Mistakes in Basel Accord Could Harm Developing Countries Stephany Griffith-Jones

The Basle Committee of G-10 banking regulators has proposed a new Capital Accord, with the expressed aim of more accurately aligning regulatory capital with the risks that international banks face. Recent detailed research shows clearly that the current Basle proposal would quite significantly overestimate the risk of international bank lending to developing countries; this would increase capital requirements excessively on such lending, leading to a sharp increase in the cost of bank borrowing by developing countries, as well as to an important fall in the supply of bank loans.

This is particularly serious issue now, as in the last five years bank lending to the developing world has already fallen sharply. This decline in bank lending and other capital flows has had a very negative impact on growth in the developing world, especially recently in Latin America. Thus, the current proposals are problematic, both in terms of the Basle Committee’s own aims (more accurate measurement of risk for determining capital adequacy) and due to their further discouragement of already insufficient bank lending to emerging markets, which damages growth of their economies. The latter impact is manifestly against one of the aims of the G-10, which is actively to encourage private flows to developing countries and use them as an engine for stimulating and funding growth.

 How do the current Basle proposals overestimate risk of lending to developing countries? How could they be modified to be both more precise and less damaging to developing countries?

One of the reported major benefits of lending to –and investing in – developing countries, is their relatively low correlation with mature markets. In our research, we have carefully tested this hypothesis empirically and found very strong evidence – for a variety of variables, and over a range of time periods – that correlation between developed and developing countries is significantly lower than correlation only amongst developed countries. For example, spreads on syndicated loans – which reflect risks and probability of default – tend to rise and fall together within developed regions more than between developed and developing countries; similar results are obtained for the correlation of profitability of banks. Furthermore, broader macro-economic variables (such as growth of GDP, interest rates, evolution of bond prices and stock market indexes) show far more correlation within developed economies than between developed and developing ones.

Finance theory and practice tells us that the clear implication of these empirical findings is that a bank’s loan portfolio that is diversified between developed and developing countries has a lower level of risk, than one focussed exclusively on lending to developed economies. In order to test this more directly we simulated two loan portfolios, one with diversification only across developed economies, and another that diversified across developed and developing regions. We found that the estimated unexpected losses for the portfolio focussed only on developed country borrowers was almost 23% higher.

Reflecting risk

Given that the capital requirements which Basel regulators determine should precisely help banks cope with unexpected losses, it is extremely unfortunate that the current Basle proposals do not incorporate explicitly the benefits of international diversification. The surprising fact that at present the Basle proposal does not do so implies that in this aspect, capital requirements will not clearly reflect risk, and thus will be both incorrectly and unfairly penalising lending to developing countries.

It therefore seems imperative that the Basle Committee in its next (and almost final) revision of the proposed Basel II, incorporates the benefits of international diversification, for example by explicitly reducing capital requirements, to take account of these diversification benefits.

Lending to small and medium enterprises

It is encouraging that there is a clear precedent, as the Basle Committee has already made such a change with respect to lending to small and medium enterprises (SME’s). After the release of the consultative document in January 2001, there was widespread concern – especially in Germany – that the increase in capital requirements would sharply reduce bank lending to SMEs, with very negative effects on growth and employment. The technical case was made that the probability of a large number of SMEs defaulting simultaneously was lower than for a smaller group of large borrowers. Intensive lobbying by the German authorities implied that this technical argument was recognised, and the Basel Committee agreed to lower average capital requirements by about 10% for smaller companies.

Our empirical research implies that at least as large a modification is justified with respect to international diversification, related to lending to developing countries. There are no practical, empirical or theoretical obstacles to such a change, which could potentially greatly benefit the developing world and ensure more precise measurement of risk and capital adequacy requirements. This, after all, is the aim of the entire process.

 Recognising the benefits of international diversification

The Basle Committee has always emphasised the technical nature of its proposals and the technical case for including the benefits of diversification is extremely strong. Furthermore, G-10 governments have a strong commitment to encouraging private flows, and therefore an incentive to avoid measures being taken that actually discourage such flows. Developing and transition countries are unfortunately not represented at all in the Basle Committee, so they have limited leverage to make the case. However, given the Basle Committee’s technical expertise and fair-mindedness, hopefully a change will be introduced to the current proposal, to take account of the benefits of international diversification. It would be technically wrong, economically unwise and politically insensitive not to do so.

(Source: www.ids.ac.uk)