View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

efficient resolution of thrift insolvencies once the
FSLIC became insolvent. Our suggestions for reform
arise from an analysis of the bankruptcy law as it
applies to unregulated nonfinancial firms, which do
not have access to the kinds of government guarantees provided by deposit insurance. Recommended
changes include incentives to discourage depositors
from funding insolvent institutions together with a
system of judicial oversight of bank and thrift failure
resolution proceedings similar to legal bankruptcy
proceedings established to deal with financially
troubled firms.
The paper is organized as follows. Section I provides general background on private lending arrangements and the nature of bankruptcy proceedings.
These arrangements are compared with the system
of government regulation and failure resolution proceedings for insured deposit-taking
institutions.
Section II examines the evolution of federal deposit
insurance and provides a detailed history of the
savings and loan crisis. Section III explores different
reform proposals. Section IV presents a summary and
conclusions.

I. MARKET DISCIPLINE,
DEPOSIT
INSURANCE, AND BANK FAILURES
The present-day financial regulatory system is in
part a legacy of the waves of Depression-era bank
failures. Legislation enacted in response to the events
of that period created a “financial safety net,” comprised of federally sponsored deposit insurance
together with increased government regulation and
supervision of financial intermediaries.
(A third
important element of this safety net, access to the
Federal Reserve discount window, had been established in response to earlier financial crises.) The
government assumed responsibility for protecting
depositors in the resulting system, with the federal
deposit insurance funds (the FDIC and the now
defunct FSLIC) assuming the role of creditor to
insured depository institutions. Acting in this role,
government regulatory agencies assumed responsibility for ensuring the safe and sound operation of
insured institutions. To facilitate this task, these
agencies were given the authority to issue regulations
restricting the activities of insured banks and thrifts
and also to supervise them to ensure that the rules
were followed.
When viewed from this perspective, it appears that
the goals and interests of government policy with
regard to bank regulation should coincide with those
4

ECONOMIC

REVIEW,

of depositors and other private creditors. But recent
history suggests otherwise. In an attempt to understand why the system failed, the analysis that follows
will compare the incentives created by the federal
financial safety net with the incentives inherent in
purely private financial arrangements. We analyze
market mechanisms designed to cope with problems
that arise when private funds are managed by others.
In particular, we concentrate on the methods employed by private creditors to “regulate” the activities
of borrowers and examine the resolution of creditor
claims under the legal bankruptcy proceedings. We
also describe the self-regulatory practices of the
nineteenth century American clearinghouses, which
offered depositors a form of private deposit insurance. The description of private financial arrangements provides a model that can be used to critically evaluate the federal system of deposit insurance
and regulation.

Risk, Market Discipline,

and Bankruptcy

The contemporary view of the modern business
firm emphasizes the diverse interests of the different
parties participating in the operation of the organization (Coase 1937; Alchian 1968; Jensen and
Meckling 1976; Fama 1980; Fama and Jensen 1983a,
1983b). Firms are viewed as a nexus for a set of contracting relationships among different economic
agents. This “property rights” view treats suppliers
of productive inputs, such as labor, as well the holders
of financial claims (shareholders and creditors), as
stakeholders whose claims against the firm are
governed by either implicit or explicit contractual
arrangements. The managers of a firm constitute a
special type of labor input responsible for coordinating
the activities of others and executing contracts among
suppliers of productive inputs.
Most large organizations are characterized by a
separation of risk-bearing and decision-making. Individuals who bear the residual risks associated with
the operation of an organization typically delegate
decision-making
responsibility
to professional
managers. The modern business corporation provides
the most familiar example of this type of organizational structure. Corporate managers make decisions
for the firm, taking risks whose costs are borne by
shareholders as well as others with a stake in the firm.
Since managers rarely hold a significantfraction
of
corporate equity, they do not bear the full cost of
bad decisions nor reap the full benefits of good ones.
Financial mutuals such as mutual insurance companies and, notably, many savings and loan associations,
MARCH/APRIL

are also
1990

characterized

by a separation

of

decision-making and risk-bearing. Residual profits of
mutuals accrue to their customers, who therefore bear
the residual risks stemming from the operation of
those organizations (although deposit insurance limits
the extent to which depositors bear residual risks at
insured savings and loans). In this sense, the
policyholders of a mutual insurance company or
depositors in a mutual savings and loan can be
thought of as “owners” (Fama and Jensen 1983a,
1983b).
Since managers do not bear the full costs resulting from their decisions, their interests may differ
from those of shareholders and creditors. To ensure
that managers have incentives to act in the interests
of shareholders, large firms typically rely on hierarchical organizational structures to monitor and
evaluate performance. A board of directors consisting
in part of individuals outside the firm’ management
s
hierarchy evaluates the performance of its senior
management.
Markets play an important role in providing both
managers and board members with incentives to act
in the interests of shareholders. Managers have an
incentive to acquire a reputation for effective performance to enhance their career prospects. Outside
directors often sit on more than one board and have
an incentive to discharge their duties effectively so
as to secure invitations to join other boards of directors. The market for corporate control also provides
a powerful incentive for corporate boards of directors and managers to act in the interests of
shareholders. Poor performance by management is
often reflected in a corporation’
s share price,
making the organization susceptible to a takeover
from another management team.
For financial mutuals, such as savings and loan
associations and insurance companies, the channels
through which the market disciplines the firm’
s
decision-makers are somewhat different. The residual
claims of mutuals are redeemable on demand at a
price determined by a prespecified rule. Thus, the
policyholders of a mutual insurance company can
redeem their policies before they mature according
to terms specified in the policy. Similarly, depositors
at mutual savings and loans can withdraw their
deposits, receiving the amount deposited plus the
stated interest. In the absence of deposit insurance,
depositors would be expected to withdraw their funds
upon learning that the association’ management had
s
embarked upon a risky and imprudent investment
strategy. As Fama and Jensen (1983a, p. 317) exFEDERAL

RESERVE

plain, “The decision of a claim holder to withdraw
resources is a form of partial takeover or liquidation
which deprives management of control over assets.”
The role of risk-bearing is most often associated
with the shareholders of a firm, but the limited liability feature of common equity imposes some of the
residual risk on a firm’ other stakeholders, most
s
notably its creditors. Private lending arrangements
reflect a recognition on the part of lenders that the
borrowers can potentially benefit by undertaking
actions that shift risk to the lender after a loan is
made. A borrowing firm can effectively transfer risk
to lenders by siphoning off assets to the stockholders
through excessive dividend payments, by increasing
the riskiness of the business, or by pledging its assets
to another creditor.
For this reason, the extension of credit is often
accompanied by a legally binding agreement limiting
the uses of borrowed funds. Banks typically extend
credit only after gathering extensive information about
the borrowing firm, and typically continue to monitor
the activities of borrowing firms after funds are
disbursed (Stiglitz 1985). Other creditors, such as
outside bondholders, commonly require covenants
limiting the actions of the borrowing firm. In addition to restricting the use of the borrower’ assets,
s
such “bonding” agreements typically require the borrower to disclose certain events to the lender and
may provide for direct supervision of the borrower’
s
business by the lender (Black, Miller, and Posner
1978). Legal bankruptcy proceedings provide the
ultimate means of enforcing the interests of creditors
by alleviating important incentive problems that arise
when a firm is insolvent or nearing insolvency.
When shareholders hold a substantial stake in a
firm they bear much of the residual risk stemming
from its activities. But once shareholder equity is
dissipated, the limited liability feature of common
stock makes added risk-taking consistent with the
interest of shareholders. Under such conditions a
risky investment strategy may actually benefit
shareholders because even a small probability of a
large gain can result in large residual profits and
restore the firm to solvency, while any losses stemming from such a strategy are borne by creditors.
At the same time, the threat of pending bankruptcy
can affect the incentives faced by the managers of
the firm. Managers advance their careers by demonstrating competence at coordinating the activities of
firms. While the insolvency of a firm need not always
BANK

OF RICHMOND

5

be due to managerial incompetence, bankruptcy proceedings typically damage the career prospects of a
fum’ managers. Thus, the managers of a failing firm
s
may perceive themselves as having little to lose from
pursuing a strategy of excessive risk-taking, viewing
it as the only opportunity left to rescue the firm as
well as their reputations. For these reasons, creditor3
typically seek to take control of a firm away from its
existing management at the first sign of insolvency.

The Resolution of Claims Under
Bankruptcy Laws
Legal bankruptcy proceeding3 can be initiated
by the management of a debtor firm as well as by
its creditors. A firm can be forced into legal bankruptcy proceedings by its creditors when it can no
longer meet its debt obligations as they come due,
or when it violates certain debt covenants. In practice, bankruptcy proceedings are often initiated by
a firm’ management when default is imminent. When
s
a firm files for protection from its creditors under
Chapter 11 of the.Bankruptcy Code, its management
nominally retains control of the organization. But
although management remains responsible for supervising the day-to-day operation of the firm, its decisions are subject to judicial review and approval by
creditors. Creditor committees form to oversee the
operations of a firm. As Todd (1986) notes, these
creditor committees hold the real power over all important operating decisions. In effect, the creditor
committees become co-managers of the bankrupt
firm, with their legal representative3 meeting frequently with management. A. trustee may be appointed to administer the operations of the bankrupt
firm if there is evidence of fraudulent behavior on
the part of management.
A firm need not be insolvent to file a voluntary
petition for Chapter 11 protection from its creditors.
Modern bankruptcy law provide3 for the rehabilitation of debtors. The idea behind Chapter 11 proceedings is to effect a reorganization of financially
troubled firms where possible. Once a bankruptcy
petition is filed, the firm is granted an automatic stay
permitting it to stop payment3 to its unsecured
creditors. Secured creditors are prohibited from
taking possession of property from the bankrupt’
s
estate unless they can obtain relief from the automatic
stay. In cases where the bankruptcy judge deems the
property securing a loan to be necessary for the continued operation of a bankrupt organization, a secured
creditor may be effectively forced to renew an extension of credit to the bankrupt firm.

6

ECONOMIC

REVIEW,

The management of a firm in Chapter 11 proceedings is given an opportunity to draw up a
reorganization plan specifying a new financial structure along with a revised repayment schedule for
outstanding debts. Creditor3 are sometimes called
upon to forgive a portion of the firm’ debt to ens
sure the viability of the reorganized firm. They may
agree to such a restructuring of the firm’ debts if
s
it seem3 likely to yield a greater repayment than the
amount that could be realized under any other course
of action, including liquidation.
If a firm’ management does not offer it3 own
s
reorganization plan, or cannot produce a plan acceptable to creditors and to the bankruptcy judge,
creditors can propose an alternative reorganization
plan. The creditors’ plan may call for a new management team to be installed.
A bankruptcy judge acts as a mediator or referee
between management and the different parties with
claims against the firm. Judicial decisions are governed by a set of Bankruptcy Rule3 (See Treister,
et al. 1988). If creditors cannot agree on a reorganization plan, the bankruptcy judge may under certain
circumstances impose a reorganization plan. In Some
cases the court may order the liquidation of a
bankrupt firm.
Liquidation of a bankrupt firm’ assets is govs
erned by Chapter 7 of the bankruptcy code. When
a firm enters Chapter 7 proceedings, a trustee is
appointed to legally represent and administer the
estate. The Bankruptcy Code establishes a schedule
of priorities for the distribution of liquidation proceeds
among unsecured creditors. Administrative expenses
of managing the bankrupt’ estate receive first prior3
ity. Unpaid wages and benefits, up to a certain limit
come next, followed by claims of governmental units
for taxes, customs duties, and accrued penalties. The
claims of holders of investment securities are subordinate to all other unsecured creditors. Thus,
holders of subordinated debt, which includes bond
and note holders, are reimbursed only after the claims
of all other unsecured creditors are satisfied. Preferred shareholders are next, with common equity
shareholders
receiving lowest priority. Secured
creditors are not subject to the schedule of priorities I
(Todd 1986; Treister, et al. 1988, chapter 6).
To summarize, private lenders employ a number
of strategies, including loan covenants, monitoring,
and bankruptcy proceedings when necessary, to protect their claims against a borrowing firm. Although
MARCH/APRIL

1990

these safeguards do not prevent insolvencies, they
do help to limit losses when a borrower becomes
financially distressed.

Private Regulation of Commercial

Banking

Today, regulation is most often viewed as a governmental activity. However, private regulatory organizations often evolve to provide for the orderly
functioning of market activity in the absence of
government
intervention.
Notable examples of
private regulation include the futures and securities
exchanges, which evolved as purely private organizations formed to set and enforce trading rules. The
nineteenth century American commercial bank clearinghouses, which essentially regulated a significant
part of the banking industry before the advent of the
Federal Reserve, provide another example of private
regulation. These clearinghouses provided an informal system of deposit insurance to depositors at
member banks. The historical lessons offered by the
operation of the clearinghouse system therefore seem
relevant to the study of deposit insurance reform,
and the system merits comparison with present-day
deposit insurance arrangements.
Commercial bank clearinghouses were first organized to conserve on the transactions costs associated
with clearing checks. Banks, as organizations that
specialized in information-intensive loans based on
the evaluation of the creditworthiness of individuals,
had a natural advantage in monitoring the creditworthiness of other banks. Moreover, their need to
exchange checks with other banks gave them an
incentive to engage in some form of monitoring.
Thus, the clearinghouses developed into a form of
private regulatory agency.
Because private regulatory arrangements
are
based on the premise that participation is motivated
by self-interest, the most common penalty for failure
to abide by the rules is expulsion from the system.
Thus, futures and securities traders who systematically violate trading rules are banned from trading
on the exchanges. Likewise, the early clearinghouses
denied access to banks that failed to meet the financial standards established by the clearinghouse
member banks.
As a prominent example of how such regulation
was effected in practice, Gorton and Mullineaux
(1987) describe the operations of the New York clearinghouse. Admission to the clearinghouse required
banks to meet an admissions test that required banks
to be well-capitalized and to submit to periodic
FEDERAL

RESERVE

examinations. In times of panic, the clearinghouse
organized suspensions of deposit convertibility and
issued loan certificates to member banks that they
could use in the clearing process in place of specie.
Through the issue of such loan certificates, member
banks essentially pooled their resources to assure
depositors of the ultimate safety of individual member
bank liabilities. In effect, the clearinghouse insured
the deposits of its member banks through this
mechanism.
Such pooling arrangements exposed clearinghouse
members to the threat of losses if a bank proved insolvent. Clearinghouse members therefore had an
incentive to ensure that only sound banks were part
of the clearinghouse. To this end, the clearinghouses
closely monitored member banks and expelled those
that did not satisfy rigorous standards.
Denial of access to the clearinghouse made it much
more difficult and costly for banks to clear checks,
so the threat of expulsion provided banks with a
strong incentive to conform to clearinghouse rules.
Moreover, expulsion was a signal that the banking
community had determined that there was a high
probability that the affected bank would not be able
to meet its deposit obligations. Thus, clearinghouses
became credible suppliers of information about the
financial condition of member banks.
On balance, the nineteenth century clearinghouses
appear to have functioned as effective private
regulatory organizations. Available evidence indicates
that the ultimate losses suffered by depositors of failed
banks during this period were negligible (Timberlake
1984). Despite its effectiveness in this regard, this
private system of regulation was replaced with
government regulation with the formation of the
Federal Reserve, and, after the collapse of the
banking system in the early 1930.5, with federally
sponsored deposit insurance.
The Federal Reserve System was created to
impose greater centralized government control over
the banking system (Timberlake 1984). Under the
clearinghouse system there were recurrent financial
panics and.bank suspensions that were viewed as a
source of macroeconomic instability.’ In addition,
there was some concern that the clearinghouse structure led to industry cartelization and monopoly
1 The assertion that banking panics have been a primary cause
of macroeconomic instability in U.S. economic history has been
disputed by recent research, however (see Benston, et al. 1986,
chapter ‘
2).
BANK

OF RICHMOND

7

profits (Carosso 1973). The power of exclusion gave
clearinghouse members the potential ability to limit
entry into their markets.
Yet, it is worth emphasizing again that the clearinghouse system actually worked quite well at limiting
depositor losses. Moreover, demands for greater
governmental control over monetary policy and concerns over macroeconomic instability, while justified,
do not necessarily provide an argument in favor of
government regulation of banking. As Goodfriend
and King (1988) point out, the exercise of monetary
policy only requires a central monetary authority empowered to carry out open market operations. Thus,
monetary policy should be able to prevent widespread
bank suspensions 30 long as the monetary authority
stands willing to supply reserves to stabilize the
relative price of currency and bank liabilities. To the
extent that preventing financial panics and bank runs
is perceived as an important goal of public policy,
available evidence suggests that liberalizing regulatory
restrictions that limit the ability of banks to establish
branches would be the most effective solution
(Calomiris 1989b). Finally, in the area of antitrust
concerns, existing antitrust laws should be adequate
to guard against anticompetitive behavior in the banking system.2

Deposit Insurance and
Bank Failure Resolution
Because a deposit insurer effectively becomes a
creditor to banks and thrifts, a system of government
regulation and supervision is a necessary adjunct to
a system of government-sponsored deposit insurance.
Government regulation and supervision in this instance is analogous to the monitoring behavior and
other protective devices employed by creditors in
private financial arrangements. The scope of this
regulatory system is comprehensive and extends to
legal arrangements for dealing with bank and thrift
failures, which differ from bankruptcy proceedings
for nonfinancial firms.
Commercial banks and savings and loan associations, along with certain other heavily regulated
financial firms such as insurance companies, are not
subject to the bankruptcy laws that apply to commercial firms. Responsibility for closing an insolvent
bank or savings and loan rests with its chartering
agency. In the case of national banks, the charter2 Kuprianov (1985) gives an account of how antitrust laws
assured savings and loans access to the Automated Clearing
Houses operated by the commercial banking industry.
8

ECONOMIC

REVIEW.

ing agency is the Office of the Comptroller of the
Currency (OCC). Before being disbanded in 1989,
the Federal Home Loan Bank Board chartered federal
savings and loan associations. The Bank Board’
s
chartering authority has since been delegated to a
new agency, the Office of Thrift Supervision (OTS).
In addition to the federal chartering agencies, each
state also charters commercial banks and savings and
loan associations. The state banking or savings and
loan superintendents are responsible for closing statechartered institutions.
Roie of th depositinsurer When an insured bank
or thrift is declared insolvent, the deposit insurer pays
off insured depositors and, in most cases, becomes
the receiver for the failed institution.3 Once the
insured depositors are paid, the deposit insurer
assumes their claims against the failed institution.
Thus, the role of the deposit insurer in dealing with
a failing bank or savings and loan differs considerably
from that of a bankruptcy judge or trustee. Rather
than acting solely in the role of a mediator between
different claimants, as a bankruptcy judge does, the
deposit insurer assumes the dual role of receiver and
claimant.
As receiver, the deposit insurer assumes responsibility for administering the assets of the insolvent
firm and has a fiduciary responsibility to all other
claimants, such as uninsured depositors and nondeposit creditors. In its role as a claimant, the deposit
insurer attempts to secure repayment of deposits from
the failed institution on behalf of insured depositors.
Federal banking law does not grant the deposit insurer preference over other unsecured creditors,
although some states have enacted “depositor
preference” statutes (Hirschhorn and Zervos 1990).
The law gives the deposit insurer substantial discretion in dealing with a failing institution. Nevertheless,
the insurer must seek the cooperation of other
creditors when attempting to reorganize and restructure the debts of an insolvent bank. Although commercial bank and savings and loan failures are not
subject to the same kind of stringent judicial over3 In the past, the FSLIC bore responsibility for administering
federally insured savings and loan institutions when they were
declared insolvent. However, with the enactment of FIRREA,
the FSLIC was dissolved and the FDIC was given responsibility for administering both the Bank Insurance Fund (BIF),
which insures the deposits of commercial banks, and the
Savings Association Insurance Fund (SAIF), the new deposit
insurance fund for savings and loans. In its new role, the FDIC
is responsible for handling insolvent savings and loans as well
as commercial bank failures.
MARCH/APRIL

1990

sight mandated by bankruptcy laws, creditors who
feel they have been treated unfairly do have recourse
to the courts.
FSLJCfaihn ‘
RFOlution~rvct&r~~~Before it became
insolvent and was itself dissolved, the FSLIC enjoyed
a large degree of discretion in the way it dealt with
failing savings and loans. It could: [ 11 liquidate. the
organization and pay off its depositors and other
creditors; [Z] reorganize the enterprise and return it
to private sector control; or [3] extend direct
assistance to enable a troubled institution to remain
in operation.
In a liquidation, or payout, a failing savings and
loan would be closed, its insured depositors paid,
and its assets liquidated. Receivership expenses had
first priority against liquidation proceeds, with remaining proceeds distributed to the association’ creditors.
s
In states with depositor preference statutes, the
claims of the FSLIC and any uninsured depositors
against the failed institution received preference over
other unsecured creditors. In states with no depositor
preference statutes, the FSLIC was forced to share
the liquidation proceeds with other unsecured
creditors. Most often, a troubled savings and loan
was not closed before it had accumulated large losses,
so that liquidation proceeds rarely covered all outstanding creditor claims in full. Only a relatively small
number of failed savings and loans have been liquidated. Between 1980 and 1988 only 78 of the 489
insolvencies officially resolved by the FSLIC were
liquidated.4
In reorganizing a failing savings and loan, the
FSLIC could: [l] directly augment the net worth of
the enterprise, either through direct cash contributions or through the issue of its own promissory notes;
121purchase subordinated debt or preferred stock as
part of a recapitalization; [3] provide the acquirer of
an insolvent institution with financial guarantees and
yield maintenance agreements guaranteeing the performance of the troubled organization’ assets; or
s
14) purchase the impaired assets of a troubled institution at a negotiated price (Zisman and Churchill
1989). Thus, the FSLIC sometimes maintained an
explicit financial stake in an institution after it was
reorganized. In addition to directly augmenting the
4 The number of official failure resolutions understates the true
number of thrift insolvencies during this period. Some troubled
institutions were handled through “supervisory mergers,” described below, while hundreds more insolvent institutions have
been taken over by regulators but have not yet been closed or
reorganized. Barth, Bartholomew, and Bradley (1989b) give data
on thrift failure resolutions and costs from 1934 to 1988.
FEDERAL

RESERVE

net worth of a troubled institution in these ways, the
Federal Home Loan Bank Board, which administered
the FSLIC, would often grant acquirers of troubled
thrifts special permission to acquire other institutions at a later date. To augment the franchise
value of financially troubled institutions, the Bank
Board sometimes provided acquirers with enhanced
branching opportunities or permission to acquire
healthy savings and loans in other states, actions that
state-mandated branching restrictions would otherwise prohibit.
Reorganizations
of troubled thrifts often took
the form of a stlperwirory
merger. In a supervisory
merger, regulators would arrange the merger of a
financially troubled institution with another institution deemed to be in better financial condition.
Supervisory mergers were accomplished without the
explicit financial assistance of the FSLIC. Normally, the FSLIC would have been expected to
recapitalize a failing institution before arranging a
merger. But as the deposit insurer’ financial resources
s
became strained, the Bank Board was forced to grant
regulatory forbearances to arrange mergers of insolvent organizations. In some cases, regulatory forbearance amounted to a waiver from regulatory
minimum net worth requirements. In many cases,
however, such forbearances involved permission to
employ liberal accounting procedures that authorized the acquirer to defer recognition of the losses
of the insolvent thrift almost indefinitely. Thus,
supervisory mergers often simply consolidated losses
into larger organizations that were permitted to continue operating without private capital. Between 1980
and 1988, 333 institutions were involved in supervisory mergers (Barth, Bartholomew, and Bradley
1989b, Table 1).
Since supervisory mergers did not require explicit
action on the part of the FSLIC, they are not officially
counted as failure resolutions. As Kane (1989b) notes,
however, the grant of regulatory forbearance made
the FSLIC the residual risk-bearer for undercapitalized enterprises that would otherwise have been
unable to attract funding. To the extent that supervisory mergers were based on promises of regulatory
forbearance, the FSLIC maintained an implicit equity
stake even in cases where its stake in the merged
firms was not made explicit.
The Bank Board had the authority to assume control of a financially troubled organization until it could
be reorganized and sold to private investors. It exercised such “conservatorship” powers in its Management Consignment Program, which is discussed in
greater detail in Section II.
BANK

OF RICHMOND

9

Finally, the FSLIC (as well as the FDIC) was
authorized to extend direct assistance to a financially troubled institution if such an action, was
deemed less costly than any of the other available
courses of action, or in cases where the institution
was judged to be “vital” to its community (Benston,
et al. 1986, chapter 4).
Ostensibly, then, FSLIC failure resolution procedures resemble legal bankruptcy proceedings, in
that they are meant to bring about the reorganization of a failing firm or provide for its liquidation in
cases where reorganization is not deemed worthwhile.
In practice, however, the savings and loan regulatory
system has proved ineffective at limiting the losses
incurred by insolvent institutions. Whereas legal
bankruptcy proceedings ensure that ,the debts of an
insolvent firm are restructured in such a way that
shareholders have an equity stake before the firm is
returned to private control, the same has not always
been true of FSLIC failure, resolution procedures.
Moreover, the system proved ineffective at curbing
the risks taken on by the management of failing
institutions. The practical importance of these differences will become evident in the ensuing account
of the evolution of the savings and loan crisis.

II. HISTORY OF THE
SAVINGS AND LOAN CRISIS
The origins of the savings and loan crisis are rooted
in the system of regulation imposed on the industry.
The ensuing account describes the evolution of the
system and highlights the characteristics that later
precipitated an industry-wide crisis. We then proceed
to a detailed account of the crisis itself.

