Full text of Economic Review (Federal Reserve Bank of San Francisco) : March 1958

View original document

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

T W E L F T H F E D E R A L R E S E R V E DI S T R I C T

FEDERAL

RESERVE

BANK

March 1958

OF SAN

FRANCISCO

The Current Reporting Series as a
Guide to Business Activity. . .
Review of Business Conditions . .

The Current Reporting Series as a Guide
to Business Activity
task of predicting movements in business
activity is a hazardous one, and the way of the
predictor is hard. Nevertheless, that branch of
theoretical economics dealing with the explana
tion of business cycles and forecasting fluctu
ations in the level of activity has not wanted for
laborers in the vineyard. The theoretical ap
proaches have varied from explanations of cy
clical fluctuations in terms of sunspots to descrip
tions of the economic system in terms of formal
mathematical structures of a forbidding number
of equations. Truly, the investigator in this area
will not lack for guidance.
h e

A series such as the W eekly Reporting Mem
ber Bank Series prepared and released by the
Federal Reserve System could easily be con
strued as an important indicator of business
conditions. The series is not intended to be in
terpreted in this fashion. It is a measure of bank
ing operations— nothing more. However, the line
of distinction between purely banking statistics
and indicators of general economic activity tends
to become blurred. Banking is such an integral
part of the general scheme of production that the
movements of banking data follow a path roughly
parallel to that of the level of business activity.
It is the work of this second part of the study
to determine how rough this path is and how
smooth the correspondence between the two sets
of data.
In brief summary, the following relationships
have been found to exist between the current
series for Weekly Reporting Member Banks and
certain indicators of business activity:
1. Total loans and business loans of reporting
banks show a tendency to lag behind down
turns in general activity.
1This is the second part of a study of banking statistics made avail
able by the Federal Reserve System and taken from the “ Principal
Resource and Liability Items of Reporting Member Banks in Lead
ing Cities,” a report released each Wednesday noon by the twelve
Federal Reserve Banks for their respective districts and by the Board
of Governors of the Federal Reserve System for the nation. The first
part of the study was published under the title “ The Current Re
porting Series as a Guide to Banking Activity” in this Review, Sep
tember 1957, pp. 114-126.
We wish to acknowledge our indebtedness to Dr. Geoffrey H.
Moore of the National Bureau of Economic Research for making
available data of the National Bureau and for substantive and edi
torial comments on a draft of this article.

Digitized
34for FRASER

2. No significant relation was found to exist
between Weekly Reporting Member Bank
Series and business upturns.
3. Industrial stock prices, wholesale prices of
commodities, and the index of industrial .
production all tend to turn down before
total loans and business loans.
4. Industrial stock prices generally turn down
before loans for carrying securities and
usually turn up before security loans do.
5. Business loans tend to lead inventories.
The business cycle

What is the business cycle? First, we may
disavow a persistent notion that there is a singu
larly unique pattern of fluctuations, a regular
rhythmical movement of economic activity that
corresponds in some manner to the cyclical pat
terns of the natural processes. Statistical studies
have shown, however, that there is a periodicity
attachable to many types of business activity. It
has been suggested that there is not one business
cycle, but rather three: a long cycle of about
fifty years duration, a cycle about eleven years
long, and a short cycle of three to four years.
Many business series, of course, do not follow
any of these general fluctuations with regularity.
This study will be concerned with an examina
tion of the behavior of the Weekly Reporting
Member Bank Series, hereinafter abbreviated
W R M B S , in the shorter or “ minor” cycle as it is
sometimes called.
There are clearly perceptible patterns of cy
clical behavior in particular industries or sectors
of the economy. Fluctuation in the production of
pig-iron is a classical example of cyclical move
ments. It is widely considered that the construc
tion industry has a cycle peculiarly its own with
a duration of about 18 years.1 But still the ques
tion of why there are business cycles is a much
disputed one and even the positive identification
of cycles is an uncertain business. It is clear to
most observers that the economy as a whole is
usually in a stage of expansion or contraction,
1 See, for example, Long, C. D., Jr., Building Cycles and the Theory
of Investment (Princeton: Princeton University Press), 1940.

March 1958

MONTHLY REV IEW

but it is much less clear where one stage passes
into the other. The turning points of crisis and
contraction at the peak of the cycle and depres
sion and expansion at the trough of the cycle are
not readily discernible. There is no surefire indi
cator that will at all times tell the turning point.
It is an easier task to determine whether the
economy is prosperous or depressed— all one
need do is look about him. If all who seek it can
find employment, if prices are advancing and
production is increasing, we are in the expansive
phase. If jobs are scarce and production in gen
eral is low, if prices are falling and unsold goods
glut the shelves of sellers, we are in the contrac
tive phase of the cycle. This much is apparent
even to the most casual observer. But to the
questions of how far advanced we are in the par
ticular phase and how soon we may expect a
turning point, the answers are seldom clear and
never certain. The spectres of the predictions of
unending prosperity in 1929 and the auguries
of mass unemployment in 1946 return again and
again to haunt the prophets.
Indicators o f activity

The W R M B S have what would appear to be
ideal qualities for business indicators since they
are available on both a national and a regional
basis and are published with a reporting lag of
only one week. H ow well they will serve in this
capacity depends, of course, upon the relative
strength or weakness of the relation between
banking activity and general business activity.
In the complexity of a modern commercial and
industrial economy, it is only a single remove
from the money market to the market place
where final products are bought and sold. Ours
is a money economy and information concerning
the money market implies information about the
business sector of the economy. The direct con
tribution of banking activity to the total produc
tion of final goods and services— the Gross Na
tional Product— is very small, amounting only to
net interest paid and wages and salaries paid by
banks. However, the banking system is a very
necessary adjunct to production and, in general,
an expansion or contraction in general economic
activity will be paralleled by an expansion or
contraction in banking activity. The most sig

nificant indicators of business activity are often
available only on a monthly or quarterly basis
and then only after a considerable time lag. The
movements which they describe may already
have taken place, thus nullifying their value as
predictive indicators. Therefore, if it could be
established that the W R M B S describe a typical
pattern in the business cycle, current banking
data would provide a basis for interpreting the
cycle. If this should prove not to be the case, it
would prescribe limitations to the use of such
data in analyzing the current situation.
The National Bureau approach

This part of the analysis will be based in large
part upon techniques developed and used by the
National Bureau of Economic Research in their
studies of the cyclical behavior of certain series of
economic data. In analyzing 800 time series, some
of them spanning 75 years and more, the Na
tional Bureau has recognized certain specific
cyclical patterns (o r the lack of such patterns).
The behavior of these patterns taken in the ag
gregate determines the behavior of business
cycles. There is no one particular series whose
fluctuations may be called “ the business cycle.”
The business cycle as defined by the National
Bureau represents the aggregative movement of
all economic series; that is, an expansion is de
fined as a phase in which the majority of eco
nomic series are expanding, though not all of
them nor at the same rate. And, similarly, the
contraction of activity represents the aggregative
contractive movement produced by the decline of
the majority of all series. T o serve as a touch
stone for comparing various series, the National
Bureau has set up “ reference cycles.” These con
sist of the dates at which the peaks and troughs
of total activity occur and the phases of expan
sion and contraction included between these
turning points.
It should be emphasized that the term “ ref
erence cycle” is an abstraction bearing no precise
correspondence to particular economic series,
though many broad aggregates, such as the Fed
eral Reserve index of industrial production,
usually move very closely in accordance with it.
The reference cycle is indicative of the aggrega
tive movement of the total of all “ specific cycles/’

FEDERAL R ESE RVE BANK OF SAN F R A N C I S C O

the designation given to the individual cycles pe
culiar to each series. The National Bureau ref
erence cycle dates have been determined by an
exhaustive study of business annals and of the
turning points of specific series. Chart 1 gives a
simple illustration of how the reference dates are
selected.
C hart 1

I L L U S T R A T I O N OF T HE D E T E R M I N A T I O N
OF R E F E R E N C E C Y C L E T U R N I N G PO IN TS
FROM SPEC IFIC C YC LES

REFEREN CE
TROUGH

R EFER EN C E PEA K

REFEREN CE
TROUGH

The reference peaks and reference troughs
represent the points in time around which the
peaks and troughs of the specific series tend to
be clustered. No attempt is made to construct a
composite curve from the specific curves. The
reference cycle is a consensus rather than an
actual cyclical pattern. A reference peak depicts
the month by which most of the series have
started to turn down and similarly the reference
trough represents the date by which a majority
of specific series have started to turn up. Table
1 gives the National Bureau reference turning
points for the period to be considered.

