«Adopting international financial standards in Asia: Convergence or divergence in the global political economy? 1 Andrew Walter The last major wave of ...»
Adopting international financial standards in Asia:
Convergence or divergence in the global political economy? 1
The last major wave of emerging market crises triggered a global reform project, led by the
US and UK, to bring financial regulation in emerging countries into line with the regulatory
standards and practices prevailing in the major countries. As noted in the introductory
chapter, emerging market and developing countries had little input into the development of these international standards. Although such standards therefore lacked substantive input legitimacy, the asserted and increasingly widely perceived superiority of the Anglo-Saxon approach to financial regulation and governance gave them a wider degree of output legitimacy. This was certainly the assumption of the G7 countries and the major international financial institutions (IFIs), which put considerable effort into the global dissemination and implementation of international standards and codes on the assumption that this would strengthen the weakest link in the global financial system. 2 This assumption was also shared by many important actors in emerging countries, notably in Asia, who saw the adoption of I wish to thank participants in the Garnet conference, the editors, and various colleagues who participated in the political economy workshop at the London School of Economics for helpful comments on a first draft of this paper. This chapter draws on a larger research project on East Asian compliance with international financial regulatory standards, published in Walter (2006, 2008).
The Financial Stability Forum (FSF), established by the G7 in 1999, referred to twelve ‘key standards’ listed on its website as ‘the various economic and financial standards that are internationally accepted as important for sound, stable and well functioning financial systems’ (Financial Stability Forum, ‘About the Compendium of Standards’, http://www.fsforum.org/compendium/about.html, accessed 2 May 2008).
international standards as a means of importing superior regulatory practices and restraining what they saw as destructive behaviour in their domestic political economies. How successful has this ambitious regulatory reform project been?
In this chapter, I argue that the quality of financial regulation in some emerging market countries has improved considerably since the 1990s, but that there has not been systematic convergence upon western regulatory standards. As in the area of exchange rate policy – where researchers have unearthed a large gap between official policies and actual behaviour (Reinhart and Rogoff 2002) – there is often a similarly large gap between words and deeds in financial regulation. Regulatory convergence has in practice been gradual, limited and variable across countries and areas of regulation, and often superficial rather than substantive (what I have elsewhere called ‘mock compliance’). 3 This is largely because the legitimacy of international standards and associated behavioural practices are often highly contested in the countries that have imported them. The costs of substantive compliance for some actors in developing countries in particular can be high, encouraging these actors to resist compliance.
Often, such actors are sufficiently influential that governments have found solutions that fall somewhere between purely formal and substantive compliance but which can be difficult for outside observers to detect.
Walter 2008. In what follows, I use the terms convergence and compliance interchangeably. In a stricter sense, convergence refers to a process by which previously different practices and institutions in national financial systems become more alike, whereas compliance signifies that the behaviour of actors who are the targets of an international rule or standard conforms to its prescriptions.
This argument has three main implications for the broader debates addressed in this volume.4 First, the relationship between input and output legitimacy in global financial governance is more complex than is sometimes supposed. While developing countries had little input into the standards and codes, there has been less resistance to formal adoption than might have been expected from the low degree of input legitimacy. Formal adoption appeared consistent with best (western) practice, but the considerable scope for de facto domestic adaptation acts as a counterbalance to the undoubted dominance of the major western countries and global financial firms in the process of global financial governance. The assumed mechanisms promoting convergence – specifically market and official incentives – have proven much weaker than the G7 and the IFIs, along with various scholars, initially assumed. 5 If the pressures for convergence were as powerful as some have claimed, we would likely see much more resistance to this form of westernisation at global and regional levels. Second, the formal adoption of international standards has not eviscerated national ‘policy space’ in financial regulation, either because international standards are flexible in their application or because enforcement by international actors is of limited effectiveness. Regulatory forbearance remains an important option for policymakers; so does window dressing for many private sector actors. At the same time, fuller convergence remains an option for those actors who perceive gains from substantive compliance. Third, it suggests that national-level private sector actors with little influence in global forums can constrain, modify, and sometimes block the implementation of international standards at the domestic level.
For official claims about the role of both official and market incentives, see FSF 2000. For academic claims of this kind, see Ho 2002; Simmons 2001; Soederberg 2003.
None of the aforementioned points constitute grounds for complacency since the approximate political equilibrium produced by national adaptation need not be economically optimal.
Indeed, a long historical view casts considerable doubt on the idea that anyone can know what constitutes (or will produce) optimal financial regulation.
The rest of the chapter proceeds in three steps. First, I sketch briefly the unevenness - both across standards and across countries - of convergence upon international regulatory standards. 6 I begin with evidence of convergence at a global level before focusing in more detail on East Asian countries. Second, I discuss some theories that do not adequately explain this outcome, before offering my own. Third, I ask whether this divergence in patterns of financial governance can continue. A final section briefly concludes.
The unevenness of financial regulatory convergence The focus is on four main areas of financial regulation: the Basle Core Principles for Effective Banking Supervision (BCP), the OECD’s Principles of Corporate Governance (PCG), the International Financial Reporting Standards (IFRS), 7 and the IMF’s Special Data Dissemination Standard (SDDS). Note that these standards range in degree of specificity from the very general to the relatively detailed. At the very general end of the spectrum are For those interested in a more detailed account of financial regulatory reform in East Asian countries after the crisis, see Walter 2006, 2008.
Strictly speaking, since 2001 the International Accounting Standards Board (IASB) issues IFRS, but existing International Accounting Standards (IAS), issued by the IASB’s predecessor, the International Accounting Standards Committee (IASC), remain valid until replaced or withdrawn.
the PCG; the BCP include a mixture of general principles and more detailed standards, while IFRS and SDDS are both relatively detailed.