Federal Regulation of Savings and Loans
The savings and loan regulatory system of the
1980s was a product of legislation enacted during the
Great Depression. Before 1932, the federal government had little involvement in thrift regulation.
Savings and loans shared in the financial distress that
afflicted commercial banks during this episode. In
an attempt to assist the thrift industry, which had
begun to contract due to heavy deposit withdrawals,
Congress passed the Federal Home Loan Bank Act
in 1932. The Act created the twelve Federal Home
Loan Banks and the Federal Home Loan Bank Board
as their supervisory agent. The goal of this legislation was to provide thrifts with an alternative source
of funding for home mortgage lending, much in the
same way that the Federal Reserve Banks provided
temporary funding for commercial banks. While the
10

ECONOMIC

REVIEW,

Federal Reserve Banks only provided short-term
credit, however, the Federal Home Loan Banks were
created to provide longer-term credit in support of
mortgage lending.
The federal government
became involved in
chartering savings and loans for the first time in 1933
with the passage of the Home Owners’ Loan Act,
which authorized the FHLBB to charter and regulate
savings and loan associations. In 1934, a year after
a system of deposit insurance was established for
commercial banks, the National Housing Act of 1934
created a deposit insurance fund for savings and loan
associations. Unlike the FDIC, which was established
as an independent organization separate from the
Federal Reserve System and the Comptroller of the
Currency,, the FSLIC was placed under the auspices
of the FHLBB.
The legislation creating the FSLIC called for the
establishment of a reserve fund equal to five percent
of all insured accounts and creditor obligations within
20, years, and empowered the agency to assess an
annual insurance deposit of l/4 of one percent on
the total deposits of insured S&Ls. The FSLIC was
further authorized to collect an additional emergency assessment of l/4 percent if it needed additional funding. At first, deposits were insured up to
a maximum of $5,000 per depositor.
When federal deposit insurance was first established, both the FDIC and the FSLIC were expected
to accumulate and hold reserves sufficient to pay off
all insured depositors under any foreseeable circumstances. The legislated deposit insurance assessments
and reserve fund targets were based on estimates of
the historical losses of depositors. But federal deposit
insurance had not been in existence long before the
deposit insurance assessments were cut and coverage
expanded. In 1935, a year after the FSLIC was
established, statutory deposit insurance assessments
for insured savings and loans were cut in half, to
l/8 of one percent of deposits. The emergency
assessment authority was similarly cut to l/8 of one
percent. That same year, the FDIC’ assessments
s
were cut from l/2 of one percent to l/12 of one percent, and its emergency assessment rights were
rescinded.
The argument for lowering deposit insurance rates
was based upon the assertion that enhanced regulation and supervision would keep future losses of
insured banks below the historical averages. At the
same time, however, there appears to have been
MARCH/APRIL

1990

some awareness that lowering deposit insurance
assessments could result in future funding problems
for the deposit insurance funds. FDIC deposit insurance assessments were reduced to l! 12 of one percent by the Banking Act of 1935, which also provided the agency with the right to borrow from the
U.S. Treasury. The FSLIC was granted similar borrowing authority in 1950, when deposit insurance
assessments for S&Ls were cut to 1112 of one
percent.
Over the ensuing years, basic insurance coverage
for S&L depositors was raised several times: to
$15,000 in 1966, $20,000 in 1969, $40,000 in 1974,
and, most recently, to $100,000 in 1980. These increases in coverage, together with a rapid growth in
deposits throughout most of the postwar period, far
outpaced the accumulation of reserves in the FSLIC
insurance fund. The five percent reserve fund target
originally mandated by the National Housing Act was
never attained. The FSLIC’ primary reserve fund
s
never exceeded two percent of insured deposits
(Barth, Feid, Riedel, and Tunis 1989).
Thus, historical data on bank losses suggests that
neither deposit insurance fund has had the necessary
reserves to deal with the contingency of widespread
bank failures. Both the FDIC and the FSLIC have
faced a chance of insolvency almost since their inception. Moreover, both agencies received the
authority to borrow from the U.S. Treasury as part
of legislated reductions in deposit insurance rates and
increases in deposit insurance coverage. As Barth,
Bradley, and Feid (1989) note, however, no formal
procedures were ever established for dealing with the
insolvency of one of the deposit insurance funds, even
though the funds were structured in a way to make
such a contingency distinctly possible, if not inevitable. Thus, the stage for the present-day savings
and loan crisis was set as early as 1950.

Origins of the Savings and Loan Crisis
The first signs of trouble surfaced in the mid 196Os,
when rising inflation and high interest rates created
funding problems for savings and loans. Regulations
prohibited federally insured savings and loans from
diversifying portfolios that were concentrated in longterm, fixed-rate mortgages. Thrift industry profitability eroded as deposit rates crept above the rates
of return provided by thgir existing holdings of home
mortgage loans. Congress attempted to address the
problem by placing a ceiling on maximum deposit
rates paid by thrifts in 1966. Thrifts were given a
slight competitive advantage, being authorized to pay
FEDERAL

RESERVE

114 of one percent more on savings deposits than
commercial banks were allowed to pay, to encourage
deposit flows to the industry.
But interest rate controls led to periods of
disintermediation whenever market interest rates rose
too far above statutory deposit rate ceilings. The
problem became increasingly severe as the inflation
and accompanying high interest rates that characterized the economic environment of the late 1970s
made the existing system of interest rate controls
unworkable. Misguided regulation was blamed for the
thrift industry’ woes, and lawmakers began to debate
s
the merits of financial deregulation.
The first significant step to deregulate the thrift
industry came in 1980 with the passage of the
Depository Institutions Deregulation and Monetary
Control Act (DIDMCA). The DIDMCA provided
for the phase-out of interest rate regulations and
permitted thrifts to diversify their asset portfolios to
include consumer loans other than mortgage loans,
loans based on commercial real estate, commercial
paper, and corporate debt securities. The act also
raised the limit on federal deposit insurance applicable
to individual accounts from $40,000 to $100,000.
This first attempt at deregulation came too late
to help thrifts cope with the steep rise in interest
rates that began in 1981 and continued into 1982.
Federally chartered S&Ls were not given the legal
authority to make variable-rate mortgage loans until
1979, and then only under severe restrictions. They
did not receive the authority to freely negotiate
variable-rate mortgage loans with borrowers until
198 1. By that time, deposit rates had risen well above
the rates most institutions were earning on their
outstanding fixed-rate mortgage loans. As funding
costs rose, many thrifts experienced heavy losses.
Federally insured savings and loans collectively lost
over $4.6 billion in 1981 and $4.1 billion in 1982.5
By one estimate, 8.5 percent of all thrifts were
unprofitable in 198 1, and most were insolvent on an
economic basis (Barth, Bartholomew, and Labich
1989). From the start of 1980 through year-end
1982, the number of FSLIC-insured thrifts fell almost
20 percent, from 3,993 to 3,287.
s Net operating income, which more accurately reflects the true
losses suffered by thrifts during this period because it excludes
nonrecurring gains, presents an even more devastating picture
of losses suffered bv savings and loans durine this neriod.
According to Barth, Bartholomew, and Bradley 0989, kppendix I-8), aggregate net operating income for the U.S. thrift
industry in 1981 was -$7.1 billion and -$8.8 billion for 1982.
BANK

OF RICHMOND

11

The industry’ staggering losses overwhelmed the
s
resources of the FSLIC. Hundreds more institutions
that had become economically insolvent were not
closed because the FSLIC lacked the resources to
deal with them. Many economically insolvent thrifts
were able to maintain the appearance of solvency
even though they were economically insolvent
because generally accepted accounting practices
(GAAP) permitted them to report their net worth
based on historical asset value,, instead of requiring
them to recognize the true market value of assets.
But as interest rates continued to rise, a significant
number of institutions soon accumulated such
massive losses that some action was required. The
FSLIC resolved 32 thrift insolvencies in 1980,
another 82 in 1981, and 247 in 1982. During the
same period, another 493 savings and loans voluntarily merged with other institutions (Barth, Bartholomew, and Bradley 1989b, Table 1). In spite of
this record-breaking
caseload,
Kane (1989b)
estimates that 237 FSLIC-insured
thrifts were
GAAP-insolvent at the end of 1982. The number
of insolvent insured thrifts in operation continued to
climb through 1988.

Regulatory Forbearance
Once the crisis in the savings and loan industry
had begun, it was perpetuated by policies of regulatory forbearance, which permitted insolvent institutions to remain open in the hope that they could grow
out of their financial problems. The policies adopted
to deal with the growing number of insolvent savings and loans during this episode stand in stark contrast to the restrictions on management typically
imposed in the course of legal bankruptcy proceedings for nonfinancial firms.
FHLBBpohiis Lacking the resources to deal with
all the problem institutions under its supervision, the
Bank Board adopted a policy of regulatory forbearance. Minimum net worth requirements
were
lowered in 1980 and 1982. Regulatory accounting
principles (RAP) were liberalized in 198 1, and again
in 1982, to permit distressed savings and loans to
defer recognizing their losses. These permissive rules
encouraged thrifts to record inflated net worth values
so as to present an appearance of solvency. Together,
lenient net worth requirements
and permissive
regulatory accounting principles lowered the number
of official “problem” institutions the overburdened
Bank Board staff was forced to deal with, although
only for a short time (Brumbaugh 1988).
By this time, many thrifts had accumulated such
large losses that even these new and permissive
12

ECONOMIC

REVIEW,

accounting rules could not conceal the fact that they
were insolvent. Concerned that acknowledging the
large number of insolvent savings and loans could
bring about a crisis of confidence among depositors,
the FHLBB implemented its income-capital certificates (ICC) program. Under this program, insolvent
thrifts could issue income capital certificates to the
FSLIC to supplement their regulatory net worth. The
idea behind the program was for the FSLIC to purchase the certificates to restore troubled institutions
to solvency. Because the FSLIC lacked the money,
it most often exchanged its own promissory notes
for the certificates. Institutions receiving such
promissory notes could include them on their balance
sheets as assets, while income capital certificates were
reported as an equity item. Such transactions
amounted to the purchase of equity in an insolvent
enterprise by the FSLIC using its own credit.
Income-capital certificates gave the FSLIC a financial interest in these troubled thrifts. If participating
institutions eventually regained profitability, as it was
hoped they would, the income-capital certificates
would entitle the FSLIC to a share of their profits.
But in the event a participating institution was
declared insolvent, the FSLIC had virtually no
chance of regaining its investment. FSLIC claims
based on income capital certificates were subordinate
not only to the claims of depositors, but of other
creditors as wel16
Where possible, the FSLIC used income-capital
certificates to facilitate mergers and reorganizations.
Prospective buyers were hesitant to assume the
liabilities of insolvent thrifts when it appeared that
the value of the institutions’ assets fell far short of
deposit obligations. Sometimes, the FSLIC transferred assets from thrifts it was in the process of
liquidating to other institutions it was trying to sell.
This latter course was typically pursued where purchasers of insolvent thrifts were reluctant to accept
FSLIC promissory notes. Many prospective acquirers
either could not or would not invest enough of their
own resources to fully recapitalize a failing institution. In such cases, the FSLIC would help effect a
recapitalization by exchanging its promissory notes
for income-capital certificates, which were transferable to the acquiring institution. In essence, the
6 Income-capital certificates did not have any stated maturity,
and were not collateralized or secured. Thus, in the event of
legal insolvency, income-capital certificates gave the FSLIC
essentially the same status as those of a holder of preferred
equity and not those of a creditor (see GAO, Th Management
Cons&ment Program, September 1987; and American Banker,
12/26/85).
MARCH/APRIL

1990

FSLIC became a partner in the new, reorganized
institution.
In many of these reorganizations, the thrift’ new
s
owners had very little of their own financial resources
at stake. Many times, the acquirer was a marginally
solvent thrift with little or no capital of its own.
Such institutions were able to expand rapidly by
taking over other thrifts in even worse financial condition. In the end, the FSLIC bore virtually all
residual risks while the management and shareholders
of the acquiring institution stood to profit handsomely
if their attempts to expand their operations proved
profitable.
The Bank Board pursued such policies out of a lack
of good alternatives. It lacked the resources to close
the insolvent institutions, and because only the
chartering agency - which was the Bank Board itself
in the case of federally chartered thrifts-could
declare a savings and loan legally insolvent, financially troubled thrifts could be kept open indefinitely. Unfortunately, the Bank Board also lacked the
resources to adequately monitor the many insolvent
savings and loans for which the FSLIC had become
the residual risk bearer. At the same time, deposit
insurance made it possible for even the most poorly
managed and unprofitable thrifts to continue expanding their operations. Keeping insolvent thrifts
open under these circumstances permitted the
FSLIC to defer recognizing its losses, but exposed
the fund to the risk of very large future losses.
Th Garn-St. GermainAct Lawmakers responded
to these developments by enacting the Garn-St. Germain Act of 1982, which combined a program of
regulatory forbearance together with further thrift
industry deregulation. To encourage greater regulatory forbearance toward financially troubled thrifts,
the Act created the “net worth certificate” program.
The net worth certificate program was essentially a
derivative of the income-capital certificates program
devised earlier by the Bank Board. Net worth certificates differed from income capital certificates in
that they did not constitute a permanent equity investment, but were issued only for a set time period
authorized by the legislation. Unlike income-capital
certificates, net worth certificates were not transferable and so were not useful in reorganizing insolvent institutions or arranging mergers. In fact, the
stated purpose of the net worth certificate program
was to forestall forced mergers or other regulatory
actions against insolvent thrifts (see GAO, Net wart/l
Certajicate
Asistance Pqyams, ‘ 84).
19
FEDERAL

RESERVE

At the same time, the Garn-St. Germain Act
attempted to reform the elements of the regulatory
structure most often blamed for the industry’ probs
lems by liberalizing investment powers of federally
chartered thrifts. Some states such as California took
the initiative to deregulate savings and loans even
further, authorizing state-chartered thrifts to engage
in activities such as direct participation in real estate
development.
Other states, notably Texas and
Florida, had granted their state-chartered savings and
loans liberalized investment powers years earlier.
Thus, the Garn-St. Germain Act attempted to
forestall action in the hope that the combined policies
of forbearance and deregulation would facilitate a
return to profitability and financial health among
insolvent thrifts. These policies were adopted in an
effort to avert the need for a federally financed rescue
of the FSLIC. Rather than providing the Bank Board
with the resources needed to begin closely monitoring and closing problem institutions, the net worth
certificate program discouraged regulators from
acting. But the added risks that continued regulatory
forbearance posed to the FSLIC fund were underestimated. Those risks were soon to become
apparent.
Eariy attemptsat reregulation Instead of improving with time as policymakers had hoped, the financial condition of insolvent thrifts continued to
deteriorate. Market interest rates had begun a pronounced and sustained decline by the end of 1982,
and economic conditions improved as the severe
recession that had begun a year earlier ended. Lower
interest rates and favorable economic conditions
throughout the nation as a whole did facilitate the
recovery of some thrifts, but a large and rapidly
growing segment of the industry continued to incur
heavy losses. Although rising interest rates had
triggered the savings and loan crisis, the subsequent
decline in interest rates to more normal levels failed
to restore financial health to many of the insolvent
institutions that had been kept open.
It was apparent to all in the industry by this time
that the FSLIC did not have the resources to give
attention to more than a few of the most financially
troubled institutions. The number of Bank Board and
FHL bank examination and supervisory personnel
actually declined between 1981 and 1984, even as
the number of thrift insolvencies soared (Barth and
Bradley 1988, 46-47). Attempts by the Bank Board
to augment the supervisory staff were discouraged
by the Office of Management and Budget. Armed
BANK OF RICHMOND

13

with an unlimited government financial guarantee and
new investment powers, many insolvent thrifts found
it easy to engage in a variety of risky and imprudent
investment schemes. As time went on, evidence surfaced that the losses at many institutions were
attributable to gross mismanagement, and in some
cases to outright fraud.
The rapidly deteriorating financial condition of the
many insolvent S&Ls that had been kept open had
begun to become apparent as early as 1983, when
the Bank Board began taking steps to limit the risks
that poorly capitalized but aggressively managed
thrifts imposed on ‘
the FSLIC. The agency proposed rules to limit the use of brokered deposits by
undercapitalized,
rapidly expanding thrifts. That
attempt ultimately proved unsuccessful, however,
when the courts ruled that the agency lacked the
legal authority to impose such a rule and lawmakers
refused to grant the necessary authority.
Capital requirements
were raised for newly
chartered institutions,
but the new capital requirements were not applied to existing institutions.
The income-capital certificate program was briefly
discontinued, only to be revived again two years later.
In 1985, the Bank Board proposed to effectively raise
minimum net worth requirements by rescinding some
of the liberal accounting rules introduced in 198 1.
It also proposed to limit the investment powers of
undercapitalized federally insured thrifts.
Unfortunately,
these initiatives proved largely
ineffective in stemming the growing losses incurred
by insolvent and ‘
inadequately supervised thrifts.
Attempts by the Bank Board to restrict the activities
of state-chartered thrifts drew considerable resistance
from legislators and regulators in. states such as
California, Florida, and Texas, where those institutions had been granted broad investment powers.
Managers of insolvent thrifts, aware that the FSLIC
lacked the resources to closely supervise more than
a fraction of all the undercapitalized institutions it
insured, proved difficult to control. Thrift industry
assets grew almost 20 percent in 1984 alone (See
GAO, Thr$Indusny Rtxtnxtming, 1985, p. 8). Unlike
the initial financial difficulties of most insolvent
thrifts, which were largely attributable to the effect
of high interest rates on the value of their mortgage
portfolios, most losses after 1982 stemmed from
credit quality problems. According to Brumbaugh
(1988, p. 67), asset quality problems were the principal cause behind the losses experienced by 80 percent of the institutions comprising the FSLIC’
s
14

ECONOMIC’
REVIEW.

caseload of problem thrifts in 1984. In contrast, asset
quality problems were seen to be the primary cause
of the losses experienced by only 20 percent of
problem thrifts between 1980 and 1984.
In certain respects the Garn-St. Germain Act can
.
be judged to have achieved its goals. Mortgage assets
declined as a proportion of all assets held by savings
and loans after 1982, with insolvent institutions taking
greatest advantage of their new investment powers.
Unfortunately, the institutions most aggressive in
exploiting their new powers also experienced the
greatest deterioration in asset quality. Those institutions subsequently exposed the FSLIC to large
losses (Barth and Bradley 1988, Tables 4 and 5).
Th ManagementCons&zment
Program By 1985, it
was becoming apparent that the combined policy of
regulatory forbearance and deregulation first adopted
in response to the thrift industry crisis had failed to
restore financial health to the industry. Instead, it
was proving to be a prescription for disaster. In an
attempt to gain greater control over insolvent thrifts
that continued to experience growing losses, the Bank
Board instituted its “Management Consignment Program” (MCP). An institution brought into the MCP
typically had its management replaced with a conservator selected by the Bank Board. The program
was conceived as a means of temporarily warehousing hopelessly insolvent institutions until they could
be sold or liquidated by the FSLIC. Many institutions placed in the MCP in 1985 were still in the
program and still incurring losses two years later (see
GAO, The ManagementConsignment
Pnpam, 198 7).
The income-capital certificates program was re-.
introduced for institutions placed in the MCP.
Using its own promissory notes to “recapitalize”
insolvent thrifts, the FSLIC attempted to sell or
merge those institutions. But as industry conditions
grew worse, it became increasingly apparent to
market participants that the FSLIC lacked the financial resources to deal with the heavy losses accumulated by troubled S&Ls. Potential acquirers
became reluctant to accept the FSLIC’ promissory
s
notes, further hampering the agency’ efforts to sell
s
off insolvent thrifts. Investor reluctance to accept
FSLIC notes stemmed at least in part from a ruling
by the Financial Accounting Standards Board that
such notes could not be counted as assets in determining net worth under Generally Accepted Accounting Principles (see GAO, Th ManagementConsignment Pmgcam, 198 7).
MARCH/APRIL

1990

The Demise of the FSLIC
By 1985, the rapidly deteriorating condition of
many insolvent thrifts had so strained the resources
of the FSLIC that the Bank Board finally had to admit that the insurance fund needed outside funding.
But as long as depositors at insolvent thrifts felt confident that the U.S. Treasury would ultimately
guarantee the safety of their deposits, they had no
reason to withdraw their funds. And as long as insolvent thrifts could continue to attract deposits, there
was no incentive to appropriate the funds needed to
recapitalize the FSLIC. As a result, hundreds of insolvent institutions were permitted to continue accumulating losses until the condition of the FSLIC
become so critical that private investors began to
question whether the U.S. government would ultimately honor all the debts accumulated by the
FSLIC. In the end, the actions of private investors
ultimately forced lawmakers to recapitalize the savings and loan industry’ insurance fund.
s
EMy attemptsto recapitalizeth FSLIC In 1985,
a study published by staff members of the FHLBB
concluded that 400 to 500 FSLIC-insured thrifts
were GAAP insolvent and estimated the cost of
resolving those insolvencies at $15.8 billion. The
FSLIC’ official reserves in 1985 were less than
s
$6 billion. The report concluded that closing or
reorganizing even a fraction of the insolvent thrifts
insured by the FSLIC would deplete the insurance
fund’ reserves (Barth, Brumbaugh, Sauerhaft, and
s
Wang 1985).
Later that year, FHLBB Chairman Edwin Gray
acknowledged to Congress that the FSLIC lacked
the funding it needed to deal with its caseload of
problem institutions. To raise the necessary funds,
he proposed imposing a one-time assessment of one
percent on all FSLIC-insured thrifts, as the Bank
Board was authorized to do by law (see American
Banker-,10/17/85). But Gray’ proposal encountered
s
a great deal of resistance from the savings and loan
industry and was subsequently withdrawn. Instead,
the FSLIC exercised its authority to impose a l/8
percent special deposit insurance assessment. The
special assessment generated an additional $1 billion
in 1985, but that amount fell far short of providing
the FSLIC with the funding it needed to continue
operating (Brumbaugh 1988, p. 51).
Alarmed by the Bank Boards bleak assessment
of financial condition of the thrift industry and its
insurance fund, Congress asked the General Accounting Office to prepare a report on industry
FEDERAL

RESERVE

conditions and the implications for the FSLIC fund.
The GAO report, released in February of 1986, concluded that the cost of closing insolvent FSLICinsured thrifts in operation at the time could be as
high as $22.5 billion, an amount well in excess of
the FSLIC’ reserves (see GAO, PotentialDemands
s
on the FSLIC Fand, 1986). In a subsequent Congressional hearing, a GAO official concluded that most
of the insolvent thrifts being “warehoused” by the
FSLIC were unlikely to ever recover. He went on
to estimate that it could take anywhere from 5 to
20 years to work out the problems of insolvent thrifts
(see WashingtonFinancial Reports, 311 O/86).
Obstaclescon$wztingrecapitaliwtion Bank Board
and U.S. Treasury officials had begun meeting in late
1985 to devise a recapitalization plan for the FSLIC.
FHLBB Chairman Edwin Gray unveiled the Reagan
administration’ plan in March of 1986. The stated
s
goal of the plan was to effect a recapitalization of the
FSLIC without taxpayer funding. The plan relied on
a transfer of resources from the Federal Home Loan
Banks and a continuation of the special deposit insurance assessment against thrifts as part of an
elaborate arrangement devised to keep funding costs
off the government budget.’
Enactment of the recapitalization measure was
delayed for over a year, however, because it encountered a great deal of opposition from the thrift
industry. There were two reasons for this opposition.
The first was the plan’ reliance on an indefinite cons
tinuation of the annual l/8 percent special deposit
insurance assessment. Thrift industry spokesmen
maintained that the plan’ reliance on a continuation
s
of the special deposit insurance assessment to service such a debt load placed an unfair burden on the
solvent institutions. Industry representatives argued
further that the proposed $15 billion funding authority
would give the FSLIC much more than it needed
to deal with its caseload of troubled institutions.
7 The plan called for the creation of a shell funding corporation
that would issue bonds to fund the FSLIC. The funding corporation was to be capitalized through the transfer of a portion
of the excess capital of the Federal Home Loan Banks. The
initial capitalization was to be used to purchase zero-coupon
Treasury bonds. These bonds were to provide collateral securing the repayment of the bond principal. Interest payments on
the bonds were to be serviced by revenues to be generated by
continuing the special deposit insurance assessmen? imposed on
FSLIC-insured thrifts. This complicated funding scheme was
chosen because it avoided the direct appropriation of federal
funds and so permitted the cost of the plan to be kept off the,
government’ budget. The plan provided the basic framework
s
behind the Competitive Equality Banking Act of 1987, which
established the Financing Corporation (FICO) to issue off-budget
debt obligations. See Brumbaugh (1988, ch. 3) for more details.
BANK

OF RICHMOND

15

As an alternative to the administration-sponsored
initiative, industry representatives proposed a plan
that would require less borrowing by delaying the
reorganization of some insolvent thrifts for ten years
or more. They also lobbied for a formal timetable
for the phaseout of the special deposit insurance
assessment
(see Washington Financial Reports,
7128186).

