number of financial services and performing loans a firm has in this bank, ownership broadness of different loans, and the level of commitment.
This study uses the length of relationship in years to analyze the relationship duration (duration). The other relationship variables deal with relationship scope. These factors represent different aspects of the breadth of the relationship. We use the following variables: number of financial services used (financial services), number of performing loans (loans), group of firms under the same ownership (multiple), and level of commitment (commitment). All of these variables indicate the nature of the relationship between the borrowing firm and the lender bank.
Financial services include the total number of the subject bank's financial services that the firm is currently using. Multiple indicates whether the firm belongs to a so-called firm group. It simply indicates if there is more than one ongoing relationship (either of the firm or family member) under the same ownership with this bank. For instance, such is the case when one person owns two firms, both of which have relationships with the bank. Another example would be a married couple who governs two or more firms, either mutually or individually.
Commitment is the ratio denoting a firm's total liabilities (to this bank) after the current loan decision per total assets of the firm. The more debt the bank has issued to the firm, the bigger share of the firm's total assets the bank holds.9 A higher value in commitment indicates closer ties and a stronger relationship between the bank and the firm.
There are five firm-specific control variables that are associated with the following sectors of firm performance: the leverage of the firm (leverage), the age of the firm (firm age), the size of the firm (firm size), the legal form of the firm (legal form), and the internal risk rating of the firm (R3-5). Originally, five risk classes are internally determined by the bank, and we use a dummy variable to analyze the effect of two risk classes (high/low) on the effective loan rate.10 R1 denotes the lowest risk level for a firm, and R5 denotes the highest risk level. Two risk groups are established, R1-2 representing low risk firms, and R3-5 representing high-risk firms.
Maturity of the loan (maturity) and year dummies (year-dummies) are used as loan-specific control variables. For each year, there is a dummy variable indicating the year when the loan was granted. The loan size variable indicates the size of the loan.
The summary statistics in Table 3 show the logarithm of the level of effective rate for the full data and the relationship-based sample and the transaction-based sample for the mixed-group. It is notable that the mean and the standard deviation of log (effective rate) are at the lowest in the relationship-based sample compared with other groups, especially to the transaction-based sample. The average relationship duration is approximately nearly 9 years, and the average number of the bank's financial services used by firms is about 10 years. The mean of loans is 1.28, which means that a normal quantity of loans is between one and two. Over 50 percent of firms or firm owners have multiple client relationships with the subject bank, that is, the value of multiple is 0.642. Personal guarantee (pergua) is used in 29.7 percent of the loans, and the nonbank's guarantee exists in 12.5 percent of the loans.11 The level of collateralization is 100.7 percent, which means that, on average, loans are fully covered. The firms are relatively highly leveraged, as the debt-to-assets ratio is 92.9 percent. The average age of the firms is 11 years. Twenty-four percent of the firms are limited liability companies and the rest are either partnerships or proprietorships. Summary statistics for the subsample of 279 firms is presented in Table 3.
|Log (effective rate)||Log Ratio|
|Log (commitment)||Log ratio||279||−2.240||4.323|
|Log (firm size)||Log ratio||279||7.031||1.498|
|Log (loan size)||Log ratio||279||5.495||1.184|
|Hotel & restaurants||0,1||279||0.047||0.211|
The explanatory variables are divided into five categories. The endogenous variable is the effective interest rate of the loan, that is, the interest rate spread that includes the total amount of loan arrangement fees for the bank. The effective loan rate is the actual and true price of the loan that the lender and the borrower have negotiated in the contract.
Our primary interest lies with the personal guarantee and collateral variables. Relationship variables are formed to capture the effects related to the strength of the bank–loan relationship, and the firm-characteristics are control variables. As for relationship characteristics, the duration of this relationship has been one of the basic measurements of the relationship strength over the years in relationship banking literature, but unfortunately it does not reveal enough about the intensity of the relationship. This is why we include various measurements for the relationship scope. Other collateral-, firm-, and loan-specific factors are control variables in the analysis. The subsequent analysis of the data has been conducted by emphasizing relationship characteristics, that is, duration and scope. Table 2 contains complete variable definitions and information on the construction of the variables.
In the following models, the characteristics of personal guarantee, collateral, and relationship banking are estimated with the cost of the credit or the collateralization as dependent variables. The general literature of relationship banking estimates the loan spread by several variable groups including relationship, collateral, firm, and loan characteristics (Berger and Udell 1995; Degryse and van Cayseele 2000; Petersen and Rajan 1994).