from the Federal Reserve Bulletin on a currently
reported basis since 1942. They have not been ad
justed for seasonal variation or trend because
they are not currendy published nor used by an
alysts in adjusted form. W e are concerned with
their interpretation as presently published. Chart
2 shows the W R M B S for the period following
W orld W ar II.
The series "Total Loans and Investments,”
most comprehensive of the asset items of Report
ing Member Banks, has not been included for ex
ploration in this portion of the study because
the behavior of loans differs from that of invest
ments during the business cycle. Movements in
the loans and investments of banks may and do
occur in opposing directions. A fitting example
of this is the behavior of total loans and invest
ments of commercial banks in the period 1945 to
1950. The series showed a sharp decline from the
1945 high, reaching this level again only in 1950.
This movement masked two significant changes
in the structure of commercial bank assets: an
abrupt decline in the holdings of United States
securities and a brisk rise in total loans extended
by banks. Chart 2 gives a striking illustration of
these movements since 1945. This series, then,
may conceal about as much as it discloses.
Total loans

T o gain a perspective of how fluctuations in
the level of general banking activity conform to
movement in over-all activity, we will start with
a consideration of total loans of reporting banks.
T able 1

D A TES OF R E FER E N C E C Y C LE P E A K S AND
T R O U G H S IN T H E U N ITED S T A T E S , 1919-54
Monthly Seftrence Dates
Trough
Puk

Expiruion

Duration In Months
Contraction

{jd»

A pril 1919

WRMB Series To Be Considered

Ja n . 1920

July 1921

M ay 1923

July 1924

The procedure to be followed in this inquiry
will be to examine the various series of the
W R M B S for the cycles for which data are re
ported. The W R M B S to be studied have been
plotted on a monthly basis from 1919 to the
present, using the average figures for the month.
The data are drawn from Banking and M one
tary Statistics for the period 1919 to 1941, and

Oct. 1926

Nov. 1927

June 1929

M ar. 1933

M ay 1937

June 1938

Feb. 1945

Oct. 1945

Nov. 1948

Oct. 1949

July 1953

Aug. 1954

Digitized for
36FRASER

Source: Arthur F. Burns and Wesley C. Mitchell, Measuring Business
Cycles, (New York: National Bureau of Economic Research) Table
16. Revisions of some dates and dates subsequent to 1938 were
supplied by the NBER.

March 1958

MONTHLY REV IEW

But suppose that the
S E L E C T E D REPORTING S E R I E S
scope of bank lending is
BILLIO N S OF D O LLARS
19 4 5 -5 5
widened, as it has been
since this early tradition
was established. Suppose
that loans are made for
purposes other than for
trade and by persons out
side the business commu
nity. Consider, for exam
ple, consum er loans. If
loans are contracted to
buy nezv autos, appliances,
etc., this is a fair indica
tion that business is ex
panding (1 ) because con
sumer income is high or
rising and (2 ) because
sales of consumer goods
are high or rising. But
suppose a loan is used to
buy an old house or a used
car, or to buy shares of
outstanding securities. In
this case, the money sup
Source: Board of Governors of the Federal Reserve System, Federal Reserve Bulletin.
ply has been expanded by
the amount of the loan, but there is no corre
Loans represent the extension of credit by
sponding increase in the production of new
banks to the other sectors of the economy. Mak
goods. Inclusion of these categories, then, weak
ing loans is a major part of the normal business
ens but does not destroy the case for using total
operations of a commercial bank. The bank serves
loans as a measure of business activity. Since
also as a depositary for the public’s funds, but
production is used as an important gauge of eco
its principal activity is trafficking in claims on its
nomic activity, total loans, some of which are
own good name. An increase in loans is reflected
made merely to facilitate a transfer in the owner
in an increase in demand deposits, since the bank
ship of existing goods, may generally be broadly
usually credits the proceeds of the loan to the
sympathetic with the movement in the business
checking account of the borrower. Traditionally,
the “ proper” sort of loan that a bank should
cycle but can not be expected to pin down its
turning points.
make is the self-liquidating type of loan to a
businessman. This loan enables the businessman,
Business loans
say a merchant, to lay in a stock of goods, the
W e may obtain a closer correspondence to
subsequent sale of which provides receipts out of
business
activity by shifting our attention from
which the loan is retired. Bank loans were never
total loans to the W R M B series “ Commercial
to be made for the purpose of providing capital
and Industrial Loans” (to be referred to from
to the firm, since this was the function of the se
this point on as “ business loans” ) . 1 There are
curities market. In other words, every loan pre
Chart 2

sumably led to an expansion of trade. If the total
of loans outstanding was high, it was so because
the level of economic activity was also high.

‘ From June 1937 to January 1956 the business loan series used was
“ Commercial, Industrial and Agricultural Loans.” Prior to June
1937, it is assumed that business loans make up the greatest part
of the difference between reported “ Total Loans” and “ Loans on
Securities."

FEDERAL RE S ER VE BANK OF SAN F R A N C I S C O

260 standard classifications of types of business
in this grouping, so it cannot be expected that
the aggregate movement of the series will reflect
changes in each and every component class of
business. The traditional short-term, self-liqui
dating business loan has been modified consider
ably over the years. One of the qualifications is
the increased resort to the business “ term loan”
in recent years, accounting for over a third of
all business loans in member bank portfolios in
1955.1 These are loans having a maturity of from
one to five years. They are usually repayable in
instalments rather than in lump sum payments.
H ow would the increasing use of term loans
affect cyclical movements in loans outstanding?
It would seem that upward movements of loans
in an expansion of the economy would be more
pronounced, while, if the loans are to be retired
in instalments, declines would be more gentle
than if loans were to be repaid in a single pay
ment. Term loans also tend to be larger than
single-payment, self-liquidating loans.
A s a further modification, not all short-term
business borrowings are transacted with the
banking system. Businesses may obtain credit
from their suppliers, and book credit may be ex
tended by business firms to their customers; or,
if the firm is sufficiently well known, it may make
use of its own paper in the money market. H ow
ever, there is no reason to believe that the cyclical
behavior of these types of non-bank credit will
show a significant difference from the pattern of
bank credit extended to the business community.
Loans fo r carrying securities

This series was the largest specific loan series
in the W R M B S prior to 1934. In the postwar
years it has been relatively unimportant in the
total of loans outstanding, possibly because of the
higher margin requirements imposed upon secu
rities purchases. Pronounced movements in stock
market activity are not solely the result of sharp
increases or decreases in production, but rather
are due in large part to reactions to heightened
or depressed expectations of the future course of
affairs. A breath of uncertainty about the politi
cal situation or about events overseas may trigger
a short but sharp turn and, as a result, a change
1 “ Business Loans of Member Banks,” Federal Reserve BulUtin,
April 1956, pp. 332-334.

in loans for securities. This again is for the most
part a loan to facilitate an exchange of existing
assets. If loans for carrying new securities issues
could be separated out from all loans on securi
ties, this series might move sympathetically or
precede rises or declines in production activity.
Since this breakdown is not feasible, little may
be expected from loans on securities as an indi
cator. It should be noted that changes in margin
requirements may affect the demand for loans
for carrying securities quite apart from changes
in business conditions.
Bank holdings of United States securities

A consideration of United States securities
holdings by reporting banks would appear at first
glance to hold no relevance for a study of the
business cycle. It should be remembered, how
ever, that commercial banks are profit-motivated
institutions and their earnings derive from the
assets that they hold. These assets may consist of
loans to business or to individuals, or they may
consist of bonds purchased by the bank. The pru
dent banker will arrange his asset portfolio with
an eye to earnings as well as safety, so that he
will have a source of income at all times. T o do
this, he may purchase securities, either govern
ment issues or corporate bonds when borrowers
are disinclined to borrow, as in the 1930’s, or he
may sell securities or let them mature when the
demand for loans is brisk, as has been the case
since W orld W ar II. Since banks will alter their
portfolios in response to changing business con
ditions, this series may reflect these changing
conditions.
Demand deposits

Thus far we have mentioned only the asset
side of the banks’ balance sheet. H ow does the
cyclical behavior of demand deposits compare
with that of loans and investments? Since, for
the banking system, increases in loans outstand
ing give rise to increases in demand deposits, it
would seem the one might be as reliable an indi
cator as the other. This is true, although consid
eration of demand deposits brings to light an
important additional factor which is attributable
only to deposits, not loans. A loan represents a
claim against the individual, the borrower, which

MONTHLY R EV IE W

March 1958

is not negotiable or transferable, and which is
offset by an initial claim against the bank— the
increase in his checking account— in favor of the
borrower, which is negotiable and transferable.
It is this element of transferability, or the prop
erty of passing as money, that is important. The
demand deposit arising out of the loan trans
action can turn over slowly or quickly, while the
loan itself is fixed. This possible variation in the
rates of use of demand deposits is not reflected
in the total figures for deposits outstanding, but
must be sought in estimates of the velocity or
turnover of money, most of which consists of
bank deposits.