A second objection to the recapitalization plan
stemmed from the prospect of an end to policies of
regulatory forbearance. Regulatory forbearance had
become a politically popular policy. So many thrifts
had become financially troubled by this time that the
group constituted a powerful special interest lobby.
The majority of thrift insolvencies were concentrated
in geographic areas experiencing severe regional
economic problems. Congressional representatives
from economically depressed areas argued that closing or reorganizing the financially troubled institutions in their districts would further exacerbate
economic problems in those regions (Brumbaugh
1988, p. 174). Attempts by some lawmakers to link
recapitalization
of the FSLIC with a broader
regulatory reform proposal further slowed down action on the measure.
FSLIC declared insolvent As debate over the
recapitalization measure dragged on into 1987, the
FSLIC’ need for funding began to grow critical.
s
Insolvent thrifts in Texas and the Southwest, where
most problem institutions were concentrated, were
forced to pay rising premiums over market rates in
an effort to attract deposits (Brumbaugh 1988, pp.
70-74; Hirschhorn 1990). As public concernover the
FSLIC’ financial condition grew, the risk premiums
s
paid by insolvent institutions rose significantly
(Hirschhorn 1989a, 1989b).
In an effort to find an alternative funding source,
the Bank Board had turned to the Federal Home
Loan Banks. The FHL banks typically extended advances to member institutions under the security of
certain collateral, most often home mortgages. Insolvent thrifts experiencing the greatest difficulty
attracting deposits could not easily expand their borrowing from the FHL banks, however, because they
could not post the necessary collateral. The Federal
Home Loan Bank System had been established to
provide a source of funding for home mortgages, not
to supply capital to insolvent thrifts. To facilitate
lending to insolvent S&Ls, the Bank Board authorized the FHL banks to extend advances secured by
promissory notesissued by the FSLIC (see GAO,
16

ECONOMIC

REVIEW,

Forbearancefor TroElbl’ Institutions,
ed
May 1987; and
The Management ConsignmentPn+yam, September
1987). By the end of 1986, the Dallas FHLB had
issued over $1 billion in advances to insolvent thrifts
secured only by FSLIC notes.
Early in 1987, the GAO announced that the
FSLIC had become officially insolvent, with its deficit
estimated to exceed $3 billion at the end of 1986
(see “Statement of Frederick D. Wolf” in U.S.
Congress, House, March 1987; and WaliStmetJournal, 3/4/87). The announcement by the GAO raised concerns over the creditworthiness of the FSLIC’
s
promissory notes. A few days after the GAO’ public
s
statement, the accounting firm of Delloite, Haskins
and Sells, which had been hired to audit the financial statements of the Federal Home Loan Banks,
threatened to issue a qualified opinion on the financial condition of the Federal Home Loan Bank of
Dallas.
The $1 billion the Dallas FHLB had advanced
solely on the security of FSLIC notes constituted
a significant fraction of the bank’ capital. Based on
s
the GAO audit of the FSLIC, the Dallas bank’
s
auditor had concluded that the fund might be unable
to back the guarantees securing the bank’ advances
s
to insolvent thrifts. A qualified auditor’ opinion would
s
have made it virtually impossible for the Dallas bank
to raise funds in private capital markets. The FSLICs
mounting financial problems had come to threaten
the financial stability of the entire Federal Home
Loan Bank System.
To avoid receiving a qualified opinion, the Dallas
FHLB demanded immediate repayment of the $1
billion in FSLIC notes it was holding. Fearing that
a qualified opinion on the condition of the Dallas bank
could cast doubt on the creditworthiness
of the
entire FHLB system, the Bank Board quickly
acceded to the Dallas FHLB’ demand and instructed
s
the FSLIC to repay the notes it had issued.8 Repayment of the notes left the FSLIC with less than $1
billion in cash reserves.
During this period, the Dallas FHLB instituted a
program to secure an alternative funding source for
a The Federal Home Loan Bank System funds the advances
it extends to member institutions through the sale of bonds to
private investors. Obligations issued by the Federal Home Loan
Bank System are the joint liability of all twelve Federal Home
Loan Banks. Moreover, the Dallas FHLB was not the only bank
in the system that had lent against the security of FSLIC notes;
it just had a relatively greater exposure to loss in the event of
a default by the FSLIC (see American Banks, 3/16/87).
MARCH/APRIL

1990

insolvent thrifts. Relatively healthy thrifts that could
still attract deposits were induced to place insured
deposits with insolvent thrifts experiencing funding
problems (see Wail StreetJourxai, 3/Z/87). But this
program by itself failed to provide sufficient funding
for insolvent thrifts. In June, the outflow of deposits
from troubled Texas thrifts began to accelerate.
Officials of the Dallas FHLB were forced to negotiate
with deposit brokers in an effort to ensure that
troubled thrifts in that district could continue to raise
funds’ through brokered deposits (see American
Banker, 611 l/87). It was only a few years earlier that
the Bank Board had attempted to curb insured thrifts’
reliance on deposit brokers. Ironically, the agency
found itself relying on the same brokers to continue
funding the problem institutions it was struggling to
keep open while waiting for the enactment of a
recapitalization measure.
Th Competitive
Equa&y Banking Act of 198 7 For
a time, it appeared that opponents of the recapitalization bill would be successful in limiting the amount
of funding approved by Congress to $5 billion, an
amount the GAO had concluded would be insufficient to deal with the magnitude of losses accumulated by insolvent thrifts (see GAO, The
TreasrylFeahal HomeLoan BankBoard Plbn& FSLIC
Recapitalization, March 1987). However, revelations of large-scale fraud at a number of financially
troubled thrifts that had been kept open through
regulatory forbearance created pressure to enact a
larger recapitalization measure. The Competitive
Equality Banking Act (CEBA), enacted in the summer of 1987, authorized the issue of $10.8 billion
in bonds to recapitalize the FSLIC. The bill also
included language mandating the extension of forbearance to financially troubled thrifts operating in
certain designated economically depressed areas of
the country.
Legal status of FSLIC Notes questioned Within
months of the passage of the recapitalization bill,
articles discussing the ultimate necessity of a
taxpayer-funded bailout of the FSLIC began appearing in the financial press (see American Banker,
1 l/ 18/87). In November, the American Institute of
Certified Public Accountants (AICPA) issued its Practice Buletin 3 warning auditors to consider the risks
associated with any FSLIC notes appearing on the
balance sheets of thrifts because of the insurer’ quess
tionable financial condition. A provision pledging the
full faith and credit of the U.S. government behind
all federally insured deposits had been included in
the CEBA. But whether this pledge extended to
FEDERAL

RESERVE

promissory notes issued by the FSLIC to private
investors was uncertain. As the ensuing events show,
the AICPA warning marked an important turning
point in the unfolding crisis. By limiting the FSLIC’
s
ability to continue issuing debt, the AICPA bulletin
helped to precipitate a funding crisis that ultimately
forced lawmakers to recapitalize the savings and loan
industry’ insurance fund.
s
In April of 1988, the Federal Home Loan Bank
of Dallas was forced to issue its 1987 annual report
without an auditor’ opinion. Since its last audit, the
s
Dallas bank had once again begun lending on the
security of FSLIC notes and had $500 million in such
advances outstanding. Its accounting firm, Delloite,
Haskins and Sells, withheld its opinion on the bank’
s
financial condition pending the release of the GAO’
s
audit of the FSLIC (see BNA? Banking Report,
4/18/88).
When the FSLIC released its preliminary 1987
annual report a week later, it acknowledged that
despite the additional funding the agency had received in 1987, it was still insolvent at the end of
the year. According to Bank Board officials, the extent of the FSLIC’ insolvency had almost doubled,
s
to $11.6 billion, during 1987 (see BNA’ Banking
S
Report, 4/25/88).
Based on its audit, the GAO concluded that the
FSLIC had understated the extent of its insolvency.
The government’ auditors projected the cost of
s
resolving the FSLIC’ existing caseload of insolvent
s
thrifts would be in excess of $17 billion, leaving the
agency with a deficit of $13.7 billion at the end of
1987. The GAO report went on to warn of the costs
of dealing with the more than 300 insolvent thrifts
that the FSLIC had yet to formally place under
receivership, which it cautioned could reach as high
as $19 billion. Based on these cost projections, a
GAO spokesman concluded that “further congressional action, beyond that already taken under the
Competitive
Equality Banking Act of 1987 to
recapitalize the Corporation [FSLIC], may well be
needed to enable the Corporation to continue to meet
its obligations (see U.S. Congress, Senate, May
1988). Later that year, the GAO would acknowledge
that its earlier estimates had grossly underestimated
the extent of the FSLIC’ insolvency.
s
In July, the accounting firm of Delloite, Haskins
and Sells finally released an unqualified opinion on
the financial condition of the Dallas FHLB. However,
its report voiced concerns over the ultimate collectibility of the FSLIC notes the bank held as collateral
BANK

OF RICHMOND

17

for its advances to insolvent thrifts, and warned the
bank to limit such advances in the future (see
American Banker, 7129188).
Although the Dallas FHLB had received an unqualified opinion on its financial condition, there was
still considerable concern over the ultimate creditworthiness of the FSLIC’ promissory notes. The
s
Bank Board had announced an ambitious plan to
reorganize and sell a record number of insolvent
thrifts during 1988, but the plan depended on the
willingness of private investors to accept FSLIC
promissory notes and other financial guarantees. But
news of the FSLIC’ deteriorating financial condis
tion made buyers increasingly reluctant to accept the
fund’ notes. Because the AICPA had warned
s
auditors to consider the ultimate collectibility of
FSLIC notes as questionable, potential acquirers
faced the risk that auditors would not grant the
institution an unqualified opinion if its.balance sheet
included FSLIC notes among its assets. Securities
and Exchange Commission regulations made it virtually impossible for a firm that received a qualified
auditor’ statement to sell its securities to investors.
s
To facilitate the issue of more FSLIC notes,
FHLBB Chairman Wall asked the U.S. Congress to
pass a resolution placing the full faith and credit of
the U.S. government behind notes issued by the
FSLIC. The Senate voted in favor of such a measure
in August, but the proposal encountered resistance
in the House of Representatives.
At issue was the question of whether the issuance
of notes and financial guarantees by the FSLIC constituted unauthorized borrowing in excess of the
amount the FSLIC was legally permitted to borrow
under the CEBA. Confidence in the Bank Board had
been undermined by the fact that the agency kept
revising its estimates of the ultimate cost of resolving insured thrift insolvencies. Between the start of
the year and July of 1988, the Bank Board revised
its estimates of the cost of resolving thrift insolvencies on at least three separate occasions, almost
doubling its projected costs from $22.7 to $42.5
billion. Some members of Congress felt that the Bank
Board had not been forthcoming with details of its
planned expenditures. Rep. John LaFalce clearly
summarized the issues surrounding the debate over
granting FSLIC notes full faith and credit status: “We
are now in a position where the Bank Board has, in
effect, issued at its whim unlimited Treasury debt
at levels in excess of its FICO bond authority which
the Congress is now being pressured to belatedly
18

ECONOMIC

REVIEW,

guarantee in order to keep the FSLIC and the industry afloat (see BNA’ Bank-ing
s
Report, 8115/88).”
The prospects for a favorable vote on the resolution requested by the Bank Board were, therefore,
dubious at best. In an apparent effort to avoid an
explicit rejection of the full faith and credit resolution by Congress, Chairman Wall announced he had
withdrawn his request for a vote on the resolution
on September 8. Although Reagan administration
officials had supported enactment of the resolution
at first, Treasury Department
officials later announced that the request was withdrawn because they
had determined that notes issued by the FSLIC
already enjoyed U.S. government backing (see BNA’
s
Banking Report, 9119188).
The GAO publicly supported the Treasury Department’
s position
(see BNA’ Banking Report,
s
1 l/l llSS[Z]). But the AICPA was not satisfied by
these pronouncements.
The organization told the
Bank Board that in the absence of a congressional
resolution, it would require an opinion by the U.S.
Attorney General on the legal status of FSLIC notes
before it would reconsider its warning to auditors on
the status of FSLIC notes. At first the Bank Board
agreed to ask the Attorney General to issue an
s
opinion (see BNA’ Banking Report, 9119188 and
1117188). However, in November a Bank Board
spokesman announced that FHLB Chairman Wall
had decided not to seek the Attorney General’
s
opinion after all. Instead, legislation clarifying the legal
status of FSLIC notes would be sought from the
10 1st Congress when it convened the following year
(see BNA’ Banking Report, 11111/88[ 11).
S
By this time the FSLIC’ situation had become
s
desperate. The 1987 recapitalization measure had
failed to provide enough funding and the 100th Congress had refused to authorize the issue of more
promissory notes. The Bank Board had estimated
it could not service more than $16 billion in notes
and guarantees (see American Banker, 9119189). But
by November the agency had committed itself to
nearly $25 billion in obligations, which included
various financial guarantees to purchasers of insolvent thrifts as well as promissory notes (BNA’Banking
s
Report, 1 l/l 1/88[3]). For almost a decade the Bank
Board had struggled to keep insolvent thrifts open
in an effort to forestall the need to close those institutions and pay off insured depositors. Insulated
from the discipline that the market normally places
on risk-taking, many of those institutions had embarked upon questionable and risky investments that
MARCH/APRIL

1990

had produced staggering losses. Now a default by
the FSLIC appeared imminent. Private investors
would no longer accept the insurance fund’ promises
s
and financial guarantees, but insisted on firm evidence
that it would be given the resources to meet those
obligations. In the end, it was the discipline imposed by private investors that finally forced action
to restore the thrift industry’ insurance fund to
s
solvency.

The Financial Institutions Reform,
Recovery, and Enforcement Act of 1989
Projected costs of dealing with the growing backlog
of hopelessly insolvent thrifts continued to climb
throughout 1988. By the end of the year, the GAO
had raised its estimate to over $I100 billion (see BNA’
s
Banking Report, 12119188). But the 100th Congress
had adjourned without providing additional funding
for the FSLIC, and so one of the first problems facing the incoming Bush Administration was that of
devising a plan to rescue the insurance fund from an
impending default.
The Bush Administration unveiled its plan to deal
with the burgeoning crisis in the savings and loan industry on February 6, 1989. In addition to asking
Congress to authorize funding to recapitalize the
FSLIC, the Bush Plan also mandated a complete
reorganization
of the federal savings and loan
regulatory system. The FDIC was called upon to
assume supervisory control of insolvent- savings and
loans until the proposed legislation was ratified by
Congress (see BNA’ Banking Report, 2113189). The
s
Bush Plan became the model for the Financial Institutions Reform, Recovery and Enforcement Act,
or FIRREA, enacted in August of 1989.
The new savings and loan regulatory system
created by the act is noteworthy in at least two
respects. First, FIRREA represents an effort to reregulate savings and loans by restricting their investment powers and requiring them to specialize more
in mortgage lending. It also calls for an end to the
capital forbearance policies instituted in the 198Os,
requiring savings and loans to meet capital requirements at least as stringent as those imposed on
commercial banks. The new regulations are to be
enforced through enhanced supervisory controls and
stricter penalties in cases involving fraudulent or
criminal activities.
Second, FIRREA brought about a complete
reorganization
of the federal savings and loan
regulatory agencies. The law dissolved the FSLIC
FEDERAL

RESERVE

and established a new deposit insurance fund, the
Savings Association Insurance Fund, or “SAIF,”
under the auspices of the FDIC. It created a new
agency, the Resolution Trust Corporation (RTC),
to take control of the FSLIC’ caseload of insolvent
s
savings and loans. FIRREA also disbanded the
Federal Home Loan Bank Board, replacing it with
a new federal chartering agency under the direction
of the Secretary of the Treasury, known as the
Office of Thrift Supervision, or OTS. The goal
behind this restructuring was to eliminate perceived
conflicts of interest inherent in the old system,
whereby the chartering agency was also responsible
for administering the deposit insurance fund. As the
history of the savings and loan crisis revealed, that
organizational structure created a situation where the
chartering agency had both the incentive and the
means to delay resolution of the problem for a protracted period.
Unlike earlier attempts to resolve the financial
difficulties facing the savings and loan industry, the
enactment of FIRREA was accompanied by a
recognition that government
funding would be
needed to resolve the crisis. In addition to allocating
funds to pay off the obligations incurred by the
FSLIC before its dissolution, the RTC is to receive
$50 billion in additional funding.9 The law also
imposed higher deposit insurance assessments for
commercial banks as well as thrifts to raise the
reserves of each industry’ deposit insurance fund.
s

III.
DEPOSITINSURANCE REFORM
With the demise of the FSLIC, government
regulators have been left to deal with a backlog of
almost 600 insolvent savings and loans. Estimates
of the ultimate cost of resolving the remaining thrift
9 In addition to providing for continued funding of FSLIC obligations incurred prior to the dissolution of the fund, FIRREA
authorized the RTC to borrow $50 billion to use in dealing with
insolvent thrifts. A new funding agency, the Resolution Funding Corporation (REFCORP), was created to borrow $30 billion.
Like the funding corporation used to borrow the funds allocated
by the CEBA, REFCORP was capitalized by a transfer of surplus
capital from the Federal Home Loan Banks and was created to
minimize the impact of the deposit insurance rescue plan on the
government’ budget deficit. Because the Federal Home Loan
s
Banks provided the funding to guarantee repayment of the princioal. funds borrowed bv REFCORP are not officially classified
asUS. Treasury debt. The Treasury was authorized to borrow
the remaining $20 billion and to transfer the proceeds to the
RTC. To the extent that deposit insurance assessments levied
against savings and loans fail short of the amount needed to
service REFCORP debt. the Treasurv bears resoonsibilitv for
providing the funds needed to maintain the inteiest payments
on such debt.
BANK

OF RICHMOND

19

insolvencies have continued to rise since the enactment of FIRREA. The crisis created by the collapse
of the savings and loan industry’ insurance fund sugs
gests that the deposit insurance system is in need
of reform. In this section we critically analyze alternatives for regulatory and deposit insurance reform,
beginning with the reforms put in place by FIRREA.

A Critical Review of FJRREA
FIRREA represents the most sweeping financial
regulatory legislation enacted since the Great Depression. It not only created a new deposit insurance fund,
but completely restructured the savings and loan
regulatory system established in the 1930s. FIRREA
also marks at least a temporary halt in a trend toward
financial deregulation evident in legislation enacted
earlier in the decade.
The new law’ emphasis on stricter regulation and
s
enhanced supervision represents an attempt to limit
potential future losses stemming from bank and thrift
insolvencies, but such measures address the symptoms of the present crisis rather than its causes. The
financial problems that beset savings and loan institutions earlier in the decade were rooted in restrictive
regulations that prohibited thrifts from diversifying
their investments, making them vulnerable to interest
rate risk. While it is prudent to limit the investment
powers of insolvent institutions until they can be
reorganized, the events of the last decade give cause
to question whether such a regulatory structure can
assure a financially sound and profitable industry over
the longer run.
Recently, some analysts have begun to question
whether depository institutions limited to investing
predominantly in residential mortgages can remain
viable. Brumbaugh and Carron (1989), for example,
argue that recent changes in the financial markets
have made funding mortgage lending less profitable
for insured deposit-taking institutions. As the market
for mortgage-backed assets has become more efficient, with investors bypassing financial intermediaries by buying and holding mortgage-backed
securities directly, there appears to be less of a need
for specialized,
deposit-taking
intermediaries
dedicated to warehousing mortgage loans.
To be certain, intermediaries
specializing in
residential housing finance will continue to play an
important role in the U.S. economy. But it now
appears that only a fraction of existing savings and
loans will find it profitable to continue specializing
in mortgage
20

lending.

What

this means

is that the
ECONOMIC

REVIEW.

industry may welI need to contract. Much of that contraction will come about through consolidation. But
the contraction of an industry is often accompanied
by the withdrawal of firms from that industry. If the
new, more restrictive regulatory structure makes it
difficult for insured thrifts to earn profits, the industry
could continue to experience financial difficulties in
the future. Financial intermediaries specializing in
residential lending may prove viable only if affiliated
with larger, diversified financial firms. FIRREA permits commercial banks to acquire financially healthy
thrifts for the first time (in the past, commercial banks
were only permitted to take over failing savings and
loans). And, as Brewer (1989) observes, simply requiring savings and loans to specialize more in mortgage lending will not prevent excessive risk-taking
if that is the goal of an institution’ management.
s

Can “It” Happen Again?
One area of regulation FIRREA did not address
is the mechanism for resolving failures of insured
depositories. New rules specify higher minimum net
worth requirements for savings and loans, but there
is no statutory provision ensuring that insolvent
institutions will be closed more promptly in the future
than they have been in the past. As long as deposits
are fully insured there is no market mechanism to
ensure the prompt closing of insolvent institutions.
In the end, how thrift insolvencies are handled will
still depend on the resources available to the deposit
insurance fund.
An important lesson emeiging.from the savings
and loan crisis is that the deposit insurance funds
themselves can become insolvent. As Barth, Bartholomew, and Bradley (1989a) have noted, the
system as it is presently organized lacks certain
important safeguards that one would expect to be
present in private insurance arrangements. Government-sponsored deposit insurance was not intended
to be self-financing, as privateinsurance arrangements
are, but ultimately relies on government guarantees
to provide depositors with assurances of the safety
of the funds they place with banks. At the same time,
existing laws do not mandate immediate action to
recapitalize the deposit insurance fund if it becomes
insolvent, nor do they specify how the claims against
an insolvent fund are to be resolved. Thus, the conditions that made the present-day crisis in the savings and loan industry possible are still present.

Regulation and Deposit Insurance
The rationale most often given for government
bank regulation centers around the importance of
MARCH/APRIL

1990

promoting the safety and soundness of the banking
system. But much of the existing financial regulatory
system cannot easily be rationalized on these
grounds. A growing body of historical research makes
clear that the existing regulatory structure developed
to address many different public policy goals, with
bank safety and financial stability constituting only
one of those goals.10
When legislation sets out complex rules governing economic relations and market structure, it is
common for government regulatory agencies to be
established to interpret, administer, and enforce those
rules. Because legislation rarely specifies exact
responses to every conceivable set of circumstances,
regulatory agencies typically are granted a certain
amount of discretion in interpreting policy guidelines
and engaging in rulemaking. But when the underlying goals of an agency are vague or seem to conflict, the grant of discretion gives regulatory agencies the power to establish the relative importance
of different policy goals.
With discretionary powers, the incentives facing
regulators become important factors determining the
primary goals of regulation. As Posner (1974) points
out, employees of government agencies have strong
incentives to please their legislative overseers and
to perform competently to increase the value of their
future prospects in the private sector. The incentives
and priorities of lawmakers, in turn, are determined
by political forces.
Because the actions of regulators can and often do
result in a redistribution of economic resources,
regulated firms have a considerable incentive to lobby
for rules that they perceive to be in their own selfinterests. Thus, regulators invariably face political
pressures when setting goals and priorities, though
these pressures are not always explicit.
r” Kareken’ (1986) comprehensive analysis of the present-day
s
system of bank regulation led him to conclude that the system
could not be rationalized by an appeal to concerns over safety
and soundness, especially in the area of regulatory restrictions
on bank branching and geographic expansion. Shull (1983)
produces historical evidence that early laws mandating the
separation of banking and commerce were rooted in concerns
unrelated to safety and soundness issues. Other authors have
concluded that the securities underwriting activities of commercial banks had little to do with the widespread bank failures that
accompanied the Great Depression, and have attributed the
motivation behind the legal separation of commercial and investment banking to factors unrelated to safety and soundness
concerns (Huertas 1984; Flannery 1985; Kaufman 1988; and
Shughart 1987).
FEDERAL

RESERVE

Deposit insurance requires some form of regulation and supervision to contain the incentives for risktaking inherent in the system. Therefore, the issue
of deposit insurance reform cannot be addressed
separately from that of regulatory reform. To address
the issue of regulatory reform, one must first ask
whether the existing regulatory structure imposes
conflicting goals that compromise the ability of
regulators to limit risk-taking by banks and thrifts.
A review of the events leading to the present thrift
crisis reveals that early resolution of the industry’
s
financial problems was hampered by conflicting goals
embedded in the regulatory system. The regulatory
structure imposed on the savings and loan industry
was designed in large part to subsidize credit flows
for residential housing by increasing the supply of
mortgage lending. In addition to being the agency
that chartered federal savings and loans, the Federal
Home Loan Bank Board also bore responsibility for
managing the FSLIC. The Bank Board was also explicitly charged with promoting private home ownership as well as the interests of the savings and loan
industry.
The situation is further complicated by the fact that
state legislatures also have the authority to charter
and regulate insured savings and loans. These
legislatures can gain much of the political benefits
derived from the subsidization of thrifts and local construction interests while allowing the FSLIC to underwrite much of the risk.
Once the crisis began, deposit insurance was
used to keep many insolvent thrifts open in an
attempt to prevent the reallocation of resources from
those institutions and the regions they served. Debate
over how much of the cost of recapitalizing the
FSLIC should be borne by the thrift industry itself
paralyzed action to resolve the crisis for a number
of years. A reluctance on the part of lawmakers to
appropriate the funds needed to close insolvent thrifts
and recapitalize the FSLIC further delayed a resolution of the crisis.
Excessive risk-taking on the part of insolvent thrifts
was tolerated because the regulatory system gave no
one the incentive to take the decisive steps that would
have been necessary to stop it. When hundreds of
savings and loans began to fail, industry regulators
lacked the resources to close those institutions and
pay off depositors, or, for that matter, to adequately
monitor them. At this point, the FSLIC itself was
insolvent and its management began behaving as any
BANK OF RICHMOND

21

other insolvent organization would be expected to
behave. Under the circumstances, the only alternative
the Bank Board had to keeping insolvent institutions
open would have been to impose losses on insured
depositors, an action that was never seriously considered. Legislation enacted during this period,
notably the Garn-St. Germain Act, made it clear that
lawmakers preferred accepting the risks that came
with taking no action against insolvent institutions
to other available alternatives.
The response of the federal regulatory system to
events as they unfolded in the course of the thrift
crisis stand in stark contrast with the way insolvencies are resolved in unregulated private market
arrangements. In periods of financial distress, the
nineteenth century clearinghouses sometimes found
it necessary to suspend payments. But those organizations continued to monitor all members closely and
acted promptly to force banks that exposed other
clearinghouse members to excessive risks out of the
system. Although bank and thrift regulators have the
right to revoke deposit insurance, they rarely exercise this right as a practical matter.
Simply giving regulators more discretionary powers
to deal with failing institutions does not appear to
offer a solution to the problem of limiting losses borne
by the deposit insurance funds. Administrators of the
FSLIC had greater discretionary powers in choosing how to deal with financially troubled savings and
loans than did the FDIC in its dealings with failing
banks. But the historical record shows that the grant
of greater discretionary powers did not ensure that
the losses insolvent institutions were permitted
to impose upon the insurance fund would be contained. As Kaufman (1989, p. 1) notes:
bank regulators . . . avoid taking actions that could put
them in conflict with powerful parties who would experience large dollar losses, such as uninsured depositors or
other creditors, management,
owners, and even large
borrowers. In addition, the regulators frequently believe
that such actions would be an admission of failure not only
of the bank but also of their own agency, which is charged
with bank safety and evaluated by many on its ability to
achieve this condition.

Insulating the economy from the potentially disruptive effects of bank and thrift failures remains an overriding goal of regulators. While it is hard to take issue
with this goal, history shows that when attempts to
minimize disruption are permitted to completely
subvert the normal market forces that would otherwise act to close insolvent institutions, the results
can be disastrous. Unless market participants are
22

ECONOMIC

REVIEW.

forced to internalize some of the risk associated with
their actions, they have no incentive to limit
risk-taking.
That thrift industry regulators were hampered by
conflicting goals that interfered with their ability to
protect the resources of the deposit insurance fund
now seems to be widely acknowledged. Avoiding a
repetition of the current thrift industry crisis depends
on our ability to devise a system that will guarantee
the prompt closure of institutions once they become
insolvent while limiting the potential disruptive
effects of such occurrences.

Lessons from Bankruptcy Law Reform
The present system of bankruptcy laws were
enacted by Congress in 1978 and amended in 1984
and 1986. This legislation instituted sweeping
reforms to the administration of bankruptcy courts
and the system of bankruptcy resolution. Before these
recent reforms, the bankruptcy judge had duties
much broader than those of an impartial referee.
Under the old law, the bankruptcy judge (originally
called the “bankruptcy referee”) was given the role
of administering bankruptcy cases under the general
supervision of the district judge, who held the ultimate legal authority to adjudicate any cases arising
from the bankruptcy proceedings. But over time, the
role and authority of the referee grew until the
“referee” became a bankruptcy judge who exercised
judicial power to decide disputes among different
parties.
Thus, the role of the bankruptcy referee, or administrator, had grown beyond that envisioned by
the laws that created the position. The authors of
the earlier law had envisioned a court-appointed
administrator acting under the oversight of an independent and impartial judicial authority. But oversight and administrative duties had come to be
delegated to a single agent, one who lacked the
insulation from outside influence normally provided
to members of the judiciary. According to Treister,
et al. (1988, pp. 5-7), this dual role came to be
perceived as the “most glaring defect of the former
bankruptcy system.”
Dissatisfaction with this system led Congress to
establish a special Commission on Bankruptcy Laws
of the United States to study, analyze, and recommend changes in the bankruptcy laws in 1970. The
Commission’ findings, published in 1973, noted
s
that:
MARCH/APRIL

1990

making an individual [the bankruptcy judge] responsible
for conduct of both administrative and judicial aspects of
a bankruptcy case is incompatible with the proper performance of the judicial function. Even if a paragon of integrity
were sitting on the bench and could keep his mind and
feelings insulated from influences which arise from his
previous official connections with the case before him and
with one of the parties to it, he probably could not dispel
the appearance of a relationship which might compromise
his judicial objectivity (as cited in Treister, et al. 1988,
p. 7).

