Full text of Economic Perspectives (Federal Reserve Bank of Chicago) : September/October 1988, Volume XII, Issue 5

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

QD
PS-

AliJJVV

ECONOMIC

PERSPECTIVES




A review from
the Federal Reserve Bank
of Chicago

SEPTEM BER O CTO BER 1988
C apital m arket im p e rfec tio n s and
in vestm en t flu c tu a tio n s
Financial services in th e year 2000
N e w d irectio n s fo r econom ic
d evelo p m en t—th e banking industry

ECONOMIC PERSPECTIVES
September/October 1988
Volume XII, Issue 5
Karl A. Scheld, senior vice president
and director of research
Editorial direction

Edward G. Nash, editor
David R. Allardice, regional studies
Herbert Baer, financial structure
and regulation
Steven Strongin, monetary policy
Anne Weaver, administration

Production

Kathleen Solotroff, graphics coordinator
Roger Thryselius,
Thomas O ’Connell, graphics
Nancy Ahlstrom, typesetting coordinator
Rita Molloy,
Yvonne Peeples,
Gloria Powell,
Stephanie Boykin, typesetters

Economic Perspectives is
published by the Research Depart­
ment of the Federal Reserve Bank
of Chicago. The views expressed are
the authors’ and do not necessarily
reflect the views of the management
of the Federal Reserve Bank.
Single-copy subscriptions are
available free of charge. Please send
requests for single- and multiplecopy subscriptions, back issues, and
address changes to Public Informa­
tion Center, Federal Reserve Bank
of Chicago, P.O. Box 834, Chicago,
Illinois 60690, or telephone (312)
322-5111.
Articles may be reprinted pro­
vided source is credited and The
Public Information Center is pro­
vided with a copy of the published
material.
ISSN 0164-0682



C ontents
Capital market imperfections and
investment fluctuations
Bruce Petersen
Firms tend to reinvest their own money before
they go to market for new funds; as a result,
industrial investment closely tracks the busi­
ness cycle, but with bigger ups and downs
Financial services in the year 2000
Christine Pavel
The 24th annual Conference on Bank Struc­
ture and Competition considers the banking
industry’s options as the millenium draws near
New directions for economic
development—the banking industry
Eleanor H. Erdevig
Innovative andfast on their feet, a couple of
small states have captured a big chunk of
banking business andforced their neighbors to
liberalize their banking regulations

Capital m arket im perfections and
investm ent fluctuations
Bruce Petersen
It is well known that investment is a very
volatile and procyclical component of Gross
National Product. Recent studies indicate that
in the United States, investment fluctuations
have been approximately four to five times
greater than fluctuations in output over the
post-war period.1 For example, between 1973
and 1975, the peak-to-trough change in the
investment-to-capital ratio was approximately
-20 percent. A change of similar magnitude,
but of opposite sign, occurred between 1975
and 1979. These changes were much larger
than output deviations over this time period.
Investment is also more volatile than output in
other countries, such as Japan, the United
Kingdom, and West Germany, although the
difference is less pronounced.2
Because the volatility of investment is a
key aspect of the business cycle, economists
have become increasingly interested in provid­
ing a sound microeconomic explanation for this
aggregate phenomenon. That is, an attempt is
being made to understand why business enter­
prises find it in their best interest to invest in
such a pronounced procyclical pattern over
time. One of the most promising theories is
based on the premise that there are serious
imperfections in capital markets.3
The logic of this theory can be briefly
summarized. As a consequence of capital
market imperfections, external finance (debt
and new share issues) costs the firm consider­
ably more than internally generated finance
from earnings and depreciation allowances.
Thus, firms may be either unable or unwilling
to offset reductions in available internal finance
with external finance. As a result, a firm’s op­
timal response to a reduction in its internal fi­
nance may be to reduce its investment.
Because fluctuations in internal finance are
highly correlated with fluctuations in aggregate
output, imperfections in capital markets may
explain why aggregate investment is so volatile.
Capital market imperfections should not,
however, have a uniform impact on the invest­
ment behavior of all firms. This point has not
been emphasized in the models employing
Federal Reserve Bank of Chicago



capital market imperfections to explain aggre­
gate fluctuations in investment. Many corpo­
rations generate quantities of internal finance
well in excess of their demand for finance—that
is, they do not depend on external finance at
the margin. The existence of capital market
imperfections may be of little consequence for
the investment behavior of these firms. On the
other hand, a large fraction of the firms in the
United States do exhaust all, or nearly all, of
their internal finance. Investment of these
firms should be the most sensitive to fluctu­
ations in internal finance if capital markets are
imperfect. This is the basic idea behind the test
described in this paper.
A panel of publicly listed manufacturing
firms is grouped according to what fraction of
their earnings they retain in the firm. If the
cost disadvantage of external finance is slight,
then corporate retention behavior should con­
tain little or no information about investment
behavior—firms can simply use external finance
to smooth investment when internal finance
fluctuates. If, however, there is a pronounced
difference between the cost of internal and ex­
ternal finance, the investment of firms retaining
all of their income may be driven by fluctu­
ations in their internal finance.
The first section of the paper describes
possible sources of capital market imperfections
and the resulting cost-of-capital schedule. Im­
plications for investment behavior are devel­
oped and related to a standard model of
investment. The next two sections of the paper
present the test results, which indicate that in­
vestment is much more sensitive to fluctuations
in internal finance for firms which retain all of
their income.4 In addition, these firms, as an
aggregate, exhibit an extremely pronounced
procyclical pattern of investment. This is not
the case for firms in the sample with a high
dividend-payout ratio.
Bruce Petersen is a senior economist at the Federal Reserve
Bank of Chicago. He thanks Ed Nash and Steve Strongin
for helpful comments. Valuable research assistance was
provided by Charlie Himmelberg.
3

Capital market imperfections
Early investment research often empha­
sized the importance of financial factors such
as liquidity and access to internal finance as
determinants of investment spending.0 Indeed,
financial effects on many aspects of real eco­
nomic activity received considerable attention
during the early post-war period. Over the last
twenty years, however, most research on in­
vestment behavior has proceeded under the as­
sumption that the investment decision of the
firm can be separated from purely financial
decisions. The theoretical basis for this ap­
proach was provided by Modigliani and Miller
(1958) who demonstrated the irrelevance of fi­
nancial policy for real investment under certain
(very restrictive) conditions.
More recently, some economists and cor­
porate finance specialists have seriously ques­
tioned how closely the predictions of the
Miller-Modigliani Theorem match the actual
stylized facts about corporate financing. Myers
(1984) has proposed an alternative framework
which he refers to as a “pecking order” theory:
There is a financing hierarchy, with internal
finance dominating external finance.
There are several explanations for a cost
advantage of internal finance over external fi­
nance, including issue costs, the taxation of
capital income, and asymmetric information
between managers and potential investors. In
this paper, we will limit our examination to
taxation and asymmetric information.
Many countries tax income from capital
gains at much lower effective rates than the
rate on dividends. Such was the case in the
United States until quite recently.6 A large
number of studies have examined the cost of
equity finance in light of the above provisions.
A thorough review of this literature can be
found in Auerbach (1983). The central con­
clusion of these studies is that internal finance
has a tax advantage over new share issues. The
basic intuition is that no tax savings occur from
the issue of new shares, while tax savings do
occur when earnings are retained, because a
dividend tax is avoided for a lower tax on cap­
ital gains.
A measure of the cost advantage of
internal finance can be readily calculated.
Using the “q” model of investment (utilized
and explained in the next section of the paper),
consider how the favorable taxation of internal
4




finance alters the breakeven q value. (Tobin’s
q is the ratio of the stock market value of the
firm to its replacement cost.) The essential in­
sight underlying Tobin’s q theory of investment
is that, in a taxless world, firms should invest
as long as each additional dollar spent pur­
chasing capital raises the market value of the
firm by at least one dollar; that is, as long as
marginal q is at least equal to unity. This
breakeven condition of q = 1 changes when
the tax rate on dividends (d) exceeds the tax
rate on capital gains (c). Consider what q value
will make shareholders indifferent to SI of re­
tained earnings or SI of dividends. The after­
tax return on SI of dividends is $ 1(1 — d) while
the after-tax return on SI of capital gains is
S1q( 1 — c). These returns will be equated only
if: q = (1 — d)/(l — c), which is a value
clearly less than unity if c < d.
Thus, the breakeven q value on internal
finance is q = (1 — d)/(l — c) while firms
should issue new shares only if q > 1. The
breakeven q value on retentions is shown on the
vertical axis of the cost-of-capital schedule ap­
pearing in Figure 1. Internal finance is ex­
hausted at R on the horizontal axis, the point
of discontinuity on the cost-of-capital schedule.
A second reason for financing hierarchies
is asymmetric information. This is a situation
in which potential suppliers of finance have less
complete or less accurate information about a
firm’s prospects than the firm itself. Important
recent papers by Myers and Majluf (1984) and
Greenwald, Stiglitz, and Weiss (1984) explain
why asymmetric information either eliminates
Figure 1
Investm ents and financing decisions

