«Sangkyun Park and Stavros Peristiani* October 2006 ABSTRACT: In moral hazard models, bank shareholders have incentives to transfer wealth from the ...»
Are bank shareholders enemies of regulators or a potential source of
Sangkyun Park and Stavros Peristiani*
ABSTRACT: In moral hazard models, bank shareholders have incentives to transfer wealth from
the deposit insurer-- that is, maximize put option value-- by pursuing riskier strategies. For safe
banks with large charter value, however, the risk-taking incentive is outweighed by the
possibility of losing charter value. Focusing on the relationship between Tobin’s q and an ex ante measure of the failure probability, this paper develops a semi-parametric model for estimating the critical level of bank risk at which put option value starts outweighing charter value. From these estimates, we infer the prevalence of moral hazard. Examining publicly held bank holding companies (BHC) during the tumultuous 1986-92 period, we find that shareholders’ risk-taking incentive were confined to a small fraction of highly risky institutions. Furthermore, our analysis shows that the inflection point at which banks begin to tilt in favor of moral hazard increased substantially in 1993-2005. These findings are encouraging to bank regulators and legislators because they indicate that higher capital rules and more rigorous supervision introduced by several legislative initiatives in the 1990s have helped squeeze a lot of the moral hazard incentives out of the banking system.
JEL classifications: G21, G28.
Keywords: Moral hazard; Tobin’s q ratio; deposit insurance; bank risk; implied default probability.
SANGKYUN PARK is an Economist at the Office of Management and Budget. STAVROS PERISTIANI is a Research Officer in the Research and Market Analysis Group of the Federal Reserve Bank of New York.
*Corresponding author, Main 3, Federal Reserve Bank of New York, 33 Liberty Street, New York, NY 10045. E-mail: firstname.lastname@example.org. We wish to thank Gijoon Hong for excellent research assistance. We are also grateful to Astrid Dick for her insightful comments and suggestions. The views expressed in this paper are those of the authors and do not necessarily represent the views of the Federal Reserve Bank of New York, the Federal Reserve System or the Office of Management and Budget.
Are bank shareholders enemies of regulators or a potential source of market discipline?
1. Introduction Many regulators and academic researchers have emphasized market discipline as a means to improve the safety and soundness of the banking system. Perhaps this cannot be more pertinent than in today’s complex financial landscape. Because of ever increasing complexity of the banking business, it is difficult to effectively regulate banks solely based on prescribed rules.
The importance of market discipline is underscored by several banking and financial crises experienced in the last two decades across several emerging markets as well as more industrialized countries worldwide. In many instances, the inability of bank regulators and market forces to effectively discipline financial institutions was deemed as the missing ingredient for ensuring financial stability.
Not surprisingly, there is renewed interest among policy makers in enacting changes that would encourage more market disclosure and transparency. Directed by the Gramm-LeachBliley Act of 1999, the U.S. Treasury and the Federal Reserve Board have explored the possibility of using mandatory subordinated debt as a catalyst to strengthening market discipline.
More recently, the new capital adequacy proposals of the Basel Committee for Banking Supervision (Basel II) consider market discipline, along with capital requirements and supervision, as one of the three pillars of support for the banking system (Basel Committee on Banking Supervision (2003)).
In a market discipline framework, debtholders, depositors, and shareholders can apply pressure on banks, raising the funding premium on debt, deposits, and equity. The direct and indirect effects of market discipline have been extensively studied in the academic literature (see survey article by Flannery (1998)). The interest in market discipline has been largely stimulated by the moral hazard literature describing the conflict between shareholders and debtholders (or the deposit insurer in the case of insured banks). In a moral hazard framework, bank management acts in the interest of shareholders that have voting power. Merton (1977) shows that deposit insurance gives insured banks a put option, that is, the right to sell the bank’s assets at the face value of its liabilities. When the deposit insurance premium is fixed or does not fully reflect a bank’s risk, the value of the put option increases with the bank’s risk, typically, represented by a larger return variance of the asset portfolio and a lower capital ratio. If the shareholders of a bank are interested mainly in the put option value, managers may accommodate them by increasing the bank’s risk. In this case, shareholders are enemies of bank regulators, and the burden of market discipline falls on the shoulders of debtholders.
Several studies of moral hazard have shown that bank shareholders are also responsive to the bank’s charter value or intangible capital (e.g., Keeley (1990), Ritchken et al. (1993), and Park (1997)). In the event of failure, shareholders have to forfeit charter value. Their incentive to preserve charter value should therefore outweigh their desire to increase the put option value when the bank’s risk is low or moderate, while the opposite is true at high levels of risk.
Consequently, bank shareholders can be allies of regulators and a source of market discipline when a bank is reasonably safe.
The moral hazard theory raises an intriguing empirical question: At what level of risk do shareholders turn into enemies of regulators? Several empirical studies have attempted to shed some light on the relative importance of put option value and charter value. Most notably, Keeley’s (1990) study attributes the sharp rise in failure among banks and thrift institutions to the gradual deterioration of bank charter value. Keeley argues that during the 1980s the rapid deregulation of the banking and thrift sectors, coupled with intense competition from nonbank institutions, resulted in the deterioration of the charter value of banks and thrifts. As a result of lower charter value, shareholders were compelled to switch to riskier strategies, which in turn brought about the increased incidences of bank failure in this period. To measure charter value, Keeley examines the relationship between regulatory changes in bank entry and the market power of bank. As an estimate of market power, Keely uses Tobin’s q, defined as the market value of assets divided by the book value of assets.