One of the principal reforms brought about by the
Bankruptcy Reform Act was to free the bankruptcy
judge from acting in the role of administrator and
enhance his judicial role. Before the Act, bankruptcy “referees” were only appointed to serve
“during good behavior.” The Bankruptcy Reform Act
provided for the appointment of bankruptcy judges
to fixed 14-year terms. Appointments were made by
the president, with the advice and consent of the
Senate. The Act also provided for an independent
system of United States Trustees under the auspices
of the Justice Department for cases where an administrator for the bankruptcy estate needed to be
appointed.”
What lessons do these events hold for deposit
insurance reform? The bankruptcy code explicitly
recognizes the possibility of conflicts of interest
inherent in a system where an agent appointed to
resolve firm insolvencies is given roles that may create
conflicting goals. To avoid such potential conflicts,
bankruptcy law provides for a separation of the different roles, separating the referee, or bankruptcy
judge, from the role of the trustee appointed to administer the estate. The role of the bankruptcy judge
is intentionally limited to mediating disputes among
different parties with claims against the firm.
Banking law gives the deposit insurer the dual role
of receiver and claimant in the event of a bank failure.
The potential for conflicting goals arising from such
a system would appear to greatly exceed those inherent in the old bankruptcy system. Recent bankruptcy law reforms suggest an alternative framework,
one based on judicial oversight that would sharply
limit the discretion of the deposit insurer in dealing
with failing institutions.
rr At first Congress was not convinced that an administrative
apparatus such as this, outside the Judicial Branch, was needed.
Accordingly, a pilot project was established. The United States
Trustee system was made a permanent part of the bankruptcy
system in 1986 (Treister, et al. 1988, pp. 85-91).
FEDERAL

RESERVE

A Role for Enhanced Judicial Oversight
One necessary ingredient for providing successful
deposit insurance is a precommitment to closing insolvent depository institutions promptly.
With
unregulated commercial firms, this precommitment
is achieved through legal bankruptcy proceedings in
which claims against the insolvent firm are resolved
under the auspices of an independent judiciary. This
observation suggests that one way to credibly commit to close failing banks and thrifts would be to
expand the role of the judicial system to make the
resolution of bank and thrift insolvencies subject to
the same kind of judicial oversight that characterizes
regular bankruptcy proceedings.
Posner (1974)
emphasizes that many features of law are designed
to pursue overall efficiency gains. By its very design,
the legal system is more immune to political pressures than government regulators. Using the judiciary
to limit the discretion of regulatory agencies may be
one way of ensuring that the regulatory process is
governed by legislative guidelines.
To ensure that failing banks and thrifts are forced
into legal insolvency proceedings, some depositors
must be put at risk of loss. Otherwise, market participants will have no incentive to force a failing
institution into insolvency proceedings. The distinction between the insured and uninsured depositor
must be restored. As Todd (1988) notes, deposit insurance was never intended to prevent all bank
failures, only to provide for the prompt resolution
of such failures.
Boyd and Rolnick (1988) have forwarded a plan
to administer federal deposit insurance more like
private insurance arrangements by instituting a
system of coinsurance. Under this plan, deposits
would be fully insured up to some amount sufficient
to protect small, unsophisticated depositors. Large
depositors would be subject to some risk of loss,
receiving perhaps 90 or 95 cents for every dollar on
deposit in the event of a bank failure. The advantage of this plan is that it would place known limits
on the maximum extent of losses borne by depositors, while still giving large, sophisticated depositors
the incentive to monitor their banks.
In the event of a bank failure, depositors could be
given prompt access to most of their funds through
a procedure similar to the modified payout procedure
used by the FDIC before the failure of Continental
Illinois National Bank (Benston, et al. 1986, ch. 4).
In a modified payout, uninsured depositors were
given immediate access to’ most of their funds
BANK

OF RICHMOND

23

based on preliminary estimates of expected losses
resulting from the liquidation of the failed bank. But
whereas a modified payout involved liquidation of the
affected institution, regulators could place the failed
institution in a conservatorship
and continue to
operate it until it could be reorganized and returned
to the private sector. Kaufman (1989) argues that insured depositors would have no incentive to withdraw
their funds, and uninsured depositors would have no
incentive to run on the financially troubled institution after it had been “failed” because they could be
assured of no further losses.
Bank failures could be administered under a system
of judicial oversight with such a system. The deposit
insurer would represent one of the claimants against
the firm. In cases where retaining present management is not deemed desirable, a conservator could
be appointed to run the institution. Under such
judicial proceedings, the deposit insurer would be
limited to paying only insured depositors.

Other Alternatives
Simply placing bank failure resolution under a
system of enhanced judicial oversight is unlikely to
provide a panacea for all the problems currently
facing the banking and thrift industries. But it would
bring about an improvement in bank failure resolution methods, and would be consistent with other
reforms now under debate. Two sets of reforms are
noted briefly below.
100 percent reseme banking Some analysts and
policymakers have argued that imposing market
discipline on depositors is not practical because it
would disrupt banking markets. One argument says
that there are too many potential externalities involved with the operation of the payments system
to risk letting a large depository institution fail. If
safety and soundness is truly an overriding policy goal,
then that goal can be achieved by requiring banks
to hold only safe assets. This is the 100 percent
reserve banking proposal, advocated by Mints, and
later, Friedman, and most recently resurrected by
Kareken (1985), and, in a slightly different form,
by Litan (1986). Such a system would truly be safe
because it would remove all private credit risk from
the payments system, substituting instead the credit
of the government, the ultimate guarantor of the
safety of the system. Kareken (1985), Gorton and
Pennachi (1989), and Jacklin (1989) postulate that
this type of banking, which amounts to a money
market mutual fund in short-term safe securities,

24

ECONOMIC

REVIEW.

would be a natural product of free-market competition under current technology and modern financial
market arrangements.
With the institution of “safe banks,” lending activities would be conducted by uninsured affiliates.
Such uninsured affiliates would still face a risk of
insolvency. The proposed insolvency,resolution procedures outlined above could be adopted to deal with
failing lending affiliates.
An enhanced r&e j&- ma&et forces As experience
with the nineteen century clearinghouse system
shows, banks have a natural advantage in monitoring the creditworthiness of other banks. If given the
proper incentives, private monitoring by banking
firms could substantially augment government supervisory efforts. Banks would then be expected to
police themselves as they did prior to the advent of
deposit insurance.
Certain kinds of deregulation
could actually
enhance the safety and soundness of the banking
system and lessen the danger of bank runs. As
Calomiris (1989) points out, nationwide branching
would probably go a long way toward providing
additional safety and soundness. Canadian history is
instructive in this regard, since Canada’ nationwide
s
branching system proved immune to bank runs during the Great Depression. Haubrich (1988) notes that
there were no bank failures in Canada during the
193Os, even though their depression was as severe
as that of the United States. In the event of deregulation, normal application of antitrust laws could ensure that competition in banking markets is preserved. Monetary policy could provide banks with
liquidity in the event of a financial panic leading to
an aggregate change in desired holdings of currency.

IV.

CONCLUSION

This paper has provided a detailed analysis of the
savings and loan crisis. To understand the events and
the needed reforms, we have drawn heavily on the
operation of private market relationships. Like private
bondholders, the deposit insurance agencies bear the
risk associated with bank failures and, therefore, have
an incentive to promptly close or reorganize failing
banks or savings and loans. But as recent events have
clearly demonstrated, the deposit insurance funds
themselves bear some risk of insolvency. As long
as no formal mechanism for dealing with the insolvency of a deposit insurance fund exists, there is some
chance that the crisis that beset the savings and loan
industry could be repeated.
MARCH/APRIL

1990

The history of the crisis suggests that simply giving regulators more discretionary authority will not
be sufficient to guarantee against future insolvencies
of one of the deposit insurance funds. The Bank
Board and the FSLIC had more discretionary power
than the FDIC, yet thrift industry regulators were
not able to prevent the insolvency of the FSLIC.
These considerations, as well as the lessons learned
from looking at the operation of the early clearinghouses, point to a number of key ingredients that
must be present in any successful publicly administered deposit insurance scheme. The clearinghouse
system was successful in maintaining safety and
soundness among banks because the members of the
system had the incentive to enforce minimum net
worth standards. Through the threat of expulsion,
the clearinghouses could discipline members that
failed to meet the conditions of membership.
Of course, administrators of the deposit insurance
funds also have such power in principle, but do not
always have the incentive to exercise it. As Milgrom
and Roberts (1988) point out:
Even if the executive authority is unusually competent,
public spirited, and immune to bribes . . . it may still be
desirable to limit its discretion, for two reasons. First, in
order to provide correct incentives to others in the organization, the authority must be able to make commitments
to act against its own interests in the future, and these
commitments
are not credible unless there are some
effective limits on the centre’ powers. . . . The second
s
reason to limit the discretion of an honest, competent
decision-maker is to discourage rent-seeking behavior by
others who are affected by the centre’ decision. . . . the
s
mere willingness of the centre to consider seriously a
decision with large redistributional
consequences
will
cause other economic agents to waste significant resources
in attempts to influence or block it or to delay its implementation. In public decision-making, for example, enormous resources are spent in proposing legislation or regulations and in advocating or opposing these proposals,
as well as in filing and maneuvering for advantage in
lawsuits.

Deposit insurance as it is presently administered
removes all elements of market discipline from banking markets, making it a political rather than an
economic decision to let an institution fail. With the
potential transfer of such large amounts of resources
at stake, sqme form of breakdown in regulatory
discipline should not be surprising.
In the case of the FSLIC, the agency was forced
to exercise regulatory forbearance because it lacked
the resources it would have needed to close insolvent institutions. Acknowledging the fund’ insols
vency and forcing insured depositors to bear a part
FEDERAL

RESERVE

of the cost was never regarded as an acceptable solution to dealing with the crisis. But lawmakers, while
not wishing to impose losses on insured depositors,
proved reluctant to appropriate the funding needed
to deal with the problem. The strategy chosen was
one of tolerating greater risk-taking on the part of
insured savings and loans, in the hope that the need
for government funding could be obviated.
While recently enacted reforms place limits on the
ability of failing thrifts to take on excessive risks, they
do not change the incentives facing market participants and regulatory agencies, and cannot guarantee
that one of the deposit insurance funds will not
become insolvent in the future. Therefore, the
reforms enacted to date cannot ensure that failing
institutions will always be dealt with promptly in the
future.
Deposit insurance reform should include legislative
guidelines specifying how bank and thrift failures are
resolved and how the insolvency of one of the
deposit insurance funds is to be resolved. A central
conclusion of this paper is that such legislative
guidelines could be enforced through a greater role
for judicial oversight. There may be good reasons
for exempting banks and thrifts from the same
bankruptcy laws applied to unregulated firms, but
increased market discipline and enhanced judicial
oversight of bank failure resolution proceedings could
play a constructive role in deposit insurance reform.

References
Alchian, A.A. “Corporate Management and Property Rights.”
In Henry G. Manne, ed., Economic Policyand the Regularion
of Secutiir& Washington, DC: American Enterprise Institute,
1968.
Atneri~anBanker. “Bank Board Seeks Authority
FSLIC Fund.” October 17, 1985.
“FSLIC Revives Income-Capital
Program:” December 26, 198.5.
“Dallas Home Loan Bank Seeks
Paybackfrom FSLIC.” March 16, 1987.

to Boost its
Certificates
$1 Billion

. “US Regulators Seeking Funds for Texas S&Ls:
Ask Help from Brokers in Gathering Deposits.” June 11,
1987.
. “FSLIC May Need Taxpayer Bailout, Gould
Says: Treasury Official Admits Recapitalization Might Not
Keep Insurer Afloat for Long.” November 18, 1987.
“Dallas Bank Warned on Uncollateralized
Advances: Accounting Firm Opinion Discourages FSLICGuaranteed Loans to Ailing Thrifts.” August 29, 1988.
BANK OF RICHMOND

2.5

September

“FSLIC Near Ceiling on Notes: GAO Aide.”
19, 1989.

American Institute of Certified Public Accountants. Accounting
Standards Executive Committee and Savings and Loan
Associations Committee. Practice Bufietin 3: Prxpaymenrs
Into the Secondary Reserweof the FSLIC and Contingencies
Related to Other Obligationsof the FSLIC. New York:
November 1987.
Barth, James R., Philip F. Bartholomew, and Michael G.
Bradley. “The Determinants of Thrift Institution Resolution Costs.” Office of Thrift Supervision Research Paper
No. 89-03, Department of the Treasury, November 1989b.

“FSLIC Notes Are Backed by Full Faith and
Credit of the U.S., GAO Concludes.” 5 1 (November 11,
1988[2]): 786-7.
. “White Says FSLIC Should Issue Promissory
Notes As Needed to Stop S&L Hemorrhaging.”
51
(November 11, 1988131): 787-8.
. “GAO Says FSLIC Needs $85 Billion More
than FHLBB Estimates Over Next Ten Years.” 51
(December 19, 1988): 1013.
“White House Fact Sheet on President Bush’
s
Proposal’
for Savings and Loan Industry.” Reprinted in 52
(February 13, 1989): 406-7.

. “Reforming Federal Deposit Insurance: What
Can Be Learned From Private Insurance Practices?”
Federal Home Loan Bank Board Research Paper No. 161,
June 1989a.

Boyd, John H., and Arthur J. Rolnick. “A Case for Reforming
Federal Deposit Insurance.” Federal Reserve Bank of
Minneapolis 1988 Annual Report: 3-16.

. “Thrift Deregulation
and Federal Deposit
Insurance.” Federal Home Loan Bank Board Research Paper
No. 150, November 1988.

Brewer, Elijah III. “Full-Blown Crisis, Half-Measure Cure.”
Federal Reserve Bank of Chicago Economic Perspectives
(November/December
1989): 2-17.

Barth, James R., Philip F. Bartholomew, and Carol G.
Labich. “Moral Hazard and the Thrift Crisis: An
Analysis of 1988 Resolutions.” Federal Home Loan Bank
Board Research Paper No. 160, May 1989.
Barth, James R., Dan Brumbaugh,‘
Jr., Daniel Sauerhaft, and
George H.K. Wang. “Insolvency and Risk-Taking in the
Thrift Industry: Implications for the Future.” Contemporary
PolicyIssues(Fall 1985): 1-32.
Barth, James R., John J. Feid, Gabriel Riedel, and M. Hampton Tunis. “Alternative
Federal Deposit ‘
Insurance
Regimes.” Federal Home Loan Bank Board, January 1989.
Barth, James R., and Philip J. Wiest. “Consolidation and
Restructuring of the U.S. Thrift Industry Under the
Financial Institutions Reform, Recovery, and Enforcement
Act.” Office of Thrift Supervision, October 1989.
Benston, George J., Robert A. Eisenbeis, Paul M. Horvitz,
Edward J. Kane, and George G. Kaufman. Perspectives
on
Safe and Sound Banking. Cambridge, Massachusetts: The
MIT Press, 1986.
Black, Fischer, Merton H. Miller, and Richard A. Posner.
“An Approach to the Regulation of Bank Holding Companies.” Journal of Business51 (July 1978): 379-412.
BNA” BankingRepori. “Dallas FHL Bank Releases 1987 Annual
Report That Lacks Auditor’ Opinion.” 50 (April 18, 1988):
s
653.
“1987 FSLIC Audit Paints Bleak Picture, Fund’
s
Insolvency Level Nearly Doubles.” 50 (April 25, 1988):
693-4.

Brumbaugh, R. Dan, Jr. Th$ts Under Siege: Restoringorder to
Ballinger
Amaicon Banking. Cambridge, Massachusetts:
Publishing Company, 1988.
Brumbaugh, R. Dan, Jr., and Andrew S. Carron. “The
Future of Thrifts in the Mortgage Market.” Paper presented at the 1989 summer meeting of the Western Economic Association.
Calomiris, Charles W. “Deposit Insurance: Lessons from the
Record.” Federal Reserve Bank of Chicago Economic
Penpectives 13 (May/June 1989): 10-30.
. “Do ‘
Vulnerable’ Economies Need Deposit
Insurance?: Lessons from the U.S. Agricultural Boom and
Bust of the 1920s.” Processed, October 1989.
Calomiris, Charles W., and Larry Schweikart. “Crisis and
Containment in the Panic of 1857.” Processed, 1989.
Carosso, Vincent P. “The Wall Street Money Trust from Pujo
through Medina.” BuGhess
HktoryReview 47 (Winter 1973):
42 l-37.
Coase, R.H. “The Nature of the Firm.” Ekonomica,New Series,
IV (1937): 331-51.
Diamond, Douglas W., and Philip H. Dybvig. “Bank Runs,
Deposit Insurance, and Liquidity.” Journal of Pokica/
Economy 91 uune 1983): 401-19.
Edwards, Franklin. “Can Regulatory Reform Prevent the
Impending Disaster in Financial Markets?” Federal Reserve
Bank of Kansas City EconomicReview 73 Uanuary 1988):
29-39.

. “Rep. LaFalce Pans Senate Backing of FSLIC
Notes with Full Faith and Credit Pledge.” 51 (August 15,
1988): ‘
264-S.

Fama, Eugene F. “Agency Problems and the Theory of the
Firm.” Th Journal of Po&ical fionomy 88 (April 1980):
288-307.

“GAO Calls for Full Faith and Credit Backing
of All FSLIC Obligations.” 51 (September 19, 1988): 473-4.

Fama, Eugene F., and Michael C. Jensen. “Separation of
Ownership and Control.” Th JournaLof Low and Economics
26 (June 1983): 301-26.

. “Bank Board Chairman Wall to Seek Full
Faith and Credit Legislation.” 51 (November 11, 1988[ 11):
827.
26

ECONOMIC

REVIEW,

“Agency Problems and Residual Claims.” The
Journal of low and Economics26 (June 1983): 327-49.
MARCH/APRIL

1990

Flannery; Mark J. “An Economic Evaluation of Bank Securities
Activities Before 1933.” In Ingo Walter, ed., hpbting
Wall Street: Comme&ai Bank Penetrationof the Corporate
Sea&es Ma&et. New York: John Wiley and Sons, 1985,
pp. 67-87.
Friedman, Milton. A Ptvyam for MonetaryStability.Bronx, NY:
Fordham University Press, 1959.

Jensen, Michael C., and William Meckling. “Theory of the
Firm: Managerial Behavior, Agency Costs and Ownership
Structure.” Th Journal of Finantial Economics3 (October
1976): 30560.
Kane, Edward J. “How Incentive-Incompatible
Deposit Insurance Funds Fail.” NBER Working Paper No. 2836,
February 1989a.
T/reS&L InsaranceMess.Washington,
Urban Institute Press, 1989b.

GAO, see U.S. General Accounting Office.
Goodfriend, Marvin S. “Money, Credit, Banking, and Payment
System Policy.” In David B. Humphrey, ed., Th U.S.
Payment System: E&Gmy, Risk and the Role of the Federal
Reserve. Boston: Kluwer Academic Publishers, 1990.
Goodfriend, Marvin S., and Robert G.’King. “Financial Deregulation, Monetary Policy, and Central Banking.” Federal
Reserve Bank of Richmond Economic Review 74 (May/June
1988): 3-22.
Gorton, Gary. “‘Public Policy and the Evolution of Banking
Markets.” The Wharton School, June 1989.
Gorton, Gary, and Joseph Haubrich. “Bank Regulation, Credit
Markets, and the Control of Capital.” Canregie-Rwh&ererS
on Public Policy 26 (1987): 289-333.
Gorton, Gary, and Donald J. Mullineaux. “The Joint Production of Confidence: Endogenous Regulation and Nineteenth
Century Commercial-Bank
Clearinghouses.” Journal of
Money, Credit and Banking 19 (November 1987): 457-68.
Gorton, Gary, and George Pennacchi. “Transaction Contracts.”
The Wharton School, September 1988.
Haubrich, Joseph. “Nonmonetary Effects of Financial Crisis:
Lessons from the Great Depression in Canada.” Processed, June 1989.
Hirschhorn, Eric. “Depositor Risk Perceptions and the Insolvency of the FSLIC.” Office of Thrift Supervision, Department of the Treasury, December 1989a.

D.C.: The

Kareken, John H. “Ensuring Financial Stability.” In T/re
Search for Financial Stability: Th Past F@y Yean. Federal
Reserve Bank of San Francisco, 1985, pp. 53-77.
. “Federal Bank Regulatory Policy: A Description
and Some Observations.” Jbuma~ of Business59 (January
1986): 3-47.
Kaufman, George G. “Are Some Banks Too Large to Fail?’
Federal Reserve Bank of Chicago Working Paper WP-89-14, .‘
,:
August 1989.
Kuprianov, Anatoli. “The Monetary Control Act and the Role
of the Federal Reserve in the Interbank Clearing Market.”
Federal Reserve Bank of Richmond Economic Review 71
(July/August 1985): 23-35.
Litan, Robert E. “Taking the Dangers Out of Bank Deregulation.” The BmokingsRtiew (Fall 1986): 3-12.
Milgrom, Paul, and John Roberts. “Economic Theories of the
Firm: Past, Present, and Future.” Canadian Jbumal of
Economics21 (August 1988): 444-58.
Posner, Richard A. “Theories of Economic Regulation.” Th
and ManagementSciences5 (Autumn
BellJournal of Economics
1974): 335-58.
“Taxation by Regulation.” Th Bell Journal of
EconomicsandManagement science 2 (Spring 1971): 22-50.

. “The Determinants of Interest Rates on Certificates of Deposit.” Office of Thrift Supervision, Department
of the Treasury, December 1989b.

Shughart, William F. II. “A Public Choice Perspective of the
1933 Banking Act.” In Catherine England and Thomas
Huertas, eds., Tire Financial Services Rewlution: PO&V
Directionsfw the Future. Boston, MA: Kluwer Academic
Publishers, 1988.

. “Interest Rates on Thrift Certificates of Deposit.”
Office of Thrift Supervision, Department of the Treasury,
February 1990.

Shull, Bernard. “The Separation of Banking and Commerce:
Origin, Development, and Implications for Antitrust.” Tke
AntitrustBulletin (Spring 1983): 25579.

Hirschhorn, Eric, and David Zervos. “Policies to Change the
Priority of Claimants: The Case of Depositor Preference
Laws.” JoumaL of Financial Sties
Reseaxh, forthcoming,
1990.
Huertas, Thomas F. “An Economic Brief Against GlassSteagall.” Journal of Bank Research 1.5 (Autumn 1984):
148-59.
Jacklin, Charles J. “Demand Equity and Deposit Insurance.”
Research Paper No. 1062, Graduate School of Business,
Stanford University, September 1989.
. “Demand Deposits, Trading Restrictions, and
Risk Sharing.” In Edward C. Prescott and Neil Wallace,
eds., Contractual Amangements for Intet?empora/ Trade.
Minneapolis: University of Minnesota, 1989.
FEDERAL

RESERVE

Stigler, George. “The Theory of Economic Regulation.” The
Bell Journal of Economicsand ManagementScience 2 (Spring
1971): 3-21.
Stiglitz, Joseph E. “Credit Markets and the Control of Capital.”
.ibuma~ofMonq, Cr&itand.Banking 17 (May 1985): 133-52.
Timberlake, Richard H., Jr. “The Central Bank Role of Clearinghouse Associations.” Journal of Money, Credit and
Banking 16 (February 1984): 1-15.
Todd, Walker F. “Aggressive Uses of Chapter 11 of the Federal
Bankruptcy Code.” Federal Reserve Bank of Cleveland
EconomicReview (Quarter 3 1986): 20-26.
. “Lessons of the Past and Prospects for the
Future in Lender of Last Resort Theory.” Federal Reserve
Bank of Cleveland Working Paper No. 8805, August 1988.
BANK

OF RICHMOND

27

Treister, George M., J. Ronald Trost, Leon S. Forman,
Kenneth N. Klee, Richard B. Levin. Fundamental of
Bankuptcy Low. Philadelphia, PA: American Law Institute,
1988.
U.S. Congress. House. Committee on Banking, Finance and
Urban Affairs. Statement of Frederick D. Wolf, Director,
Accounting and Financial Management Division, GAO, in
Federal Sawtzgsand Loan Insurance CorporationRecapitali
zation Act of 1987. Hearings BeJbre the Subcomminee on
Financial Institutions
Supemkion, Regzkztion,and Inwance.
100th Cong., 1st sess., March 3 and 4, 1987.
U.S.

Congress. Senate. Statement of Frederick D. Wolf,
Director, Accounting and Financial Management Division,
GAO, in tight
on the Conditir-wof the Finan&
Semkes Indwny. Hearings Before the Committee Banking,
on
Housing, and Urban Affaiin. 100th Cong., 2nd sess.,
May 18, 25, and 26, 1988.

U.S. General Accounting Office. Net Worth Cert@ate A&vane
Pmgrams: T/reirDesign, Major LXJktues, and IGarb Implementation. Washington: November 5, 1984.
Thnijr Industry Restructuringand the Net Worth
Cen#cat~ Program. Washington: September 24, 1985.

. Thnji Indkstry Th Tma.wy/Federa~ Home Loan
Bank Board Plan fw FSLIC Recapitaliaotion.Washington:
March 1987.
. VmjftIndurtry:Forbearance&r TroubledInstituons,
f 982-1986. Washington: May 1987.
Thriji Industry: Th Management Cons&ment
Pmgram.‘
Washington:
September 1987.
Wall Street humal. “Southwest Thrift Regulators Plan to Put
Funds from Strong S&Ls into Weak Ones.” March 2, 1987.
“FSLIC Technically Bankrupt at the End of 86,
Preliminary Audit by GAO Finds.” March 4, 1987.
Wdtagton Fkancia~Reports.“GAO Says Insolvent S&Ls Beyond
Hope, Dismisses Regulatory Moves.” 46 (March 10, 1986):
373-4.
Industryis
116-8.

Zisman, Barry S., and William 0. Churchill. “Federal Assistance
to Failing and Failed Thrifts.” Banhing Eqansion Repotter
8, no. 16 (August 21, 1989): 1, 9-14.

Thn$? Industy Problems: Potential Lkmands on
the FSLIk Fund. Washington: February 1986.

