Comment on "International Standards for Strengthening Financial Systems: Can Regional Development Banks Address Developing Country Concerns by Liliana Rojas-Suarez"
Paper for conference on "Financing for Development: Regional Challenges and the Role of Regional Development Banks" Washington, DC
© Peterson Institute for International Economics
Avoiding the Standards Constipation Blues
Since the emerging market crises 1997-1998, the so-called 'international community' has attached increasing importance to the design, agreement, implementation and assessment of standards and codes as a core element of crisis prevention. The Financial Stability Forum (FSF), itself established in April 1999 as part of the effort to strengthen financial systems and improve coordination among the agencies responsible for them, posts on its website the Compendium of Standards (http://www.fsforum.org/Standards/Home.html, last update: June 2001), citing 69 (!) standards. Of these, twelve have been highlighted as key for sound financial systems (see Table 1 in Rojas-Suarez, this volume).
In her excellent paper, Liliana Rojas-Suarez avoids to play the Standards Constipation Blues and focuses on arguably the most widespread of all recent financial standards, the bank capital adequacy standard enshrined in the Basel Capital Accord, to which more than 100 countries claim to adhere (although it was originally designed for internationally active banks in G 10 countries). In my comments, I will follow her script by first discussing general concerns about applying standards to developing countries, then by setting the focus on the Basel Capital Accord in its current 1988 and its proposed new version, finally by providing two policy suggestions.
Standards: Why, How, When
The current international effort to codify best practices and to disseminate them widely should help the seamless integration of local economies into global markets. When a country gets globalized, her institutional, legal and other structures need to move towards international best practices if she wants to provide the appropriate market signals and information in the beauty contest for global capital supply. If designed appropriately, standards confer efficiency-enhancing value by themselves; hence, countries should have a great self-interest in incorporating them while developing institutions and markets. As for bank capital standards, they aim at reducing bank insolvencies to safeguard a country's banking system, immunizing taxpayers in the event of bank insolvencies, and aligning incentives of bank owners and managers with those of uninsured claimants (Rojas-Suarez, 2001).
Whether international standards are appropriate for crisis prevention in developing countries, depends much on how they are designed and 'owned' and when they are implemented and updated. Rojas-Suarez lists and discusses four concerns: a) just what holds for exchange rate regimes (Jeff Frankel's Graham Lecture at Princeton, 1999), holds for standards: no single standard is right for all countries or at all times; b) the sequencing of standards implementation in view of a country's institutional and market development; c) lack of sufficient participation of developing countries in setting the standards ('ownership'); and d) lack of effectiveness. Let me just add that other concerns have been voiced (UN, 2001; Griffith-Jones, 2001; Park, 2000; Persaud, 2000; Reddy, 2001; UNCTAD, 2001):
- standards and codes designed to discipline debtor countries distract attention from the capital supply side which has contributed to the 1997-98 crises and contagion, notably bank credit reversals, and thus slow true progress towards a crisis-resistant global financial architecture,
- ignorance of investors' herding behavior in standards design, notably for market sensitive risk management and transparency, risks to raise rather than reduce the crisis proneness of the world financial system,
- too heavy assignment in standards design and assessment to international financial organizations, introducing a conflict of interest between their assessment and their lending programs and putting in doubt the emphasis on voluntary adoption as incentives and sanctions linked to standard setting risk to become features of surveillance and conditionality.
Rojas-Suarez stresses the fact that, just like for the process of opening the capital account, standards implementation faces the issue of sequencing and capacity constraints. The issue of a poor country's capacity constraints, the proliferation of financial standards listed in the FSF Compendium and the need for prioritization of standards are intimately linked. This makes it important to think about the parameters for prioritization. Rojas-Suarez shows, very convincingly, that the depth of local financial markets and the quality of the legal and institutional frameworks are parameters of outstanding importance if unwanted side-effects are to be avoided and standards to be effective. One can think of further parameters: purely technical standards may be more easily amenable to implementation than those with policy implications, which would require standards implementation to form an integral part of the economic reform process; standards that take an important socio-cultural dimension may pose the most enduring challenge of - credible - implementation. But the trade-off between the tendency of international organizations to proliferate standards and poor countries' capacity constraints will remain. It is urgent to start thinking about opportunity costs in the area of standards and codes (see below for a suggestion).
I wonder whether the concern about 'ownership' has not been overemphasized. Andrew Crockett (2000) has, for example, worried early on about the trade-off between the legitimacy of the FSF (rising with the number of countries and shades of opinion sitting in the three working groups) and its effectiveness. Emerging markets with relevant policy experiences have been part of the process early on and regional groupings have been established.
What I am more worried about is the potential lack of effectiveness of the entire standards and codes process. Argentina shows, in my view and words, Gresham's Law in action, paying debt with ROSCs rather than dollars. As Rojas-Suarez points out, Argentina has been one of the developing countries with the most Reports on the Observance of Standards and Codes (ROSCs) published in the IMF website (http://www.imf.org/external/np/rosc/rosc.asp). Yet it is in deep crisis. In its "Experimental Report on Transparency Practices: Argentina" (http://www.imf.org/external/np/rosc/arg/index.htm), the IMF praised in 1999 (when markets started to send the country's dollar bond spreads higher) the country for 'exceeding' the requirements of the IMF Code of Good Practices on Fiscal Transparency, just when the rating agencies started to downgrade its sovereign debt for fiscal concerns.
The other observation that feeds my concerns about the effectiveness of the process is that private capital flows to developing countries have been declining steadily since the process began. Either investors do not (yet?) pay attention to ROSCs, or no progress has been made. You choose.
