GFS and Other Macroeconomic Statistics 379
institutional units are aggregated into sectoral bal-
ance sheets, which contain the comprehensive data
for the fi nancial corporations subsectors. At a second
level, the data in the sectoral balance sheets are con-
solidated into surveys.
A7.101 Financial statistics, on the other hand,
consist of a comprehensive set of fl ows and stock
position data on fi nancial assets and liabilities of all
sectors of an economy. Th ese data are organized and
presented in formats designed to show fi nancial fl ows
among the sectors of an economy and corresponding
fi nancial asset and liability positions. Also included in
monetary and fi nancial statistics are the fl ow of funds
data, presented in a matrix format. A detailed fl ow of
funds accounting cross-classifi es fi nancial assets ac-
quired by each sector, by instrument with the coun-
terpart debtor sector. It also cross-classifi es liabilities
incurred by each sector by instrument and coun-
terpart creditor sector. Th erefore, this matrix shows
the fi nancial transactions among all subsectors of an
economy and the rest of the world. Such a presenta-
tion is particularly useful to analyze the allocation of
fi nancial resources and users in an economy.
Linkages between GFS and Monetary and Financial Statistics
A7.102 Linkages between GFS and the monetary
and fi nancial statistics (MFS) result from the fi nancial
relations between government and fi nancial corpora-
tions. As clients (and in addition to holding currency),
governments hold deposit assets with fi nancial corpo-
rations and contract liabilities by borrowing from and
selling debt securities to the corporations. As inves-
tors, governments generally are oft en the sole owner
of public fi nancial corporations or hold equity in other
fi nancial corporations. Th ese fi nancial relationships
result in either a net claim of government on the fi nan-
cial corporations or a net claim of these corporations
on government. Th e net asset/liability position be-
tween the general/central government sector and the
fi nancial corporations sector should be consistent, and
reconcilable in the two datasets. Th e extent to which
these data are similar is oft en a good indicator of the
consistency in macroeconomic statistics in a country.
A7.103 Diff erences in the amounts reported as net
claims between the government sector and the fi nan-
cial corporations sector could be used to check the
accuracy and consistency of the respective datasets.
Where the two sets of data are materially diff erent,
the reasons for the diff erences must be ascertained,
and documentation on the size and reasons for the
discrepancy should be provided to users of the data.
Good statistical practice is for the compilers to inves-
tigate and try to resolve diff erences. Reasons for dif-
ferences can oft en be found in:- Coverage—In many cases, governments have nu-
merous accounts held in several fi nancial institu-
tions. Th e institutional coverage of general/central
government should be the same in both datasets.
A common case exists where certain government
institutions have accounts with fi nancial institu-
tions and MFS cover these accounts, but the ac-
counts of these institutions are not covered in GFS
because these GFS data are confi ned to budgetary
accounts, thereby not covering the data of the ex-
trabudgetary units. Diff erences may also arise if
government has accounts with a fi nancial institu-
tion, but this fi nancial institution is not covered in
the monetary and fi nancial statistics. - Sectorization—Some of the statistical institu-
tional units may not be appropriately and con-
sistently identifi ed and classifi ed as general
government or public sector units or the sector
classifi cation of the subsectors may be diff erent
in the two datasets. For example, an institutional
unit that manages and organizes externally fi -
nanced projects and foreign grants may not be
appropriately designated as a government ac-
count in fi nancial corporations’ records. - Classifi cation and coverage of fi nancial instru-
ments—Th e classifi cation of fi nancial instruments
included in fi nancial assets and liabilities may dif-
fer, or an instrument may not be consistently clas-
sifi ed in the two datasets. For example, diff erences
may arise when an instrument such as accounts
receivable/payable is not treated the same way in
the data, or when a loan is incorrectly reported as
equity investment in one of the datasets. - Time of recording—Complementary periods used
in government accounting may result in transac-
tions being recorded at a time other than when
economic ownership changed hands. - Accrual versus cash recording—Although concep-
tually both datasets should be recorded on an ac-
crual basis, GFS compilers oft en use cash-based