28

ECONOMIC

REVIEW,

“Treasury Circulates a Memo Claiming S&L
Subverting FSLIC Plan.” 47 (July 28, 1986):

MARCH/APRIL

1990

Takeovers

and Stock Price Volatility
Je&

M. Lmkm’ and John A. Weinberg’*

I.
INTRODUCTION
There is now a large literature documenting the
statistical relation between stock prices and dividends
at the aggregate level. A robust finding is that stock
prices are too volatile to be explained by subsequent
changes in dividends. Observations of large market
swings, like the crash of October 1987 and the minicrash of October 1989, encourage the popular
perception that stock prices are excessively volatile.
While these observations have provoked a great deal
of analysis, there has been little discussion of the
possible link between excess stock price volatility and
the fact that changes in the control of large corporations often take place via market acquisition of the
These
transactionsoutstanding
shares.
takeovers-are
often associated with dramatic increases in the price of the shares of the firm being
acquired; these are called “takeover premia.” In fact,
some commentators argue that movements in the
stock market in the 198Os, including the large market
declines of October 1987 and October 1989, were
linked to changes in takeover activity. In this article
we explore the possible link between takeover activity and stock price volatility.
The explanation we propose relies on recent advances in our understanding of imperfections in the
monitoring of firm managers. These imperfections
imply that there is a “value of control” (we make this
term more precise below) that is appropriable by the
managers of a large corporation. This private value
of control arises out of the delegation of decisionmaking authority that is intrinsic to the separation
of ownership and control in the modern corporation.
The value of control explains, in part, the premium
often paid to shareholders to acquire control of a firm.

We will argue that the value of control, along with
the probability that someone will be willing to pay
it, can vary independently of the expected present
value of dividends. This adds an independent source
of variation to the price of the traded shares of
publicly held corporations.
The plan of the paper is to first describe the
Martingale Model of stock prices, often referred to
as the “efficient markets theory.” This serves as a
benchmark, both for the excess volatility findings and
for the alternative model we propose. We then survey
some of the key empirical regularities concerning
stock prices: these include the excess volatility
finding in time series of aggregate stock prices, as
well as the behavior of individual stock prices before
and after control change transactions.
We then proceed to outline the essential elements
of our model of the link between takeovers and stock
prices. First, we describe imperfections in the relationship between a large firm’ managers and the peos
ple who hold claims issued by the firm. Next, we
describe the implications of these imperfections for
some of the characteristics of the claims issued by
the firm-specifically,
the legal control mechanism
associated with them. We argue that traded shares
are bundled claims giving the holder the right to help
determine the control of the firm as well as a claim
on a stream of dividends. We then show how such
shares can display excess volatility because of variations in the expected future value of the control right
embedded in the claim. The final section describes
some of the implications of these insights for policy
and for economic theory. The appendix provides a
more rigorous derivation of our model of excess
volatility.

II.
THEMARTINGALEMODEL
OFSTOCKPRICES

’ Research Department, Federal Reserve Bank of Richmond,
P.O. Box 27622, Richmond, VA 23261.
Department of Economics, Krannert Graduate School of
Management, Purdue University, West Lafayette, IN 47907.
l

l

The authors would like to thank Tim Cook, Tom Humphrey,
Tony Kuprianov, and Stacey Schreft for very helpful comments
on an earlier draft. The views expressed are the authors’ and
do not necessarily reflect the views of the Federal Reserve Bank
of Richmond or the Federal Reserve System.
FEDERAL

RESERVE

As a benchmark, consider a simple but general
model of the determination
of stock prices, the
Martingale Model. The empirical findings of excess
volatility that we describe below are essentially
BANK

OF RICHMOND

29

contradictions between the properties of the Martingale Model and those of actual stock market data;
stock prices are more volatile relative to dividends
than is predicted by the Martingale Model. When
we describe an alternative explanation for stock price
volatility, a comparison of the predictions of the alternative model with those of the Martingale Model will
be useful.
According to the Martingale Model-often
referred to as the “efficient, markets theory” or the
“expected present value relation”-the price of a given
stock at any given time is equal to the expected
present value of the stream of future dividends that
will accrue on that stock’ To be more explicit, let
.
pt be the price of a share of stock at time t (after
payment of dividends .due at time t); let dt.+ s be the
amount paid in dividends paid at time t +s, where
the index s takes on the values 1,2,3, . . . . We
abstract fromlinflation, and so we assume ihat dt + s
is the real ‘
value of dividends at time t $ s. We also
abstract frsm stock splits or repurchases, and so the
sequence of dividends,’ dt +.s for s ; 1,2,3 . . .,
captures the total return to an investor who purchases
the share at time ; and holds it to eternity. Note that
from the point of view of an investor at the current
date, the future stream of dividends is a sequence
of random variibles:
The Martingale Model asserts that there is a con;
stant rate r, where r :> 0, at which future elipected
returns are discounted back to the present,‘
and that
the current price is related to next period’ price and
s
next period’ dividend by the ‘
s
equation‘

(1) pt = (1 + r)-‘
EMpt+l + c&+11,
where Et[wt +‘ is notation for .the expected value
J
of a random variable wt+s, with the expectation
taken using only information available’ at period t.
Equation (1) states, that the current price of a Stock
equals the expected value of the sum sf next pe6o$s
price and dividends, discounted back to the present
at rate r.z
Equation (1) can be used to derive an equation
relating the current stock price to the entire stream
I Stephen F. Leroy (1989) calls this theory the Martingale
Model. His article also contains an excellent deskriotion. historv.
and survey of empirical tests of the theory’ This section follows
.
his exposition.
1

,

~,I

* A martingale is any randoni series {wt} that always satisfies
wt = &[wt+ I]. The model is called the Martingale Model
because there is a simple variable that is a, martingale-the
present value of the value of a mutual fund that reinvests all
dividend earnings. See Leroy (1989, pp. 1589-90).

30

ECONOMIC

REVIEW.

of ,future dividends. First, update equation (1) one
period, replacing t + 1 by t +2 and t by t + 1, and
substitute the resulting expression in (1) for pt + I to
get
(2)

pt = (1 + r)-‘
Et[(l

+ r)-‘
Et+I[pt+z

+ dt+21 + c&+11,
where Et + l[wt + 21 ‘ the expected value of the
is
random variable wt + 2 given information available at
time t + i. The law of iterated expectations implies
that IMEt + dpt + 211= E&t + 21.‘
Equation (2) can
then be rewritten as
(3)

pt = (1 + r)-lEt[d;+ll

+ (1

+‘
ds2Et[pt+2

+ &+21.
If one repeats this substitution n times, the result
is an equation relating pt to the stream of dividends
from period t + 1 to period t + n, plus the term (1 +
r) - nEt[pt + ,I. One can assume that this term converges to, zero as n approaches infinity. This assumption rules out speculative bubbles. (We’ discuss this
ll
assumption below.) Under this assumption, the equation obtained as the,limit of this repeated substitution is
(4). Pt = d,
where v,” =

5 (1 + r)-.SEtIdt+sl.
‘ 1
s=

Equation (4) states that the current price .of a stock
equals v:‘ the present value of expected future
,
dividends.
Thismodel was first advanced by Paul,Sainuelson
(1965), and is often .called the’ “efficient capital
mark&s inodel,” a teim associated with Eugene
Fama’ (1970) exposition. The model can be
s’
vietied as arising in particular’ classes of artificial
ecoiomies; An artificial’
economy is just a particular
mathematical
specification of the’ preferences,
technological opportunities, and informational abilities
of ‘
economic agents, together ‘
with sonii= notion -of
the mutual consistency of plans, or equilibrium. In
one class of artificial economies that gives rise to the
Martingale Model; agents are risk-neutral, discount
the future at the same rate, and share common
information and beliefs about future returns (see
Lucas (1978)). In another such. class there is a
p&rf&tly risk-free asset, and all randomness in stock
returns is idiosyncratic to individual stocks (see
Connor (1984)).
MARCH/APRIL

1990

In many other artificial economies, equation (1)
does not always hold. However, there is usually a
more general version of equation (1) that does hold.
In general, the current stock price is related to the
entire probability distribution governing the sum of
next period’ price and dividends, rather than just
s
the mean, as in (1). This implies that risk premia
can affect the current price of a stock, as in the
Capital Asset Pricing Model or the Arbitrage Pricing Theory (see Connor (1984)). More general
economies also imply that discount rates can vary
over time, rather than remain constant as in (1). It
turns out, however, that empirical contradictions of
(1) or (4) do not seem to be attributable to risk premia
or time-varying risk premia (see West (1988b)).
Even in economies in which equation (1) holds,
the stock price may not satisfy equation (4) because
of the presence of a speculative bubble. A stock price
is said to contain a bubble if it can be written as

III.

SOME EMPIRICALREGULARITIES
IN STOCK PRICES
Aggregate Stock Prices
The Martingale Model has some strong implications for the joint behavior of stock prices and
dividends. One of the most striking of these is an
upper bound on the variability of stock prices relative
to dividends. There is now a large literature, beginning with the seminal papers by Leroy and Porter
(1981) and by Shiller (1981), that documents the
failure of empirical data on stock prices and dividends
to satisfy this inequality; see West (1988b) and Leroy
(1989) for recent surveys.
To understand this variance bound, first define a
variable et + s as the difference between the actual and
expected dividends in period t +s. In other words,
(6)

(5)

p! = pt + bt,

where pt is given by e uation (4), and bt is the bubB
ble term. In order for pt to satisfy (l), it must be true
that bt = (1 + r) - ‘
Et[bt + r]. In fact, any random
bt series that satisfies this condition implies that p!
satisfies (1). There are an infinite number of bt
random variable series that satisfy this condition, so
there are an infinite number of solutions, p:, that
satisfy equation (1). Only one solution is consistent
with (4), however, and that is the solution in which
bt = 0. Recall that in deriving equation (4) we
assumed that the expression (1 + r) - “Et[pt + ,]
converges to zero as n grows very large. This effectively rules out any solution other than p! = pt.3
A negative value for bt implies that there is a positive
probability that the stock price is eventually negative,
which is inconsistent with the free disposal of stocks.
This case is conventionally ruled out. A positive value
for bt implies that the stock price is always above
the fundamental value, given by equation (4). It is
useful to keep in mind the properties of bubble solutions to equation (1) because our model of takeovers
and stock prices predicts that an econometrician
would be unable to reject the hypothesis that stock
prices contain a bubble term.

3

To see this, note that:
(1 + 19-~Etlpi’
+.l = (1 + r)-"EtIbt+n pt+nl
+
= bt + (1 + r)-nEtfpt+.l,

which converges

to bt if pt +n is the series defined by (4).
FEDERAL

RESERVE

et+s dt+,=

Et[dt+,], for s=1,2,3,.

. . :

Then define a variable d;as the present discounted
value, of acmal dividends. Shiller called this the “ex
post rational” stock price. This is what the price of
the stock would be if the entire stream of future
dividends were known with perfect certainty, and the
Martingale Model, equation (4), determined the
price. More explicitly,
(7)

d;

=sFl(l

+ r)-Sdt+s.

Using these two definitions, we can obtain the following relation between pt and dt?
(8)

d; = pt + xt,

where x:’ = El(l

+ d-Set+s.

Equation (8) states that the ex post rational price is
equal to the actual current price plus the present value
of the unexpected component of future dividends.
One immediate implication is that the current price
is an unbiased forecast of the ex post rational price;
in other words, pt = Et[dt’ This follows from the
].
fact that Et[et + ,] = 0 because of the optimality of
forecasts of future dividends. The optimality of
forecasts also implies that the forecast errors, et + s,
s=l,Z,
. . . , are uncorrelated with pt, and this
implies that pt and x? are uncorrelated. Therefore,
BANK

OF RICHMOND

31

we can derive the following relation between
variances of pt and d;:
(9)

var(d;)

the

= var(pt) + var(xtd).

Since variances are positive, equation (9) implies that
the variance of stock prices has an upper bound:
(10)

var(d; ) 1 var(pt).

The variance of the actual stock price can be no
greater than the variance of the present value of actual future dividends.
The original tests of the inequality (10) were first
published in 1981 by Leroy and Porter, and by
Shiller; both papers reported violations that were large
in magnitude and statistically significant. A large
number of papers have appeared since developing
or applying this inequality test (see the recent survey
by West (1988b)). S ome initial work questioned the
finding of excess volatility on econometric grounds,
arguing that small sample bias and/or the presence
of unit roots in dividends may explain the results (see
Flavin (1983), Marsh and Merton (1986), and
Kleidon (1986)). Subsequent studies taking account
of the possibility of unit roots and small sample bias
“still tend to find substantial excess volatility” (West
(1988b), p. 639).
Recent work has examined the possibility that
risk aversion causes stock prices to deviate from the
Martingale Model, as might be expected from more
general theories of asset pricing (see Singleton
(1987)). Allowing risk averse investors, however, fails
to explain excess volatility. Other recent work has
examined the possibility that the expected rate of
return, r, varies over time (see Campbell and Shiller
(1988a and 1988b), and West (1988a)). Although
this line of work is at a very preliminary stage, initial
results suggest that the variance of the expected rate
of return would have to be implausibly large to explain the excess volatility results. Consequently,
many of the simplifications inherent in the Martingale
Model do not seem to be responsible for the inconsistency between the model and the data.
Some researchers have examined whether the
finding of excess volatility could be caused by
speculative bubbles. It appears that empirical
evidence on stock prices is consistent with the
presence of bubbles, which is not surprising, because
bubbles can take many forms (see West (1987 and
1988a), and Shiller (1984)). Bubbles are often
32

ECONOMIC

REVIEW,

associated (in many people’ minds) with large suss
tained increases in asset prices followed by a sharp
collapse, as in Tulipmania, the South Sea Bubble,
and other famous cases (see Mackay (18X), but see
also Garber (1989)).
Bubbles need not take such a spectacular form,
however. In the model of takeovers and stock prices
that we consider below, an econometrician examining data generated by the model would be unable
to reject the conclusion that the stock price includes
a bubble. But ‘ our model, what’ appears to the
in
econometrician to be a bubble term is uniquely determined and has an economic rationale-it
is actually
part of the fundamental of the stock, properly defined. Therefore, one way of interpreting our explanation of stock price volatility is that the characteristics
of the. financial claims of the modern corporation
could give rise to whatappears to be a bubble in stock
prices. This exemplifies the point made by Hamilton
and Whiteman (1985) and Hamilton (1986) that
movements
in the true fundamental
that are
unobserved by the-econometrician are indistinguishable from bubbles.

Takeovers

and Individual Stock Prices

The research discussed above focuses on the
behavior of the aggregate stock price and dividend
series. At the level of the individual firm, the relationship between the market for corporate control
and stock prices has been extensively investigated
using the “event study” methodology. This approach
examines the behavior of share prices of participating
firms around the date of the announcement
of a
takeover or other change in control. To the extent
that stock price changes cannot be explained by a
market model (the Capital Asset Pricing Model, for
example), these abnormal changes are attributed to
the takeover event. Much of the event study literature
on takeovers was surveyed by Jensen and Ruback
(1983).4 Averaging over the results of a large
number of studies, Jensen and Ruback find that
there is a 30 percent abnormal increase in the stock
price of a target firm in the event of a tender offer
takeover (a takeover executed by a direct purchase
of shares). In the case of mergers, when there is
agreement on the acquisition between the managements of the acquiring and target firms, the gains in
the target’ stock price are substantially lower (20
s
percent). One might conclude, in these cases, that
part of the premium that the acquirer is willing to
4 Also see the recent survey by Jarrel, Bri&ey
MARCH/APRIL

1990

and Netter (1988).

pay is going in some form to the incumbent management. When a change in control is executed through
a proxy contest with little or no direct purchase of
shares by those acquiring control, the abnormal stock
price change is much smaller (8 percent). In the cases
of tender offers and mergers, Jensen and Ruback
report that the abnormal changes in the stock prices
of bidding firms are much smaller than those for the
target firms; there is a 4 percent change for bidding
firms in tender offers and no significant change in
mergers.
Jensen and Ruback interpret the results from the
event study .literature as providing evidence that the
market for corporate control reallocates productive
resources from less to more efficient users (managements). That is, takeovers create value for shareholders because they result in an improved use of
resources. One might call this the “inefficient management hypothesis.” This hypothesis suggests a world
in which some managements are better matched. than
others to the assets and activities of any given firm.
Hence, in this view, the market for corporate control is a market in which managers search for and
acquire firms to which they are well matched.
Like the inefficient management
hypothesis,
the process described in this paper is also one of
searching and matching. In our view, however, a
manager can earn private benefits from an improved match between management and assets. If
managers are motivated by this private value, then
one would expect to see acquiring managements pay
a premium for control. At the same time, one would
not necessarily expect acquisitions to generate value
for shareholders of the acquiring firm. These expectations are supported by the distribution of stock price
gains observed in the event study literature; large
gains accrue to target firm shareholders in tender offer
takeovers and little or no gains accrue to acquiring
firm shareholders. Similarly, one would not necessarily expect acquisitions driven by the value of
control to result in improved profitability after the
acquisition. An extensive literature, surveyed by
Mueller (1987), has examined post-merger performance using accounting data. The most notable result
is the failure to find evidence of improved performance after mergers. While this evidence has been used
to discredit the inefficient management hypothesis,
it is consistent with the approach described in this
paper based on the private value of control.

Takeovers and Aggregate Stock
Price Movements
If one accepts the existence of a control premium
in a takeover transaction, there are sharp implications
FEDERAL

RESERVE

for the time series behavior of an individual firm’
s
stock price; the price would fluctuate not only ,with
information about future dividends, but also with
information about the probability of a change in control of the firm. The existence of a control premium
does not, by itself, have any implications for aggregate
stock price behavior. If the probability of a takeover
were independent across firms and over time, then
the effect on stock prices would average out across
firms. Stock price indices would, then, be expected
to vary only with information about expected future
aggregate dividends. If, however, there are systematic
movements in aggregate takeover activity over time,
then takeover activity (or expected future levels of
takeover activity) will affect aggregate stock prices.
There is evidence suggesting that aggregate
takeover activity is subject to systematic movements
over time. Shughart and Tollison (1984) examine
annual data on the number of takeovers in the U.S.
from 1895 to 1979. They find that they cannot
reject the hypothesis that merger activity follows a
random walk. If this is so, then an unexpected rise
in takeover activity has persistent effects. Hence,
future expected rates of takeover activity will depend
on the current rate. If a higher aggregate rate of
takeovers implies a higherprobability that a randomly
selected firm will face a challenge for control, then
the random walk behavior of takeover activity has
implications for the behavior of aggregate stock
prices. A rise in takeover activity implies a rise in
the rate at which control premia are realized in
changes of control. This, in turn, implies higher stock
prices in the aggregate.
The notion that there is a link between takeover
activity and aggregate stock prices is certainly consistent with casual observation of the behavior of
stock prices in the 1980s. The decade witnessed an
unprecedented wave of activity in the market for corporate control, coinciding with a sustained and
substantial rise in stock prices. The two large declines
in the market in the late 1980s in October 1987 and
October 1989, both came at times when many were
beginning to suspect that the takeover and buyout
boom might be coming to an end. In fact, much of
the discussion surrounding the mini-crash of October
1989 centered on the collapse of a single deal, the
UAL buyout. It was feared that the failure of the
pilots’ union to raise the financing for their offer
was a signal of similar problems arising for future
deals. Many commentators attributed the preceding
increase in overall stock prices from January to August
of 1989 in part to expectations of increased takeover
activity. Most notably, some recent research seems
to indicate that the over 10 percent decline in the
BANK

OF RICHMOND

33

stock market on October 14-16, 1987, which
arguably triggered the crash of October 19, 1987,
was caused by U.S. House Ways and Means Committee consideration and approval of a tax bill containing restrictive antitakeover provisions (Mitchell
and Netter (1989)).
Iv.

ANALTERNATIVEEXPLANATIONOF
STOCK PRICE VOLATILITY
The previous section summarized the empirical
literature on the volatility of aggregate stock prices
and argued that volatility is too large to be consistent with the Martingale Model described in Section II. In this section we present a theory of stock
price volatility that is based on takeovers. The theory
is also consistent with the empirical regularities
displayed by individual stock prices around control
change events. In addition, the theory offers an
explanation for the broad comovements in stock
prices and control change activity described above.
The key to the relationship between takeovers and
the volatility of stock prices is the value of control
of a firm. In this section we discuss the concept of
“the value of control,” and describe how the value
of control can affect stock prices.

The Nature of the Firm and
the Value of Control
To make precise just what we mean by the term
“value of control,” we briefly describe some important features of the way the modern, publicly held
corporation is organized.
The diverse activities associated with the modern
large corporation involve a large number of people:
employees, directors, and the individuals and institutions holding the contractual liabilities of the firm,
to name just a few. We focus on two main groups.
We refer to the individual or group of individuals
exercising effective control over the firm’ operations
s
as the management or managers: the chief executive
officer, for example. We will refer to the people or
institutions that own the explicit financial claims
issued by the firm as claimholders: for example,
shareholders, bondholders, or banks that have made
loans to the firm.
The relationship between managers and claimholders is a complex one, governed by a variety of
legal (and other) arrangements. For example, loan
and bond contracts often contain explicit covenants

34

ECONOMIC

REVIEW,

that restrict future actions of the firm, including
investment decisions, financial restructuring, or excessive dividend payouts (see Smith and Warner
(1979)). Publicly held firms generally have a rather
elaborate and explicit governance structure. Holders
of shares of stock have the right to vote periodically
on various matters affecting the firm. A board of
directors, formally elected by the shareholders, is
charged with the responsibility of overseeing the
operation of the firm, and has the vested authority
to hire and dismiss the managers of the firm.
Managers submit important policy decisions to the
board at regular meetings for formal approval. While
holders of various forms of claims do have some
ability to monitor and, perhaps, affect the actions of
managers via these mechanisms, managers in the
typical large corporation have wide discretion over
how they use the firm’ productive resources.
s
A more detailed description of these complex
arrangements is beyond the scope of this paper.
There is an extensive literature on the design of the
arrangements between managers and claimholders,
much of which draws its inspiration from Berle and
Means (1932) (see, for instance Jensen and Meckling (1976) and Fama and Jensen (1983). From
this literature, one can identify an important tradeoff
between two opposing forces: sharing risk widely
versus minimizing conflicts of interest.
The desire to allocate risk efficiently leads to widely
dispersed ownership of the (risky) residual claim
usually associated with ownership of the firm. The
dispersion of ownership leads immediately to the
need for delegated decision making authority. The
communication and coordination costs which would
be associated with direct decision making by a large
number of claimholders makes the appointment of
professional managers (with relatively small ownership stakes) a virtual necessity. This is a key
characteristic distinguishing the modern corporation
from the sole proprietorship in which the owner and
manager are one individual.
The delegation of decision-making authority is not
without its costs. The fact that management’ owners
ship stake is relatively small suggests that the goals
and incentives of managers may not always coincide
perfectly with those of the claimholders. In addition,
managers, who are directly involved in the operation
of the firm, are likely to have a significant informational advantage over claimholders regarding alternative uses of the firm’ resources. The delegation
s
of decision-making allows managers to pursue private
objectives that might harm the long-term interests
of the firm.
MARCH/APRIL

1990

Many of the legal arrangements between claimholders and the firm’ managers alluded to earlier are
s
designed to mitigate the misalignments of incentives.
Managerial compensation schemes are often explicitly
tied to the performance of the firm. This strategy
imposes part of the residual risk associated with
managerial decisions on the managers themselves.
This type of compensation, however, works against
the goal of efficient risk sharing which originally led
to the dispersion of ownership and the delegation of
decision-making authority, since managers are made
to bear the risk rather than claimholders. Some
managerial decisions can be directly mandated by
claimholders through, for instance, covenants in bond
and loan contracts. More specifically, covenants
give the claimholder certain rights-to
declare
bankruptcy for example-in
certain predetermined
circumstances. This. presumably discourages the
firm’ managers from taking the undesirable actions.
s
The manager’ informational advantage, however,
s
makes the monitoring of such agreements imperfect
at best. And finally, the board of directors, ostensibly representing shareholders’ interests, supervises
managers and attempts to ensure that managerial
decisions are in the interest of shareholders. The
limitations of the supervisory role of boards of directors are apparent: because they devote very little time
to a given firm, they are unable to duplicate the
managers’ knowledge, and so must rely on limited
and self-serving reports by managers in evaluating
managers’ performance. In short, the nature of the
relationship between corporate management and corporate claimholders leaves management with wide
discretion
in allocating the firm’ productive
s
resources.
The problems associated with the separation of
ownership and control suggest that managers may
be able to extract private benefits, or “rents,” from
their insider positions. There may be actions that
managers can take that benefit themselves without
adding to the value of the firm and, therefore, to the
wealth of the claimholders. The value of control,
then, is the value of the stream of benefits which
necessarily accrue to those in control of the firm. This
is a private value in the sense that those in control
cannot credibly commit to transfer these benefits to
claimholders. These benefits may take the form of
private consumption of “perks” or of the pursuit of
private goals distinct from value maximization. It has
also been suggested by Jensen (1986) that managers
can derive private benefits from the discretionary control over the firm’ free cash flow. For instance, in
s
order to pursue firm growth as an end in itself, a
manager may use retained earnings to fund investFEDERAL

RESERVE

ments with negative net present value. More generally, access to internal funds for investment shelters
managers’ decisions from the scrutiny they would
receive in obtaining external sources of finance.
Allowing management to extract private value may,
in fact, be part of the (imperfect) scheme for providing managers with correct incentives. If managers
are able to extract more rents during good (profitable)
times than bad-because,
for example, managers’
actions come under more’
direct scrutiny during bad
times-then
managers have an incentive to take
actions that make good times more likely. In addition, control of a large organization may be valuable
in and of itself, quite apart from any resources directly
obtained thereby. It could provide utility directly for
managers in the form of enhanced prestige or ego
gratification.

Corporate Financial Claims
We can now describe how the value of control of
a firm affects the nature of the financial claims issued
by the firm. It is essential to our argument that a
financial claim is a contract between the issuer (the
corporation) and the holder of the claim. This contract specifies payments to be made by the corporation under a variety of contingencies. Sometimes
these specifications are explicit, as in the case of bank
loans or corporate bonds. In other cases, promised
payments are implicit, as in the expectation of dividend payments to equity holders based on an announced dividend policy. In addition to stipulating
payments, the financial claim gives the holder certain rights. A debt holder may have the right to
directly monitor some of the actions taken by corporate management,’ as specified in a bond covenant.
Debt claims also carry important rights in the case
of bankruptcy. The main right attached to a standard common stock equity claim is the right to vote
on some corporate governance matters on a oneshare-one-vote basis. Most important, shareholders
have the collective ability to choose corporate
management through the election of the board of
directors.
Debt and common stock equity are the predominant forms of financial claims issued by the modern
corporation. Other forms of claims can be viewed
as hybrid varieties, such as preferred stock or convertible debt. Uncovering the determinants of the
mix of claims issued by corporations remains one of
the major challenges of financial economics. A recent paper by Harris and Raviv (1988) is particularly relevant to the concerns of this paper. They
BANK

OF RICHMOND

3.5

assume that managers derive private value from the
control of a firm and examine the implications of this
assumption for the design of securities. They find
that if claims are to be issued with an interest in promoting efficiency-enhancing changes in control but
deterring efficiency-reducing changes, then rights
to vote on changes in control should be attached to
equity claims and not debt claims. This is, of course,
exactly the allocation of rights observed.
Given our arguments above that managers derive
private rents or value from the control of firms, it
is useful to view equity claims as bundled claims. The
voting feature of tradeable equity shares implies that
control can be acquired through the purchase of
shares; buy enough shares, and you can install
yourself or anyone of your choosing in top management positions. Hence, the claim to a stream of
dividends is bundled with a claim to the premium
that a potential manager might pay to acquire enough
votes to take control of the firm. Note that this feature
is unique to equity; one cannot acquire control of
a firm by buying all of its debt. Hence, the equity
claim is necessarily linked to the process of change
in control, regardless of how those changes come
about.
It is interesting to note that firms often issue voting
and nonvoting classes of equity. While nonvoting
equity is relatively unimportant in publicly held firms
in the U.S., in some other countries it is more important. The relative prices of voting and nonvoting
shares often reflect the value of control. For instance,
Hermann and Santoni (1989) show that when Swiss
firms began allowing foreign investors to hold voting
shares, the value of the voting shares increased
relative to the value of outstanding nonvoting shares
by as much as 20 percent. While there may be other
explanations of this increase, allowing foreign purchases of voting shares may have increased the
likelihood of an acquirer buying shares in order to
obtain control.

