LITERATURE REVIEW
Long-term and intensive ties between banks and their client firms generate value and increase economic efficiency. Little is known, though, on how this value is divided among the stakeholders involved in such relationships. In the course of building the relationship, the lender accumulates borrower-specific information which gives him significant benefits. To the extent that the lender passes these benefits to the borrower, relationships will also be valuable from the borrower's point of view. Based on existing literature, most researchers and scholars agree that a close bank relationship has both advantages and disadvantages to companies. In different economies, banks control and affect firms not only through loans, but also through equity ownership and proxy voting.
Bank-Firm Relationship Advantages
Diamond (1984) develops a model indicating that under a close bank-firm relationship, not only the bank monitoring costs would be mitigated at its lowest level, but does it suggest a solution to principal agent problem. It can effectively prevent transfer of wealth from lenders to shareholders, and improve corporate governance (Limpaphayom and Polwitoon, 2004). Theorists also suggest that this crucial relationship reduces the information asymmetries and incentive problems. It is broadly agreed that such problems (asymmetry and incentive) affect corporate investment to a large degree. Therefore, firms with such a relationship should improve their access to capital and invest more in the market.
This argument is also well supported by Diamond's model (1991). Reputation effect is a key factor mentioned in his study. Firms acquiring loans from their banks could build a reputation and then disclose this information to the market. Diamond (1991) assumes that if moral hazard occurs in many places, this behavior indeed provides a certification of good credit; it is also viewed as a signal to outside investors. Ultimately, reputation effect helps firms to raise more funds on public markets either through equity or arm's length debt in the future. Empirical finding indicates that investment for firms with close ties to a bank are less sensitive to liquidity than others financed with arm's length debt. Petersen and Rajan (1994) find a phenomenon of easy availability of funds to firms with close bank ties based on small U.S. business data. Best and Zhang (1993) illustrate that there is a positive and statistically significant relationship between the bank loan announcements in the stock market and the firm value.
The strong bank-firm relationship maintains a significant role in the modern financial markets, especially during the financial distress. Many researchers agree that debt burdens make companies more unstable in economic turmoil. Hoshi, Kashyap and Scharfstein (1990) argue that a close bank relationship can reduce the costs of financial instability. Moreover, it is possible that a firm's main bank is also the main bank of its suppliers and even its customers. Therefore, this relationship can overcome the issues of credit expansion and trustworthiness among customers, suppliers, firms and banks.
Empirical analysis in Hoshi et al. (1990) proves that this sort of relationship helps to reduce the costs of a financial crisis. They find a main bank would like to invest more in firms with such bank affiliations after business financial crisis than those without. There is another reason why firms prefer a bilateral financing arrangement, i.e. to reduce the costs of information disclosure. Campbell (1979) points out that the small unknown companies with creative and promising ideas are viewed as major participators. They do not have a lot of internal funds or external opportunities raising money. More importantly, they do not need to disclose the proprietary technological information related to their ‘leap-frog' product innovation to their competitors either in direct or indirect ways.
Bank-Firm Relationship Disadvantages
Theoretical analyses done by Rajan (1992) indicate that the relationship-building process between firms and banks is very costly to the borrowing side. Since the lending side flexes the borrowing agreement and gives the borrower financial aid at a favorable rate at the beginning, in return, the financial institution can accumulate a great amount of non-public material about the firm through this relation and finally have a bargaining power over the firm. Furthermore, by using this monopoly power, a bank could even threat to cut off a firm's loan or simply charge it at a higher rate during the process of relationship building (Diamond, 1991). This is so, because there is an extra searching cost for the firm looking for an alternative bank to replace the current one. Also, if a firm discontinues the relationship with its current bank, others would view this as a negative sign due to information capture and adverse selection problems (Castelli, Dwyer Jr. & Hasan, 2006).
These theories suggest another agency cost to borrowing firms. If a close bank relationship makes firms' access to capital easier than those without such a relationship, better performance should be observed i.e. firms enjoy faster growth or high profitability among their peers. However, some empirical evidence shows opposite stories. Weinstein and Yafeh (1998) document a negative relationship between profitability and the degree of bank-firm relationship. Their sample period is prior to the Japanese financial market liberalization from 1977 to 1986. The evidence shows that although the strong bank ties did improve access to funds, this phenomenon has vanished since deregulation in early 1980s. Meanwhile, they conclude that banks conduct rent-seeking behavior which means banks use their monopoly power to siphon profits from their client-firms.
Further, they find that banks tend to shy away from risky but profitable investment and put pressure on their client-firms. This could be one of the reasons that companies with close bank ties cannot beat non-relationship firms in the Japanese atmosphere. Moreover, Agarwal and Elston (2001) examine a hundred large listed German firms from 1970 to 1986 and they do not find any evidence showing the benefits of German universal banking relationship.
Nevertheless, there is a negative relation between interest payment and bank-influenced firms although it is only significant at the 10% level.
Intensive Bank-Firm Relationships and Bank Performance
The modern literature on financial intermediation has long emphasized the value- creation function of lending relationships. In a context of asymmetric information in credit markets, lending relationships facilitate the information exchange between the borrower and the lender through repeated interaction over the duration of the relationship and through the provision of multiple financial services. Lenders invest in generating information from their client firms and borrowers are more inclined to disclose information (Boot 2000). Consequently, the information asymmetries between the bank and the firm are lessened as time goes by.
