“A Scorecard for Bank Performance: the Belgian Banking Industry.”

“A scorecard for bank performance:

the Belgian Banking Industry.”

Prof. Dr. André Thibeault, Corneel Defrancq and Jessie Vantieghem

Point of Contact: [email protected] (+32 9 210 9860)

Table of Contents

THE KPMG – VLERICK PRIME FOUNDATION PARTNERSHIP ...... 1 Acknowledgment ...... 3 Executive Summary ...... 5 Introduction ...... 7 Structure of the report ...... 7 Chapter 1: Industry overview ...... 9 Contribution to the Belgian economy ...... 10 Trends over time ...... 10 Savings ratio of households ...... 14 Banking sector’s contribution compared with European peers ...... 16 Internationalisation of the Belgian banking landscape ...... 18 Chapter 2: The value chain ...... 19 Sample selection and segments description...... 20 Sample Selection ...... 20 ECB segmentation ...... 20 Size segmentation ...... 22 Fortis Bank SA/ NV-BNP Paribas Fortis ...... 22 ING Belgium SA/NV-ING ...... 22 BankScope segmentation...... 23 Operational income to total assets ...... 24 Net interest income to total assets...... 27 Net commissions and fees to total assets...... 32 Financial transactions to total assets ...... 36 Operational expenses to total assets ...... 40 Loan loss provisions to total assets...... 45 Conclusion ...... 48 Chapter 3: The financial performance ...... 51 Return on equity ...... 53 Leverage ...... 57 Return on assets ...... 61 Asset yield ...... 65 Profit margin ...... 69

Cost to income ...... 73 Conclusion ...... 78 Chapter 4: The risk-return trade-off ...... 79 Chapter 5: What did we learn in this study? ...... 83 References ...... 85 Appendices ...... 88 Appendix 1: The ratios ...... 88 Appendix 2: The variables ...... 88 Appendix 3: A38 Code ...... 89 Appendix 4: NACE BEL Code ...... 91 Appendix 5: List of abbreviations...... 92 Appendix 6: Contribution in terms of percentage of both the banks individually and per sector. .. 93 Appendix 7: The segmentation between universal and investment banks ...... 95 Appendix 8: The evolution of the natural logarithm of the total assets ...... 96 Appendix 9: Ratios for each bank individually ...... 98 Agricaisse - Caisse Coopérative de Dépôts et de Crédit Agricole ...... 99 Banca Monte Paschi Belgio SA ...... 99 Bank J. Van Breda en Co NV ...... 99 Banque de la Poste-Bank van de Post ...... 99 Deutsche Bank SA-Deutsche Bank NV ...... 99 F. van Lanschot Bankiers Belgie ...... 100 Fortis Bank SA/ NV-BNP Paribas Fortis ...... 100 ING Belgium SA/NV-ING ...... 100 Onderling Beroepskrediet-OBK-Bank C.V.B.A...... 100 Société Générale Private Banking N.V...... 100 Agricaisse - Caisse Coopérative de Dépôts et de Crédit Agricole ...... 101 Banca Monte Paschi Belgio SA ...... 101 Bank J. Van Breda en Co NV ...... 101 Banque de la Poste-Bank van de Post ...... 101 Deutsche Bank SA-Deutsche Bank NV ...... 101 F. van Lanschot Bankiers Belgie ...... 102 Fortis Bank SA/ NV-BNP Paribas Fortis ...... 102 ING Belgium SA/NV-ING ...... 102

THE KPMG – VLERICK PRIME FOUNDATION PARTNERSHIP

This report on the mapping of the Belgian banking industry is part of a 3-year project conducted in the format of a Prime Foundation Partnership between KPMG and the Vlerick Centre for Financial Services.

The research content of the Prime Foundation Partnership will be filled in as follows:

Year 1: Helicopter View – International Scorecard. What is the performance (from many different angles) of the participants in the banking sector and in the insurance sector in Belgium, and how do they rank under these dimensions?

Year 2: Cluster analysis – survey – who performed the best and what drives their performance? What differentiates these banks and these insurance companies? Look at periods before/during/after the crisis. Make links to different dimensions like: distribution channels, business models, asset structure, financial aspects and managerial aspects.

Year 3: Wait to see the results from years 1 & 2 and then identify interesting issues to be investigated further. Points of specific research interests raised by KPMG could also be investigated in the course of this part of the research project.

ACKNOWLEDGMENT

This research project would not have been possible without the support of many people.