Takeovers

and the Value of Control

When the control of a corporation changes hands,
the value of control is often transferred as well. The
way in which a change in control takes place determines how the value of control is transferred and how
the financial claims on the corporation are affected.
One form of change in control is, of course, internal
succession to the top management positions. When
a vacancy at the top is filled by promotion from
within, the value of control need not be “purchased” from shareholders. The internal transfer of

36

ECONOMIC

REVIEW,

control might represent an implicit contract between
generations of managers; new managers may have
“paid for” control through a period of apprenticeship.
Alternatively, one might view the value of control
as accruing to a coalition or team of managers (such
as the CEO, the board of directors, and other top
executives). Internal succession then amounts to
keeping control in the hands of the same coalition.
Similarly, the board of directors hiring a CEO from
outside the firm, for instance, is a transaction between the controlling coalition and an individual who
is joining the coalition.
In the cases of internal succession and external
hiring discussed above, there is no change in the
designation of the delegated decision-making authority. There is, therefore, no need for those engaged
in the change of control to purchase control through
the acquisition of shares. However, sometimes a
change in the delegation of control becomes desirable
to at least some shareholders. They may feel that
incumbent management has not responded well to
a change in the economic environment or that an
alternative management would perform better. In
such cases; the shareholders’ voting rights become
important.
The various ways in which a change in the delegation of control might be brought about were
discussed by Manne (1965) in an effort to outline
the economic role of corporate takeovers. Manne
views all changes in control as attempts to replace
less efficient with more efficient management. The
nature of the equity claim gives an unsatisfied
shareholder a number of options. First, one could
try to unseat the incumbent board of directors
through a proxy contest. Proxy contests, however,
are relatively infrequent. This may be because of
the costs involved in soliciting votes; incumbent
management can use corporate resources to fight its
battle, but dissidents must use their own resources.
Having incurred the expense, the outcome of the
contest remains uncertain until the actual vote is held.
One way in which a challenger for control can reduce
the uncertainty is through his or her own ownership
of shares. This, of course, suggests an alternative
route to obtaining control. By acquiring enough
shares, one can dispense with the need for a pro-.
longed and potentially unsuccessful proxy contest.
Faced with a challenge to its (valuable) control,
incumbent management can be expected to spend
resources resisting the change. This is true in the
case of a proxy contest or an acquisition of shares.
When a challenger attempts to gain control through
MARCH/APRIL

1990

the acquisition of shares, or when an incumbent seeks
to protect control through the acquisition of shares,
the share price is bid up to reflect all or part of the
private value of control. In a friendly merger as
opposed to a hostile takeover, shareholders may
realize a smaller part of the value of control; this
would be so if the acquiring management obtained
the incumbent management’ consent through some
s
form of implicit or explicit payment. In short, the
extent to which a change in control results in value
accruing to shareholders depends on the extent to
which there is competition for control.

in the market for control. Both the private value and
the public profitability that a manager can achieve
with a firm may depend on how well-matched that
manager is to the firm’ organization, array of acs
tivities, and “corporate culture.” Time and resources
may be required to investigate the quality of such
a match. Hence, the potential acquirer’ behavior may
s
best be viewed as a process of costly search. Both
the costs of search and the likely costs of making
an acquisition (once a match is found) affect the
raider’ willingness to search for targets.
s

In the absence of frictions or barriers to open
competition in the market for control, the market
price of equity would always fully reflect the value
of control. There are, however, some important frictions built into the market for control. Many of these
derive from the very nature of the relationship between corporate ownership and management. An
unrestricted
market for control could expose
managers to too much employment risk; managers
might then have an insufficient incentive to accumulate firm-specific human capital. Shareholders
have an interest in giving their, delegated decision
makers an incentive to make themselves well
matched to the particular firm they are managing.
On the other hand, complete protection from the
market for control is not good from the shareholders’
point of view. Entrenched management can receive
the private benefits of control with no concern for
the firm’ performance on shareholders’ behalf. These
s
opposing forces suggest an optimal intermediate
degree of protection for incumbent managers. Such
protection may take the form of golden parachutes,
or provisions in the corporate charter giving the
manager the right to take certain defensive actions
in the event of a takeover attempt.

Viewing the market for control as a market in which
buyers or raiders search for targets has implications
for the effects of the value of control on stock prices.
The extent to which a share price reflects the value
of control depends on the probability that a potential raider finds the firm to be worth challenging for
control. This probability, in turn, depends on the
overall level of ongoing search activity. In addition
to the artificial and natural frictions suggested above,
the level of search activity is likely to depend on what
might be called the “infrastructure” of the market for
corporate control. By this we mean, for instance, the
conditions under which a raider could obtain financing for a deal. Casual observation suggests that the
takeover boom of the 1980s was fed, in part, by
innovations in the market for below-investment-grade
corporate debt (junk bonds). In short, the availability of a full array of financial and legal services
facilitates the search process. Variation over time in
these infrastructure services might contribute to variation in the level of search and takeover activity, and
thus to variations in stock price volatility over time.

In addition to the frictions built into the form of
corporate governance, government regulations can
create barriers to takeover activity. A variety of federal
and state regulations restrict the actions of a raider
in a contest for control. A prime example at the
Federal level is the 1968 Williams Amendment,
which restricts the actions of bidding firms by, for
instance, requiring that tender offers be outstanding
for a minimum number of days. Such restrictions can
add to the cost of attempts to acquire control, thereby
making such attempts less frequent.
One might label. barriers to takeovers that arise
from legal restrictions or the contractual relationship
between ownership and management “artificial” barriers. There may also be important “natural” barriers
FEDERAL

RESERVE

We are not aware of a theoretical explanation of
the variations in aggregate takeover activity, although
it has been suggested (e.g., by Gort (1969)) that
waves of mergers are driven by large disturbances
to the economic environment. For our purposes, it
is enough to take as given that takeover activity varies
over time according to a random process which can
reasonably be described by a random walk. With this
assumption, in periods of high .takeover activity, such
activity is expected to remain high. Hence, the
perceived probability that a randomly selected firm
faces a challenge for control in the near future is high,
and the value of a stock price index significantly
exceeds the expected value of the underlying stream
of dividends. By similar reasoning, in periods of low
takeover activity, stock prices are closer to the value
of future dividends. These arguments lead directly
to our excess volatility results.
BANK OF RICHMOND

37

Takeovers

and Stock Prices

(12)

The descriptive analysis above can be made quite
rigorous. To be specific, one can formally specify an
artificial economy that displays the forces described
above, albeit in relatively stark and simple form. We
have done this in a forthcoming paper (Lacker, Levy
and Weinberg 1990), where we specify agents’
preferences,
their production
and investment
technologies, and, most crucially, the informational
opportunities available to them. The critical feature
of the economy is that the agent that manages an
asset also has the ability to manipulate the observed
return on the asset. The appendix of this paper
describes a similar economy in more detail. Here we
present the main implications.

where vi’ =

(11)

pt

=

(1

+

r)-%[pt+1

+

var(pt) = var(v:‘ + var(v4) + Zcov(v~,v~).
)

The possibility of excess volatility in stock prices
is now easily demonstrated:
Pmposition I: If var(vi ) + Zcov(v:‘ ) > 0,
,v~
then var(pt) > var(v:‘ and the variance of the
),
stock price is greater than the variance of the
present value of expected
dividends. For
example, if v? is not negatively correlated with
v?, then var(pt) > var(v:‘
).
Therefore, the price of a stock can vary by more
than is justified by variations in expected future
dividends. The condition that v:’and v? are positively correlated is stronger than required; all that
is needed is for the correlation between v:‘
and v? to
be not too large a negative number. This condition
seems reasonable. if the actual real’ value of cqnixed
trol is positively correlated with reakd dividends,
this assumption is satisfied. One would think that
the expected value of controlling a firm would be
larger if the firm is expected to do better.

dt+l

+ 76+1rlt+11.

The variable ?r: + 1 is the probability that a takeover
occurs during period t + 1, given information available
during period t. Equation (11) states that the current price of a share equals the expected discounted
value of the sum of the price, dividends and the value
of control, with the latter weighted by the probability of a takeover.
As in the Martingale Model, we can use this equation to derive an expression for the current stock price
in terms of the entire stream of future dividends. As
before, the derivation requires repeated substitution
for pt + 1, pt + 2, and so on. The result is

38

E (1 + r)-SEtId+Sr)t+SJ
s=l

and v? is defined as before. Comparing equation (12)
to equation (4) reveals that the present value relation is now augmented by a term related to the value
of control. The current stock price is equal to the
expected present value of dividends plus the expected
present value of the premium associated with control, adjusted by the probability that shareholders
realize that premium. One immediate implication of
(12) is that the variance of pt can be written in terms
of the variances and covariance of v:’and vZ:
(13)

For our economy, we can derive the equilibrium
price of shares of stock in any given asset. Let pt
now be the price of a share ;fno takeoveroccursduringpmbd t, and let qt be the price paid if there is
a takeover in period t. Both pt and qt are determined by general equilibrium conditions in our economy:
pt by the value investors place on a share kriowing
no takeover will occur until next period at the earliest,
and qt by the value to a new manager of acquiring
and subsequently controlling the firm. In equilibrium
qt > pt, meaning that a new manager is willing to
pay a premium to acquire control of the firm. This
is the value of control in our economy, and we denote
it r)t = qt - pt. We can find an expression for the
equilibrium value of pt, the current stock price, that
is analogous to equation (1). The result, derived in the appendix, is that the current stock price
depends on the expected value of control in a
takeover as well as on expected dividends and expected price as before:

pt = VP + v;

ECONOMIC

REVIEW,

We have not yet shown how the variance in stock
prices compares with the variance of dt’, the ex post
rational price. The Martingale Model predicts that
var(pt) I var(d;), but this inequality is violated empirically. Can our economy display violations of this
inequality? To find out, first recall that because d;
= v:’ + x;‘ and cov(v~,x~) = 0 because of the
,
optimality of forecasts of d;, we know that
(14)

var(d;)

= var(vP) + var(x:‘
).

The variance of pt can be written as follows:
(15)

var(pt) = var(v:‘ + var(v?) + Zcov(v2,vl)
>
= var(d;) - var(xf) + var(vY)
+ 2cov(v&v4).

MARCH/APRIL

1990

This expression gives us a condition under which the
variance bounds condition in the Martingale Model
is violated.
Pmposition If var(vl) + Zcov(v?,vl) > var(xP),
2
then var(pt) > var(d;), and the variance of the
stock price is greater than the variance.of the
present value of actual dividends.
The condition in Proposition 2 states that the
variance of the expected value of control plus the
covariance of the expected value of control and the
expected value of dividends must exceed the variance
of the error in forecasting the present value of
dividends. This condition can be understood by comparing var(pt) in our model, equation (15), with
var(pt) = var(d;) - var(xf) from equation (9) in the
Martingale Model. In the latter, var(pt) is equal to
var(vtd), which is less than var(d;) by the amount
var(x?). In our model, var(pt) is pester than var(v:‘
)
by the amount var(vZ) + 2cov(vt ,vt”). For this effect
to dominate, making var(pt) > var(d;), var(vZ) +
Zcov(v?,vY) must be larger than var(x!).
More intuitively, a share of stock in our model is
a bundled claim, consisting of the right to a stream
of dividends plus a share of the right to control the
firm. The latter is a value that can be realized by the
shareholder in the event of a takeover, and it adds
a variance to the stock price above and beyond the
variance in expected dividends. It contributes a
variance of its own to the price of the stock, and in
addition could well be correlated with the expected
present value of dividends. These two effects could
add enough to the variance of the stock price to make
it larger than the variance in the present value of
actual dividends, consistent with the empirical violations of the Martingale Model’ variance bounds
s
condition.
This explanation of excess stock price volatility
does not rely on some other explanations that have
recently been advanced. Some economists have suggested that fads or irrational “noise traders” are
responsible for observed anamolies in stock prices
(see Shiller (1984), Black (1986), DeLong et al.,
(1987), and Campbell and Kyle (1988)). In our
economy, all agents are fully forward-looking and
expectations are rational. There are no unexploited
arbitrage opportunities because the future control
premia are rationally anticipated and incorporated
into the current price of the stock. There are no
externalities, and no restrictions on the contracts
agents can write except those that follow from the
technological and informational constraints agents
FEDERAL

RESERVE

face. In fact, our equilibrium is Pareto optimal, meaning that no agents can be made better off without
making some other agents worse off. The key feature
of the economy that gives rise to excess volatility is
the friction affecting the contractual arrangements
between managers and claimholders; managers’
privileged position in control of the asset implies a
positive value of control.
The evidence from event studies of tender offers
and mergers, described in Section III, subheading
“Takeovers and Individual Stock Prices” above, is
consistent with the model presented here. The large
abnormal increase in the stock price of the target firm
represents the control premium qt. The fact that the
stock price of the bidding firm changes very little
suggests that a substantial part of the increased
productivity or private value of control associated with
the acquisition is captured internally by the acquiring firm and is not passed on to the acquiring firm’
s
shareholders.
Our model is also consistent with one of the most
striking features of the empirical variance bounds
literature. Shiller’ first paper contained graphs plots
ting d;, the present value of actual dividends (he
called it p;), against pt, actual stock prices, for the
Standard and Poor’ Composite Price Index and for
s
the Dow Jones Industrial Average. The path of pi
is fairly smooth, while the path of pt takes large persistent swings away from pi. An analogous graph,
using more sophisticated techniques for removing
trends, appears in a recent paper by Campbell and
Shiller (1987, Figure 2, p. 1083). Both Shiller’ and
s
Campbell and Shiller’ plots show that the difference
s
pt - pi was largest during four time periods: the first
decade of this century, the late 1920s the mid-1960s
and the early 1980s. The peak in the mid-1960s is
particulary large. All four of these periods correspond
to merger waves, periods in which changes in corporate control were particularly frequent. This suggests that the economy might experience periods in
which the probabilities of takeover for a broad range
of stocks move together and exhibit long persistent
swings. These swings might be caused by accelerations of technological shifts as some have argued
(Gort 1969), periodic shifts in the regulatory environment affecting changes in corporate control, or
innovations in the infrastructure of financial markets.

V.
SOMEIMPLICATIONSOFOURTHEORY
In this section we briefly discuss some of the
implications of our theory, first for recent events and
BANK

OF RICHMOND

39

trends in financial markets, and then for proposals
to alter the regulations governing takeovers and
markets for traded financial claims.

Recent Developments

in Financial Markets

Dramatic changes have occured in the markets for
corporate financial claims in the last decade. Stock
prices displayed a broad upward trend through the
198Os, albeit with setbacks in the fall of 1987 and
the fall of 1989. It is a widely held perception that
volatility has increased. An entirely new market has
emerged for below-investment-grade,
tradeable corporate debt, or “junk bonds.” And the pace of changes
in corporate control via acquisition of outstanding
shares has increased dramatically.
Simultaneously explaining all of these trends is far
beyond the scope of the present paper. However,
our theory is able to cast a new light on many of these
developments and their interrelations.s One plausible interpretation is that, for some reason, perhaps
linked to technological improvements in the ability
of investors to monitor firm performance, investors
are now much more willing to hold risky, high-yield
corporate debt such as junk bonds. While not all of
these securities have been associated with corporate
takeovers, it seems clear that they were essential to
many of the control transactions of the 1980s. The
shift in investor demand for these securities facilitated
takeovers that would not have been possible without the market for these securities. This improvement in the ability of acquirers to finance takeovers
led in turn to a secular rise in the probability of a
takeover for a broad range of stocks, n: + 1 in our
setup, and so led to a broad upswing in stock prices.
The theory might also illuminate some recent
short-run swings in stock prices. In recently published
research, Mitchell and Netter implicate Congressional
consideration
of antitakeover
legislation in the
October 19, 1987 crash in stock prices. They argue
that “a tax bill containing antitakeover provisions proposed by the U.S. House Ways and Means Committee on October 13, 1987, and approved by the
Committee on October 1.5 was the fundamental
economic event causing the greater than 10% decline
in the stock market on October 14-16, which
arguably triggered the October
19 crash.” By
making takeovers more costly, such a bill would
reduce the probability of future takeovers and thus
depress current stock prices, consistent with our
theory.
5 There

40

may be plausible alternative explanations,

of course.

ECONOMIC

REVIEW.

Analogously, the role of the junk bond market in
facilitating changes in corporate control might explain
why information about the willingness of investors
to hold below-investment-grade
securities would
affect stock prices so strongly, as they seemed to in
1989. At many times during that year, particularly
during the late summer and early fall, reports of broad
stock price declines cited sharp declines in junk bond
prices as the proximate cause. Similarly, the collapse
of one well-publicized deal, the bid for UAL, was
cited often for the broad decline in stock prices in
the fall. Finally, we note that the fall of broad
measures of stock prices since Summer 1989 has
coincided with a rise in the use of proxy fights in
corporate control contests, a method of control
change that does not provide shareholders with an
immediate monetary payment.

Regulations to Curb Takeovers
Stock Price Volatility

and Reduce

The finding of excess volatility of stock prices
is often taken as evidence of capital market imperfections or the presence of irrationality in the determination of asset prices (see Shiller (1984), Black
(1986), DeLong et al., (1987), and Campbell and
Kyle (1988)). Such imperfections, in turn, are often
adduced in support of various policy proposals that
would legislatively alter the way financial markets
currently operate. For example, some advocate that
“circuit breakers” or “collars” be imposed on the stock
market to halt or restrict trading in the event that
prices change by more than some prespecified
amount (see, for example, Greenwald and Stein
(1988)). The argument is that such restrictions
would reduce price volatility and improve the efficiency of financial markets. Similarly, some have suggested policy changes to discourage takeovers, either
by making the financing of takeovers more difficult
or costly, or by erecting barriers to changes in control via acquisition of shares (Scherer (1988), for
example).
A complete evaluation of these many proposals
is beyond the scope of this paper. We can point
out, however, that in our model takeovers regularly
occur, and are responsible for excess stock price
volatility. Excess volatility arises because of the
mechanisms by which the complex agency problems
inherent in the management and financing of the
modern corporation are resolved. These mechanisms
thus have a positive allocation role. In fact, excess
volatility is consistent with full market efficiency in
our model, and there is no constructive role for
government intervention. The lesson, then, is that
MARCH/APRIL

1990

the empirical finding that stock price volatility is larger
than can be explained by the Martingale Model does
not by itself justify regulatory intervention in financial markets.
Of course, a wide range of government policies
already in place have important effects on the
phenomena our model attempts to describe. The
requirements imposed on corporate charters constrain
the legal forms that corporate governance can take.
The Securities and Exchange Commission significantly constrains the financial structure and conduct
of publicly held firms, requiring, for example, that
votes be strictly proportional to shareholdings. SEC
regulations also impose severe restrictions on tender
offers. Underlying all financial claims, of course,
the structure of bankruptcy law has an important
and sometimes neglected influence on financial
arrangements.

thereby have important effects on the market for corporate control. Altering these regulations may well
reduce stock price volatility, but would most likely
alter the efficiency with which the control of assets
is allocated. Any assessment of the impact of altering such regulations must look far beyond the effect
on stock price volatility.

VI.
CONCLUDINGREMARKS
Our analysis contains a broader message for the
understanding of financial markets. Traditional approaches to asset pricing treat an asset as nothing
more than a claim to a stream of payments. The
starting point of our analysis is the view that a financial asset is a contractual relation between various
parties. A direct implication of this view, as our model
illustrates, is that financial assets in general, and
traded stock shares in particular, are bundled claims
tying together fragments of governance rights with
titles to streams of payments. Building upon this view
may provide us with new insights into the diverse
financial arrangements characteristic of developed
economies.

Our model is not rich enough, as yet, to be able
to fully assess the role of these and other regulations
affecting the market for control. We suspect that they
have important effects on the way various legal rights
are allocated among the claimants of a firm, and

APPENDIX

In this appendix, we develop a simple model that
delivers an equilibrium pricing equation of the form
of equation (11). The model is similar in spirit to the
one in Lacker, Levy and Weinbeig (1990). That
paper was concerned with demonstrating that excess
volatility was possible in principle. The model
described here is somewhat more general in that it
allows for periodic swings in takeover activity and
shows how these might lead to coincident swings in
stock prices.

person who controls the firm, and cannot be contractually transferred to claimholders. These are the
rents that accrue to managers and correspond to the
private value of control posited by Harris and Raviv
(1989).
The per period value of control, yi, and dividends,
dt, are assumed to follow stochastic processes given
by
(A. 1) yt = aoy’- 1 + ei, and
t

In this economy there is a large number of durable
productive assets (projects) and an even larger
number of people (agents). Some people are
claimholders, and others are managers. Together with
the services of a manager, a project can produce a
stream of putput {zt}, t = 1,2, . . . , where
zt = dt + y’ The portion dt of the project’ output
*.
s
is publicly observed. A manager can commit to paying out (all or a part of) dt to claimholders. The
remainder of the project’ output, y;, is privately
s
observed by the maqager and is not verifiable by any
outsider. Hence, yt is simply consumed by the
FEDERAL

RESERVE

dt = aldt- 1 + I.&
where ao, al I 1, and ei and ui are independent,
mean-zero random variables, independently and identically distributed over time.6
6 One could assume a more general joint process for (y’
,d,}
without altering the results. Under more general assumptions,
claimholders would need to be able to form expectations about
future values of unobservables, y, based only on publicly observed variables. Our assumptions allow us to avoid the filtering problem which arise with a more general specification.
BANK OF RICHMOND

41

Claimholders hold claims to the dividend stream
{dr}, and these claims are attached to voting rights
allowing claimholders, collectively, to delegate control of the productive project. For simplicity, we
assume that a change in control requires a unanimous
vote. Hence, a raider can acquire control by purchasing all claims to a particular project. We assume,
however, that there are agents engaged in search
activity to obtain information about the value of controlling projects. We do not model this search
behavior explicitly. Rather, we simply assume that
at any point in time there is a probability &, that
a raider arrives on the scene and obtains information about the value of control. We assume that r#~t
follows a first-order stationary markov process, that
is, that the probability distribution of & + 1, given the
entire history of realizations up to and including
period t (4t - j for j = 0, 1,2,. ..), depends only on +r.
The raider observes the incumbent’ current value,
s
y:, and also learns what his own value would be if
he took control; call this Y;. We assume that if a raider
arrives in period t, then y] = yi - r + e], where e]
satisfies the same assumptions as does &, but is drawn
independently of ei. Thus the raider’ current-period
s
value of control could be different from the incumbent manager’ current-period value of control.
s
The value of control is the present discounted value
of the stream of per period values of control,
weighted, for each future period, by the probability
that the manager will still be incumbent in that period.
The value of control is calculated by the incumbent
manager, yielding the amount the incumbent
manager would accept to forego continued control
of the asset. The value of control is also calculated
by the raider; it is the amount the raider would pay
to acquire control of the asset. Both quantities are
influenced by the past exqerience of the project
through the influence of y; - I, and for both the
incumbent manager and the raider the future of the
value of control evolves according to (A.l). But
because y] can differ from yi, and because these influence the expected values of y: + S and yt + s, there
can be a discrepancy between the value of control
to the incumbent and the value of control to the
raider.
Once y] and Y; are observed, the raider can choose
to initiate a bid for control through the acquisition
of shares. We assume that there is an arbitrarily small
but nonzero cost of initiating a challenge for control.
Hence, the raider only does so if his own value of
42

ECONOMIC

REVIEW,

control is greater than the incumbent’
s.7 Define Qt
as the probability that a raider appears in period t
and has a greater value of control than the incumbent. The value of @t depends on the raider’ exs
pectations of future per period values of the control,
yi + s, and the probability that some other raider will
come along and acquire the asset in the future. We
take the series 9t as given for now. Let q” be the
value of losing control: the expected present value
of the manager’ earnings from the next-best alters
native occupation in the event of losing control of
the asset. Then the value of control of an asset can
be written as
(A.2)

vf = (1 + r) - ‘
Et{&+ 17’
+ Cl-@t+d(yf+1

+ rlf+1)}.

ri + r is the value of being in control at the end of
period t + 1. Equation (A.2) states that the value of
control is the present value of the value of losing control, multiplied by the probability of losing control
next period, plus the value of remaning in control
at the end of next period, multiplied by the probability of remaining in control. An identical expression determines the value of control for a raider, $.
Note that if a raider assumes control this period, at
the end of the next period he is an incumbent, so
ri + r appears on the right side of the expression for T$Z
(A.3)

$ = (1 +r) - ‘
Et(at + 17’
+ (I-@t+d(yf+1

+ rlf+1)}.

If a raider arrives in period t, a change of control
takes place only if q; > & the value of control to
the raider exceeds the value of control to the incumbent. Because yi + s evolves according to a stationary
process, one can show that T,I~ 7; if and only if
>
e5 > e:, the current-period value of control is larger
for the raider than for the incumbent. Therefore, the
probability that a change in control occurs ifa raider
arrives in period t is Pr[e] > e:]. The probability that
a change in control actually occurs in period t is then
& = &Pr[e] > e]], the probability that a raider
arrives times the probability that a change occurs
given that a raider has arrived. Given our assumptions about tit, e:, and ei, the expected future rate
of change in control depends only on the current
value of Qt.
7 Relaxing this assumption, so that there is a dontest for control
whenever a raider “arrives,” would not change the nature of the
results but would complicate the computation of the present
discounted value of control.
MARCH/APRIL

1990

We define ?T:+~ to be the probability, given that
the firm does not face a challenge to control before
or during period t, of the first such challenge occurring in period t+s. For s = 1, a:+, = Qt+r. For
s 1 2, n:+, is given by
n:+s

Using equation (A.4) and solving forward, we have
pt = v:’ + v:,
where

s-l
= +‘ j41 (I-@t+j).
t+s

0:

=

Et

s

s~l(l+r)-s*:+sd+s
I

1

Equation (A.2) can now be solved forward to yield
T$ = Et

For convenience,
$+s
Section IV, subheading
Prices.”

c” (1 +r)-s[7r:+sqo
s=l

+ (n: + s(1 -at

+ s)/@t + s>Yi + s
I

Notice that 7: depends on the expected future rate
of takeover .activity, as well as on the expected future
values of y:.
If a raider arrives in period t and draws a current
value of control, e:, that is larger than the incumbent’ then the raider outbids the incumbent by pays,
ing a premium of $ for the equity. shares of the firm.
In the event of a takeover, the purchase price (ex
dividend) of the shares is
(A.4)

qt = vt’ + 7:.