Economic Perspectives

any reliance on external equity finance or
causes suppliers to demand a large premium.7
Myers and Majluf consider a situation in
which managers (or current owners) are better
informed than potential shareholders about the
true value of both the firm’s investment oppor­
tunities and the existing assets in place. In
addition, managers are assumed to act in the
interest of existing shareholders, and potential
new investors are aware of this. Since external
investors cannot distinguish the quality of
firms, they value them all at the population
average. Consequently, new shareholders im­
plicitly demand a premium to purchase the
shares of relatively good firms to offset the losses
that will arise from inadvertently funding
below-average firms, sometimes referred to as
“lemons”.
The intuition behind the “lemons” pre­
mium also can be described in terms of the q
model of investment. Following the example
in Myers and Majluf, let the true q value of
“good” firms be qG and the true q value of
“lemons” be qL and the percentage of good
firms be p. Because of asymmetric information,
all firms are initially valued at a weightedaverage value, qA = pqG + (1 —p)qL. It can be
shown that if the market does not collapse be­
cause good firms drop out, the breakeven q
value for good firms is approximately:
q = qG/qA. The breakeven q will exceed unity
by an amount that depends on the percentage
of “lemons” and the difference between the
value of good firms and “lemons”. The ratio
qG/qAindicates how much dilution occurs when
good firms issue new shares; the lemons pre­
mium, Q, is equal to qG/qA —l.8
A financing hierarchy depicting the com­
bined effects of taxation and asymmetric infor­
mation is shown in Figure 1. Firms exhaust
internal finance first and issue new shares only
if the marginal project has a q of at least
1 + Q. Also appearing in Figure 1 are three
possible demand schedules for new investment,
where projects are ranked according to their
Tobin’s q value. If a firm has available an
amount of internal finance R and an invest­
ment schedule depicted by demand curve D1}
it would finance all desired investment
internally and pay out some dividends. If its
investment demand schedule was D2 instead of
Du it would exhaust all internal finance but not
issue new shares. Finally, only if a firm’s in­
vestment demand schedule intersects the ex­
Federal Reserve Bank of Chicago



ternal finance portion of the financing
hierarchy, as depicted by investment demand
Z)3 will it issue new shares.
Debt considerations can be incorporated
into the cost schedules depicted in Figure 1.
It is often assumed that firms have some “debt
capacity” which is determined by the cost of
financial distress and by agency costs.9 It is well
known that debt finance creates agency prob­
lems and that the greater the debt-equity ratio,
the more distorted are the firm’s investment
incentives. Additionally, managers have in­
centives to issue new debt, which will raise the
riskiness and lower the value of existing debt.
Debt-holders understand these conflicts of in­
terest and rationally demand covenants which
restrict the behavior of managers, particularly
with respect to new debt issues.
In addition, recent work by Stiglitz and
Weiss (1981) and others emphasize how asym­
metric information between borrowers and
lenders can cause distortions similar to those
discussed above for new share issues. Asym­
metric information may increase the cost of
new debt, or even result in “credit rationing”.
One simple way to include debt finance
in the cost-of-capital schedule is to allow firms
to leverage every dollar of equity finance by
some fraction of a dollar of debt finance. In
this case, a dollar contraction in internal fi­
nance would cause R to decline by more than
one dollar. A second, more general way to in­
troduce debt finance is to include an upward
sloping schedule which connects the internal
and external finance segments (see for example
Auerbach, 1983). The position of this schedule
shifts in tandem with shifts in the quantity of
internal finance. The slope of the debt supply
schedule determines the extent to which firms
can offset reductions in internal finance with
greater leverage.
The financing patterns of most corpo­
rations are consistent with the predictions of a
financing hierarchy. Most corporations rely
very heavily on internal finance, particularly
small corporations where asymmetric informa­
tion problems are likely to be most pronounced.
New share issues account for only a small frac­
tion of new equity finance in the United States.
Srini Vasan (1986) examines the financ­
ing behavior of corporations engaged in manu­
facturing over the period 1960-80. He finds
that corporations with assets of under $100
million raised 85 percent of their finance from
5

internal sources. The balance came from bank
debt (10%), corporate bonds (3%) and new
share issues (2%). In addition, Srini Vasan
finds that the average retention ratio of small
corporations is very high and that many cor­
porations pay no dividends at all for long peri­
ods of time. This evidence indicates that it may
be very common for corporations to operate at
or near point R in Figure 1.
Fluctuations in investment at the firm
level
Consider a firm which has an investment
demand schedule like Z>2 in Figure 1. It is im­
portant to note that D2 does not intersect either
the internal finance segment or the new share
issues segment. Ignoring debt finance, the
firm’s optimal position is R; that is, it exhausts
all internal finance but does not issue new
shares. If the cost differential between internal
and external finance is large enough, the in­
vestment demand schedule can shift a consid­
erable distance without any investment
response.
Now consider what happens if internal fi­
nance expands or contracts. This amounts to
an increase or a decrease in the length (OR) of
the internal finance segment in Figure 1. Be­
cause the firm is exhausting all internal finance,
changes in earnings can cause a dollar-fordollar change in investment. For example, if
internal finance declines, the firm will contract
investment by moving up its investment de­
mand schedule. External finance will not be
used to smooth investment until the marginal
project has a return of at least 1 + Q.
Such a prediction is quite contrary to
standard neoclassical models of firm investment
behavior. In these models, capital markets are
assumed to be perfect, thus firms’ cost-ofcapital schedules are not discontinuous as
shown in Figure 1. A firm’s optimal level of
investment is determined by the cost of capital;
vertical shifts in the cost-of-capital schedule will
change the optimal capital stock and the rate
of investment.
The investment model that will be con­
sidered in the remainder of the paper is the q
theory of investment. The intuition of the
model is that, absent considerations of taxes or
capital-market imperfections, a firm will invest
so long as the value of an additional unit of
capital—marginal q—exceeds unity. In equi­
6




librium, the value of an extra unit of capital is
just its replacement cost, so that marginal q is
unity. The conceptual advantage of this
framework is that it is forward looking; that is,
investment is driven by the stock market’s
evaluation of the firm as measured by q. This
has become a standard model of investment
behavior and it has been estimated by many
researchers, usually for highly aggregated data.
Empirical implementation of the q theory
of investment requires rather strong assump­
tions about technology and adjustment costs.
Following Summers (1981) and several subse­
quent papers, adjustment costs are assumed to
be zero until some normal level of investment
is reached, after which marginal adjustment
costs rise linearly with investment.10 This can
be shown to produce the following equation:
0)

C W * = a, +

*i d „

+

Hu

where / is investment, K is the replacement
value of the capital stock, i and t denote the
firm and time period, respectively, a, is the
normal value of (I/K), and pit is an error term,
(^represents the value of q at the beginning of
the period adjusted for corporate and personal
tax considerations. (The tax adjustments are
calculated following the procedures outlined by
Summers [1981].)
An alternative model is required to de­
scribe the investment behavior of firms who
may not be able to respond to fluctuations in
Q because of capital market imperfections. In
the simplest alternative, investment is con­
strained by available cash flow (CF). The basic
model estimated in the paper is:
(2)

(//A),, = a, + a! & „ + a2 (CF/fC)lt + pit

Tax-adjusted q is included in the model to
control for variation over time in investment
opportunities. Tests of robustness of this basic
model are discussed in the next section.
To summarize, in a world of no capital
market imperfections, variations in Tobin’s q
should lead to variations in investment. How­
ever, for firms exhausting all internal finance
and facing a high shadow price on external fi­
nance, q could vary over a considerable range
with no investment response. Thus, variations
in cash flow, not q, may drive investment for
some firms.
Economic Perspectives

The data and test results
Value Line data is used to implement the
test described above. (The detailed definitions
of the empirical measures can be found in
Fazzari, Hubbard, and Petersen [1988].) At­
tention here is limited to firms within the
manufacturing sector. The selection of the time
period over which to conduct the test is very
important. Enough years are needed to obtain
adequate time-series variation. However, too
long a time period would permit firms that may
initially be constrained by capital market
imperfections to mature. With these consider­
ations in mind, and taking into account the
data availability, the time period 1970 to 1984
was selected.11 Subintervals within this period
are also analyzed.
The sample of firms was obtained as fol­
lows. Firms with missing or inconsistent data
were deleted. In addition, firms with major
mergers were deleted because mergers can
cause inconsistencies when constructing lags.
Finally, firms with negative growth rates in
sales were excluded.1' The resulting sample
consisted of 422 manufacturing firms.
The tests described above require that the
sample be partitioned into groups of firms; the
obvious selection criteria is retention behavior.
If capital market imperfections lead to financ­
ing constraints on investment, they should be
most evident for firms that retain all of their
income. If, however, internal and external fi­
nance are nearly perfect substitutes, then re­
tention behavior should contain little or no
information about investment behavior, in­
cluding fluctuations in investment. The classi­
fication scheme chosen divides the sample into
three groups as follows.13
Class 1: Dividends < 0.1 for at least
10 years
Income
Class 2: Dividends < 0.2 for at least
10 years, but not
Income
in class 1;
Class 3: All others.
Several summary statistics for the firms in
each class are reported in Table 1. Class 1
firms—those that are most likely to be affected
by capital market imperfections—retained an
average of 94 percent of their income, and paid
a dividend on average in only 33 percent of the
Federal Reserve Bank of Chicago



Table 1
Sum m ary statistics: Sam ple o f m anufacturing
firm s, 1970-1984
Category of firm
Class 1
Number of firms
Average
retention ratio
Percent of years
w ith positive
dividends

