...offered by the bank, and the other is commitment from the bank to extend loans in times of crisis. Previous empirical studies of relationship lending focus primarily on the price of services, specifically loan rates, offered by the bank and do not control for the possibility that bank relationships can be intertemporal in nature. (1) In other words, the interest rate a firm pays can be a function of a number of factors, including the borrower's expectation of lower rates in the future, its purchases of other services, and the probability of being supported by the bank in the future, as well as the lender's commitment to doing business in times of crisis. Such diverse possibilities can cause loan rate and relationship lending to have an ambiguous correlation. To illustrate, although Berger and Udell (1995) report that relationship duration has a negative effect on the loan rate, Degryse and Van Cayseele (2000) provide evidence that the correlation is positive, and Petersen and Rajan (1994) and Elsas and Krahnen (1998) report no significant correlation.

In contrast, our study examines the association between relationship lending and credit availability in times of crisis. By focusing not on the price of services but on the extension of loans in times of crisis, our approach seeks to observe whether a bank relationship has any value when it is needed the most.

The intuition that underlies our empirical model is that a firm enters relationship lending through a concentrated business relationship. The firm's objective in doing so is to increase the likelihood that it will have access to bank credit in tough times (Petersen and Rajan 1994). Alternatively, a firm may seek multiple lenders to secure more stable financing because a single bank might experience its own liquidity problem (Detragiache, Garella, and Guiso 2000). We examine the validity of these competing objectives by investigating whether firms in Indonesia, the Republic of Korea, the Philippines, and Thailand that had strong lending relationships benefited from better access to credit during the Asian financial crisis that lasted from July 1997 through the end of 1998.

We measure the intensity of the bank relationship by the number of financial institutions with which the firm does business, and we define intensity as highest if a firm uses only one bank. Our data are from a survey conducted by the World Bank; the survey includes questions about whether firms had sufficient access to credit during the crisis and how many financial institutions the firm used in the course of its business.

Another problem with prior studies--besides not controlling for the possible intertemporal nature of bank relationships--is that they do not distinguish between opaque and transparent borrowers, a distinction based on the quality of the borrower's financial statements. The effects of financial transparency and disclosure on lending are important: to the extent that banks function as information processors, the transparency and disclosure reflected in independently audited financial information can either substitute for or complement a long-term lending relationship. Berger and Udell (2005) hypothesize that the value of relationship will be weaker for transparent firms because for those firms the lender's information problem in underwriting can be addressed more cost effectively with certified financial statements (for example) than with relationship lending. Our data set allows us to test this hypothesis because the World Bank survey includes questions about whether the firm voluntarily supplied audited financial statements or whether the bank required

the firm to supply audited financial statements. Such data are rarely available, and they make it possible for us to discern the effect of transparency and disclosure on credit availability.

Our main finding, after we control for the endogenous decision to post collateral, the endogenous choice of number of lending relationships, and several firm-specific characteristics, is that during the Asian financial crisis, strong lending relationships increased both the economic and the statistical likelihood of obtaining credit for Korean and Thai firms. For example, during the crisis, on average, firms in Korea with one lending relationship have a 17.7% higher chance of obtaining credit than do firms with two lending relationships. In Thailand, the comparable average is 24.2%. For Indonesian and Philippine firms, however, we observe no significant association between relationship lending and credit availability.

When we formally test the Berger and Udell (2005) hypothesis that the value of relationship is weaker for transparent firms, we find that in Indonesia, even with weak lending relationships, firms that undergo mandatory audits face less credit constraint than do firms that voluntarily provide an audit. This finding suggests that for Indonesian firms with weak lending relationships, banks replace relationship-lending technology with a financial-statement lending technology. We also find that the value of a mandatory audit is stronger for firms with weaker lending relationships than for firms with stronger relationships. In contrast, in Korea and the Philippines, mandatory audits prove not to be significant. The data do not allow us to explore the Berger and Udell hypothesis for Thai firms.

