Full text of Introduction to Flow of Funds, 1980

View original document

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

Introduction to Flow of Funds

Board of Governors of the Federal Reserve System

Introduction to Flow of Funds

Board of Governors of the Federal Reserve System

June 1980

Copies may be obtained from Publications Services, Board of
Governors of the Federal Reserve System, Washington, D.C.
20551, and remittance should be made payable to the order of
the Board of Governors of the Federal Reserve System in a form
collectable at par in U.S. currency. (Stamps and coupons are not
accepted.) Per copy price, $1.50; 10 copies or more to one ad
dress, $1.25 each.

Library of Congress Catalog Number: 80-600-127

Contents

Preface

Section 1: Concept of Accounts

Section 2: Organization of Accounts

Section 3: Relation to National Income and Product Accounts

Section 4: Definition of Sectors and Transaction Categories

Section 5: Publications of Flow of Funds Data

Section 6: Data Sources

Bibliography

Text Tables
1
2
3
4
5
6
7
8

Summary of flow of funds accounts for the year 1978
Financial assets and liabilities, December 31, 1978
Private claims on U.S. government institutions
Gross saving and investment, comparison of two systems, 1972
Saving and physical investment in flow of funds accounts, 1972
Sector structure
Financial transaction categories
Sources of financial data for flow of funds accounts

3
7
21
27
30
32
39
50

Appendix Tables

Summary of flow of funds accounts for the year 1978
Financial assets and liabilities, December 31, 1978
Funds raised in credit markets
Credit market supply of funds
Credit market debt outstanding
Claims against credit market debt
Household sector saving and investment
Household sector assets and liabilities
Nonfinancial business sector saving and investment and assets and liabilities

S. 1
S.2
S.4
S.5
S.6
S.7
S.8
S.9
S.10

Preface
Flow of funds accounting is a national statistical
system that has been under development by the
Federal Reserve since 1947 and that has been pub
lished in several forms. Descriptions and definitions
of the successive forms of accounts have appeared
in the Federal Reserve Bulletin and in separate pub
lications.1 In the following pages that descriptive
material is pulled together and brought up to date
with the current form of the accounts.
Section 1 of this publication describes broadly
the purposes of the flow of funds system and the
aggregative view of recent decades that emerges
from the data in their present form. Section 2 gives
the conceptual framework of the system, while sec
tions 3 and 4 present a detailed description of statis
tical relationships and definitions. The structure of
the system described in section 4 is of the form
published in late 1979 and differs in some respects
from earlier presentations. The structure will inevi
tably continue to change to keep up with the evolu
tion of financial practices, and after 1980 some of
the descriptions may no longer fit the current form
of presentation.
1. For earlier publications, see the bibliography.

This is not a statistical publication, although a
few tables in the appendix underlie the historical
material in section 1. Nor does it describe specific
statistical sources and derivation procedures used in
compiling the accounts. The data are published and
made available in several forms as described in
section 5. A separate publication on the derivation
methods is mentioned in section 6, which discusses
briefly the problems of adapting data sources to the
account structure.2
This publication is at a general level and is in
tended merely to indicate the purposes of the sys
tem and what it encompasses. The bibliography in
cludes several review papers that go into further
detail on uses of financial accounts for judgmental
and econometric analysis and that include their own
extensive bibliographies on macroeconomic work in
financial analysis. This publication is an introduc
tion to that work as well as to the flow of funds ac
counts themselves.

2. All requests for data and for information about the accounts
should be addressed to Flow of Funds Section, Division of
Research and Statistics, Board of Governors of the Federal Re
serve System, Washington, D.C. 20551.

Section 1: Concept of Accounts
The flow of funds system of national accounts is
designed to bring the many financial activities of
the U.S. economy into explicit statistical relation
ship with one another and into direct relation to
data on the nonfinancial activities that generate in
come and production. The purpose of the accounts
is to provide, systematically, the aggregate mea
sures of transactions needed to identify both the in
fluences of the nonfinancial economy on financial
markets and the reciprocal influences of develop
ments in financial markets on demand for goods
and services, sources and amounts of saving and in
vestment, and the structure of income. The ac
counts are intended to provide an empirical base for
exploring such questions as the sensitivity of bor
rowing to interest rates as against other influences,
the effects of the cost and supply of credit on de
mand for physical investment, the role of money
holdings in the public’s structure of assets and lia
bilities, and the relation of financial positions—
levels of assets and liabilities—to demands for
goods and services, for credit, and for investment
in financial claims.
The flow of funds system focuses on such ques
tions in a macroeconomic setting that covers, as far
as possible, all of the institutional groups and all
types of transactions in the economy. Some ele
ments of the system exist elsewhere and separate
from the structure—as statements of, for example,
corporate finance, government finance, balance of
payments, money and banking activities, individ
uals’ saving, residential finance, and security mar
ket activity. These elements are incorporated into
the system as integral parts, together with informa
tion from income and product accounts on saving
and capital formation. Each such element is one as
pect of an integrated economy, and each connects
with the others in several ways. The security mar
kets, for example, are a point of intersection among
business developments generating long-term credit
needs, international capital movements, the avail
ability of bank credit, flow of credit through finan
cial institutions, and the financing of government
deficits. When all of the relevant information is put
together in mutually consistent fashion in one
framework, each element is seen as part of a
broader macroeconomic system, and its connection
with the other parts can be made explicit enough
for analysis. Like the national income and product
accounts, published by the U.S. Department of

Commerce, the flow of funds system is a social ac
counting structure that records both the payment
and the receipt aspects of any transactions included
in the system and that includes a balance in each
account of the structure between total payments and
total receipts. The flow of funds accounts can, in
fact, be viewed as a direct extension of the Com
merce income and product structure into the finan
cial markets of the economy, with the purpose of
establishing direct linkages between the Commerce
data on saving and investment—the capital account
in the income and product structure—and the lend
ing and borrowing activities that are associated with
saving and investment.

Model account structure
The nature of those linkages and, more generally,
the relation of financial markets to the nonfinancial
activities of the economy are portrayed in table 1.
The table is a severely condensed and simplified
form of the flow of funds matrix found on page
S.l, which maps the basic structure of the flow of
funds system.3 The arithmetic of the matrix is fairly
simple, and when applied to the full system of ac
counts it lays a basis for understanding both the ac
counting relationships among the time-series tables
in this publication and the analytic approach under
lying the system.
The accounting rules of the matrix, their conse
quences for analysis, and the relation of the matrix
to the tables are spelled out in section 2. A very
simplified form of the matrix structure appears in
table 1, which is a statement of the capital account
for the economy as a whole showing investment in
assets in the use columns and the means of financ
ing that investment in the source columns. The
table divides the economy into several groups of
sectors, each of which has a column in the matrix;
and all transactions of each group are recorded on
one or another row of the matrix. Row 1—saving—
is for each sector the net sum of current receipts
from income less current outlays for consumption,
operating expenses, and so forth. Saving appears as
a net amount available from current operations for
investment purposes. Other amounts are borrowed
3. The note to table 1 is important for working out the rela
tion between the general description in this section and the more
complex form that financial activity takes in the actual economy.

Section 1: Concept

row 1 for the sector, and (2) that on any one row of
the matrix the sum of all uses of funds shown
across the columns is equal to the sum of all
sources of funds in that row. (Rows 1 and 2 of
table 1 are the source and use sides respectively for
a single account covering all nonfinancial transac
tions together.) With balance vertically between
saving and investment and horizontally between
payments and receipts, each column constitutes one
full sector account of the structure and each row a
transaction account. The relationships among
columns, among rows, and between columns and
rows express the interlocking nature of the account
ing system as a whole.
As one illustration of the structure, the govern
ment column in table 1 can be seen to be a particu
lar form of budget statement, with a nonfinancial
deficit in the first row offset by a net sum of

in financial markets by each sector, and borrowings
and saving together are the sources of funds used to
acquire physical and financial assets.
A distinction is drawn between nonfinancial
transactions reflected in rows 1 and 2 of table 1—
purchases and sales of goods and services, trans
fer payments and receipts, and taxes—and financial
transactions in rows 3 through 6—net changes in
the capital amounts of claims owed as liabilities or
held as assets by each sector. All of the financial
transactions of a sector are combined into a net fi
nancial investment that is the excess of the sector’s
lending (financial uses) over its borrowing (finan
cial sources).
The two basic constraints in the matrix are
(1) that for each sector total investment, which is
the sum of capital outlays plus net financial invest
ment, is by definition equal to the saving shown in
Table 1. Model flow of funds matrix1
Hypothetical data

Sector

Type of transaction

Private
domestic
nonfinancial
Use

Nonfinancial
1. Saving............................................
2. Capital outlays..............................
Financial
3. Net financial
investment..............................
4. Total financial uses
and sources (5 T 6)................

5. Deposits at financial
intermediaries........................
6. Loans and securities......................

Source

Govern
ment

Use

Source

Financial
inter
mediaries
Use

Use

- 10

179

Source

Memo:

Rest of
world

Total

Source

- 10

Use

-4

170

-4
65

Domestic
totals

Source

Use

170

Source

174

170

147

144

140

147

1. This table compresses twenty-six sectors in the full system into four columns for sector types that are to be distinguished in the present discussion; the rows are a similar grouping
of transaction categories. In addition, the matrix is simplified by omitting the row and the column for discrepancies and a number of items peripheral to the mainstream of financial
transactions. These omitted items are treated as nonexistent in the simple economy shown. The relation of transactions in this table to the full matrix on page S. 1 is as follows;
FULL MATRIX

MODEL

Gross saving
Gross investment
Private capital expenditures, net
Net financial investment
Financial uses, net, and financial sources
Gold, foreign exchange, Treasury currency
Deposits and currency
Insurance and pension reserves, interbank items
Credit market instruments
Security credit, trade credit, taxes payable, equities,
miscellaneous, and sector discrepancies

Saving
Omitted
Capital outlays
Net financial investment
Total financial uses and sources
Omitted
Deposits at financial intermediaries
Omitted
Loans and securities
Omitted

The government sector should be interpreted as central government only, with state and local governments omitted as another simplification. Of the omissions, the most important for
the discussion that follows is insurance and pension reserves, which are a major form of intermediation. This item is left out because part of such reserves are liabilities of governments
and complicate the relation between intermediation and financial institutions. The present section focuses only on the broad outlines of structural relationships, and a more detailed de
scription requires many qualifications and additions to the broad form to incorporate these governmental reserves and the other omitted items.

Introduction to Flow of Funds

changes in cash balances, loans, and debt outstand
ing in the rows below. This column is a balanced
account that differs from other budget statements
for governments mainly in that it distinguishes
sharply between nonfinancial and financial transac
tions and arranges transactions so that they can be
identified across the rows in the accounts of other
parties to the transactions. The rest-of-the-world
column is similarly a particular form of balance of
payments statement, arranged so as to connect it
with other sector columns along specific rows. A
loan from the government to a foreign borrower,
for example, appears on row 6 as a government use
of funds and a foreign source of funds, regardless
of how it may be treated in other budget statements
or balance of payments statements. Each of the two
loan entries—the use and the source—is then play
ing a double role in the matrix: vertically as a com
ponent of its sector’s column balance, and horizon
tally as part of the row balance. Horizontally they
of course balance each other, but vertically they in
tegrate in more complex ways with the other trans
actions of the two sectors separately. The interlock
of the system consists of establishing such double
roles, horizontally and vertically, for all transac
tions of all sectors in the system simultaneously.
The result is an integrated structure that can be used
to measure linkages either vertically or horizontally
or, in the most complete forms of analysis, in both
directions simultaneously.
The condition that saving equal investment for
each sector is identical in form with the well-known
equality of saving and investment for the overall
economy in income and product accounting. For the
total economy, investment on a consolidated basis
consists of outlays for capital goods plus net foreign
investment, the excess of lending abroad over bor
rowing from abroad. In the flow of funds accounts,
similarly, each sector’s investment consists of its
purchases of capital goods plus a net financial in
vestment that includes net lending to the rest of the
domestic economy as well as abroad. In the model
matrix (table 1), the first three rows state the equal
ity of saving and investment for each sector, and a
separate memo column gives totals for the three do
mestic sectors. This memo column is one form of
the capital account in the Commerce Department in
come and product system. It can be seen that the
matrix, in its domestic sector columns, constitutes a
deconsolidation of that capital account, with capital
outlays distributed among the domestic sectors and
with a detailing of each sector’s net financial in
vestment that is a more general form of the net for
eign investment in the consolidated total.

This relation to the Commerce domestic capital
account is expressed in the first three rows of the
matrix. In the lower rows, the matrix goes into de
tail on the forms of lending and borrowing by each
sector that underlie the sector’s net financial invest
ment. Only two types are shown in the model—
deposits at intermediaries and loans and securities—
but the full matrix (page S.l) has many more cate
gories, as shown by table 7 on page 39. For each
of these financial categories a full accounting of
purchases and sales of the particular type of instru
ment is required in the system. It is this detailing of
credit transactions in the capital account that brings
the financial statistics of the economy into coherent
relation to one another and into direct relation to
the nonfinancial statistics in the income and product
accounts. The accounting link to nonfinancial trans
actions is net financial investment (row 3 of table
1), but the economic substance of the information is
in the interactions among specific types of credit
flows—deposits and loans—and between such flows
and specific forms of capital outlay, income genera
tion, and saving, all within the accounting con
straints of the system.
The matrix goes beyond the Commerce capital
account in that it incorporates the foreign sector
(rest-of-the-world) explicitly. This form requires
that the consolidated domestic capital account be
shown as a memo column, but it has the advantage
that for each category of financial transaction the
matrix states directly all of the transactions in the
market, whether domestic or foreign. Alternative
forms tend to obscure market conditions without
adding information.

Analytic role of matrix
The matrix is an essential framework for both cal
culating and using financial market statistics on an
economy-wide basis. It is general enough in form
to assimilate new types of financial instruments,
new forms of relationship, and the continual
changes in emphasis or practice in individual finan
cial markets. The explicit constraints of the system
enforce a consistency of analysis not easily reached
without the framework, particularly in questions at
a macroeconomic level, at which all market forces
interacting with one another are to be accounted
for. Such questions can be handled only when the
transactions involved have been stated within the
matrix on a complete basis but without double
counting.

Section 1: Concept

The role of the matrix for such purposes can be
demonstrated by the simple exercise of assuming
some major financial development, such as a sharp
rise in government borrowing or in deposit flows to
banks; placing that flow in its appropriate cell of
the matrix, and working out even a minimum possi
ble conjunct set of entries necessary to keep the
matrix in balance. If the example is in government
borrowing, that source of funds to the government
must be mirrored in the government column in
some combination of a nonfinancial deficit (nega
tive saving) and government lending as an offset
ting use of funds, because the money raised is obvi
ously being absorbed in one way or another. At the
same time, the borrowing itself must be matched by
an equal amount of lending somewhere along the
row that carries government securities. And in
whatever column that lending appears, there must
be a source of funds available for this use. In the
simplest situation, that source can be the positive
private saving and borrowing from the government
that are already implied in the government account.
For example, a minimum complete accounting for
the government borrowing might be as follows:
ITEM

PRIVATE
Use

Saving
Government loans
Government securities

GOVERNMENT

Source

Use

7
3

Source

-7

When this form of speculation is extended from
the merely possible to the likely, economic analysis
enters the exercise. In the example above there are
questions of the probable demand by private sectors
for other types of financial assets, such as cash and
other deposit claims, that are competitive with gov
ernment securities as investment forms and that af
fect the volume of flow into financial intermediaries
and the volume of credit supplied by intermediaries.
More broadly, analysis raises questions as to the
circumstances that generated the government defi
cit, including income distribution and private de
mand for capital goods, and the resulting influences
on credit market flows. Each aspect interacts with
the others, and as the various tendencies are item
ized they are to be fitted into the frame of the
whole in mutually consistent forms.
The operation of the matrix is also illustrated by
considering what happens when the money supply
increases. Money is a liability of the banking sys
tem and an asset of the public; any increase in it
must be accompanied by some combination of a de

crease in other bank liabilities, an increase in bank
assets, and offsets in the accounts for other sectors.
The organization of the accounts forces these contraentry questions to the surface and in the process
spells out the initial question in a complete form.
Analysis of this kind can be applied to an actu
ally expected set of developments by using the ma
trix structure as a device in forecasting or projecting
the future, with the specific function of keeping in
dividual parts of the forecast in touch with one an
other. The merit of such constrained systemwide
forecasts is that each element can be tested by the
plausibility of its counterparts in other areas of the
matrix. The structure as a whole is reasonable only
when all of its parts are reasonable. Whether the el
ements are derived econometrically from empirical
models or assembled judgmentally by hand, the
procedure has room for successive approximations
that approach the final result by working out the ef
fects of each change on the rest of the structure and
by then working back from the effects to revised
versions of the initiating change.
A complete forecast on this basis illustrates the
integral role of financial market behavior in capital
theory and in general theories of income, produc
tion, and economic growth. The choices each indi
vidual in the economy makes as to consumption,
physical investment, financial investment, and bor
rowing are related to one another and are confined
only as a group within the limits of his income and
net worth. The option of borrowing lets a person
shift his consumption and investment patterns over
time, and higher levels of debt allow him to carry
higher levels of either physical or financial assets at
any time. He may in his mind attach priorities to
one or another use of his income, but in practice all
of his demands work against one another to some
extent and indeed also influence the amount of in
come he tries to earn.
The columns of the matrix reflect these relations
among the activities of an individual transactor by
putting all of his transactions together in the general
form of a statement of sources and uses of funds.4
The system becomes macroeconomic when the
columns for all sectors are aligned with one another
to generate the market summary rows in which the
demands of different transactions impinge on one
another. The effect is a joining of financial invest
ment analysis directly to theories of production, in
come, saving, and physical investment in a manner
that adds generality to the model as a whole.
4. By statistical necessity, however, they are combined with
similar transactions of many other individuals.

Introduction to Flow of Funds

Stocks and flows

The generality of the matrix tends to obscure cer
tain structural aspects of the financial system that
are of continuing interest in analysis. These struc
tural aspects are concerned with concepts such as
intermediation, “primary” demands for credit, and
“ultimate” sources of credit—or more broadly with
“double-counting” of credit flows and the position

of financial institutions in the system. In a general
sense, intermediation consists of borrowing for the
purpose of lending rather than for nonfinancial out
lays. The term is usually associated with financial
business, such as banks, savings institutions, insur
ance companies, and investment companies that
concentrate on such activities. The distinction be
tween intermediary and nonintermediary sectors
must be recognized as an institutional one, how
ever, rather than a concept definable in theory; on
the one hand households, nonfinancial business,
and governments also engage in intermediation to
some extent, and on the other intermediaries are
subject to the same general investment principles as
nonfinancial sectors. Nevertheless, the difference in
degree is extreme and the distinction justified in
practical analysis.
Intermediaries tend to specialize in the forms of
debt they offer, or the forms of credit they extend,
or both. Insurance companies, for example, raise
funds primarily through policy premiums but invest
broadly in credit markets, while finance companies
specialize in their lending but not their borrowing.
Savings and loan associations are specialists both in
borrowing—through deposit accounts—and in lend
ing—through mortgages. In whatever way they spe
cialize, however, these institutions are filling a gap
between the types of claims the nonfinancial public
wants to hold as assets, such as liquid deposits and
insurance reserves, and the very different types of
claim the public wants to (or is able to) owe as
debts, such as bank loans, consumer credit, and
mortgages. With or without intermediaries, the total
of claims held as assets by nonfinancial transactors
is nearly equal to the total of their debts, because
directly or indirectly they owe the debt to one an
other. But with intermediation the composition of
their assets becomes very different from the compo
sition of their debts. The intermediaries are thus
performing a transformation process within the fi
nancial markets between the asset and the liability
sides of the public’s balance sheet.
Intermediaries are important to analysis in many

5. As the matrix on outstandings shows, stock data are pre
sented only for financial claims in the accounting structure and
do not include parallel data for holdings of tangible assets or for
sector net worth, which are the balance sheet counterparts of
capital expenditures and saving in the flows. A more complete
set of sector balance sheets that include tangibles and net worth
has been compiled recently at the Federal Reserve using Com
merce Department estimates of stocks of reproducible assets, es
timates of land values based mainly on data from the Census Bu
reau and financial positions from flow of funds data. These full
balance sheets are part of the flow of funds accounting system
and include statements of the relation between gross investment

as a flow and changes in net worth, both for principal sectors
and for national totals. Tangible assets are valued at current
prices in the balance sheets, and as a result much of the differ
ence between investment flows and movements in net worth
arises from valuation changes that are reflected in net worth but
not in saving or investment.
The complete balance sheets have been available so far only
on an informal basis, and their description has not been included
in this publication. They are maintained on a current basis in an
nual form, however, and copies can be requested from the Flow
of Funds Section at the Federal Reserve.