Class 2

Class 3

49

39

334

0.94

0.83

0.58

33%

83%

98%

Median capital
stock-1970 (millions
of 1982 dollars)

27.1

54.2

401.6

Median capital
stock-1984 (m illions
of 1982 dollars)

94.9

192.5

480.8

Average real
sales growth

13.7%

8.7%

4.6%

Source: These calculations are based on samples selected from
the Value Line database.

years. The typical class 1 firm paid no divi­
dends for the first seven to ten years and a small
dividend in the remaining years. In fact, 21
firms in class 1 never paid a dividend over the
entire time period, although these firms are, on
average, very profitable. Going across classes,
there is a pronounced increase in the percent­
age of time that a positive dividend is paid and
a corresponding decrease in the retention ratio.
The classes are effectively sorted by firm
size as well, as the capital stock figures show.
While class 1 firms are small relative to firms
in class 3, they are still large relative to U.S.
manufacturing corporations in general.14
Table 2 presents information on new
share issues and debt finance for each of the
classes. Ceteris paribus, one would expect firms
in class 1 to rely more heavily on new share is­
sues than firms in the remaining classes. The
typical firm in class 1 has an investment de­
mand schedule like D2 or Z)3 in Figure 1. In
contrast, the typical firm in class 3 has a de­
mand schedule like Dx and should not simul­
taneously pay dividends and issue new shares,
given the taxation of corporate income. Con­
sistent with the cost-of-capital schedule in Fig­
ure 1, firms in class 1 issue new shares more
frequently—approximately one year in every
four—than do firms in the other two classes.
Even for class 1, however, the amount of fi­
nance raised from new share issues is small
7

Table 2
N ew share issues and deb t u tilizatio n
Category of firm
Class 1

Class 2

Class 3

Average percent of
years w ith new
share issues

28%

19%

10%

Average value of share
issue as a percentage
of cash flow

23%

13%

8%

Average ratio of debt
to capital stock

0.57

0.52

0.33

Correlation of the
earnings-to-capital ratio
and the change in total
debt-to-capital ratio
(averaged over firms)

0.23

0.15

0.09

Source: These calculations are based on samples selected
from the Value Line database.

compared to funds generated from internal
cash flows.
The last two lines of Table 2 provide in­
formation on debt utilization. Although one
would expect the firms in class 3 to have higher
debt capacities, the debt-to-capital ratios are
much higher for classes 1 and 2. These results
are consistent with the existence of a financing
hierarchy; i.e., constrained firms appear to
borrow up to their debt capacity.
Table 3 reports the estimates of the con­
tribution of internal finance toward explaining
investment after controlling for movements in
Qj Equation 2 is estimated for each retention
class with Q and CF/K as explanatory vari­
ables. Fixed firm and year effects are included,
and the equations are estimated over three time
periods: 1970-75, 1970-79, and 1970-84.
Given the method of construction of the
Value Line database, the strongest case for
asymmetric information between firms and
outside investors can be made for the shorter
time periods, particularly 1970-75. A firm is
not added to the database until it is “of interest
to subscribers and the financial community.”
Once a firm is added, however, observations
on items from its income statements and bal­
ance sheets are collected as far back as possible;
in practice, for at least 10 years prior to the
date it is added to the Value Line database.
The majority of class 1 firms were not recog­
nized until near the end of the full-sample pe­
riod, even though the data for these firms
extend back to 1969. Thus, if asymmetric in­
8



formation is an important impediment to firms
obtaining external finance, then the sensitivity
of investment to fluctuations in internal finance
for class 1 firms should be greatest in the earlier
time periods.
The results in Table 3 show large esti­
mated cash flow coefficients for class 1 firms.
As expected, the cash flow coefficient is largest
(0.670) in the earliest period. The coefficient
is the smallest (0.461) for 1970-84. Further­
more, as the sample period is extended one
year at a time from 1970-75 to 1970-84, the
estimated class 1 cash flow coefficients decline
monotonically. The cash flow coefficients in
classes 2 and 3 are positive and approximately
stable over time.
It is the difference in the estimated coef­
ficients across the three classes that should be
stressed. Comparing classes 1 and 3, the dif­
ferences in the estimated coefficients range from
0.416 for 1970-75 to 0.231 for 1970-84. These
differences are always statistically significant at
very high confidence levels. It is also important
to note that including internal finance in the
investment model (Equation 2) explains a
Table 3
Effects o f Q and cash flo w on investm ent
Dependent variable: (IIK )it
Variable

Class 1

Class 2

Class 3

Sample period: 1970-75

On

-0.0010
(0.0004)

0.0072
(0.0017)

0.0014
(0.0004)

(O F IK ) jf

0.670
(0.044)

0.349
(0.075)

0.254
(0.022)

R *.

0.55

0.19

0.13

Sample period: 1970-79

Oit

0.0002
(0.0004)

0.0060
(0.0011)

0.0020
(0.0003)

(O F /K ) it

0.540
(0.036)

0.313
(0.054)

0.185
(0.013)

0.47

0.20

0.14

Sample period: 1970-84

Oit

0.0008
(0.0004)

0.0046
(0.0009)

0.0020
(0.0003)

(O F IK ) it

0.461
(0.027)

0.363
(0.039)

0.230
(0.010)

R*

0.46

0.28

0.19

Note: Standard errors appear in parentheses. The equations
were estimated using fixed firm and year effects (no t reported).

Economic Perspectives

much greater proportion of the variance of the
investment-to-capital ratio (UK) in class 1
than in the other two classes. In class 1, 46 to
55 percent of the variance in I/K is explained,
depending on the time period analyzed, pri­
marily by the variation in cash flow alone.
Several tests of robustness of these findings
were undertaken.15 In addition, alternative
specifications of Equation 2 were estimated,
including lags of cash flow and current and
lagged sales. A number of investment studies
in the literature have found that models which
include sales or output (“accelerator” models)
often outperform q models. Thus, the results
in Table 3 could arise because cash flow and
sales are correlated. When current and lagged
sales are included in the model, the cash flow
coefficient declines in all three classes. For
firms paying out a large share of their income
as dividends (class 3 firms), the coefficient falls
so close to zero that it is insignificant, with one
exception.16 However, for firms retaining all of
their income (class 1 firms), the cash flow coef­

ficient is 0.392 for 1970-75, 0.360 for 1970-79
and 0.301 for 1970-84; these coefficients are
highly statistically significant. Thus, for class
1 firms, a large fraction of the fluctuation in
investment continues to be explained by fluc­
tuations in cash flow even after including sales
in the model.
Aggregate investment fluctuations
The results in Table 3 strongly suggest
that capital market imperfections do not have
a uniform effect on investment behavior across
firms. An implication of these results is that
researchers looking for explanations of why ag­
gregate investment is so volatile should focus
their attention on specific types of enterprises.
There are solid theoretical reasons for using
retention behavior as a selection criterion in
future studies. Additional empirical support is
provided below.
The results in Table 3 indicate that at the
firm level, fluctuations in cash flow appear to

Figure 2
Average in v es tm en t-to -c ap ita l ratio (by dividend payout class)
ratio

Federal Reserve Bank of Chicago



9

cause pronounced fluctuations in investment
for firms exhausting all of their earnings. The
important question for macroeconomics is
whether this leads to pronounced aggregate
fluctuations in investment.
Figure 2 presents a plot over time of the
average investment-to-capital ratio (I\K) for
each class, constructed by averaging across
firms. It is apparent that IjK is procyclical in
each of the classes. It is further apparent that
the fluctuations in ///f are much more pro­
nounced for the class 1 aggregate than for the
class 3 aggregate. While not shown here, for
the class 1 aggregate, the plot of the cash-flowto-capital ratio is almost identical to the //K
plot. This is not the case, however, for the class
3 aggregate.
For the class 1 aggregate, there are peakto-trough changes in //A* of 0.21 between 1972
and 1975 and 0.14 between 1979 and 1983.
Peak-to-trough changes in the average cashflow-to-capital ratio are of almost identical
magnitude. In contrast, for the class 3 aggre­
gate, there are peak-to-trough changes in //A*
of 0.03 between 1973 and 1975 and 0.04 be­
tween 1980 and 1983. While not reported here,
simple regressions of I/K on various measures
of the aggregate economy reveal the same pat­
tern.17 These regressions indicate that invest­
ment is much more procyclical for firms which
retain all of their income.
While not as interesting as the peak-totrough movements in I\K, the variance of the
//K series also gives a rough impression of just
how much more volatile is investment in the
class 1 aggregate than in the other two classes.
The variance of I/K for the class 1 aggregate
over the 1970-84 period is four times greater
than the variance of the class 2 aggregate and
ten times greater than that of the class 3 ag­
gregate. While not shown in Figure 2, two
average I\K series were constructed, one for all
of the firms in the sample and a second which
excluded all class 1 firms. While class 1 firms
make up only ten percent of the sample, ex­
cluding them causes the variance of the I/K se­
ries to fall by nearly one-third.
Conclusion
It is beyond the scope of this paper to give
an estimate of what fraction of the fluctuation
in aggregate investment in recent decades can
be explained by imperfections in capital mar­
10



