Improving Transparency in External Positions of Oil Exporters
John F. Wilson 1
In recent years the International Monetary Fund has been a vigorous promoter of
increased transparency in all types of economic statistics. Some progress has been recorded,
but various problems remain. Among issues that are relevant to the Middle Eastern
Department (MED) is the practice by the U.S. Treasury Department of concealing capital
flows and asset/liability positions of the United States with certain Middle Eastern countries.
Data are obscured for the following: Bahrain, Iran, Iraq, Kuwait, Oman, Qatar, Saudi
Arabia, and the United Arab Emirates.2 Statistics for these countries are combined inside an
oil exporters group, rather than being displayed bilaterally, as is done for all others. In the
Treasury Bulletin, for instance, the asset/liability positions and capital flows reported with
these countries, by both banks and non-banks in the United States, are merged into an “Other
Asia” group, which masks the bilateral detail. Reference is made to the attached table in the
following discussion about the desirability of achieving greater transparency in these U.S.
The practice of affording special confidentiality in U.S. capital flow statistics vis-à-
vis selected oil producers originated after the 1973 oil crisis. Certain of these producers were
enjoying large surpluses and were loath to have the size of their external assets disclosed.
The United States obliged them in this undertaking. Whatever the merits of the case for
abetting this practice in the mid-1970s, however, surely the time has passed when it has any
justification. With public commitments by the Fund and most national authorities to greater
transparency in economic policy and data, the costs and disadvantages of obscuring routine
statistics for a few oil producers now outweigh the benefits of continuing this practice simply
for reasons of history and inertia.
The U.S. authorities’ practice in this regard is not limited to standard published
sources such as the Treasury Bulletin.3 The United States is also a reporter to the Bank for
International Settlements (BIS) and one of the most important in that system. The essence of
Middle Eastern Department, International Monetary Fund. My appreciation to Rainer
Widera of the BIS for helpful comments and corrections on an earlier version of this note.
The U.S. reporting system compiles individual data for all eight countries, so there is no
question that details are available.
Naturally, all other U.S. sources, such as the Federal Reserve Bulletin, Survey of Current
Business, etc., are constrained by these disclosure limitations.
D:\Docstoc\Working\Pdf\Bcd735c4-3af1-4f3a-B0d7-1472318d177b.Doc September 28, 2011 (3:49 PM)
the BIS framework is bilateral identification of countries with which reporting banks have
claims and liabilities. Other BIS reporters include data for individual countries in the oil
exporter group. No other reporter suppresses bilateral information for specific, pre-defined
countries in its submissions.4 The result is that published BIS-bank-related assets and
liabilities of all eight countries in the oil-exporter group are regularly understated by amounts
of the U.S. omissions. The missing numbers are shown as a residual in the BIS Quarterly
Review, e.g., in Table 6A on “External Positions of Reporting Banks vis-à-vis Individual
So far as BIS reporting is concerned, the Treasury’s motivation for suppressing this
detail is especially unclear, because presentations in BIS publications do not display bilateral
positions in any case. That is, even if full detail were provided by the authorities, U.S.
positions would simply be aggregated with 27 other reporters in deriving individual results
for partner countries in the oil exporter group.5 BIS officials would much prefer to have full
bilateral disclosure by all reporters. Indeed, the subject of incomplete U.S. locational
statistics has been on the agenda several times at BIS biannual meetings of central bank
statisticians (and will be again for the October, 2001 meeting), but so far this has not brought
the desired response.6
There are other contexts in which this practice has adverse or embarrassing
consequences. For instance, in 1997 the Fund sponsored a Coordinated Portfolio Investment
Survey especially to measure bilateral portfolio positions (from the holding/asset side) on a
worldwide basis and thereby assist BOP compilers to do a better job in the difficult area of
measuring portfolio flows. In the published results, U.S. positions with these eight countries
were blank and footnoted with “not disclosed to the IMF for reasons of confidentiality.”7
There are now 28 BIS reporters in the locational statistics, which are published quarterly.
Admittedly, several BIS reporters; including Australia, Denmark, the Netherlands Antilles,
and Singapore do not provide full country detail for all geographic areas. Some of these gaps
may be intrinsic to their compilation systems. For instance, the bank returns in Bahrain
presently omit Finland because positions with that country were de minimus when the returns
were developed. Australia suppresses detail, mostly for smaller countries, but only when the
number of bank-reporters showing positions with those countries is very small.
On the other hand, U.S. authorities do provide to the BIS a full country breakdown for oil
exporting countries in the context of the BIS’s consolidated international banking statistics.
These differ from the more familiar locational statistics (and, in my opinion, are less
reliable). Two of the significant differences are a) the consolidated statistics are not
consistent with balance of payments measurements, and b) they only cover claims, not
liabilities, of reporting banks.
Another Coordinated Portfolio Investment Survey is scheduled for end-2001, and the
number of participants will be significantly higher than in 1997.
As seen in the attached table, which gives results for December, 2000, the numbers
in question are not trivial. Without going into detail about the Treasury International Capital
Reporting System (TIC), U.S. bank and non-bank claims on the oil-exporter group totaled
about $12.5 billion, and their reported liabilities to this group came to about $27.7 billion.
