The lender-borrower relationship has long been studied in prior studies. There are two sides of credit borrowing in the lender-borrower relationship, the demand side and the supply side. One would expect that factors from both sides have effects on the lender-borrower relationship. An important strand of research focuses on borrower effects on this relationship and on the setting of loan contract terms. Taking the determination of collateral as an example, Chan and Kanatas (1985) show that, in cases where the lender and the borrower have different opinions about the borrower’s project, collateral will be offered by the borrower when the lender’s valuation of the project is lower than the borrower’s.

Higher quality borrowers signal their creditworthiness by offering more collateral. Besanko and Thakor (1987) also find a positive relationship between collateral and borrower creditworthiness. In contrast, Berger and Udell (1990) and Harhoff and Korting (1998) find a positive relationship between collateral and borrower risk in the context of small business loans. Aside from collateral, a number of studies also examine the impact of borrower characteristics on the determination of loan price (Angbazo, Mei and Saunders (1998), Gorton and Kahn (2000), among others). In this strand of the literature, borrower effects on the determination of loan contract terms have been widely explored while lender effects have not.

Another strand of research addresses the effects of lender characteristics on loan contract terms. Studying different types of financial intermediaries, Carey, Post and Sharpe (1998) find evidence that compared to banks, finance companies seem to be more likely to make secured loans frequently and lend to riskier borrowers. Hubbard, Kuttner and Palia (2002) incorporate another lender attribute, bank financial health, and show that low-capital banks tend to charge higher loan rates than well-capitalized banks. Coleman, Esho and Sharpe (2002) examine further lender characteristics, and state that bank monitoring ability, bargaining power, risk and syndicate structure have significant influence in determining loan maturity and pricing. However, studies about lender effects on the lender-borrower relationship and especially the determination of loan contract terms remain scarce in this literature.

This paper is an empirical study of bank effects on the setting of loan prices, taking into account the influences of both bank and borrower attributes. Further lender characteristics are included in this study of bank effects on loan prices. Notably, a new dimension of bank characteristics, the number of lenders at the loan level, is introduced. This is motivated by recognizing the important influence of single versus multiple banking relationships documented in prior studies. It has been well documented that the number of banking relationships a borrower maintains at a given moment plays an important role in the lender-borrower relationship. Petersen and Rajan (1994) find that small firms borrowing from multiple banks are of lower creditworthiness than those borrowing from a single bank.

For such firms, borrowing from fewer banks generally increases credit availability and lowers the cost of funds. Houston and James (2001) observe that firms relying on a single bank exhibit greater sensitivity of investment to cash flow than firms maintaining multiple bank relationships or borrowing from public debt markets. Harhoff and Korting (1998) empirically investigate the role of lending relationships in determining the costs of external funding, and document that the number of relationships increases with the firm’s age, size, and leverage. Carletti (2000) studies the link between the number of bank relationships and banks’ incentives to monitor, along with the effect of such link on loan rates and firms’ choice between single and multiple relationships.

Recognizing the important influence of the number of lenders on the lender-orrower relationship, the current paper incorporates a new variable, the number of lenders at the loan level, along with other lender characteristics, to examine bank effects on loan yield spreads. The number of lenders at the loan level is expected to affect the setting of loan contract terms. The effects of risk diversification, monitoring duplication, negotiation complexity and bargaining power constitute the main issues. As for multiple-lead-lender loans, the negotiation process between a borrower and multiple lenders becomes more complex than is the case with a single lender.

Considering the potential for monitoring duplication and benefit sharing in cases of loans with multiple lenders, it is expected that the lenders’ monitoring effectiveness is affected by the presence of multiple lenders, as is the settings of loan contract terms. The bargaining power of each party will change according to its level of commitment to the loan contract. Furthermore, the presence of multiple lenders in a given loan contract could suggest that there is unfavorable information about the borrower and thus that the original bank is unwilling to lend to the borrower on its own. Including more lenders in a loan contract could diversify risk and reduce each lender’s exposure to firm-specific risk, while also serving to discourage strategic default on the part of the borrower (Esty and Megginson (2003)). In brief, the number of lenders at the loan level is expected to affect the setting of loan contract rates.

This study extends the existing literature by emphasizing the significant influence of bank characteristics on loan yield spreads and provides evidence that borrower characteristics are important determinants of loan yield spreads. Bank effects on loan yield spreads are examined after controlling for the effects of borrower and non-yield-spread loan characteristics. Our main findings are that banks with greater monitoring power and riskier banks with lower capital-asset ratios extract higher rents, which is consistent with the findings in prior studies.

Importantly, the new dimension of lender characteristics the number of lenders for a given loan contract is shown to have a significant influence on loan yield spreads. The positive relationship between the number of lenders at the loan level and loan yield spreads suggests that the presence of multiple lenders is associated with the duplication of monitoring, complex negotiation processes, the possibility of unfavorable information about the borrower, and the intention to discourage borrowers’ strategic default.

The contributions of this paper to the existing literature are threefold. First, considering influences from both the demand side and the supply side of credit borrowing, we assemble bank and borrower financial variables in order to fully investigate bank effects on the determination of loan yield spreads, controlling for the effects of non-yield-spread loan features and borrower characteristics. Second, we incorporate a new dimension of bank characteristics in the study of bank effects on loan yield spreads. Specifically, we introduce the number of lenders at the loan level as a variable affecting the setting of loan yield spreads, along with bank size, bank risk, and bank monitoring power.

Third, continuing our focus on the role of multiple lenders, in the case of syndicated loans we include all lead banks for a given loan contract in this study. For multiple-lead-bank syndicated loans, we assign each lead bank a weight according to its contribution to the loan facility based on its share of the syndicated loan, data which is available from DealScan. By so doing, we avoid omitting data on multiple lead banks which provide valuable information about bank effects on loan yield spreads. In contrast, Coleman, Esho and Sharpe (2002) study only the lead bank which contributes the largest portion of the syndicated loan. This could lead to a biased understanding of the effects of lender characteristics due to the omission of substantial lender information in the case of multiple-lead-bank syndicated loans.

The remainder of the paper is organized as follows. In the next section, we discuss proxies for bank, borrower, and non-yield-spread loan characteristics. The central testing hypotheses in this study are also discussed in section II. Section III describes the data and the empirical approach we use. Our empirical tests are reported in Section IV. Section V concludes.

Download
PDF Ebook Bank Effects and the Determinants of Loan Yield Spreads