kets. Much depends, of course, on what frac­
tion of investment comes from firms that are
exhausting all, or almost all, of their internal
finance. In the sample of firms employed in
this study, that fraction is not particularly large
(see Table 1). This is because the Value Line
database is heavily weighted toward large,
mature corporations; that is, towards firms for
which public information is readily available.
Statistics indicate, however, that a sub­
stantial fraction of investment in the manufac­
turing sector is coming from firms with
characteristics similar to the class 1 aggregate.
For example, manufacturing firms of under $10
million in asssets accounted for approximately
14 percent of total investment in manufactur­
ing over the period 1970-84. (The average firm
size in the class 1 category is considerably
larger than $10 million.) The average re­
tention ratio of these firms is very
high—approximately 80 percent—and they
raised a neglible fraction of finance from new
share issues.18 If these firms have //K ratios that
are as volatile as those making up the class 1
aggregate reported here, they could easily ac­
count for a major fraction of the investment
volatility in manufacturing.
Finally, it is important to point out that
the sample of firms utilized in this study were
drawn entirely from the manufacturing sector.
The average size of enterprises in other sectors
of the economy, such as trade and agriculture,
is much smaller than firms in manufacturing.
For this reason, capital market imperfections
such as asymmetric information may be even
more important, and access to external finance
more restricted in these sectors. Further re­
search is needed to assess the volatility of in­
vestment in these sectors.
1 See, for example, the findings in Greenwald and
Stiglitz (1988).
2 See the findings in Greenwald and Stiglitz (1988).
3 See, for example, Greenwald, Stiglitz, and Weiss
(1984).
4 The test results presented- in this paper draw
heavily on the statistical results in Fazzari,
Hubbard, and Petersen (1988).
5 An extensive list of citations on this investment
research can be found in Fazzari, Hubbard, and
Petersen (1988).
Economic Perspectives

6 The difference between the tax rate on dividends
and capital gains has been quite large because of
(1) the exclusion of 60 percent of long-term capital
gains; (2) the taxation of such gains only upon re­
alization; and (3) forgiveness of the tax if the gain
is not realized before death. Recent tax reform has
greatly reduced the difference.
7 These theoretical arguments draw heavily on the
“lemons” problem first considered by Akerlof
(1970).
8 For example, suppose qG= 5 and qA = 2, then Q
is 1.5 and a new project must have a q of at least
2.5 before managers will seek external equity fi­
nance.
9 Financial distress refers to the set of problems that
arise whenever a firm has difficulties in meeting its
principal and interest obligations. Agency costs
arise from the efforts of creditors of the firm to en­
sure that the firm honors its contractual obli­
gations.
10 For an overview of these assumptions, see the
discussion in Summers (1981).
11 Manufacturing firms were included in the sample
only if they had observations from 1969 to 1984.
The number of firms and data items available on
Value Line increased substantially in 1969. The
number of firms that had observations on the nec­
essary variables dropped significantly after 1984.
675 firms had some data from 1969 to 1984. The
sample was reduced to 422 firms for reasons dis­
cussed in the paper.
12 The objective is to consider the investment be­
havior of firms constrained because of capital mar­

ket imperfections, as opposed to financial distress
due to poor market performance.
13 This approach limits the sensitivity of the classi­
fication scheme to outliers of the dividend-income
ratio. In a particular year, this ratio could be very
high due to abnormally low income, even though
the firm generally retains most of its earnings.
14 Based on information from the Quarterly Financial
Reports of the Securities and Exchange Commission,
approximately 85 percent of manufacturing corpo­
rations had smaller capital stocks in 1970 than the
average class 1 firm.
15 For example, Equation (2) was estimated using
first and second differences (as opposed to the con­
ventional within-group estimation) to address
measurement-error considerations. The coefficient
estimates are quite similar. These tests are reported
in Fazzari, Hubbard, and Petersen (1988).
16 The exception is the full sample period of
1970-84.
17 The regressions mentioned in the text are of the
form:
(UK), = T l + Y2 At
where A is either capacity utilization in manufac­
turing or unemployment. The regressions were run
with and without time trends. The coefficient on
A is, in absolute value, three to four times greater
for the class 1 aggregate.
18 Firms between $10 and $100 million in assets also
have, on average, very high retention ratios. In­
formation on total investment and retention ratios
of manufacturing firms grouped by asset size can
be found in the Quarterly Financial Reports of the Se­
curities and Exchange Commission.

Bibliography

Abel, Andrew B. and Olivier J. Blanchard. “The
Present Value of Profits and Cyclical Move­
ments in Investment.” Econometrica, vol. 54
(March 1986), pp. 249-273.
Akerlof, George A. “The Market for Lemons:
Qualitative Uncertainty and the Market
Mechanism.” Quarterly Journal of Economics,
vol. 84 (August 1970), pp. 448-500.
Auerbach, Alan J. “Taxation, Corporate Finance,
and the Cost of Capital.” Journal of Economic
Literature, vol. 21 (September 1983),
pp. 905-940.
Fazzari, Steven, R. G. Hubbard, and Bruce
Petersen. “Financing Constraints and Corpo­
rate Investment.” Brookings Papers on Economic
Activity, forthcoming.
Greenwald, Bruce and Joseph Stiglitz. “Examining
Alternative Macroeconomic Theories.”
Federal
 Reserve Bank of Chicago


Brookings Papers on Economic Activity, forth­
coming.
Greenwald, Bruce, Joseph Stiglitz, and Andrew
Weiss. “Information Imperfections in the
Capital Market and Macroeconomic Fluctu­
ations,” American Economic Review, vol. 74
(May 1984), pp. 194-199.
Modigliani, Franco and Merton Miller. “The Cost
of Capital, Corporation Finance and the
Theory of Investment,” American Economic
Review, vol. 48 (June 1958), pp. 261-297.
Myers, Stewart C. “The Capital Structure
Puzzle.” Journal of Finance, vol. 39 (May
1984), pp. 575-592.
Myers, Stewart C. and Nicholas S. Majluf. “Cor­
porate Financing Decisions When Firms
Have Investment Information That Investors
11

Do Not.” Journal of Financial Economics, vol.
13 (June 1984), pp. 187-220.
Srini Vasan, Philip Vijay. Credit Rationing and Cor­
porate Investment. Unpublished Ph.D. Disser­
tation, Harvard University, October 1986.
Stiglitz, Joseph E. and Andrew Weiss. “Credit
Rationing in Markets with Imperfect Infor­
mation.” American Economic Review, vol. 71
(June 1981), pp. 393-410.

72




Summers, Lawrence H. “Taxation and Corporate
Investment: A q-Theory Approach.”
Brookings Papers on Economic Activity, 1981, vol.
1 pp. 67-127.
Tobin, James. “A General Equilibrium Approach
to Monetary Theory,” Journal of Money,
Credit, and Banking, vol. 1 (February 1969),
pp. 15-29.

Economic Perspectives

24th Bank Structure Conference
Financial Services in the Year 2000
“While the expansion of banking powers
is consistent with a flexible, safe, and efficient
financial system and increased real benefits to
consumers, there still remain reasons for policy
makers to be cautious about such changes in
financial structure,” said Alan Greenspan,
Chairman of the Federal Reserve Board and
keynote speaker at the 24th annual Conference
on Bank Structure and Competition, sponsored
by the Federal Reserve Bank of Chicago on
May 12-13, 1988. The Conference offered
many experts from regulatory agencies, the
banking industry, and academia the opportu­
nity to present their views and recommen­
dations for balancing the benefits of increased
efficiency from expanded powers against possi­
ble increases in bank risk.
In a number of speeches and panel dis­
cussions, the participants discussed a variety of
issues, including the effects of October 19, fi­
nancial restructuring, corporate separateness,
new powers, and bank risk. The last session of
the Conference assembled a panel of industry
experts who attempted to sum up the two days
of discussion and build a framework for re­
structuring. The panel included Donald
Crawford, senior vice president and director of
government relations for the Securities Indus­
try Association; Robert Litan, senior fellow at
the Brookings Institution; S. Waite Rawls III,
vice chairman of Continental Illinois Corpo­
ration; and Kenneth Scott, professor of law at
the Stanford Law School.
New powers
While many at the Conference believed
that banking firms would be granted broader
powers, there was considerable disagreement
about the specific powers banking firms would
and should have.
Kenneth Scott, of the Stanford Law
School, provided some insight into the deter­
mination of new bank powers. According to
Mr. Scott, if new bank powers are determined
in the political arena where special interest
groups need a “super majority” to effect change
but a simple majority to keep the status quo,
little change will occur. If new powers for
Federal Reserve Bank of Chicago



banking firms are left to the regulators, only
those powers that are easily understood or eas­
ily measured will be granted. If new powers
are decided by economists, new powers will be
granted only if there are synergies between
banking and the new activities. Finally, said
Mr. Scott, if new powers are determined by the
market, those firms that correctly assess the
opportunities for expansion will be rewarded
and those that do not will be punished. This
“marketplace calculus,” according to Mr.
Scott, is the best way from an efficiency stand­
point to determine which new powers are ap­
propriate for banking firms.
Donald Crawford, from the Securities In­
dustry Association, argued that the politcal
process, in fact, was directing the push for new
powers in an inappropriate direction—toward
securities activities. Citing profitability figures
for the underwriting of various securities, Mr.
Crawford argued that competition, tax reform,
and deregulation have narrowed spreads in
virtually every area of investment banking.
Therefore, if banks entered this industry, the
competition would be ruinous to both banking
firms and securities firms. “Combining the two
industries will exponentially increase the po­
tential for mismanagement on both sides of the
fence,” said Mr. Crawford. Earlier in the day,
William T. Gregor, a senior vice president at
the MAC Group, had made the same point:
“For many banks underwriting is going to be
an economic Vietnam.”
Furthermore, Mr. Crawford noted that
the securities industry is one in which “mistakes
are made easily,” and is “unforgiving” because
assets are marked to market daily. To support
this contention he pointed to the Stock Market
Crash of October 1987. As a result of the
Crash, the securities industry lost $2.2 billion
in two days, $1.7 billion of which was from
trading accounts. This produced the worst
quarter in the history of the securities industry.
“You can’t underwrite unless you make mar­
kets,” Mr. Crawford warned the bankers in the
audience, “and if you make markets, you will
occasionally have to take hits.”
These losses, however, did not impress S.
Waite Rawls III, Continental Illinois Corpo­
13

