A Theoretical and Empirical Assessment

of Bank Risk-Shifting Behavior

James R. Barth, Mark Bertus, Jiang Hai and Triphon Phumiwasana*

All China Economics International Conference

APEC Study Centre, Faculty of Business, City University of Hong Kong

December 18 - 20, 2006

* James R. Barth is
Lowder Eminent Scholar in Finance at Auburn University, and Senior Fellow at the Milken Institute, Mark Bertus is Assistant Professor of Finance at Auburn University, Jiang Hai is Professor in Finance and Deputy Department Head of Finance at Jinan University, and Triphon Phumiwasana is an Economist at the
Milken Institute.

     摘要：
        Banks are widely acknowledged to be important players when it comes to mobilizing savings and then channeling those funds to productive investment projects. Unlike many other financial intermediaries, moreover, they also facilitate payments by providing liquidity services to individuals and firms. In the process of providing these and other services that contribute to economic growth and development, banks take on various types of risks with the expectation that the return they receive will compensate for the risks. There is nothing improper with this type of bank behavior per se, but what is improper is when bank owners attempt to shift part of this risk without adequate compensation to other stakeholders like depositors.

The purpose of this paper is twofold.
First, it develops a theory of the risk-taking behavior of a bank that may be termed “prudential” risk taking.
Banks, by their very nature, are in the risk business.
They operate with substantial leverage and the returns on their assets are volatile.
Some banks will choose to operate in such a situation by allocating their assets solely to achieve the highest expected return possible per unit of risk, with both depositors and bank owners (investors) being fully compensated for the risk they bear.
Other banks may choose to operate in the same situation by allocating their assets in a similar but more risky manner by shifting the additional risk unbeknownst and hence without adequate compensation to depositors. In contrast to the prudential banks, these “risky” or “gambling” banks are only able to shift risk to depositors because of asymmetric information.
The theory that is developed examines the risk-taking behavior of banks both under symmetric and asymmetric information to compare and contrast risk-taking behavior that may be deemed
prudential and that which may be deemed gambling due to the shifting of risk to depositors that are unaware of the risk they are being forced to bear.
A bank regulatory authority is introduced into the analysis that can set minimum capital requirements and provide deposit insurance to limit risk taking and to protect depositors from risk shifting.
However, asymmetric information between the bank and the regulator still allows some shifting of risk by a bank but now to the deposit insurer.
The theoretical model is therefore modified to examine the extent to which the regulator can limit the shifting of risk by estimating the probability of risk shifting by a bank through effort extended to detect such behavior.
The implications of various modifications in the theoretical model for the interest rate offered to depositors and, ultimately, social welfare are examined.
The second purpose of the paper is to empirically examine one of the hypotheses that results from the theoretical model.
The hypothesis is that the shifting of risk to depositors, or the depositor insurer, due to asymmetric information results in a lower interest rate being paid on deposits and hence a smaller net interest margin than under symmetric information.
An empirical model is specified to examine the relationship between the net interest margin of a bank and a measure of the degree of market information asymmetry under the hypothesis that there is a negative relationship.
The hypothesis is tested based upon data for 3,115 banks in 98 countries.
We control for bank-specific and country-specific factors to better assure that any relationship found is not due to omission of other relevant factors that may influence bank net interest margins. In addition, several robustness checks are performed to provide additional confidence in our findings.
Overall, the empirical results support the hypothesis that the greater the degree of market information asymmetry the higher the net interest margins of banks.

The remainder of the paper proceeds as follows.
In the next section the theoretical model is presented.
To better understand the implications of risk-shifting behavior by banks, we consider first the case of symmetric information between banks and depositors (or a deposit insurer) and then consider the case of asymmetric information.
In the third section we discuss the empirical model and the data used in the estimation of the model.
The main empirical results and the results of several robustness tests are presented in the fifth section. The conclusions are reported in the sixth and last section.