7 Internationally Active Large Banking Groups

Jörg Decressin, Wim Fonteyne, and Hamid Faruqee
Published Date:
September 2007
  • ShareShare
Show Summary Details

This chapter1 focuses on the international diversification of and associated risks for the 12 most internationally active large banking groups (ILBGs) in the European Union. The chapter relates general trends among these ILBGs to accounting-data-based and market-based measures of risk, both at the level of individual institutions and as a group. The chapter finds that diversification has proceeded and generally benefits individual ILBGs, but that risks to the group of ILBGs as a whole have not necessarily declined.

The 12 institutions examined here were selected from the top 30 European banks by choosing those with less than 50 percent of business in their home country.2 Together, they account for about one-fourth of total assets of the more than 7,000 European credit institutions (Table 7.1).

Most of these 12 banks have recently rebounded from a weak spell in the early 2000s. Financial market weakness following the deflation of the high-tech stock market bubble and the global and European slowdowns affected the ILBGs’ performance considerably in 2001–03. Since then, bank profitability has generally recovered and both financial soundness indicators and market-based indicators of risk have strengthened.

Table 7.1.Assets of Internationally Active Large Banking Groups in Sample(In billions of euros)
Total Assets
ABN Amro Holding543.2597.4556.0560.4608.6
Banco Bilbao Vizcaya Argentaria292.6305.5274.9282.6306.2
BNP Paribas693.1825.3710.3783.0905.9
Deutsche Bank929.0917.7758.3803.6840.0
HSBC Holding714.5779.0723.4818.9937.4
HVB Bayerische Hypo-und Vereinsbank694.3716.2526.5473.2467.4
ING Groep650.2705.1716.4778.8866.2
KBC Bank176.9215.9208.5210.7230.2
Nordea Bank AB224.5241.5249.6239.3251.8
Banco Santander Central Hispano347.3355.9319.0346.4569.8
Share of Total Assets of EU-15 Credit Institutions
(in percent)28.128.324.624.725.6
Memorandum items:
Total assets of top 30 EU banks12,285.813,497.313,003.413,479.414,888.9
Total assets of EU-15 credit institutions22,570.424,398.724,991.226,151.228,250.01
Total assets of euro area MFIs12,408.013,576.713,931.214,557.315,725.6
Sources: BankScope; and European Central Bank.


Sources: BankScope; and European Central Bank.


Among the 12 banks, international exposure, balance sheet composition, and the degree of international diversification differ substantially. The financial soundness indicators, while generally moving in the same direction, also are at different levels across these banks. This indicates that the banks might be following different business strategies.

At the same time, the risk profiles of the ILBGs do not exhibit a clear trend toward improvement. Despite their increased diversification, the volatility in the banks’ returns has increased markedly. Hence, in the market’s perception at least, the banks are intermediating more volatile risks.

Banking Sector Trends

In 2001–03, most of the European banking sector suffered a pronounced hit from declining asset prices and the accompanying slow growth of the euro area economy. These developments were manifest in slower or negative balance sheet growth, higher nonperforming loan ratios and loan losses, and lower or even negative returns on equity. Average balance sheet growth declined from about 25 percent per year in 1999–2000 to –3 percent per year in 2001–02.

In 2003–04, however, most large banks rebounded. As asset markets recovered and the economic cycle turned, ILBG balance sheets started expanding again (by about 8 percent annually) and credit quality improved. Combined with the impact of cost-cutting, this enabled a return to profitability. Pretax profits and returns on equity rebounded to the levels seen before the run-up to the tech bubble or higher. In parallel, financial soundness indicators improved.

Investment behavior since the late 1990s differs sharply among the ILBGs. Some banks have expanded their fee-earning and financial market activities at the expense of traditional lending. Deutsche Bank, HVB, ING, and HSBC, and to a lesser extent WestLB, have all considerably raised the share of securities in their asset portfolios since 1998, to a level above 60 percent in the case of Deutsche Bank. In contrast, BBVA, Santander, KBC, and Nordea have expanded their loan books as a share of total assets.

