Have more strictly regulated banking systems fared better during the recent financial crisis?

Abstract

We assess whether during the recent financial crisis banking systems in countries with more stringent prudential banking regulation have proved more stable. We find indicators of regulatory strength to be relatively well correlated with the extent to which countries have escaped damage during the recent crisis, as measured either by the degree of equity value destruction in the banking sector or by the fiscal cost of financial sector rescue.

Acknowledgements

The authors are indebted to Sven Blöndal, Romain Duval, Jørgen Elmeskov, Sebastian Schich, Jean-Luc Schneider and several members of the OECD Economics Department for useful comments.

Notes

¹The World Bank surveys, which constitute the by far largest part of the underlying information set, were released in 2007, 2003 and 2001 and cover the periods 2005/2006, 2001/2002 and 1998/2000, respectively. The last update of the panel database was published in June 2008. The information used in the construction of the indicators was subsequently verified and corrected by OECD member countries, thus improving the reliability of the available information.

²This overall indicator is obtained by taking the average of the normalized area indicators. Normalized indicators are obtained by subtracting, for a given area indicator, for each country the cross-country mean from the indicator value and dividing the result by the cross-country standard deviation.

³A set of intermediate outcome indicators of financial stability has been assembled from the ‘Financial Soundness Indicators’ and the ‘Global Financial Stability Report’ of the International Monetary Fund, and tested for cross-country correlations with the regulation indicators. See Ahrend et al. (2009) for details.

⁴Only banks that are included in the Datastream Banking Share Price Index DSBANKS for the respective country were retained.

⁵These results, which are only described verbally here, are available in Ahrend et al. (2009).

⁶Alternatively, this analysis can be replicated using the third quarter of 2009, with very similar results. These results are available from the authors upon request.

⁷See IMF Fiscal Affairs Department (2009), table 4. ‘Recovery rates’, which critically influence the cost of direct interventions, are estimated based on historical data from former banking crises. Estimates of the expected cost of guarantees are broadly based on the Contingent Claims Approach. The size of rescue packages likely reflects not only the severity of the financial crisis but also other factors including inter alia the responsiveness of policy makers.

⁸As such, the financial centre dummy would pick up a bias in the indicators. Alternatively, governments with financial centres may be particularly willing to put up the resources to guarantee the survival of a strong financial sector.

⁹The degree to which imbalances (as bubbles in housing or credit markets) build up prior to a crisis may be thought to affect the strength of the crisis. During the recent pre-crisis period, a (downward) deviation from a Taylor rule, which under certain circumstances could be interpreted as overly accommodating monetary policy, has been associated with strong price increases in real estate and concomitant developments of housing investment and mortgage credit (see Ahrend, 2010). A (downward) pre-crisis deviation from a Taylor rule may hence be expected to result in a larger need for financial sector bail-out, and hence a larger net fiscal cost, and this result is indeed found in our analysis.

¹⁰As measured by the standard deviation of the normalized area indicators. See Ahrend et al. (2009) for these results.