ration and the only commercial banker on the
panel: “Shoot, a billion seven. Citibank]
charged off twice that in a day. I thought we
were talking about risk here.” Mr. Rawls also
asked Mr. Crawford, “If what you’re protect­
ing is worth so little, why do you defend it so
doggedly?”
In a previous session of the Conference,
Larry Mote, a vice president and economic
adviser at the Federal Reserve Bank of
Chicago, may have provided an answer to Mr.
Rawls’ question. He noted that average re­
turns and levels of compensation in the securi­
ties industry are relatively high. Moreover,
there is a high degree of concentration and
barriers to entry are significant in the securities
industry. These characteristics, along with
long-run stability of some spreads, are sugges­
tive of market power.
In his presentation, Robert Litan, of the
Brookings Institution, suggested that if banks
broaden their securities activities, prices for
underwriting services are going to come down
and profits will decrease. This effect will be
most pronounced in merger and acquisition
services, according to Mr. Litan. Currently,
banks can provide advice on mergers and ac­
quisition, but cannot underwrite corporate se­
curities. However, underwiting capabilities are
very advantageous to the M&A business. Ear­
lier, Thomas G. Labrecque, president and chief
operating officer of Chase Manhattan Corpo­
ration, had commented on that very issue. He
stated that his organization recently lost busi­
ness to Deutsche Bank because Chase cannot
underwrite corporate securities. “In my hum­
ble opinion,” Mr. Litan opined, “those fat,
outrageous M&A fees would come down if
banks were in that business and could also
underwrite securities.”
Robert Litan, however, conceded to Mr.
Crawford that securities may not be the most
important area for banking firms to enter. Mr.
Litan felt that bank entry into insurance would
have a greater impact on consumers than bank
entry into the securities industry. Citing studies
of the Consumers Federation of America and
the American Insurance Association, Mr. Litan
estimated that more competition in insurance
agency would reduce premiums by $5 billion
annually. Banking, he argued, is a logical
source for this new competition. As John Boyd,
a research officer at the Federal Reserve Bank
of Minneapolis, contended earlier in the day,
14



life insurance underwriting is a low-risk activ­
ity, and if banking firms were to engage in this
activity, their overall level of risk would likely
decrease.
Restructuring
If banks are going to be granted broader
powers—securities as well as real estate and
insurance—how should the financial services
industry be restructured so that safety and
soundness are preserved; the safety net is not
extended to nonbank sectors; and efficiency is
not sacrificed? In other words, restructuring
requires walking a tight rope between risk and
efficiency.
“In this industry [banking],” said Mr.
Rawls, “risk is a four-letter word,” but without
risk, a company would have “zero potential for
revenues or growth.” Mr. Rawls continued,
“Today, risk is adapting to a new reality, or
failing to adapt.” That new reality is that “the
needs of business have changed faster than
banks’ capability of serving those needs. Being
a reliable provider of funds just isn’t enough
anymore.” Earlier in the day, Bert Ely, a fi­
nancial institutions consultant, had stated that
the financial services industry is changing more
rapidly than banking regulation, due to elec­
tronic technology and financial innovation.
The issue, said Mr. Rawls, “is what are we
going to do about it?”
Three conference participants, Robert
Litan, Robert Laurence of the Federal Finan­
cial Institutions Examination Council, and
Samuel Talley, a banking consultant, would
allow a banking firm to engage in any nonbank
activity it chooses so long as those activities are
carried out in subsidiaries of the holding com­
pany and the banking subsidiaries are “narrow
banks.” A narrow bank is one that accepts
deposits and invests them only in government
securities.
Mr. Litan conceded that having nonbank
activities operated as a bank subsidiary would
be more efficient, but it would also be riskier.
In other words, the temptation for the bank to
come to the rescue of a nonbank subsidiary
would be great since the performance of the
subsidiary directly affects the bank’s financial
statements. Therefore, in the interest of safety,
Mr. Litan said that he preferred that nonbank
activities be carried out by subsidiaries of the
bank holding company rather than the bank.
Economic Perspectives

Locating nonbank operations in subsid­
iaries of the holding company, however, was
not enough for Mr. Litan, nor was it enough
for the other panelists. All agreed that
“firewalls” are needed. How high, how thick,
and of what substance, however, were major
issues yet to be resolved.
According to Mr. Litan, the choices are
imperfect. One of those choices is the
“lawyers/regulators approach.” This approach
entails making rules and regulations that gov­
ern transactions and affiliations between banks
and their nonbank affiliates.
Mr. Rawls thought that this “approach”
was not so much a means to control risk but a
battle over turf. The Glass Steagall Act sepa­
rated investment and commercial banking, but
“the insidious thing is,” explained Mr. Rawls,
“it also created separate regulatory bodies—the
SEC and the Fed. And it created separate
Congressional committees to oversee the sepa­
rate regulatory bodies.” The problem, accord­
ing to Mr. Rawls, is that “the distinction
between banking and securities has really
blurred; the distinctions between the bodies
that regulate them have not.” Consequently,
“the issues are discussed from two different
points of view. Compromise is hard to come
by. Firewalls, functional regulation, and sub­
sidiaries are products to serve the regulators
and Congress, not to serve bankers or their
customers,” said Mr. Rawls.
Mr. Rawls as well as the other panelists
felt it necessary to distinguish between financial
firewalls and management and marketing
firewalls. While Mr. Crawford was accused of
favoring management and marketing firewalls,
such as a ban on cross marketing, the other
panelists generally agreed that only financial
firewalls were appropriate and necessary. As
Mr. Rawls pointed out, “you have to keep
bank deposits away from other activities, but
not marketing and management.” In fact, he
argued that marketing and management
firewalls would increase risk and reduce effi­
ciency. Furthermore, as for financial firewalls,
Mr. Rawls said that if barriers are erected,
deposits should not necessarily be with loans on
one side of the firewall and securities on the
other side.
That solution flies in the face the “narrow
bank” proposal. The narrow bank proposal is
another firewall alternative and the one that
Federal Reserve Bank of Chicago



Mr. Litan believes to be the best among the
imperfect choices.
Because a narrow bank does not make
loans, all lending as well as other activities
would be carried out in nonbank affiliates un­
der the umbrella of a bank holding company.
According to Mr. Litan’s proposal, bank hold­
ing companies that converted their banks to
narrow banks could engage in any nonbank
activities, not only those deemed permissible
by the regulators or Congress. The deposits of
narrow banks would be federally insured, but
they would have relatively low deposit insur­
ance premiums because they are virtually risk­
free entities. Narrow banks could invest in
both long- and short-term government securi­
ties. Conversion to narrow banking would be
purely voluntary and gradual (over a ten-year
period), and small banks would be exempt be­
cause, in Mr. Litan’s opinion, small banks do
not pose a risk to the system.
Referring to his narrow bank proposal,
Mr. Litan said “I think it solves all the prob­
lems, or most of the problems that have been
leveled against the banking industry in terms
of going out to broader powers.” He then
elaborated on that point, “It solves the conflict
problem because a narrow bank can’t loan; it
solves the run problem because a narrow bank
is liquid . . . .”
Whether firewalls be in the form of rules
or narrow banking, Mr. Scott questioned
whether any firewalls would be effective as long
as the fundamentals of the current federal de­
posit insurance system go unchanged.
Firewalls are supposed to protect against “un­
acceptable risk.” But protect whom? The in­
surance fund, said Mr. Scott, and accordingly
any discussion about new powers and risk must
include a discussion about federal deposit in­
surance. The current flat-rate system inher­
ently has a “perverse incentive system,” said
Mr. Scott. Furthermore, he said that if regu­
lation and supervision were adequate, then the
current deposit system would not be in the poor
condition that it currendy is in. Therefore, said
Mr. Scott, “if the present system is going to be
bailed out but not otherwise materially
altered,” then the thickness of firewalls and
distinctions between banking and nonbank
subsidiaries becomes important. But Mr. Scott
conceded, “maybe that’s all that is politically
possible now.”
15

Politically possible
Other panelists at the Conference’s last
session also spoke of the politically possible.
They all seemed to have agreed that the Con­
gress will grant banks broader securities pow­
ers. If they don’t, commented Mr. Litan, the
“states will take it upon themselves to broaden
securities powers,” especially New York. Mr.
Crawford further pointed out, and Mr. Litan
concurred, that political forces may have banks
trade insurance and real estate powers for se­
curities powers. “Is the trade-off worth it?”
Mr. Crawford queried.