We are aware of the possible endogeneity of nonrate terms in the case of simultaneous determination in the type of models that we use in our econometric specifications. In many cases, simultaneous equations are run in order to control for instance potential endogenous problems of the exogenous variables, loan spread, and collateral. Generally, the estimated Ordinary Least Squares (OLS) parameter values may not be consistent if endogeneity is suspected. In our model, this problem may exist specifically in personal guarantee, collateralization, and loan size. However, we assume that according to the discussions and the credit process in this specific bank, the loan rate is determined after other variable values are set. Thus, the bank negotiates with its clients about the size of the loan and the level and quality of collateral before making a decision on the loan rate.
As the previous sections indicate, personal guarantee or real assets as collateral are used to separate transaction-based loans from relationship-based loans. When a personal guarantee or collateral is required, it is an assumption that bank may want to have protection from a possible moral hazard situation in the near future (Brick and Palia 2007; Calcagnini, Farabullini, and Giombini 2007; Ortiz-Molina and Penas 2008; Pozzolo 2004; Scheelings 2006) or that the bank–firm relationship has a strong nature with a high level of mutual trust, which is confirmed by a personal guarantee (Hernández-Cánovas and Martínez-Solano 2006; Ono and Uesugi 2005).
According to the literature, duration and scope of the relationship can be used as measurements for the strength of relationship (Angelini, Di Salvo and Ferri 1998; Berger and Udell 1995; Blackwell and Winters 1997; Boot and Thakor 1994; D'Auria, Foglia, and Reedtz 1999; Mörttinen 2000). When the bank–firm relationship is strong, there is less asymmetric information between the bank and the firm. This should be indicated through lower loan spread decisions (or alternatively through lower collateralization requirement). Consecutively, these relationships should produce more efficient loan pricing and less loan defaults, which both are beneficial not only to the customer firm but also to the bank (better reputation as a lender) and finally to the society (less bankruptcies, less loan defaults), that is, a higher level of corporate responsibility to the civil society and to the various stakeholders of the firm and bank (cities, communities, and government).
The general estimation model in this study is in line with the current literature on relationship banking:
Model 1 is applied to test H1. By analyzing the variables that emphasize the characteristics of lending structure (transaction- versus relationship-based lending), Model 1 tests whether the use of personal guarantee (pergua) is oriented with either transaction-based or relationship-based lending. The lending structure characteristics, structure, are established from four variables: duration, financial services, loan size, and loan maturity. This is done by forming 10 groups from each variable, ranking the loans by the groups. Finally, one continuous joint variable (structure) is created, where each of these variables are equally weighted. The total group values of these four variables are scaled between 0 and 1 (0 indicates the orientation for transaction-based lending, whereas 1 indicates the orientation for relationship-based lending). These four variables are assumed to increase with the orientation for relationship-based lending (Berger and Udell 1995, 1998, 2003; Blackwell and Winters 1997; Bodenhorn 2003; Petersen and Rajan 1994).
Model 2 and Model 3 are applied for the second hypotheses. In Model 2, the cost of the specific lending structure can be observed through the relationship between pergua and the dependent variable, effective rate. Model 3 is an OLS regression, which tests the second hypothesis in separate samples of transaction-based and relationship-based data. These samples are formed by identifying transaction-based customer firms from relationship-based customer firms. The firm is identified as a transaction-based customer if at least 75 percent of its loans are granted as transaction based (structure ≤0.25), and the firm is identified as a relationship customer if at least 75 percent of its loans are granted as relationship based (structure ≥0.75). In addition, a mixed group of transaction- and relationship-based samples is established (0.25 < mixed group < 0.75). The mixed group contains firms with no clear orientation for either transaction- or relationship-based lending.
The specific models estimated in this study are as follows:
|Log (financial services)||0.078||0.677|
|Log (firm age)||−0.291||0.157||−0.127||0.416|
|Log (firm size)||−0.140||0.223||−0.113||0.244|
|Log (loan size)||−0.240||0.10|
Model 3, OLS regression/H2
where α is the constant, γ represents the collateral characteristics, β represents the relationship characteristics, δ represents the firm characteristics, θ represents the loan characteristics, and ε is the error term.