The immediate connection between financial mar
kets and nonfinancial activity is in terms of net
flows of claims, since it is as flows that financial
markets absorb funds from income and supply
funds to spending. These financial flows are always
increments in amounts of assets and liabilities out
standing, however, and the levels of these claims in
existence are as much a part of the picture as the
flows themselves. Economic equilibrium (in any
sense of that term) must be a balance simultane
ously among stocks, among flows, and between the
stocks and flows, a consideration that is reflected
not only in advanced models but also in such ruleof-thumb indexes as liquidity ratios, turnover rates,
and debt-service coverage by income.
Over the period covered by the flow of funds ac
counts, several types of credit have shown fairly
stable relations to expenditures or receipts in terms
of flows; but the flows have been at such rates as to
generate strong secular drifts, relative to activity, in
levels of debts and assets, either upward or down
ward. The meaning of these drifts in stock relation
ships, or even whether they have meaning, is an
important aspect for financial analysis; for such
questions, data on stocks of financial claims out
standing are included in this publication on a basis
parallel to the tables on flows, including both a ma
trix of claims as assets and as liabilities, on pages
S.2 and S.3, and time-series compilations for indi
vidual rows and columns.5

Intermediation and primary credit flows

Section 1: Concept
ways. Their presence in the market broadens enor
mously the forms of both financial investment and
borrowing available to the public: capital formation,
saving, income, and consumption are all higher
than they would be without the catalyst of interme
diaries in raising financial flows. In the U.S. econ
omy a large part of all credit goes through interme
diaries. In the short run much of financial analysis
is concerned with how well intermediaries are meet
ing demands for the specialized forms of credit they
offer with the funds they are able to attract from
savers. Legal constraints on their rate structures,
lending practices, and forms of borrowing often
prevent intermediaries from adjusting fully to cur
rent conditions and decision patterns, causing siz
able shifts of funds into and out of them. On the
other hand, intermediaries frequently introduce new
practices or new credit instruments that also have
major effects on the structure of flows. Both the
constraints and the innovations of intermediaries
have to be taken into account in even simple and
summary analysis of economic development.

Flow of funds summary tables
Intermediation implies some basic or primary flow
of credit in the economy that in part passes through
these institutions. The idea of such a flow raises the
question as to what it might be or, more specifi
cally, how it might be defined in an analytically
useful form. The matrix itself is too general in form
to show such a concept, inasmuch as it puts finan
cial intermediaries parallel to other sectors and
gives totals along the rows that include all sectors
indiscriminately. This matrix accommodates both
the intermediary-type debt owed by nonfinancial
sectors, particularly governments, and the market
able debt of intermediaries, such as bonds and open
market paper, that are not distinguishable as debt
instruments from the same types of claim owed by
nonfinancial sectors.
Nevertheless, the elements needed to approximate
a concept of a basic flow of credit that may or may
not be intermediated can be abstracted from the ma
trix. One such approximation is put together in the
two summary tables on credit markets that appear at
the beginning of most flow of funds presentations,
one on the structure of borrowing in credit markets
and the other on sources and forms of supply of
funds to credit markets. They appear in this publi
cation beginning on page S.4 for flows and page
S.6 for outstandings. The summary tables are not

an explicit part of the flow of funds structure, but
they are useful in outlining relationships among
forms of borrowing, among forms of lending be
tween credit demand and supply, and between inter
mediary and direct lending in markets.
Using figures from the model matrix in table 1,
the two summary tables are illustrated together in
table 2, with borrowing in the upper part and credit
supply in the lower. Table 2 identifies the interme
diation process through an institutional distinction,
isolating groups of firms that are mainly in the
business of borrowing funds for the purpose of re
lending and treating these groups as a channel of fi
nancial flows rather than a primary source of credit
demand or supply. The primary credit flow, on the
first row of the table, is thus borrowing by every
one else, that is, nonfinancial sectors, itemized by
sector in the next three lines. The total of 82 is less
than the matrix total of 92, on the bottom row of
table 1, by the amount of credit market borrowing
by intermediaries. The full form of the table, on
page S.4 includes the types of instruments used as
well as a listing of borrowing sectors.6
Part II of table 2 is a summary structure of the
sources of supply for the credit flows listed in part I.
It is more complex than part I in that it shows
6. Notes to table 1 list several types of financial claims that
have been omitted from the model matrix for simplification, such
as trade credit and taxes payable. These omitted forms are not
part of organized credit markets and are omitted from the full
form of table 2, as on p. S.4. Types of credit included in the
table 2 total are listed in the full version on page S.4.

Table 2. Model credit market summary
Figures from table 1

1.
2.
3.
4.

I Funds raised by nonfinancial sectors
Government

Foreign
Private domestic nonfinancial

5. II Sources of credit supplied
6.
Government
Foreign
7.

82
15
7
60
82
2
1

Financial intermediaries
Total funds advanced
Sources of funds
Private domestic deposits
Funds raised in loans and securities
Other sources

50
10
10

13.
14.

Private domestic nonfinancial sectors
Direct purchases of loans and securities
(lines 5-6-7-8+10)
Deposits in intermediaries (line 9)
Total financial investment (lines 12+13)

19
50
69

15.

Credit sources not in line 14 (lines 6 + 7+11)

9.
10.
11.
12.

Introduction to Flow of Funds

supply at more than one level simultaneously: direct
lending in credit markets by nonfinancial sectors
and by intermediaries as well as the sources of
funds to intermediaries to finance their part of the
direct credit supply. The first item (row 6 in table 2)
shows lending activity by federal government units,
federally sponsored intermediaries, and the Federal
Reserve, a source of funds to credit markets that is
almost entirely directed by public policy objectives
such as assistance to particular credit markets and
open market operations by the Federal Reserve.
Foreign lending on row 7 includes foreign private
funds, but in recent years it has been dominated by
foreign official transactions in U.S. government se
curities as a reflection of balance of payments de
velopments. Both row 6 and row 7 are largely ex
ternal influences on credit markets in that most of
the amounts that appear do not reflect profitmaking
decisions in the narrow sense but rather are policy
directed or almost automatic.
The middle section of part II of the table summa
rizes credit market lending by intermediaries, in
cluding commercial banks, and the main types of
financing for that lending. In the table 2 example,
intermediaries supplied credit of 70, mainly from
private deposit growth but also from their own bor
rowing in credit markets and from other sources,
such as foreign and government deposits, insurance
and pension reserve growth, and their own retained
income. The several sources of intermediary funds
vary greatly in relative proportions as credit condi
tions change, and shifts in their sources are re
flected elsewhere in the table, such as in govern
ment lending to intermediaries and the forms of
financial investment by the private nonfinancial sec
tors.
The bottom section of part 11, beginning with
row 12, integrates into the preceding picture the
structure of financial investment by private domes
tic nonfinancial sectors: households, businesses, and
state and local governments. Row 12 is direct lend
ing by these sectors in credit markets, which con
sists of any of the credit flow in row 1 not supplied
by government or foreign sources, or by intermedi
aries, together with credit market borrowing by in
termediaries on row 10. From the viewpoint of the
private sectors, this direct lending is but one of sev
eral possible forms of financial investment, with al
ternatives that in the model table are wrapped to
gether as deposits, such as checking accounts,
savings deposits, and negotiable certificates of de
posit. These deposits, shown here as a use of
funds, are the same flows that appear on row 9 in

the intermediary section of the table as a source for
intermediary credit supply. The distribution of pri
vate funds between direct lending and deposit flows
has sizable influence on although not total control
of the flow of credit through intermediaries and
thus on supply of the specialized forms of credit,
such as mortgages and bank loans, that come
mainly from intermediaries. Line 14 contains the
total flow from the private nonfinancial groups that
is distributed between direct and deposit flows. It is
somewhat less than total borrowing on row 1 al
though closely related. The difference, on row 15,
consists of direct flows from government and for
eign sources and the “other” sources of intermedi
ary lending.7 In recent years a major element of
change in row 15 has been foreign official lending
that reflects shifts in the U.S. trade balance and
capital outflows in the balance of payments state
ments. Most movements in government lending are
reflected directly or indirectly in the private total on
row 14, while intermediary “other” sources are rel
atively stable over the long run.

Diagram of financial structure
The view of the economy that is reflected in table 2
is indicated in diagram D-l. The diagram is a
picture of the nation’s capital account and, again,
uses the model matrix (table 1), abstracting in the
same way from the full complexity of the system.
The diagram is specific to the data in the model in
that it shows current-account deficits for the gov
ernment and foreign sectors on the right, in parallel
with physical investment, as “uses” of private sav
ing entering the capital account on the left; if either
of these sectors had a positive current balance, it
would appear at the left in the diagram. In a fully
detailed picture, the dissaving of any individual
transactor with a negative current-account balance
would also appear on the right, with treatment of fi
nancial flows parallel to that for a deficit govern
ment. To simplify the diagram, all saving of the
private domestic sectors is on the left, even though
the total for the sector is a net sum for savers and
dissavers together.8 The diagram takes on precision
if each sector is viewed as a single person in an
economy made up of only four different persons.

7. Row 15 is not included in the full form of the table.
8. Saving in the diagram is the total for private sectors, in
cluding intermediaries. Intermediary saving is an internal source
of funds for lending that is identified in the diagram.

Section 1: Concept
This primitive view can be extended easily to a
more general picture with many separate units in
each sector.
The diagram pictures saving entering the capital
account as a diversion from the current payments
stream and passing through a number of channels to
finance physical investment outlays and government
and foreign deficits that inject spending back into
the current stream. Part of the saving goes directly
into physical assets, to the extent that people buy
ing capital goods pay for them without recourse to
borrowing, and this appears as “internal finance,”
the excess of capital outlays (170) over private bor
rowing (60). Another part of saving goes into finan
cial assets, however, such as cash and deposits for
liquidity and marketable securities for capital gains.
Part of this financial investment goes directly to
nonfinancial borrowers (12), but most of it is put
into deposit (55) and security claims on intermedi
aries (10), who relend the funds to nonfinancial
sectors through quite different forms of credit in
struments, such as mortgages and bank loans (70).9
These credit flows from intermediaries, combined
with direct lending from savers, make up the total
9. Intermediaries use their own retained income (5) to lend
somewhat more than they borrow.

9
volume of borrowing (82) by nonfinancial sectors
that it used mainly to cover the deficits of dissavers
and to pay for investment outlays not covered by
internal finance.
The bottom of the diagram shows a reverse flow
of funds from the borrowing to the lending side of
the structure that is equal to the investments in fi
nancial assets made by the two deficit sectors. In
the model matrix (table 1) these two sectors have
not limited their borrowing to the minimum needed
to cover their deficits but rather have borrowed ex
tra amounts to add to their asset positions (8). They
borrow in part for liquidity purposes and in the
government’s case to help carry out lending pro
grams for public policy purposes. These extra
credits represent borrowing in order to relend and
as such constitute a form of intermediation by non
financial sectors. The U.S. government does in fact
act as an intermediary to financial markets, floating
its own securities to assist agriculture, small busi
ness, the home mortgage market, and other private
markets; a more complex diagram could show that
activity explicitly. The more general point, how
ever, is that quasi-intermediation occurs in many
forms in nonfinancial sectors, that all of its forms
create reverse flows in the diagram, and that they

Introduction to Flow of Funds

add an element to the relation between the total
flow of credit as defined in the diagram and the as
sociated totals of saving and investment.
Typically, this kind of intermediation is difficult
to identify, as is illustrated by taking the private do
mestic nonfinancial sector in the diagram to be a
single individual. Even though this person’s saving
(179) was greater than his capital outlays (170), he
chose to put a substantial amount into financial in
vestments (69) and then to borrow some of these
funds on the other side of the market in different
forms (60). This is reasonable behavior, since the
combination of assets and liabilities he now has
suits his short-term and long-term needs better than
do lower totals of both assets and liabilities. How
ever, it raises the question whether he borrowed to
invest in physical or in financial assets, and the an
swer is that he borrowed for the two purposes
jointly in unidentifiable and even undefinable pro
portions. Only in special cases, such as the two
deficit sectors in this model or a borrowing total
that is larger than total investment outlays, can such
intermediation be even partially measured.
This discussion illustrates the ambiguities in the
concept of intermediation and thus in the concept of
a basic or primary flow of credit that is to some ex
tent intermediated. Intermediation is nevertheless a
useful construct for analysis when it is given an in
stitutional sense that is based on the characteristics
of a set of financial businesses, including the legal
constraints on their operations, their typical prac
tices as borrowers and lenders, and their flexibility
in responding to changing economic conditions.
Isolating financial institutions as an intermediary
group brings out the broad division of financial
flows between those that enter this area of special
ized and constrained lending operations and those
that are available only in more generally marketable
instruments. It is this institutional foundation for
analysis that underlies table 2. As applied to nonfi
nancial sectors, the concept of intermediation is too
ambiguous to be useful, and it is well replaced by
the integrated balance-sheet view of physical invest
ment, financial investment, and borrowing that is
implicit in this discussion.
In comparison with the accounts as published,
the diagram is primitive although accurate as far as
it goes. In both the model matrix and the diagram,
credit flows are limited to the main-stem group of
financial claims that are handled in organized credit
markets, such as securities, mortgages, consumer
credit, and bank loans. In flow of funds publica
tions these are labeled credit market instruments.

This central group of claims is the focus of the
summary tables in publications as well as in the
model used for this discussion. A glance at the full
matrix on page S.l reveals that the financial struc
ture as a whole includes a variety of other claims,
such as gold, foreign exchange, trade credit, and
equities in noncorporate business. These are more
specialized instruments that are also part of the fi
nancing of the economy and that appear in the ac
counts where appropriate, but for summary pur
poses they are treated as outside the credit markets
proper. The diagram is also primitive in that it can
not easily show negative financial flows, such as
debt repayment or reductions in asset holdings, and
because it ignores the layers of intermediation
among financial firms, such as bank loans to secu
rity dealers and insurance company purchases of fi
nance company bonds.
The diagram serves only to illustrate in broad
outline the relation between saving and investment
on the one hand and the aggregates of borrowing
and intermediation shown in table 2 on the other.
The financial markets absorb part of saving and
supply part of the funds for spending; but the total
volume of credit flows as defined here has no nec
essary relation, dollar for dollar, either with saving
or with investment because of the opportunities for
internal finance and for borrowing to carry financial
assets that the diagram suggests. The effects of re
stricting or expanding credit flows are thus not nec
essarily or immediately on saving and investment
but rather tend to be diffused throughout the system
inside and outside financial markets. Such relations
as exist between credit flows and nonfinancial ac
tivity must be found analytically and empirically,
with credit seen both as borrowing by nonfinancial
sectors and as lending by those same sectors.

Historical relationships
The charts on the following pages give examples of
the empirical relationships over time that have em
erged from the data at the broadest levels of the
economy. The charts do not go into specific sector
analysis, which would consist of relations among
shifts over time in sources of funds, uses of funds,
prices, and interest rates, and which can become
more complicated than is suitable for this historical
survey. The focus is rather on broad aggregates and
on the story they tell of changing financing prac
tices and changing financial structure. Some of the
relations are close, but they do not point directly to

Section 1: Concept
conclusions about the workings of the economic
system as a whole. With credit flows dependent on
demands by borrowers for financial assets as well
as their nonfinancial outlays, the problem of distin
guishing supply and demand is more complex for
financial markets than for many other areas of anal
ysis, and has yet to be worked out.
The charts start, somewhat arbitrarily, at the
point of household and business borrowing, proceed
to the total borrowing by nonfinancial sectors that is
pictured in the diagram, connect that total to private
financial investment, and summarize the results of
these flows in terms of debts and assets outstanding.
The statistical relations of the charts to flow of
funds tables are listed in the notes to charts (page
23), which refer to the appendix tables (pages
S.1-S.10). Except for the matrices and the two
summary tables discussed earlier, the only tables in
cluded are those reflected in the charts.10 Before
charting, all of the data used were deflated by a
compound index of prices and the 1952-79 trend of
gross national product in order to highlight cyclical
relationships apart from the strong growth trends
over the period in most of the data. A single defla
10. The tables carry annual data for a few years only. They
are taken from Flow of Funds Accounts, 1949-78 (Board of
Governors of the Federal Reserve System, 1979).

tor was used for all of the time series, and a rising
trend in any of the plotted series indicates a rate of
growth faster than the trend of GNP, although not
necessarily faster than GNP growth in any short
period. A series with a falling trend is not necessar
ily decreasing in actual dollar amounts; it may be
only increasing at a slower rate than the GNP trend.
The deflator itself is presented and described in the
notes to charts; it is not directly part of the ac
counts, which show only current-dollar quantities.
Charts 1 and 2 illustrate the associations that
have existed between short- and long-term borrow
ing in credit markets by business and households,
and their spending for capital goods. Capital ex
penditures here cover all private domestic invest
ment in the national income and product accounts
except capital outlays by financial sectors, and they
also include purchases of consumer durables, which
are consumption spending in the national income and
product accounts (NIPA). Chart 1 shows the closely
parallel movements of changes in short-term busi
ness credit and net inventory movements over the
period except for a short time in the late 1960s and
during 1973. The relation exists not with the NIPA
measure of inventory change, however, but rather
with book inventory movements, which include
complicated interactions among price movements,
inventory turnover rates, and inventory accounting

Introduction to Flow of Funds

methods such as FIFO and LIFO (first-in-first-out
and last-in-first-out). The difference between the
two inventory measures is small before the later
1960s on the scale of this chart, and the NIPA mea
sure, which reflects standardized pricing and ac
counting methods, is shown in the chart only for
the later years, when higher inflation rates made the
difference substantial. On an a priori basis there is
no reason for the closer relation of short-term fi
nancing to book rather than NIPA inventory move
ment because the higher prices that buyers pay
during inflation to maintain inventory on a book
basis is offset, in general, by higher prices in their
sales, and the increased cash flow from sales are
available to be used directly to finance higher costs
of inventory maintenance. The chart indicates that
this is not dominant business practice, however,
and that in general price effects of inflation on cash
flow appear to be isolated from inventory financing
operations. The 1969 and 1973 aberrations in the
chart are strong cases of the general tendency for
short-term borrowing to grow somewhat more than
inventories during cyclical expansions and to fall
somewhat more during recessions. This tendency is
consistent with increased expectations, as activity
approaches a peak, that short-term financing can
soon be refunded in longer-term forms at more at
tractive long rates.
Chart 2 gives a similar pairing of longer-term
private investment and credit, which in this chart is

a mixed collection of corporate bonds, mortgages,
and consumer credit.11 In the early years shown in
the chart the high rate of investment relative to
credit raised was centered in housing, in which pur
chases of new construction were considerably larger
than net mortgage borrowing. Beginning in 1961
home mortgage financing shifted upward relative to
construction, possibly reflecting an increase in
transactions in existing houses; but later in the
1960s this shift was offset by a marked reduction in
consumer credit flows relative to consumer durables
purchases and by a restraint in business long-term
financing relative to outlays. In the later 1960s
business relied heavily on short-term credit, as chart
1 shows, but in 1970 long-term financing shifted
upward again relative to spending, at first to fund
short-term debt and later to take advantage of the
expanded supply of both business and housing
credit. The years after 1974 contrast with preceding
periods in the high rates of private borrowing rela
tive to both capital spending and GNP. After 1974
net borrowing by the group started running at rates
well above net fixed investment, primarily because
of unusually large financings in residential mort
gages. While these flows accelerated the growth
rates in debt outstanding during the later 1970s,
the high inflation rates of those years acted to mod-

Chart 2

11. Investment appears in charts 2 and 3 net of depreciation
charges, which are measured on a current-cost basis, as in NIPA.

Section 1: Concept

erate the rise in debt relative to income and prod
uct, as the later charts show.
Chart 3 combines the data from charts 1 and 2
and illustrates the extent to which movements in
private capital outlays have been dominant com
ponents of GNP fluctuations. The two vertical
scales in chart 3 have the same gradient while their
absolute values are very different, and only during
the Vietnam period of the later 1960s are the move
ments in GNP substantially larger than those in the
investment series. Total private borrowing has al
most the same volatility as net investment and thus
is almost as closely related to fluctuations in GNP.12
Chart 4 completes the structure of borrowing by
adding to the business and household components
net funds raised by foreign borrowers, state and
local governments, and the U.S. government. Net
borrowing by the U.S. government is much the
most volatile of these elements and is the source of
most of the changes in differences between the pri12. A common practice in current analysis is to measure
either investment or net borrowing as a percentage of GNP. The
relationship illustrated in the chart—that is, roughly equivalent
amplitudes on very different base levels—produces volatile per
centage movements relative to GNP that are sensitive indicators
of cyclical movements. The chart suggests, however, that the
movements in percentages are somewhat beside the point and
that comparisons of absolute movements indicate more directly
regular and irregular cyclical developments.

vate and total series. In general its effect has been
to shift peaks and troughs into earlier quarters,
making total borrowing a slightly leading series at
GNP turning points.13 Over the 1970s federal net
borrowing also contributed considerably to the up
ward tendency of total credit flows relative to GNP
trend, for reasons that are apparent in later charts.
State and local governments are a hybrid group in
the economy and hence in the social accounts: as
governments they are attributed no physical invest
ment outlays in income and product data but as in
dependent decisionmakers these units base their fi
nancial planning on much the same market
considerations as households and businesses and in
this respect are part of the “private” segment of
credit markets. Although borrowing by state and lo
cal governments is excluded from the private totals
in charts 1 and 2 that are related to physical invest
ment, their investment in financial assets is in
cluded with other private financial investment in
later charts on supply of funds to credit markets.