Behavior of the WRMBS
in the Business Cycle
Having defined the business cycle as the Na
tional Bureau reference cycle for the purpose of
this study, the W R M B S will be overlaid upon
this cycle and compared with it in three ways.
The first and perhaps most significant test will
be to compare the timing of the turning of the
W R M B S to the turning points of the reference
cycle. The second comparison will be made be
tween the general movements of the W R M B S
and the reference cycle between turning points.
The third test is a comparison of the average duTa b l e 2
ILLU STR A T IO N O F M E T H O D
D I F F E R E N C E S IN M O N T H S B E T W E E N B U S I N E S S
LO A N A N D R E F E R E N C E C Y C L E TU R N IN G PO IN TS
T W O P O S T -W O R L D W A R II C Y C L E S ,
PEA K S AND TROUGHS
Differences
at Peak
Month*’

Deviation c l
Observation
From Average
of Observations*

Differences
at Trough
Months3

Deviation of
Observation
from Average
of Observations

2.7

Cycle 1

Cycle 2

A V ER A G E

+ 2 .5

(1.5)

- 2 .3

(1.8)

1The minus and plus signs refer to the number of months by which
the business loan turning points precede or follow the reference
turning points. A zero indicates that the two turning points coin
cide, that is, occur in the same month.
1 For explanation, see text and Chart 3.
3There is always one more trough than the number of cycles, since
the fluctuations are measured from trough to trough.

C hart 3

COMPARISON OF B U S IN E S S LOAN TROUGHS
TO R E F E R E N C E TROUGH IN TWO
R E F E R E N C E CYCLES, 1945-54

rations of the W R M B S cycles and their phases
with the average length of the reference cycle
and its component parts.
The method of comparison used in the first
test will be to compare the dates of the peaks and
troughs of the W R M B S to the corresponding
reference peaks and troughs. This is done by
taking the average of the algebraic sum of the
differences in months between the specific W R
M BS turning points and the reference turning
points. An examination of the relationship of the
turning points of business loans and reference
turning points in the two cycles following World
W ar II should serve to illustrate the method of
comparison that is employed.
T o illustrate the method of comparison further,
the business loans troughs or lower turning
points are pictured in Chart 3 using the reference
trough as a levelmark and plotting the business
loan troughs relative to this mark. Reading from
the preceding table, we see that in the two cycles
following W orld W ar II, business loans turned
up an average 2.3 months in advance of the gen
eral business recovery. The mean or average de
viation defines a lead range of from —4.1 months
to — 0.5 months about this average. This is ob
viously not the range of observations, since some
of the original observations lie outside of it. But
as the average is a measure of central tendency,
the mean deviation describes the average amount
by which the figure given, the average lead in
this case, varies from the original observations.

FEDERAL RE SERVE BANK OF SAN F R A N C I S C O

able

C O M PA R ISO N O F TURNING PO IN TS O F R E F E R E N C E
C Y C L E S AN D S E L E C T E D R EPO R TIN G S E R IE S
Toll! loanj
Average
Months
Lead (— } or
Lag (+ )

Reference
Peak
Reference
Trough

Mean
Deviation
Around
Average

Business Loans
Average
Months
Lead (— ) or
Lag (+ )

Mean
Deviation
Around
Average

Loam For Carrying Securities
Average
Months
Lead I— ) or
lag (+)

Mean
Deviation
Around
Average

Holdings of li. S. Securities
(Straight)
(Inreried)
Average
Hean
Average
Moan
Months
Deviation
Months
Deviation
Lead (— ) or
Around
Lead {— ) or
Around
Lag (+ )
Average
Lag (+ )
Average

Demand Deposits
Average
Mean
Months
Deviation
Lead (— ) or
Around
Lag (+)
Average

7 Cycles, 1919-38, 1945-54
+ 4 .8 ( 1.8)

+ 4.811.8)

2.6 ( 6.9>

+ 6 .0 ( 10 .0 )

+ 7.019.3)

+ 18.8(13,9)

-1 8 .8 (1 3 .2 )

+ 0.7(12.7)

3.814.5)

2.0 ( 8.8)

+ 3.7 ( 5.9)

3.1 (2.8)

2 Cycles, 1945-54
Reference
Peak
Reference
Trough

+ 3 .0 ( 2.0)

+ 2.511.5)

- 1 .0 ( 1 1 .0 )

- 3 7 .5 ( 4 . 5 )

+ 1.51 3.5)

-1 0 .0 ( 0>l

- 3 . 0 < 2.0)

-2.3(1.81

+ 17.0(18.0)

- 1 0 .5 ( 4.5)

+ 3.31 0.4)

- 4 5 .0 (1.5)1

-2 0 .6 (1 3 .7 )

+ 6 .8 I 5.2)

1.515.0)

+ 6.9( 5.0)

3.813.4)

6 Cycles, 1919-27, 1933-38, 1945-54
Reference
Peak
Reference
Trough

+ 5.0 ( 2.0)

+ 4.812.2)

2.5 ( 8.5)

+ 0.41 5.0)

+ 1.2(5.61

+ 14.2(16.2)

1.2(11.3)

3 Only one cycle.

The mean deviation serves two purposes. First,
it gives a clue to the degree of dispersion of ob
servations about the average. Second, if the mean
deviation does not exceed the average lead (as
it does not in the above case), the turning points
of the tested series have generally led the refer
ence turning point. In the event that this aver
age deviation exceeds the average lead or lag of
the W R M B S turning point, the range straddles
the reference turning point and the W R M B S has
both led and lagged the reference turning point in
the course of the several cycles examined.
Behavior o f the W RM BS at the turning points

W e are now ready to examine Table 3, which
summarizes the comparisons of the timing of the
turning points of five selected W R M B series
with those of the reference cycles in three dif
ferent combinations of cycles between 1919 and
1954. The first comparison includes all seven of
the cycles of this period with the exception of the
W orld W ar II years. In the second, the two post
war cycles are considered apart from the prewar
cycles and presented in the fashion in which they
best correspond to the reference cycle. Third, the
cycle from 1933 to 1938, which began with the
trough of the Great Depression, is omitted from
this grouping of cycles because the low level of
activity during this cycle makes it very unlike the
others examined. The differences apparent in a
Digitized 40
for FRASER

single series caused by a selective grouping of
cycles clearly illustrate that each cycle differs
from every other cycle. The omission of the 193338 cycle has a considerable effect on the timing
of the lower turning point of total loans and
business loans.
In an examination of the selected W R M B
series in seven cycles, there are three series which
demonstrate a distinctive and consistent pattern
at the turning points. Total loans and business
loans lag the reference peak by identical inter
vals of 4.8 months. Demand deposits lead the
reference trough by an average of 3.1 months.
These are the only instances in which the points
all fell on one side of the reference turning point.
The deviations of the loans for carrying securi
ties describe such a wide range within which the
turning points may precede or follow the refer
ence turning points that nothing can be said with
certainty about their characteristics.
Bank holdings of United States securities led
the reference cycle by such a wide margin that
it was thought advisable to test this series by
inverting it, that is, by comparing the peaks of
the specific series with reference troughs and the
troughs of this series with reference peaks. While
the straight peak to peak comparison showed
that in every instance bank holdings of Gov
ernment securities turned down before the ref

March 1958

MONTHLY R EV IEW

trough, and tend to la g ; but, as the mean devia
tion demonstrates, they have led the reference
upturn in some cyclical recoveries. The inverted
series, holdings of United States securities,
shows the most consistent timing. In the six
cycles examined, the turning points of this series
lag the reference cycle at both the peak and the
trough, though within a rather wide range of
from about 1.5 months to almost a year. Indeed,
this series is the only one of the five W R M B S
that shows a consistent pattern of behavior at
both of the cyclical turning points. Demand de
posits of reporting banks turn up at the trough
before general activity experiences a recovery,
but within a rather wide range.

erence peak, the lead was at least 18 months with
a mean deviation of over a year. When the series
is shown as inverted, most of the leads now
become lags, with the average lag being much
closer to a reference turning point than was the
average lead with a peak to peak and trough to
trough comparison. The typical behavior of the
peaks of the specific cycles for bank holdings
of United States securities would thus appear to
occur soon after a reference trough and far in
advance of a reference peak. The smaller mean
deviations that are obtained by using the inverted
series justify this manner of treatment.
In the two postwar cycles total loans and
business loans show a much more distinctive
and consistent pattern than any of the other series
at both the upper and lower turning points. Total
loans, for example, lag by three months at the
peak with a mean deviation of two months and
lead at the trough by exactly the same amounts
and deviation. Business loans in these two cycles
have moved closer to the reference turning points
in these two cycles than in the other comparisons
and show a tendency to lead at the trough. These
two were the only series in the two most recent
cycles studied that showed a consistent pattern
of behavior at both of the turning points.