The main estimation results for the personal guarantee and collateral characteristics are presented in Tables 4-6. Tables 4-6 report the empirical results for H1 and H2, respectively. Furthermore, the results in Table 6 refer to the testing of H2 with a pooled cross-sectional OLS regression. Every loan is treated independently. The regressions are subsample examinations with up to 279 observations available for the reported variables. The statistical significance is defined at 10 percent level with bolded parameter values. We control the heteroskedasticity of models by applying the White test (1980).
|Log (financial services)||0.039||0.341|
|Log (firm age)||0.050||0.256||0.142||0.0004|
|Log (firm size)||−0.011||0.667||−0.038||0.092|
|Log (loan size)||−0.208||<0.0001|
|Relationship based||Transaction based||Mixed|
|Log (firm age)||0.142||0.018||0.069||0.398||0.200||0.305|
|Log (firm size)||−0.095||0.002||−0.071||0.194||−0.159||0.089|
Our main emphasis in characterizing collateralization and the strength of the bank–firm relationship relies on two collateral variables: personal guarantee (pergua) and collateral (collateralization), and six relationship variables: relationship duration (duration), the number of the bank's financial services used by a firm (financial services), the number of loans (loans), existence of multiple client relationships (multiple), the level of commitment (commitment), and alternatively a joint variable (structure) that indicates the collective impact of the lending structure (whether it is more transaction-based oriented or relationship-based oriented).
Empirical Results for H1
The regression analysis begins by exploring the effect of relationship, collateral, firm, and loan characteristics on the use of personal guarantee in Table 4. Both regressions are tested with Model 1.
In Model 1 (logistic regression) in Table 4, the dependent variable is personal guarantee (pergua), having values 0 (guarantee) and 1 (no guarantee). Regression (1) shows that multiple is the only relationship characteristic that affects pergua. The parameter value, −0.625, indicates that if the customer firm owner is identified as having multiple business relationships with this bank, this incidence seems to lower the probability of using personal guarantee in the loan contract. If the level of collateral is increasing, the probability for personal guarantee will be lower. However, the nonbank guarantee increases this incident. This may happen because of the lending policy of the nonbank (risk-sharing nature). As for firm characteristics, unlimited liability firms are more likely to be linked with the use of personal guarantees than limited liability firms. Loan characteristics show that both maturity and loan size are negatively related to the use of personal guarantees, which may be associated with the lending structure separation.
When transaction-based loans are generally shorter and smaller (and vice versa for relationship-based loans), loan characteristics (maturity and loan size) indicate that relationship-based loans are less likely to be associated with personal guarantees than in transaction-based loans. This result supports the theoretical approach in Pozzolo's (2002) study, where the use of personal guarantee is associated with higher loan premiums and with riskier and smaller firms, that is, related to transaction-based lending. In regression (2), the structure variable has a significantly negative parameter value (−2.755), which confirms the results in regression (1); when the emphasis for relationship-based lending is increasing, the use of personal guarantee becomes less probable. One percentage increase in structure lowers the loan spread by 2.76 percent. These results are supported in similar findings in Ono and Uesugi (2005), Calcagnini, Farabullini, and Giombini (2007), and Brick and Palia (2007). In sum, the tests of aggregate data support the transaction-based orientation in association with the use of personal guarantees in the first hypothesis.
Empirical Results for Hypothesis 2
The estimation results for the second hypothesis are reported in Tables 5 and 6. Table 5 reports the results in OLS regressions, where the dependent variable is the cost of credit (effective rate). From these regressions, it is possible to test how the use of personal guarantees or the lending structure (transaction based or relationship based) in aggregate data affects loan pricing. It is found that the use of personal guarantees (pergua) has a positive parameter sign, indicating that personally guaranteed loans are more expensive than loans without a personal guarantee (statistically significant parameter value at 1 percent level). Additionally, transaction-based loans are more expensive than relationship-based loans (parameter value for structure is −1.655 and statistically significant at 1 percent level). From single variables, the duration of relationship (duration) lowers the cost of credit (−0.126), the number of loans (loans) increases the cost of credit (0.037), collateral (collateralization) slightly lowers the cost of credit (−0.001), higher firm risk (R3-5) increases the cost of credit (0.085), maturity of the loan (maturity), and the size of the loan (loan size) lowers the cost of credit (−0.038 and −0.208), in regression (1).
Three subsamples of relationship based, transaction based, and mixed-group are analyzed separately with Model 3, in Table 6. The mixed-group is made up of customer firms that have identifications for both relationship-based and transaction-based loans within 25–75 percent range (e.g., two relationship-based loans and one transaction-based loan). All three samples are analyzed with OLS regressions and with the number of observations, that is, 58, 88, and 133, a total of 279. The customer firm is identified as relationship based for structure–variable values that exceeds 0.5, and for transaction-based customers less than 0.5. The mixed-group includes all loans with structure–values within 0.25 and 0.75 range (as stated previously).
