Constraints and Risks of Relationship Banking

Banks face several constraints as they attempt to exercise corporate governance in an efficient way. First, given their own poor corporate governance, they may have weak incentive for genuine relationship banking or may use this practice with perverse or distorted incentives. Second, a relationship bank may face a potentially serious conflict of interest when it intervenes in the management of a financially distressed firm. Third, with deregulation and increased competition in the financial markets, the incentive for banks to invest in relationship-specific information production is weakening. Finally, relationship banking may be jeopardized in a financial crisis, when banks cannot fulfill their implicit obligation to provide rescue and restructuring services to distressed firms.

4.1. Corporate Governance in Banks

The realization of whatever benefits can be expected from relationship banking depends on sound corporate governance in the banks. Otherwise, they will not have much incentive to invest in relationships, or relationship banking practices might simply serve as a convenient tool for distorting the allocation of bank resources. If banks are directed (by the government, bureaucrats or politicians) to lend to specific sectors and firms, the room for relationship-based banking is limited. The same is true if banks are owned and controlled by non-financial business groups. The controlling owners will have a strong incentive to use bank loans and other services to benefit their group or family rather than all the shareholders of the bank. There is little incentive in this case to have a close banking relationship with this bank for firms that are not subsidiaries of the group or otherwise have close business linkages.

Some of the external mechanisms of corporate governance are weaker in banking than in non-financial industries. For instance, the market for corporate control for banks is restricted by the regulations on the qualification of potential acquirers or the prior approval of the bank regulator. The discipline of the product market is also weak due to the oligopolistic markets for some banking services. The typically diffuse ownership structure and the consequent free-rider problem also make it unrealistic to expect effective monitoring by shareholders. Meanwhile, the deposit insurance system often encourages moral hazard behavior by banks such as excessive risk-taking. In return for the provision of deposit insurance, the financial regulatory authorities are justified in monitoring and disciplining bank management. Bureaucracy and political interference, however, makes the regulator far from an efficient and effective monitor.

Analyzing 234 cases of corporate control changes among bank holding companies in the U.S. over the period 1987-92, Prowse (1995) finds that hostile takeovers or friendly mergers did not play an important role as a means of disciplining bank managers, and that the internal control device of board-induced management turnover was not used as frequently as in manufacturing, particularly for manager-entrenched banks. Regulators providing last-resort control mechanisms were found to exercise their regulatory intervention mainly over banks with markedly low ownership concentration. Other studies show that banks with entrenched management tend to most actively engage in acquisition programs that are likely to increase the perquisites available to management, simply because of their bigger size (Allen and Cebenoyan, 1991). In addition, banks with managements that are relatively free from outside shareholder control tend to make the riskiest and most unprofitable investments (Gorton and Rosen, 1992).

Spong, Sullivan, and De Young (1995) identify a number of characteristics (financial, ownership, and management) of the most efficient and least efficient banks in the Tenth Federal Reserve District of the U.S. They find that efficient banks are associated with:

• Board of directors – more frequent meetings; higher director fees; and members with a higher average net worth, a greater equity share, and better attendance rate
• A strong ownership group or management with a vested interest in the bank (rather than diffuse ownership with hired managing officers)
• Higher compensation for managing officers.

The Japanese banking problems since the early 1990s are also often attributed to a corporate governance vacuum in banks where managers enjoyed wide latitude (Hanazaki and Horiuchi, 1998).

• Banks are generally more diffusely held than other exchange-listed firms. Small shareholders suffer from the free-rider problem, which is further worsened by the comprehensive safety net – in the form of the rescue of distressed financial institutions – provided by the government.
• Competition – another powerful means of discipline – has not been effective due to the delayed deregulation in the financial markets with the restriction of new entries and the compartmentalization of the sector.
• The supervision by regulatory authorities over banks has been also problematic. The practice of post-retirement service for high-level government officials on the boards of private firms (amakudari) has been common for banks, undermining the effectiveness of monitoring by the regulatory authorities (Horiuchi and Shimizu, 1998).

Hanazaki and Horiuchi (1998) analyze the performance of 125 regional banks (capital/asset ratio and bad loan ratio) for the period of 1985-89, and find that firms with amakudari officers from the MOF had significantly worse performance. This evidence is contrary to the argument of Aoki, Patrick, and Sheard (1994) that the amakudari system disciplines financial authorities to effectively monitor bank management. Their argument is that regulatory authorities have strong incentives to rigorously monitor banks in order to secure good jobs for their officials after retirement. Rixtel and Hassink (1998) is another empirical study on the efficiency of the practice of amakudari. They find that banks accepting amakudari officers tend to have poor profitability, deteriorating solvency or declining growth in lending to construction, real estate, and non-bank financial industries. They also found that MOF/BOJ amakudari appointments led to increased lending to these industries. They suggest that recapturing the lost market share in these industries might have been the strongest motivation for the practice, casting doubt on the credibility of the hypothesis that it has been used as an instrument of prudential policy.