Section 1 explains the rationale for our empirical model, and Section 2 outlines the model's design. Section 3 describes the sample and data, Section 4 presents the empirical results, and Section 5 extends the study by examining the effect of accounting disclosure on credit availability in the presence of relationship lending. Section 6 describes the robustness testing of our results, and Section 7 concludes.

1. RATIONALE FOR MODELING ACCESS TO BANK CREDIT

We model the likelihood of having access to bank credit in tough times as a function of lender diversity and collateral pledged to obtain bank loans. The control variables are the firm's size, leverage ratio, profitability, growth opportunities, ownership structure, and the industry in which it is operating.

The Lender Diversity variable is a proxy for relationship strength and measures the number of financial institutions with which a firm does business. We define intensity as highest if a firm uses only one bank. The implicit assumption in this definition is that there is an inverse monotone link between number of bank relationships and intensity of bank--firm relationships. We realize that this assumption can be considered as being strong because recent theory argues that in addition to the number of relationships, the composition of lending also matters for intensity (Minetti 2004, Minetti and Guiso 2004). However, the survey does not provide data on intensity of relationships like duration of lending relationships, or firm-specific market shares of banks.

Collateral is our proxy for the cost of lending. Collateral includes assets such as land and buildings, machinery and equipment, and stocks. Since the survey data focus on the type of asset rather than the ownership of the asset, we cannot identify whether the variable measures "inside collateral" (assets owned by the firm) or "outside collateral" (assets not owned by the firm). (2)

Lender Diversity and Collateral can be endogenously determined along with the credit decision. Therefore, we allow for their endogeneity in our empirical estimation. Lender Diversity is endogenous because when firms choose the number of banks they will do business with, they consider the costs and benefits of having multiple bank relationships. For example, Thakor (1996) argues that in a multiple banking relationship each lender's incentive to screen is reduced, which in turn reduces the probability of the firm's obtaining a loan. Internal monitoring (as signaled by the existence of a board of directors or of independent directors on the board) can substitute for screening by lenders, and in this case having multiple lenders may not lead to a lower probability of obtaining loans. Guiso and Minetti (2004) underscore another dimension of the decision about the number of banks to do business with by showing that firms with more valuable and deployable assets tend to choose multiple banking relationships. Finally, Elsas, Henemann, and Tyrell (2004) argue that the optimal number of bank relationships is determined when firms balance the risk of lender coordination failure against the bargaining power of a pivotal relationship bank.

Collateral is endogenous because it is part of the implicit cost of borrowing (Brick, Kane, and Palia 2003). A firm decides on a collateral pledge at the same time it makes the borrowing decision.

We choose several variables to serve as instruments for the decision on the number of banks to do business with (Lender Diversity) and on the posting of collateral. These instruments are existence of a board of directors, independent directors on the board, trade with foreign countries, government incentives used by the firm, credit obtained from suppliers, and the level of the firm's tangible assets.

Board of Directors and Independent Directors on the Board capture the strength of the firm's internal monitoring. A strong monitoring mechanism can reduce the free-rider problem that lenders face with firms that have multiple bank relationships and can thus allow the firm to have banking relationships with several banks. These two variables can also be associated with Collateral since the board can monitor the firm's business activities, thereby affecting the firm's performance (Hermalin and Weisbach 2003). Overall, the presence of a board, and of one with outside or independent directors, can provide better corporate governance.

Trade with foreign countries is likely to be associated with both Lender Diversity and Collateral. Firms that trade internationally have greater investment opportunities, which can affect the number of lending relationships the firms build. On the one hand, when a firm has high growth opportunities, a bank that has an informational monopoly on the firm can extract greater rents (Petersen and Rajan 1994, Houston and James 1996). On the other hand, the firm can mitigate the lender's informational monopoly by building several lending relationships. Trade is correlated with Collateral because firms with greater investment opportunities may be more complex, so that lenders have greater informational problems. For instance, a firm's prospects depend on the health not only of the domestic market but also of foreign markets. Collateralization can address these complexities and the associated informational problems.