In the first regression (1) (Table 6), all loans are for customer firms identified as relationship-based lenders with 133 observations and an adjusted explanatory power of 0.4. The use of personal guarantee (pergua) increases the loan premium by 0.233 percent for these loans. Collateral does not seem to have any effect on the cost of the loan. For relationship characteristics, only the number of loans (loans) is statistically significant with a positive parameter value of 0.046. Firm age (firm age) and firm risk (R3-5) are positively related with the cost of credit, and the total assets of the firm (firm size) are negatively related to the cost of credit. It seems that matured firms pay slightly higher premiums, and larger firms receive a small discount in loan premiums. High-risk firms pay higher premiums than low-risk firms.
The second regression (2) (Table 6) includes transaction-based loans only with 88 observations and an adjusted explanatory power of 0.105. One of the evident results for the differences between relationship-based and transaction-based customer loans is that transaction-based customers pay higher loan rates when a personal guarantee is used compared with relationship-based customers (0.423 versus 0.233). These parameter values are statistically different at 5 percent significance level. In addition to previous results in H1, this confirms that not only are transaction-based loans more expensive than relationship-based loans but also the effect of the use of personal guarantees on a loan premium is higher for transaction-based loans than relationship-based loans. This result is supported by Bodenhorn (2003). In the transaction-based sample, loans and multiple (relationship characteristics) are statistically significant with the parameter values of 0.145 and 0.275, respectively. The parameter sign is consistently positive in both relationship-based and transaction-based samples. The parameter value for collateralization is nearly equivalent to zero with no statistical significance.
The third regression (3) (Table 6) results refer to a mixed-group with consistent parameter signs. Interestingly, pergua (0.658) is statistically significant and clearly exceeds the values of relationship-based (0.233) and transaction-based (0.423) samples. This indicates that if the firm cannot clearly be identified either as a relationship-based or transaction-based customer, it pays higher loan rates than relationship- and transaction-based loans in case a personal guarantee is used in the loan contract. Other parameter signs are in line with previous regressions; loans are positively related, and firm size is negatively related with the cost of credit.
The empirical results for H2 support the suggestion that personal guarantees are connected with higher loan premiums. Although the use of personal guarantees does not lower the loan premium, which would indicate risk mitigating, the results show a positive relationship between the use of personal guarantees and a loan premium. A similar association between the use of personal guarantees and the cost of loan is found in Brick and Palia (2007) and Calcagnini, Farabullini, and Giombini (2007). In addition, Hernández-Cánovas and Martínez-Solano (2006) find that the use of personal guarantee is increased with the duration of relationship. In this paper, however, no statistically significant association between the personal guarantee and the duration of relationship is found.
The results also confirm that transaction-based loans are more expensive than relationship-based loans, measured by both the personal guarantee (pergua) and the lending structure emphasis (structure). This connection to the relationship-based loans, that is, beneficially priced loans, is supported in the literature (Angelini, Di Salvo and Ferri 1998; Berger and Udell 1995; Blackwell and Winters 1997; Boot and Thakor 1994; D'Auria, Foglia, and Reedtz 1999; Mörttinen 2000; Niskanen and Niskanen 1999; Peltoniemi 2007a). It seems that the use of personal guarantees in transaction-based lending is not a mitigating tool as means to manage risk but rather a consequential indication of a high-risk firm, or a strengthening act to confirm the transaction-based lending between bank and firm. As a robust test, it is confirmed that a personal guarantee is more likely to be associated with a high-risk firm than a low-risk firm (probit model results not reported).
A sensitive analysis is conducted in Table 7, where unsecured loans are separated from secured loans. This is important information when analyzing the role of personal guarantees and collateral, as the differences between secured and unsecured loans should be existent through their unique nature. When the loan is secured, the total collateralization level exceeds the total of liabilities, and for unsecured loans, the total of liabilities exceeds the total collateralization.
