VoxEU Column International Finance

Determinants of banking system fragility: A regional perspective

As we are learning time and again, if a big bank goes down it takes a lot of other things down with it. This column looks at the effect of bank fragility and failure on other nearby countries and outlines what can be done to mitigate the cross-border contagion.

It is well known that banks face shocks both on their asset and liability side. A shock that initially affects one institution can become systemic and infect the larger local economy. The globalisation of banking implies further that shocks affecting a particular bank or country can now affect not only the local real economy but also the financial system and real economy in other countries (Peek and Rosengren 1997, 2000, for example, show that shocks hitting Japanese banks generate supply-side effects on the real economy in the US).

The current academic literature on financial fragility, however, has mainly focused on stability of individual banks or stability of individual countries’ banking systems (see eg Allen et al 2009 for a review) but has disregarded regional banking system fragility. This column describes the findings of a recent empirical study of the determinants of regional banking system fragility (Degryse et al 2012). Regional banking system fragility is referred to as a situation when countries’ banking stock indices in a region have jointly very low returns. The 2007–09 financial crisis has shown that a nation with a fragile banking system may affect countries in the region through cross-border linkages and common exposures, and raise concerns for regional banking system fragility. This column studies which regional banking characteristics can alleviate regional banking fragility and which regional banking characteristics help attenuate the impact of cross-regional contagion.

Likely effects of regional banking system characteristics

We consider liquidity, capitalisation, competition, diversification, and presence of foreign banks in a region. These are motivated as follows:

  • The liquidity on a bank’s balance sheet serves as a first line of defence against liquidity shocks. Allen and Gale (2000) and Freixas et al (2000) model banks facing regional liquidity shocks stemming from consumers who are uncertain about where they will consume. A common implication is that greater regional banking system liquidity enhances the stability of the regional banking system.
  • A higher capital base provides a cushion against insolvency. Our motivation to use the capital base of the region’s banking system comes from Freixas et al (2000) and Allen and Gale (2000) who argue that a better capitalised banking system helps in reducing possible contagion effects from individual bank failures in the same country or region.
  • The relationship between competition and financial stability is complex. The ‘competition-fragility’ theories – based on the idea of ‘charter/franchise value’ argue that more bank competition erodes market power, leading to more bank risk-taking. Alternatively, the ‘competition-stability’ view suggests that more market power in the loan market may result in higher bank risk. In terms of the geographical scope of the competition measure, a region’s degree of competition may be a more relevant statistic than the national degree of competition (see also Liu et al 2010).
  • Diversification of bank activities may improve or deteriorate banking stability. Financial conglomerates, for example, face a diversification discount (Laeven and Levine 2007). In contrast, Stiroh (2004) and De Jonghe (2010) find that banking system fragility worsens when a bank engages in non-traditional activities. We therefore test whether diversification in banking activities increases or decreases regional banking fragility.
  • The presence of foreign banks in a region may impact the fragility of the regional banking system in different ways. On the one hand, a greater foreign bank presence may lead to greater banking efficiency and competition in the domestic financial system. On the other hand, foreign banks may provide a channel for cross-border contagion when they transmit shocks from one region to another.
Methodology, research question, and data

We follow the methodology in Bae et al (2003) to study regional banking system fragility. The latter is measured by joint occurrences of negative extreme returns in the banking system indices of multiple countries in the region. The joint occurrences of negative extreme returns are called ‘coexceedances’.

Our research question is whether regional banking system characteristics determine regional banking system fragility. We further study cross-regional contagion by evaluating the effect of coexceedances in one region on banking system fragility in other regions. We are particularly interested in which key regional banking system characteristics in the host region help to dampen the impact of contagion from the triggering region.