The W RMBS exam ined between turning points

Table 4 is concerned with the behavior of the
W R M B S examined between the turning points ;
that is, during the expansive and contractive
phases of the reference cycle.
Chart 4 serves to illustrate the behavior of
total loans in the reference cycle with the indi
vidual cycles identified by date.
The behavior of total loans in the expan
sion phase supplements what was said about the
timing of the turning points of this series. In the
expansion phase—-from reference trough to peak
— this series either rose throughout the entire
interval, or fell at the beginning of the phase and
then started to rise for the rest of the expansion.
In the first instance, that of a continual rise, there
were no turning points in the expansion phase.
The trough of the series might have coincided
with the trough of the reference cycle, or it
might have preceded it; and the peak of this

Comparison of the turning points of the five
W R M B series in six cycles shows the most dis
tinctive patterns of behavior. Total loans lag
behind the reference peak within the narrow
range of three to seven months. Business loans,
on the average, decline at the upper turning
point before total loans but show a wider range
of variability relative to the reference peak. Both
of these series are much less reliable at the
T

arle4

G E N E R A L BEH A V IO R O F S E L E C T E D R EP O R TIN G BAN K S E R IE S
DURING T H E R E F E R E N C E C Y C L E , 7 C Y C L E S , 1919-54

Behavior

Rose

Total Loans
Expansion
Contraction

Rose-Fell

Bull ties: Loans
Expansion
Contraction

Fell-Rose
Fell

Loans for Carrying
Securities
Expansion1
Contraction

Holdings of U .S.
Securities
Expansion
Contraction

Demand Deposits
Expansion
Contraction

1 Data available for only six expansions.

FEDERAL R ES ER VE BANK OF S A N F R A N C I S C O

C hart 4

CYCLICAL BEHAVIOR OF TOTAL LOANS
D U R IN G S E V E N R E F E R E N C E C Y C L E S

the cyclical behavior of this series since W orld
W ar II. If the turning points for the postwar
series are inverted, a somewhat better fit is ob
tained for this period. Bank holdings of United
States securities conform rather well to the ref
erence cycle with four of seven peaks in the ex
pansion phase and four of seven troughs in the
contraction. But as was mentioned, the peaks of
this series tend to occur so early in the expansion
that they might better be treated as lagging the
reference trough.
Comparison of W RM BS with
reference cycle lengths

R E F E R E N C E TROUGH

R E F E R E N C E PEA K

R E F E R E N C E TR0UGH

series either coincided with the reference peak
or followed it. In the case where the series fell
and then rose during the expansion phase, the
lower turning point of the series clearly lagged
the reference trough. In no case did the upper
turning point or peak of the total loans series
precede the peak of the reference cycle. The be
havior of total loans in the contraction of the
reference cycle, while less characteristic than
the expansion, also has a story to tell. The most
typical behavior of this series is to rise and then
fall in the contraction. This means that the upper
turning point most often falls in the contractive
phase, which squares with what has been said
earlier about the timing of the peaks of this series.
The instance in which this series rose during a
reference contraction was in the 1920’s when
both total loans and business loans expanded
from 1923 to 1929, extending over more than
two complete reference cycles.
Demand deposits, while rising throughout
most of the expansion phases, also showed a
tendency to rise in a good share of the contrac
tive phases. In two cases this series moved contra-cyclically in the contraction, rising through
out. In three other cases the series reached a
trough early in the contraction and rose for the
rest of this phase.
Loans for carrying securities give the appear
ance of contracting during a considerable part of
the boom. There seems to have been a change in
Digitized42
for FRASER

The timing of the turning points and their
behavior during the separate phases of the ref
erence cycle indicate that the specific cycles of
the W R M B S do not correspond or coincide with
the reference cycle. Table 5 gives the average
length of the expansion and contraction phases
of the reference cycle, together with the average
length of the specific cycle for the seven cycles
considered and for the postwar cycles.
The average length of the reference cycle dur
ing seven complete cycles from 1919 to 1954
was 48 months, with the expansion phase about
one and one-half times the length of the contrac
tion. In the two postwar reference cycles, the
cycle averaged 53 months, with the expansion
about four times as long as the contraction. These
postwar contractions were, of course, of a much
milder nature than the sharp dips of 1920 and
1937 and the severe and protracted decline from
the 1929 heights. They were more in the nature
of pauses in the general postwar expansion than
T

able

COMPARISON OF THE AVERAGE DURATION OF
EXPANSIONS AND CONTRACTIONS OF SELECTED
REPORTING SERIES WITH REFERENCE CYCLES
7 Reference Cycles, 1919-54
Expansion Contriction
[y d s

Humber »l

Specific

» ! ! e r « c e C jc U l, 1 94S-54

Circles

fip a w o n

C w itrsctieit

[jrte

. 4 4 .9

1 2 .0

5 3 .0

Reference C ycle

2 9.9

18.1

4 8 ,0

Total Loans

4 5.8

25.2

7 1 .0

4 8.5

7 .5

5 6 .0

B u sin ess Lo an s

4 4.8

2 4 .6

6 9 .4

4 7 .0

8 .5

5 5 .5

Lo an s for C arrying
Securities

3 7.5

3 3 .2

70.4

2 0 .5

3 0 .0

9 .5

. ::

Holdings of U. S.
Securities
D em and Deposits

1 One cycle.

22.1
3 6 .2

22.3
18.6

4 4 .4
5 4.8

1 3 .5

I; 30.0 '

3 9 .0

52.5

1S.O>

45.0'
•

March 1958

MONTHLY REV IEW

retreats. In general, the specific cycles for the
loan series are longer than the reference cycles.
The relative proportions of the separate phases
of the total loans and business loans cycles are
nearly identical to the proportions of reference
expansions to contractions, while the expansions
of the loans for carrying securities cycle were
shorter relative to the contractions. In the two
postwar cycles, total loans and business loans
expansions were both absolutely longer and also
proportionally greater relative to their contrac
tions than the reference.
The phases of holdings of United State se
curities are almost evenly divided between ex
pansion and contraction when all of the cycles
are considered. In the two cycles following
W orld W ar II, contractions averaged about
three times as long as expansions due to the fact
that there have been heavy sales by banks since
1946. The demand deposit series in the 191954 interval shows a complete cycle shorter than
the reference cycle, though the lengths of expan
sions relative to contractions is about the same.
Summary o f com parisons:
W RM BS with reference cycle

What general conclusions may be drawn from
the relationship of the W R M B S to the National
Bureau reference cycles as representative of the
business cycle? It has been seen that the timing
of the W R M B S in certain cases seems to fall
into a distinctive pattern: total loans and busi
ness loans turn down from their peaks after the
reference cycle peak has done so. Demand de
posits usually recover from a contraction trough
before the reference cycle does. Bank holdings
of United States securities follow a well defined
course if the series is considered as inverted, but
still this series lags behind the reference cycle.
None of the W R M B series considered can give
us that clue more sought after than the Holy
Grail— an accurate and dependable guide to the
end of the boom.1
The behavior of the W R M B S in the separate
phases of the reference cycle is somewhat varia1This is consistent with the results in a study by Geoffrey Moore,
Statistical Indicators of Cyclical Revivals and Recessions, Occas
sional Paper No. 31, National Bureau of Economic Research (New
York, 1950) in which most of these series were tested and found
not acceptable as indicators.

ble, again except for these three series, total
loans, business loans, and demand deposits, each
of which exhibits a fairly stable pattern of be
havior in only one phase of the cycle. The ex
pansion and contraction phases of the W R M B
series differ sufficiently in duration from those of
the reference cycle so that, as in the 1920’s, more
than an entire reference cycle may be contained
within the cycle of one of the W R M B series.
(Chart 4)
It is perhaps not the most appropriate com
parison to put the W R M B S against a series
which actually represents a consensus of general
activity and which is not available on a current
basis, containing as it does many series which are
available only after a wait of several months. It
might prove more enlightening to compare the
W R M B S with series of general scope which are
available on a more current basis.

The WRMBS and Specific Indicators
In an effort to isolate certain specific series
which might call the turns of fluctuations in gen
eral business activity, Geoffrey Moore of the Na
tional Bureau of Economic Research has selected
21 series of which 8 showed a tendency to lead
general activity, 8 coincided with reference cycle
turning points, and 5 lagged.1
It would be too ambitious an undertaking
for this study to attempt to match the selected
W R M B S against all 21 of M oore’s indicators,
so it is proposed to compare two of the W R M B S ,
total loans and business loans, against three
of these indicators: the Federal Reserve index
of industrial production, the Bureau of Labor
Statistics index of wholesale prices excepting
food and farm products, and the Dow Jones
index of industrial stock prices. These two loan
series have been chosen because, of all the
W R M B S , they show the strongest link with
business activity. The method of comparison will
be the same as that used in examining the timing
of the turning points of the selected W R M B S
relative to the reference cycle. In this instance,
however, the turning points of the total loans
and business loans series will be used as the point
of reference for the turning points of the selected
1Ibid.