13. Turning-point relationships among series are affected by
the deflation of the data that was described earlier. With undefla
ted data, the leading characteristic of total borrowing is more
pronounced, because adding growth trends to the data shifts
peaks in GNP further into later quarters than peaks in borrowing.
The reason is that GNP is a proportionately more stable quantity
than net credit flows and has a relatively larger trend component.

Chart 3

1953

1957

1961

1965

1969

Introduction to Flow of Funds

Charts 5 and 6 indicate the cumulative effect on
debt outstanding that has resulted from the structure
of borrowing since 1952 that appears in preceding
charts. Within a total debt owed by nonfinancial

sectors that has risen only slightly in relation to
GNP, there was a major shift from federal into pri
vate liabilities outstanding over the whole period
1953-74, with U.S. government debt shrinking

Chart 4

Chart 5

Section 1: Concept

from 47 percent of the total in 1952 to 17 percent
at the end of 1974.14 The explanation for the shift
lies in a combination of circumstances: the legacy
still remaining from depression and war at the be
ginning of the period in the form of high public
(and low private) debt, the strength of private in
vestment demand stemming partly from the same
cause and tending to generate private debt, and the
favorable government budget position that reflected
strong private demand for goods and services.
The resulting trends in debt structure were strong
but could not be expected to continue indefinitely.
Growth of private debt, particularly that of house
holds and state and local governments, slowed
markedly after 1965 and has been roughly on a pla
teau relative to GNP trends since then. A sharp
burst of private borrowing in 1972-73 was followed
14. Deflating flows and levels by a single index of growth
and prices for the charts creates a special relation between the
deflated figures for net borrowing and changes in debt outstand
ing. If the deflator increases 5 percent a year, borrowing must
equal 5 percent of outstanding debt merely to keep deflated debt
constant. A borrowing rate of more than 5 percent will raise the
debt level, but if the rate is less than 5 percent, deflated debt
goes down even with positive borrowing. In the data used for the
charts, average growth in the deflator was 6.8 percent per year
from 1952 to 1974. U.S. government debt was growing in abso
lute terms by only 2.2 percent, however, which deflation con
verted to a 4.5 percent rate of decrease. Private debt, on the
other hand, was growing at a 9.4 percent annual rate, well above
the deflator. These figures exclude corporate equity issues from
both debt and borrowing.

Chart 6

15
by an equally sharp correction in 1974-75, and
while those were turbulent years burdened by food
production problems, oil price increases, and an ep
isode of price controls, they also saw a strong
runup in capital outlays (charts 1 and 2), and an
equally strong reversal in capital spending during
the 1974-75 recession. The downtrend of federal
debt appears to have come to an end in 1975 and,
after a short-run compensation for the drop in pri
vate debt, federal debt maintained a fairly stable re
lation to GNP trends. In the most recent years
shown in chart 5 there is some tendency toward re
sumed growth of private debt offset by decreases in
federal debt, and the chart suggests that if private
capital formation is to be maintained at adequate
rates financing may have to shift from heavy reli
ance on debt toward greater use of internal or exter
nal equity funds.
Chart 6 divides private debt into four types, in
cluding state and local government securities. All of
these forms were growing relative to GNP trends
until 1965, but from then to 1974 only business
debt continued an upward tendency. After the
1974-75 correction residential and business loans
were the sources of strength that underlay the up
ward tendency of the total in chart 5. Growth in
these debt forms was less than charts 1 and 2 might
suggest, however, because much of the flow in
those charts was offset by the high inflation rates of

Introduction to Flow of Funds

the period and is removed by deflation in chart 6.
State and local government debt shows a steady de
clining trend even after the sizable advance refund
ings of 1977-78 that are included in the total.
Charts 7 through 12 shift to the supply side of
credit markets and summarize aspects of private
nonfinancial sectors as lenders rather than as bor
rowers. As diagram D-l and table 2 illustrate, most
of the total borrowing by nonfinancial sectors that
appears in chart 4 has a counterpart in accumulation
of financial assets by private domestic nonfinancial
sectors, either directly through security purchases in
markets or indirectly through investment in deposits
or other claims on intermediaries that are lending
directly. The relationship is presented in amounts
outstanding in chart 7, in which the top line is total
debt of nonfinancial sectors, the middle line is
holdings of such debt by private domestic sectors
including intermediaries, and the bottom line is de
posits and security holdings by households, nonfi
nancial business, and state and local governments.15

In terms of the model table 2, these are respectively
item 5, items 8 plus 12, and item 14. The gap be
tween total debt and the middle line consists of fed
erally related and foreign direct holdings, and the
increase in that gap after 1975 reflects mainly the
large increases in foreign official holdings of U.S.
government securities related to international trade
deficits. The gap between the middle and bottom
lines consists of net holdings by intermediaries fi
nanced by sources other than private domestic de
posits and securities: mainly insurance and pension
reserves, federal and foreign deposits, and the eq
uity funds of intermediaries. That gap also widened
markedly after 1975 as a result of an accelerated
flow of funds into pension reserves and the shift of
those flows into investments in credit instruments.
These developments produced a clear contrast in the
chart between the definite upward trend of total
debt after 1975 and the essential flatness in private
nonfinancial assets, reflecting a strong growth in
these other sources of credit market funds.

15. Insurance and pension reserves appear as assets of house
holds in the total accounting structure and on that basis could be
included in the bottom line as assets of the nonfinancial group.
Such reserves are, however, more remote from day-to-day in-

vestment decisions of households than are their deposit and secu
rity portfolios and on that basis are set aside in summary tables
and in these charts as a separate financial relationship in the sys
tem.

Chart 7

Section 1: Concept
Chart 8 shows the marked shift over 25 years in
the relative positions of private financial assets and
debt in the economy. This shift combines the rela
tion in chart 7 between private assets and total debt,
which had little trend over the last 20 years relative
to activity, with the shift in chart 5 from federal to
private debt outstanding. The chart illustrates that,
to an increasing extent since the early 1950s, the fi
nancial assets that private investors hold have come
to be based directly or indirectly on claims owed by
themselves rather than by others. For the most re
cent years the widening gap between the two lines
in the chart reflects primarily the accelerated flow
of pension reserve funds into debt securities men
tioned earlier.
Paralleling this change in asset and debt structure
has been a sizable increase in the degree of inter
mediation in financial markets. Private sectors as
borrowers must look mainly to financial institutions
for mortgages, consumer credit, and business loans,
all of which are usually too specialized and too
small in individual loan size to be broadly market
able; and even for their marketable debt, such as
municipal and corporate bonds, the institutions are
important suppliers of funds. As their debts have
grown, private sectors as lenders also have turned
increasingly to institutional deposits as an invest
ment alternative to the diminishing supply of fed

Chart 8

17
eral securities outstanding. Over the period covered
in the charts, institutional holdings of credit instru
ments increased from 68 percent of the total held
privately in 1953—the middle line of chart 7—to
84 percent at the end of 1979. Most of this growth
was financed by the shift of private investment
away from direct market instruments and toward
deposits in intermediaries, which increased from 58
percent of private portfolios in 1953—the bottom
line of chart 7—to 70 percent in 1979.
That shift in private investment is broadly indi
cated in chart 9, which divides the private asset to
tal in chart 8 (and chart 7) into its deposit and
credit instrument components. The two components
have been rather more volatile over the short run
than their total has been, reflecting the sizable
short-run changes in yield relationships between de
posits and market instruments and the shifts of
funds in response to changes in the yield spread.
Interest rates on most deposits have been much
more stable cyclically than have market rates, partly
because of regulatory ceilings that have restrained
those rates from following market rates upward
when credit has been tight. In the sequence of tight
financial conditions covered by the chart—
principally 1959, 1966, 1969, and 1973-74—high
rates on market instruments drew increasing
amounts of funds out of institutions and into direct

Introduction to Flow of Funds

investment, reflecting widened yield spreads and
heightened investor sensitivity to those spreads.
Some of the deposit growth in the first half of the
1960s resulted from the development of large nego
tiable certificates of deposit by commercial banks,
which introduced the CDs as money market instru
ments to attract funds from large investors, such as
major corporations, that had been holding liquidity
mainly in Treasury bills. With the diminishing posi
tion of Treasury debt in the market, the CD became
a major investment medium and an important con
duit for converting liquidity holdings into bank
credit for the growing volume of private debt.
The effect of rate ceilings on deposits was most
extreme in 1969, as chart 9 indicates, when CDs in
particular fell from $23 billion at the beginning of
the year to $9 billion at the end (in actual dollars).
Following that episode ceilings were lifted on CD
rates, and in the 1973 tight-credit period the effect
of yield spreads was much more moderate than in
1969.
In 1973 there was also a marked increase in se
curity repurchase agreements held by nonfinancial
investors and included in the deposit total. RPs are
a short-term form of bank liability that have never

been subject to rate ceilings and that can yield the
investor money market rates of return on maturities
as short as one day. They are included in the de
posit total in chart 9 and are another reason why
deposits held up fairly well during 1973-74. After
1977 the introduction of six-month money market
certificates of deposit, the growth of money market
mutual funds, and other new channels opened ac
cess to market yields to much smaller deposits, as
little as $1,000 in some cases; and by the end of
the charted period most of the yield-spread incen
tives toward disintermediation had vanished from
credit markets. The effect was to give deposit insti
tutions resilient lending positions during tight credit
conditions but at the same time to put an increased
part of their liabilities at market rates of interest.
This change in their cost of funds was carried
through to the rates that institutions charged, which
also became more volatile and more market related
for all types of loans, including mortgages and con
sumer credit. The shift toward stronger and more
market-related lending positions for intermediaries
fitted closely as a structural change with the re
duced supply of federal debt as a liquidity instru
ment for private investment.

Chart 9

1953

1957

1969

1973

1977

Section 1: Concept
Chart 10 combines again the deposits and credit
instruments from chart 9 to compare the total hold
ing—the same total that is in chart 7 and chart 8—
with GNP. In spite of the shifts in composition of
total assets, the relationship to GNP is unusually
close over most of the chart in both movement
and—in what is presumably only a coincidence—
absolute value. The asset total tends to have some
what smaller cyclical movements than GNP and
shows a slight leading tendency. These differences
are small within the total relation, however, and
over the span of years covered no significant drift
in trend is apparent. The relation is closer than that
between GNP and debt totals, with the differences
absorbed in connective elements of credit supply,
such as government, foreign, and pension reserve
sources of supply, that constitute the gaps between
the three lines in chart 7.
This connection between private financial assets
and activity is, like the rest of the material in the
charts, an empirical “black box” in that it neither
supports nor is explained by any broadly accepted

19
analytical system. Without analytic support there is
no basis for predicting whether it will continue un
der other economic circumstances, such as chroni
cally higher inflation rates or slower economic
growth rates. The persistence of the relation on a
quarterly basis nevertheless suggests a connection
between activity and asset holdings that may be a
macroeconomic constraint of importance both to
forecasting and to policy. That it is closer than the
debt relation to GNP suggests further that changes
in other forms of credit supply may be the source
of changes in the debt relation and whatever conse
quences follow from the debt movements. During
1976-79, when a marked increase occurred in both
net foreign and pension reserve supply, private bor
rowing was high (chart 3), and private debt rose
faster than trend (chart 5), with very little effect on
the asset connection in chart 10. A continuance of
such conditions could lead into dangerously burden
some levels of private debt without the correspond
ing rise in private liquidity that accompanied the
1972-73 debt increase.

Chart 10

Billions of 1972 dollars

1220

1180

1140

1100

1060

Introduction to Flow of Funds

Chart 11 breaks out the household component of
private assets to illustrate the drift in sector distribu
tion of holdings within the total. Over the period of
the chart the household share increased from 71
percent of the total to 82 percent, with an offsetting
decrease, from 19 to 10 percent, in holdings by
nonfinancial business—mainly business liquid as
sets. State and local government holdings rose from
6 percent of the total to 8 percent, even with a de
clining trend in their debt outstanding. Statistically,
the total is more reliable than its parts, particularly
the distribution between household and business as
sets, and the shift may be weaker than source data
suggest. For corporations, however, a drop in liquid
asset position from the early 1950s is fairly well
supported, and there is little question that over 25
years corporations have held diminishing amounts
of a total that has stayed almost constant relative to
GNP.
Chart 12, finally, introduces corporate equity
holdings into the picture. The preceding charts have
focused on debt instruments, both as liabilities and
as assets, and the relationships have appeared for
this total of claims. Equities have a separate posi
tion in the financial system in that as liabilities they
are only residual claims and, in a legal sense at
least, are not burdens on issuers. In the flow tables
net new stock issues appear as external sources of
funds to business and as net financial uses of funds
Chart 11

by investors. In tables on outstanding assets and lia
bilities, however, they appear only as assets valued
by market prices, and no specific liability for them
is attributed to issuers. While net purchases of equi
ties are a small component of household financial
investment, holdings valued at market are a major
part of household assets, as large as all deposits and
debt securities over much of the period of the chart.
Chart 12 illustrates that such holdings are also far
more volatile in amount than other financial assets
because of movements in stock market prices.
There is a mild correlation between changes in the
two that is obscured by the compression of the ver
tical scale of chart 12 needed to cover movements
in stock values. In the deflated dollars of the chart,
equity values have moved by roughly $7 for each
$1 change in holdings of deposits and credit instru
ments, neglecting the many short-term perturbations
in the equity series and some irregularities in the
timing of movements. There may be interaction be
tween the two totals that is exaggerated for equity
values because of the lack of net flows of new is
sues, but to some extent the two series are probably
also responding in parallel to other conditions in the
economy.
The shift in credit structure from directly held
central government debt to intermediated private
debt, which is a dominant feature of the charts, is
put into another perspective in table 3. This table

Section 1: Concept

Table 3. Private claims on U.S. government institutions
Amounts outstanding at year end in 1972 dollars, trend removed

1952

1965

1974

1978

1979

807

500

362

423

418

98
64
2
643

71
31

60
20

58
16

57
14

397

282

348

346

6. Holdings, by group

807

500

362

423

418

7.
8.

Foreign
Private domestic

15
792

23
477

45
317

71
352

57
361

9.
10.

Financial
Nonfinancial

399
393

220
257

136
181

166
186

165
196

89
304

63
194

52
130

51
136

49
147

571
1,134

740
1,213

738
1,128

751
1,131

731
1,121

1,149

1,648

1,493

1,518

1,528

Item

1. Total claims, by type
Currency in circulation
Member bank reserves
Foreign deposits at Federal Reserve Banks
U.S. government securities

2.
3.
4.
5.

Currency outside banks
U.S. government securities

11.
12.

Memo: Private domestic nonfinancial sector

13. Total deposits at financial institutions
14. Total money, deposits, and credit instruments
Memo: Financial business holdings of reserves and credit instru

ments
* Amount insignificant.

pulls together, again in detrended and deflated dol
lars, a set of claims on the U.S. government and
closely related institutions as a proxy for a total of
basic reserve assets that are available for private in
vestment. The set of claims is broader than govern
ment debt in chart 5 in that it includes reserve
money and also the debt of federally sponsored

credit agencies and mortgage pools; these have
played a greatly expanded role in recent years as
government-related intermediaries, lending housing
and farm credit on the basis of public issues of their
own debt.16
16. The total is also a little narrower than that in chart 5 in
that it omits government debt held by the Federal Reserve.

Chart 12

1953

1957

1961

1965

1969

Introduction to Flow of Funds

The total of these reserves fell sharply from the
1950s to 1974 but then moved upward again with
the reversal in trend of U.S. government debt. Part
of the rise after 1974 was absorbed by foreign inter
vention operations in international exchange mar
kets. Domestic holdings, on row 8, also went up,
however, in both Treasury and federally sponsored
securities. The rise in domestic holdings of Trea
sury securities beginning in 1975 was mainly in fi
nancial institutions, which for various reasons had
gone into extended and illiquid positions by the end
of 1974. Much of the increase reflected a shift of
pension fund and insurance investment flows away
from equities and toward high-grade debt instru
ments. That shift was prompted by legislation and
may have been a one-time adjustment toward an as
set structure that will be maintained in the future
with a smaller current demand for Treasury securi
ties. Banks almost doubled their Treasury security
holdings in 1975-76 as part of a recovery from a
heavily loaned position at the end of 1974 that was
the counterpart of the high private debt that appears
in chart 5. The picture here is of an unstable finan
cial condition at the end of 1974 that was followed
by abrupt but interrelated corrections by both finan
cial and nonfinancial sectors in both assets and lia
bilities. These corrections were followed after 1976,
however, by a return to many of the earlier trends
that had led into 1974, with the possibility of an
early repetition of the 1974 experience.
One of the effects of the increasing exposure of
private finance in credit markets has been the bur
geoning of the federally sponsored credit activities
mentioned earlier. While operating in specialized
credit markets—mainly housing and farm credit—
the sponsored agencies expanded their debt on row
5 of table 3 from $7 billion in 1952 to $102 billion
at the end of 1978 (in the deflated dollars of the ta
ble) to finance an almost equal increase in their
holdings of private credit. If private debt continues

to grow relative to activity, these agencies will
probably absorb a rapidly increasing share of the to
tal growth accompanied by growth in a variety of
government guarantee programs for private credit.
Direct lending and loan guarantee operations by the
government have focused on particular kinds of
credit that have been in difficulty, and they are seen
mainly as a method by which the government can
help those loan markets to compete against other
kinds of demand. While the agencies appear some
times to be “draining” loanable funds from private
markets and creating credit tightness that would not
exist otherwise, they also perform an important
function in supplying an investment and liquidity
instrument—in the form of their own debt—of a
kind that has become increasingly scarce.
This review of the historical data has not tried to
explain analytically the cycles and trends of postwar
financial developments or to point up trends with
alarm. Its purpose has been rather to illustrate main
connections within the flow of funds accounts
among lending, borrowing, balance sheet positions,
and nonfinancial activities—connections that are
close over the period covered and that are basic
data for analysis of the economy as a whole. Some
of the connections have shown tendencies to shift
over the period, and financial markets have re
sponded with facility and speed in adapting to new
practices and new financial instruments. These
changes in financial flow structure usually appear in
individual markets or sectors and can be explained
to some extent by detailed analysis of those markets
or sectors. They occur within the frame of the
whole, however, and are in part reactions to
changes in that frame. Whether as a framework that
constrains particular markets or as elements acting
on one another within a system, the main members
of the structure appear to have an empirical exist
ence that should be recognized explicitly in the data
and in analysis of the data.
□

Section 1: Concept

Notes to charts
All data in charts 1 through 12 have been deflated by a single
compound index (1972 = 100) that is the product of (1) the
U.S. nonfinancial population 16 years of age and over, (2) the
GNP price deflator, and (3) an exponential growth trend in de
flated GNP per capita found by least-squares regression to be 1.8
percent per year for the period 1952—79. After deflation, all flow
data are plotted as centered two-quarter moving averages (three
quarters weighted 1-2-1), with 1953 Q1 and 1979 Q4 omitted
from the charts. Assets and liabilities outstanding are deflated by

Table

Chart Item

1
2

4
5

S.10
Net change in inventories
S.10
Business loans
Private net fixed investment
Households + business borrowing
S.4
except business loans
Private net capital outlays
S.4
Households + business borrowing
S.4
Total nonfinancial borrowing
S.4
Households + business borrowing
S.6
U.S. government debt
S.6
Other nonfinancial debt
S.10
Business loans
S.6
Residential mortgages
S.6
Other business + household debt
S.6
State & local government debt
S.6 & S.7
Total debt
Privately held credit market
S.7
instruments
Private domestic nonfinancial sector
S.7
deposits and credit market assets
S.7
Deposits and credit market assets
S.6
Credit market debt
S.7
Deposit assets
S.7
Credit market assets
S.7
Deposits and credit market assets
holdings
S.7
Deposits and credit market assets
S.9
Household deposits and credit market assets
S.9
Deposits and credit market assets
S.9
Corporate equity holdings

the same index, but they are plotted directly, not as moving av
erages. To eliminate quarterly seasonality of outstandings the
year-end levels were incremented to other quarters of each year
by the seasonally adjusted quarterly flows. Corporate stocks are
omitted from all data in the charts both as assets and as liabilities
except in chart 12.
Data for the charts are derived from tables at the back of this
publication (pp. S.1-S.10) and are identified in the following
list:

Line
9, upper section. Book value = line 9 less total IVA
18 4- 19, upper section
Table S.8, 13 - 10 + Table S.10, 5 less capital consumption
27 + 28 upper section less business loans in chart 1
From charts 1 and 2
27 + 28 upper section
2, upper section
27 + 28 upper section
2
5
12 + 13, lower section
10 + 11
21+22 less business loans and residential mortgages
20
1, total credit market debt owed by nonfinancial sectors

12
48, credit instruments, deposits, and currency
48
5
38
32
48
48
2
2
19 (at market value)

Fourth-quarter values of the deflator are presented below in reciprocal form, as multipliers against actual data:

1952
1953
1954
1955
1956
1957
1958

3.233
3.136
2.995
2.833
2.661
2.516
2.408

1959
1960
1961
1962
1963
1964
1965

2.280
2.176
2.095
1.986
1.885
1.794
1.693

1966
1967
1968
1969
1970
1971
1972

1.581
1.483
1.366
1.250
1.144
1.051
.972

1973
1974
1975
1976
1977
1978
1979

.872
.758
.681
.627
.570
.509
.453

Section 2: Organization of Accounts
Section 1, on the concept of flow of funds ac
counts, describes the system only in broad terms
and does not constitute an operating description of
the system. Sections 2, 3, and 4 define the accounts
in terms of the rules that organize the system, the
relation of the accounts to income and product in
formation, and descriptions of individual sector and
transaction categories. Statistical derivation proce
dures for individual items in the accounts and pro
cedures for processing source data are described in
a separate publication.
Section 1 emphasizes that the matrix organization
of data is fundamental to the calculation, under
standing, and analysis of flow of funds information.
The matrix is also the organizing principle for the
statistical tables in flow of funds publications, each
of which is a statement, in time-series form, of one
column or one row of the matrix taken as a bal
anced account of debits and credits. The tables for
individual columns are sector statements of sources
and uses of funds, while the tables for rows sum
marize purchases and sales in markets for individual
transaction categories. Tables of contents are orga
nized to indicate the matrix structure of the timeseries tables as directly as possible. This section de
scribes in more specific terms the organization of
the matrix of accounts and hence of the system it
self. References to the matrix are to the table on
page S.l of the statistical section.
As a device in social accounting, the flow of
funds matrix has the following characteristics:
1. Sectors. The economy is divided into major
groups of transactors, such as households, busi
nesses, and governments. These groups are termed
sectors in flow of funds discussions and consist of
sets of commonly identifiable economic units. The
term sector thus always has an institutional meaning
in these accounts, contrasted with many other
bodies of data, economic models, and analytic dis
cussions in which it sometimes refers to types of
activities, as in the investment sector or the financ
ing sector. Investment and financing are forms of
activity that any institutional group might undertake
and are referred to here as types of transactions (see
item 3 below).17

17. Noncorporate business is an exception to this principle, as
discussed in section 4.

2. Sector uses and sources. A pair of columns,
one for payments (U for uses of funds) and one for
receipts (S for sources of funds), is established for
each sector, and all transactions by the members of
the sector are reflected in one or the other of these
two columns.