Government Incentives used serve as a cushion and decrease the participating firm's probability of default. Alternatively, only firms in weak or poor condition can qualify to participate in government programs. For these reasons, Government Incentives are related to Collateral. Government incentives used can also affect the number of lending relationships a firm builds in that if government programs provide financial assistance, the firm's need to rely on bank loans may decrease.

Since a supplier is an additional source of credit along with financial institutions, obtaining credit from suppliers can affect Lender Diversity. Supplier Credit can measure the mitigation of lender coordination failure (Elsas, Heinemann, and Tyre1 2004), lender's negative liquidity shock (Detragiache, Garella, and Guiso 2000), and alleviation of the hold-up problem (Rajan 1992).

Finally, the tangible asset ratio (the sum of machinery, equipment, and buildings to total assets) can be considered a proxy for the liquidity cost (Guiso and Minetti 2004) and the value of collateral.

2. MODEL DESIGN

The probability of a firm experiencing credit constraint as a function of relationship strength (Lender Diversity), Collateral, and other control variables is

Credit Constraint * = [[beta].sub.0] +[[beta].sup.*.sub.1] Lender Diversity + [[beta].sup.*.sub.2] Collateral + [X.sub.1][[beta].sub.3] + [mu] (1)

Credit Constraint = 1 [Credit Constraint * > 0] (2)

Credit Constraint = [Credit Constraint* [less than or equal to 0], (3)

where Credit Constraint * is the unobserved latent variable measuring the degree of deterioration in credit availability since the onset of the crisis. In equation (1), [[beta].sub.0] is the constant, [[beta].sub.1] and [[beta].sub.2] are coefficients, and [[beta].sub.3] is a vector of coefficients; and [mu] is the error term. [X.sub.1] is a set of exogenous variables that include firm size, debt capacity, profitability, growth opportunities, ownership structure, and the industry binary variables.

The standard two-stage least squares method is not appropriate for dealing with the potential endogeneity of Lender Diversity and Collateral since the dependent variable is binary. Instead, we adopt the two-step estimation method developed in Rivers and Vuong (1988).

To apply the Rivers and Vuong method, we model Lender Diversity and Collateral in equation (1) as reduced-form equations

Lender Diversity = [[beta].sub.10] + [X.sub.1][[beta].sub.11] + [X.sub.2][[beta].sub.12] + [v.sub.1] (4)

Collateral = [[beta].sub.20] + [X.sub.1][[beta].sub.21] + [X.sub.2] [[beta].sub.22]+ [v.sub.2], (5)

where [X.sub.1] and [X.sub.2] are a set of exogenous variables, and [v.sub.1] and [v.sub.2] are the error terms. We use the exogenous variables [X.sub.2] only in equations (4) and (5), where they serve as instruments for the...

NOTE: All illustrations and photos have been removed from this article.



More articles from Journal of Money, Credit & Banking
Emergence of captive finance companies and risk segmentation in loan m..., February 01, 2008
Money market integration.(Statistical table)(Report), February 01, 2008
Information contagion and bank herding.(Report), February 01, 2008
Can interest-bearing money circulate? A small-denomination Arkansan ex..., February 01, 2008
The effect of bank credit on asset prices: evidence from the Japanese ..., February 01, 2008

Looking for additional articles?
Search our database of over 3 million articles.

Looking for more in-depth information on this industry?
Search our complete database of Industry & Market reports by text, subject, publication name or publication date.

About Goliath
Whether you're looking for sales prospects, competitive information, company analysis or best practices in managing your organization, Goliath can help you meet your business needs.

Our extensive business information databases empower business professionals with both the breadth and depth of credible, authoritative information they need to support their business goals. Whether it be strategic planning, sales prospecting, company research or defining management best practices - Goliath is your leading source for accurate information.