Why do the BIS data on public sector foreign claims differ from the CEBS data on sovereign exposures?

BIS Quarterly Review  | 
13 December 2010

(Extract from page 19 of BIS Quarterly Review, December 2010)

In July 2010, as part of the EU-wide stress testing exercise, the Committee of European Banking Supervisors (CEBS) asked the 91 participating banks to disclose their sovereign exposures to EEA countries. Since then several media outlets have attempted to compare the CEBS stress testing numbers with the figures on public sector foreign claims obtained from the BIS consolidated banking statistics (on an ultimate risk basis). As noted in the CEBS Statement on the disclosure of sovereign exposures in the context of the 2010 EU-wide stress testing exercise,1 the data obtained from these two sources are not directly comparable.

There are several potential sources of variation in the numbers derived from the two datasets. First, the two reporting populations are not the same, as more banks report to the BIS consolidated banking statistics than took part in the CEBS stress testing exercise. Second, in their individual disclosures accompanying the publication of the stress test results, banks were allowed to deduct offsetting short positions (where the immediate counterparty was the same sovereign) from the gross exposures recorded on their trading book. This is generally not the case when banks report their positions for the BIS consolidated banking statistics. Third, the numbers disclosed as part of the CEBS stress testing exercise are on an immediate borrower basis. The BIS consolidated banking statistics contain data on both an immediate borrower basis and an ultimate risk basis, but the figures that are most often referred to in the context of sovereign debt exposures, including all of the public sector foreign claims numbers in this section of the BIS Quarterly Review, are on an ultimate risk basis.2 Fourth, the two datasets also differ in the levels of consolidation that they use in order to assign the holdings of various banking units across national jurisdictions.

1 The CEBS statement is available at www.c-ebs.org/documents/News---Communications/2010/CEBS-2010-194-rev2-(Statement-on-disclosures-of-so.aspx.
2 For a discussion of the reasons for using the BIS consolidated banking statistics on an ultimate risk basis (as opposed to those on an immediate borrower basis) when measuring banking systems' exposures to specific public sectors, see Box 1 ("Measuring banking systems' exposures to particular countries") in the highlights section of the June 2010 BIS Quarterly Review, page 20.