Using a similar Tobin’s q measure, Demsetz, Saidenberg and Strahan (1996) explore more directly the relationship between charter value and bank risk. The authors find a negative relationship between charter value and different stock market measures of risk. Consistent with theoretical predictions, they also discover that banks with higher charter value are motivated to take safer strategies and tend to hold more capital. Brewer and Mondschean (1994) compare the behavior of low- and high-capital savings and loan associations (S&Ls). Their analysis finds that poorly capitalized S&Ls exhibit a positive relationship between stock market returns and junk bond holdings. This result indicates that the market may be looking favorably at high-risk strategies for firms whose option value is likely to be higher than charter value.
This paper differs from the previous empirical studies in that we focus on the tradeoff between put option value and charter value in relation to the level of bank risk. In particular, our analysis aims to highlight the bipolar behavior of bank shareholders -- on one hand, as allies of regulators, protecting their stake in a low-option value institution by penalizing risky strategies, and on the other hand, as enemies of regulators, condoning more risk-taking strategies for institutions whose option value outweighs charter value.
In this paper, we gauge bank risk using two different but related methods. First, we estimate a bank’s probability of failure from historical bank failure records. To verify the robustness of empirical findings, we also look at the likelihood of default using a contingent claims approach proposed by Merton (1974). Condensing the measure of risk into a single dimension greatly simplifies our analysis. As suggested by Keeley (1990), a bank’s option value and charter value can be jointly inferred from Tobin’s q ratio (henceforth referred to for brevity as the q ratio) measuring the market value of assets relative to the book value of assets. A negative relationship between the failure probability and the q ratio would mean that the expected loss of charter value outweighs the increase in the option value, and a positive relationship would indicate the opposite.
The next section illustrates that the theoretical relationship between bank risk and the q ratio is nonlinear and convex, reflecting the changing relative importance of charter value and put option value. Moral hazard theory, however, establishes a convex functional relationship between bank risk and market-to-book value without providing an explicit well-defined parametric structure for empirical testing. We resolve this challenge by using a semi-parametric spline estimation technique to estimate the link between bank risk and the q ratio.
The principal findings of our analysis are interesting in several respects. When we estimate the relationship between the q ratio and risk for the period 1986-1992, we discover that the threshold at which the marginal contribution of the option value starts outweighing the expected loss of charter value is around a 17-percent annual probability of failure (or 22-percent for the implied probability of default). The analysis also reveals that the q ratio for institutions with a strong core deposit base decreases at a higher level of the failure probability. This result is consistent with the theoretical prediction that banks with more valuable charters are more averse to risk.
This estimated point of transition is at a fairly high level with only a very small fraction of bank holding companies (roughly 3 percent) attaining failure scores greater than this threshold level. Among publicly held BHCs, therefore, moral hazard appears to have been confined to a small set of highly risky ones even during this sample period when moral hazard was believed to be prevalent.
The banking crisis led to tighter regulations, including higher capital requirements, risksensitive deposit insurance premiums, and prompt corrective action. In addition, many barriers to competition, such as restrictions on intrastate branching, interstate banking, and inter-industry financial affiliations, were removed in the 1990s. These developments may have influenced option value and charter value and hence the relationship between the q ratio and risk. Allen and Rai (1996), for example, show that more rigorous regulation and supervision encourage banks to protect charter value.
To explore this possibility, we estimate the relationship between the q ratio and risk (measured now by the implied default probability) for the period 1993-2005. We find that the implied probability of default inflection point at which banks begin to tilt in favor of moral hazard increased from 0.22 in 1986-1992 to 0.50 in 1993-2005. These findings are encouraging for bank regulators because they demonstrate that tighter regulation and supervision have helped squeeze a lot of the moral hazard incentives out of the banking system.
The rest of the paper is organized as follows. The next section describes the hypothesis to be tested based on the findings of the moral hazard literature. In Section 3, we present a semiparametric model for estimating the nonlinear relationship between bank risk and the q ratio.
Section 4 briefly describes the two methods used in measuring the riskiness of banks. Section 5 presents our empirical findings. The final section summarizes the paper’s findings.
2. Option value and the q ratio
2.1. The effects of option value and charter value on shareholders’ wealth The moral hazard literature modeling the bank shareholders’ incentive to take advantage of deposit insurance (e.g., Keeley, 1990, Ritchken et al., 1993; and Park, 1997) derives an implied U-shaped relationship between shareholders’ wealth and risk. When deposit insurance premiums do not reflect bank risk fully, bank shareholders with limited liability can transfer wealth from deposit insurance to themselves by pursuing riskier strategies, namely, lower capital ratios and more volatile asset returns. Gains from this put option value, however, are mitigated by the possibility of losing charter value that cannot be recovered in liquidation. The option value continuously increases with risk, while the charter value is fairly stable in the short run. At a low level of risk, therefore, shareholders’ wealth decreases with risk because the value of the option is not large enough to offset the expected loss of charter value. If risk increases sufficiently, however, the value of the option dominates the charter value, and shareholders’ wealth increases with risk.
To illustrate the link between this theoretical prediction and our empirical study, we use the analytical framework of Park (1997), which is relatively simple and intuitive. The results of other studies can be applied in a fundamentally similar way. Park (1997) models the maximization of bank shareholders’ wealth under the following simplifying assumptions: All agents are risk neutral; all projects have zero net present value; and the cost of debt does not increase with risk thanks to deposit insurance charging fixed premiums. In period 1, bank managers maximize the second-period wealth of shareholders by choosing the capital ratio (ratio of equity to assets) and the portfolio share of the risky asset (two assets in the model: a risky asset producing a random return and a risk-free asset yielding a certain return). Expected shareholders’ wealth in period 2 is defined by