Economic analysis of banking regulation

Chapter 11  An Economic Analysis of Banking Regulation
In this chapter we use asymmetric information (leading to adverse selection and moral hazard) to study the effects of banking regulations. We study 8 categories of banking regulation.

Government safety net:  Deposit insurance
Restrictions on Asset Holding
Bank Capital Requirements
Bank (Prudential) Supervision: Chartering and Examination
Assessment of Risk Management
Information Disclosure Requirements
Restrictions on Competition

Government safety net:  Deposit insurance (FDIC and中央存款保險公司)

I. Theories of Bank Runs (擠兌)
Random Withdrawal Risk:  A run can occur based on self-fulfilling beliefs, implying that the banking system is inherently unstable.
Information-based Runs:  A run occurs when depositors observe a signal correlated with the value of banking system assets (indicators of recession, decline of net worth of a particular class of bank borrowers). This fragility is intensified by the fear that banks invest in assets which are opaque, illiquid, non-marketable, and containing private information (asymmetric information).

Some Implications on the theory of information-based runs:
1. Since banks operate on a “sequential service constraint” (a first-come, first-served basis), those depositors who monitor their bank more closely would get to the bank earlier when a bad signal arrives and recover their funds in full, while the losses suffered by late-coming depositors are the penalty they must pay for neglecting monitoring. Thus, the possibility of bank runs gives incentives for depositors to monitor.
2. When a bank becomes terminally ill, a bank run can drive the bank into default very quickly and take away the control of the management. Otherwise, the managers who no longer have enough incentives to manage its remaining properties, might sell off the assets in a hurry at lower prices (“fire sale”), bet on take an even greater risk bet, or commit fraudulent conducts. Therefore, it would not be undesirable that the "panicky" depositors race each other for demanding redemption, and this is in fact an essential feature of the bank's efficient liquidation. Thus, runs on banks may be viewed as a form of monitoring and an optimal response (Calomiris and Gorton (1994)).
3. Chari and Jagannathan (1988) argue that uninformed agents learn the state of the bank only by observing the line at the bank. If there happens to be a long line at the bank, more depositors will join in to withdraw. Thus, runs on banks as a means of market discipline may be imperfect.

II. Evidence on the Cause of Bank Runs
1. Evidence from the National Banking System (1863-1913) indicates that periods of bank failures were associated with macroeconomic instability (Bernanke (1983), Gorton (1988), and Kaufman (1994)).
2. Studies by Saunders and Wilson (1996) covering 1929-1933 and Calomiris and Mason (1997) concentrating on June 1932 Chicago banking panic both find strong contagion effects. Moreover, they also stress the significant role of economic shocks and thus they support the information hypothesis of bank runs.
3. Park (1991) argues that a review of bank panics in US history supports the importance of bank-specific information. He finds that the government or banks have stopped widespread bank runs mainly by providing information about banks’ solvency rather than liquidity. He also finds evidence that clearing house loan certificates and suspension can convey bank-specific information to depositors and succeed in tranquilizing bank runs and contain contagious effect, while equalization of reserves, implemented only once in 1873, fails to stop the run because equalization of reserves is an arrangement to ask sound banks to share problems with problem banks. The public knew that some banks are in trouble. When they learned that all banks shared the problems by pooling reserves, depositors conjectured that the entire banking system would become fragile. This conjecture prompted system-wide bank runs.
4. Countries with banking systems offering the same debt contracts did not experience the same occurrence bank panics. Empirical research has identified the institutional structure such as branch bank laws, bank cooperation arrangements, and formal clearing houses, for the probability of panic and the resolution of crisis. Cross-country comparisons lead to the conclusion that bank panics are not inherent in banking contracts, rather, institutional structure matters. Bordo (1985) and Williamson (1989) attribute a large part of US bank failures to the absence of branch banking (institutional factor). Canada experienced no panic after 1830s. Canada allowed nationwide branch banking (number of banks was in 40s in 19th century and falls to 10 in 1929). Suspension of convertibility did not occur. (During 19th and early 20th centuries the American banking system suspended convertibility 8 times). During crisis, the largest bank in Scotland acted as a leader and the Bank of Montreal in Canada acted as a lender of the last resort. Failure rates and their costs were much lower. In US failure rate during 1870-1909 was 0.36%, while in Canada, was 0.1%.
5. Institutional arrangement that affect the likelihood and severity of panics: 
(1) During Scottish free banking era, banks form large networks and enter voluntary coinsurance arrangement.
(2) Starting 1853 clearing houses associations (CHA) was originated by banks in New York City. The clearing houses clear liabilities, respond to panics, act as lender of the last resort, issue private money, provide deposit insurance, audit risk-taking activities, set capital requirement and penalize members who violate rules. During panics, it issues clearing house loan certificates in small denominations which are a liability of the association. This brand name effect did actually contain some of the bank panics. In 1913, Congress created the Federal Reserve System, which is a federal clearing system modeled on the many private clearing houses that had sprung up across US.
6. What are more likely to trigger runs nowadays?
(1) Runs on managed liabilities (large CDs, interbank loans, commercial papers, Repos): Sudden withdrawal of large CDs, fail to roll over commercial papers and Repos, etc.
(2) Losses on derivatives trading.