76




All the panelists seemed to have agreed
that the approach to firewalls would be rulesoriented. While this is not the best alternative
as far as Mr. Litan was concerned, he conceded
that he would not see his narrow bank proposal
adopted in his lifetime. Referring to the regu­
latory approach to firewalls, he said that this
seems to be “the direction we’re headed.” Mr.
Rawls lamented that this approach may very
well mean that when all is said and done the
restrictions placed on banks’ securities activities
will be onerous. Then bankers will say to
Congress, “thanks, but you didn’t do
anything.”
—Christine A. Pavel

Economic Perspectives

New directions for economic
developm ent—the banking industry
Eleanor H. Erdevig
Projected job growth in the service in­
dustries has stimulated interest in those indus­
tries as a source of economic development and
employment growth.
One of the service industries that has
generated particular interest has been the
banking industry. Technological improve­
ments no longer require that banking oper­
ations be located in close proximity to a bank’s
customers. The elimination by some states of
restrictions on banking, such as geographic
limits and usury laws, has provided a means of
attracting banking operations from other states.
The results have been an acceleration of the
deregulation trend as states affected by the
possible or actual loss of banking employment
enact similar measures to retain and attract
banking operations and economic benefits for
consumers.
South Dakota and Delaware were pio­
neers in targeting the banking industry for
economic development. The efforts of South
Dakota were directed toward credit card oper­
ations. Delaware targeted not only credit card
operations, but also wholesale banking and
international banking.
General interstate banking legislation
provides an opportunity for bank holding
companies to transfer operations to banking
subsidiaries located in states with the greatest
opportunities for profitable operations. It also
provides geographic diversification of risk.
This paper looks at the efforts and results
in South Dakota and Delaware in developing
the banking industry in their states. It also
looks at the current status of interstate banking,
i.e., where bank holding companies are estab­
lishing or acquiring new out-of-state banking
subsidiaries and the resulting effects on asset,
deposit, and employment growth in the indi­
vidual states.
Projected job growth all in services
More than 21 million jobs are projected
to be added to the United States economy be­
tween now and the end of the century, accordFederal Reserve Bank of Chicago



Table 1
P rojected e m p lo ym e n t g ro w th , 1986-2000
Industry

Number

Percent
change

(mil.)
Total
Nonfarm wage and salary
Goods-producing
Mining
Construction
Manufacturing

21.4
20.1
*
*
.9
-.8

19
20
*
-8
18
4

Service-producing
Transportation
and public utilities
Wholesale trade
Retail trade
Finance, insurance,
and real estate
Services
Government

20.1

27

.5
1.5
4.9

9
27
27

1.6
10.0
1.6

26
44
10

-.3
*

-10

Private households
Nonfarm self-employed
and unpaid family workers

1.7

20

Agriculture

-2

‘ Less than .05
Source: Kutscher, Ronald E., "Projections 2000: Overview and implications of the projections to 2000,"
M onthly Labor Review , Vol. 100, pp. 3-9, September
1987.

ing to the Bureau of Labor Statistics. (See
Table 1.) Of the increase, virtually all will be
in the service-producing industries. Although
some of the goods-producing industries, which
include manufacturing, mining, and con­
struction, are projected to grow, others will
decline, and as a result, no net change in em­
ployment is expected in the goods-producing
sector.
Financial services are among the serviceproducing industries expected to continue to
show substantial rates of output growth. Al­
though employment in finance is expected to
grow less rapidly than in the recent past, there
are expected to be 262,000 more jobs in bank­
ing, 495,000 more in credit agencies and inEleanor H. Erdevig is a regional economist at the Federal
Reserve Bank of Chicago.
17

vestment offices, and 134,000 more in security
and commodity brokerages and exchanges by
the year 2000.
State opportunities for financial centers
Projected growth in employment in fi­
nancial services has encouraged states and cities
interested in economic development to establish
financial centers. Among the incentives offered
to encourage financial services firms, partic­
ularly banking institutions, to locate in an area
have been the elimination of interest rate ceil­
ings on loans, lower tax rates on bank net in­
come, and permission for interstate banking.
In 1978, the Supreme Court, in Marquette
National Bank v. First of Omaha Service Corpo­
ration, 439 W.S. 249 (1978), affirmed the right
of a national bank to charge interest rates to
out-of-state credit card customers at the rate
permitted by the law of its home state. This
meant that national banks in states with higher
or nonexistent ceilings on consumer lending
rates could export that rate to consumers in
states with lower ceiling rates.
The ability to offer more pricing flexibil­
ity in consumer lending presented an opportu­
nity for a state to increase commercial banking
transaction activity and employment. By in­
creasing or eliminating ceiling rates on con­
sumer loans and granting permission to
out-of-state bank holding companies to own
national bank subsidiaries, state legislators
could encourage such companies to establish
subsidiaries engaged primarily in national
credit card operations. Subsidies or tax incen­
tives, frequently used by states to attract man­
ufacturing companies, were usually not
involved. Minimum employment levels were
usually the only requirement in implementing
the legislation. In addition the legislation usu­
ally included restrictions on the operations of
the bank subsidiary of the out-of-state bank
holding company to protect existing in-state
banks, which also benefited from the changes
in usury rates.
In addition to nonexistent or nonbinding
usury rates, some states offered other induce­
ments in targeting commercial banks. These
included permission to charge annual fees for
credit cards or loans and lower income tax rates
on bank income. The soliciting state could also
offer lower cost operations and a plentiful, ed­
ucated labor force.
18



Competition for the out-of-state bank’s
operations came from other states. These states
sought either to retain the operations of their
resident banks or to attract operations of banks
located in other states.
Technological improvements have facili­
tated the ability of bank holding companies to
locate certain banking operations in other
states. Thanks to electronic data transmission
and funds transfers and other communications
technologies banking operations no longer need
be carried on in close proximity to the majority
of a bank’s customers.
The marketing strategy of some states was
to be financial pioneers. If successful, the pio­
neer states would become established financial
centers. Once established, bank operations
were unlikely to move back to the home state
or elsewhere unless a new location could dem­
onstrate distinct advantages.
States adjacent to the pioneers have gen­
erally been follow-the-leader states. They have
enacted similar legislation when faced with the
prospect of losing banking operation facilities
to neighboring states.
Pioneer states: South Dakota
and Delaware
South Dakota was the first state to enact
commercial banking legislation specifically
aimed at bringing out-of-state banking oper­
ations to the state to create jobs, expand the
economy, and increase tax revenues. In Feb­
ruary 1980, South Dakota removed all usury
ceilings for credit card loans and other types of
consumer lending effective May 1. Commer­
cial banks, savings banks, savings and loan as­
sociations, and credit unions were previously
held to a 12 percent usury ceiling. In March,
the state further amended its banking laws to
permit an out-of-state bank holding company
to establish a single state or national de novo
bank and move its credit card operations there.
Such a bank was limited to a single banking
office and was to be operated in a manner and
at a location that would not attract customers
from the general public. (Subsequent legis­
lation has eliminated most of these original re­
strictions.)
New York’s Citicorp was the first out-ofstate bank holding company to establish a new
national bank in South Dakota. The bank,
Citibank (South Dakota), N.A., at Sioux Falls,
Economic Perspectives

Table 2
G ro w th a t com m ercial banks
1980 - 1987
United States
Amount
Percent
Total domestic assets (bil.)
Total loans
Loans to individuals
Credit card loans
Employment

Delaware
Amount
Percent

$1,068.9

70

$15.3

300

$39.3

1,067

782.7

96

13.2

453

35.3

1,953

153.3

85

12.7

2,549

26.3

4,100

72.4

243

12.3

207,876

25.0

24,375

96,566

7

3,281

75

13,536

347

was organized to engage principally in nation­
wide consumer credit card lending activities
then currently conducted by Citibank’s New
York banks.1 At the end of 1987, it was the
largest commercial bank in South Dakota, with
domestic assets of SI2.0 billion, total loans to
individuals of $11.6 billion, and 3,462 employ­
ees. (See Figure 1.)
Other out-of-state bank holding compa­
nies from Texas and Nebraska also established
subsidiaries in South Dakota, primarily to offer
credit card services. At the same time, two
large bank holding companies with headquar­
ters in Minnesota expanded consumer loans
and employment at existing subsidiary banks
in South Dakota. Currently, four of the five
largest commercial banks in South Dakota are
subsidiaries of out-of-state bank holding com­
panies and all are located in Sioux Falls.
As a result of the acquisitions and expan­
sions of subsidiary banks by out-of-state bank
holding companies, South Dakota experienced
the fastest rate of growth in the U.S. in loans
to individuals for credit cards at commercial
banks. Additionally, its rates of growth in total
domestic assets, total loans, loans to individ­
uals, and employment were second only to
Delaware. (See Table 2.)
Delaware has long been a state with gen­
erally less restrictive requirements for business
corporations. About 179,000 companies, in­
cluding 56 percent of the Fortune 500 firms and
45 percent of the companies listed on the New
York Stock Exchange, are incorporated in
Delaware.
Since 1981, legislation has been directed
more specifically toward the development of
the financial services industry, with emphasis
on commercial banking.
Federal Reserve Bank of Chicago