3. Financial transaction categories. All pay
ments and receipts of each sector are classified into
standard categories, which constitute the rows of
the matrix. Just as each family, firm, or govern
mental unit is classified entirely in one or another
column, so each individual financial claim—such as
a savings account passbook or a single Treasury
bill—is exclusively in one or another row, and all
transactions in that claim are recorded in that row.
Summation of all uses of funds along a row and
across the sectors of the matrix gives a total of out
lays made to acquire a particular kind of asset,
whereas summation of sources along a row yields a
total of funds raised in that particular manner.
4. Financial sources and uses. Financial claims
are shown in the S column of a sector only to the
extent that members of the sector issued such
claims as liabilities to raise funds. Correspondingly,
transactions in the U columns refer only to dealings
in the claim as an asset. Sale of the claim as an as
set is a negative offset against acquisitions of
claims in the U columns, and debt repayment is an
offset to borrowing in the S columns.
Gold and special drawing rights (SDRs) are
treated as financial assets but not as liabilities. Gold
is a metal widely used as a monetary reserve, but it
is not owed by anyone to the holder, and SDRs are
of the same nature. All transactions in these items
appear as purchases and sales of assets in the U
columns of the matrix.
Net new issues of corporate equities appear as
sources of funds to issuing sectors, but in tables on
outstanding assets and liabilities most corporate
equities, like gold and SDRs, are shown only as as
sets. As residual claims on corporate net worth, eq
uity securities do not constitute specific liabilities of
business that can be stated in the accounts.
5. Financial market summaries. Each purchase
of a claim is always someone else’s sale of that
same claim. Hence, taking the economy as a whole

Section 2: Organization
and including transactions with foreigners, total
funds raised by issuing a particular type of claim
are necessarily equal to total funds used to acquire
that claim as an asset.
Total borrowing then equals total lending in any
type of claim and for any set of claims taken to
gether. Each row or set of rows for financial claims
therefore is a summary of all funds coming into and
going out of a particular financial market or set of
markets.

6. Floats in financial transactions. Because in
many instances a single transaction is not entered
into the books of the buyer and seller on the same
day, there are many discrepancies in the basic ac
counting records of the economy between total as
sets and liabilities outstanding. The result is a cer
tain amount of floating supply of claims as assets or
liabilities that is an exception to item 5 above. In
the flow of funds accounts, the floats that can be
estimated appear in the Discrepancy column of the
matrix (discussed under item 10 below).
7. Nonfinancial transactions. The first ten rows
of the matrix are for each sector a condensed sum
mary of all nonfinancial transactions—payments
and receipts for wages, goods and services, taxes,
and transfers. Current receipts and payments are
netted into a sector total of saving, while purchases
of physical capital are shown separately.
Cumulated across the columns for domestic sec
tors, the row for saving adds to total saving in the
U.S. economy, which is shown in a memo column
(National saving and investment). The physical in
vestment rows add across in a similar way to total
capital formation in the economy.
8. Sector balances—saving and investment. As
an accounting matter, every receipt of funds by an
individual or a sector is reflected in one or more
uses of funds, if only to increase cash balances.18
18. In the flow of funds context the terms source of funds and
use of funds mean no more than the standard terms credit and
debit in double-entry bookkeeping. The sector statement is not a
traditional sources and uses of funds aimed at explaining move
ments in a single item such as working capital, bank reserves,
Treasury cash, or gold and foreign exchange. Any such single
item in the accounts is a concept of funds special to one or an
other activity in the economy. When parallel statements are set
up for all sectors, there is no one concept of funds that can be
useful uniformly across the matrix. Even cash loses its generality
in this setting, because cash of the public is different from cash
of a bank or monetary reserves of a central bank. Hence the flow
of funds statement evolves to a generalized form in which the
funds themselves vanish, and there remains only the balance be
tween total debits and total credits.

For each sector, then, a balance exists (except for
statistical discrepancies) between total sources and
total uses of funds. This balance can be shown in a
variety of ways, but in the matrix presented on
page S. 1 sector-account balances are shown as an
equality between gross saving of each sector and its
gross investment (rows 1 and 4 of the matrix). The
concepts of saving and investment used here for
sectors are the same as those applied to national ag
gregates, and for each sector saving equals invest
ment in the same sense as for the total economy.
For each sector saving equals that sector’s physical
capital formation plus a net financial investment
that measures the sector’s excess of lending to other
sectors over its borrowing from other sectors. At
the national level, similarly, total saving equals
capital formation plus net foreign investment, where
the latter is the excess of lending abroad over bor
rowing from abroad.
With this accounting structure, the particular
types of financial transactions by a sector, both bor
rowing and lending, are subcategories under net fi
nancial investment. The totals shown for financial
sources and uses of funds by a sector in general in
clude financial flows within as well as between sec
tors. It is only in the net of the two totals, where
intrasector flows are canceled out, that the financial
figures become intersector flows. This netting is
carried across to the national total of net financial
investment, where all domestic flows are washed
out and where net financial investment of the econ
omy becomes identically equal to net foreign in
vestment.19
9. Balance of the matrix as a whole. The effect
of the preceding eight items is to produce in the
matrix a severely constrained accounting system
that undertakes to place every transaction of the
economy into direct juxtaposition to its counter
parts, both vertically in sector accounts and hor
izontally in transaction or market-summary ac
counts. Horizontally the matrix is constrained by
the equalities between saving and investment, be
tween total nonfinancial sources and nonfinancial

19. Net foreign investment is measured in flow of funds ac
counts from capital flows (that is, the net of financial flows) in
the international transactions statement, whereas net foreign in
vestment in the national income and product accounts is mea
sured from the current account—exports less imports and net
transfer payments. The difference between the two measures is
the statistical discrepancy in the international statement, shown
in the matrix as the sector discrepancy (row 44) for the rest-ofthe-world sector.

Introduction to Flow of Funds

uses of funds, between net financial investment and
net foreign investment, and between total borrowing
and total lending in each financial market. Verti
cally it is constrained by the equality between sav
ing and investment by each sector and for the econ
omy as a whole.
The upshot of these constraints is that in using
this organization of data as a framework for anal
ysis—construction of models, simulation exercises,
forecasting, or estimation of the data—no one cell
of the matrix can be altered without changing at
least three others: one in another row of the same
sector column, one in another column of the same
row, and at least one other for the second column
and second row.

11. Matrix as capital account. The most general
and most important characteristic of the matrix is
that it constitutes a capital account for the economy
as a whole deconsolidated among a number of insti
tutional sectors. It is a capital account in the sense
that it is a statement of acquisition of assets—both
physical and financial—together with the sources of
funds used to acquire those assets. For each sector
the entry for gross saving is the net sum of internal
sources of funds—a residuum of current receipts
less current outlays—and constitutes in the matrix
an addition to sector net worth plus capital con
sumption reserves. Investment is stated gross of
capital consumption and net of borrowing and is
thus a use of funds consistent with the saving con
cept as a source.
The matrix deconsolidates among sectors the cap
ital account of the national income and product sta
tistics. The nature of the matrix as an expansion of
that capital account into individual sectors, and into
individual financial markets, is central to the con
cept of flow of funds accounting and analysis. The
position of the income and product capital account
in flow of funds is discussed in section 3.
As already mentioned, each table in the sector
and transaction accounts is a statement in timeseries form of one column or one row of the ma
trix. The sector tables are statements of sources and
uses of funds, and the transaction tables cut across
sectors to summarize flows into and out of individ
ual markets. Any one cell of the matrix appears in
both a sector table and a transaction table and is a
link between the two.
This simple matrix organization of flow of funds
tables allows flexibility in grouping of the data for
specific purposes. Flow of funds data lend them
selves to many views of economic activity, and
each view characteristically needs its own summary
structure, with particular items or relationships em
phasized. The matrix itself provides a map from
which more condensed systems can be designed
with explicit indication of where each cell will fall
and with assurance that balance of the accounts as a
whole can be maintained to the degree necessary. □

10. Discrepancies. The matrix includes a Discrep
ancy column and a Discrepancy row to absorb un
accounted entries in transaction rows and sector
columns. The Discrepancy column carries the net
sums of sources of funds less uses of funds across
rows, and the sector-discrepancy row carries corre
spondingly the net sums vertically. Because all ele
ments of the matrix are reflected in both of the two
accounts, they add to identical net totals in the cor
ner of the matrix.
While they are net totals in the matrix, the two
discrepancy accounts can also be viewed as a final
sector and a final transaction account in a matrix
that identically adds to zero in both directions. With
that viewpoint, the statements in item 9 on con
straints take on added generality, since one option
in changing the matrix is to alter sector or transac
tion discrepancies. Indeed, if any single cell within
the matrix is altered without explicit offsetting ad
justments, the three other changes will automati
cally be in a sector discrepancy, a transaction dis
crepancy, and the joint sum of sector-transaction
discrepancies.
All discrepancy entries have the sign of net uses
of funds (the net sum of all sources minus allocated
uses in an account). This is an arbitrary convention;
it happens to be the same as that used in interna
tional transactions statements and the opposite of
the convention in the national income and product
accounts, in which the statistical discrepancy is on
the saving side of the capital account as a net source.

Section 3: Relation to National Income and Product Accounts
As stated earlier, a major purpose of the flow of
funds accounts is to relate developments in financial
markets directly to the nonfinancial activities of the
economy. For that purpose, the nonfinancial econ
omy is taken to be measured by the scope, defini
tions, and data of the U.S. national income and
product accounts (NIPA) published by the Bureau
of Economic Analysis (BEA) in the Department of
Commerce. The capital accounts for individual sec
tors of the economy that are pictured in the matrix
on page S.l are in accounting form essentially a deconsolidation of the single capital account published
by the Commerce Department in, for example, the
July 1979 Survey of Current Business (SCB), as
table A-5, “Gross Saving and Investment,” on
page 25 and as table 5.1 on page 47. The account
appears in quarterly income and product estimates
in SCB as table 15. In that account all financial
claims within the United States are offset against
one another in a national consolidation of all capital
account activity. There is thus no recording of fi
nancial flows within the economy or of financial in
vestment by individual sectors. The consolidation
leaves a measure of net financial investment for the
economy as a whole that is conceptually the same
as net foreign investment—the excess of U.S. lend
ing abroad over U.S. borrowing abroad.
The flow of funds deconsolidation distributes the
national totals of saving and tangible investment

among domestic sectors. It introduces explicit re
cording of financial flows among sectors, detailed
by type of instrument, that indicates the routes—
direct or through intermediaries—by which sectors,
such as households that have excesses of saving
over physical investment, lend to sectors, such as
business and governments that may have an excess
of spending.
The position of the Commerce Department capi
tal account in the flow of funds system is described
in tables 4 and 5, by using data for the year 1972.
These tables refer to Commerce Department catego
ries of saving and investment as presented in The
National Income and Product Accounts of the
United States, 1929-74, published as a supplement
to SCB in 1977.

Saving and investment totals
Table 4 gives the relationship between the national
totals of saving and investment in the two systems
of accounts. Total gross national saving (line D) in
the flow of funds accounts is equal to the Com
merce Department total, with the major exception
that purchases of consumer durables have been
treated in the flow of funds accounts as capital ex
penditures rather than as consumption. This shift
produces a smaller amount of current outlays and a

Table 4. Gross saving and investment, comparison of two systems, 19721
Millions of dollars

Income and
product
accounts

Flow of
funds
accounts

Difference

Source of difference

A. Gross private saving
B. Government surplus
C. Capital grants to the United States

180,354
-3,516
710

303,560
-15,152
—

123,206
-11,636
-710

Rows B, E, and I
Insurance and pension reserves
Omitted in F/F flows

D. Gross national saving

177,548

288,408

110,860

Rows C, E, and I

E. Gross private domestic investment
F. Net foreign investment

188,275
-9,046

299,516
-11,676

111,241
-2,630

Consumer durables
Rows C and J

G. Gross national investment

179,229

287,840

108,611

Rows E and F

H. Statistical discrepancy (D minus G)

-1,681

568

2,249

Item

I. Wage disbursements less accruals
J. Statistical discrepancy in
international transactions

1. F/F = Flow of funds.

329

1,920

Rows I + J
Omitted in F/F flows
F/F row F is based on capital
flows rather than currentaccount balance

Introduction to Flow of Funds

larger amount of saving' in flow of funds accounts
than in NIPA.20
This treatment of consumer durable goods is
based on the consideration that expenditures on
those goods are, in a financing context, closely
similar to those on producers’ durables: (1) a house
hold purchase of durables typically represents an in
vestment in a product that will be useful over a per
iod of several years; (2) consumer durable goods
substitute to a significant degree for related business
capital equipment; and (3) purchases of durable
goods are debt-financed to a large extent. To bring
consumer durable goods into the complex of saving
and investment, a total for household saving is
taken before deduction for these purchases, and
total saving and investment are correspondingly
higher.
A second, smaller difference in the two totals of
saving is in the treatment of allocations to the
United States of special drawing rights when these
are created and distributed by the International
Monetary Fund. Such allocations occurred in 1970,
1971, 1972, and 1978, and in the Commerce ac
counts they appear as capital transfers from abroad,
a component of saving that offsets the rise in SDR
holdings included in net foreign investment. SDR
allocations are reflected in asset holdings in flow of
funds accounts, but they are omitted from flows as
not actually transactions.
Flow of funds accounts have a somewhat differ
ent distribution from Commerce accounts between
private and public saving. This difference arises
from the treatment of government life insurance and
retirement fund activities. In the Commerce ac
counts transactions of government life insurance
and retirement funds with households are treated as
social insurance contributions and transfer payments
in the current account, both part of personal in
come. In flow of funds, however, life insurance
and pension claims by households are established as
part of household assets, and claims of these types
against government funds are treated the same as
claims on private insurance and retirement funds.
Growth of government reserves is thus taken to re
flect not a current surplus in government accounts
but rather an increase in liabilities owed to house
holds. This difference in distribution shifts saving
from governments to households relative to the

Commerce Department accounts but has no effect
on total saving.21
Federal government insurance funds are consoli
dated directly into the flow of funds sector account
for the U.S. government, where net growth in in
surance reserves is deducted from current surplus
and appears as a financial source of funds under lia
bilities. State and local employee retirement funds,
however, are shown as a financial sector separate
from the operating accounts of these governments.
The treatment there is to transfer saving from gen
eral government (in the flow of funds state and lo
cal governments—General funds sector) to house
holds, and to impute a lending from households to
the retirement funds (in the flow of funds state and
local government employee retirement funds sec
tor). The amount of both transactions is measured
by total net growth in the funds’ assets.
An occasional minor difference between the two
totals of national saving occurs when there is an ex
cess of wage accruals over disbursements, usually
arising from retroactive wage rate changes and en
tered in the accounts ex post. While these could be
shown in the household statement as a nonfinancial
source offset by a receivable as a financial use, the
added complication would not contribute to anal
ysis, and such adjustments are merged with the
nonfinancial discrepancy, as shown in table 4, row I
The only difference in table 4 between the mea
sures of gross private domestic investment (row E)
in the two systems is the presence of consumer du
rables in the flow of funds total, as discussed
above. In the total of national investment another
difference arises in the measurement of net foreign
investment. In the Commerce Department accounts,
net foreign investment is measured as the net of
current-account transactions in balance of pay
ments—imports, exports, and unilateral transfers.22
In flow of funds accounts, on the other hand, net
foreign investment is measured as the net of capital
account transactions, the excess of U.S. funds ad
vanced to foreigners in all financial forms over
U.S. funds raised from foreigners. The current and
capital balances are conceptually equal, but in prac
tice they differ by the amount of the statistical dis
crepancy in the international transactions statement,
and this discrepancy is thus part of the difference
between the two measures of gross national invest-

20. The shift includes introducing capital consumption allow
ances for consumer durables but is made without imputing in
come from use of durables to total income or services from dura
bles to product.

21. Government retirement funds here cover government em
ployees and persons covered by railroad retirement. Old-age and
survivors insurance is treated the same in the two accounting
systems; in neither does it give rise to household saving.
22. With the adjustment for SDR allocations mentioned earlier.

Section 3: Relation to NIPA

ment. The capital account balance is used in flow
of funds accounting because it is the foreign coun
terpart (with sign reversed) of the net financial in
vestment that is measured for domestic sectors from
financial transactions and that appears in the model
matrix in table 1 as row 3. Table 4 shows on row J
that this difference in foreign investment totals is
reflected in a difference between the national in
come and product statistical discrepancy and flow
of funds discrepancy between saving and invest
ment (row 44, last column of the matrix, page S.l).23

Distribution of totals among sectors
Table 5 spells out the allocation of national saving
and tangible capital formation among flow of funds
sectors. Part A.l shows the allocation among the
flow of funds of each component of total saving as
published in the national income and product ac
counts (total column). Part A.2 shows the changes
in the total and their distribution in the flow of
funds accounts occasioned by differences in treat
ment of specific transactions. Part A.l is based en
tirely on Commerce Department data underlying the
national income and product accounts, whereas A.2
is based on flow of funds estimates except for con
sumer durables. In A.l a few specific points of al
location should be mentioned. Corporate farms are
in the farm sector, and household capital consump
tion on row 8 is on owner-occupied housing and
nonprofit facilities.
Gross saving of nonfarm nonfinancial corpora
tions in the flow of funds is different from the
Commerce Department nonfinancial corporate total
(for example, SCB, July 1979, page 32) only in
that the flow of funds total includes net receipts
from foreign branches and excludes farm corpora
tions.
The major differences in transaction treatment be
tween the two accounting systems, recorded in part

23. The discrepancy between saving and investment in the
flow of funds is to be distinguished from the nonfinancial dis
crepancy that appears in the matrix, row 1, discrepancy column.
The latter matches total gross saving (the net on current nonfi
nancial transactions), including that of foreign sectors, with total
nonfinancial investment (row 5). In this matching, the foreign
component is the balance of payments current-account balance
(with opposite sign) used as net foreign investment in the Com
merce Department statement. The flow of funds nonfinancial dis
crepancy is thus equal to the Commerce Department statistical
discrepancy, although opposite in sign, plus the excess of wage
accruals over disbursements.

