Bank-Firm Relationships and Contagious Banking Crises

Abstract

This paper argues that in an open economy a banking system with close bank-firm relationships may be easily subject to contagious banking crises because it is difficult to distinguish between "crony capitalism" and "good" main bank relationships. I show that, if international investors cannot distinguish the bank type, the distinction between crony capitalism and good main bank relationships becomes very fuzzy. In particular, the model can explain sequences of bank defaults within a country, even if the insolvent banks are very few ex ante, as well as sequences of banking crises among countries that are equally rated by international investors, but indeed differ in the ex ante solvency of their banking system.

In developing countries, external finance to firms is provided mainly by banks. Furthermore, bank-firm relationships are very close and borrowers usually rely on a single bank.¹ This situation is usually condemned for causing problems of crony capitalism and connected lending because banks would have incentives to fund negative-net-present-value projects (Dewatripont and Maskin 1995). Most importantly, it is considered the main factor behind the processes of accumulation of bad loans and, consequently, of banking crises (Diaz-Alejandro, 1985, Krugman, 1998, and Corsetti, Pesenti, and Roubini, 1999a and 1999b).

However, the corporate finance literature has emphasized a number of desirable features of close bank-firm relationships, often referred to as main bank relationships. Notably, these are considered to allow financiers to take a longer view on investments and to reduce financial constraints for firms in temporary difficulties (Rajan, 1992, Hoshi, Kashyap, and Scharfstein, 1991). For this reason, main bank relationships have been proposed as a model for the development of the financial system in developing and transition economies (Aoki and Patrick 1994), where sources of external finance other than bank loans are limited.

This paper shows that in a small open economy, even if banks perform desirable functions and do not lend to negative-net-present-value projects, a banking system based on close bank-firm relationships may be easily subject to phenomena of contagion which undermine its stability.

The instability of financial systems where firms borrow predominantly from a main bank arises from the difficulty of distinguishing between insolvent and illiquid banks. On one hand, banks may renew loans to insolvent projects and, as a consequence, accumulate losses. On the other hand, banks may be only temporarily illiquid because they lend to projects in temporary difficulties, but that are solvent in the long run. In this case, close bank-firm relationships indeed allow financiers to take a longer view on investments (Rajan 1992) and must be assessed positively. However, if international investors cannot observe the quality of the investment opportunities funded by banks, the distinction between illiquid and insolvent banks becomes very fuzzy. Contagion may thus arise either between illiquid and insolvent banks within a country or across countries, because international investors, unable to distinguish across banks, demand the same interest rate on deposits whatever the bank type.

In this paper, I show that incomplete information provokes a uniformly low equilibrium interest rate in the aftermath of the liberalization of capital flows, when the lending boom starts, and an apparently sudden increase in the cost of funds, which causes the defaults of the insolvent banks, at a date to be endogenously determined in the model. Even if illiquid banks do not default immediately, revealing that they are not insolvent, the temporary increase in the interest rate burden may drive them to insolvency and cause them to default after a few periods.

These problems would not arise if firms had many financiers, as they do in advanced economies, because insolvent projects would not be financed (see, for instance, Hart, 1995 and Dewatripont and Maskin, 1995). In this case, even if investors do not observe a bank type, they know that banks are not renewing loans to insolvent projects. As a consequence, there are no sudden increases in the interest rate and the financial system is more stable.

Moreover, banking crises are expected to be more frequent after the liberalization of capital inflows and the deregulation of the banking system, as...