In the event that there is no takeover attempt in
period t, the (ex dividend) stock price is
pt = (l+r)-‘
EtIdt+r

+ a:+1qt+1

+ (1 -d+1)pt+11.

is written.
“Takeovers

as ~t+~ in
and Stock

Suppose that there are a large number of identical
versions of the asset that we have just described. The
stochastic processes governing dt, ei, e:, and 9t are
the same, although the realizations of these random
processes are independent across assets. If’ the
number of these assets is quite large, then the fraction that experience a change in control is very close
to the population probability that a change in control occurs (by the Law of Large Numbers). Now
define ri + s as the probability that a takeover occurs
in period t +s to any given firm selected at random,
given the information known in period t. Imagine
calculating a stock price index as a weighted average
of individual stock prices; the weights are not
important-any
arbitrary weights will do. Then the
formula derived above will also apply to the stock
index, where pt is the value of the stock price
index, and qf, VT,and v? are interpreted as weighted
averages across stocks.

REFERENCES
Berle, Adolf A., Jr., and Means, Gardiner C., 1932, The
Modem Corporationand Private Pmpen+y,Chicago: Commerce Clearing House.

Connor, G., 1984, “A Unified Beta Pricing Theory,” Joumdof
Economic Thory 34, 13-3 1.

Black, Fischer, 1986, “Noise,“JoumalofFinance 41(3), 529-543.

DeLong, J. Bradford, et al, 1987, The Economic Consequences
of
Noise Trades, NBER Working Paper No. 2395.

Campbell, John Y., and Kyle, Albert, 1988, SmartMoney, Noise
Trading and Stock Price Behavior, Princeton University
Financial Research Center Memorandum No. 95.

Fama, Eugene F., 1970, “Efficient Capital Markets: A Review
of Theory and Empirical Work,” Journal of Finance 225(Z)
(May),,383-417.

Campbell, John Y., and Shiller, Robert J., 1987, “Cointegration
and Tests of the Present Value Model,” Joumai of Political
Economy 95(S), 10621088.

Fama, Eugene F., and Jensen, Michael C., 1983, “Separation
of Ownership and Control,” Journal of Low and Economics
26 (June), 301-325.

, 1988a, “The Dividend-Price Ratio and Expectations of Future Dividends and Discount Factors,” Rex&w
of Financing Studies 1, 195228.

Flavin, Marjorie A., 1983, “Excess Volatility in the Financial
Markets: A Reassessment
of the Empirical Evidence,”
Journal of Politico/Economy 91(6) (December), 929-956.

1988b, “Stock Prices, Earnings, and Expected
Dividend’
s,” &ma/ of Finance 43(3), 661-676.

Garber, Peter M., 1989, “Tulipmania,”
Economy 97(3) (June), 535-560.

FEDERAL

RESERVE

BANK

OF RICHMOND

Journal of Po/iticol

43

Greenwald, Bruce, and Stein, Jeremy, 1988, “The Task Force
Report: The Reasoning Behind the Recommendations,”
Journal of Economic Peqpectives Z(3) (Summer), 3-23.

Mackay, Charles, 1852, Memoirs of Extraordinaty Popuk~
Delusions and tke Madness of Crmz& 2d. ed., 2 vols.,
London: Office Nat. Illustrated Library.

Gort, Michael, 1969, “An Economic Disturbance Theory
Mergers,” QuarterlyJournal of Economics83, 624-642.

Manne, Henry G., 1965, “Mergers and the Market for Corporate
Control,” Journal of Pohical Economy 73, 110-120.

of

Hamilton, James D., 1986, “On Testing for Self-Fulfilling
Speculative Bubbles,” Intematiotzal
EconomicRevim 27(3)
(October), 54.5552.
Hamilton, James D., and Whiteman, Charles H., 1985, “The
Observable Implications of Self-Fulfilling Expectations,”
Journal of MonetaryEconomics 16, 353-373.
Harris, Milton, and Raviv, Artur, 1988, Th Design of
senrrities, Department of Finance, Kellogg Graduate School
of Management, Northwestern Univ., Working Paper No.
64.
Hermann, Werner, and Santoni, G.J., 1989, “The Cost of
Restricting Corporate Takeovers: A Lesson From Switzerland,” Federal Reserve Bank of St. Louis Review,
November/December,
3-l 1.
Jarrell, Gregg A.; Brickley, James A.; and Netter; Jeffry M.,
1988, “The Market for Corporate Control: The Empirical
Evidence Since 1980,” Journal of fionotnic Perspectives
2(l) (Winter), 49-68.
Jensen, Michael C., 1986, “Agency Costs of Free Cash Flow,
Corporate Finance, and Takeovers,” AmeriiGan Economic
Rewim 76(2) (May), 323-329.
Jensen, Michael C., and Meckling, William H., 1976, “Theory
of the Firm: Managerial Behavior, Agency Costs and
Ownership Structure,” Journal of Financiai Economics 3,
305-360.
Jensen, Michael C., and Ruback, R., 1983, “The Market for
Corporate Control: The Scientific Evidence,” Jouma/ of
Fiuankal Economics 11 (April), S-SO.

Marsh, Terry, and Merton, Robert C., 1986, “Dividend Variability and Variance Bounds Tests for the Rationality of
Stock Market Prices,” Amerikan Economic Revim 76(3)
(June), 483-498.
Mitchell, Mark L., and Netter, Jeffry M., 1989, “Triggering
the 1987 Stock Market Crash: Antitakeover Provisions in
the Proposed House Ways and Means Tax Bill?,” Joumaf
of Financbl Economics24, 37-68.
Mueller, Dennis C., 1987, The Corporation: Ghwtk, Dioenisicot;on and Mergen, New York: Harwood Academic
Publishers.
Samuelson, Paul A., 196.5, “Proof That Properly Anticipated
Prices Fluctuate Randomly,” IndiAFtriolManagement
Rewiew
6, 41-49.
Scherer, F. M., 1988, “Corporate Takeovers: The Efficiency
Arguments,” Journal of Economic Penpectives2( 1) (Winter),
69-82.
Shiller, Robert J., 1981, “Do Stock Prices Move Too Much
to be Justified by Subsequent Changes in Dividends?
Amehan Economic Rewie~ 7 l(3) (June), 42 l-436.
1984, “Stock Prices and Social Dynamics,”
Bmokings’
Paperson Economic Activity 2, 457-498.
Shughart, William S., II, and Tollison, Robert D., 1984, “The
Random Character of Merger Activity,” Rand Journal of
Economics 15, 500-509.

Kleidon, Allan W., 1986, “Variance Bounds Tests and Stock
Price Valuation Models,” Journal of Pohicai IGonomy
94(S) (October), 953-1001.

Singleton, Kenneth J., 1987, “Specification and Estimation of
Intertemporal
Asset Pricing Models,” in Handbook of
Monetary Economics, B. Friedman and F. Hahn, eds.,
Amsterdam: North-Holland.

Lacker, Jeffrey M.; Levy, Robert J.; and Weinberg, John A.,
1990, “Incentive Compatible Financial Contracts, Asset
Prices, and the Value of Control,” Joumai of Financial
Zntemzediation (forthcoming).
1,

Smith, Clifford W., Jr., and Warner, Jerold B., 1979, “On
Financial Contracting: An Analysis of Bond Covenants,”
Journ& of Financial Economics7, 117- 16 1.

LeRoy, Stephen F., 1989, “Efficient Capital Markets and
Martingales,” humai of Economic L.&rzr~m 27 (December),
1583-1621.
LeRoy, Stephen F., and Porter, Richard D., 1981, “The
Present-Value Relation: Tests Based on Implied Variance
Bounds,” Econometrica49(3) (May), 555-574.
Lucas, Robert E., Jr., 1978, “Asset Prices in an Exchange
Economy,” Economenica 46(6) (November), 1429-45.

44

ECONOMIC

REVIEW,

West, Kenneth D., 1987, “A Specification Test for Speculative
Bubbles,” Ibuorrerly
Journal of Economics 102(3) (August),
553-580.
, 1988a, “Dividend Innovations and Stock Price
Volatility,” Econometrica56( 1) (January), 37-61.
1988b, “Bubbles,
Fads and Stock Price
Volatilit; Tests: A Partial Evaluation,” Journal of Finance
43(3), 639-660.

MARCH/APRIL

1990

A Mandate

for Price Stability
Robeti L. Hem/’

I.
INTRODU~I~N
Stephen Neal, Chairman of the House Banking
Subcommittee on Domestic Monetary Policy, has
introduced legislation (H. J. Res. 409) requiring
that the Federal Open Market Committee of the Federal
Reserve System shall adopt and pursue monetary policies
to reduce inflation gradually in order to eliminate inflation
by not later than 5 years from the date of this enactment
of this legislation and shall then adopt and pursue monetary policies to maintain price stability.

This paper argues for passage of the Neal Resolution, which would make price level stability the dominant goal of monetary policy. The alternative to a
rule that mandates price stability is the exercise of
ongoing discretion over the desired price level. This
discretion, it is argued, encourages groups that benefit
from high and variable inflation to lobby the political
system. A rule is desirable primarily because it limits
the incentives for special-interest politics.
An earlier experience with discretionary monetary
policy occurred under the Articles of Confederation
(178 l-l 789). On the basis of this experience, James
Madison concluded that discretion creates political
pressures
from special interest constituencies.
Madison and the other authors of the Constitution,
therefore, took discretionary control over. the price
level away from government. Article I, Section 8 of
the Constitution
empowered
Congress to “coin
money” and “regulate the value thereof.” Today, this
language appears general. At the time, however, it
was clearly understood as restricting Congress to
specifying the metallic content of coins. [See
Timberlake (1989) and Christainsen (1988), especially the references in footnote 2 of the latter paper.]
The first part of the paper reviews the importance
the authors of the Constitution placed on constraining discretionary issue of paper money. The second
part of the paper argues that the recent experience
with discretion vindicates Madison’ judgment that
s
discretion nurtures special-interest politics. In replacing discretion with a rule, the Neal Resolution would
Economist and Vice President, Federal Reserve Bank of Richmond. The author gratefully acknowledges helpful criticism from
Milton Friedman, J. Huston McCulloch, and from colleagues
at the Richmond Fed.
l

FEDERAL

RESERVE

reestablish the original intent of the authors of the
Constitution and return price level determination to
a constitutional framework.

II.
PRICELEVELDETERMINATIONWITHIN
A CONSTITUTIONALFRAMEWORK
By 1787, James Madison and his correspondents,
including James Monroe, George Washington, and
Edmund Randolph, had concluded that the ascendancy of parochial political interests over the national
interest was spreading disorder and leading to a
disintegration of the Union. A primary manifestation
of these parochial interests was overissue of paper
money. State legislatures were pressured by debtors
to pass laws making paper money legal tender and
then to issue large amounts of it. By 1786, seven
states had adopted paper money as legal tender.
Madison wrote to his brother on August 7, 1786
(Madison 1975, p. 89):
. . . the States are running mad after paper money, which
among other evils disables them from all contributions of
specie for paying the public debts, particularly the foreign
one. In Rhode Island a large sum has been struck and
made a tender, and a severe penalty imposed on any
attempt to discriminate between it and coin. The consequence is that provisions are withheld from the Market,
the Shops shut up-a general distress and tumultuous
meetings.
Shortly thereafter, he wrote to Thomas Jefferson
complaining of the “warfare & retaliation” among
states that were passing laws enabling their citizens
to pay out-of-state debts in depreciated paper money
(Madison 1975, pp. 94-S).
In Spring 1787, Madison wrote the memorandum
“Vices of the Political System of the United States”
in preparation for the Federal Convention to be held
at Philadelphia in May. In “Vices” Madison addressed
the problem of how to prevent a national legislature
from following the examples set by state legislatures,
where majorities had violated the rights of individuals
and minorities. Madison first described how unrestrained majority rule encouraged
majorities to
exploit minorities (Madison 1975, pp. 354-5):
These causes lie 1. in the Representative
the people themselves.
BANK OF RICHMOND

bodies.

2. in

45

1. Representative appointments are sought from 3 motives.
1. ambition. 2. personal interest. 3. public good. Unhappily the two first are proved by experience to be most
prevalent.
2. A still more fatal if not more frequent cause lies among
the people themselves. All civilized societies are divided
into different interests and factions, as they happen to be
creditors or debtors-rich or poor-husbandmen,
merchants
or manufacturers-members
of different religious sectsfollowers of different political leaders-inhabitants of diierent
districts-owners
of different kinds of property &c &c. In
republican Government the majority, however composed,
ultimately give the law.

Madison argued that appeals made on the basis
of the “general and permanent good of the Community, ” “character,” or “religion” would do little to
prevent majorities formed out of these special interest
groups from exploiting minorities (Madison 1975, pp.
355-6):
Is it to be imagined that an ordinary citizen or even an
assembly-man of R. Island in estimating the policy of
paper money, ever considered or cared in what light the
measure would be viewed in France or Holland; or even
Massts or Connect.? It was a sufficient temptation to both
that it was popular in the State; to the former that it was
so in the neighbourhood. . . . Place three individuals in a
situation wherein the interest of each depends on the voice
of the others, and give to two of them an interest opposed
to the rights of the third. Will the latter be secure? The
prudence of every man would shun the danger. The rules &
forms of justice suppose and guard against it. Will two
thousand in a like situation be less likely to encroach on
the rights of one thousand? The contrary is witnessed by
the notorious factions & oppressions which take place in
corporate towns limited as the opportunities are, and in
little republics when uncontrouled by apprehensions of
external danger.

interests the defect of better motives” (TireFedwahit
No. 51).
The Constitutional Convention ended the discretion of state legislatures over the price level and the
issue of paper money. Article I, Sec. 10 of the Constitution states that “No state shall . . . coin money;
emit bills of credit [paper money]; make anything
but gold and silver coin a tender in payment of debts.”
Article I, Sec. 8 gave the Federal government the
power “to coin money, regulate the value thereof,
and of foreign coin, and fii the standard of weights
and measures.” To the framers of the Constitution,
this language clearly committed the United States
to a specie standard.’
In many states during the Confederation period,
state legislatures had arbitrarily set aside commercial contracts. Through inflation caused by printing
paper money, states had abrogated contracts in favor
of debtors. Article I, Sec. 10 of the Constitution prohibits states from “impairing the obligation of contracts.” (Later, in the same spirit, the Fourteenth
Amendment stated “nor shall any State deprive any
person of life, liberty, or property without due process of law.“) Removing discretionary control over
the price level from government was a key device
for enforcing the principle that government should
not impair contractual obligations.

The Society becomes broken into a greater variety of
interests, of pursuits, of passions, which check each other,
whilst those who may feel a common sentiment have less
opportunity of communication and concert.

The authors of the Constitution carefully compromised between the need to give government the
power to raise revenue and the need to protect private
property from arbitrary seizure. The Constitution
separates the branch of government that spends
public monies from the branch that levies taxes. It
safeguards this separation by giving Congress exclusive rights “to borrow money on the credit of the
United States.” The Executive Branch cannot spend
money “but in consequence of appropriations made
by law.” By reserving to Congress the power to tax,
the authors of the Constitution ensured that the
exercise of this power would be accompanied by
public discussion. Furthermore,
“bills for raising
revenue shall originate in the House,” whose

Inevitably, citizens will form political groups in an
attempt to use the coercive power of the state to further their own self-interests, rather than the general
interest. In Th Fedmakt No. 10, Madison accepts
the reality of factionalism in government promoted
by self-interest. The separation of powers, checks
and balances, and the federal system embodied in
the Constitution were designed to restrain selfinterest through “supplying by opposite and rival

i Christainsen (1988, p. 427) writes: The first draft of the Constitution gave the legislature of the United States the power to
“emit bills” [paper money]. On August 16, 1787, however, the
convention moved to strike this power from the Constitution,
and in Madison’ account, “striking out the words . . . cut off
s
the pretext for a paper currency, and particularly for making the
bills a tender either for public or private debt.” Of the eleven
delegates whose remarks Madison reported, ten clearly put forth
the view . . . that striking the phrase in question would deny
Congress any power, under any circumstances, to create paper
money.

Madison concludes by expounding the famous idea
of Essays No. 10 and No. 5 1 in Tire Federaht. In
a national legislature in a large country, the general
interest is protected because the large numbers of
disparate groups make it difficult to form exploitive
majority coalitions (Madison 1975, p. 357):

46

ECONOMIC

REVIEW,

MARCH/APRIL

1990

members were subject to elections every two years.
A specie standard was one of the checks imposed
to assure’ that taxes were imposed only through
explicit.legislation. Congressional responsibility “to
coin money” was designed to prevent the Executive
Branch from copying the behavior of sovereigns who
levied taxes through debasement of the coinage..

actual fact, changes in the .money stock far exceeded changes in money demand.2 Discretion was
exercised primarily in trading off the goal of price
stability against other goals.

Iii.

Inflation generates revenue directly through the
increase in fiat money that creates the inflation. More
important, inflation ‘
interacts ‘
with the lack of indexing in the tax code to increase tax revenue. Finally,
unanticipated inflation reduces the real value of the
taxes the government must impose to pay holders
of existing government debt..

THE RECENT EXPERIMENTIN DISCRETION
Although the specie standard lapsed under the
pressure to finance the Civil War with greenbacks,
it was reestablished in 1878. When the Federal
Reserve System was established in 191.3, it was subjected to the discipline of the gold standard. Federal
Reserve notes were subject to a 40 percent gold
reserve. Battered by the Debression and two world
wars, the gold standard metamorphosed
into the
Bretton Woods system, under which the Federal
Reserve felt constrained to raise interest rates in
response to gold outflows. Because domestic inflation was viewed as the major cause of gold outflows,
the Federal Reserve kept inflation at a low level. To
a considerable degree, the Bretton Woods system
limited government discretion over the price level.
This limitation on discretion began to break down
in the 1960s however, when the Federal Reserve
System stopped raising interest rates to prevent gold
outflows. In 1963, Allan Sproul(1980, pp. 12 1, 126),
president of the Federal Reserve Bank of New York,
made an early, eloquent plea for discretion:
[The Federal Reserve Act] was a determination that there
was to be a degree of monetary management in the United
States. But because of ancient prejudices and still lively
suspicions . . . it was thought that this power could be
substantially divorced from acts of discretion. . . . Changes
in the production of gold, the international balance of
payments, and the rise and fall of the self-generated credit
needs of agriculture, commerce, and industry were to
determine, pretty largely, the amounts of Reserve Bank
credit which would come into being or go out of existence.
. . . It seems to me patent that the uncertain hand of man
is needed in a world of uncertainties and change and
human beings, to try to accommodate the performance of
the monetary system to the needs of particular times and
circumstances and people. I here agree with Professor
Samuelson, of the Massachusetts Institute of Technology,
who has written that “a definitive mechanism, which is to
run forever after, by itself, involves a single act of discretion which transcends, in both its arrogance and its capacity for potential harm, any repeated acts of foolish
discretion that can be imagined.”

Later, discretion came to be defended primarily
as allowing the Federal Reserve to vary the money
stock in line with changes in money demand. In
FEDERAL

RESERVE

IV.

THEINFLATION TAX

After 1964, the political system was under constant pressure to increase revenue. The 1964 general
election provided the congressional votes to undertake a broad expansion of income redistribution programs.3 Two years later, the Vietnam War defense
buildup began. After the mid-1960s, a rapidly
growing economy that would generate continuous increases in revenue for defense and domestic spending programs became a dominant political concern.
Initially, the political system accepted inflation as the
cost of high real growth and the government revenue
generated by that real growth. Later, the political
system came to depend directly upon inflation for
revenue.
Before indexation in 1985, inflation increased the
real revenue raised by the personal income tax. Inflation pushed individuals with unchanged real income
out of tax-exempt into taxable status. It eroded the
real value of the standard deduction. Most important, due to the progressive rate structure of the personal income tax, inflation increased real revenue by
moving individuals with unchanged real income
into higher marginal tax brackets. Inflation still
2 From 1965 to 1989, real GNP doubled. Because the public’
s
demand for the purchasing power represented by M2 rises in
line with real GNP. the demand for real M2 also doubled. In
contrast, the stock of M2 rose sevenfold. According to the ouantity theory, the excess supply of M2 should cause %e price ]evel
to rise bv a factor of 3.5 (7/Z = 3.5). Over the oeriod 1965
to 1989, the implicit price deflator increased by almost exactly
that factor.
3 In the election, Democrats had campaigned for a national
medical care program (Medicare) and a Social Security program
with universal coverage. In contrast, Republicans had campaigned for Social Security coverage limited to the needy
elderly and financed out of general revenues. The elections gave
the Democrats a 295-146 majority in the House and a-net
increase of 4’ Northern Democrats. The conservative coalition
2
of Republicans and Southern Democrats that had blocked social
legislation in the 1950s crumbled.
BANK OF RICHMOND

47

increases revenue through the absence of indexation
in other parts of the tax code. The capital gains tax
is levied not only on real gains, but also on paper
gains that only compensate for inflation. Revenue
from estate taxes rises as inflation lowers the real
value of the estate tax exemption. Inflation raises
corporate taxes by eroding the real value of depreciation allowances, which are based on historical cost,
rather than replacement cost. It also raises corporate
taxes through increases in the dollar value of inventories that augment measured profits, but not real
profits.
Studies done for the year 1974, when the inflation rate was 11 percent, yield the conclusion that
inflation increased federal tax revenue in that year
by 17 percent. (See Appendix. Because of the complexity of the federal tax code, construction of an
annual series on revenue increases produced by inflation would require considerable work.) Although
the revenue raised by inflation varied over time with
the inflation rate, this revenue contributed significantly to total revenue until the reduction in the
inflation rate in the 1980s and the indexing of the
personal income tax in 1985.

V.

INCOMETFUNSFERSAND

INFLATION

The combination of inflation and government price
fiing allows the political system to circumvent legal
prohibitions against arbitrary confiscation of private
property. Revenue transfers imposed by this combination are not subject to the checks and balances
and public discussion that constrain the enactment
of explicit tax legislation. By reducing public discussion, such transfers avoid criticism for providing
benefits to groups that are well-off. The relative ease
of effecting income transfers through government
price fting in an inflationary environment encourages
the formation of special-interest lobbies. Inflation thus
increases the incentive to use government-regulated
prices to redistribute income.
After the mid-1960s, in response to pressure from
the politically potent housing lobby, Congress increasingly subsidized credit to the housing industry.
In September 1966, Congress passed legislation extending interest rate ceilings to S&Ls. These Regulation Q ceilings, administered jointly by the Fed, the
FDIC, and the FHLBB, were set at a higher level
for S&Ls than for banks. The original intention was
to allocate credit directly to housing by making
deposits more attractive at S&Ls than at banks.
48

ECONOMIC

REVIEW,

Because Reg Q ceilings were not raised with the rise
in inflation and market rates after 1966, Reg Q
became an instrument for transferring income from
holders of small deposits to the housing industry.4
Holders of small deposits, who did not have access
to money market instruments paying a competitive
rate of return, were in effect taxed at a rate equal
to the difference between the market interest rate
and the Reg Q ceiling rate.
Reg Q ceilings subsidized credit to housing by
keeping interest rates on thrift deposits below
market rates. In combination with the prohibition
of adjustable-rate mortgages, these ceilings constrained thrifts to borrow short-term
through
passbook savings accounts, while making them lend
long-term. The rise in inflation in the late 1970s and
early 1980s produced a rise in market rates and in
the rates at which thrifts borrowed. Their old mortgages, however, continued to pay the lower rates
offered in the less inflationary past. Consequently,
a majority of thrifts became insolvent. In the absence
of inflation, there would have been no thrift crisis.5
The Nixon wage and price controls, imposed in
August 1971 in response to 4 percent inflation,
created extensive new opportunities for the political
system to redistribute income among different groups
without explicit legislation. Inevitably, administration
and enforcement of wage and price controls require
considerable discretion. Wage and price controls
create a shadow fiscal system of implicit taxes and
transfers.
The controls on the energy industry were a good
example of how the political system combined inflation with legislated price fling to redistribute income.
Price controls on oil were kept after other price controls were eliminated. In his book review of Th
Eio~omics and Politicsof Oil Price Regdation, Henry
Jacoby (1984, p. 1176) comments:
When the first oil shock occurred there was a system of oil
price controls already in place-a hangover from the Nixon
anti-inflation scheme of 1971. They were modified and
4 The ceiling rate on commercial bank savings deposits was set
at 4 percent in 1966, 4.5 percent in 1970, 5 percent in 1973,
and 5.25 percent in 1979. In contrast to this 1.25 percentage
ooint rise from 1966 to 1979. over the same oeriod. the threemonth Treasury bill rate rose’
almost 5 percentage pbints, from
about 5 percent to 10 percent. In May 1970, this inflation tax
was effectively restricted to holders of small deposits as a result
of the exemption from Reg Q ceilings of certificates of deposit
in denominations of $100,000 or greater.
5 Because deposit insurance allowed insolvent thrifts to continue
to attract deposits, the decision whether to close an insolvent
thrift became a political decision rather than a market decision.
MARCH/APRIL

1990

extended and used to hold down the price of domestic
crude oil so that people downstream (oil refiners, distributors, and the ultimate consumers) got a lower average price
of domestic-plus-imported
supplies. . . . A shadow system
of public finance, unique to the oil sector, was createdcomplete with taxes, transfers, and (no surprise) deadweight
loss. In practice the system grew to mind-bending complexity as the various players (regions, consumers, refiners, and
producers holding various classes of oil reserves) fought
over the goodies.6
A very contentious issue at the time . . . was the question
who actually benefited from the $15$45 billion (depending
on the year) producers were denied. In the mid-l 970s there
was a group of analysts who held that the oil price controls
were a fraud to the consumer: U. S. product prices were
set in world product markets . . . and there was no way
for controls on crude oil to affect prices at the pump. The
rents were being transferred to refiners in the form of
increased margins.