A.2 of table 5, have been discussed in relation to
table 4: consumer durables (row 12), which affects
total saving, and government life insurance and re
tirement funds (rows 13 and 14), which affect only
distribution among sectors. Another adjustment to
allocation of saving is capital-gains dividends of
open-end investment companies, which are treated
in the Commerce Department accounts as a capital
transfer rather than as a dividend component of per
sonal income. In the flow of funds these are a cur
rent-account payment from investment companies to
households in order to avoid using a capital-transfer
account in the system for this one item. Saving is
reallocated accordingly.
Row 18 introduces capital consumption on con
sumer durable goods needed to derive net house
hold and national saving in the flow of funds ac
counts, where such durables are viewed as capital
goods rather than as consumption at the time pur
chased. The estimate is from BEA and like NIPA
capital consumption is straight-line depreciation
stated in current-year prices.
Part B.l of table 5 shows sectoring of totals of
gross private domestic investment by type. Both to
tals and details are estimated by BEA. An impor
tant part of this section is the allocation of residen
tial construction purchases directly to households.
This is the net amount spent by households on
owner-occupied dwellings, including new houses,
mobile homes, condominiums, brokerage fees, and
additions and alterations to those dwellings. The na
tional income and product accounts treat owneroccupied housing as a business activity, in which
owners pay imputed gross rents to themselves and
then as landlords deduct depreciation, property
taxes, mortgage interest, and so forth as business
expenses to yield a residual imputed net rental in
come that is part of personal income.24 The flow of
funds accounts continues these imputations in per
sonal income, consumption, and personal saving,
which are directly from the national income and
product accounts, but in the rest of the household
statement consolidate landlord activities of owner
occupants directly into other activities: (1) capital
consumption allowances are added back to personal
saving to get household gross saving (in table 4,
row 6) because gross saving is a cash-flow concept,
and capital consumption is only a book-entry cost
rather than a cash outlay; (2) capital outlays for
owner-occupied housing appear directly as a house24. Details are shown in, for example, SCB, vol. 59 (July
1979) p. 68, lines 61-68.

30
Table 5. Saving and physical investment in flow of funds accounts, 1972'
Nonfinancial business

Item

House
holds

Total

Non
State
Farm corporate Corporate
and local
nonfarm
A.l

1. Personal saving
2. Undistributed corporate
profits
3. Corporate IVA
4. Corporate capital
consumption adjustment
Capital consumption allowances
5.
Corporate
6.
Noncorporate
7. Wage accruals less disbursement
8. U.S. government surplus
9. State and local government
surplus
10. Capital grants to the United States
11. Gross national saving (NIPA)

22,4832
-6,597

7,427

2,522

-8

2,695

- 165

65,362
40,002
-329
-17,263

410
17,504 7,029

62,195

2,757
-329

13,747
66,874 7,545

111,241

.....................

3,145

.....................

8,491

.....................

1,420

.....................

381
288,408

191,171

76,107

97,560

106,937

15,469

80,776

- 17,263

10,019

710
381

7,545

15,469

.......................
-8,491

-3,145

...................

... - 1,420
80,776

5,256

-20,408

8,599

381
0

18,581

5,256

-20,408

5,842

.....................

106

Allocation of gross private domestic investment in NIPA among F/F sectors

62,006
4,049
36,783
21,174

40,715
4,049
34,866
1,800

664

15,344

4,911

372

664

6272
14,717

6263
4,285

372

116,827
9,442

5,5414 6,629
610

13,364
1,271

86,869
7,561

188,275

46,256 7,903

29,979

99,341

299,516

13,747

Transactions differences between NIPA and F/F affecting saving

4,424

. . .

4,796

Transaction differences between NIPA and F/F affecting investment
111,241

.....................

157,497 7,903

1. NIPA = National income and product accounts published by Department of
Commerce, Bureau of Economic Analysis.
F/F = Flow of funds accounts.
IVA = Inventory valuation adjustment.

15,469
-17,263

13,747
710
177,548

111,241

Not
allocated

Allocation of NIPA saving among F/F sectors

114

B.2
Consumer durables
Mineral rights from U.S. government
Gross private domestic
investment (F/F)

Finance

30,024
-6,597

B.l

20. Residential construction
21.
Mobile homes
22.
One- to four-family structures
23.
Multifamily
24. Nonresidentia! plant and
equipment
25. Change in business inventories
26. Gross private domestic
investment (NIPA)

Federal

49,370

A.2
12. Consumer durables
13. U.S. government and insurance
and pension reserves
14. State and local government
pension reserves
15. Capital gains dividends of
investment companies
16. Less: not allocated
17. Gross national saving (F/F)
18. Depreciation on consumer
durables
19. Net national saving (F/F)
(lines 17-5-6-18)

Government

29,979

912

-912

100,253

-912

2. Includes foreign branch profits.
3. Change in builders’ work in process.
4. Nonprofit organizations.

4,796

Section 3: Relation to NIPA

hold investment expenditure; and (3) mortgage bor
rowing against owner-occupied houses appears in
credit market funds raised by households. This con
solidation of housing activity into the household
sector differs from flow of funds accounts in some
other countries and must be remembered in interna
tional comparisons.
Buying and selling of existing houses within the
household sector is netted out of the outlay total ex
cept for brokerage fees. Such transactions are both
a use and a source of funds in household capital ac
counts and are not included in gross capital forma
tion in GNP. The dollar volume of trading in exist
ing houses has in recent years been much larger
than that for purchases of new houses, however,
and with rising average house prices the in
crease has been accompanied by net growth in
mortgage debt unrelated to new capital formation.
This is a substantial consideration in comparing
household mortgage borrowing with purchases of
new homes.
Business investment in one- to four-family units
represents only changes in work in process on
houses for sale to households and is essentially an
inventory-change component of the total residential
figure. Multifamily construction allocated to house
holds consists of condominium purchases. Farm
residential construction is allocated to farms as pur
chasers, since it is commingled with other farm ex
penditures in financing.
Nonresidential plant and equipment (row 20) is
allocated as a single figure among sectors rather
than separately for construction and producers’ du

rable goods. The household allocation is for plant
and equipment of schools, churches, and other non
profit organizations.
Like existing-house transactions, purchases and
sales of all types of land and existing plant and
equipment are omitted from the sector distribution
of capital outlays, as are transactions in intangibles
such as leaseholds and patents. This omission pro
duces statistical imbalances in the accounts insofar
as there are net transfers among sectors in tangible
assets, and the basis for omitting such transactions
is only the lack of substantial information on the
quantities. In general there is probably a net sale of
land and intangibles by households and noncor
porate business and a net purchase of these assets
by corporate business and finance, causing im
balances of opposite sign in the two sets of sector
statements. There may have been several billion
dollars of such transfers in recent years that are not
in the accounts.
Table 5 carries no allocation of net foreign in
vestment among domestic sectors. Each sector’s net
foreign investment is part of its net financial invest
ment, but not yet entirely identifiable as such. To
complete identification would require allocation of
miscellaneous financial sources and uses of funds in
the balance of payments statements that are occa
sionally sizable but not specified as to nature. Pend
ing further specification of those items, net foreign
investment can be viewed only as the consolidated
total of net financial investment for the United
States, mixed for individual sectors with similar net
investment in domestic claims.
□

Section 4: Definition of Sectors and Transaction Categories
each. Flow of funds publications frequently carry,
as on page S.3, a submatrix for nonbank financial
sectors at the most detailed level available.
The sector structure from the most detailed level
to the broadest groupings used in the sector and
transaction accounts tables is shown in table 6.

Any institutional group for which a complete state
ment of sources and uses of funds can be estimated
in flow of funds accounts may be viewed as a sec
tor for analytic purposes. At the most detailed level
there are about 26 such sectors for which data are
maintained on a continuing basis. The list changes
over time, however, to reflect either new institu
tional arrangements that appear in markets or ex
panded availability of data that allow more detailed
presentation of activities. The concept of a sector is
very general, moreover, and in table presentations
and discussions of the data, the elemental sectors
are often combined into broader sector groupings
that can also be treated as sectors analytically.
The matrix on page S.l simultaneously shows
two levels of sector detail, of which one is a very
broad summary of the accounts into four sector
groups—private domestic nonfinancial, U.S. gov
ernment, finance, and foreign—while the other
breaks private domestic and finance into three parts

Sector definitions
Households consists predominantly of individuals
as consumers, as owners of houses that they live in
and as personal investors in financial claims and in
business activities. Unlike the treatment in some
other countries, this sector does not include directly
the business activities of noncorporate proprietors,
such as investment in producers’ capital goods and
rental real estate or borrowing directly related to
those business activities. Noncorporate business
transactions are set off, in this account structure, in

Table 6. Sector structure
Households
Farm business
Nonfarm noncorporate business
Corporate nonfinancial business
State and local governments—General funds

1 Noncorporate
business

Nonfinancial
business

Private
domestic
nonfinancial

Non
financial

Rest of the world
U.S. government

Federally sponsored credit agencies
Mortgage pools
Monetary authorities
Domestic commercial banks
Domestic affiliates of commercial banks
Foreign banking offices in the United States
Banks in U.S. territories and possessions

Savings and loan associations
Mutual savings banks
Credit unions
Life insurance companies
Other insurance companies
Private pension funds
State and local government
employee retirement funds
Finance companies
Real estate investment trusts
Open-end investment companies
Money market funds
Security brokers and dealers

Commercial
banking

}

Savings

institutions

► Insurance

►

Finance
not elsewhere
classified

►

Private
► nonbank
finance

Finance

Section 4: Definitions

separate business sectors that allow transactions to
appear in concise, related form and allow summary
sector statements for business as a whole, combin
ing corporate and noncorporate forms. While many
business proprietors mix their household and busi
ness financing activities in ways that make separate
statements artificial for single proprietors, the dis
tinction between household and business positions
is generally much sharper for partnerships and large
proprietorships; and in the national totals the pres
ence of larger firms gives substantial meaning to
the separate business sector statements.
The household sector includes as marginal com
ponents of the group all personal trusts and non
profit organizations serving individuals, such as
churches, schools, labor unions, and charitable or
ganizations. Trusts and nonprofit groups are in
cluded mainly because the data for separate state
ments are historically thin and sporadic. Beginning
in 1968 there have been systematic tabulations of
trust assets that have been administered by insured
commercial banks, which include a substantial part
of personal trust assets, and insured-bank totals for
personal trusts and estates are presented as an ap
pendix to Flow of Funds Accounts, 1949-78, page
171. During the 1970s, personal trust assets in
cluded about 6 to 7 percent of total financial assets
of the household sector, including security holdings
in the range of 15 percent of the sector totals. For
security market analysis, personal trusts thus appear
to be a significant element of household invest
ments but far from a dominant one.
Data on nonprofit organizations are thinner and
more scattered. Recent estimates give them about 3
percent of total household financial assets and 5
percent of the sector’s corporate equity holdings for
the mid-1970s.

Farm business covers all farming activities in the
United States including corporate farms. The sector
includes farm credit cooperatives consolidated with
the farms that own them, and it includes farm hous
ing activities. Consumption activities of farmers are
in the household sector. Noncorporate farm income
in the accounts is as defined and measured for na
tional income purposes and it includes imputed in
comes. Except for retained profits of corporate
farms, income is transferred entirely to the house
hold sector and is reflected in household saving.
Owner equity investments in noncorporate farming
enter the farm sector through the transaction ac
count, “equity in noncorporate business.’’
To the extent that farmers commingle household

33
and business activities in their own accounts, this
sector departs somewhat from the principle that all
activities of a unit are to be in a single sector ac
count. Because the corporate component is small,
the farm business sector can be viewed almost as an
activity subaccount of the household sector that iso
lates in a separate statement the investment and
business financing of this aspect of household
activities.
Nonfarm noncorporate business consists of part
nerships and proprietorships in nonfinancial enter
prises, including individuals’ rental activities and
the professions. Like farming, this sector is treated
in the accounts as an activity subaccount of the
household sector: all current income is transferred
to households, net saving is shown as zero, gross
saving is equal to capital consumption allowances,
and all changes in equity capital appear as net in
flows in “proprietors’ net investment.’’
Corporate nonfinancial business comprises all pri
vate corporations not specifically covered in finan
cial sectors or farming. It includes holding com
panies and closed-end investment companies on a
consolidated basis, and also includes real estate
firms. It is identical with the nonfinancial corporate
group shown in Commerce tables except that it ex
cludes farm corporations.
Like the Commerce group this sector covers only
the domestic activities of corporations and does not
represent worldwide financial position or investment
and borrowing by U.S. corporations. Operations of
foreign subsidiaries and foreign branches are re
flected only in the income items for branch profits
received and net dividends paid (which are net of
dividends received from foreign subsidiaries) and in
the asset item for foreign direct investment, which
combines on a net basis all capital-account transac
tions between domestic head offices and foreign op
erations. All of these links are on a cash basis of
payments and receipts and do not, in particular, in
clude the foreign earnings retained abroad that are
part of the international transactions statement as
current-account receipts and capital-account invest
ment outflows.25 U.S. operations of foreign corpo
rations are treated symmetrically: they are included
25. Balance-sheet equities in overseas operations that appear
in statements of assets and liabilities for the sector include the
effect of earnings retained abroad, however, and this difference
in treatment between flows and outstandings causes sizable dif
ferences between the net flows and changes in the outstandings
shown.

Introduction to Flow of Funds

in the sector, their earnings payments to foreign
parents are positive in net dividends or negative in
branch profits received, and all capital-account
flows are netted into a single net financial source of
funds as “foreign direct investment” in the United
States in the corporate business statement.
This treatment of the sector as a domestic consol
idation differs markedly from financial statements
published by corporations, which are almost always
worldwide consolidations. Balance sheet ratios can
differ substantially on the two bases, depending on
whether foreign operations have been financed
heavily by foreign borrowing or by parents’ domes
tic borrowing that has moved abroad through the di
rect investment outflow item. The two types of con
solidation have different purposes, with the
domestic oriented toward a view of the U.S. econ
omy as such and the worldwide focused on the con
dition of companies as such. A national accounting
system is by its nature concerned with activities of
a country and thus incorporates a domestic state
ment for business, as in the form used here.

the Postal Service and the Federal Financing Bank.
Many of these agencies operate lending programs,
and a few have debt outstanding with the public
separate from Treasury securities. The sector does
not include the Federal Reserve System and certain
Treasury monetary accounts that constitute the mon
etary-authority sector, and it does not include a set
of federally sponsored credit agencies described be
low. The sector account is consolidated, and trans
actions and claims among agencies are not shown.
Statistically, this sector has the same coverage of
agencies and the same degree of consolidation as
the Monthly Treasury Statement of Receipts and Ex
penditures (MTS) with the exception of the District
of Columbia, which is included in state and local
governments. With that exception, its credit market
debt is defined by “net borrowing from the public”
as shown in MTS and as tabulated in the Treasury
Bulletin, table FO-1.

State and local governments—general funds com
prises the 50 state governments, their political sub
divisions, and all corporations, enterprises, debt
issuing authorities, and trust funds operated by
these units other than employee retirement funds;
these last are shown separately as a financial sector.
The basic data for the sector are the aggregates in
the U.S. Census Bureau’s Census of Governments.
Rest of the world is as defined in the statement of
international transactions for the United States, and
the data in this sector account are from that state
ment, with capital-account flows classified into
flow of funds financial transaction categories and
with nonfinancial transactions as published in the
national income and product accounts. The sector
discrepancy is conceptually the “statistical discrep
ancy” from the international transactions statement
and differs from that item only by statistical differ
ences between NIPA net foreign investment and the
current-account balance in international transactions
statements.

U.S. government covers, for all years, the agen
cies and funds that are in the government’s unified
budget as of 1969, except the District of Columbia.
Included are the Exchange Stabilization Fund, em
ployee retirement funds, life insurance funds, all
corporations that are wholly or partly owned by the
government, and all “off-budget” activities such as

Federally sponsored credit agencies is a financial
sector consisting of six specialized lending institu
tions that have close legal connections with the
government but that are now excluded from the
government budget accounts as private institutions.
In the flow of funds accounts they are separate
from the government sector for all years. These
agencies finance their lending activities mainly
through issues of their own debt securities, and
such issues have been closely coordinated with
Treasury debt operations. The agencies are as fol
lows:
AGENCY

PRINCIPAL TYPE OF
CREDIT

Federal Home Loan Banks

Loans to savings and loan
associations

Federal Home Loan Mortgage
Corporation

Residential mortgages

Federal National Mortgage
Association

Residential mortgages

Federal Land Banks

Farm mortgages

Federal Intermediate Credit Banks

Short-term farm credit

Banks for Cooperatives

Short-term farm credit

Federally sponsored mortgage pools comprises a
set of arrangements authorized by the Housing Act
of 1968 whereby negotiable securities are issued to
the public against collateral of specific pools of
mortgages that have been insured or guaranteed by
federal agencies. The securities are “pass-through”

Section 4: Definitions

in that holders receive periodically scheduled inter
est and amortization payments related to the pools
as well as any prepayments of principal on the
mortgages in the specific pools. The securities
themselves are guaranteed by the managing agen
cies, which have so far consisted of the Govern
ment National Mortgage Corporation (GNMA), the
Federal Home Loan Mortgage Corporation
(FHLMC), and the Farmers Home Administration
(FmHA). Pooling and security issues have been un
dertaken both by these three agencies and by pri
vate businesses such as mortgage bankers. The
FmHA securities—certificates of beneficial owner
ship—have, since March 1975, been sold only to
the Federal Financing Bank and appear in the tables
as federal government purchases of U.S. govern
ment agency securities.
This sector is unusual in that it consists only of
these arrangements, with mortgages as assets and
pool securities as liabilities, rather than of a set of
institutions in the normal sense. The pooled mort
gages do not appear on any lender balance sheets as
assets, and the securities appear nowhere as institu
tional liabilities. The sector thus exists to fill an im
portant role in matching asset and liability positions
in the system.

Monetary authorities covers the Federal Reserve
System and certain monetary accounts of the Trea
sury: the gold account, the silver account, and an
account constructed to record other currency liabili
ties of the government and the assets behind those
liabilities. The sector is identical with the group of
institutions and accounts for which the “Member
Bank Reserves, Federal Reserve Bank Credit, and
Related Items” table in the Federal Reserve Bulle
tin is a balance sheet. “Factors supplying reserves”
in that table are assets, and “Factors absorbing re
serves” are liabilities. The principal liabilities are
thus bank reserves and currency in circulation, and
the principal assets are U.S. government securities,
gold, bank borrowings from the Federal Reserve,
Federal Reserve float, and Treasury currency—
assets that are backing for the reserve money of the
economy.26
26. The structure of the sector is described in detail in
“Member Bank Reserves and Related Items,’’ Banking and
Monetary Statistics, 1941-1970, Section 10 (Board of Governors
of the Federal Reserve System, 1976), in particular pp. 522-23.
The present form of the published table is somewhat different
from that published before the 1970s, and the flow of funds sec
tor conforms to the present version of the table.

35
Domestic commercial banks comprises all banks
that have head offices in the 50 states. It conforms
in coverage of institutions to the domestic-bank
subset shown in the Federal Reserve’s monthly sta
tistical releases on bank credit. It omits, in particu
lar, U.S. branches of foreign banks, which are in
the “foreign banking offices” sector described be
low. These branches were included in semiannual
statements of condition for banks (call reports)
as a major component of noninsured banks through
June 1978. The presence of foreign branches in the
call reports means that the call report totals that are
benchmarks for this domestic sector are somewhat
lower than published call report totals but are not
identifiable separately. Domestic noninsured banks
are a small group, however, and benchmarks for
this sector are only slightly larger than call reports
published for insured commercial banks.
The sector is on a consolidated basis: all deposit
and loan relationships among domestic commercial
banks have been “washed out.” Interbank items in
general add to different totals as assets and as lia
bilities because of items in transit and classification
variances, however, and the net differences are in
cluded in the sector account as a residual interbankclaim liability.
As with nonfinancial corporations, foreign
branches and foreign subsidiaries of these banks are
not included in the consolidation. These offices are
classified as foreign in international transaction
statements and in the rest-of-the-world sector state
ment and are linked to the domestic bank statement
through a single net interbank liability to foreign.
Domestic subsidiaries, other than Edge Act corpora
tions, are consolidated into the sector; but domestic
parents such as bank holding companies are not.

Domestic affiliates of commercial banks covers
mainly holding-company parents of banks. The data
included for the group are at present limited to spe
cific assets and liabilities related directly to banking
activity: loans to banks, loans purchased from
banks, and commercial paper issued to finance such
activities.

Foreign banking offices in the United States is a
combination of Edge Act and Agreement corpora
tions, U.S. branches of foreign banks, and U.S.
agencies of foreign banks. Edge Act corporations
are subsidiaries of American banks that are mainly
in international banking under specific laws but that
have certain domestic deposit liabilities and money
market positions. Branches and agencies of foreign

Introduction to Flow of Funds

banks operate in the United States under special
bank charters and have substantial domestic posi
tions in business loans and deposits or deposit-like
liabilities, but they are not corporate entities sepa
rate from their parent banks.
These three groups are included in international
transactions statements as banks, and in Federal Re
serve bank credit statistics their credit holdings are
added to domestic bank credit to arrive at total bank
credit. Their deposit-type liabilities are included in
the monetary aggregates published by the Federal
Reserve.
The group does not include U.S. banks that are
subsidiaries of foreign banks. Subsidiary banks are
included in the domestic bank sector described ear
lier.

also excluded; they are in the U.S. government sec
tor account.