FEDERAL RE S ER VE BANK OF SAN F R A N C I S C O

indicators. Table 6 presents a comparison of the
timing of the turning points of the three indica
tors in seven cycles, the two postwar cycles, and
six cycles, omitting the Great Depression.
T a b l e 6

C O M P A R IS O N O F TURN ING PO IN TS O F S E L E C T E D
EC O N O M IC IN DICATO RS WITH
T O T A L L O A N S S E R IE S
Federal tesorye Indei
of Industrial Production
Avenge
Hein
Months
Deviation
lead (— ) or Around
Lag ( + )
Arerage

Bureau ol Labor
Statistics index
ol Wholesale Prices
Average
Months
Lead (— ) or
Lag ( + )

Dow Jon« laden of
industrial Slock
Prices

Moan
Deviation
Around
Average

Average
Hue
Months
Deviation
Load (— ) or Around
Lag ( + )
Average

7 Cycles 1919-38, 1945-54
Total Loans
Peak
Total Loans
Trough

- 4 .2 1 1.8)

-14 .0(10 .41

-7.6 (1 1.8 1

-1 4 .8 (7 .7 )

8.5(11.51

6.413.7)

2 Cycles 1945-54
Total lo a n s
Peak
Total Loans
Trough

—3.0 ( 2.0)

-1 7 .0 (1 4 .0 )

8.012.0)

+ 3 .3 ( 4.4)

6.0(5.01

5.5110.5)

6 Cycles
Total Loans
Peak
Total Loans
Trough

1919-33, 1945-54

- 4 .2 1 2.2)

-1 6 .2 (1 0 .8 )

7.5(3.51

- 1 . 0 ( 6.0)

9.514.2)

6 .0 ( 9.0)

In only one case is there a tendency for a series
to lag behind the total loans of reporting banks.
This is for the index of industrial production,
which for the two postwar cycles shows an av
erage lag of some three months at the trough.
This lag is washed out when averaged in with
either six or seven cycles. All of the series turn
down from their peaks before total loans makes
its turn. The indexes of industrial production
and industrial stock prices lead quite consistent
ly at the peak with a relatively narrow range
of deviation. Wholesale prices lead by the great
est average figure, but the mean deviation around
the upper turns of this series in 6 cycles ranges
from 5 months to 27 months, centering about
an average lead of 16 months. The index of in
dustrial stock prices shows a consistent ten
dency to lead total loans of reporting banks at
both the peak and the trough, while the other
two series generally lead at the trough also; but,
as seen by their average deviation, they have on
occasion lagged at the trough. It might be noted
that rather than using the W R M B S as an indi
Digitized for
44FRASER

cator of the future movements of these selected
business series, these series might be used to
predict downturns in total loans. Note that
wholesale prices, industrial stock prices, and in
dustrial production turned down an average 16
months, 8 months, and 4 months, respectively, in
advance of the downturn in total loans.
It is to be expected that W R M B business
loans will show much the same pattern of tim
ing at the turning points relative to the indicators
to which they are being compared as did total
loans. Table 7 shows the turning points of these
series in the three groupings of cycles using busi
ness loans of reporting banks as a point of ref
erence.
The same general pattern of turns at the peaks
and troughs is present, even to the single in
stance of an average lag of industrial production
at the trough in the two postwar cycles. It is to
be noted that business loans show a somewhat
closer correspondence than did total loans. Look
ing at the indicators just tested, it might be
concluded that if the wholesale price index has
turned down, prices of industrial shares have
been falling, and industrial production is falter
ing, that total loans or, more particularly, busi
ness loans may well be expected to turn down.
TA B L E 7
C O M P A R IS O N O F TURN ING P O IN T S O F S E L E C T E D
EC O N O M IC IN D ICATO RS W ITH B U S IN E S S
L O A N S S E R IE S
Federal Reserve Indei
otf Industrial Production
Average
Months
load (— ) or
Lag (+ )

Kean
Deviation
Around
Average

fcuroao of Labor
Statistics Index
at Wholesale Prices

Dow Jones Indei ol
Industrial Sleek
Prices

Average
Mean
Months
Deviation
Lead (— ) or
Around
la g (+ )
Average

Average
Mean
Months
Dovlalioa
load (— ) or Around
Lag (+ )
Average

7 Cycles 1919-38, 1 945-54
Business Loans
Peak
- 4 . 4 ( 1 .9 )
Business Loans
Trough
- 7 .3 ( 1 1 .0 )

-1 4 .0 (1 0 .0 )

- 1 4 .3 (7 .9 )

7.7(12.0)

7.4 (2.7)

2 Cycles 1945-54
Business Loans
Peak
- 3 .0 ( 2.0)
Business Loans
Trough
+ 2 .7 ( 4.2)

-16 .51 13 .5)

7.5(1.51

6.5(4.51

-1 6 .0 (1 0 .5 )

7.2 (3.2)

9.5 (3.8)

5.0(11.0)

6 Cycles 1919-33, 1945-54

Business Loans
Peak
- 4 .2 ( 2.2)
Business Loans
Trough
—1.2( 5.51

5 J ( 9.0)

March 1958

MONTHLY R EV IEW

Securities loans and stock prices

The nature of these two series suggests that
there may be a close correspondence in the course
of their cyclical movements. Table 8 compares
the turning points of the index of industrial stock
prices with the turns of loans for carrying secur
ities in the seven-cycle, two-cycle, and six-cycle
comparisons made earlier.
T

able

TURN ING PO IN TS O F IN D USTRIAL S T O C K P R IC E S
C O M P A R E D W ITH LO A N S F O R CAR RYIN G
S E C U R IT IE S
Industrial Slock Prices
J Cycles 1919-M, 1945 54
Average
Mean
Months
Deviation
teed (— ) or Around
Lag (+)Avtrag*

lo a n s for
Carrying
Securities
Peak

lo a n s for
Carrying
Securities
Trough

-20.8113.4)

5.2 ( 5.8)

2 Cycles 1945 54
Average
Moults
load (— ) or
Lag (+ )

Mean
Deviation
Around
Average

4.0111.0)

-1 5 .0 (1 4 .0 )

t Cycles 1919-33, 1945 54
Average
Mean
Months
Deflation
Lead {— ) tr
Around
Lag (+ )
Average

5 .0 ( 7.0)

-1 2 .3 (1 1 .1 )

Industrial stock prices turn up from a trough
on the average of about a year before securities
loans hit bottom. But the mean deviation de
scribes a range from a one-month lead to a lead
of almost two years about this average. W ith a
variation of about two years, even the knowledge
that the turning points of the two series follow
this prescribed sequence is not sufficient ground
to accept this series as an indicator. Looking
back to Tables 6 and 7 it is seen that there is a
closer correspondence between both total loans
and business loans and industrial stock prices
than there is here. The reason, perhaps, for this
considerable disparity in the timing of the turns
of the two series is that loans for carrying se
curities accounts for loans to buy bonds as well
as stocks, and the motivations for buying bonds
are quite different from those for buying stocks
and will not necessarily respond only to factors
affecting trading in stocks.
Business loans and inventory stocks

Inventories play a very vital role in the pro
duction process. T o maintain continuous opera
tions, the manufacturer will try to accumulate

stocks of raw materials to offset irregularities in
delivery of these items, and he will also hold a
certain stock of finished goods to ensure that
he may meet the requests of retailers, wholesal
ers, or other manufacturers for his product.
Wholesalers, in their turn, will build stocks of
goods in order to be able to meet the demands of
retailers with a minimum of delay. Retailers will
want to stock quantities of a variety of goods to
accommodate their customers. Inventories held
at the various stages of production and distribu
tion allow sufficient flexibility to smooth out the
day-to-day fluctuations in operations that would
otherwise disrupt and hinder the flow of goods
from producers to consumers.
The stocks of inventories held by the business
community represent a substantial investm entsome $90 billion in December 1957. The net
changes in the level of inventories held is among
the most volatile elements contributing to changes
in business investment and, consequently, may be
looked upon as a key factor in fluctuations of
business activity. Loans to carry inventories
would seem to exemplify the “ correct” loan to be
made by banks, discussed in the introduction. An
examination of the relationship of business loans
to inventories would seem to be a proper sub
ject for this study. Unfortunately, compara
ble data are not available for the entire period
1919 to 1954, and the investigation must proceed
on a piecemeal basis. The Department of Com
merce series dates back only to 1939 and can only
be compared to loans in the postwar cycles. The
National Industrial Conference Board supplies
an index of the value of manufacturers’ inven
tories for the period 1929 to 1940, which ex
cludes industries tied to agriculture or the ex
tractive industries. This series is not comparable
with the later Commerce series. T o provide some
common ground for analysis, inventories of dur
able goods are used here for both series. Table 9
gives the timing of the turning points of these
series relative to business loans.
The first set of turning points, which covered
the period of the Great Depression, appears to be
at odds with the postwar series in which the in
ventory series showed a lag at all four turning
points. Chart 5 shows how closely the inventories