III. Contagion
No matter what the cause of bank runs is, it is widely believed that bank runs are contagious, that is, a run on one or a few (may be insolvent) banks can trigger off runs on other (otherwise liquid and solvent) banks. According to this view, bank runs on an individual or a few banks can spread like infectious diseases or forest fires and shatter confidence in the banking system and ultimately provoke a system-wide run, or a bank panic. Thus, the banking system is susceptible to systemic risk.
We say that a bank panic occurs when bank debt-holders of many banks suddenly demand that banks convert their debt claims into cash at par to such a extent that banks suspend convertibility of their debt into cash. Because it is believed that bank run contagion is a major feature of the historical record, the widespread fear of bank runs makes the public and regulators to consider bank runs an evil to be avoided at any cost.
The argument for bank run contagion can be divided into two categories:
(1) Pyramiding and web of interbank transactions and payment clearing, leading to a chain of liquidity squeeze between creditor-debtor banks. This may be due to regional economic problems (agricultural disaster or regional recession), or some financial institutions are heavily exposed to credit risk.
(2) Due to asymmetric information, depositors lack information about the quality of their banks’ assets. Upon observing a bank failure, a depositor is unable to tell whether his own bank is financially sound. If the two banks are similar in some ways, this may cause depositors at other banks to suspect their own banks’ soundness and decide to run on their own bank as well. Since banks operate on a “sequential service constraint,” depositors have strong incentives to show up at the bank as early as possible. When many depositors react the same way, this leads to bank runs on both insolvent and solvent banks.

IV. Deposit Insurance and other forms of safety net
The purpose of deposit insurance is to short-circuits bank failures and contagion effect. Banks pay deposit insurance premium to the insurer (FDIC in Us and 中央存款保險公司 in Taiwan) and depositors are insured up to a certain limit in case the bank fails. There are two main methods to handle a bank failure:
(1) Payoff method
(2) Purchase and assumption method:  This is equivalent to guaranteeing all deposits.

Besides deposit insurance, some governments often provide liquidity to banks when they face runs (lender of last resort) or immediately extend deposit insurance to complete insurance and/or all types of deposits, or even guarantee non-deposit liabilities. For example, in Taiwan all troubled banks are merged to public banks or taken over by the government (中央存款保險公司 or 金融重建基金), thereby guaranteeing all deposits and other bank creditors (“government guarantee”).

V. Moral Hazard
1. With deposit insurance (or even “government guarantee”), depositors have no incentives to impose market discipline on their banks by withdrawing deposits when they perceive that the bank is taking too much risk. Therefore, banks with a government safety net have an incentive to take on greater risks. A most likely result is that taxpayers pay off the bill.
2. The structure of differential deposit insurance premiums must be able to effectively differentiate the risk born by banks.

VI. Adverse Selection
Risk-lovers will find banking industry an attractive business. They will raise deposit rate to attract more deposits, increasing the costs of good bankers. Since depositors are insured (or even guaranteed), they will deposit at banks that offer the highest deposit rate. Good banks will suffer deposit outflows.
Even worse, outright crooks will also find banking attractive because insured depositors have no incentive to monitor their banks, and it is easy for them to get away with fraud and embezzlement.

VII. Too Big to Fail
FDIC guarantees all deposits and bond debts when Continental Illinois became insolvent in 1984.
Government provides guarantees of repayment to large uninsured creditors (bondholders) of the largest banks even when they are not entitled to this guarantee. If large depositors of a big bank know that their bank is guaranteed, they have no incentive to monitor. This increases moral hazard incentives for big banks. The consequence of too-big-to-fail is that big banks take on even greater risks, thereby making bank failures more likely.
Financial Consolidation creates more mega size banks. These larger and more complex banking organizations challenge banking regulation;
(1) Increased “too big to fail” problem
(2) Extends safety net to new activities, increasing incentives for risk taking in these areas.

Restrictions on Asset Holding
Even without government safety net, banks have incentives to take on too much risk. Since most small depositors are incapable of monitoring their banks, this provides a rationale for government regulation on reducing banks’ risk-taking.
1. Promote diversification (貸款關係人限制)
2. Restrict holdings of risky asset (common stocks)

Bank Capital Requirements
1. Minimum leverage ratio
2. Basel Accord: risk-based capital requirements

Bank (Prudential) Supervision: Chartering and Examination
1. Chartering (screening of proposals to open new banks) to prevent adverse selection.
2. Examinations (scheduled and unscheduled) to monitor capital requirements and restrictions on asset holding to prevent moral hazard
CAMELS ratings (Capital adequacy, Asset quality, Management, Earnings, Liquidity, Sensitivity to market risk). The scale is from 1 to 5 with 1 being strongest and 5 being weakest.

目前我國金融監理之金融預警系統採用CARSEL指標，主要包含：「資本適足性（ capital adequacy）」、「資產品質（ asset quality ）」、「守法性（ regulations' compliance ）」、「經營策

略與穩定性（strategies and stability）」、「盈利性（earnings）」與「流動性（liquidity）」

Assessment of Risk Management

1. Changes in financial environment for banking lead the bank examiners to place more emphasis on evaluating soundness of management processes for controlling risk.

2. Promote sound corporate governance.

Information Disclosure Requirements
Requirements to adhere to standard accounting principles and to disclose wide range of information.

Restrictions on Competition

1. Regulation Q (Interest Rate Ceilings)
2. Branching restrictions (eliminated in 1994)
3. Glass-Steagall Act (repeated in 1999)