The conference, cosponsored by the Financial Stability Forum’s Working Group on Capital Flows, was timed to complement an upcoming IMF Executive Board discussion of initiatives to strengthen the IMF’s Special Data Dissemination Standard (SDDS) and General Data Dissemination System (GDDS) in the area of external debt and to provide wide-ranging commentary to the working group, which was finalizing its report on cross-border capital flows, external positions, and the analytical bases for risk and liquidity management.
Participants emphasized the key role that better and more timely data can play in facilitating more knowledgeable market decisions and more effective policymaking. But the exchange of views also highlighted the resource constraints that face already overburdened national statistical offices. The conference represented an important first step in expanding the degree of understanding that compilers and users in the public and private sectors have of each other’s needs. It helped delineate many of the critical issues before the statistical and policymaking communities, and frank exchanges helped clarify the differences in perspectives that shape user and compiler positions.
In his welcoming remarks, IMF Acting Managing Director Stanley Fischer noted that this conference represented a signal opportunity to improve understanding and coordination between data users and compilers. He asked attendees to consider a number of questions, including whether the push for improved data should not also encompass a press for greater resources for statistical operations and whether goals should be more clearly defined in the quest for better debtor and creditor data. On the debtor side, he asked participants to be mindful of the distinction between residents and nonresidents, look at both domestic and foreign debt, and address the question of reserves. From the creditor’s perspective, the question remains how to interpret international banking statistics.
The resources brought to bear on these data issues will be enhanced, Fischer added, if international and national efforts are coordinated and if international organizations complement—not duplicate—their efforts. Good data play a key role in formulating effective policies and avoiding crises, he said, and that lesson was vividly in evidence for the IMF during the Asian crisis, where the absence of timely data allowed debt problems to fester until they reached crisis proportions.
Mario Draghi, Director General of the Italian Ministry of the Treasury and Chair of the Financial Stability Forum’s Working Group on Capital Flows, briefed participants on the perspective his working group brought to these issues. He voiced concern that policymakers and market participants did not spend sufficient time communicating with one another and that data compilers were frequently not fully aware of users’ requirements. Draghi said the conference was a good starting point for improved coordination, but he also pointed to the need for better coordination at the national level. Statistical offices needed to keep pace with the rapidly increasing sophistication of the international financial markets, he cautioned. Progress had been too slow in such central areas as compiling comprehensive data on short-term debt; filling gaps in information on offshore centers, derivatives, and other off-balance-sheet items; and addressing data discrepancies.
The task was admittedly an ambitious one, Draghi noted, but the price of inadequate data—in terms of market uncertainty and crises—was simply too high to be considered a viable alternative.
Why good data matter
One of the critical lessons of the past five years is that in a world of increasingly integrated capital markets, the importance of reliable and timely data—especially on external debt and capital flows—cannot be overestimated. This point was stressed by both Caroline Atkinson, Deputy Assistant Secretary for International Monetary and Financial Policy at the U.S. Treasury, and Nouriel Roubini, Director of the Policy Development and Review Department of the U.S. Treasury, in presentations that proved broadly reflective of the views of a wide range of data users.
Atkinson outlined the case for improved data, arguing that without solid information, policymakers simply operated blind. When there was bad news, countries might instinctively want to hide it, but markets had a way of imagining terrible things in a news vacuum, and this can precipitate a crisis. Greater transparency and better information, she said, form the core of the new financial architecture.
Roubini acknowledged the real constraints many statistical agencies confront but insisted that recent experience indicated that consistent, comparable, and analytically useful measures of external debt could serve as a modern equivalent of the canary in a coal mine—providing early warning of rising dangers and averting more costly problems down the line. He argued, in particular, for enhanced measures of liquidity risk and greater attention to balance sheet risk.
What price better data?
The IMF’s SDDS and GDDS have provided a systematic structure for data dissemination in the global community, according to Paul Cheung, Chief Statistician of Singapore’s Department of Statistics. In a presentation that struck a chord among a number of statistical compilers, he contended that the global community has bought into the concept of data dissemination standards because of their general applicability. But broad support for these initiatives might be jeopardized and lead to nonobservance or even withdrawal, Cheung cautioned, if proposed additions to the SDDS and GDDS were designed too specifically to address the problems of a subset of countries.
Cheung agreed, for example, that external debt issues could pose systemic risks but questioned the relevance of requiring quarterly external debt reporting from net creditor countries. Creditor countries, he explained, would have to divert scarce resources to meet this tight monitoring system and would have to make unnecessarily burdensome demands on respondents.
Annual data, he suggested, would suffice for net creditor countries. Some participants disagreed, noting that the distinction between debtor and creditor countries may be a fluid one and that even major industrial countries are subject to external debt problems.
In addition to proposed improvements in core data, the conference examined whether other types of debtor information could usefully, and practically, be compiled. Paul Tucker, Deputy Director, Financial Stability, Bank of England, focused his remarks on national risk management, suggesting better information and analysis would be needed on financial derivatives, given their growing role, and that debt managers should make greater use of stress tests to gauge the impact of exchange or interest rate changes—for example, on balance sheets. He also suggested that the results of individual stress tests could be aggregated to an experimental measure of systemic stress. But he, too, acknowledged the limited resources available for data initiatives. Often, the data needed depends on the question being asked. For the United Kingdom, service sector data had become a priority, he said. Debtors, on the other hand, must prove themselves to creditors and are in a more natural position to need, and want, to collect external debt data.