Banks in U.S. territories and possessions are
classified as domestic in international transactions
statements and are included in the banking group
here. They are not, however, included in bank
credit totals or the monetary aggregates published
by the Federal Reserve. The group consists of those
currently published by the Federal Deposit Insur
ance Corporation. It includes branches of U.S. and
foreign banks in these areas.
Savings and loan associations are mutual and
stock institutions chartered by states and federal
government to accept share capital and to lend pri
marily in mortgages. The group consists of associa
tions covered in Federal Home Loan Bank Board
statistics, including noninsured associations.
Mutual savings banks are institutions operating
under savings bank charters in 19 states with de
posit insurance from the FDIC. Data for the group
are those published by the National Association of
Mutual Savings Banks.
Credit unions are employee organizations related
to individual firms or agencies that are organized
under state or federal charter to accept share funds
from members and to lend consumer credit to mem
bers. The group consists of all state and federal
credit unions in statistics published by the National
Credit Union Administration.
Life insurance companies are those covered in the
annual Fact Book published by the American
Council of Life Insurance but excluding fraternal
orders. Government life insurance programs are

Other insurance companies are the fire, casualty,
and other companies covered in Best's Aggregates
and Averages, a private publication.

Private pension funds are defined in the annual
statistics on self-administered pension funds pub
lished by the Securities and Exchange Commission.
They include retirement funds of nonprofit organi
zations and multiemployer plans shown in those
data. Their total assets are treated as a holding in
trust for the household sector and are the measure
of a pension reserve liability to households. By this
treatment pension funds have zero saving by defini
tion. The current-account transactions that affect
pension fund assets are imputed to households and
are reflected in personal saving. This money is
then advanced by households to pension funds in
the financial accounts.
State and local government employee retirement
funds are the group of such funds reported in the
Census of Governments. They have the same posi
tion in the accounts as private pension funds, with
zero saving and a liability to households equal to
their assets. A current-account transfer of saving
from governments to households is required to fi
nance this household investment, however, because
in the national income and product accounts the
saving is attributed to governments. This NIPA
treatment is described in section 3.

Finance companies comprises sales finance, con
sumer loan, and commercial finance companies
covered in the Federal Reserve’s five-year survey of
finance companies.27 The group also includes mort
gage companies.

Real estate investment trusts (REITs) are a recent
form of intermediary that, through 1960 legislation,
are exempt from federal corporate income tax pro
vided that they distribute most of their ordinary in
come to shareholders and that most of their invest
ments and gross income are from real estate or
mortgages. They can be either open end or closed
end, but in practice all trusts created have been
closed-end companies. Their investments have been
mainly in construction and development loans, and
27. The 1975 survey was published in the Federal Reserve
Bulletin, vol. 62 (March 1976), pp. 197-207.

Section 4: Definitions
their funds have been raised through diversified pat
terns of bond and share issues, bank loans, and
commercial paper issues.

on page 26 in its relation to the discrepancy trans
action row.

Open-end investment companies (mutual funds)
are the group reported by the Investment Company
Institute, other th^n money market funds and mu
nicipal bond funds. Money market funds are a sepa
rate sector described below, and municipal bond
funds are consolidated into the household sector.
Mutual funds issue shares on a continuing basis and
redeem shares on demand at values based on cur
rent market values of net assets. Their capital is in
vested in a variety of debt and equity instruments,
with wide differences among funds in investment
objectives. Closed-end companies are consolidated
with the nonfinancial corporate business sector.

Transaction categories

Money market funds are a form of mutual fund
that became prominent during the 1970s. They are
open-end investment companies that invest primar
ily in short-term money market claims and offer to
small investors the yields that are available from
commercial paper, large CDs, and other instruments
that typically come in minimum denominations of
$100,000 or more. The group covered is that re
porting to the Investment Company Institute.

Security brokers and dealers data are based on
aggregates published by the Securities and Ex
change Commission in its annual reports and cover
ing all such firms registered with the commission.

Discrepancy, the last column in the matrix, records
the residual excess of total sources over total uses
of funds along transaction rows of the matrix.
These discrepancies have the sign of a net use of
funds. In an accounting sense the discrepancy
column is the last sector account needed to com
plete the matrix. As indicated in the descriptions of
transaction accounts below, many of these discrep
ancies have substantive meaning and are not solely
the result of statistical deficiencies. The discrepancy
for nonfinancial transactions is equal to the statisti
cal discrepancy in the national income and product
accounts (with sign reversed), reflecting the integra
tion of Commerce data into the system discussed in
section 3. Transaction accounts with zero discrepan
cies have residual estimates along the row for trans
actions of some actual sector in the account. Fre
quently the residual is in the household account, but
not always. The discrepancy column is discussed

Transactions in the flow of funds accounts fall into
three major groups: current nonfinancial, capital
nonfinancial, and financial. In addition, there are
several internal entries, subtotals, and transfers be
tween current and capital subaccounts, such as capi
tal consumption charges, current surplus, saving,
investment, corporate profits, and unincorporated
business net income. Many sectors also have a re
sidual discrepancy item—the excess of saving over
investment in the data.
Flow of funds accounting, as sketched in table 1,
is not directly concerned with current transactions—
income, transfers, current spending—except through
the gross-saving measure of excess current receipts
flowing into capital account. Current transactions
yielding that surplus are covered systematically in
the national income and product accounts and are
not repeated or paralleled in this system. Some of
the sector tables include a few current items from
NIPA data as a convenience in analyzing relation
ships with capital account activities, but not on a
systematic basis across sectors. The household
statement includes a derivation of gross saving from
NIPA personal saving, and the statement of nonfi
nancial corporate business includes the distribution
of profits, the position of inventory valuation and
capital consumption adjustments, and the relation of
NIPA gross saving and investment to counterparts
on company books. Statements on the government
sectors include some detail on taxes, purchases, and
transfers, and the rest-of-the-world statement sum
marizes exports, imports, and transfers. The first
table in most flow of funds presentations gives the
NIPA data needed to show the sector distributions
of the gross saving totals published in NIPA.

Capital nonfinancial
Sector gross saving is a direct accounting link be
tween the current account, where it is a residual use
of funds that clears the account, and the capital ac
count, where it enters as a source of capital funds.
Other capital nonfinancial components of the matrix
are the purchases of capital goods that are discussed
in section 3. in their sectoring and their relation to
the national income and product accounts. The flow

Introduction to Flow of Funds

of funds table mentioned earlier that presents sec
toring of gross saving includes as well a sectoral
distribution of capital outlays on a time series basis.

ample, household mortgage assets and liabilities are
entered separately), nor in different types (such as a
deduction of security credit from security hold
ings).28 Certain time-series tables of the accounts
show such deductions, but they are within special
formulations and not part of the general structure of
the accounts.
Net financial investment for each sector is the ex
cess of net acquisitions of financial assets over net
increases in liabilities. It measures net funds ad
vanced by each sector to all other sectors. Net fi
nancial investment for each sector plus the statisti
cal discrepancy for that sector equals the sector net
surplus on all nonfinancial transactions.
Table 7 lists the types of financial claims for
which separate transaction accounts are maintained
in the flow of funds accounts. The items listed are
categories normally shown in the published tables.
Some are sums of subcategories for which accounts
are also maintained; subcategories are indented. The
groupings are those frequently used to summarize
transaction accounts.

Financial
For flow of funds accounting the term “financial
transaction” refers broadly to transfers of claims
among transactors, mainly debt claims—promises to
pay—but also money, which is a claim on the
banking system, and business equities. Common
examples are borrowing, in which the lender’s cash
(or sometimes deposit liability) is exchanged for the
borrower’s debt instrument; security trading, in
which lenders exchange cash and securities; and
debt restructuring, in which a lender accepts a new
debt instrument from a borrower in exchange for an
old one. In each of these cases the balance sheet of
each transactor has offsetting entries among liabili
ties, among assets, or in assets matched against lia
bilities, and the transaction is entirely in financial
accounts.
Beyond these are transactions in which a finan
cial claim is transferred in exchange for goods and
services. Such a transfer occurs, for example, when
the seller of a house accepts a mortgage from the
buyer in exchange for title to the house, when de
ferred payment for a purchase results in a tradecredit claim of the seller on a buyer, or finally
when a purchase is immediately paid in cash. These
transactions also affect balance sheets, but the off
sets to the changes in claims are in physical assets
or in net worth if the transaction is in the current
account for the buyer or seller.
A third type of financial transaction included in
the system consists of accrual of claims, which are
bookkeeping entries within single balance sheets
that usually offset current-account entries and thus
net worth. These are not transfers in the plain
sense, but the sector statements contain many of
them in the “miscellaneous” category described be
low. Tax liabilities are a specific accrual item that
appears as a separate category. The reasons for in
cluding accruals in transactions statements are men
tioned in the discussion of those categories.
All financial transactions are entered from the ac
counts in a particular form of net basis: asset sales
by a sector are entered as a negative use of funds—
deductions from purchases of the same kind of
asset—whereas debt repayments are entered under
sources as deductions from new borrowing of the
same type. There are in the matrix no deductions of
liabilities against assets either within a type (for ex

Gold and special drawing rights consists of gold
held as a monetary reserve and SDR holdings.
Transactions in these assets are recorded only for
monetary authorities, the Exchange Stabilization
Fund in the U.S. government sector, and the rest of
the world. All transactions are treated as uses of
funds, and no liability is imputed for not holding
gold or for SDRs. All monetary gold is in assets of
the monetary authorities, while gold held by the
Exchange Stabilization Fund is in U.S. government
assets together with all SDR holdings. As men
tioned in section 3, SDR allocations to the United
States, which began in 1970, are included in asset
holdings but not in flows. Revaluations of the U.S.
dollar also appear as changes in gold and SDR
holdings not reflected in flows. The flow data in
clude only purchases and sales of these assets.
Official foreign exchange position is defined as in
international transactions accounts: convertible for
eign currencies and the net IMF gold tranche posi
tion. This is a liability of the rest of the world and
a net asset distributed between the U.S. government
(Treasury holdings of currencies plus IMF subscrip
tion less IMF notes and letters of credit) and mone
tary authorities (Federal Reserve holdings of curren
cies less certain deposits of the IMF).
28. The one exception to this rule in the matrix is the net In
ternational Monetary Fund position (capital subscription less cer
tain IMF claims on the United States), which is counted in the
U.S. foreign exchanged position as an asset on a net basis.

Section 4: Definitions

Table 7. Financial transaction categories
Gold and special drawing rights
Official foreign exchange position
IMF gold tranche position
Convertible foreign exchange
Treasury currency

Monetary reserves

Demand deposits and currency
Private domestic
U.S. government
Foreign
Time deposits at commercial banks
Savings accounts at savings institutions
Money market fund shares
Federal funds and security repurchase agreements

Deposit claims on
financial institutions

Interbank claims

Life insurance reserves
Pension fund reserves

Corporate equities

}

Insurance and pension
reserves
Corporate equities

U.S. government securities
Treasury issues
Short-term
Other marketable
Savings bonds
Federal agency issues
Loan participation certificates
Sponsored agency issues
Mortgage pool securities
State and local government obligations
Corporate and foreign bonds
Mortgages
Home (one- to four-family) mortgages
Multifamily residential
Commercial
Farm
Consumer credit
Installment
Noninstallment
Bank loans not elsewhere classified
Other loans
Open market paper
Finance company loans to business
U.S. government loans
Sponsored credit agency loans
Loans on insurance policies
Security credit
Owed by brokers and dealers
Owed by others
Taxes payable
Trade credit
Equity in noncorporate business
Miscellaneous
Foreign claims
U.S. government claims
Insurance claims
Unallocated claims

Sector discrepancies

Credit market
instruments

► Other claims

Introduction to Flow of Funds

Treasury currency consists of silver held as mone
tary reserves by the domestic economy and certain
asset-debt relationships between the banking system
and the federal government in connection with the
monetary system—seigniorage on silver, deposits
with the U.S. government for redemption of Fed
eral Reserve Bank notes and national bank notes,
and the liability of the U.S. government in connec
tion with minor coin and United States notes
backed by gold reserves.29 Transaction flows for
this category occur only between the Treasury and
the monetary authorities. Beginning with 1970, this
account also includes SDR certificates as an asset
of monetary authorities and a Treasury liability.
SDR certificates are a domestic claim, separate
from SDRs themselves, through which some of the
Treasury’s SDR holdings become part of the asset
backing for the monetary base.
The large difference between total assets and
total liabilities in the estimates of amounts outstand
ing reflects the fact that gold and silver are shown
in the accounts as assets but not as liabilities (ex
cept seigniorage revaluations on silver, which are
treated as a U.S. government liability). Gold and
silver are treated as tangible assets rather than as
claims.

The totals used here also differ from M-1A in
that they include deposits held by foreign banks and
foreign governments as part of rest-of-world hold
ings, while these are omitted from the M-l that was
introduced in early 1980. The item excludes all
forms of checkable time deposits at banks and sav
ings institutions such as NOW and ATS accounts,
which are included in M-1B but are not in M-l A.
The matrix on page S.l indicates in the discrep
ancy column differences in this category between li
abilities as seen in bank records and assets as re
corded in holder-sector accounts. These differences
are mail float, representing checks in the mail that
are moneys no longer on the books of senders but
not yet on the books of receivers. Mail float relates
to checks that have not yet entered the banking
system’s clearing procedure. It exists in parallel
with and separate from cash items in process of col
lection and Federal Reserve float. Cash items and
Federal Reserve float are deducted from gross de
mand deposits liabilities of banks to consolidate the
bank liability down to an amount owed to non
banks.31 Mail float is a further deduction taken to
arrive at holder records of money balances.
This deduction of mail float is necessary to bring
holder entries for cash into consistent timing with
the other entries in nonbank accounts. It is mainly
an accounting requirement, however, and does not
imply that holder records are analytically more im
portant than bank records. In general the public
looks at the bank record of its deposits as more rel
evant in managing cash than the balance on its own
books. Were it possible statistically to shift timing
of all noncash entries in sector accounts to a basis
consistent with bank records of money supply lia
bility, the entire body of accounts would perhaps be
improved for analysis. Short of this, the deduction
of mail float is necessary.32

Demand deposits and currency covers demand de
posits at commercial banks in the United States,
government and foreign deposits at Federal Reserve
Banks, and U.S. currency outside banks. The con
cepts and estimating methods used to calculate
these monetary liabilities are the same as those in
M-1A data published by the Federal Reserve in
treatment of floats, interbank deposits, and so forth.
There is a substantial statistical difference, how
ever, in that the flow of funds accounts are calcu
lated for the single last day of each quarter rather
than for weekly, monthly, or quarterly averages.
The one-day numbers are needed for consistency
with all other information entering the system and
are frequently very different from the averages be
cause of specific events or even the day of the
week that ends a quarter.30
29. For a detailed discussion of these relationships, see Flow
of Funds in the United States, 1939-1953 (Board of Governors
of the Federal Reserve System, 1955), chap. 17. The form of
Treasury accounting for these relationships has changed over the
years, in particular in the elimination of the gold and silver ac
counts, and the flow of funds scope of the monetary authority
sector has changed. The general principles described these never
theless still hold.
30. A series on ownership of demand deposits was introduced
in Federal Reserve Bulletin, vol. 57 (June 1971). For the rela
tion between those data and flow of funds estimates, see that
Bulletin, table 8, p. 463.

31. The role of these items in the money stock is described in
Federal Reserve Bulletin, vol. 46 (October 1960), pp. 1108-12.
The money stock as published by the Federal Reserve is a bank
ing-system liability record.
32. Statistically, mail float is estimated directly and used in
calculating household cash as a residual. The nature and meaning
of household cash as an “other-party” record are discussed in
George Garvy, “The Float in Flow of Funds Accounts,” in The
Flow of Funds Approach to Social Accounting, Studies in In
come and Wealth, vol. 26 (National Bureau of Economic Re
search, 1962), pp. 431-61.
A further note on the meaning of the bank-record liability: If
all check-writing were to cease for a fortnight and all checks in
the clearance system to reach their final destination, both the
bank gross records of liabilities and holder record of assets
would settle at the level of demand deposits shown in money
supply statistics—that is, net of cash items and Federal Reserve
float. Bank records would come down and holder records would
come up. It is this ultimate view of the present state of balances
plus checks in transit that in general has most meaning to the
public as a cash balance.

Section 4: Definitions

Mail floats are show in the matrix for private do
mestic and for U.S. government deposits. Foreign
deposits are on a bank-record basis in international
transactions statements (and hence here), consistent
in timing with at least the large bulk of capitalaccount transactions.
A mail float in demand deposits implies corres
ponding floats in many if not all other transaction
categories. As a general matter, records of sales
and purchases and of lending and borrowing are not
timed simultaneously, and it is not possible to bal
ance both sector accounts and transaction accounts
without float items. Statistically, most of these
floats cannot be estimated. The largest volume of
transactions generating float is undoubtedly in trade
credit, however, and as noted below, a float exists
in the system for that account.
Time deposits and savings accounts consists of all
time deposits at commercial banks (including nego
tiable certificates of deposit) and all deposit and
share accounts at mutual savings banks, savings and
loan associations, and credit unions. The item in
cludes all checkable savings deposits such as NOW
and ATS accounts. Flows include crediting of inter
est and dividends as well as deposits and with
drawals. Postal Savings System deposits are in the
miscellaneous category, and savings bonds are in
U.S. government securities.
Money market fund shares are the highly liquid
claims that investors hold against money market
mutual funds (q.v.). Payments into and out of these
shares are typically at no cost to the investor, and
withdrawals in particular can be very fast, through
telegraphic transfers and special checking accounts.
The funds require minimum balances and minimum
amounts for withdrawal, but these requirements are
fairly low, such as $2,000 and $500 respectively.
These shares are included in the M-2 introduced by
the Federal Reserve in 1980.

Federal funds and security repurchase agree
ments represents net borrowings by the banking
system and other finance from nonbank lenders. A
large proportion of these loans have one-day matu
rity and they are thus highly liquid investments, but
minimum denominations are typically large, a mil
lion dollars or more. Overnight security RPs are in
cluded in the Federal Reserve’s M-2, and longerterm RPs are in M-3. Federal funds are excluded
from these monetary aggregates altogether as assets
either of deposit institutions or of federally related
agencies. This flow of funds category has broader

scope than the M components in that, while inter
bank lending has been netted out, the quantities do
include federal funds and RP lending by savings in
stitutions and by federally related agencies.
Federal funds loans are unsecured borrowings of
today’s balances at Federal Reserve Banks, and
Federal Reserve regulations limit the range of par
ticipants in the market to banks, savings institu
tions, and federally related financial institutions.
The market is dominated by interbank lending that
is consolidated out in the flow of funds accounts.
The amounts included are thus net borrowings by
banks from nonbanks.
Security repurchase agreements are loans that
from the borrower’s viewpoint consist of a security
sale out of portfolio with an agreement to repur
chase the security at a named future date at a
named higher price. For the lender the loan is col
lateralized by title to the security. The RP market
has a wide range of both borrowers and lenders, but
the claims included here, limited by data availabil
ity, are only those against banks and savings insti
tutions.
The net liability of banks to nonbanks in these
instruments is measured by the excess of their re
ported borrowings over their reported lending to
U.S. banks, and because the data combine federal
funds and RPs the two types of claim are combined
in this category. The implicit net borrowing by
banks is statistically much larger than the amounts
of lending reported by nonbanks in recent years,
and the excess is included in the category as a dis
crepancy. Further development of data sources on
lending can be expected to reduce this discrepancy.
Interbank claims are a set of relationships between
the Federal Reserve and commercial banks, among
the several subsectors of commercial banking, and
between domestic and foreign banks. The item pulls
together several kinds of claim among these sectors
and combines them in forms that are consistent with
the standard statement for monetary authorities,
with a consolidated statement for domestic banking
as a whole, and with an item for net international
capital inflows. The table below lists the types of
claim that are combined into this category and their
treatment for each sector as an asset (A), a negative
asset (-A), a liability (L), or a negative liability (-L).
Claims among the commercial banking subsectors
appear explicitly on a net basis in the subsector
statements but for the group as a whole are washed
out in a process similar to consolidation. For sev
eral reasons, including timing discrepancies, domes
tic interbank claims come to somewhat different to-

Introduction to Flow of Funds

tals statistically as tabulated on the asset and
liability sides of balance sheets, and the net differ
ence is included in this category as an unallocated
net liability of the commercial banking sector. This
item can be interpreted mainly as a float in interbank transactions.

plementary contracts. Borrowing by policyholders
on policies from insurance companies and from
government insurance programs is a positive liabil
ity of households in the “Other loans” category
rather than as a deduction from policy reserves.
Changes in policy dividend accumulations and acci
dent and health reserves are in the miscellaneous
transactions category as liabilities to policy
holders.33

ITEM

Loans to member banks
Federal Reserve float
Member bank reserves
Vault cash

FEDERAL
RESERVE

COMMERCIAL
BANKING

A
A
L
L

L
L
A
A

Due to foreign affiliates
Due from foreign affiliates
Deposits at foreign banks

L
-L
-L

Due to domestic affiliates
Domestic interbank deposits
Domestic interbank loans

L, -L
L, -L
L, -L

Unallocated interbank claims

Total interbank claims

A, L

REST OF
WORLD

A
-A
-A

The amounts due to and due from foreign affili
ates—branches, subsidiaries, and foreign head of
fices—are reported by banks in published state
ments on a net basis, as either an “other” asset or
an “other” liability, depending on the sign of the
balance for all types of claim and all types of affili
ate. Because the totals of the reported items have
no separate significance, they are further netted
here to a liability item for the total banking system,
as in the Federal Reserve’s statistical series on non
deposit sources of bank funds. Statements on inter
national transactions use the separate gross flows of
claims on and liabilities to affiliates for individual
banks as capital outflows and inflows respectively.
For consistency with banking, the rest of the world
sector statement combines these into a single net
capital inflow to the United States and in this re
spect shows smaller total inflows and outflows than
Commerce Department statements on international
transactions. Deposits at unaffiliated foreign banks
are in published statements for domestic banks as a
component of cash balances.
Life insurance reserves are established in the ac
counts as a claim by households as policyholders
against life insurance companies and U.S. govern
ment insurance programs. Statistically, the category
is estimated to be equal to policy reserves against
private and U.S. government life insurance policies,
including individual and group annuities and sup

Pension fund reserves are in the accounts as a
claim of households as beneficiaries against retire
ment programs. They cover private pension plans
(both those administered by insurance companies
and other private plans, and both vested and un
vested plans), government employee retirement
funds, and the Railroad Retirement Fund. They do
not cover the OASI social insurance programs. Sta
tistically, the category is estimated as equal to re
serves of private plans administered by insurance
companies and total assets of other private plans,
government employee retirement funds, and the
Railroad Retirement Fund.34
Corporate equities represent net issues of and trans
actions in equity securities of private domestic cor
porations and U.S. net purchases of stocks of for
eign corporations. The category includes shares
issued by the open-end investment company sector
and covers both common and preferred stock. Fig
ures for asset levels of sector holdings are stated at
market value, and annual changes in levels differ
from net purchases because of fluctuations in mar
ket price. No estimates of liabilities for corporate
stock are attributed to issuing sectors except openend investment companies. These companies differ
from other corporations in that they undertake to re
deem shares on demand at values based on current
values of portfolio assets.