South Dakota
Amount
Percent

The Financial Center Development Act
of 1981 (FCDA) was signed into law on Feb­
ruary 18, 1981. It permitted an out-of-state
bank holding company to establish a de novo
bank with a single office operated in a manner
and at a location not likely to attract customers
from the general public in Delaware. How­
ever, the bank could operate to attract and re­
tain customers with whom the bank, the
out-of-state holding company, or such holding
company’s bank or nonbanking subsidiaries
had business relations. The bank was required
to employ within one year not less than 100
persons in the state in its business.
In addition, FCDA essentially eliminated
interest rate ceilings on all types of loans in­
cluding bank revolving credit (i.e., credit cards)
and bank closed-end credit and permitted
banks to charge fees “for the privileges made
available to the borrower under the plan” (i.e.,
annual card fees).
The Delaware legislation also included an
attractive bank tax structure. The rate of tax
on bank net income was revised to 8.7 percent
of net income not in excess of $20 million; 6.7
percent of net income over $20 million but not
over $25 million; 4.7 percent of net income
over $25 million but not over $30 million; and
2.7 percent of net income over $30 million.
The combined state and local marginal tax rate
in Wilmington, Delaware, has been calculated
at 4.5 percent, compared to 24.2 percent in
New York City, 11.7 percent in Philadelphia,
10 percent in Pittsburgh, and 6.5 percent in
Chicago.
The first acquisition approved by the
Federal Reserve Board of Governors under
Delaware’s FCDA of 1981 was that by J.P.
Morgan & Company, Incorporated, a New
York bank holding company, of Morgan Bank
79

Table 3
D e la w a re banking industry changes
D ecem ber 1980 - D ecem ber 1987

Bank group

Total domestic assets
Amount*
%

Total domestic loans
Amount*
%

Employees
Amount
%

4,062.5

10.3

3,498.7

9.9

1,218

9.0

FCDA banks

23,263.4

59.1

20,688.7

58.6

6,400

47.2

CCBA banks

1,390.0

3.5

1,328.1

3.8

158

1.1

Nonbank banks

10,659.7

27.1

9,781.3

27.7

5,793

42.7

Subtotal

39,375.6

100.0

35,296.8

100.0

13,569

100.0

Continuing banks

(41.0)

Total

(22.6)

(33)

39,334.6

Discontinued banks

35,274.2

13,536

Loans to individuals
Amount*
%
Continuing banks
FCDA banks

Credit card loans
Amount*
%

Total deposits
Amount*
%

1,298.1

4.9

747.2

3.0

3,473.0

20.6

14,064.6

53.6

13,575.4

54.4

7,332.9

43.4

CCBA banks

1,328.8

5.1

1,328.8

5.3

612.4

3.6

Nonbank banks

9,570.8

36.4

9,305.0

37.3

5,474.8

32.4

26,262.3

100.0

24,956.4

100.0

16,893.1

100.0

Subtotal
Discontinued banks
Total

(2.7)

0

(36.0)

26,259.6

24,956.4

16,857.1

*mil.$

(Delaware), on November 19, 1981.2 The
newly chartered bank was formed to engage in
wholesale banking to domestic corporations
and financial institutions nationally and inter­
nationally. The bank also planned to partic­
ipate in large loans made by Delaware banks
and in loans to Delaware banks. In its order,
the Board stated, “This increase in available
capital should have a positive impact on eco­
nomic development in Delaware.”
On June 6, 1983, Delaware enacted two
additional state banking laws, the Consumer
Credit Bank Act of 1983 (CCBA) and the
International Banking Development Act of
1983 (IBDA). The CCBA permitted an outof-state bank holding company to establish a
consumer credit bank which was limited to
conducting a nationwide credit card or con­
sumer loan business. Capital requirements
were minimal and there was not an employ­
20



ment requirement. The bank however was re­
quired to be allied with a qualified credit card
processing association that must hire at least
250 employees in its first three years in
Delaware.3 The International Banking Devel­
opment Act of 1983 was specifically aimed at
attracting foreign banks and foreign capital but
it also encouraged the establishment of
internationally-oriented Edge Act banks and
international banking facilities. It removed the
usury ceiling on extensions of credit by inter­
national banking facilities, eliminated reserve
requirements for such facilities, and also ex­
empted their net income from Delaware’s state
franchise tax.
The Delaware legislation eliminating in­
terest rate ceilings on consumer loans and im­
plementing an attractive tax-rate schedule on
bank net income has also encouraged the ac­
quisition of so-called nonbank banks by out-ofEconomic Perspectives

state companies. Such nonbank banks have
usually been acquired or established by non­
bank holding companies, primarily for the
purpose of offering consumer loans and credit
cards, or alternatively, offering commercial
loans but not accepting demand deposits.4
The major contributors to the growth in
the commercial banking industry in Delaware
have been the FCDA banks.(See Table 3.) By
the end of 1987, 17 FCDA banks had opened
and one was pending. Eight are subsidiaries
of bank holding companies located in New
York, and are variously engaged in wholesale
banking, cash management services, nation­
wide commercial lending, as well as consumer
lending and credit card operations. FCDA
banks that are subsidiaries of bank holding
companies in Georgia, Maryland, Virginia,
North Carolina, and Pennsylvania, are prima­
rily engaged in consumer lending and credit
card operations.
Eight nonbank banks were in operation
at the end of 1987 and have been the second
major source of growth in commercial bank
assets and employment, particularly recently.
Increases in total domestic assets and total
loans at nonbank banks have been primarily
the result of the increases in loans to individuals
for credit cards. Growth in employment has
been particularly strong at nonbank banks
compared to other banks in Delaware. Of the
13,536 employment increase at Delaware com­
mercial banks from 1980 to the end of 1987,
5793 or 43 percent was at nonbank banks.
Employment increased 3168 alone at
Greenwood Trust Company in New Castle,
Delaware, after it was acquired by a subsidiary
of Sears, Roebuck and Company in January
1985 and began offering the new Discover
credit card.
The limited-purpose CCBA banks have
had a smaller impact on the growth of assets,
total loans, and employment. Part of this is
attributable to the more recent enactment of
the CCBA legislation. Most of the impact of
the CCBA has been the result of the aggregate
growth of credit-card related loans to individ­
uals. The total effect of the CCBA banks on
employment does not appear in the banks’ fig­
ures because the increases in employment are
primarily at the credit-card processing associ­
ations with which the CCBA banks are usually
affiliated. In fact, only 158 employees were
Federal Reserve Bank o i Chicago



Figure 1
The rew ards o f innovation
millions of dollars

South Dakota

thousands of employees

Delaware

South Dakota

Delaware

reported at the nine CCBA banks in operation
at the end of 1987.
Delaware’s eleven continuing commercial
banks also benefited from the legislation to en­
courage the expansion of banking employment.
They experienced aggregate growth in domes­
tic assets, loans, deposits, and employment well
above the national average. Growth in credit
card loan balances was a major contributor to
the increase in total domestic assets at the con­
tinuing banks. A substantial amount of the
increase was at Mellon Bank Delaware, N.A.,
which had become a subsidiary of the Mellon
National Corporation, Pittsburgh, following
approval of the merger of Mellon with The
Girard Company on March 7, 1983. The
elimination of the usury ceiling in Delaware
provided an incentive for the Pennsylvania
bank holding company to expand credit card
operations at the Delaware subsidiary bank.
In addition, CCBA provided an opportu­
nity for Delaware banks to serve as a qualified
credit card processing association. Growth in
commercial and industrial loans and real estate
loans at the banks was also well above the na­
tional average.
Nearby states: playing follow-the-leader
After the successes in South Dakota and
Delaware, other states, particularly those
nearby, found it necessary to play follow-theleader. Some of the banks in these nearby
states were either moving their credit card op­
erations to South Dakota or Delaware or were
27

threatening to do so. These states, particularly
New York, Nebraska, Virginia, Maryland, and
Pennsylvania found it necessary to take imme­
diate action to retain a competitive edge in the
banking industry.
New York was among the first states af­
fected by the moves, real and threatened, of its
banks to South Dakota and Delaware. New
York’s response was to enact legislation, effec­
tive January 1, 1981, that eliminated ceilings
on interest rates for credit cards and most per­
sonal loans and permitted credit card fees.
The New York legislation, however, did
not reverse the planned move by Citicorp of its
credit card operations from New York to South
Dakota. Furthermore, the state was unable to
compete very successfully for some of its other
banks’ operations with Delaware, which was
much closer to New York and offered both the
absence of a usury rate and much lower tax
rates on bank net income. Consequently, many
of the large New York bank holding companies
established subsidiary banking operations
there.
Nebraska, South Dakota’s neighbor, found
it necessary to enact banking legislation similar
to that in South Dakota. In early 1983, First
National of Nebraska, Inc., Omaha, agreed to
acquire Valley State Bank, Yankton, South
Dakota, just across the South Dakota-Nebraska
border.5 The acquisition’s purpose was to di­
rect the expansion of the credit card business
of its subsidiary, First National Bank of
Omaha, to the Yankton bank. First National,
the largest credit card issuer in Nebraska,
wanted to take advantage of the absence of
usury restrictions in South Dakota.
In response to the First National move
and the apparent success of South Dakota in
attracting out-of-state banking operations,
Nebraska approved similar legislation on April
18, 1983. The legislation eliminated all usury
ceilings on credit cards and allowed out-of-state
bank holding companies to acquire a single de
novo bank in Nebraska whose services were
limited to credit card operations at one office.
Following the passage of the Nebraska
legislation, First National Bank of Omaha said
that it would expand its credit card operations
in Omaha as well as open the credit card sub­
sidiary bank in Yankton, South Dakota.
Virginia, near Delaware, also sought to
retain the credit card operations of its banks.
It eliminated the interest rate ceiling on bank
22




