The Design of Financial Regulation in the G20 Debate: Tata or Ferrari?

The urgency of strengthening the regulation and supervision of financial institutions was highlighted in the first two G20 meetings in November 2008 and in April 2009. Over the last few months, a large number of academic papers, official reports and white papers have been published in the US and in Europe, and international organizations have also contributed.  Much less has come from the non-G7 G20 countries. Hence, rules are being set by the G20 but at the initiative of advanced countries. But are they appropriate for emerging economies? Shouldn’t we opt for different rules for ‘Tata’ and ‘Ferrari’?*

The G20 agenda in financial regulation focuses on three issues: (i) filling the gaps of regulation and supervision, (ii) tackling ‘too-big-to-fail’ financial institutions, and (iii) reducing the pro-cyclicality of capital requirements and accounting standards.

The first issue is probably the most important one. The key cause of the crisis is that banks escaped capital regulation by using off-balance-sheet special investment vehicles (SIVs) to buy asset-backed securities (ABSs) while financing these investments mostly through short-term asset-backed commercial paper. However, the corresponding risk was not transferred since banks extended guarantees and credit lines to their SIVs. When in the wake of the crisis the short-term funding dried up, ABSs (now called ‘toxic assets’) were transferred back to banks’ balance sheets (where capital requirements apply), leading to a sudden undercapitalisation of the banking sector and to the subsequent disruptions in financial markets. Failure to regulate the ‘shadow’ banking sector is thus at the root of the crisis. The G20’s response so far has been to regulate hedge funds, fight regulatory havens and re-organize financial supervision within each country or region. Although SIVs are mainly a creation of advanced economies, emerging countries are concerned by regulatory arbitrage. They should take the question of supervisory scope very seriously, and be aware of the risk encountered when giving in to the temptation to favor the development of national ‘champions’ through light supervision.

The second issue – that of ‘too-big-to-fail’ institutions – also concerns emerging countries, since a number of their banks have reached the size (if not still interconnectedness) of the systemically important institutions at the global level. This primarily concerns Chinese banks, whose new lending, in response to the crisis, has surged and is expected to increase by 33.9 percent of GDP in 2009, a level of re-leveraging unprecedented in the world. This calls for a more adequate regulation and supervision. The goal is to have ‘systemic’ banks at least partially internalize the social cost of bank bail-outs either by applying more conservative prudential standards (proposed by the US Treasury) or by forcing the biggest banks to pre-plan their own demise by writing ‘living wills’ (suggested by Lord Turner from the FSA). However, it is difficult to even imagine that these measures will be considered in China, where banks are state-owned and are often viewed as an extension of the fiscal policy. As direct government subsidies have decreased over time, bank lending on preferential conditions has been used to help state-owned firms. Such behavior has become especially prominent in the Chinese response to the crisis, but there are concerns about the future solvency of these institutions.

On the third question, the G20 considers introducing counter-cyclical capital or reserve buffers, based on a macro-prudential assessment of risk. This agreement is striking because economists do not know when to start building buffers and when to draw on them.  A number of methods have been proposed (autoregressive adjustment of capital, as well as a multiplier to adjust capital, which could depend on lending growth, credit spreads, GDP growth or asset prices), but the research has shown that none of these measures performs perfectly as each crisis is different in nature. As for mark-to-market accounting, a number of politicians have called for reform of accounting standards, insisting that ‘market’ value is not always ‘fair’ value, and have expressed support for model-based or historical-cost accounting. However, in most previous crises, non-transparent balance sheets tended to actually deepen the crisis due to late recognition of problems, and to unnecessarily delay the recovery. Consequently, many economists propose to continue relying on fair-value accounting, while applying filters to mitigate the pro-cyclicality of accounting standards, or supplementing them with stress-testing exercises that would provide more long-term views of solvency.

The question of countercyclical buffers is where perhaps Tatas and Ferraris should be differentiated for at least two reasons. First, the introduction of higher and countercyclical capital requirements is likely to increase the cost of financial intermediation. While this cost is justifiable in countries with deep financial markets in order to achieve financial stability, emerging economies – with a significant share of the population still excluded from banking markets – would prefer a differential country-by-country approach. Second, in open emerging economies, the credit cycle is very much dependent on capital inflows. To the extent that capital inflows are substitutes for domestic credit, counter-cyclical buffers may prove inefficient.

Finally, the current reform proposals build on the Basel II framework, which was created by developed countries but is being implemented by all international banks. Some of its proposals are difficult to carry out in practice even in developed economies. For example, very few banks are able to compute “through-the-cycle” internal ratings, which would be a way to mitigate the pro-cyclicality of capital requirements. Among emerging countries, only South Africa has implemented Basel II, but both banks and the supervisor are struggling due to short span of available data, which does not cover the whole economic cycle and, thus, makes calculation of “internal ratings”, “stress-tests”, or future counter-cyclical buffers unreliable. It remains to be seen how the work of Basel II will change now when non-G7 G20 countries have been given membership in 2009.


* “The expression is from Andrew Sheng, University of Malaya”?