International Diversification

The ILBGs have acquired substantial foreign investments and loans. Total consolidated foreign (that is, outside the home country) exposure of banks in EU-15 countries more than doubled since 2002 (Figure 7.1).

Figure 7.1.Consolidated Foreign Claims of EU-15 Countries

(In trillions of U.S. dollars)

Source: Bank for International Settlements.

Note: RoW = Rest of the world.

Cross-border exposures of EU-15 banks in EU-15 countries have risen by some 160 percent since 2000, an even more stark increase than in the total cross-border exposures of these banks. Most of the banks’ foreign investments are also in the EU-15 (Table 7.2). Again, the major banks have followed different strategies, with many still concentrated in their home markets, with others having more of a presence outside the European Union, and with some being invested heavily in the new member states. The latter strategy seems to have contributed substantially to profits, as profit margins in the new member states generally have been higher than in home markets.

Table 7.2.Cross-Border Exposures of ILBGs, 2004(In percent of total assets or operating revenues)
Home CountryTotal EU-15New Member States



ABN AmroNetherlands37.89.860.653.50.00.0
BNP ParibasFrance58.262.968.392.00.30.4
Deutsche BankGermany13.730.035.645.00.30.5
HSBCUnited Kingdom41.3n.a.42.730.30.00.0
Sources: BankScope; and banks’ annual reports.Notes: Includes subsidiaries of which the bank owns at least 50 percent of shares. Data for HSBC are for 2005.
Sources: BankScope; and banks’ annual reports.Notes: Includes subsidiaries of which the bank owns at least 50 percent of shares. Data for HSBC are for 2005.

A majority of the ILBGs have operating revenues that are geographically more diversified than their assets, suggesting that higher returns are earned on assets abroad. The analysis looks at revenues and assets of banks’ subsidiaries in EU countries as well as worldwide. Using 2004 data from Bankscope and the banks’ annual reports, total assets and operating revenues are derived for those subsidiaries in which the banks have majority ownership.3 Minority-owned subsidiaries are not considered in the analysis (as many of these investments are financial rather than strategic), and foreign branches are not identified separately.4 Hence, the total foreign investment of the banks is likely to be underestimated, and potentially more so within the European Union, but most of the strategic investment is captured. Some two-thirds of the banks turn out to have more geographically diversified operating revenues than assets (Table 7.3), which could indicate that banks’ foreign operations tend to be more profitable than their home-country businesses.5 This, of course, has been a major driver of the banks’ expansion abroad in the first place.

Table 7.3.Operating Assets and Revenues of ILBGs: EU and Worldwide, 2004(Herfindahl index)
ABN Amro1,9492,0043,0303,008
BNP Paribas3,4094,5413,8764,564
Deutsche Bank3809814,0702,936
Sources: BankScope; and banks’ annual reports.Notes: The Herfindahl index is defined as the sum of squared market shares in the countries under consideration. It hence ranges from 0 (complete dispersion) to 10,000 (complete concentration). HSBC data are for 2005.
Sources: BankScope; and banks’ annual reports.Notes: The Herfindahl index is defined as the sum of squared market shares in the countries under consideration. It hence ranges from 0 (complete dispersion) to 10,000 (complete concentration). HSBC data are for 2005.

Banks with a more internationally diversified stream of revenues can be shown to achieve higher returns on average and have more securities on their balance sheet. Figures 7.2 and 7.3 plot banks’ Herfindahl index for diversification across EU countries (2004 data) against average returns on assets and returns on equity during 1994–2004. The figures show that geographical concentration of revenues, and to a lesser extent of assets, is negatively correlated with returns. Hence, more diversified banks (lower Herfindahl index) tend to have higher returns. The direction of causality is, of course, ambiguous: banks that find more profitable expansion opportunities abroad than at home are the most internationally diversified, which means that international expansion may reflect institution-specific competitive advantages or relatively low profitability in the home market. A similar analysis for the standard deviation of these returns reveals a slight upward trend (not shown), indicating that the more diversified banks also tend to have a lower variation in their returns. This means that the higher returns are, on average, not the result of more risky portfolios (as captured by the standard deviation of returns).6 On average, more internationally diversified banks also have more securities on their balance sheets (Figure 7.4). Hence, there seems to be a positive correlation between the international diversification of banks and their focus on fee-earning and financial market activities. More internationally diversified banks therefore seem to be less focused on traditional intermediation.