|Unsecured Loans||Secured Loans|
|Log (firm age)||0.098||0.248||0.080||0.117|
|Log (Firm size)||−0.017||0.705||−0.123||<0.0001|
Comparing the parameter values of pergua in regression (1) and regression (2), it is found that the effect of personal guarantees on the cost of the credit is more strongly in the sample of unsecured loans. For both samples, the pergua is positive and statistically significant at 5 percent level (0.315/unsecured, 0.193/secured). The magnitude is large especially for unsecured loans. It seems that personal guarantee matters and is important when the total collateralization does not reach the total of liabilities. This may be an indication of a supplementary rather than complementary role that the personal guarantee has in these loan contracts. For collateralization, no significant deviation from value zero is observed. The structure variable, which measures the orientation of the lending structure, is negatively associated with the cost of credit in both regressions and with a similar magnitude. There seems to be no differences with the loan pricing for unsecured or secured loans whether the lending structure is transaction-based or relationship-based oriented. In both unsecured and secured loans, relationship-based lending is related with lower loan rates than transaction-based loans.
This paper analyzes the characteristics of the personal guarantee in small business loans in Finland. As a bank-based financial system, the Finnish data used in this study are comparable with countries with similar type of financial systems, such as Germany, Italy, and Japan. The data contain unique information of the credit files, such as relationship, collateral, firm, and loan characteristics.
We address two main research questions: (1) are personal guarantees associated with different lending structures, and (2) do personal guarantees affect the cost of the credit? The investigation itself and the results of these research questions make a definite contribution to the current literature in three main respects. First, this study is, to the best of the authors' knowledge, the first empirical examination to analyze personal guarantees and the cost of credit in different lending structures. Second, in this study, we are able to introduce a factor new to the literature in order to identify the lending structure—whether its orientation is transaction based or relationship based. This type of contextual lending structure component is very rare or even nonexistent in previous research. Third, we are able to use accurate measurements of the cost of credit and the level of collateralization in the analysis. The cost of credit is calculated as an effective rate that includes the arrangement fees and/or provision for the bank in addition to the loan premium. This information provides a more accurate measurement of the cost of credit compared with the loan premium without the arrangement and/or bank's provision. The previous literature often applies the loan premium without calculating the effective rate, mainly due to data restrictions. Also, in this study the level of collateralization is measured as a continuous measurement, whereas in many studies it is available as a dichotomy variable (more restricted information content for research purposes).
In transaction-based lending, the level of asymmetric information is typically higher than in relationship-based lending, where information flow is more constant and ongoing. There are differences also in the length and scope of the relationship and in the maturity and size of the loans between these lending structures. In this study, a continuous joint variable is formed from these four variables, which is assumed to capture the orientation of the lending structure for the loan whether it emphasizes more transaction-based or relationship-based lending. The joint variable indicates whether the use of personal guarantees is associated with transaction-based or relationship-based lending and further analyzes the effect on the cost of the credit. In the analysis, it is tested whether a personal guarantee is used to mitigate problems related to asymmetric information (moral hazard and adverse selection), or to confirm the existing relationship between bank and firm. The mitigating “tool” refers to risk management if the use of personal guarantees lowers the cost of credit, and confirmation “tool” is linked to the higher cost of credit when the loan is personally guaranteed. Related literature supports both views. The empirical analysis has found answers to the research problem where the association between the role of personal guarantees (mitigation/confirmation) and lending structure is previously unknown.
The main results of this paper present new knowledge and understanding on how the personal guarantee affects the cost of credit when the lending structure is correctly identified and assessed. Our findings support that (1) personal guarantees are generally related to transaction-based lending, and the use of personal guarantee increases the cost of credit, and (2) when separating relationship-based firms from transaction-based firms, it is found that transaction-based customers pay higher loan premiums when the loan is personally guaranteed than when compared with similar cases for relationship-based customers. Thus, transaction-based loans are more expensive than relationship-based loans when personal guarantee is considered. These results are considered robust after assessing the relationship, firm, loan, and collateral characteristics. The results suggest that the personal guarantee is used as a confirmation tool in transaction-based firms on stronger level than in relationship-based firms. Another interpretation suggests that a personal guarantee increases the loan premium in transaction-based loans because it simply relates to higher firm risk.
A sensitive analysis shows that the effect of a personal guarantee on the loan premium is strong especially in unsecured loans. No statistical significance is found in secured loans. This implies that the personal guarantee may be a supplementary component of the total collateralization when real assets as collateral do not cover the total bank liabilities of the firm. This type of setting indicates an unsecured loan.