The authors would like to convey thanks to KPMG for providing the financial means and specially to Bart Walterus, Partner Financial Services Consulting / KPMG Advisory, who has been providing managerial support and professional advices of great help throughout the report elaboration. Furthermore, the authors would like to thank Erik Clinck, Olivier Macq, Stijn Broekx, Jorn Deneve and Ingrid Stoffels for their valuable feedback at various occasions throughout the duration of the study.

At the Vlerick Centre for Financial Services, Maureen O’Hare has played a major role in the discussions leading to the Prime Foundation Partnership of KPMG and Diane von Teufenstein, as business manager of the Vlerick Centre for Financial Services, has been coordinating this project with a professional and dedicated approach.

The authors,

Prof. Dr. André E. Thibeault Corneel Defrancq Jessie Vantieghem

EXECUTIVE SUMMARY

The goal of this study is to develop a scorecard for assessing bank performance in the Belgian banking industry. The performance will be measured by using a multi-factors approach to make sure performance is assessed under different dimensions and these performance measures will be applied to different segments of the Belgian banking landscape.

The dimensions to be considered are: the value chain, the breakdown of the ROE and the risk-return trade-off.

Four different segments will be defined, based on the size of the banks (defined by their total assets) or based on their type of activities or revenues:

The importance of the financial sector in Belgium is substantial – but it has decreased slightly during the period 2001 – 2010. The financial sector’s contribution to Belgian employees’ employment compensation went from about 6% in 2001 to 5% in 2010. The same trend is observable in the number of people employed in the financial sector. However, while the number of people employed decreased during this period, the average salary steadily increased from €58,000 in 2001 to €70,000 in 2008. After 2008, the financial crisis caused wages to freeze at about €70,000. Another striking feature of the Belgian economy is the savings rate, one of the highest in the world, which fluctuated between 15% and 20% between 2001 and 2010.

When looking more specifically at the Belgian banking sector, one has to realise its international standing. In Belgium, almost half of the total assets (held by Belgian financial institutions) are in the hands of foreign controlled credit institutions.

When breaking down the value chain, it is interesting to note that private banks are consistently exhibiting the highest ratio of operating income to total assets. In the investment / universal banks segment, the investment banks have a better ratio than the universal banks; and when looking at bank size, the smallest banks are achieving the best performance.

When we broke the total operating income into its components, we found that net interest income was negatively correlated with the size of the bank as measured by total assets. As expected, the private banks are dominating the other banks when net fees and commissions are used as the benchmark. When using financial transactions as the benchmark, no conclusion can be drawn, because this value driver is so volatile. On the costs side of the value chain, all of the banks have been able to reduce their operating expenses during the period under study. Private banks have experienced the highest level of operating expenses to total assets (4%), while universal banks have been able to bring this figure down to 1%.

While looking at the performance of the banking industry based on the breakdown of the ROE, the four big banks dominate the market – not because of their good operational profitability (ROA), but because of a bigger risk appetite, which is reflected in a higher leverage compared to the other banks. The comparison of investment banks and universal banks shows that the investment banks perform better in terms of ROE. However, based on their total assets in 2009, the investment banks account for not even 1% of the sample, while the universal banks account for the rest. In the segment based on “BankScope”, the private banks have the best ROE, due to their better performance in operational profitability (ROA).

The analysis of the risk-return trade-off shows a cluster of banks in an area from low to medium risk and from a low to a high ROE. The banks presenting the best risk-return trade-off over the period under study are Banque Delen and Europabank. An interesting point of this risk-return analysis is the inverse relationship for the Big Four banks. ING shows a good risk-return profile, followed by Dexia and KBC with higher risk and lower return, and finally by Fortis with the worst risk-return profile for the period under study.

INTRODUCTION

How to measure performance in the Belgian banking industry? Participants in the financial services industry worldwide have been using many different benchmarks to assess bank performance. The goal of this study is to develop a scorecard for assessing bank performance in the Belgian banking industry. The performance will be measured by using a multi-factors approach to make sure performance is assessed under different dimensions and these performance measures will be applied to different segments of the Belgian banking landscape.

