Body
In a 2013 working paper , Aurel Schubert—the head of the Statistical Department at the European Central Bank (ECB)—and his coauthors wrote that: "Growing current account deficits, especially in the United States and some EU countries, have contrasted with large current account surpluses in some other EU countries, in the oil-exporting economies and in many East Asian economies, notably China. The availability of reliable, accurate and timely statistics has been crucial to the assessment of all these developments [author's emphasis]." It is puzzling then that when the ECB recently released a report on the holdings of euro area securities, it shed little to no light on the largest public knowledge gap with respect to assessing the geographical distribution of financial risk in the global economy.
The ECB report, entitled "Who holds what? New information on securities holdings," outlined some of the admirable efforts that it has made towards reporting the holdings of its member states' securities. The ECB now collects security-by-security data on the holders of those assets, which is similar to what is contained in the US Treasury's Treasury International Capital (TIC) survey. The coverage is quite remarkable: For debt and equity securities, the report claims that the survey can account for 92 percent and 73 percent, respectively, of the total outstanding. Moreover, the country of residency of the holders for each of these securities is reported; indeed according to the report, approximately one third of the securities accounted for in the database are held by nonresidents.
The report does not attribute the stock of securities by country despite the data having been collected, however. Instead, it unhelpfully groups certain countries together; the most notable example is the grouping together of Japan and China, two of the largest capital exporters in the world. The ECB cites confidentiality restrictions as the reason for such reporting practices. The rules1 state that attributed data must a) cover at least three "economic agents" or, b) where one or two economic agents make up a sufficiently large portion, the published data should be arranged in such a way to prevent indirect identification. These rules are far too restrictive. While the US Treasury—itself not without blemish regarding transparency—is also subject to confidentiality rules, they are interpreted to simply ensure that any single observation is not made up of a single respondent, allowing for the publication of the indispensable TIC survey.
These data allow the public to know with some degree of confidence who their creditors are. When the sovereign debt crisis in the euro area entered adolescence in 2010, market and policymakers' concerns were exacerbated by an apparent lack of awareness of where the ultimate risk exposure lay. The lack of official reporting of country holdings similar to that reported by the TIC survey in the United States has given rise to a cottage industry of private data firms tracking portfolio flows of large pension and insurance investors as a way to measure cross-border capital flows. Given the considerable returns to scale possessed by the supranational ECB, it is disappointing that it has not made more a serious effort to disclose this data. Many smaller and less well-resourced countries such as Korea, Japan, and Australia already disclose such information on their portfolio investment liabilities.
As for why this information is important, take the example of China's foreign exchange reserves. Another post by this author in 2013 offered a rough estimate of the currency allocation of China's near $4 trillion worth of foreign reserves. The statistics published by each country's statistical department account for only about $2.1 trillion of those reserves. The largest missing piece of the puzzle in tracking down China's reserve allocation is the euro area, the second deepest asset market in the world. Drawing on the data disclosed in the ECB's report, of the €26.9 trillion in debt and equity securities accounted for in the securities holdings statistics survey, €8.5 belonged to nonresidents of the euro area. The TIC data indicate that US residents held about €1.6 trillion.2 That leaves nearly €7 trillion of assets held by nonresidents that are publicly unaccounted for. A nontrivial portion is held by close, non–euro area neighbors such as Denmark, Switzerland, and Sweden. But almost certainly a large portion is held by other large capital exporters, namely China.
This does not excuse China or any other state such as Taiwan from disclosing at least the currency allocation of their foreign exchange reserves to the IMF's COFER survey.3 It is, however, understandable why they would resist revealing such information. For the European Union, as with the United States and other economies with deep domestic capital markets, it is in their interest to promote greater transparency and thus better price discovery and risk management.
The ECB is the single most powerful institution in the euro area—its only truly federalist institution—and in a stand-alone monetary union made up of independent political entities/states, it is arguably the most important institutional guardian of the overall interest of the common currency. The lack of a single democratic overseer of the ECB with the power to sack or politically sanction the central bank moreover bestows upon the ECB a particular responsibility to wield its power in an accountable and transparent manner. The ECB must—in a manner commensurate with its vast institutional powers—therefore lead other EU institutions on transparency issues. Failure to disclose highly relevant country asset data for "confidentiality reasons" that seem rooted in bureaucratic inertia rather than sound judgment indicates that Frankfurt is failing to live up to this responsibility. The ECB should make a more earnest effort in this regard, informing policymakers and the public of the efficacy of greater transparency and urging lawmakers to give the ECB more institutional latitude to do so.
Notes
1. See ECB/1998/NP28 Article 2 for confidentiality procedures. Just to note, an amendment (EU/2015/373) to this law was passed in March 2015, but which does not appear to enhance the reporting latitude of the ECB in this regard.
2. US residents' holdings of euro area assets are reported in US dollars in the TIC survey. We used the end-of-2014 exchange rate of $1.3297 per euro to calculate the amount in euros, though the face value of such assets in euros may be different. Of course the value in dollars of those assets is considerably lower today.
3. COFER stands for Composition of Foreign Exchange Reserves. See related post by Allie Bagnall.