Credit market instruments is a core group of debt
claims that is the principal medium used by nonfi
nancial sectors in raising funds through formal
33. Measurement of life insurance claims is discussed in Fed
eral Reserve Bulletin, vol. 45 (August 1959), p. 837.
34. Treatment of pension funds claims is discussed in Federal
Reserve Bulletin (August 1959), p. 838. With corporate equities
valued at market prices in pension fund assets, year-to-year
changes in stock market prices can cause substantial differences
between movements in total reserves shown as outstandings and
the net flows into reserves shown in flow tables. The net flows
represent premium receipts and investment income less benefit
payments and operating costs.

Section 4: Definitions

credit channels. It excludes trade credit arising in
the normal course of business, tax liabilities, secu
rity credit, and proprietors’ equities in noncorporate
business. It also excludes miscellaneous claims,
which are mainly accruals for private sectors and
various trust deposits for the U.S. government.
Credit market instruments are used by financial
as well as nonfinancial sectors as a source of funds
but to a much smaller extent relative both to bor
rowing in this form by nonfinancial sectors and to
borrowing in other forms by financial sectors. In
the matrix borrowing by financial sectors in credit
markets is included in the credit market rows, but
the principal summary tables on credit flows, dis
cussed in section 1 and illustrated in table 2, focus
on the use of these markets by nonfinancial sectors.

U.S. government securities consist mainly of what
security markets refer to as Treasury and federal
agency securities. Following market usage, the cat
egory includes issues by sponsored credit agencies
that are not direct liabilities of the U.S. government
and not guaranteed by the government. The group
also includes mortgage pool securities, which are
federally guaranteed, and several kinds of nonmarketable Treasury debt held by the public. More spe
cifically, the types of security included are the fol
lowing.
1. All Treasury issues, both marketable and nonmarketable,
including savings bonds, foreign series, and foreigncurrency series. Excluded are special issues to government
agencies and trust funds, on the basis that these are intra
sectoral claims to be eliminated in a consolidated sector
statement, but the total includes special issues to spon
sored agencies that arise from earlier periods when such
agencies were part of the government.35
2. Securities issued by agencies of the government other than
Treasury, such as the Tennessee Valley Authority, Postal
Service, and Export-Import Bank.

3. Loan participation certificates issued by the Government
National Mortgage Association, the Export-Import Bank,
and several other agencies. Most of these issues were
floated during the 1960s and at the time were treated in
government accounts as sales of loan assets. Because of
firm government guarantees, however, they are treated
here as government borrowing collateralized by the loans
held.

35. Where maturity detail is shown, “short-term marketable” consists of
all bills, certificates, notes, and bonds due within a year of the date shown,
regardless of original maturity. The amounts also include part of issues due
within two years on a sliding-scale basis. “Other” issues are marketable
issues not classified as short-term and all nonmarketable issues.

43
4. Commodity Credit Corporation certificates of interest and
CCC-guaranteed bank loans. CCC price-support opera
tions for farm commodities have changed form somewhat
over the years, but in the national income and product ac
counts most farm inventories that are placed under CCC
loan programs are treated as federal government purchases
in GNP, and the CCC loans are here treated as federal
borrowing to finance those purchases.

5. Security issues by the sponsored credit agencies sector
(q.v.). As mentioned above, most of these are not guaran
teed by the government, but they are closely associated in
markets and in debt-management operations.
6. Securities issued by the mortgage pool sector (q.v.).
These “pass-through mortgage-backed’’ securities are col
lateralized by pools of mortgages and are further guaran
teed by the Government National Mortgage Association or
the Farmers Home Administration (both government agen
cies) or by the Federal Home Loan Mortgage Corporation
(a sponsored agency). As mentioned in the sector descrip
tion, the securities are separate from the mortgage pools
and are not directly the liability of any other sector. Fed
erally related guarantees are the basis for including them
in U.S. government securities. An important feature of
“ pass-through’’ securities is that all amortization and pre
payment of debt principal on the underlying mortgages is
paid out immediately to security holders. This feature
gives the effect that the maturity of the securities con
forms to the maturity characteristics of mortgages, with an
element of unpredictability caused by early retirement of
mortgages.

For each of these types of securities the amounts
included are those held outside the issuing sector.
The Treasury debt that appears in the accounts is
exactly the Treasury figure for “public debt securi
ties held by the public’’ as published, for example,
in the Treasury Bulletin, table FD-1, where “the
public” includes Federal Reserve Banks. Govern
ment agency debt in the accounts is the amount of
agency securities held by the public that appears in
table FD-1 with the exception of a small amount of
mortgage debt that has been transferred to the mort
gage category. Totals for sponsored credit agencies
are as published in the Treasury Bulletin, table
TSO-5, and for mortgage pools as published in the
Federal Reserve Bulletin, table 1.56, “Mortgage
debt outstanding.” In recent years mortgage pool
securities based on Farmers Home Administration
mortgages have been purchased entirely by the Fed
eral Financing Bank, a U.S. government agency,
and in the accounts this appears as sizable pur
chases of agency securities by the government sec
tor.
Liabilities of the U.S. government not covered
by this category are shown in the following list:

Introduction to Flow of Funds

U.S. GOVERNMENT LIABILITY

TRANSACTION CATEGORY

Special notes issued to the IMF

Negative in official foreign exchange position

Defense Department and Coast Guard housing mortgages

Multifamily residential mortgages

Trust and deposit liabilities
Certain accrued interest (beginning fiscal year 1956)

► Miscellaneous financial

Postal Savings System deposits

Currency items in the public debt
Other Treasury currency liabilities

Treasury currency

Certain accounts payable

Trade debt

State and local government securities cover the
total debt of all state and local government units,
except loans from the U.S. government (which are
in the other loans category) and trade debt. State
and local obligations held by the state and local
government sector are included in both assets and
liabilities of that sector. Both short-term and long
term securities are included, conforming in amount
and maturity division to data shown in the Census
Bureau’s annual Survey of Government Finances.
Interest income from all of these securities is ex
empt from federal income taxation.
Some of the debt in this category is shown as
owed by the nonfinancial corporate business sector.
These are tax-exempt issues by state and local gov
ernment agencies that finance projects for specific
private corporations, mainly to improve pollution
control, and the corporations guarantee payment of
both interest and principal. Although these securi
ties are legally government liabilities, the treatment
here gives more directly the relation of business to
the securities.
Corporate and foreign bonds consist of the
funded debt of U.S. private corporations and for
eign (private, governmental, and international
agency) bonds held in the United States. The do
mestic liability has the coverage reflected^ in the
Securities and Exchange Commission series on
bonds and notes in “Net Change in Outstanding
Corporate Securities’’ as published before 1974. It
thus includes convertible issues until converted into
equities. Conversions appear as debt retirements
and equity issues.
Mortgages consist of loans secured by real estate
collateral and are divided into four categories ac
cording to type of collateral. Home mortgages are
those secured by one- to four-family residential
properties, and as long-term loans they are assumed
to be owed entirely by households; they include

loans on individual units of condominium struc
tures. Business liabilities for home mortgages are
estimated construction loans on work in process,
while the savings and loan association liability is
“loans in process’’ in their balance sheets, which is
an offset against mortgages already in their assets
that have not yet been disbursed. “Commercial’’
mortgages consist of all nonfarm nonresidential
mortgages, including those of nonprofit organiza
tions such as churches and schools. Mortgage data
are statistically the same as those published in the
Federal Reserve Bulletin, table 1.56, “Mortgage
Debt Outstanding.’’

Consumer credit comprises short- and intermedi
ate-term consumer installment and noninstallment
credit and is statistically the same as the consumer
credit series published monthly in the Federal
Reserve Bulletin, table 1.57 and in separate statisti
cal releases.

Bank loans n.e.c. (not elsewhere classified)
covers the following types of bank loans:
1. By the commercial banking sector (in terms of call re
port classifications):

a. Business loans, except open market paper (in the
other loans category)
b. Farm loans, except CCC-guaranteed loans and CCC
certificates of interest (included as a government li
ability in the U.S. government securities category)

c. Loans to foreign banks (loans to domestic commercial
banks are in the interbank claims category
d. Loans to other financial institutions except commercial
paper (in the other loans category)

e. All other loans (mainly to foreign official institutions,
nonprofit organizations, and individuals).
2. By Federal Reserve Banks:

a. Foreign loans on gold
b. Industrial loans.

Section 4: Definitions

Real estate and security loans are excluded en
tirely from bank loans n.e.c. as credit in the flow
of funds mortgage and security credit categories.
Consumer credit is also excluded from this cate
gory.
Both the asset and liability sides of the category
are measured gross of valuation reserves.

Other loans is the final grouping within credit mar
ket instruments and consists of the following types:
1. Directly placed finance company paper.

2. Dealer-placed commercial paper.
3. Bankers acceptances.

4. Finance company loans to business mainly for financ
ing of equipment purchases or to carry inventories of
receivables.
5. Loans from U.S. government (other than mortgages
and trade credit, both included in other financial cate
gories, and cost CCC loans, treated as government
purchases of inventories), 36 such as student loans,
small business loans, and foreign aid loans.
6. Loans other than mortgages by federally sponsored
credit agencies.
7. Policy loans on life insurance policies.

The first three of these types have in common
that they are short-term money market investments
from the lender’s view and are combined in tables
as “open market paper.’’ Sector holdings are
shown only for the combination of the three types
of instrument, in absence of data for further break
down of holdings. It is thus as a combination that
the three instruments constitute a transaction cate
gory.
Security credit consists of loans subject to Federal
Reserve regulation for the purpose of purchasing or
carrying securities. It includes loans to security
dealers from the banking sector and customer debit
and credit balances with brokers and dealers. This
credit is, in the first instance, an indirect form of
supply of funds to credit markets, rather than a
credit market demand for funds. On the main stem
of the relationship, banks finance private security
holdings through direct security loans and loans
covered by broker and dealer credit to customers; in
addition banks finance dealer direct holdings of se
curities. It does not include all loans with security
collateral, many of which are in bank loans n.e.c.

Taxes payable is the excess of taxes accrued from
past operations, as measured in NIPA, over taxes
that have been paid. Both U.S. government and
36. CCC loans to cooperatives for tobacco and CCC storage
facility loans are treated as loans and included in the other loans
category.

45
state and local taxes are included. At present the
item covers only corporate profits taxes, but it
would be useful and relevant to include parallel lia
bilities for personal income, social insurance, and
indirect taxes. Unlike most other financial items in
the accounts, this is not a claim that has been for
mally recognized by both debtors and creditors. Un
til final settlement on a year’s liability, each party
makes his own estimate of the amount involved.
Taxes payable are nevertheless recognized in finan
cial planning by both business and governments and
in business accounting.
Because estimates of the amount of claim can
differ, the discrepancy in this transaction account is
different in concept from the mail floats discussed
above. Statistically, the liability side is estimated
from corporate balance sheets, whereas the asset
side is the excess of Commerce Department esti
mates of accruals over governmental reports of ac
tual receipts. While part of the discrepancy between
the two arises from data problems, a conceptual el
ement remains.
The data discrepancy in taxes appears in the ac
count for the corporate business sector. Algebrai
cally, accruals less payments of profit taxes should
equal the change in the sector’s profits tax liability,
but in the statistics this is not the case. Accruals,
receipts, and balance-sheet liabilities are taken from
independent data sources, derived from separate
tabulations of profit estimates, governmental re
ceipts data, and corporate balance sheets; and there
are inevitable inconsistencies in timing, coverage,
and estimating procedures among the three. In addi
tion there is always some amount of payments or
refunds in tax settlement cases that has not been en
tered into either balance sheets or accrual estimates.
For these reasons the three tax items shown in the
corporate sector table typically do not balance ex
actly in the statistics.

Trade credit is an approach to a book credit cate
gory; it consists broadly of receivables and pay
ables, other than consumer credit, arising from
sales of goods and services. It covers mainly claims
among domestic businesses, but it includes corres
ponding claims to and from foreign buyers and
sellers and claims on governments and nonprofit or
ganizations. The trade credit asset of the U.S. gov
ernment arises from prepayments to business on
items not yet delivered, while its liability includes,
for recent years, similar prepayments by foreign
governments on items not yet delivered by the U.S.
government. These foreign prepayments are in rest-

of-world trade credit. In the flow tables noncor
porate receivables are netted against payables, but
in tables on outstandings they are shown separately.
A large mail float exists between receivables and
payables in trade credit for two reasons: receivables
are recorded by sellers before buyers have received
and recorded amounts payable, and buyers write
down payables when checks are mailed and before
sellers have received them. This float is in the
transaction discrepancy along with statistical incon
sistencies of the estimates.

Equity in noncorporate business represents net
flows of equity funds invested by proprietors in un
incorporated businesses, both farm and nonfarm.
Statistically, however, this business source of funds
is measured as a residual in noncorporate sector
statements, the amount necessary to balance sources
and uses of funds. As a residual its interpretation
depends on the treatment of all other items in the
noncorporate statements, and its meaning as an eq
uity flow is specific to the structure of the state
ments.
As mentioned in the sector descriptions, noncor
porate business sectoring in these accounts is actu
ally a segregation into separate statements of certain
household transactions that are plainly business re
lated. The purpose of the sectoring is to present as
fully as possible statements for nonfinancial busi
ness as a whole, but the segregation cannot include
all flows that have a business element because
many of them are mingled with household financing
and liquidity management. A business bank loan
that is used to finance higher education results sta
tistically in a withdrawal of proprietors’ equity in
this item, while a house mortgage loan to pay for
business construction is reflected as an equity in
flow. Both of these are appropriate representations;
on the other hand noncorporate business liquid as
sets are left mainly in the household sector account,
and any business-related transactions for liquidity
management may produce equity flows in the data
that are more obscure. Because segregation of non
corporate business activities is artificial to some ex
tent, there is no “correct” procedure but rather a
selection of items to segregate that minimizes the
artificiality of both the statement and the residual
net equity item. The two noncorporate statements in
the system are constructed with this purpose.
An important determinant of the meaning of the
proprietors’ net equity flow is the amount of saving
that is shown for the noncorporate business sectors,
since segregation of saving is one of the more arti

Introduction to Flow of Funds
ficial elements of the sectoring. For the annual esti
mates in the present treatment, all net income of
noncorporate business is shown as withdrawn by
proprietors and becomes part of household income,
while net saving (retained income) of the firms is
arbitrarily put at zero. Gross saving, by this device,
becomes identically equal to capital consumption al
lowances. This means that all investment in physi
cal and financial assets by noncorporate sectors be
yond the amount of capital consumption is to be
viewed as financed externally in the accounts. Such
funds as are not raised from credit markets or trade
debt enter the sectors as net equity investment by
proprietors in the household sector. To the extent
that noncorporate business has in fact an identifi
able retained income, this treatment overstates
household saving as a source of funds (by overstat
ing income receipts); but it also overstates house
hold equity flows to business as a use of funds by
the same amount. Discrepancies in household or
other accounts are thus unaffected by the treatment.
For the quarterly estimates, it is assumed that in
come withdrawals and equity inflows are more uni
form over the year than business income and that in
unadjusted quarterly accounts there are positive and
negative retained earnings that add to zero over the
year. In seasonally adjusted accounts, retained earn
ings are zero quarterly as well as annually.37

Miscellaneous financial claims consists of several
forms of specific claims together with a variety of
unallocated sources and uses of funds in sector sta
tistics. Among identified claims the largest types in
recent years have been business direct foreign in
vestment and customers’ claims on insurance com
panies.
Nonbank direct investment by business in foreign
affiliates appears separately in the foreign-claims
section of the tables on miscellaneous claims. As
amounts outstanding such direct investments are as
defined and reported in international investment po
sitions of the United States published annually by
the Commerce Department, and they include both
U.S. investments abroad and foreign investments in
the United States. As flows however, they conform
to NIPA treatment in that they exclude earnings re
tained abroad both from investment income in the
current account and from net foreign investment in
the capital account; only actual net investment pay
ments to foreign affiliates are included in direct in37. These remarks apply to noncorporate farms, but it should
be noted that the farm business sector has a small net saving rep
resenting retained income of corporate firms.

Section 4: Definitions
vestment flows. The effect is that year-to-year
changes in outstanding direct investment positions
have typically been much larger than the net flows
because of growth in equity in retained incomes.
Other specific items in this category are foreign
deposits held by business and government, equity
and deposit claims on U.S. government-related
agencies, and accrual items arising in the course of
insurance business such as dividend accumulations
and accident and health reserves in life insurance,
and prepaid premiums and benefits payable in fire
and casualty insurance.
The unallocated items arise in the course of sec
tor accounting, when known totals of financial
sources or uses of funds are adopted as controls for
the sector’s financial accounts. Any component of
the totals that cannot be attributed to one of the
specific transaction accounts then falls residually
into the unallocated items. As a social-accounting
practice this is arbitrary, since an alternative proce
dure is to leave unknown items in a sector’s dis
crepancy. Treating them as miscellaneous claims,
however, keeps them within the bounds of financial
transactions and sharpens the meaning of most sec
tor discrepancies.
At the simplest level, the principle is illustrated
by sector accounts for commercial banks, life insur
ance companies, savings and loan associations, and
mutual savings banks. There exists an established
universe estimate of the balance sheet and financial
transactions for each of these industries as a
whole.38 For each, the bulk of financial assets and
liabilities is clearly identifiable in terms of flow of
funds transaction types, but for each there is a
minor remainder of assets and liabilities—mainly
income receivable and expenses payable—that is
left unspecified. These claims are generated by the
calculation of income on an accrual basis and must
be included in financial accounts to maintain con
sistency with income statistics. When they are in
cluded, the sector discrepancy for each of the
groups then becomes a measure of the statistical in
consistency between, on the one hand, the body of
the income and product data from which saving and
physical investment are derived and, on the other
hand, the body of balance-sheet data that constitutes
financial accounts. That some of the balance sheet
is of an unknown nature can be approached within
the framework of financial statistics.

38. Each, in fact, is defined operationally in terms of universe
data available.

Unallocated claims for the rest of the world are
only slightly different. Here the control totals are
from international transactions data, and preserving
them maintains the discrepancy in the statement.