Four different segments will be defined, based on the size of the banks (defined by their total assets) or based on their type of activities or revenues:

 A first segment will compare the banks relative to the source of their income as defined by the ECB (investment banks or universal banks).  A second segment will be based on the size of the banks. Four different categories will be created based on total assets.  A third segment will study Belgium’s Big Four banks individually: Dexia, Fortis, ING and KBC.  The last category is based on the BankScope segmentation. Furthermore, three dimensions will be defined on the basis of a series of variables in order to measure performance. The report will map the Belgian banking sector with regard to these three performance dimensions. These dimensions are:

1) The value chain 2) The financial performance 3) The risk-return trade-off

STRUCTURE OF THE REPORT

To answer the key question – “How to measure performance in the Belgian banking industry?” – we have structured the report as follows:

In Chapter 1, we start by outlining the industry context. In this chapter, we present some well-known figures about the position of the Belgian banking industry in the Belgian and European economies. This chapter sets the context for the remainder of our study.

Chapter 2 details the value chain in the Belgian banking industry. We leave the global view of the industry and move down to the segments level, comparing the value chain of the banks in Belgium (as measured by assets) over the last 10 years. Examining the years 2001 to 2010 allows us to cover the boom years of the early 2000s as well as the credit crunch of 2007 - 2009.

In Chapter 3, we break down each market segment’s ROE according to the Dupont analysis. Thus, the segments are compared on the basis of their profitability – and more precisely, on their capacity to generate income as well as their cost management.

While Chapter 3 compares performance on the basis of the return component, Chapter 4 tackles the combination of both risk and return. The risk-return trade-off will be portrayed by means of graphs.

Chapter 5 presents the major findings of the study.

CHAPTER 1: INDUSTRY OVERVIEW

In order to provide an overview of the industry, we use the two reports of Febelfin, the Belgian Financial Sector Federation (Febelfin, 2010; Febelfin, 2009). Created in 1993, Febelfin consists of five financial trade organisations:

1. ABB-BVB: Belgian Bankers’ and Stockbroking Firms’ Association 2. BEAMA: Belgian Asset Managers Association 3. UPC-BVK: Professional Union of Credit Providers 4. BLA: Belgian Leasing Association 5. BASEM: Belgian Association of Stock Exchange Members

In a first phase, we scrutinise the financial sector’s contribution to the economy in terms of gross added value, employment and employee wages. Secondly, we break down the gross added value according to four distinguishable sub-sectors: (1) insurance corporations and pension funds, (2) financial auxiliaries, (3) other financial intermediaries except insurance corporations, and (4) the central bank and other monetary financial institutions. As a third topic of research, we compare the total assets on the balance sheet of the credit institutions with the GDP of the countries that are part of the European Monetary Union. Fourth, we assess the internationalisation of the industry by ranking the EMU countries according to their respective share of foreign-owned banks.

CONTRIBUTION TO THE BELGIAN ECONOMY

Trends over time To assess the importance of the financial sector in the Belgian economy, we have calculated the Gross Value Added (GVA) of the financial intermediation sector from 2000 up to 2010 (see Figure 1). We clearly see the decline in added value in the years from 2005 to 2008. From 2008 to 2010, we see the shift to previous levels of added value. In 2010, 6.14% of the Belgian economy (as measured by GDP) was built on added value from the financial intermediation sector.

Gross Value Added (GVA) measures the contribution to the economy of each individual producer, industry or sector. The gross value added at basic prices is defined (according to the ESA95 standard) as the output valued at basic prices less intermediate consumption valued at purchaser prices, before deduction of consumption of fixed capital (European Central Bank).

5,90%

5,70%

5,50%

5,30%

5,10%

4,90% 1999 2001 2003 2005 2007 2009 2011

Figure 1: Gross Value Added of the financial sector (as a percentage of the Belgian economy) Source: National Bank of Belgium (Belgostat)

During this same period of time, the total sum of wages paid by the financial service industry (Figure 2) has been relatively in decline in the Belgian economy. In 2010, 4.97% of the total sum paid to employees came from the financial industry.

6,50%

6,00%

5,50%

5,00%

4,50% 1999 2001 2003 2005 2007 2009 2011

Figure 2: Financial sector's contribution to Belgian Employees' employment compensation Source: National Bank of Belgium (Belgostat)

One could conclude that the average salary paid within the financial industry is dropping, but when we also consider the number of people active in the financial industry (see Figure 3), we can counter this conclusion.

2010 2009 2008 2007 2006 2005 2004 2003 2002 2001

130 135 140 145 150

Figure 3: Employment: Number of people employed in the financial industry (x1000) from 2001 to 2010 Source: National Bank of Belgium (Belgostat)

When the total amount of employee compensation1 for the financial industry is divided by the total number of employees in the financial industry, we arrive at Figure 4. This highlights that the average wage has been growing at a fairly slow pace in the 2000 to 2010 timeframe. However, no downward trend is observable.