As a managerial implication in small business bank lending, the personal guarantees do affect loan pricing, which may have consequent effects on firm behavior in bank loan contracts. As the role of personal guarantees seems to be different depending on the lending structure, it is important to notice some key issues. In transaction-based loans, where the length and scope of relationship and maturity and size of loans are usually “small,” it is vital to assess the relationship between the use of personal guarantees and loan premium so that no incentive can be created for the firm to commit opportunistic behavior (moral hazard) as a result of this assessment. To do so, a managerial recommendation to banks is to use personal guarantees in transaction-based loans if possible, to commit the business owner to participate in the responsibility. In addition, the loan premium should reflect the firm risk appropriately but not excessively in order to prevent the business owner to commit potential moral hazard problems.
Moreover, in relationship-based loans, it would be rewarding to provide a balanced trade-off for the firm between the use of personal guarantees and a loan premium in such a way that firm owners would be encouraged to pledge personal guarantees against lower loan premiums. These implications for small business bank lending are subjectively deduced from the empirical findings of this study.
The scope or breadth of bank–firm relationship is a dimension that measures the strength of relationship, in addition to the duration of relationship. For instance, the scope of relationship can define the relationship's breadth by the number of financial services (loans, deposits, savings accounts, financial management services, etc.) the firm has in the bank.
In return to getting access to the loan files of this bank, we have promised confidentiality to the bank and the customers regarding identity and location.
All the firms are small businesses with fewer than 30 employees, except one with 160 employees (descriptive statistics, not an explicit restriction).
These include 55 loans with fixed rates and 53 lacking loan rate information.
The number of observations in our regression models drops dramatically mainly because of the lack of firm-specific information from financial statements. Thus, the reduced sample size is due to missing data. The regressions include 279–285 observations.
Personal guarantee as a part of total collateralization has minor significance in a monetary sense compared with the real assets that are used for collateralization.
Our data include 55 observations (4.6 percent) of collateralization that exceeds the value of two (overcollateralized). We have set these values as equal to two for simplicity. We assume that the bank is indifferent to the extent of overcollateralization when it is greater than two.
The firm's liabilities to the subject bank include all performing loans as well as bank guarantees to the firm.
We use two aggregate risk levels to generate practical interpretation of the information content whether the firm is identified as high- or low-risk firm. We tested regressions also with the five different risk classes and report that the identification to high/low riskiness is more understandable for the reader with the aggregation. A similar setting with the data examination has been implied in Peltoniemi (2007a, b), which is referred to in this approach.
The number of observations in our regression models drops dramatically mainly because of the lack of firm-specific information from financial statements. Thus, the reduced sample size is due to missing data. The regressions include 279–285 observations.
The variable structure is used in Table 4/regression 2, replacing the variables duration, financial services, maturity, and loan size, as established accordingly.
The variable structure is used in the Table 4/regression 2, replacing the variables duration, financial services, maturity, and loan size, as established accordingly.
- 1998). “Availability and Cost of Credit for Small Businesses: Customer Relationships and Credit Cooperatives,” Journal of Banking and Finance 22, 925–954. , , and (
- Bank of Finland (1996). Financial Markets in Finland. Bulletin. Special issue, ISBN 951-686-523-2.
- 2009). “Does Debtor Protection Really Protect Debtors? Evidence from the Small Business Credit Markets.” AFA 2009 San Francisco Meetings Paper. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1101309. , , and (
- 1998). “The Efficiency Cost of Market Power in the Banking Industry: A Test of the ‘Quiet Life’ and Related Hypotheses,” Review of Economics and Statistics 80, 454–465. , and (