Figure 7.2.Herfindahl Index for the Euro Area and Return on Assets

Sources: BankScope, Annual Reports; and IMF staff calculations.

Figure 7.3.Herfindahl Index for the Euro Area and Return on Equity

Sources: BankScope, Annual Reports; and IMF staff calculations.

Figure 7.4.Herfindahl Index for the Euro Area and Securities

Sources: BankScope, Annual Reports; and IMF staff calculations.

Market-Based Financial Soundness Indicators

Diversification of the ILBGs affects the stability of both the individual institutions and the financial system as a whole. As discussed in Chapter 6, increased diversification, both internationally and across different business lines, should yield better risk profiles for financial institutions. However, when all or most banks diversify in the same way, financial stability may not improve as the system may become more vulnerable to large, common shocks.7 Moreover, economies themselves are becoming increasingly integrated, in particular within the European Union, thus reducing the stability-enhancing effects of international diversification. This section evaluates the stability of individual financial institutions and the system as a whole using the distance to default, explained in Chapter 6, as a market-based soundness indicator.8

Evolving business strategies have not notably altered the risk profile of individual ILBGs. The DD (distance to default) analysis indicates that there is no clear trend in the portfolio DD for the ILBGs, 9 nor in the DD of any of the 11 individual ILBGs (Figures 7.5 and 7.6).10 The individual DDs, as well as the portfolio DD, declined sharply toward the end of the 1990s and again during 2001–03, but recovered to near peak levels afterward.11

Figure 7.5.Portfolio Distance to Default of 11 ILBGs

(Standard deviations)

Source: IMF staff calculations.

Note: ILBGs = Internationally active large banking groups.

Figure 7.6.Bank Distance to Default of Individual ILBGs

(Standard deviations)
(Standard deviations)

Source: IMF staff calculations.

This lack of a clear change in the ILBGs’ risk profile reflects a combination of higher capitalization and higher exposure to risks. Until the mid-1990s, the banks’ capitalization was relatively low, but this was matched by low volatility. Toward the end of the 1990s, the capital-to-assets ratio generally improved, but, because volatility rose simultaneously, the DD declined. As volatility stayed high through 2004, the decline in the DD during 2001–03 was caused by declining capital ratios, which, in turn, was related to an economic downturn and a simultaneous financial market correction. The sharp decline in volatility since 2004 and slowly recovering capital ratios are the main drivers behind the recent improvements in the DD. But even now, for most banks, the volatility is higher than it was a decade ago, indicating that, in the market’s perception at least, the banks have used their improved capital base to intermediate more volatile risks. In other words, the banks have moved toward more high-risk/high-return projects, and have boosted their capital to do so.

The ILBGs in the sample tend to be riskier than other large European banks. Figure 7.7 shows that the portfolio DD for the 11 ILBGs in the sample has generally been lower than the portfolio DD for a broader sample consisting of major euro area banks (53 in total).12 This indicates that, on average, the market considers risks among the 11 ILBGs to be somewhat higher (and less diversified) than risks among the entire group of large euro area banks. During a few years in the early 1990s, the opposite was true, as a result of a relatively strong performance of the 11 ILBGs (featuring low volatility).

Figure 7.7.Difference between the Distance to Default of the Portfolio of All Large Euro-Area Banks and the 11 ILBGs

(Standard deviations)

Source: IMF staff calculations.