BIS data for end-2000 show reporting bank claims on these countries at about $96.8 billion,
and liabilities at about $171.2 billion.8 That is, the absence of U.S. participation in bilateral
reporting for these countries understates claims on them by about 13 percent of published
BIS amounts, and more than 11 percent of the combined total. Similarly, liabilities to the oil-
exporters are understated by about 16 percent of the BIS figures and 14 percent of the
One obvious consequence of these omissions is that assessments by Fund missions
and vulnerability exercises for these MED oil exporters are distorted by the missing statistics.
Indeed analytic conclusions may be outright misleading. Notably, while bank-related
external liabilities of these countries are understated, it can also be seen that their external
assets are understated by roughly twice as much. That is, proper bilateral attribution in the
U.S. statistics would improve the perceived external net asset position of this group, rather
than worsen it, at least so far as the year 2000 is concerned.9
In addition to burying bilateral positions in geographic aggregates, the U.S.
authorities do the same with transactions (flow) statistics for long-term securities. Data from
the Treasury Bulletin at the bottom of the attached table show that net purchases of such
securities by the countries in question came to about $14.8 billion in the year 2000. Again,
while displaying bilateral detail for U.S. securities transactions is the norm in U.S.
publication policy, no country detail is revealed for this group. Depending on trends through
time, of course, such portfolio flows can cumulate to very sizeable positions, thus
multiplying the distortions in the banking numbers.
The practice of concealing both bank positions and securities transactions for
selected countries not only hinders Fund surveillance of these countries, it also contributes to
distortions in analyzing their balance of payments. For example, over the last decade,
progress has been made in using certain data sources external to a given country—mostly
notably the BIS banking statistics themselves—to help fill gaps in their own BOP
compilations. But the suppression of detail in the U.S. statistics robs this effort of a valuable
component in putting together statements for countries in this group, most of which by any
The BIS totals were derived by simply adding up the bilateral results for each of the eight
countries in the group.
Of course, we cannot here say what might be the quantitative impact on individual members
of the group.
measure have deficient local BOP compilations, which could benefit from the outside
A further consideration is that the position and flow aggregates shown in the
attached table may mask divergent situations among the eight countries in the group, because
individual external positions cannot be discerned from these totals. It cannot be assumed that
each country moves in tandem with the others. Put more dramatically, inside the “other
Asia” aggregate one country might be getting into trouble, while the others are doing just
fine. U.S. non-disclosure practice, however, denies Fund staff and all other analysts any
statistical basis for making judgments on the subject.
In summary, there is a strong case that concealment of these international capital
flow statistics has outlived its usefulness and has become counter-productive. It would
therefore seem appropriate for the Fund to encourage U.S. authorities to re-think their present
policy, and treat these countries in the same open manner as others are presented in U.S.
international statistics. While MED would be a clear beneficiary of this change in treatment,
reform would also serve much broader ends of transparency and candor in general.
As a new U.S. Executive Director is about to take up his position at the Fund, now
would seem to be an opportune time to address this matter. It is therefore suggested that
MED raise this issue with the new U.S. ED and suggest that the Treasury’s disclosure policy
be liberalized. As the Fund’s main defender of international statistical standards, it seems
likely that STA should also interested in joining these representations. Not only are the
analytic grounds for concealment of such data no longer convincing, it may even prove to be
the case that only inertia is keeping the old system going at all. Some small initiative in this
regard may be just what is needed to bring it to a close.10 At the least, pressing for candor in
the U.S. international accounts will give MED a chance to demonstrate its commitment to
promoting transparency, and it will present the U.S. Treasury with a challenge to embrace a
higher level of transparency in its statistical policy.
Note, for instance, that the United Arab Emirates is still annotated as the “Trucial States”
in the Treasury Bulletin, which suggests this system has been on autopilot for some years.
Claims and Liabilities with MED Oil-Exporters in U.S. Statistics
and the BIS Statistics, December 2000 1/
(In billions of U.S. dollars)
A. Claims and Liabilities (positions)
Claims on oil exporter group
Banks (TB) 11.4
Of which: claims of domestic customers 0.2
Nonbanks (TB) 1.1
Total 2/ 12.5
BIS total 3/ 96.8
U.S. total/BIS total (in percent) 12.9
U.S. total/Combined total (in percent) 11.4
Liabilities to oil exporter group
Banks (TB) 24.9
Of which: custody liabilities 5.7
Nonbanks (TB) 2.8
Total 2/ 27.7
BIS total 3/ 171.2
U.S. total/BIS total (in percent) 16.2
U.S. total/Combined total (in percent) 13.9
B. Securities Transactions (flows) (TB)
Net transactions in long-term domestic securities 4/ 14.8
Foreign purchases from U.S. residents 112.0
Foreign sales to U.S. residents 97.3
Sources: U.S. Treasury Bulletin (TB), June, 2001; and BIS, Quarterly Review:
International Banking and Financial Market Developments, June, 2001.
1/ "Oil exporter" group includes Bahrain, Iran, Iraq, Kuwait, Oman, Qatar,
Saudi Arabia and the United Arab Emirates.
2/ Shown as a "residual" for Middle East and Africa in the BIS volume (Table 6A).
3/ Sum of published BIS claims/liabilities with countries shown in footnote 1.
4/ Treasury and Agency securities, and domestic bonds and stocks.