- 1995). “Relationship Lending and Lines of Credit in Small Business Finance,” Journal of Business 68, 351–381. , and (
- 1998). “The Economics of Small Business Finance: The Roles of Private Equity and Debt Markets in the Financial Growth Cycle,” Journal of Banking and Finance 22, 613–673. , and (
- 2003). “Small Business and Debt Finance,” in Handbook of Entrepreneurship. Eds. Z. Acs and D. Audretsch . Boston, Dordrecht, London: Kluwer Academic Publishing, 299–330. , and (
- Collateral, Loan Quality, and Bank Risk,” Journal of Monetary Economics 25, 21–42. , and (1990). “
- 1999). “The Consolidation of the Financial Services Industry: Causes, Consequences, and Implications for the Future,” Journal of Banking and Finance 23, 135–194. , , and (
- 1987). “Collateral and Rationing: Sorting Equilibria in Monopolistic and Competitive Credit Markets,” International Economic Review 28, 671–689. , and (
- 1985). “Screening vs. Rationing in Credit Markets with Imperfect Information,” American Economic Review 75, 850–855. (
- 1987). “The Role of Collateral in Credit Market with Imperfect Information,” European Economic Review 31, 887–899. (
- 1994). “The Role of Collateral in a Model of Debt Renegotiation,” Journal of Money, Credit and Banking 26, 72–86. (
- 1993). “Contemporary Banking Theory,” Journal of Financial Intermediation 3, 2–50. , and (
- 1997). “The Relationship between U.K. Banks and Their Small Business Customers,” Small Business Economics 9, 167–178. , and (
- 1997). “Banking Relationships and the Effect of Monitoring on Loan Pricing,” Journal of Financial Research 20, 275–289. , and (
- 2003). “Short-Term Loans and Long-Term Relationships: Relationship Lending in Early America,” Journal of Money, Credit and Banking 35, 485–506. (
- 2000). “Relationship Banking: What Do We Know?,” Journal of Financial Intermediation 9, 7–25. (
- 1994). “Moral Hazard and Secured Lending in an Infinitely Repeated Credit Market Game,” International Economic Review 35, 899–920. , and (
- 1991). “Secured Lending and Default Risk: Equilibrium Analysis and Policy Implications and Empirical Results,” Economic Journal 101, 458–472. , , and (
- 2007). “Evidence of Jointness in the Terms of Relationship Lending,” Journal of Financial Intermediation 16(3), 452–476. , and (
- 2007). “Loans, Interest Rates, and Guarantees: Is There a Link?,” Working Paper, presented at the Workshop “Interest rates in retail banking markets and monetary policy,” European Central Bank, Frankfurt, February 5, 2007. University of Urbino, Italy. , , and (
- 2001). “The Separating Role of Collateral Requirements in Credit Markets with Asymmetric Information,” Lineex Working Papers, 23/01, 1–33. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=283179. , , and (
- 1976). Industrial Organization in Japan. Washington, DC: Brookings Institution. , and (
- 1985). “Asymmetric Valuations and the Role of Collateral in Loan Agreements,” Journal of Money, Credit and Banking 17, 84–95. , and (
- 1987). “Collateral and Competitive Equilibria with Normal Hazard and Private Information,” Journal of Finance 42, 345–364. , and (
- 1998). “Bank Lending to Small Businesses and the Use of CRA Loan Data to Measure Market Structure,” Working Paper (Federal Reserve Board, Washington, DC. , and (
- 1999). “Bank Interest Rates and Credit Relationships in Italy,” Journal of Banking and Finance 23, 1067–1093. , , and (
- 2000). “Relationship Lending within a Bank-Based System: Evidence from European Small Business Data,” Journal of Financial Intermediation 9, 90–109. , and (
- 1994). Banks, Finance and Investment in Germany. Cambridge, UK: Centre for Economic Policy Research. , and (
- 2005). “Empirical Determinants of Relationship Lending,” Journal of Financial Intermediation 14, 32–57. (
- 1999). “Impact of Participative Service Relationships on Quality, Satisfaction and Retention: An Exploratory Study,” Journal of Business Research 46, 121–132. , and (
- 1990). Hausbankbeziehungen als instrument der Dindung zwischen Banken und Unternehmen. Dissertation thesis, University of Bonn. (
- 1998). “Multiple Banking Relationships and the Fragility of Corporate Borrowers,” Journal of Banking and Finance 22, 1441–1456. , , and (
- 1989). “Equilibrium Loan Pricing under the Bank-Client Relationship,” Journal of Banking and Finance 13, 221–235. , , and (
- 1998). “Lending Relationships in Germany—Empirical Evidence from Survey Data,” Journal of Banking and Finance 22, 1317–1353. and (
- 2006). “The Price of UK Commercial Credit Lines: A Research Note,” Journal of Business Finance and Accounting 33, 932–938. , , and (