Sector discrepancies is the last row of the matrix,
the final transaction account that closes the matrix
vertically. A few sector accounts have no discrepancy
entry because data are lacking to put together inde
pendently estimated totals of saving and investment.
For such sectors—noncorporate business, pension
funds, and most elements of finance n.e.c.—one or
another source or use of funds is derived residually in
the sector account as the amount needed to balance
saving and investment. The effect is to shift whatever
discrepancy actually exists in the sector’s column of
data into some other account—first, in the transac
tion-account row that the residual is taken in and
then, perhaps, into another sector through further re
siduals. In any social-accounting system, the designer
in effect chooses where to show discrepancies or in
deed whether to show them at all. For this and other
reasons there may be a low correlation between actual
data errors and discrepancies as recorded in the sys
tem.
For the sectors mentioned in the discussion of
unallocated claims, sector discrepancies represent
inconsistencies among a few major bodies of data
for the sector. For governments and nonlife insur
ance, discrepancies are more complex because to
tals of financial sources and uses were built up for
these sectors from identifiable components rather
than broken down from clearly demarked totals
with unallocated residuals.
The discrepancy in the household sector is the
most complex in the system and in general the larg
est. Statistically, every transaction of households is
a residual, since all items in the account are derived
from the books of other sectors, including wages
and personal taxes. The household discrepancy is
thus a final resting place for data inconsistencies
throughout the system. Because a good deal of the
data in the system becomes available as coherent
sector information—for example, balance sheets of
financial institutions—data inconsistencies are to a
large extent between sector columns of the structure
and consist of such problems as differences between
borrower and lender records on the timing of credit
flows. For transaction types in which households are
either borrowers or lenders, most of these inconsisten
cies are carried directly into the residual household
estimates and thence to the household discrepancy. □

Section 5: Publications of Flow of Funds Data
The principal publication of current data for the
flow of funds accounts consists of quarterly tables
of both seasonally adjusted and unadjusted flows.
These current tables are extensions for up to five or
six quarters of base data that are produced each
year by a review and revision process. The base
data are published as tables of year-total flows from
1946 to the present and year-end outstanding claims
from 1945 to date. The base data exist in quarterly
form from 1952, and these quarterly data are made
available to the public as computer data tapes; they
are not published as quarterly tables because of the
amount of paper required. As an alternative to the
computer tapes, computer printouts are supplied on
request for quarterly data in selected sections of the
accounts that are of particular interest to individual
users.
The current quarterly tables are a separate publi
cation of the Board of Governors that is available
on request, and a mailing list is maintained for
quarterly distribution. Complete historical tables of
year-total flows and year-end outstandings, begin
ning in the late 1940s, are published approximately
every three years and are for sale by the Board of
Governors. In both types of publications the tables
include a full set of sector statements of saving and
investment and a full set of transaction account ta
bles that give net borrowing and lending in individ
ual types of claims. They also include the two sum
mary financial tables described in section 1, as well
as a table showing the sectoral distribution of data
from the national income and product accounts used
in the system. Estimates of outstanding assets and
liabilities are maintained on a quarterly basis and
are on the computer data tapes; but they are printed
only in year-end amounts in the annual publica
tions, where they parallel the flow tables in cover

age of summaries, sectors, and transaction types.
The two summary tables of flows appear monthly
in the Federal Reserve Bulletin.

Current quarterly tables become available about
six weeks after the end of the most current quarter
included in the data, but the data for that most re
cent quarter are preliminary and tentative. Each is
sue of the quarterly tables includes revisions for the
next to last quarter that result from the large
amount of data that has become available since the
quarter’s first preliminary tabulation. The tables
may also include revisions in earlier quarters of the
current calendar year to conform to revisions in
source data resulting from new benchmarks such as
the quarterly call report for commercial banks.
On a current quarterly basis, however, revisions are
not carried back to earlier years that have already
appeared in the annual publication mentioned previ
ously, even when source data are revised for sev
eral preceding years. Such longer-run revisions are
postponed until the annual revision.
Annual revisions are intended to introduce all of
the new information that has become available in
the preceding year. The revision includes as a rou
tine matter new benchmarks for earlier years or—as
in the case of NIPA—revisions of source data based
on new benchmarks; but it can also include shifts to
new data sources, changes in derivation methods,
improvements in table formats, or even changes in
sectoring and transaction categories. Such changes
can affect the accounts back to the earliest years
covered, even when no new data have appeared for
those periods. Annual revisions thus have more po
tential scope in the flow of funds accounts than in
the national income and product accounts, for
which revisions are usually limited to three years. □

Section 6: Data Sources
While a full derivation statement is outside the
scope of this publication, it is possible and useful to
list the principal bodies of data that go into flow of
funds calculations at present and to indicate briefly
the schedules on which these data are available.39
The summary list in table 8 omits many peripheral
and occasional sources of information, but it
sketches the statistical skeleton of the system. For
some areas both benchmark and current sources are
listed; where no such distinction is shown, the
sources used for current information are not subject
to revision except as indicated.
These source descriptions are relevant only to the
last two or three years as a group. Some of the
sources are available for most of the period since
1945 covered by the accounts, but others are very
recent. Statistical sources change continually in
both scope and detail, and one of the problems in
establishing a long continuous history is in transfer
ring procedures from older to newer sources. A de
scription of those transfers would go beyond the
scope of table 8 and of this section.
Table 8 has two conspicuous omissions. The first
is in sources of nonfinancial data such as saving,
corporate cash flows, government surplus, and capi
tal outlays. These data come directly from the
Commerce Department’s national income and prod
uct accounts, and while the sectoring in flow of
funds requires detail that does not appear in the
published NIPA, the Commerce Department makes
all the necessary breakdowns and supplies them to
the Federal Reserve. For the first preliminary calcu
lation of a quarter, the 15-day NIPA estimates are
used. These figures omit corporate profits and profit
tax accruals, which must therefore be estimated by
the Federal Reserve for that first run.
The other major omission from the table is the
household sector. Data for this sector are almost en
tirely residuals from the rest of the calculation in
that virtually all of the transactions and balances are
measured from reports by other parties to their
transactions. (The exceptions consist only of mort
gage and trade debt liabilities of nonprofit organiza
tions included in the sector.) The residual status of
the sector means that a listing of data sources for
39. The most recent description of methods of calculation is
“Flow of Funds Accounts—Data Sources and Derivations’’
(October 1971). This publication is available on request from the
Board of Governors of the Federal Reserve System, Flow of
Funds Section, Washington, D.C. 20551.

all other sectors together is implicitly a listing of
sources for households. Some of the chains of rela
tionships are extremely long, but a discussion of
their nature belongs in a description of the deriva
tion methods rather than in a listing of sources.
None of the inputs listed in table 8 are compiled
explicitly or exclusively for the flow of funds ac
counts, although the needs of the accounts have
been one consideration in the design of many of the
reporting forms. Rather, flow of funds accounting
consists of absorbing and digesting a wide variety
of financial information, both flows and balances,
each part of which has been constructed in isolation
from others with its own accounting procedures,
timing classifications, and institutional coverage.
The digestion process is intended to standardize the
accounting as far as possible, so that a transaction
in a financial claim appears consistently in the
seller’s and buyer’s statements in the same transac
tion category, at the same value, and in the same
period.
Problems of consistency can be illustrated in cer
tain areas. One is in federally related securities—
over $200 billion outstanding at the end of 1978—
which are known collectively as the agency-issue
market. These issues receive widely varying treat
ment in holders’ balance sheets and are frequently
combined with bonds of the World Bank and the
Inter-American Development Bank that are unre
lated to the government in any direct sense. The
problem was exacerbated by the growth of federally
related mortgage pool securities during the 1970s
and, in the later years of the decade, by private
mortgage pool securities that are similar in form.
These securities are, on the one hand, not the liabil
ity of any specific issuing institution and, on the
other, are frequently treated by lenders directly as
mortgages. As a result there is statistical uncertainty
in tracking both the volume outstanding and their
ownership.
In consistency of timing, there are major prob
lems in commercial banking, where certain balance
sheet items are highly volatile on a day-to-day
basis, including money supply liabilities. The dat
ing of bank balance sheets is frequently not coinci
dent with those of the other parties to the transac
tions, and the differences can generate sizable
discrepancies in the data.
These are a few illustrations of the statistical
problems that arise in combining a variety of sepa-

Introduction to Flow of Funds

rate accounting systems into an integrated structure
that matches payments and receipts throughout the
economy. The process of adjustment unavoidably
produces sector and market statements that differ to
varying extents from the conventional statements
used by specialists in particular financial activities.
This is part of the price of constructing the broader
system. One direction for development of the sys

tem is a deepening of the detail in financial ac
counts in ways that show continuously the relation
ships to other presentations of the same
information. However, the principal uses of the
data are in studies of intersectoral and intermarket
relationships, and for these the standard categories
of the accounts are unavoidable even when they are
somewhat unfamiliar to individual activities.

Table 8. Sources of financial data for flow of funds accounts
Area

Source

Availability

Federal government

Monthly Treasury statement of
receipts and expenditures,
Treasury Bulletin

Monthly, about 30 days after
month-end

Federally sponsored credit agencies

Statements of condition for
five groups

Monthly, quarterly, or semiannual;
25 days

Census Bureau, Governmental
Finances
Gross offerings of securities,
Public Securities Association
Banking data

Annual, about 17 months lag

State and local governments
Benchmark
Current

Monthly, 30 days

Treasury Bulletin

Weekly, 10 days revised by
quarterly call reports
Monthly, 60 days

Call reports
Several weekly and monthly
reporting systems

Quarterly, about 4 months,
10 to 20 days after period end,
revised quarterly

Savings and loan associations

Federal Home Loan Bank Board

Monthly reports, 25 days
Seviannual call reports
Semiannual surveys of
liquid assets

Mutual savings banks

National Association of Mutual
Savings Banks
Call reports

Monthly reports, 45 days

National Credit Union
Administration

Monthly reports, 24 days

Life insurance
Benchmark
Current

Life Insurance Fact Book
American Council of Life Insurance

Annual, 8 months
Monthly, 50 days, revised
after 12 months

Private pension funds

SEC Statistical Bulletin

Quarterly, 10 weeks, revisions
annually

Census Bureau, Governmental
Finances

Annual, 12 months

Census Bureau

Quarterly, 10 weeks

Best's Aggregates & Averages
Treasury Bulletin
SEC

Annual, 9 months
Monthly, 60 days
Quarterly, 10 weeks

Commercial banking
Benchmark
Current

Credit unions

State and local government retirement
systems
Benchmark

Current
Other insurance
Benchmark
Current

Semiannual, 8 months

Section 6: Data Sources

Source

Area
Finance companies
Benchmark
Current
Real estate investment trusts

Security brokers and dealers
Benchmark
Current

Investment companies, money
market funds
Rest of world
Benchmark

Current

Availability

Federal Reserve, survey of
finance companies
Federal Reserve monthly survey

Quinquennial, 28 months

National Association of Real Estate
Investment Trusts

Quarterly, 3 months

SEC Annual Report
Banking data

Annual, 12 months
Weekly, 10 days and monthly,
20 days

Investment Company Institute

Quarterly, 6 weeks and
monthly, 30 days

Monthly, 35 days

BEA (Commerce Department), International Annual, 6 to 9 months
Transactions, and International
Investment Position

BEA, International Transactions
Treasury Department
Banking data

Quarterly, 10 weeks
Monthly, 6 weeks
Weekly, 10 days

FTC, Quarterly Financial Report

Quarterly, 10 weeks

Financial reports to regulatory
commissions
FTC, mail surveys, regulatory
commissions

Annual, 1 to 2 years

IRS, Statistics of Income
Averages for other industries

Annual, 3 to 4 years

IRS, Business Tax Returns;
Partnership Tax Returns;
Corporate Income Tax Returns
Census Bureau, Census of Housing
Banking data (q.v.)
Corporate data (q.v.)
Consumer credit statistics

Annual or biennial, 3 to 4 years

USDA, Balance Sheet of the Farming
Sector
Banking data (q.v.)
Federally sponsored agency data (q.v.)
Mortgage data (q.v.)

Annual, 9 months

Mortgages

Federal Reserve

Quarterly, with preliminary in 30
days; revisions from institu
tional data

Consumer credit

Federal Reserve

Monthly, 35 days, revised at
various intervals

Open-market paper

Federal Reserve Bank of New York

Monthly, 20 days, revised occasionally

Nonfinancial corporate business
Manufacturing, mining, trade
Regulated industries
Benchmark

Current

Other industries
Benchmark
Current

Nonfarm noncorporate business
Benchmark

Current

Farm business
Benchmark

Current

Approximately decennial, 2 years

Bibliography
Data presentations in print, 1979
Board of Governors of the Federal Reserve System. Flow of Funds Accounts, 1949-1978. Washington:
Board of Governors, 1979. This contains the annual total flows and year-end assets and liabilities.
___________ “Flow of Funds, Seasonally Adjusted and Unadjusted.” Processed. Washington: Board of
Governors, Division of Research and Statistics, Flow of Funds Section. This contains the current quar
terly data, available six weeks after the end of the quarter.

Earlier forms of accounts
“A Quarterly Presentation of Flow of Funds, Saving, and Investment,” Federal Reserve Bulletin, vol. 45
(August 1959), pp. 828-59.
Board of Governors of the Federal Reserve System. Flow of Funds in the United States, 1939-1953. Wash
ington: Board of Governors, 1955.
Copeland, Morris A. A Study of Moneyflows in the United States. New York: National Bureau of Economic
Research, 1952.
“Flow of Funds Seasonally Adjusted,” Federal Reserve Bulletin, vol. 48 (November 1962), pp. 1393-1407

International presentations
Organisation for Economic Co-operation and Development. OECD Financial Statistics. Annual. Volume I
contains statistics of most OECD member countries, volume II extensive notes on country definitions
and data sources.
United Nations. Yearbook of National Accounts Statistics. Annual. New York. Includes definitions.

Other
Bain, A. D. “Surveys in Applied Economics: Flow of Funds Analysis,” Economic Journal, vol. 83
(December 1973), pp. 1055-93.
Cohen, Jacob. “Copeland’s Moneyflows after Twenty-Five Years: A Survey,” Journal of Economic Litera
ture, vol. 10 (March 1972), pp. 1-25.
Dorrance, Graeme S. National Monetary and Financial Analysis. New York: St. Martin’s Press, 1978.
Earley, James S., Robert J. Parsons, and Fred A. Thompson. Money, Credit, and Expenditure: A Sources
and Uses of Funds Approach. Bulletin 1976-3. New York: New York University Graduate School of
Business Administration, 1976.
Freund, William C., and Edward D. Zinbarg. “Application of Flow of Funds to Interest-Rate Forecasting,”
Journal of Finance, vol. 18 (May 1963), pp. 231-48.
Goldsmith, Raymond W. The Flow of Capital Funds in the Postwar Economy. New York: National Bureau
of Economic Research, 1965.
Hendershott, Patrie H. Understanding Capital Markets. Volume 1. A Flow-of-Funds Financial Model: Esti
mation and Application to Financial Policies and Reform. Lexington, Mass.: Lexington Books, 1977.
Mendelson, Morris. “A Structure of Moneyflows,” Journal of the American Statistical Association, vol. 50
(March 1955), pp. 72-92.
National Bureau of Economic Research. The Flow of Funds Approach to Social Accounting: Appraisal,
Analysis, and Applications. Studies in Income and Wealth, vol. 26. Princeton University Press, 1962.

Other—continued
Ritter, Lawrence S. The Flow of Funds Accounts: A Framework for Financial Analysis. Bulletin 52. New
York: New York University Graduate School of Business, 1968.
Robinson, Roland I., and Dwayne Wrightsman. Financial Markets, 2d ed. New York: McGraw Hill, 1980.
Taylor, Stephen P. “Revision Experience in Flow of Funds Accounts.” GNP Data Improvement Project.
Processed, December 12, 1973.
___________ “Uses of Flow of Funds Accounts in the Federal Reserve System,” Journal of Finance, vol. 18
(May 1963), pp. 249-58.

Billions of dollars
Private domestic nonfinancial sectors

Sector
Transaction
category

Households

Gross saving..........................................
2 Capital consumption........................
3 Net saving (1-2)..............................

State
and local
governments

Business

338.3
181.0
157.3

Total

195.6
172.7
22.9

7.6

380.1
298.2
200.3
92.0
5.9

174.1
249.9

11 Net financial investment (12-13). . .

81.9

-75.7

12 Financial uses........................................
13 Financial sources.................................

248J

84.6

25.1

23.5

-34.9

23.5

-34.9

.
15.0

5.4

8.4

-.2

358.0

55.7

28.9

400.8

46.7

13.3

141.2

199.6

843.4

-.3
5.9

43.0

351.4

18.2
18.2

20
21
22

Time and savings accounts...............
At commercial banks.....................
At savings institutions..................

105.2
44.1
61.1

23
24

Fed funds and security RPs............
Money market fund shares...............

6.9

7.5
6.9

25
26
27

Life insurance reserves.....................
Pension fund reserves........................
Net interbank claims........................

12.0
65.8

Corporate equities...........................

-6.2

29
30
31
32
33
34
35
36
37
38

Credit market instruments...............
U.S. Treasury securities...............
Federal agency securities...............
State and local govt, securities. . .
Corporate and foreign bonds . . .
Mortgages.......................................
Consumer credit..............................
Bank loans n.e.c...............................
Open-market paper........................
Other loans....................................

58.0
17.3
9.7
3.3
-1.4
14.5

1.4

.
.

22.5
22.5

.
.

115.2
54.1
61.1

-.2

1.1
1.1

.
.

63.5
-2.6

2.6

104.8
50.6
3.4

14.6

Security credit....................................
Trade credit.......................................
Taxes payable....................................

42
43

Equity in noncorporate business . .
Miscellaneous.................................

-20.8
7.6

44 Sector discrepancies (1-4)..................

-41.8

.
.

1.0

23.6

25.1

3.2
1.7

3.8

40
41

3.2
20.1
43.3

14.6
9.8
2.8
1.0

33.9
5.2
19.9

1.4
1.4

1.1

54.9

18.1
21.4

45.5
3.4

1.0
1.6

-20.8
4.0
7.0

2.6

2.4

-.5

311.8

37.7
28.2

32.8

.1
.1

.
.

28.3
20.1
148.2
50.6
37.3
5.2
22.2

1.6

7.7

53.7
55.1
-1.3

-.4

-.1

4.0

18.3
6.6
3.9

7.9

1.4

47.9
3.4

3.4

-.3

-20.8
25.7

-20.8
5.0

4.7
10.8

20.4

54.9
1.6

-13.3

4.0

10.8
.

13.0

2.7
3.5

2.4

2.6
2.6

28.2
24.8
-.2
3.7

7.8
9.7
-2.0

124.2
65.0
59.2

4.0

20.9
6.9

1.7

4.2

-.3
1.7

135.8

6.3
9.3
.1
-3.1

7.6

1.7
79.8

-1.1

-1.0

.3
.3

1.4

3.6

41.4

44.6
.5
.1

7.5
.9

30.6

2.8
14.6
12.5

-1.2
14.6

41.4
41.4

5.9

7.0
7.7
-.4

o
-.4

5.9

124.2
65.0
59.2

2.6

2.1
6.9

11.5
6.9

20.9
6.9

11.7
58.9

12.0
65.8
14.9

12.0
65.8
15.6

478.0
55.1
40.1
28.3
31.6
149.0
50.6
58.4
26.4
38.6

62.3
5.2

50.0
5.2

1.1

7.5

128.7
-6.5
7.0
9.6
-.3
35.0
26.9
58.4
-1.3

6.9

168.1
5.2
12.3
14.2
31.6
68.3
20.4

31.5

6.7

-2.3

7.3
.9

5.1
10.9

2.8
7.9
12.5

1.8

-1.0
1.4

9.1
-6.3

28.2
24.8
-.2
3.7

59.2

4.8

28.9
25.1
-.2
4.0

59.2

-2.9

.2
.5

7.8
9.7
-2.0

1.3

37.1

.2
.6

65.0
65.0

9.7

11.9

8.2
3.5

19.9

539.0
551.8
200.3
108.0
221.1
22.3

43.0
55.7

191.2

2.3
2.3

1.3

18.5

-3.1
-3.3

551.6
551.8
200.3
108.0
221.1
22.3

22.0
15.5
- .3
6.8

18.8 ■

1.7
-.3

*
*

11.7
58.9
15.6

348.4
6.9
19.1
23.8
31.3
133.9
47.3
58.4
2.2
25.5

9.8
1.4

1.6
.6

9.5

71.4
20.0
13.3
4.5
-1.4
15.5
3.2
16.3

-6.2

-1.6

4.0

9.6
4.2

12.6

46.2

385.7

.3
6.9

5.4

125.6

524.9
359.7
165.2

548.5
359.7
188.7

13.3
3.3
10.0

Demand deposits and currency . . .
Private domestic..............................
Foreign.............................................
U.S. government...........................

-1.2
-7.1
.7
.2

National
saving
and
investment

34
2.7
.7

-2.0

16
17
18
19

162.6

1.0

-34.7

20.7
5.6

2.0

-36.6
-2.0

1.2

5.5

12.7

24.6

8.1
8.1

U
1.0

6.7

Gold and official foreign exchange .
Treasury currency..............................

2.0
2.0

18.4
6.0
12.4

Discrepancy

AU
sectors

Private
nonbank
finance

Commercial
banking

160.4

-1.1
-1.1

Monetary
authority

12.7

16.3
209.3
22.3
2.0

5.4
5.4

Full text of Introduction to Flow of Funds, 1980

FRASER