FEDERAL R ESE RVE BANK OF SAN F R A N C I S C O

able

TURN IN G PO IN TS O F D U R A B L E G O O D S
IN VEN TO R IES C O M P A R E D W ITH B U S IN E S S LO A N S
Durable Goodj Invenlories

Iverag*
Months
Lead(— ) or
Ug (+)

Mean
Deviation
Around
Average
l ’/ i C y c le s 1 9 2 9 - 3 9

B u s in e s s L o a n s
Peaks
B u s in e s s L o a n s
T ro u g h s

( 0)

-9

(20)
2 C y c le s 1 9 4 5 - 5 4

B u s in e s s L o a n s
Peaks

+ 4 .5

(1 .5)

B u s in e s s L o a n s
T ro u g h s

(

Conclusion
Summing up, movements in the W R M B S are
a poor tool to use alone as a basis for calling the
turn in business activity. This is not said to dis
credit the series, but rather to call attention to
the fact that these series are not published with
this end in mind. It was established in the first
part of this study that the W R M B S is an accept
able indicator of changes in the disposition of the
resources of the banking system. It would impose
an unfair burden upon the series to ask of it a
task it is not equipped to perform. That the serv
ices of the banking system are indispensable to
the general scheme of production cannot be de
nied, but it is necessary to point out that all of
the sectors of the economy do not develop at the
same pace nor do they necessarily prosper at
the same time. The banking series here examined
tend to follow rather than lead the changes in
the majority of business series.
It is, nevertheless, a useful endeavor to high
light the behavior of the banking system in the
business cycle. One point of particular interest

series, together with “ Purchased Materials,”
“ Goods in Process,” and “ Finished Goods” cor
respond with business loans. The Commerce
series on inventories shows a relationship to
business loans which offers promise. Admitted
ly, the data for only two cycles are not enough to
establish a significant cyclical relationship be
tween the two series. There is, however, other
evidence that would lend support to this conclu
sion. In his examination of manufacturers’ inven
tories at business cycle turns in the period 191938, Abramovitz found a
C hart 5
tendency for inventories
B U S I N E S S LOAN S AND S E L E C T E D IN V E N T O R Y S E R I E S
to lag behind general ac
tivity by a period of three
B ILLIO N S OF C O LL A R S
to six m on th s.1 In his
study as in ours the Na
tional Bureau reference
TOTAL INVENTORIES
cycle dates are used as a
measure of business ac
tiv ity . S in ce busin ess
loans similarly lag behind
reference peaks by an av
erage of 4.8 months, this
would imply that inven
tories m ay turn dow n
either with business loans
or, as post-W orld W ar II
experience indicates (see
0 0 D S IN
Chart 5 ), slightly after
business loans have turned
down.
1 Moses Abramovitz, Inventories and
business Cycles (New York: Na
tional Bureau of Economic Re
search, 1950), pp. 80-87. See also
Doris M. Eisemann, Bank Credit
& Inventory Cycles (Santa Mon
ica, California: The Rand Corpo
ration, 1957), 12 pp.

Digitized for
46FRASER

Sources: Board of Governors of the Federal Reserve System, Federal Reserve Bulletin; Department
of Commerce, Survey oj Current Business.

March 1958

MONTHLY REV IEW

is the consistent tendency of loans outstanding
to turn down some time after business activity
in general has turned down. There is not a direct
one-to-one correspondence between production
activity and the demand for credit. Businessmen
generally demand credit upon their expectations
of the future, based in large measure on present
activity. If production is expanding and there is
a strong demand for goods, loans outstanding
will also expand. The fact that loans outstanding
continue to climb while production has started
to decline might be attributed to either of two
causes: an increase in new loans or a decrease
in repayments. It is most likely due to the latter,
for it has been found that there is a sharp rise
in the liabilities of business failures late in the
boom. And as sales fall off, collections become
slower, and requests for extensions on notes be
come more numerous. In general, the expected
receipts which served as the basis for making the
loans and out of which the loans were to be re
tired, have failed to materialize. The turning of
loans outstanding at the trough relative to the
reference cycle is not consistent enough to per
mit analysis of its behavior.
The movements of bank holdings of United
States securities are conditioned in large part
by movements in general business activity and
the resultant effects on loans outstanding. The
movements of this series are symptomatic of busi
ness contraction and expansion rather than
causal.
The relationship of the selected economic in
dicators compared to business loans as exam
ined in Tables 6 and 7 supports the earlier evi

dence of a consistent tendency of loans outstand
ing at reporting banks to lag at the peak of the
cycle and a less consistent tendency to lag at the
trough of the cycle.
It is also noted that the W R M B S is not suffi
ciently selective to pinpoint changes in specific
sectors outside the sphere of banking activity. It
is possible, however, that the business loan series,
in view of its correspondence to business inven
tories, may prove to be of value in detecting
fluctuations in inventory stocks. This relation
ship deserves further investigation.
Does this seeming inability of the W R M B S
to foresee the turns in the tide of economic ac
tivity rule out its usefulness in assessing the busi
ness situation? Not at all. Insofar as it is able to
indicate turns in the activity of the banking sec
tor, it is an adequate tool of analysis. The modern
industrial economy is such a complex organism
that there is no one single indicator that can de
scribe it accurately. The health and vigor of the
economy depends upon the functioning of its
component parts. No single indicator can give a
sufficiently accurate picture of the over-all con
dition of the economy without neglecting often
times important developments in certain areas of
activity. Any true estimate of the situation must
draw upon several sources of information, one of
which is the banking statistics, to complete the
picture.
The techniques used in foretelling future
events in the affairs of men have come a far
piece since the Etruscan oracles predicted the fu
ture by examining the entrails of the sacred
geese, but even now as then there is no single easy
answer to complex and compelling questions.

Review of Business Conditions
activity nationally and in this District
during the opening months of 1958 continued
to recede from the advanced level reached in the
late summer of 1957. January output of the na
tion’s mines, mills, and factories declined over 2
percent from December and was 9 percent below
a year ago. A further drop of 2 percent occurred
in February. A s in other recent months, the out
put declines in early 1958 were mostly among
hard goods producers— manufacturers of both
producers’ equipment and consumer durables.
Employment of wage and salary workers in dur
able goods industries fell over 11 percent in the
twelve months ending in February, compared
with a drop of about 3 percent in nondurables.
Ingot steel output continued to fall in January
and February; in fact, during February it was
below its low point in the previous business
downturn in mid-1954. Production of other pri
mary and fabricated metals and most types of
machinery also fell in January and February.
The number of automobiles assembled in the first
two months of the year was approximately 27
percent under the year-ago level, while sales of
automobiles lagged even more. Although sales of
nondurable good s m anufacturers were un
changed between December and January after
seasonal adjustment, sales of durable goods man
ufacturers recorded another sharp decline. R e
tail sales fell moderately in January, and the
preliminary estimate for February indicates an
other drop, mostly in durable goods. Inventory
holdings by manufacturers were reduced further
in January after seasonal adjustment, although
stocks at retail outlets showed no net change
from December. However, automobile dealers
experienced steady increases in their inventories
early in 1958, which reached a total of almost
900,000 cars at the end of February. Construc
tion outlays during January and February de
clined somewhat more than the seasonal amount.
Total employment across the nation fell sharp
ly between December and January. Although the
decline in February was more moderate, the
number of persons holding jobs in mid-February
was below the level of a year ago by about 1,200,000. In addition, the number of full-time non
farm job holders working less than their accus
tomed work schedules in February was over
u s in e s s