80000

70000

60000

50000

40000

30000

20000

10000

0 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 4: (Total sum of wages) / (Employment) in the financial industry (€) Source: National Bank of Belgium (Belgostat)

1 This is the total remuneration, in cash or in kind, paid by government to its employees. It includes employers’ actual and imputed social contributions. Employers’ actual social contributions are actual payments into social security schemes and into funded autonomous pension schemes by government on behalf of its employees. Imputed contributions are the estimate of the accruing pension liability in employers’ unfunded defined- benefit schemes (definition Belgostat).

When we compare the different sectors (A38 coded, see Appendix 3), we see that the number of people employed in the financial and insurance sector is roughly the same as the number of people in the accommodation and food service industry.

Mining and quarrying (BB) Manufacture of coke and refined petroleum products (CD) Scientific research and development (MB) Manufacture of computer, electronic and optical products (CI) Manufacture of electrical equipment (CJ) Electricity, gas, steam and air-conditioning supply (DD Manufacture of basic pharmaceutical products and pharmaceutical… Real estate activities (LL) Publishing, audiovisual and broadcasting activities (JA) Water supply; sewerage, waste management and remediation activities… Telecommunications (JB) Manufacture of textiles, wearing apparel and leather products (CB) Manufacture of machinery and equipment n,e,c, (CK) Manufacture of furniture; other manufacturing; repair and installation of… Advertising and market research; other professional, scientific and… Arts, entertainment and recreation (RR) Manufacture of transport equipment (CL) Manufacture of chemicals and chemical products (CE) Manufacture of wood and paper products, and printing (CC) Activities of households as employers of domestic personnel and… Manufacture of rubber and plastics products, and other non-metallic… Computer programming, consultancy and related activities; information… Agriculture, forestry and fishing (AA) Manufacture of basic metals and fabricated metal products, except… Manufacture of food products, beverages and tobacco products (CA) Other service activities (SS) Financial and insurance activities (KK) Accommodation and food service activities (II) Social work activities (QB) Transportation and storage (HH) Construction (FF) Human health activities (QA) Administrative and support service activities (NN) Education (PP) Legal and accounting activities; activities of head offices; management… Public administration and defence; compulsory social security (OO) Wholesale and retail trade, repair of motor vehicles and motorcycles (GG)

0 100 200 300 400 500 600 700

Figure 5: Employment – Belgian Sector Comparison (number of employees, x1000, 2010 data) Source: National Bank of Belgium (Belgostat)

SAVINGS RATIO OF HOUSEHOLDS

When comparing the gross savings ratios of 2010 for the European Union (Figure 6), we see that Belgium has almost the highest gross savings rate of the EU countries concerned2. Only Germany had a higher gross savings rate in 2010. In the periods before, Belgium also had a very high gross savings ratio. Therefore, Table 1 represents the gross savings ratio over the last 11 years for the European countries. This table shows us that Belgium’s gross savings rate has always been above the EU-27 gross savings ratio. Furthermore, of all the countries listed, Belgium, Switzerland and Germany have had the highest gross savings ratio.

Latvia 4,2 Lithuania 7,94 Hungary 8,15 Estonia 9,63 Portugal 9,82 Czech Republic 10,28 Netherlands 10,91 Slovakia 11,26 Finland 11,28 Cyprus 11,55 Italy 12,06 EU (27 countries) 12,31 Norway 12,43 Ireland 13,36 Sweden 13,39 Austria 13,45 France 15,63 Slovenia 15,72 Belgium 16,17 Germany 17,05

0 5 10 15 20

Figure 6: Gross Savings ratio of households in EU, 2010 data Source: Eurostat

2 The gross savings rate of households is defined as gross savings (ESA95 code: B8G) divided by gross disposable income (B6G), with the latter being adjusted for the change in the net equity of households in pension funds reserves (D8net) (Eurostat, 2011). Gross savings is the part of the gross disposable income that is not spent as final consumption expenditure. Detailed data and methodology at: http://ec.europa.eu/eurostat/sectoraccounts.