Rent control laws furnish another example of the
way inflation combines with government-regulated
prices to redistribute income, in this case, from the
owners of the housing stock to renters. Consider also
automobile insurance: in California, Proposition 103,
which was passed in a 1988 referendum, called for
a rollback in automobile insurance rates of ‘ per20
cent. The constitutionality of the rollback is now
being litigated in the courts. Proposition 103 also
mandated that the state’ insurance commissioner be
s
elected in the future. Given the extensive criticism
of the cost of car insurance in California, it is
unlikely that the next commissioner will raise rates
after taking office. Inflation will then lower the real
value of insurance rates, regardless of whether the
courts sanction a rollback.

VI.

EROSION OF SUPPORT FOR THE
PRICE SYSTEM
Inevitably, in an inflationary environment, government officials blame inflation on the special factors
that change individual prices. In an environment
where no one accepts responsibility for inflation, competition for political power encourages inflation
scapegoating, which plays on public confusion over
“high” and “rising” prices by attributing inflation to
monopoly power. This scapegoating in turn erodes
public support for resource allocation through ‘
the
price system.
6 Ironically, when the extent of pollution in Communist countries appeared in 1989, the price system of western countries
was praised for having produced efficient use of energy. An
article in the Nm Y& 7Imes (l/23/90, p. 17) commented, “The
lack of market forces kept these [Communist] countries from
realizing the impressive gains in energy efficiency registered in
the West after the oil shocks of the Seventies. . . .”
FEDERAL

RESERVE

Erosion of support for resource allocation through
the price system was especially strong in the market
for home construction. The cycle of inflation and
recession that began in the mid-1960s induced
cyclical boom and bust conditions in the home construction market. (Housing construction, like other
forms of investment, falls more sharply than aggregate
output in a recession.) Cyclical downturns in the
housing and construction industry created the impression that the free-market allocation of credit
discriminated against specific classes of users. In
particular, the concentration of unemployment in the
construction industry created the impression that construction workers had to bear a disproportionate share
of the burden of reducing inflation.
Because downturns in housing construction were
attributed to “high” interest rates, they created
pressure for “cheap” credit.’ Many believed that
lower interest rates for housing would follow from
an increase in the supply of credit to housing made
possible by higher money growth. In response to constituent pressure, some congressmen pressured the
Fed for higher money growth and lower interest rates.
These congressmen blamed financial monopolies for
“high” interest rates. “High” interest rates, they
argued, .exacerbated inflation by raising the cost of
doing business. In 1975, the cyclical downturn in
housing produced House bills that would have required the Fed to set a floor of 6 percent under Ml
growth and “to allocate credit away from inflationary
uses, and toward national priority uses, including
low- and middle-income housing” (HR 3 161).8
Rep. Jim Wright (US Cong., Z/4/75, p. 7) made the
case for one such bill, HR 2 12, produced by the
Democratic Steering and Policy Committee.
REP. WRIGHT: With any given supply of new money
overall, a credit allocation program is needed to channel
credit away from nonproductive speculative and inflationary
uses, such as corporate takeovers, excessive inventory
accumulation, and speculation in land and commodities, and
toward credit-starved priority areas of the economy. . . .
HR 212 requests the Federal Reserve to allocate credit
toward priority uses and away from nonpriority speculative
and inflationary uses.

7 Congress was especially sensitive to this pressure because
increases in deficits during recessions created the appearance
that government was the main competitor for housing credit.
* Treasury Secretary Simon, along with influential members of
the Senate Banking Committee, opposed these bills. As a
consequence, they emerged in amended form as House Concurrent Resolution 133. which reauired onlv that the Fed
periodically consult with Congress “over ranges of growth or
diminution of monetary and credit aggregates.”
BANK

OF RICHMOND

49

Fed chairman Arthur Burns countered these assertions with arguments that inflation arises from government deficits and monopoly power in labor markets.
Under pressure to lower interest rates, he defended
money markets as highly competitive:
SEN. BIDEN: Doctor, on occasion you have also indicated
that with regard to interest rates, either the Fed can’ or
t
shouldn’ concentrate on lowering interest rates. Yet we
t
are faced with that question all the time here in the Cont
t
gress. . . . If the Fed can’ or shouldn’ be the outfit that
concentrates on that, who should?
DR. BURNS: You know, you could leave interest rates
alone. After all, we have highly competitive money and
capital markets. If you are going to engage in price control
exercises, you ought to turn to those sectors of the economy where there are pockets of monopoly. . . . We have
pockets of monopoly in the field of labor, but we don’
t
talk about that. (US Cong., 4/29/75, p. 18)

As inflation created public distrust of the price
system, it also created opportunities to subsidize
users of credit. Rising rates of inflation that pushed
market rates above usury ceilings provided a subsidy
to homeowners who obtained mortgages at belowmarket rates. Homeowners with existing mortgages,
like other debtors, benefited from unexpectedly high
inflation. Furthermore,
inflation turned existing
federal credit programs into subsidies for the home
construction industry. These programs had existed
before the inflation of the mid-1960s. The rationale
for them was that they made “it possible for home
owners and rental project owners to finance the construction or acquisition of housing properties at
reasonable (italics supplied) levels of interest rates”
(US Cong., Z/28/64, p. 22). The credit extended by
these programs before 1965 was relatively small, and
it was largely extended at market rates. [See US
Cong., Z/28/64, Table 3-Z.] With inflation, “reasonable” levels of interest rates became historical levels
of interest rates, and “reasonable” rates became subsidized rates.
By lessening public acceptance of credit allocation
by the marketplace and by increasing the ease of
hiding subsidies, inflation encouraged myriad government interventions in the market for housing credit.
These interventions disguised the social cost of housing, which led to a misallocation of the capital stock.
Government intervention also produced the HUD
scandals and the S&L bailout of the 1980s.

VII.
POLITICALSELF-INTEREST
AND
THE COMMON INTEREST
Revenue generated
increase in government
50

by inflation financed an
spending relative to GNP
ECONOMIC

RNIEW,

after the mid-1960s. Because this increase in revenue
did not have to be explicitly legislated, it allowed
postponement of a political consensus over the acceptability of the increased spending. Prior to indexation of the personal income tax in 1985, inflation
continuously increased tax revenue as a percent of
GNP. Periodic “tax cuts” would return revenue as
a percent of GNP to its original base value. The practice of imposing continuous tax increases through
inflation, while legislating offsetting reductions only
occasionally, raised the average tax rate imposed over
time. The increase in the average tax rate allowed
Congress to raise taxes sufficiently to finance the
expansion of income transfer programs, while postponing a decision on whether to legislate permanently
taxes sufficient to pay for them. Inflation allowed
Congress to postpone continually its constitutional
responsibility to make explicit, publicly debated
decisions on the share of resources to appropriate
to the public sector.
The distortions produced by continual inflation and
the absence of indexing in the tax code gave Congress an incentive to rewrite the tax code periodically.
Individuals and corporations necessarily lobbied Congress on an ongoing basis to protect their own interests. The uncertainty over the long-run incidence
of taxes acted to discourage investment.

VIII.
CAN WE LEARN To LIVE
WITH INFLATION?
Is “high” inflation bad and “moderate” inflation all
right? Why not learn to live with the current 5 .percent inflation? Historical experience offers no example
where positive inflation was maintained at a steady
rate over any significant period of time. Sustained
inflation is always associated with a fluctuating rate
of inflation. The reason is that, in an inflationary
environment, the incentive for the political system
to inflate changes continually. First, the revenue
raised with a given rate of inflation tends to fall
because the public finds ways to reduce the base of
the inflation tax. For example, the revenue generated
in the 1970s by inflation and the lack of indexing
in the corporate income tax fell as firms shifted from
long-term to short-term investments, which could be
depreciated over a short time period. Second, the
income transfers to politically influential constituencies produced by the combination of inflation and
price controls tend to fall as the public finds ways
to circumvent the price controls. For example, in
the 1970s money funds allowed individuals to bypass Reg Q by holding money market instruments
MARCH/APRIL

1990

indirectly. With a given rate of inflation, therefore,
the revenue raised and the income transfers effected
by inflation fall over time. Political pressures to offset this fall through an increase in the inflation rate
create instability in inflation.
Finally, because the size of the federal government
deficit varies with changes in the rate of growth of
output, a concern over government deficits produces
pressure for expansionary monetary policy. In the
absence of a clear mandate to stabilize the price level,
large government deficits will continue to create
political pressures for the inflationary monetary policy
that has characterized the last three decades.

has avoided allocating credit among competing
private uses. The primary manifestation of the rule
not to allocate credit is an unwillingness to allow in‘
solvent financial institutions to use the discount window. Use of the discount window by insolvent financial institutions would move credit allocation away
from its free market allocation. Again, this rule has
worked well. It is evident that if either rule were made
subject to exceptions, the Federal Reserve System
would come under regular political pressure to make
exceptions. Hopefully, passage of the Neal Resolution will make price level stability a rule that is
followed with no exceptions.

Ix.

REFERENCES

CONCLUSION
The only way to assure a stable monetary environment is to replace the exercise of ongoing discretion
over the desired price level with a rule that makes
price level determination part of the constitutional
framework of government. In a recent editorial, Th
Financial Zhes of London (l/23/90, p. 16) stated,
The notion that money must fall within the domain of dayto-day politics is a ZOth-century heresy. . . . Painful
experience with the modern manipulation of monetary
policy suggests that money is more appropriately an element
of the constitutional framework of democracy than an object
of the political struggle. Monetary stability is a necessary
condition for a working market economy, which is itself a
basis for a stable democracy.

The purpose of a rule is to reduce the incentive
for special-interest constituencies to form with the
goal of either redistributing income through the
political system in a way that does not reflect a social
consensus explicitly ratified through the legislative
process or of redistributing income in an arbitrary way
away from minority groups. This rationale for a rule
means that a rule must be exactly what its name
implies-a guiding principle with no exceptions. The
central bank cannot condition the political system
to respect its independence if politicians know that
the central bank makes exceptions to its rules.
This argument has wider application than just to
a rule for price level stability. For example, unlike
most other central banks, the Federal Reserve
System has never interfered in the foreign exchange
market by allocating foreign exchange at favorable
rates to politically influential importers. This rule has
worked well. Similarly, the Federal Reserve System

FEDERAL

RESERVE

Black, Robert P. “In Support of Price Stability.” Federal Reserve
Bank of Richmond Economic Review 76 (January/February
1990), 3-6.
Christainsen, Gregory B. “Fiat Money and the Constitution:
A Historical Review.” In PO&al Bushes Cych. Edited by
Thomas D. Willet. Durham: Duke University Press, 1988,
424-34.
Financk/ Times. “Pohl Throws a Gauntlet.” January 23, 1990.
Jacoby, Henry D. “Review of Th Economic and Pohhks of Oil
P&e Reguhtion,” by Joseph P. Kalt. In Journal of PofitkaL
Economy 92 (December 1984), 1176-78.
Madison, James. Th Papers of James kadion, vol. 9. Edited
by Robert A. Rutland. Chicago: The University of Chicago
Press, 1975.
Nkw York Times, “Report Warns
Europe.” January 21, 1990.

of Pollution

in Eastern

Sproul, Allan. “Money Will Not Manage Itself.” In S&ted
Popen of Allon Sproul. Edited by Lawrence S. Ritter.
New York: Federal Reserve Bank of New York, December
1980.
Timberlake,
Richard H. “The Government’
s
License
’ Create Money.” Cato Journai 9 (Fall 1989), 29-49.

to

U. S. Congress, House, Subcommittee on Domestic Finance
of the Committee on Banking and Currency. A S&y of
Fe&al Credit Programs, vol. 1, 88th Cong., 2d sess.,
February 28, 1964.
U. S. Congress, House, Subcommittee on Domestic Monetary
Policy of the Committee on Banking, Currency and
Housing. An Act to her
Intemu Rates and A&uate Credit,
Hearings, 94th Cong., 1st sess., February 4, 5, and 6,
1975.
U. S. Congress, Senate, Committee on Banking, Housing and
Urban Affairs. Fint Meeting on the Gmduct of Monetary
Policy, Hearings, 94th Cong., 1st sess., April 29, 30 and
May 1, 1975.

BANK OF RICHMOND

51

Appendix on Revenue from Inflation

This appendix reviews quantitative estimates of
five separate increases in federal revenue in 1974 due
to the inflation that year of 11 percent.

Added Seigniorage: The outstanding stock of
base money (currency in circulation, foreign and other
deposits at the Fed, and member bank reserves) in
1974 was $111 billion. With inflation at 11 percent
in 1974, the public had to add an additional 11 percent to holdings of base money in order to maintain
its real value. (This addition to base money is
equivalent to a tax collected by the government in
that it allows the government to finance additional
expenditures.) Seigniorage in 1974, therefore, can
be put at about $12.2 billion ($111 x .ll).
Lower Real Interest on Outstanding
Treasury Debt: As of June 1974, the Treasury paid
an average rate of interest of 6.56 percent on its
outstanding debt. At this time, the average maturity
of this debt was 3 years. The market rate of interest
on a 3-year Treasury note was 8.33 percent. The
difference in the market rate and the average rate
paid (1.77) is an estimate of the extent to which past
issues of federal debt failed to incorporate adequately
a premium for future inflation. With $254.5 billion
of debt held by private investors, the gain to the
government from unanticipated inflation in 1974 was
$4.5 billion (.0177 x $254.56).

Income Tax Bracket Creep: Before the indexing that took effect in 1985, inflation increased the
real revenue raised by the personal income tax. Inflation eroded the real value of the standard deduction, the personal exemption, and the low-income
allowance. Because the rate structure of the personal
income tax was progressive before 1985 with respect
to nominaf income, inflation increased real revenue
by increasing individuals’ nomirza~
income. Fellner,
Clarkson and Moore (1975) use a stratified sample
of tax returns from the Internal Revenue Service in
order to calculate the increase in revenue in 1974
due to inflation. They apply the actual tax code in
1974 to these returns and also a hypothetical tax code
whose nominal provisions are adjusted upward by the
rate of inflation in 1974. They conclude that inflation in 1974 increased revenue from the personal
income tax by $6.7 billion.
This figure is fairly close to a rough estimate from
aggregate figures. Between 1973 and 1974, nominal
personal income increased 9.7 percent. Inflation
52

ECONOMIC

REVIEW.

(measured by both the CPI and the consumption
expenditures deflator), however, rose by 11 percent,
so real income declined by about 1 percent. An
indexed tax code that caused changes in real revenue
to reflect only changes in real personal income, then,
would have produced an increase in nominal personal
tax receipts of .about 8.7 percent (9.7 percent - 1
percent). In fact, personal tax receipts rose by 14.3
percent. These figures suggest an elasticity of real
revenue from the personal income tax with respect
to inflation of .64 [(14.3 - 8.7)/8.7]. In 1973,
personal tax receipts were $107.3 billion. The real
tax increase due to inflation, then, was about $6
billion ($107.36 x .087 x .64), which is close to the
Fellner et al. figure.

Nominal Capital Gains Taxation: Inflation increases the real revenue raised by the capital gains
tax because increases in the dollar value of assets due
to inflation are taxed as real rather than nominal
gains. Feldstein and Slemrod (1978) estimate that
inflation caused the tax on capital gains to generate
an additional revenue of $.5 billion in 1973. (This
figure is a lower estimate of the revenue gain for 1974,
when the inflation rate was higher than in 1973.)
Corporate Income Tax: Inflation raises the real
revenue from the corporate income tax. Fellner,
Clarkson and Moore (1975) also calculate the increase in corporate taxes in 1974 due to inflation.
In these calculations, they adjust corporate depreciation allowances for inflation, so that depreciation is
at replacement cost, rather than historical cost. They
also reduce profits due to the nominal gain in the
dollar value of inventories caused by inflation. They
estimate that inflation increased corporate taxes in
1974 by $10 billion. [This figure may be an
underestimate.
Feldstein and Summers (1979)
estimate that inflation in 1977 of only 6.8 percent
increased the taxes of nonfinancial corporations by
$32 billion. That is, in 1977, inflation raised the
effective corporate tax rate from 41 percent to 66
percent.]
Totals: The shares of the inflation tax contributed
by the separate parts of the tax code in 1974 were
seigniorage 36.0 percent., depreciation of existing
government debt 13.3 percent, personal income tax
excluding capital gains 19.8 percent, capital gains
1.5 percent, and corporate tax 29.5 percent. These
relative shares, however, underestimate the importance of the personal income tax component of the
MARCH/APRIL

1990

inflation tax. A constant inflation rate would generate
the same amount of revenue each year from the other
components (abstracting from reductions that occur
as the public learns how to evade the inflation tax).
In contrast, revenue increases from the personal income tax were cumulative because each year taxpayers were forced into higher tax brackets. The
cumulative increase in revenue was only limited
because taxpayers could not be forced into a marginal
tax bracket higher than 70 percent.
The figures listed above for the separate components of the inflation tax add to $33.9 billion. That
is, if the tax code had been indexed for inflation in
1974, federal revenue would have been lower by
$33.9 billion. In 1974, federal government revenue,
exclusive of social security taxes, was $198 billion.
In 1974, therefore, 17 percent of revenue was derived
from inflation. Of course, Congress reduced tax rates
on an ad hoc basis to keep the overall tax burden
relative to GNP fairly constant. These reductions,

however, occurred only sporadically. The steady
increase in real revenue produced by inflation combined with occasional reductions in tax rates raised
the average tax rate over time.

REFERENCES
Feldstein, Martin and Joel Slemrod. “Inflation and the Excess
Taxation of Capital Gains on Corporate Stock,” National
Tax Journal 31(2), June 1978, 107-18.
Feldstein, Martin and Lawrence Summers. “Inflation and the
Taxation of Capital Income in the Corporate Sector,”
National Tax Journal 32(4), Dec. 1979, 445-70.
Fellner, William, Kenneth W. Clarkson and John H. Moore.
“Correcting Taxes for Inflation,” Washington DC: American
Enterprise Institute for Public Policy Research, 1975.
Slemrod, Joel F. and Martin J. Feldstein. “Inflation and the
Excess Taxation of Capital Gains on Corporate Stock,”
National Tar Journal 31, June 1978, 107-l 18.

Maintaining

Price Stability: A Proposal
Robert L. He&

The Neal Resolution would make price stability
the dominant goal of monetary policy. This paper
proposes giving the Fed and Fed-watchers a measure
of whether ongoing policy is consistent with this goal.
This measure would require the Treasury to issue
two kinds of bonds at each maturity:

A Standard Bond: As presently issued, interest
and principal are paid in current dollars. The yield
equals a real (inflation-adjusted) yield plus the
inflation expected by the market.

An Indexed Bond: On this new bond, interest
and principal
changes in a
be of constant
indexing, the
incorporating

payments would be adjusted by
price index; thus payments would
purchasing power. Because of this
yield would be a straight real yield,
no inflation premium.
FEDERAL

RESERVE

The difference in yields on the two kinds of bonds
would offer a measure of expected inflation over the
life of the bonds.’
Investors holding fixed-income securities
an incentive to forecast inflation accurately;

have
their

r This proposal is similar in spirit to one made by Alan Greenspan
(1981), who advocated issuance of a five-year Treasury note with
interest and principal payable in gold. Milton Friedman (1974)
has long advocated indexing of all government bonds on ethical
grounds. He objected to the experience of the 1960s and 1970s
in which the government issued bonds that promised to pay
dollars in the future and then inflated away the real value of the
promised dollars. Assar Lindbeck (1989, p. 498, fn. 4) has
proposed the issuance of indexed bonds in order to permit
observation of changes in money growth on ex ante, as
opposed to realized, real rates of interest. Humphrey (1974)
discusses earlier proposals for indexed bonds, for example, proposals made by Keynes in 1924, Bach and Musgrave in 1941,
and Friedman in 195 1.
BANK

OF RICHMOND

53

consensus forecast, however,
is not signalled
clearly by market rates because these rates embody
a changing estimate of the expected real yield. Comparing the yields on standard and indexed bonds
would costlessly and continuously indicate the inflation expected by investors.
The market’ reaction to monetary policy actions
s
would be reflected in the yield spread between standard and indexed bonds. The advantage of such a
measure can be seen in the publicity accorded the
exchange rate in relatively small, open economy
countries. Headlines in a recent edition of the Financial Times of London (l/31/90, p. 3) read: “Canada
Puts Brakes on Interest Rates Fall: dollar plunge
brings caution to easing up on inflation fight.” The
article states:
An unexpected plunge in the Canadian dollar has strengthened the view that an abrupt fall in domestic interest rates
earlier this month will not be sustained. . . . The tumble in
the Canadian dollar caused by the relatively small fall in
interest rates reinforces a widely-held view that the Bank of
Canada’ watchword is likely to be caution. The central
s
bank responded to the sudden weakness in the currency
by pushing short-term rates up. By Tuesday this week,
the yield on three-month Treasury bills had climbed back
at 12.3 percent, compared with 11.9 percent when the
bank sent its initial signal that it was ready to relax its
interest-rate policy.

As this article shows, the Canadian central bank
is constrained by the behavior of the exchange rate.
Investors holding fixed-income assets denominated
both in U. S. dollars and in Canadian dollars make
portfolio decisions based in part on the expected
difference in inflation between the U. S. and Canada.
If these investors believe that a reduction in the Bank
of Canada’ discount rate will raise Canadian inflas
tion (relative to U. S. inflation), given the prevailing
interest rate differential, they will attempt to move
out of Canadian assets and into U. S. assets-the
Canadian dollar will fall immediately against the
dollar. Moreover,, because imports comprise about
a third of the basket of commodities in the Canadian consumer price index, the fall in the exchange
rate will appear quickly in inflation figures. This swift
association between the actions of the Bank of
Canada and price indices thus acts as a check on
inflationary policy actions.
Because imports are still only a relatively small
fraction of U. S. consumption, the U. S. public is
not sensitive to the foreign exchange value of the
dollar. Also, changes in the foreign exchange value
of the dollar do not solely measure changes in expected domestic inflation. Particularly over the
54

ECONOMIC

REVIEW.

198Os, the preponderance of changes in the foreign
exchange value of the dollar have reflected changes
in the real terms of trade caused by capital flows. For
these reasons, the exchange rate does not exercise
the kind of constraint in the U. S. that it exercises
in smaller, more open economies. The role the
exchange rate plays in these countries, however, does
indicate the advantages of creating a measure of
expected inflation.
First of all, a ready measure of the real (inflationadjusted) rate offered by the indexed bond would
lessen pressure for inflationary monetary policy by
eliminating public confusion over market rates and
real rates. Public perception that increases in market
rates necessarily indicate increases in real rates creates
pressures for stimulative monetary policy. If, for
example, new statistics indicated higher expected
inflation than previously forecast, a higher funds rate
would be necessary to keep real interest rates unchanged. Such an increase in the funds rate, however,
has often been seen by the public as causing an increase in real rates and, thus, as a “tightening” of
policy. With the yield on indexed bonds measuring
the real rate, the Fed can easily dispel the perception that all increases in the funds rate are increases
in real rates. Furthermore, the public will easily be
able to see how little leverage the Fed can exert over
real rates through allowing monetary acceleration.z
A measure of expected inflation would also provide a direct check to monetary policy actions (or
inactions) judged inflationary by the market-such
actions would produce an immediate rise in the yield
on standard bonds and in the differential yield between standard and indexed bonds. The rise in the
yield on standard bonds would impose a capital loss
on the holders of these bonds. Holders of variablerate mortgages with yields tied to the yield on standard bonds would incur higher interest payments.
Indeed, all creditors receiving payment in dollars in
the future would feel their financial interests threatened. A readily available measure of expected inflation that rose in response to monetary policy actions
judged inflationary by the market would make it
easier for creditors to counteract pressure on the Fed
to trade off price stability for short-term output gains.
In the U. S., the long lag between monetary policy
actions and changes in prices means that it is difficult
to associate particular policy actions with inflation.
2 Numerous economists have documented the virtual end of the
liquidity effect whereby an increased rate of growth of money
is associated with a fall in the real rate of interest. See. for
example, Mehra (1985).
MARCH/APRIL

1990

for inflation of Fed actions. The assessment would
be provided by individuals who have a financial
interest in monitoring Fed success in achieving price
level stability. The resulting constraint placed on
inflationary monetary policy would rest on the most
effective check available in a democracy-public
awareness.

As a result, inflation does not provide an adequate
check to pressures by government officials to keep
rates “low.” If an exhortation by a government
official to lower the funds rate produced an immediate
rise in the yield differential between standard and
indexed bonds, however, this rise would embarrass
the official. Officials would also realize that such
pressures are counterproductive. The Fed, concerned
about an adverse reaction in the expected inflation
measure, would be very reluctant to lower the funds
rate after an exhortation for easy money.

REFERENCES

Finally, the coexistence of standard and indexed
bonds would encourage the Fed to find ways of committing itself to a noninflationary policy in order to
eliminate a yield differential arising from a risk
premium. Even with a return to price stability, a
positive yield differential would appear in the two
kinds of long-term bonds if the public feared a future
lapse in the commitment to price stability. The Fed
would have an incentive to find ways to commit itself
to a monetary policy of price stability.

Greenspan, Alan. “Can the U. S. Return to a Gold Standard?’
Th Wall &ret Journal, September 1, 1981.
Humphrey, Thomas. me Concept of Indexation in the History
of Economic Thought.” Federal Reserve Bank of Richmond &mornic R.+rur& 60 (November/December
1974),
264-277.
Lindbeck, Assar. “Remaining Puzzles and Neglected Issues in
Macroeconomics.” ScandinavianJinma~of Eroiromic 9 l(2),
1989, 495-516.

There is a lack of agreement over specific ways
to constrain decision-making by the Federal Reserve
in order to achieve price stability. The proposal made
here leaves the operational details of achieving price
stability to the Fed. It provides, however, for a continuously available assessment of the consequences

Macroeconomic

Friedman, Milton; Charles Walker; Robert J. Gordon; and
William Fellner. “Indexing and Inflation.” American Enterprise Institute Round Table held on July 17, 1974,
Washington: American Enterprise Institute, 1974.

Mehra, Yash. “Inflationary Expectations, Money Growth, and
the Vanishing Liquidity Effect of Money on Interest: A
Further Investigation.” Federal Reserve Bank of Richmond
Ecwomic Rec.&w 71 (March/April 1985), 23-35.

Data: A User’ Guide
s

Roy H. Webb, Editor
The Federal Reserve Bank of Richmond is pleased to announce the publication of Mamveconomic Data:
A User’ Guide. This 4%page book will help a user of macroeconomic data to understand the most important
s
data series well enough to effectively interpret them. Chapters include:
l
l
l
l
l
l

The National Income and Product Accounts
Industrial Production and Capacity Utilization
Labor Market Data
Macroeconomic Price Indexes
Monetary Aggregates
Seasonal Adjustment

Copies may be obtained

free of charge by writing to:

Public Services Department
Federal Reserve Bank of Richmond
Post Office Box 27622
Richmond, VA 23261

FEDERAL

RESERVE

BANK OF RICHMOND

55


Federal Reserve Bank of St. Louis, One Federal Reserve Bank Plaza, St. Louis, MO 63102