A Key Standard: Banks' Capital Requirements
In developing countries, banks remain the most important financial intermediary. Counterproductive effects of bank capital regulation will thus be particularly harmful. Significant changes to the 1988 Basel Capital Accord are currently under discussion and a final version (Basel II) is to be published in 2002 for implementation in 2005.
Rojas-Suarez has shown very convincingly that, unlike in developed countries, bank capital fails to send warning signals ahead of crises, not least for regulatory distortions that encourage bank lending to the public sector. This holds as local-currency public liabilities have been treated with a zero risk weight in the 1988 Accord. This is set to change: The Basel Committee on Bank Supervision is now proposing two main approaches to the calculation of risk weights: a 'standardized' and an 'internal ratings-based' (IRB) approach. Under the standardized approach, , a sovereign with a AAA rating would receive a 0% risk weight; lower ratings translate into a jump in risk weights via 20, 50, 100 up to 150% for sovereigns weighted below B-. The IRB approach is based on mapping risk measures derived from probability of default, loss given default and maturity into risk weight buckets. Representative values for benchmark risk weights as a function of the default probability for corporate exposures are provided in the Committee proposal. Importantly, it does not suggest a linear, but a strongly exponential rise of risk weightings along the spectrum of higher probability of default. This again should lead to much higher risk weights imposed on bank assets against local public authorities.
The paper divides the world into three country groups as for their response to bank capital regulation. This division, which yields many helpful insights, will be overlapped by the strong distinction between investment-grade versus speculative-grade borrowers introduced by the Basel II proposals. Simulation exercises show that speculative-grade borrowers, the bulk of emerging and developing countries, will suffer from a dramatic rise in debt costs, if the 'internal ratings-based' approach prevails (Reisen, 2001). By contrast, the 'standardized' approach, which links risk weights to ratings by eligible external credit assessment institutions, would leave banks' regulatory capital charges, risk-adjusted returns and hence required yield spreads largely unchanged to most developing-country borrowers.
The concern that Basel II will raise the volatility of private capital flows to speculative-grade developing countries, and hence their vulnerability to currency crises, is based on four aspects of the New Accord. First, the rigidity of the 8% minimum capital ratio; linking bank lending to bank equity acts as an automatic amplifier for macroeconomic fluctuations: banks lend more when times are good, and less when times are bad. Rigid capital requirements reinforce that habit. Second, the lateness and the cyclical determination of agency ratings, which define regulatory capital needs under the 'standardized' approach; this means that during booms ratings improve and capital charges decline, while ratings are lowered during the bust, implying higher capital requirements. Third, the cyclical nature of the probability of default and of yield spreads, which determine regulatory capital needs and debt costs under the IRB approach; during the 1970-1999 period, one-year default rates for speculative-grade borrowers oscillated between 1% in tranquil times and 10% in crisis years, largely as a result of global, not idiosyncratic, shocks. Such fluctuations in default probability would translate into corresponding pro-cyclical shifts in risk weights, from 100 to 500 % for speculative-grade borrowers. Finally, the ongoing incentives for short-term rather than long-term interbank lending embedded in the Basel Accord; for speculative-grade developing countries, the regulatory incentives continue to tilt the structure of their capital imports towards short-term debt and makes them vulnerable to capital-flow reversals. Rojas-Suarez's paper demonstrates how this introduces a short-term bias for domestic loans in developing countries.
Regional development banks should not only devote resources to capacity building in order to help developing-country clients to cope better with adherence to financial standards and codes. They should also engage in research that investigates under region-specific circumstances whether standards are effective and productive or whether they risk producing unintended side-effects.
Rodrik (2001) reports World Bank research that has estimated that it costs a typical developing country $150 million to implement requirements under just three of the WTO agreements (customs valuation, sanitary measures and intellectual property rights). He points out that this corresponds to a year's development budget for many of the least-developed countries. Such quantification of (opportunity) costs would seem necessary in view of their proliferation for financial standards as well. The estimates should be published as part of any ROSC.
Crockett, Andrew (2000), Towards a Sustainable Financial System: A Role for the Financial Stability Forum, in R. Adhikari and U. Hiemenz (eds.), Achieving Financial Stability in Asia, OECD Development Centre Seminars, Paris.
Griffith-Jones, Stephany (2001), New Financial Architecture as a Global Public Good, mimeo, Institute of Development Studies, Brighton.
Park, Yung Chul (2000), On Reforming the International Financial System: An East Asian Perspective, in R. Adhikari and U. Hiemenz (eds.), Achieving Financial Stability in Asia, OECD Development Centre Seminars, Paris.
Persaud, Avinash (2000), Sending the Herd Off the Cliff Edge: The Disturbing Interaction Between Herding and Market-Sensitive Risk Management Practices, The Institute of International Finance, 2000 Essay Competition in Honour of Jacques de Larosiere, Washington DC.
Reddy, Y.V. (2001), Issues in Implementing International Financial Standards and Codes, BIS Review 62/2001.
Reisen, Helmut (2001), "Will Basel II Contribute to Convergence in International Capital Flows?", Oesterreichische Nationalbank, Proceedings 29. Volkswirtschaftliche Tagung 2001, pp.49-69, Vienna.
Rodrik, Dani (2001), Trading in Illusions, Foreign Policy March/April.
Rojas-Suarez, Liliana (2001), Can International Capital Standards Strengthen Banks in Emerging Markets?, Institute for International Economics Working Paper Series, Washington DC.
UN (2001), Information Note by Financing for Development Secretariat, 20. August, http://www.un.org/esa/ffd/NGO/business_site1001/meetings_at_the_federal_reserve_.htm.
UNCTAD (2001), Trade and Development Report, Geneva.