This process enhances economic efficiency through many channels. First, having a long-term horizon facilitates the design of implicit credit contracts over the duration of the relationships that may increase value. This is achieved, for instance, through reduction in welfare- dissipating collateral requirements, through the deployment of welfare-enhancing inter-temporal tax-subsidy schemes in loan pricing (Petersen and Rajan 1995), as well as through more flexible contracting terms (Boot, Greenbaum and Thakor 1993). Second, the reusability of the information generated by the lender over repeated transactions and over time is also beneficial in terms of savings on the fixed cost of screening and monitoring (Boot, Greenbaum and Thakor 1993).
Third, it avoids the free-rider problem of monitoring since the bank internalizes the benefits of such investments. Higher monitoring levels increase value since, for instance, they help solve principal-agent problems of managerial behavior. Additionally, relationship banks develop sector-specific expertise that enhances the value of financed projects (Boot and Thakor 2000). Furthermore, relationship lending contributes greatly to economic growth by promoting the efficient allocation of capital as long as better informed banks provide credit to the most productive projects first. At the same time, close bank-firm relationships entail some costs to the firm. The most significant cost is that having a single relationship gives an informational monopoly to the only informed bank, which can impose hold-up costs for the firm (Rajan 1992).
Additionally, the soft-budget constraint problem, that is inefficient loan renewal decisions, is more likely to happen when only one lender has the option to bail out the firm in case of distress (Dewatripont and Maskin 1995); managers are more inclined to default strategically to divert cash to themselves when there is only one creditor than when there are many creditors (Bolton and Scharfstein 1996). In spite of these problems, existing empirical research on relationship lending stresses that benefits outweigh the costs, that is, relationships generate value. Only to the extent that such value created is passed on to or shared with the borrower, through lower cost of borrowing, more flexible contract terms, and so on, a relationship will also be valuable for a firm that borrows from its relationship lender. That is to say, a firm will benefit from relationship lending as long as the bank shares the value with the borrower. In consequence, if lending relationships are valuable, it should be reflected in the overall firm performance.
The authors focus on small firms for various reasons. First, small firms are more likely to suffer from information problems in the capital markets. The value of relationship lending, which is based on a bank gathering soft information, is likely to be higher for the smallest, youngest and most opaque firms because of the lack of credit history, the impossibility to credibly disclose their quality, and the lack of separation between ownership and management, which increases the asymmetric information between insiders and outsiders (lenders). Second, small firms are typically restricted to obtaining external finance only from financial institutions. Public debt markets are only accessible to large firms. While little more than 500 companies access the organized capital market, more than 2.5 million small firms rely on financial intermediaries to finance their investment projects. Third, small firms are extremely important for the economy: roughly 50 percent of the 3 million firms do not have employees while 1.3 million have between one and nine employees; 180,000 firms have between 10 and 499 employees and only 1,700 firms have more than 500 employees.
Previous empirical literature has already investigated the real effects of close bank- firm relationships. Hoshi, Kashyap and Scharfstein (1990) focus on Japanese listed firms that are in financial distress. They find that firms that have close financial relationships to their banks (are in same industrial group and have a larger share of bank loans from the same lender) invest more and have a higher sales growth after a period of financial distress than non-group firms. Weinstein and Yafeh (1998) analyze 6836 Japanese firms in the period 1977-1986 and find that firms with close ties to their lenders exhibit slow growth rates and lower profitability. That shows that in Japan most of the benefits of bank-firm relationships are appropriated by the banks.
Agarwal and Elston (2001) use a sample of large listed and unlisted German firms in 1970-86 and find that bank-influenced firms do not have higher profitability or growth (bank influence is defined as financial institutions owning part of the firm.). They interpret these results as evidence that German universal banks engage in rent-seeking activities. Chirinco and Elston (2006) use data on 91 listed firms in Germany and find that bank influence is not associated with a reduction of neither finance costs nor a change in profitability. It is interesting to see that these three studies have been conducted in Japan and Germany, which are the economies that have been cited frequently as ideal to study bank-firm relationships.
None of the above papers has properly taken into account the potential endogeneity between bank relationships and firm performance in the empirical models. Therefore, the estimations could be biased. There are two studies that deal with the reverse causality issue using information on large listed firms. Degryse and Ongena (2001) use a panel dataset of 235 publicly listed Norwegian firms between 1979 and 1995 and find that firms with a bilateral relationship are more profitable. Fok, Chang and Lee (2004) examine 178 firms traded on the Taiwan Stock Exchange between 1994 and 1998. They find that the number of foreign-bank relationships is positively related to firm performance; however, the number of domestic-bank relationships is negatively related to firm performance. Since domestic-bank loans are more likely to be relationship loans, the results are interpreted as evidence that bilateral relation- ships are profitable. However, the existing papers have been conducted for listed firms. As noted above, bank relationships are particularly important for small and informationally opaque firms.
A few papers contribute to the existing literature by providing empirical evidence for small firms. Furthermore, the authors carefully design an empirical strategy to deal with identification problems inherent in this type of research. In particular, they use a very large panel and instrumental variables estimation in order to fully exploit the exogenous information present in the data. This paper is also related to a branch of empirical papers that search for evidence on the uniqueness of bank loans with respect to other sources of finance.
Typically, these empirical papers measure the impact on a firm's stock price when information about a bank relationship is revealed. The starting point of this literature is the work of Fama (1985), who argues that bank-firm relationships are important since they affect a firm's ability to raise capital, from both within the bank and from other non-bank sources. Based on that observation, James (1987) compares the stock price reaction to announcements of private and public debt and bank loans. He finds that bank loan announcements are associated with positive and statistically significant stock price reactions, while announcements of private and public debt are not followed by such a response. Numerous event studies have expanded the results in James (1987). Our research question, namely to estimate the value of close bank relationships for the borrowers, is similar to this branch of the literature. However, since these analyses rely on the stock price reaction, this type of event study can only be implemented for large, listed firms.
