Assessment of the macroeconomic impact of higher loss absorbency for global systemically important banks

Executive summary

Weaknesses at large financial institutions have often played a central role in the triggering and propagation of systemic financial crises. The 2007-09 financial crisis was only the most recent example. Since the crisis, authorities worldwide have sought ways to strengthen regulation and supervision of these institutions, including through efforts at the international level led by the Financial Stability Board (FSB) and the Basel Committee on Banking Supervision. As part of a package of measures for addressing these risks, the Basel Committee has made proposals for improving the loss absorbency of global systemically important banks (G-SIBs).

In April 2011, the FSB and Basel Committee reconvened the Macroeconomic Assessment Group (MAG) to investigate the macroeconomic costs and benefits of these proposals. The MAG comprises macroeconomic modelling experts from central banks and regulators in 15 countries and a number of international institutions. Close collaboration with the International Monetary Fund (IMF) was an essential part of this process. In its work, the MAG drew on its earlier assessment of the transitional costs of the proposals for strengthened capital and liquidity requirements under Basel III, and on the long-term cost-benefit analysis performed by the Basel Committee's Long-term Economic Impact (LEI) study.

The costs of the G-SIB proposals stem from the adverse impact on economic activity, especially investment, of banks' actions to increase interest rate spreads and cut lending in order to build up their capital buffers. The MAG estimated the impact of higher capital requirements on G-SIBs by scaling the impact of raising capital requirements on the banking system as a whole, reported by the MAG in 2010, by the share of G-SIBs in domestic financial systems. While these shares vary across jurisdictions, the share of the top 30 potential G-SIBs (using the Basel Committee's proposed methodology and end-2009 data) averages about 30% of domestic lending and 38% of financial system assets in the MAG economies.

If we use lending shares as a scaling factor, raising capital requirements on the top 30 potential G-SIBs by 1 percentage point over eight years leads to only a modest slowdown in growth. GDP falls to a level 0.06% below its baseline forecast, followed by a recovery. This represents an additional drag on growth of less than 0.01 percentage points per year during the phase-in period. The primary driver of this macroeconomic impact is an increase of lending spreads of 5-6 basis points. Soon after implementation is complete, growth is forecast to be somewhat faster than trend until GDP returns to its baseline. The aggregate figures conceal significant differences across countries, which reflect differences in the role of G-SIBs in the domestic financial system and in current levels of bank capital buffers. International spillovers are also important, and in some countries are likely to be the dominant source of macroeconomic effects.

The overall results are robust to variations in key assumptions. Using a longer list of banks, scaling by assets rather than lending, shortening the implementation period, or limiting the ability of authorities to offset slower growth with monetary or macroprudential policy were all found to increase the growth impact, but not markedly.

What will be the effect of the full package of the Basel Committee's proposals for stronger capital requirements - the set of buffers that will be required of all banks under Basel III, combined with the additional buffers to be carried by G-SIBs? The impact of the Basel III proposals, using the end-2009 global capital levels as a starting point, was calculated by the MAG in 2010. On top of this, we assume for illustrative purposes that the top 30 G-SIBs will need to raise their capital ratios by an additional 2 percentage points, and that both parts of the reform are implemented over eight years. Adding together these two components, we find that the impact is again quite small, with GDP at the point of peak impact forecast to have fallen 0.34% relative to its baseline level. Roughly 0.04 percentage points are subtracted from annual growth during this period, while lending spreads rise by around 31 basis points. As before, different assumptions lead to different effects, with faster implementation or a weaker monetary policy response increasing the impact on GDP.

The benefits of the G-SIB framework relate primarily to the reduction in the exposure of the financial system to systemic crises that can have long-lasting effects on the economy. The LEI estimated the benefits of Basel III by multiplying the degree to which it reduces the annual probability of a systemic crisis, by an estimate of the overall cost of a typical crisis in terms of lost output. Drawing on the LEI's results, the MAG estimated that raising capital ratios on G-SIBs could produce an annual benefit in the order of 0.5% of GDP, while the Basel III and G-SIB proposals combined contribute an annual benefit of up to 2.5% of GDP - many times the costs of the reforms in terms of temporarily slower annual growth.

As is the case with any economic analysis, producing these estimates required making a number of assumptions. Some of these can be imposed or removed as part of the estimation process. Other aspects are more difficult to analyse. For example, the role of G-SIBs in the financial system, through their status as market leaders or through their dominant positions in certain activities, may be greater than is implied by their share in lending or assets. In this case, the macroeconomic impact of their adjustment to higher capital levels may be greater - although the benefits from strengthening their balance sheets would be greater as well. Conversely, a pullback in activity by G-SIBs may have a relatively mild impact if credit can be provided instead by smaller institutions or capital markets; this would reduce the macroeconomic cost of the stronger G-SIB requirements, but could also reduce the benefit from reducing the risks of distress at a G-SIB.

More analysis is needed to understand these effects fully. It will also be important to engage in further study of the impact of other elements of the FSB's broader framework for global systemically important financial institutions (of which the Basel Committee's G-SIB proposals form a part) as they are implemented, such as the proposal that unsecured and uninsured creditors be "bailed in" at the point of resolution. We will only be able to fully understand these effects, and assess the relevant costs and benefits, as more experience is gained.