Miki Eran, Controller of Foreign Exchange for Israel, outlined her country’s experience, observing that some costs of data initiatives (for example, in exchange rate exposure) were more easily handled in a country such as Israel that had once had a system of exchange rate controls in place. The system set up to administer controls was adapted for new purposes, and the costs involved in new data gathering were kept in line by limiting the number of respondents (only “big players” were required to report), allowing respondents to choose the Internet or whatever reporting system suited their needs, and giving them the software needed to reply.
Credit and market data
One means of closing gaps in external debt data is to make greater use of creditor and market sources of information. Philip Turner, Head of Secretariat, Emerging Markets, BIS, noted that his organization was considering changes on three fronts: measuring exposures on an “ultimate risk” basis (banks already assess their exposure in this fashion), working cooperatively with commercial banks (demonstrating why banks and other users need this data and ensuring that banks’ reporting load remains reasonable), and enhancing the coverage and analytical usefulness of BIS data. Turner explained that the BIS was seeking to improve existing statistics (notably offshore reporting) and stood ready to facilitate the establishment of bilateral contacts between debtors and creditors to examine discrepancies between the two groups’ data.
Guillermo Le Fort, Director of International Affairs for the Central Bank of Chile, argued for a statistical system that differentiates assets and liabilities by their risk characteristics as well as by their functional classification. He underscored the importance of taking external assets into account in the analysis of vulnerability, cautioning that focusing only on liabilities could overstate vulnerabilities.
Participants generally saw value in having some means of frequent monitoring of foreign external market activities and views, but differed on how this might be carried out. James Lau, Executive Director of the Hong Kong Monetary Authority, stressed that one-size solutions and universal templates were inappropriate for this purpose. The experience of Hong Kong SAR, he suggested, argued for some type of customization for small and medium-sized markets. While most countries engage in some sort of high-frequency monitoring, practices differ widely. Developing countries tend to rely more heavily on statistical reporting by market participants, while industrial countries place more reliance on informal contacts to elicit market views and expectations. In a complementary presentation, Dino Kos, Senior Vice President of the U.S. Federal Reserve Bank of New York, reviewed the United States’ long experience with both formal and informal systems of high-frequency monitoring. For the United States, it is considered less important to know who is active in the market than why they are active.
Broadly, the conference delineated two contrasting positions. Policymakers and data users (and a few compilers) argued forcefully for better—that is, fuller and more timely—data on capital flows and external debt. They cited the key role better information plays in shaping more effective policies and in facilitating more appropriate investment decisions. In contrast, compilers from industrial countries and offshore centers typically were more conscious of the multiple demands placed on national statistical resources and were less convinced of the value of providing such detailed and high-frequency data on external debt.
Compilers and users commended the beneficial role played by the SDDS and the GDDS, but expressed less unanimity on proposals to strengthen the external debt component of both. Participants vigorously debated the merits of requiring net creditor countries to submit more frequent debt data, and some questioned more broadly the practicality of a one-size-fits-all approach to the new data proposals.
There was widespread support for the continued improvement of the statistical frameworks being developed by international organizations and for other international cooperative efforts, such as the BIS’s long-standing International Banking Statistics, the IMF’s Coordinated Portfolio Investment Survey, and the Joint BIS-IMF-OECD-World Bank Statistics on External Debt. And there was unquestioned agreement on the value of developing and improving creditor data as well as wide recognition of the usefulness of attempting to reconcile debtor and creditor data as a way of improving national data.
While a number of participants forcefully made the case for exercising restraint in adding new categories of data, there was considerable concern about the increasing importance of nonbank private institutions and the relative absence of data on their activities. Among the ideas to improve the data on nonbank private institutions were proposals to encourage data disclosure for market use, improve transparency, employ new technologies to obtain data faster and more cheaply, and make wider use of existing databases maintained by private firms and regulatory agencies.
Virtually all participants acknowledged that data initiatives would impose growing demands on countries’ statistical resources. Some worried about a diminution of quality; others advised making the best use of limited resources and finding a good solution when the ideal was out of reach.
In her summary, Carol Carson, Director of the IMF’s Statistics Department, stressed the usefulness of “getting views on the table.” The conference had highlighted the importance of perspective. She noted that a consensus between data users and data compilers on priorities for producing better data more quickly had not emerged. Data viewed as crucial at the global level may appear much less vital to national compilers. Resource constraints are real, and hard-pressed data compilers at the national level have yet to be convinced of the need for, and the urgency of, international data requests. The case for improved international data, Carson said, needs to be made more convincingly.
Jack Boorman, Director of the IMF’s Policy Development and Review Department, acknowledged that the conference discussions had revealed “a lot of operational problems” for which quick or easy solutions were not readily available. For many international organizations, the recent crises in Asia and elsewhere—and the stunning absence of key data on a timely basis—had galvanized them to consider remedial action. There was clearly value in forums that brought users and compilers together, enhanced communication, and, ideally, laid the foundation for future consensus and action in these areas. Boorman indicated that since these efforts were not starting from scratch but were rather building on existing systems, progress was likely to come to a large extent through ad hoc improvements. But he was convinced that data collection and dissemination were improving and making a positive difference.
The agenda and background material prepared for the Conference on Capital Flow and Debt Statistics: Can We Get Better Data Faster? will be posted in late March on the IMF’s website (http://www.imf.org).