4.2. Conflicts of Interests in Banks’ Governance Role

Banks play a role in corporate governance primarily through their representation on corporate boards of directors. Commercial bankers used to be attractive as outside directors of corporations for several reasons. They provided advantages in information and facilitated stable and long-term relations with the banks; the banks tended to be in the central position of the network of interlocking directors even in the market-based system of the U.S. as well as in the bank-centered system. However, the situation has changed significantly since the 1980s. Large reputable firms now rely little on bank borrowings; with keener competition in banking, they may be reluctant to give a privileged position to a particular bank; and bankers as a source of information have become less important with the progress in IC technologies and the weakening role of banks’ interlocking networks. The percentage of large firms with a commercial banker on the board is: Germany 75.0% (100 largest firms in 1974), Japan 52.9% (761 listed firms, 1992), U.S. 31.6% (Forbes 500 firms in 1992; Davis and Mizruchi, 1999).

A probably more important reason why banks are not active in their corporate governance role in the U.S. is the potentially substantial costs associated with this role in bankruptcy codes and procedures. Having a banker on the board of a firm entails a conflict of interest between the fiduciary duty to the firm’s owners and to the bank employer. If a bank, with its representative on a corporate board, effectively controls corporate decisions, it may try to protect its interests at the expense of those of other stakeholders. In the event that the bank acts "inequitably" prior to a borrower’s bankruptcy, it may lose senior creditor status and face liability for the losses to other stakeholders (equitable subordination and lender liability).²⁰

In order to avoid this complication, banks opt to maintain an "arm's length" relation with their corporate clients, and are very cautious about their officers being represented on corporate boards. To the extent that board representation helps banks' corporate governance role, this rule in the bankruptcy codes constrains effective monitoring and control activities by banks. Thus, bankers tend to be on the boards of firms where they are not likely to face the cost of lender liability: large, stable, and better performing firms that are unlikely to fall into financial distress. The bank represented is often not the biggest bank lender for the firm. Banks' representation on corporate boards might initially increase with the increasing volatility and deterioration of corporate performance, but is likely to decrease at higher levels of risk.²¹

4.3. Deregulation and Competition in the Financial Markets

Since the mid-1970s, Japanese firms have had more options and flexibility in their financing, including increased dependence on bond issues, and this has weakened their reliance on main banks (Aoki, 1994). In the 1980s, differences in capitalization between MB client firms and other firms largely disappeared with deregulation related to the issuance of unsecured foreign bonds as well as local bonds (Weinstein and Yafeh, 1998). Deregulation and keener competition in the financial market are certainly threats to relationship banking. At the same time, another obvious trend in financial regulation is a move toward universal banking, which may provide banks with a new motivation to keep investing in relationship-specific information production and monitoring.

Deregulation, competition and relationship banking

Increased competition in financial markets may reduce the value of relationships. Keen competition may prevent banks from reaping the rewards of investments in firm-specific information production or of helping client firms at an early stage in anticipation of rents in later years (Mayer, 1988; and Rajan, 1992). This uncertainty may reduce banks' incentives to invest in firm-specific information production and relationship management. Competition, however, needs not be harmful to relational financing once the practice is more or less established. Also, banks’ incentives for relation-specific investments are not necessarily weakened if competition comes largely from the capital market where relationships are not important.

First, banks require market power (a smaller number of banks and other closesubstituting financial institutions) to be able to extract monopoly rents, most likely through the intertemporal smoothing of interest rates. Competition threatens relationships as the possibility of losing corporate clients to other financial institutions reduces the value of relevant information, discouraging new investments in relationships. Thus, in order to foster the emergence of relationship banking, it may be necessary to restrict competition.²² However, as noted by Aoki and Dinç (1997), once relationship banking has become well established, the restriction may safely be removed. Because of an accumulation of information that cannot be easily transferred to other market participants, the main bank is likely to maintain its informational advantage, and its incentives for maintaining relationship banking would be continued.

Also, competition from bond markets increases the effectiveness of relationship banking, whose major role is to reduce information asymmetries. If increased competition is mainly in the area of transaction lending (such as corporate bond issues) rather than relationship lending, banks are encouraged to move to less competitive businesses and become more relationship-oriented. Banks with strong reputations as relationship lenders are least affected by competition (Boot and Thakor, 1998). For small and medium-sized firms and young firms in non-high-tech industries, a reputation mechanism that sustains a bank’s rescue commitment may not be affected negatively by the increasing competition from bond markets (Aoki and Dinc, 1997). The decline of traditional commercial banking with the trend of securitization and increased non-bank competition will have little impact on SMEs (Berger and Udell, 1995).

Universal banking: a new avenue for relationship banking?

If commercial banks are permitted to undertake investment banking services, they can continue to utilize their private information about corporate clients even though the corporate financing pattern changes from bank loans to direct financing. In a universal banking system as adopted in Germany and other continental European countries, banks offer a full range of commercial and investment banking services including underwriting, holding and trading on their own accounts, brokerage, and securities custody business. They also sell insurance, mortgages, and investment funds usually through affiliates. Banks are not necessarily legally bound in deciding the scope of their businesses: British commercial banks have always been permitted to engage in universal and relationship banking – but have apparently often refrained (Fohlin, 1999).²³

A major concern of the universal banking system is the potential conflict of interest, though "firewalls" can prevent this problem to some extent. In order to attract investors with concerns about conflicts of interest, securities underwritten by commercial banks may be traded at a discount (higher yields). On the other hand, commercial banks are supposed to have private information accumulated through their relationship banking that is not available to general investors in the capital market. If investors believe that commercial banks have greater ability to certify the values of new issues, they might actually trade securities underwritten by such banks at a premium (lower yields). Empirical evidence seems to indicate that the certification merit more than compensates for the concern over conflicts of interest.