Digitized for
48FRASER

2 million— more than twice the average of the
past few years. Unemployment increased to 6.7
percent of the civilian work force in Febru
ary after adjustment for seasonal factors; this
was the highest rate of unemployment since the
late months of 1949. The wholesale and consum
er price indexes continued to show increases in
January despite the slide in business activity;
rising prices of farm products and food have
more than offset scattered declines in the prices
of raw materials and manufactured goods. The
rather severe winter weather in many Eastern
and Southern areas during February, which
ruined crops and generally disrupted business ac
tivity, should also be reflected in higher food
prices in the February indexes.
The experience of several other sectors of the
national economy was more encouraging. New
defense contracts awarded, after being a depres
sive force for several months, increased in N o
vember and December, and they are expected to
expand rapidly during the first half of 1958. It
must be recognized, however, that there will be
an appreciable time lag before many of these
commitments are translated into actual money
expenditures. New machine tool orders picked
up in January after an extremely poor December,
and some tool makers reported an even better
inflow of orders in February. New private hous
ing starts in January, at a seasonally adjusted
annual rate of 1,030,000 dwelling units, ex
ceeded all months of 1957 except August in an
nual terms. In February, new housing starts
slumped to an annual rate of 890,000 units pri
marily because of the winter storms.
Excess reserves have exceeded member bank
borrowings at Federal Reserve Banks since the
first of the year, and the reduction in member
bank reserve requirements by 0.5 percent in late
February freed reserves amounting to about
$500 million. Short-term interest rates have de
clined further in response to a lessened demand
for credit and moderate easing in the supply of
bank reserves by the Federal Reserve System.
The gradual fall, which longer term rates had ex
perienced since October and November, changed
to fluctuation within a narrow range between late
January and early March. This was largely a
reflection of renewed Treasury borrowing at

March 1958

MONTHLY R EV IE W

extended maturities and increased corporate and
municipal security offerings.
District business decline m oderated
in early 7 958

In the Twelfth District, there are indications
that the decline in over-all business activity dur
ing early 1958 was moderate when compared
with the experience of the previous six months.
The decline so far this year also appears to
have been less than that which occurred nation
ally. Before these recent developments are dis
cussed, however, it may be beneficial to review
briefly several ways in which the Twelfth Dis
trict economy differs from the national econ
omy.1 Because of the difference in industrial
composition between the Twelfth District and
the nation as a whole, forces affecting the level
of business activity nationally are not transferred
to the economy of this region with an exactly
corresponding impact. Depending on the indus
tries involved, changes in business activity na
tionally may induce greater or lesser changes in
the level of District activity. These structural
differences between the Twelfth District and the
entire nation may be illustrated by an examina
tion of the sources of personal income.
Income payments from government and farm
sources in 1956, the most recent year for which
comparable data are available, were a larger pro
portion of total income payments in this District
than they were nationally— 24 percent versus 21
percent, respectively. Within the private non
farm sector in 1956, commodity producing in
dustries— manufacturing, mining, and construc
tion— comprised a smaller part of total income
payments in the Twelfth District (36 percent)
than they did nationally (42 percent). Manufac
turing alone contributed 26 percent of private
nonfarm income in this District in contrast with
33 percent nationally. Commodity producing in
dustries, and manufacturing in particular, are
ordinarily more subject to cyclical swings in eco
nomic activity than are distributive and service
industries.
These figures indicate that there may be some
“ built-in’' resistance to sharp fluctuations in pri
vate business activity in the Twelfth District
•See also this Review, December 1957, pp. 154-159.

compared with the nation as a whole. On the
other hand, this District is highly vulnerable to
changes in Government spending policies, par
ticularly with respect to its important aircraft
and related industries. M ajor aircraft fabrica
tors and firms subcontracting the production of
components and equipment expanded rapidly
from late 1956 to mid-1957 in response to or
ders for new weapons systems. However, revised
Government procurement policies in mid-1957
resulted in the cancellation of a substantial vol
ume of orders and the stretching out of delivery
schedules in many continued contracts. The
sharp decline in District aircraft activity which
followed more than offset the gains which had
occurred since late in 1956.
The leveling off and succeeding downturn in
business spending for plant and equipment and a
continuation of the general weakness in con
sumer spending for durable goods have affected
other parts of the nation somewhat more than
they have this District in the past six to eight
months. Downward adjustments in defense
spending have had a greater impact, though, on
the Twelfth District than on the nation. Indeed,
during the late summer and early fall of last year,
manufacturing activity in this District was fall
ing off more rapidly than in the nation because
of the initial sharp cuts in defense spending.
More recent developments have not been quite
so unfavorable; cutbacks continue in a number
of industries in the Twelfth District, but the
over-all impact appears to have moderated.
District nonfarm employment shows only
seasonal decline in January

Total nonfarm employment in the Twelfth
District declined only the usual seasonal amount
between December and January. More than seas
onal cutbacks occurred in manufacturing and
mining employment, but these were offset by
losses smaller than usual for the season in con
tract construction and wholesale and retail trade
employment. The drop in manufacturing em
ployment between December and January was
the smallest since mid-1957, affecting for the
most part firms in primary and fabricated metals,
machinery, and transportation equipment. A l
most half of the decline in transportation equip

FEDERAL RE S ER VE BANK OF SAN F R A N C I S C O

ment occurred in automobile assembly plants.
Aircraft industry payrolls on the Pacific Coast
were pared by little more than 2,000 workers,
compared with average monthly declines of about
9,000 workers between July and December. This
does not signal an end to cutbacks among aircraft
firms, however. W hile some rehiring has already
begun at Seattle’s aircraft plants, Southern Cal
ifornia aircraft firms indicate that further lay
offs should occur over the next few months.
Present indications from labor market reports,
though, are that the average reductions over the
next two to four months will be less than half
those in late 1957.
Pacific C oasi unemployment rate declines

Total employment and unemployment in the
three Pacific Coast states changed according to
their usual seasonal patterns between December
and January. After seasonal adjustment, unem
ployment was down slightly, which lowered the
rate of unemployment to about 5.3 percent of the
labor force. Although the method by which em
ployment and unemployment are estimated is
less precise than the method used for national es
timates, it appears probable that there was no
counterpart on the Pacific Coast to the much
sharper than seasonal rise in unemployment na
tionally in January. The rate of unemployment
in Pacific Coast states in December and January
was, however, about as high as during the peak
months of August and September 1954 in the
previous business downturn.
The number of persons drawing weekly bene
fit payments under state and federal unemploy
ment insurance plans in Twelfth District states
during January remained at about the level of
the previous two months after adjustment for
seasonal factors. Joblessness covered by these un
employment insurance plans averaged almost 80
percent above the level of a year ago and was
higher than in any period since early 1950.
Man-hours continue dow nw ard movement

Broad indications of current production trends
in the three Pacific Coast states are provided by
data on man-hours of manufacturing production
workers. Further nonseasonal declines in manhours were reported by most durable goods man
Digitized for50
FRASER

ufacturing firms during January. The drop of 2
percent between December and January reported
by firms making durables was about equal to the
average monthly declines during the second half
of 1957. While aircraft firms reduced their man
power usage somewhat less than in previous
months, there were continued cutbacks among
firms producing lumber, metals, autos, and ma
chinery. Nondurable goods industries showed a
small net increase in man-hours between Decem
ber and January after seasonal adjustment.
Production o f steel declin es; lumber
improves slightly

Steel producers in the Western Region (in
cluding Colorado) reduced output almost 10 per
cent between December and January. Western
utilization, which was 65.6 percent of the en
larged January 1, 1958 capacity, was 9 points
above national utilization. Since late January,
further sizable cuts have been made in output at
Twelfth District steel mills, and the rate of out
put may have fallen closer to the average rate of
mills in the rest of the nation.
The lumber industry in January and February
showed little change from early 1957. The extent
to which recovery may come for the industry
cannot be determined until spring weather al
lows a real test of the market for residential
housing across the nation. Redwood production
and orders in January were up 3 and 15 percent,
respectively, from a year ago. Western pine pro
duction and orders for the first seven weeks of
this year were running slightly above the yearago levels also. Douglas fir production, and to a
small extent orders, were down for the same pe
riod.
Douglas fir plywood production through midFebruary remained at a record level, but the in
flow of new orders in this period was substan
tially below production. Consequently, order
backlogs declined somewhat from early January.
Between mid-January and late February four
successive price cuts were made by a number of
manufacturers, dropping the price of the index
grade from $72 to $64 per thousand square feet.
The latest price cut is reported to have induced
a number of mills to cut production schedules

March 1958

MONTHLY R EV IEW

sharply or to shut down completely. Almost im
mediately, however, new orders picked up brisk
ly, and some mills have begun to quote higher
prices.

down 15 percent from December and 45 percent
from a year ago.

Building activity m ixed

Weekly department store sales in the Twelfth
District were down about 6 percent from a year
ago in the four-week period ending February 22.
Reversing the pattern of recent months, depart
ment store sales in the nation for the same period
declined more than in this District. The national
drop of 9 percent was in large part a reflection
of poor weather in many Middle Western and
Eastern areas in mid-February. Both this Dis
trict and the nation as a whole show a 4 percent
decline this year compared with January and
February in 1957.