credit card loans effective April 1, 1983. Un­
limited annual fees were also permitted.
On March 15, 1983, Virginia approved
legislation that permitted so-called Financial
Service Center Banks. The legislation author­
ized an out-of-state bank holding company to
acquire a de novo bank, provided the bank was
created primarily to engage in a significant
multi-state credit card operation. The bank
could also engage in limited deposit-taking and
commercial loan operations and was therefore
subject to regulation as a bank.
First Kentucky National Corporation,
Louisville, Kentucky, in April 1984, was the
first out-of-state bank holding company to ac­
quire a de novo bank for credit card operations
in Virginia.6 The bank was acquired to trans­
fer the credit card operations of its Louisville
Bank to the Virginia bank in light of Virginia’s
more liberal revolving credit interest rate and
credit card fee laws.
Maryland. Efforts in Maryland to increase
or eliminate interest rate ceilings and permit
credit card fees encountered strong opposition
from consumer groups and labor organizations.
Legislation was enacted, effective July 1, 1982,
to raise the interest rate ceiling to permit banks
to charge up to 24 percent interest on out­
standing balances on credit cards. Annual fees
on credit card accounts, however, were not
permitted and other restrictions on credit card
operations remained. It was not until July 1,
1983, in a major shift in state policy, that fees
were permitted for credit cards. At the same
time, legislation was enacted similar to that in
Delaware to encourage out-of-state bank hold­
ing companies to establish credit card oper­
ations in Maryland.
The Maryland legislation, however, was
too late to prevent four of its major banks from
moving their credit operations to Delaware.
The holding company of Maryland National
Bank, the state’s largest bank, established
Maryland Bank, N.A., a Delaware subsidiary,
on March 15, 1982. Maryland National Bank
then moved its credit card operations to
Delaware and sold all of the bank’s credit card
accounts to its subsidiary. By September 1982,
three more Maryland bank holding companies
had shifted the credit card operations of their
Maryland bank subsidiaries to de novo banks
in Delaware: First Omni Bank, N.A.,; Subur­
ban Bank/Delaware;7 and Equitable Bank of
Delaware, N.A.
Economic Perspectives

Pennsylvania. Early attempts to increase
the interest rate ceiling on credit card loans and
allow credit card fees in Pennsylvania were
unsuccessful. As in Maryland the opposition
was primarily from consumer groups and union
organizations. Legislation was finally approved
on March 25, 1982 which raised the maximum
interest rate on purchases made with bank and
merchant credit cards and installment con­
tracts from 15 to 18 percent. In November
1982, banks were permitted to charge a card
fee of up to $15 per year.
Opposition to an increase in interest rate
ceilings and permission to charge credit card
fees in Pennsylvania encouraged banks in the
state to move their credit card operations to
Delaware. As noted above, Mellon National
Corporation through its merger with The
Girard Corporation acquired the original
Farmers Bank of the State of Delaware at
which credit card operations were expanded.8
Other Pennsylvania bank holding companies
established FCDA banks in Delaware. PNC
Financial Corporation, Pittsburgh, established
PNC National Bank, Wilmington, on March
10, 1982 and CoreStates Financial Corpo­
ration, Philadelphia, parent of Philadelphia
National Bank, established CoreStates Bank of
Delaware, N.A., on June 1, 1982. More re­
cently, Equibank, Pittsburgh, established
Equibank (Delaware), N.A., Wilmington, on
March 4, 1987.
Effects of general interstate banking
Legislation to permit out-of-state bank holding
companies to acquire limited-purpose commer­
cial banks has been followed by an accelerated
trend toward general interstate banking. The
activities of subsidiary banks are usually not
restricted and the banks are allowed to com­
pete fully with existing instate banks. As a re­
sult, additional alternatives are available for
bank holding companies to locate operations
at subsidiary banks in states where the least risk
and the greatest opportunities for profit exist.
The extent to which the assets of a state’s
commercial banks were held by subsidiaries of
out-of-state bank holding companies at the end
of 1987 ranged from none in seven states to
over 60 percent in five states. These were
Maine, South Dakota, Washington, Delaware,
and Nevada. (See Table 4.) These states also
had a very high proportion of the credit card
Federal Reserve Bank of Chicago



loans and the employment at commercial
banks at the subsidiaries of the out-of-state
bank holding companies.
Policy implications
Emphasis by states on economic develop­
ment goals for the banking industry raised
concerns that sufficient attention was not being
directed toward any inherent safety and
soundness risks associated with interstate ex­
pansions. To date the concerns do not appear
to have been necessary.
Interstate expansion on the basis of de­
regulation in a target state may not have been
the most efficient. Expansion in the home state
or into other states might have been more effi­
cient if the regulatory environments had been
comparable.
The trend toward deregulation of interest
rates and elimination of usury rates has accel­
erated. Adjacent states faced with the move­
ment of bank operations, credit card operations
in particular, to states that had eliminated in­
terest rate ceilings on consumer loans either
raised their own usury rate ceilings or elimi­
nated them.
The trend toward interstate banking has
also accelerated but the emphasis has been on
Table 4
States w ith over 20 percent o f to ta l dom estic
assets at subsidiaries o f o u t-o f-s ta te
bank holding com panies
Percent of total at subsidiaries
of ou t-of-state bank holding companies
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

State
M aine
South Dakota
W ashington
Delaware
Nevada
Arizona
District of
Columbia
South Carolina
Oregon
Connecticut
M ontana
Idaho
Rhode Island
Georgia
North Dakota
Utah
Tennessee
Maryland
Kentucky
Florida
Indiana

Domestic
assets

Credit card
loans
Employees

86
76
73
66
66
59

93
99
92
72
98
54

86
62
70
52
53
56

53
48
45
40
40
37
37
35
32
30
30
26
22
21
21

48
67
44
46
69
53
35
64
47
17
28
57
26
14
42

43
41
48
46
32
36
40
33
27
35
29
24
18
20
22

23

acquisition of existing banks that offer a full
range of bank services. When out-of-state bank
holding companies seek to acquire existing
banks, the number of possible buyers of in-state
banks and generally the sale prices increase.
Additionally, it also encourages bank holding
companies to shift certain operations to subsid­
iary banks in states offering the most attractive
climate for operations.
Overall results in the banking industry
have been similar to those associated with fi­
nancial incentives offered by states to attract
industrial firms. Those states that are the first
to offer new incentives to attract firms generally
succeed in attracting at least a few firms.
Then, faced with the prospect of their firms
expanding elsewhere, other states soon begin to
offer similar incentives. Once a significant
number of states begin to offer similar incen­
tives, the ability of the incentives to affect the
location decision is lost.
In the case of bank deregulation, when
the regulatory environment in all states be­
comes roughly similar, the ability of states to
successfully use deregulation as an incentive for
economic development is also lost.
1 Citicorp, New York, New York, 67 Federal Reserve
Bulletin 181 (February 1981).
1 J. P. Morgan & Company, Incorporated, New
York, New York (Morgan Holdings Corp.,
Wilmington, Delaware), 67 Federal Reserve Bulletin
917 (December 1981).
3 A consumer credit bank, because it does not make
commercial loans, is not considered to be a bank for

24



the purposes of the Bank Holding Company Act of
1956, as amended, and is specifically excluded from
the definition of a bank in the Competitive Equality
Banking Act of 1987 with given restrictions. It is
therefore not subject to the Douglas Amendment
restrictions on interstate banking.
4 The Bank Holding Company Act of 1956
(BHCA), as amended, defined a commercial bank
subject to regulation under BHCA as one that ac­
cepted demand deposits and made commercial
loans. If both of these conditions were not present,
national or state chartered banks were not subject
to regulation under BHCA, and became known as
nonbank banks. The Competitive Equality Bank­
ing Act of 1987 redefined the term bank to include
an FDIC-insured institution whether or not it ac­
cepted demand deposits or made commercial loans.
Companies that had acquired nonbank banks on
or before March 5, 1987 were grandfathered and
were permitted to retain the bank and not be reg­
ulated as a bank holding company but were gen­
erally restricted to exis'ting activities and limited to
an annual rate of growth in assets of seven percent.
5 First National of Nebraska, Inc., Omaha,
Nebraska, 69 Federal Reserve Bulletin 390 (May
1983) .
6 First Kentucky National Corporation, Louisville,
Kentucky, 70 Federal Reserve Bulletin 434 (May
1984) .
7 Maryland National Corporation, Baltimore,
Maryland, 68 Federal Reserve Bulletin 203 (March
1982).
8 First Maryland Bancorp, Baltimore, Maryland,
68 Federal Reserve Bulletin 320 (May 1982); Sovran
Financial Corporation, Norfolk, Virginia, 72 Fed­
eral Reserve Bulletin 276 (April 1986).

Economic Perspectives

BULK RATE

B Q H H IS M S O e c o n o m ic

PERSPECTIVES
Public Information Center
Federal Reserve Bank
of Chicago
P.O. Box 834
Chicago, Illinois 60690
D o Not Forward
Address Correction Requested
Return Postage Guaranteed




U S. POSTAGE
PAID
CHICAGO, ILLINOIS
PERMIT NO. 1942