In our analysis we use countries’ banking indices from Datastream starting from 1 July 1994 to 31 December 2008 and cover four regions – Asia (ten countries), Latin America (seven countries), Europe (as one entity) and the United States, to compute exceedances and coexeedances. Our regional banking characteristics are based on banks reporting to Bankscope. We use a narrow definition of liquidity, which is the ratio of cash and cash-equivalent assets to total assets. At the regional level, Asia and US have the largest average liquidity ratios (2.8%) during the sample period, while Europe has the lowest (1.8%). Capitalisation of the banking system is captured by the capital-to-assets ratio. We find that the banking systems in Asia, on average, maintain low capital-to–total assets ratio (5.3%), compared to Latin America (8.7%), and that Europe has on average lower capital ratios (4.7%) than the US (7%). As an indicator of concentration, we use the ratio of total assets of the biggest five banks to total assets of all banks. We find that banking systems in Asia are, on average, relatively more concentrated than the ones in Latin America. The degree of diversification is proxied by the loan ratio, ie the ratio of net loans to total earning assets for each country. It gives the extent to which banks are involved in traditional loan-making activities compared to non-traditional activities. We find that net loans are about half of the total earning assets in almost all countries. Latin America has the lowest ratio (44%) with respect to all other regions. Finally, we explore the impact of the degree of foreign bank presence in Asia, Latin America, the US, and Europe. In general, Latin America presents the highest degree of foreign ownership, while Asia is characterised by the lowest degree among the four regions we consider.

Main findings

We find the following results:

  • A region’s banking system characteristics play a significant role in explaining regional banking system fragility:
    • Higher liquidity reduces regional banking system fragility in all regions.
    • Higher capitalisation reduces regional banking system fragility in all regions with the exception of Asia and Europe, where it has no effect.
    • Our findings are supportive of the competition-stability view in most regions as an increase in competition in the banking industry significantly reduces coexceedances.
    • A focus on traditional loan-making activities increases the likelihood of a single country in the bottom tail, but there is no significant impact on joint occurrences of extreme negative returns in the region.
    • A greater presence of foreign banks reduces regional banking fragility in Asia and Latin America.
  • Within-region contagion is higher in emerging-market regions compared to developed regions, and is stronger in Latin America than in Asia.
  • For cross-regional contagion, we find that the contagion effects of Europe and the US on Asia and Latin America are significantly higher compared to the effect of Asia and Latin America among themselves.
  • A higher level of liquidity and capital in the host region attenuate significantly the contagion effects from other regions.
Concluding remarks and policy implications

This column shows that regional banking system characteristics such as higher liquidity and capital help attenuate regional banking system fragility and reduce the impact of cross-regional contagion. Therefore, national supervisors should monitor not only the domestic banking system but also the other banking systems in the region. Finally, the column highlights the importance of coordinating policies at the regional level.


Allen, F, A Babus, and E Carletti (2009), “Financial Crises: Theory and Evidence”, Annual Review of Financial Economics, 97–116.

Allen, F and D Gale (2000), “Financial contagion”, The Journal of Political Economy, 108:1–33.

Bae, K, GA Karolyi, and RM Stulz (2003), “A new approach to measuring financial contagion”, The Review of Financial Studies, 16:717–63.

Degryse, H, MA Elahi, and MF Penas (2012), “Determinants of Banking System Fragility: A Regional Perspective”, CEPR Discussion Paper 8858.

De Jonghe, O (2010), “Back to the basics in banking? A micro-analysis of banking system stability”, Journal of Financial Intermediation, 19:387–417.

Freixas, X, BM Parigi, and J-C Rochet (2000), “Systemic Risk, Interbank Relations and Liquidity Provision by the Central Bank”, Journal of Money, Credit and Banking, 32(2):611–38.

Laeven, L, and R Levine (2007), “Is there a diversification discount in financial conglomerates?” Journal of Financial Economics 85, 331–67.

Liu, H, P Molyneux, and JOS Wilson (2010), “Competition and Stability in Banking: A Regional Analysis”, available on ssrn.com.

Peek, J and E Rosengren (1997), “The international transmission of financial shocks: the case of Japan”, American Economic Review, 87:495–505.

Peek, J and E Rosengren (2000), “Collateral damage: effects of the Japanese banking crisis on real activity in the United States”, American Economic Review, 90:30–45.

Stiroh, K (2004), “Diversification in Banking: Is Noninterest Income the Answer?”, Journal of Money, Credit and Banking, 36(5):853–82.

1,260 Reads