Construction activity in the Twelfth District
during January, as in late 1957, reflected mod
erate improvement in residential housing and
declines in other types of construction. Both
building permit awards and construction con
tract awards for residential housing showed seas
onal gains from December and slight gains from
January a year ago. The number of dwelling
units covered by Government-insured mortgages
increased between December and January, re
sponding to seasonal influences but also indi
cating renewed interest among lenders in these
programs. Applications for V A appraisals rose
over 75 percent from December’s very low level,
although they were still 36 percent below Janu
ary last year. Requests for F H A commitments,
however, were not only up 14 percent from De
cember but were also 31 percent higher than a
year ago. Permits and contract awards for nonresidential construction experienced some seas
onal rise from December but failed by a small
amount to match the year-ago level. Contract
awards in the public works and utilities sector,
which held up well during most of 1957, were

District departm ent store sales decline,
but less than nationally

Total retail sales in the Twelfth District during January may not have fallen as much as those
at department stores, since retail trade employ
ment declined less than the usual seasonal
amount from December. Automobile sales, both
in this District and across the nation, appear to
have been quite disappointing, however. In Cal
ifornia alone, sales were almost 19 percent below
January 1957, while the drop nationally was
about 22 percent. State sales figures for Febru
ary are not yet available, but automobile sales
nationally dropped an estimated 32 percent from
a year ago.

Reprints of the series of three articles on the aluminum industry
recently published in this Review are now available for distribution.
Requests for copies should be directed to the Federal Reserve Bank
of San Francisco, 400 Sansome Street, San Francisco 20, California.

FE DERAL R ESE RVE BA N K OF S AN F R A N C I S C O
BUSINESS INDEXES — TWELFTH DISTRICT*
(1 9 4 7 - 4 9 av erag e =

100)

Total
nonagrlcultural
employ
ment

Industrial production (physical volume)1
Year
and
month

Lumber

1929
1933
1939
1949
1950
1951
1952
1953
1954
1955
1956
1957

95
40
71
100
113
113
110
118
116
124
116
106

Petroleum*
Refined
Crude

Lead1

Cement

87
52
67
99
98
106
107
109
106
106
105
101

78
50
63
103
103
112
116
122
119
122
129
132

64
27
56
100
112
128
124
130
132
145
156
149

165
72
93
101
109
89
87
77
71
75
79
77

1957
January
February
March
April
May
June
July
August
September
October
N ovember
December

102
102
101
101
101
101
101
101
102
101
101
101

131
130
132
132
138
131
133
137
135
132
131
124

120
127
140
154
157
152
162
160
169
161
146
139

78r
87 r
88
82r
83r
78r
69r
75 r
75r
76 r
63
62

1958
January

100

122

135

Copper*

Electric
power

105
17
80
93
113
116
113
111
101
118
129
126

29
26
40
108
119
136
144
161
172
192
210
224

125
137
133
135
126
130
113
llo r
127
126
125
125r

220
211
221
228
229
239
238
233
217
223
221
211

•

Total
mf’g
employ
ment

Car
loadings
(num
ber)*
Exports

Dep’t
store
sales
(value)1
Imports

Waterborne
foreign
trade*- *

Retail
food
prices

It 4

30
18
31
98
107
112
120
122
122
132
141
141

64
42
47
100
100
113
115
113
113
112
114
118

190
110
163
85
91
186
171
140
131
164
195

124
72
95
121
137
157
200
308
260
308
443

237
269
267
298
283
252
188
210
173
199
210

421
417
489
534
698
511
770
572
607
684
582
-----

•••
99
103
112
118
121
120
127
134
138

55
97
105
120
130
137
134
143
152
157

102
52
77
94
98
100
100
100
96
104
104
96

138
138
138
138
138
139
138
138
138
138
137
137

157r
158r
158r
158r
158r
159r
159
158r
156 r
155r
152
151r

105
96
100
103
99
100
94
97
93
84
95
93

137
141
146
137
141
148
141
144
141
134
139
139

116
117
116
117
117
118
118
119
119
119
118
119

137

150

132

121

BANKING AND CREDIT STATISTICS — TWELFTH DISTRICT
(a m o u n t * In m illio n * o f d o lla r * )

Condition Items of all member banks'
Year
and
month

Loans
and
discounts

U.S.
Gov’t
securities

Demand
deposits
adjusted’

Total
time
deposits

2,239
1,486
1,967
7,093
7,866
8,839
9,220
9,418
11,124
12,613
13,178

495
720
1,450
6,415
6,463
6,619
6,639
7,942
7,239
6,452
6,619

1,234
951
1,983
9,254
9,937
10,520
10,515
11,196
11,864
12,169
11,870

1,790
1,609
2,267
6,302
6,777
7,502
7,997
8,699
9,120
9.424
10,679

1957
February
March
April
May
June
July
August
September
October
November
December

12,556
12,576
12,649
12,694
12,911
12,912
12,945
13,178
13,064
13,185
13,178

6,356
6,177
6,520
6,315
6,249
6,319
6,313
6,293
6,433
6,357
6,619

11,279
11,129
11,622
11,210
11,310
11,407
11,329
11,561
11,570
11,770
11,870

9,690
9,794
9,839
9,995
10,155
10,188
10,220
10,301
10,417
10,304
10,679

1958
January
February

13.106
13,002

6.573
6,884

11,601
11,305

10.761
10,992

1929
1933
1939
1950
1951
1952
1953
1954
1955
1956
1957

Member bank reserves and related items
Bank
rates on
short-term
business
loans’

Factors affecting reserves;
Reserve
bank
credit*

4.74
4.81
...........................

5.21
..........................

5.13
, . • •

Treasury"

0
- 110
- 192
-1,14 1
-1,58 2
-1 ,9 1 2
-3,07 3
-2 ,4 4 8
-2 ,6 8 6
-3 ,2 5 9
-4 ,1 6 4

+
+
+
+1
+1
+2
+3
+2
+2
+3
+3

23
150
245
198
983
265
158
328
757
274
903

41
37
—
35
56
+
29
—
49
50
+
109
76
+
14
+
— 18

816
170
445
261
374
426
145
434
322
298
454

+
+
+
+
+
+
+
+
+
+
+

—

258
427

+
+

—

3.35
3.66
3.95
4.14
4.09
4.10
4.50
4.97

34
2
2
39
21
7
14
2
38
52
31

Commer
cial18

+
+
+
+
+
+
+

16
12

Money in
circu
lation*

_
-

Reserves11

Bank
debits
Index
31 cities*'1*
(1947-49 =
100)*

6
18
31
+
14
189
+ 132
39
+
30
+ 100
96
— 83

175
185
584
2,026
2,269
2,514
2,551
2,505
2,530
2,654
2,686

42
18
30
115
132
140
150
154
172
189
203

494
170
430
209
402
320
292
480
159
447
480

139
9
— 31
54
+
20
+
6
+
39
+
30
—
8
37
+
23

2,517
2,495
2,560
2,526
2,483
2,457
2,592
2,581
2,517
2,652
2,686

200
200
202
200
203
205
197
204
200
202
217

180
298

—

137
17

2,662
2,520

211
203

—
—

1 Adjusted for seasonal variation, except where indicated. Except for department store statistics, all indexes are based upon data from outside sources, as
follows: lumber, California Redwood Association and U.S. Bureau of the Census; petroleum, cement, copper, and lead, U.S. Bureau of Mines; electrio
power, Federal Power Commission; nonagricultural and manufacturing employment, U.S. Bureau of Labor Statistics and cooperating state agencies;
retail food prices, U.S. Bureau of Labor Statistics; carloadings, various railroads and railroad associations; and foreign trade, U.S. Bureau of the CensuB.
* Daily average.
* N ot adjusted for seasonal variation.
* Los Angeles, San Francisco, and Seattle indexes combined.
6 Commercial
cargo only, in physical volume, for Los Angeles, San Francisco, San Diego, Oregon, and Washington customs districts; starting with July 1950, “ spe
cial category” exports are excluded because of security reasons.
• Annual figures are as of end of year, monthly figures as of last Wednesday
in month.
d e m a n d deposits, excluding interbank and U.S. G ov't deposits, less cash items in process of collection. Monthly data partly esti
mated.
• Average rates on loans made in five major cities.
' Changes from end of previous month or year.
10 Minus sign
indicates flow of funds out of the District in the case of commercial operations, and excess of receipts over disbursements in the case of Treasury
operations.
11 End of year and end of month figures.
u Debits to total deposits except interbank prior to 1942. Debits to demand
deposits except U.S. Government and interbank deposits from 1942.
f>— Preliminary.
r— Revised.

Digitized for52
FRASER