Table 1 measures formal compliance with SDDS, IFRS, and with one key aspect of the BCP, the ‘Basle I’ capital adequacy standard, for different groups of countries just before the recent crisis. It is restricted to these three areas because IMF and World Bank data on compliance with the BCP and the PCG (collected through the Financial Sector Assessment Programme, or FSAP) is not publicly available, and because there is no generally agreed measure of compliance in these latter two areas.
Table 1: Formal compliance with SDDS, IFRS, and Basle I standards: percentages by country group, end 2007
Source: IMF, Global Financial Stability Reports; Deloitte-Touche Tohmatsu; Barth, Caprio and Levine (2007) Notes: *The figure for the ‘IMF’ group for Basle I compliance is for those 143 countries on the Barth et al.
database, which probably overestimates compliance in this category. **Basle I figures generally are as of end 2007, updated from Barth et al. by the IMF’s Global Financial Stability Report, April 2008. ***The IMF lists twenty-six emerging market countries. ****The thirteen major crisis-hit countries are Argentina, Brazil, Hungary, India, Indonesia, Japan, Korea, Malaysia, Mexico, Russia, Thailand, Turkey and Venezuela (noncompliant). *****The ten major East Asian economies are China (non-compliant), Hong Kong, Indonesia, Malaysia, Japan, Thailand, Singapore, Taiwan Province of China, South Korea and the Philippines.
Three things stand out about table 1. First, OECD countries exhibit fairly high levels of formal compliance across these three sets of international standards. Second, formal Basle I compliance is almost universal, whereas the pattern for SDDS and IFRS is more variable.
Third, among emerging markets (for which the SDDS was primarily intended), formal compliance is high for SDDS and Basle I, but low for IFRS.
Because compliance is a continuous rather than a binary variable, these formal indicators do not fully capture either the reality of legislation or regulatory and private sector practice. For example, although most emerging market countries have not yet adopted IFRS in full, some claim that their domestic accounting standards are ‘largely’ though not completely based on them (e.g. Korea and Thailand). Even if we could pinpoint where international standards and national regulations diverge, there is the more complex question of whether regulators, banks, companies, and internal and external auditors actually behave in ways that are consistent with national rules. Mock compliance occurs when actors formally signal their adoption of specific international rules or standards but behave inconsistently (see Raustiala and Slaughter 2002: 539; Shelton 2003: 5). In effect, mock compliance can occupy a range of outcomes between the extremes of formal non-compliance and substantive (behavioural) compliance, though it does not exhaust all the possibilities on the compliance spectrum (figure 1); it is analogous to the now widely recognised phenomenon in exchange rate policy where there is often a large divergence between announced and de facto policies (Reinhart and Rogoff 2002). Mock compliance can occur for numerous reasons, including deliberate regulatory forbearance by the government or its agencies, low bureaucratic enforcement capacity or corruption, and behaviour by private sector actors inconsistent with the intent of the rules.
Figure 1: The spectrum of compliance
We can be reasonably sure that the level of mock compliance with SDDS is low in most cases. This is because the macroeconomic data that SDDS subscribers are obliged to post on the IMF’s bulletin board are based upon publicly available national statistics and must be internally consistent. In addition, the IMF publicly declares whether or not a country posting data meets the requirements of SDDS. 8 The only other financial regulatory standards for which categorical official judgements about (country-level) compliance are made are those for money laundering and terrorist financing.
In other areas, the IMF-World Bank FSAP, which laboriously assesses countries’ compliance with all international financial standards, produces reports from which very critical and quantitative judgements about compliance are (at member countries’ request) often excised. 9 A quarter of all reports are never published, and many important countries have simply refused to participate in the FSAP (though the recent G20 collective commitment to participation may change this). In areas like corporate governance, bank regulation, and accounting, reaching judgements about degrees of compliance is difficult and often The SDDS is sometimes criticised as insufficient and outdated (e.g. IIF 2006), though the question of its optimality – or indeed that of the quality of the underlying data – is different to that of compliance.
For the Reports on the Observance of Standards and Codes (ROSCs), see http://www.imf.org/external/np/rosc/rosc.asp (accessed 3 May 2008).
controversial. The FSAP acts as an interlocutor with public sector regulators, but the quality of compliance is at least as much a question of private sector behaviour. For all these reasons, published FSAP reports offer a poor guide to patterns of compliance.
Nevertheless, some FSAP reports are relatively candid and additional anecdotal evidence often does emerge that helps to give a better picture of compliance. Taken together, this can show that formal non-compliance in some areas is considerable and that mock compliance is even more significant. To illustrate, regulatory officials in Korea, among the most avid adherents of the international standards project in Asia, claimed that the country had met or exceeded most international standards by 2002 (FSS 2002: forward). But the FSAP review team in 2003 argued that the new Korean financial regulator was insufficiently independent from government and industry, as required by the first BCP. Considerable evidence also emerged of regulatory forbearance for banks willing to lend to those large, distressed corporations important to the Korean government’s industrial restructuring objectives. For example, foreign-controlled Korean banks complained of government pressure after 2000 to roll over loans to Hyundai, including to its semiconductor affiliate, Hynix. From May 2000 to June 2002, Korean financial institutions, mostly state-controlled, provided new credits to Hyundai group. Half of this financial support went to Hynix, even though it was then uncreditworthy (US ITA 2003: 18). The new financial regulator had allowed banks to classify their Hynix loans through late 2001 as ‘normal’ or ‘precautionary’, a relatively lenient treatment that required them to set aside only small provisions (Fitch Ratings 2002: 2-3).