A Theory of Conflicts of Interest in Banking Relationships

Abstract

This paper highlights a dark side of banking relationships by elucidating the conditions under which a pre-existing relationship between a lending bank and a borrower can be detrimental to positive valuation effects of loan announcements. The effect of a pre-existing relationship is more likely to be negative when the pre-existing loans are large and firms' screening costs are low. A theoretical model shows that loan announcement's positive effect on borrowers' value due to the standard information advantage can be more than offset by the bank's conflict of interest when the bank's asset quality reputation is poor, i.e. when the probability of the bank holding a bad loan is large.

Keywords:

• Conflicts of interest
• banking relationships
• soft-budget constraint
• sequential equilibrium

Acknowledgements

The author is grateful to Charles Calomiris, Mark Carlson, Richard Ericson, Jaehoon Hahn, Eslyn Jean-Baptiste, and seminar participants at Columbia University, the Federal Reserve Bank of New York, KAIST Graduate School of Management and KDI School for their helpful comments and suggestions.

Notes

¹Fama (1985) argues that information produced over time through continuing lending activities is derived from informal provision of insider information, subjective judgments on management abilities, accumulation of repayment history, and other private information associated with monitoring borrowing firms.

²The model may be extended, however, to a more general setting where the loan announcement effect on borrowers' values depends on whether the borrowers have loan balances before the loan announcement.

³The model's setting is applicable to both mergers and acquisitions (M&A) and purchase and assumption (P&A) arrangements as a way of restructuring. The difference seems to be, however, that under the P&A arrangement, no personnel are transferred in general from the failed banks to the acquiring banks, whereas under the M&A an acquiring bank typically retains most of an acquired bank's loan officers. This characteristic has important implications on the transfer of private information about borrowing firms and on the loan officers' close customer relationships.

⁴Relationship banking adds value to borrowing firms by Pareto-improving exchange of information between banks and borrowers, flexibility and discretion in contracts and value-enhancing intertemporal transfers in loan pricing.

⁵There is literature that studies the effect of changes in bank quality on client firms' stock returns. Its main finding is that a decrease in a bank's quality or an increase in the probability of a bank's insolvency reduces its client firms' market value due to possible destruction of the relationship-specific advantage and its associated information. See Slovin et al. (1993) for the US; Bae et al. (2002) for Korea; and Yamori & Murakami (1999) for Japan.

⁶It is common for financial institutions to hold equity in other companies. Korean banks, for instance, held 147 trillion won worth of equity claims in other firms (7.5% of the banks' marketable securities) as of the end of 1998 when there were bank closures and transfers of their loans to other banks. Banking Act of 1998, Article 37 forbad financial institutions to hold more than 15% of any other company's issued voting stocks except when it is necessary to promote corporate restructuring.

⁷The term ‘debt’ is used when indicating a pre-existing loan, D, to differentiate it from a transferred loan, L.

⁸A Type H bank holds neither prior debts nor transferred loans of bad firms. Therefore, by construction, only good firms' loans should be transferred to a Type H bank.

⁹The common terms—the profits from recovering the transferred loans and the good firms' pre-existing debt—are suppressed in expected profit equations (1) and (3). No matter which type (good or bad) of firms' loans the bank decides to renew, it gives the same expected revenue because the good firms' loans and pre-existing debt are recovered anyway, and the loan's interest rate is (1−δ)/δ.

¹⁰Joo & Pruitt (2006) document that changes in bond rating during Korea's financial crisis magnified stock prices' effects. The bond rating changes resulted in significantly stronger changes in stock prices than did ratings changes of identical magnitude announced during the non-crisis period.

¹¹By construction, both good and bad loans of Type N firms' would be transferred to Type H banks, as opposed to Type P firms, to which only good firms' loans are transferred.

¹²A set of firms whose transferred loans mature at t=2 is different from a set of firms whose transferred loans mature at t=1. Hence the bank needs to determine again at t=2 whether to screen the firms before making its lending decisions.

¹³This is consistent with previous studies such as Leland & Pyle (1977) and Berlin et al. (1996). They support the bank's holding of equity in the borrowing firms because it enhances the bank's credibility in certifying the firms' values.