The DDs of the different banks also reflect differences in business strategies and focus. Based on the developments in their DDs, the banks can be divided into different groups using cluster analysis.13Figure 7.8 presents the binary tree structure of a representative hierarchical cluster. Banks in the same (or nearby) cluster(s) are more similar in terms of their DDs than banks in clusters further away. In the first step, the clustering selects HSBC, KBC, and HVB as being different from the other banks in the sample. These three banks have major exposures in Asia (HSBC) and Central and Eastern Europe (KBC and HVB), which sets them apart from the other banks. The next cluster contains Deutsche Bank and BNP Paribas. Together with ING (which is selected to be part of the next cluster, possibly due to its high profitability), these are the largest of the other banks. All three have a strategy that relies heavily on noninterest income. The cluster consisting of the remaining six banks in the sample (ING, BBVA, Santander, Fortis, Nordea, and ABN Amro) features highly profitable banks. Moreover, the business of these banks (with the exception of ING) tends toward traditional lending, with interest income yielding between 55 and 65 percent of total income. Within this group, BBVA and Santander can be seen as a subgroup because of their large exposures in Latin America.

Figure 7.8.Tree Structure of a Representative Hierarchical Clustering of the ILBGs

Source: IMF staff calculations.


Cross-country diversification has proceeded in the ILBGs, even though considerable differences persist. Income sources continue to differ across banks but tend to be more internationally diversified than assets. In addition, an analysis of majority-owned subsidiaries shows that the most internationally diversified banks, on average, tend to focus more on fee-earning and financial market activities, and that they achieve higher returns with lower volatility. Geographic diversification highlights differing investment strategies, as some banks are heavily invested in the new member states, while others follow a worldwide or more domestically oriented strategy.

The main finding of this chapter is that increased diversification has not led to a decline in the systemic risk of the ILBGs as a group, despite some positive effects for individual ILBGs. The capital-to-assets ratio is generally higher than a decade ago, but, for many banks, so is the volatility. At the same time, through the process of diversification, the ILBGs have increased their exposure to common financial cycles and more similar income sources. Hence, in the markets’ perception at least, the level of risk incorporated in the ILBGs has not decreased.


Prepared with research assistance from Marianne El-Khoury.


The business of a banking group in a country is defined as an unweighted average of the percentage of assets, revenues, and employment in that country. The top 30 banks were selected from the list of Top 300 European banks published by The Banker magazine in 2002 (see Schoenmaker and Oosterloo, 2005, for details).


Subsidiaries are added by country of residence. Subsequently, these country data are used to calculate the Herfindahl index as a measure of diversification over countries. The Herfindahl index is defined as the sum of the squared shares of assets or revenues in the countries under consideration. It hence ranges from 0 (complete diversification) to 10,000 (no diversification).


Data on branches are missing, which means that the assets of foreign branches cannot be identified as foreign. Typically, foreign branches are smaller than subsidiaries, although there are exceptions. Deutsche Bank’s large-scale operations in London, for example, are organized as a branch.


Some caveats apply. For instance, foreign revenues might be generated using some assets based in the home country, or vice versa. Home-country activities may be disproportionately weighted toward financial market activities that yield small margins on large volumes of assets. Higher revenues abroad could be accompanied by higher costs. And within certain accounting limits, profits are fungible, allowing banks to choose where to book their profits.


Although the standard deviation is an indicator of volatility, and hence risk in this sense, it does not capture risks that might be present in the portfolio but have not materialized.


In addition, financial stability can also be negatively affected when diversification leads to lower capitalization or profitability or to increased earnings volatility.


As in Chapter 6, the estimation is done using the procedure described in Vassalou and Xing (2004) on daily market and annual accounting data.


The portfolio DD can be viewed as a risk profile measure tracking the evolution of the joint risk of failure of the firms in the portfolio (see Chapter 6).


No data are available for WestLB. Hence this bank has not been included in the DD analysis.


Given that the correlations between the cyclical components of bank DD with the cyclical component of GDP growth vary widely across EU countries, it seems unlikely that the similarities in the developments of the DDs of the individual banks are driven by increased synchronicity of real business cycles (see Chapter 6).


For a list of the banks included and for further background information, see De Nicolò and Tieman (2006).


For background information on cluster analysis, see Chapter 8.

    Other Resources Citing This Publication