The public appears to have rationally accounted for the possibilities of conflicts of interest, which constrain commercial banks to underwrite high-quality securities. Evidence in the pre-Glass-Steagall period (when there were no firewalls) shows that securities underwritten by commercial banks had a better default record in the long run and were traded at better prices than those underwritten by investment houses, despite the potential conflicts of interest (Ang and Richardson, 1994; Kroszner and Rajan, 1994; Puri, 1994 and 1996).

The separation between commercial and investment banking has been relaxed in the 1990s in both the U.S. and Japan. Evidence for the period 1993–95 in the U.S. shows that commercial banks, with their private information and certification role as monitors, bring small issuers to the market and that bank underwriting reduces yield spreads particularly for lower-credit issues, in a way that is consistent with the view that bank association is valuable for such issuers due to the bank’s dominant certification effect (Gande, Puri, Saunders, and Walter, 1997).²⁴ A similar study was undertaken by Hamao and Hoshi (2000), who compare the characteristics of straight corporate bonds underwritten during 1994-96. They find that for issues underwritten by bank subsidiaries, the issue size is substantially smaller and the certification effects dominate (though not statistically significantly) the conflict of interest effects.²⁵

A close relationship with a single main bank entails a risk along with the advantages. Client firms fall into trouble if the bank becomes unable to extend credit. In many cases, other banks and investors do not have full information on why firms cannot get credit from their MBs and might think they are not creditworthy. Thus, it is not easy for such firms to turn to other banks or the capital market. Evidence shows that the values of borrowing firms depend on the health of their banks (Yamori and Murakami, 1999; Slovin, Sushka and Polonchek, 1993). When the whole banking sector is in distress, high bank dependence is likely to be costly, as it makes the firm more vulnerable to shocks affecting the banking sector. If a MB is in a worse situation in this period of turmoil, its client firms may face particular difficulties in obtaining credit.

Spiegel and Yamori (2001) estimate the sensitivity of equity values of an equallyweighted portfolio of MB client firms to changes in the equity values of their MB. Introducing a structural break after the turbulence in the Japanese financial markets (the last quarter of 1997), they find that MB sensitivity is consistently higher after the break. This shows that Japanese firms are still sensitive to their MBs during episodes of financial turbulence. Kang and Stulz (2000) find that firms with high dependence on bank financing (higher share of banks in total debt), compared to those with a mere MB relationship, experienced lower stock-price returns and lower level of investment in the first three years of the 1990s, a period of credit contraction in Japan.

Gibson (1997) investigates whether the financial health of the MB (using credit ratings as a proxy) affects the investment behavior of the firm for Japanese during 1994- 95. He finds no significant effect, which might be due to the more accommodative lending attitude of financial institutions during the sample period. However, among bankdependent firms, investment was over 50% lower at firms with the lowest-rated MB. These results contrast a similar study by Gibson (1995), who using data for 1991-92, finds a small effect of poor bank health on corporate investment and no difference between bankdependent and non-bank-dependent firms.

It has been noted that the various implicit rules related to MBs have failed since the financial crisis. Recently, for instance, MB representation on corporate boards has decreased in spite of the sharp deterioration of corporate performance. This indicates that the insurance function of MBs is weakening. As many banks themselves have been in distress since 1995, they could not prevent their non-financial subsidiaries from going bankrupt either (Economic Planning Agency, 2000).

Restructuring of banks in trouble

Insolvent banks should be restructured in a timely manner in order to restore public confidence in the banking sector. However, bank bankruptcy results in the loss of the relation-specific investment for reputation building and information gathering. Investment can be saved if troubled banks are merged with other banks. In situations where many firms no longer rely on bank credit, investment can also be saved if banks are allowed to underwrite securities. Bank’s stakes in the financing needs of their client firms continue even with the new competition with the securities markets (Aoki and Dinç, 1997). If a bank generates substantial non-transferable information through relationship banking, this information capital and the relationship are largely lost if this bank fails. The loss might be much larger than the book value of the bank (Petersen and Rajan, 1994; Berger and Udell, 1995).

The views expressed in this paper are the views of the author/s and do not necessarily reflect the views or policies of the Asian Development Bank Institute nor the Asian Development Bank. Names of countries or economies mentioned are chosen by the author/s, in the exercise of his/her/their academic freedom, and the Institute is in no way responsible for such usage.

Post a Comment

We welcome your feedback on this publication. Post a comment. ADBI is not obliged to acknowledge or publish comments and may abridge or edit them before web posting.

Comment(s)

There are [0] comment(s) for this entry. Post a comment.