- 2006). “Banking Relationships: Effects on Debt Terms for Small Spanish Firms,” Journal of Small Business Management 44(3), 315–333. , and (
- 1990). “The Role of Banks in Reducing the Costs of Financial Distress in Japan,” Journal of Financial Economics 27, 67–88. , , and (
- 2006). “A Lender-Based Theory of Collateral,” Journal of Financial Economics 84(3), 826–859. , and (
- 1989). “The Incidence of Secured Debt: Evidence from the Small Business Community,” Journal of Financial and Quantitative Analysis 24, 379–394. , and (
- 2000). “The Importance of Bank Seniority for Relationship Lending,” Journal of Financial Intermediation 9, 57–89. , and (
- 2000). “Number of Bank Relationships: An Indicator of Competition, Borrower Quality, or Just Size?,” CFS Working Paper No. 2000/06 https://www.ifk-cfs.de/fileadmin/downloads/publications/wp/00_06.pdf. , and (
- 2000). “Collateral vs. Project Screening: A Model of Lazy Banks,” CEPR Discussion Paper no. 2439. , , and (
- 2000). “Essays on Loan Pricing and the Credit Channel of Monetary Policy,” Research Reports, Kansantaloustieteen laitoksen tutkimuksia 85, 1–117. Department of Economics, University of Helsinki. (
- 1984). “The Role of Financial Corporate Grouping,” in Economic Analysis of the Japanese Firm. Ed. M. Aoki . New York: North Holland. (
- 1999). “Pankkikilpailun ja pankkisuhteiden vaikutus yrityslainojen saatavuuteen ja ehtoihin. Tutkimus pienistä ja keskisuurista yrityksistä. (The effect of bank competition and bank relationships on the availability and terms of small business loans),” Tampereen yliopisto, Yrityksen taloustieteen ja yksityisoikeuden laitos, Sarja A1: Tutkimuksia 46: 1–120. , and (
- 2005). “The Role of Collateral and Personal Guarantees in Relationship Lending: Evidence from Japan's Small Business Loan Market,” RIETI Discussion Paper Series 05-E-027. http://www.rieti.go.jp/jp/publications/dp/05e027.pdf. , and (
- 1970). “A Credit Scoring Model for Commercial Loans,” Journal of Money, Credit and Banking 2, 435–445. (
- 2008). “Lending to Small Businesses: The Role of Loan Maturity in Addressing Information Problems,” Small Business Economics 30(4), 361–383. , and (
- 2004). “The Value of Relationship Banking: Empirical Evidence on Small Business Financing in Finnish Credit Markets,” Acta Universitatis Ouluensis Oeconomica, G15. Oulu University Press. (
- 2007a). “Sources of the Benefits of Relationship Banking: Evidence from Small Business Financing in Finland,” Journal of Financial Services Research 31(2–3), 153–171. (
- 2007b). “Collateral, Borrower Risk, and Relationship Banking,” The Finnish Journal of Business Economics 4, 393–414. (
- 1994). “The Benefits of Lending Relationships: Evidence from Small Business Data,” Journal of Finance 49, 3–37. , and (
- 1995). “The Effect of Credit Market Competition on Lending Relationships,” Quarterly Journal of Economics 110, 407–433. , and (
- 2002). “Secured Lending and Borrowers' Riskiness,” SSRN Working Paper series. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=302124. (
- 2004). “The Role of Guarantees in Bank Lending,” EFMA 2004 Basel Meetings Paper. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=498982. (
- 1995). “Covenants and Collateral as Incentives to Monitor,” Journal of Finance 50, 1113–1146. , and (
- 2006). “Firm Risk and Collateralized Asset Choice in Small Business Bank Lending: Theory and Evidence,” SSRN Working Paper series. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1090012. (
- 1999). “Bank Consolidation and Small Business Lending. A Small Firm Perspective,” in Business Access to Capital and Credit. A Federal Reserve System Research Conference. Eds. J. L. Blanton , A. Williams and S. L. W. Rhine . Washington, DC: Federal Reserve, 328–361. , and (
- 1990). “Asymmetric Information, Bank Lending, and Implicit Contracts: A Stylized Model of Customer Relationship,” The Journal of Finance 45, 1069–1089. (
- 1983). “Costly Information Production Equilibria in the Bank Credit Market with Applications to Credit Rationing,” The Journal of Financial and Quantitative Analysis 18(2), 229–256. , and (
- 1995). “Financial Intermediation and the Market for Credit,” in Handbooks in OR and MS, 9 Eds. R. A. Jarrow , V. Maximovic , and W. T. Ziemba . Amsterdam, The Netherlands: Elsevier, 1073–1103. (
- 1998). “On the Costs of a Bank Centered Financial System: Evidence from the Changing Main Bank Relations in Japan,” Journal of Finance 53, 635–672. , and (
- 1997). “Why Is Bank Debt Senior? A Theory of Asymmetry and Claim Priority Based on Influence Costs,” Review of Financial Studies 10, 1203–1236. (
- 1980). “A Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for Heteroskedasticity,” Econometrica 48(4), 817–838. (