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 EU (27 countries) 11,59 12,63 12,31 12,15 11,68 11,34 11,05 11 11,5 13,74 12,31 Belgium 16,77 17,87 17,46 16,79 15,51 15,09 15,73 16,4 16,79 18,42 16,17 Czech Republic 11 10,21 10,12 9,05 7,86 9,52 10,62 10,25 9,37 10,66 10,28 Denmark 4,28 9,56 9,51 9,79 6,37 3,71 5,45 4,25 5,03 7,73 : Germany 15,1 15,21 15,71 15,98 16,12 16,28 16,39 16,84 17,41 17,05 17,05 Estonia : : 0,46 -0,52 -5,78 -4,12 -6,26 -1,76 3,42 11,59 9,63 Ireland : : 4,66 6,69 8,65 9,63 8,12 7,59 11,08 14,74 13,36 Spain 11,14 11,08 11,36 11,98 11,28 11,32 11,15 10,7 13,38 18,05 : France 14,1 14,75 16,01 15 15,44 14,39 14,61 15,11 15,26 16,17 15,63 Italy 14,16 16,03 16,82 15,97 16,01 15,83 15,25 14,75 14,56 13,44 12,06 Cyprus 10,82 11,25 9,5 12,18 10,26 10,04 9,97 6,88 4,87 11,38 11,55 Latvia -3,86 -6,12 1,13 3,02 4,32 0,84 -4,11 -5,75 4,91 10,28 4,2 Lithuania 6,11 4,35 4,42 2,42 0,8 0,95 0,88 -4,21 -3,01 7,1 7,94 Hungary 11,5 11,86 10,29 7,9 10,11 11,3 11,96 8,43 8 9,93 8,15 Netherlands 12,11 14,69 13,9 13,1 13,05 12,27 12,19 13,04 12,15 12,95 10,91 Austria 14,09 12,57 12,91 13,66 14,03 14,46 15,13 16,33 16,28 15,67 13,45 Poland 12,27 14,11 10,55 10,12 9,43 9,67 9,81 8,52 3,7 9,91 : Portugal 10,58 10,6 10,3 10,72 9,97 9,97 8,01 6,99 7,08 10,86 9,82 Slovenia 14,63 16,22 16,61 14,38 15,54 17,14 17,53 16,17 15,2 15,01 15,72 Slovakia 10,98 8,96 8,56 6,77 5,92 6,61 5,6 7,23 6,14 7,81 11,26 Finland 8,12 8,1 7,92 8,65 9,83 8,46 6,8 7,24 8,07 11,91 11,28 Sweden 6,94 10,95 10,79 9,82 8,89 8,34 9,42 11,58 13,95 15,56 13,39 United Kingdom 4,67 6,04 4,8 5,12 3,7 3,95 3,45 2,64 2,04 6 : Norway 9,16 8,18 12,75 13,33 11,79 14,5 5,54 6,89 9,15 12,38 12,43 Switzerland 16,86 17,14 16,07 14,82 14,42 15,42 16,64 17,73 16,91 17,12 : Table 1: Gross savings rate for the period 2000-2010 Source: Eurostat

BANKING SECTOR’S CONTRIBUTION COMPARED WITH EUROPEAN PEERS

The following sections of Chapter 1 will focus more specifically on the banking sector (instead of on the financial sector in general).

Figure 7 shows the total assets held by the banking industry relative to the GDP for each country of the European Union at the end of 2010. The total assets held by the Belgian Credit Institutions amount to 3.53 times Belgium’s GDP.

ROMANIA 0,64 POLAND 0,78 SLOVAKIA 0,82 LITHUANIA 0,95 BULGARIA 1,01 CZECH REPUBLIC 1,09 HUNGARY 1,33 LATVIA 1,59 SLOVENIA 1,59 ITALY 1,78 ESTONIA 2,16 GREECE 2,17 FINLAND 2,58 PORTUGAL 3,08 FRANCE 3,49 BELGIUM 3,53 SPAIN 3,65 GERMANY 3,75 DENMARK 3,96 SWEDEN 4,05 AUSTRIA 4,20 NETHERLANDS 4,82 UNITED KINGDOM 6,58 MALTA 8,37 CYPRUS 9,28 IRELAND 9,29 LUXEMBOURG 19,64 0 5 10 15 20 25

Figure 7: Banking sector's contribution to the economy: Consolidated total assets of all Credit Institutions compared to GDP Source: Eurostat

According to ECB data, 106 banks are established in Belgium (end of December 2010, see Figure 8) – 48 of those banks are under Belgian law, the rest operates under foreign law: branches of non-EEA (European Economic Area)-based banks (9), branches of the euro-area-based credit institutions (41) and branches of EEA-based credit institutions (outside the euro area) (8).

branches of EEA-based credit insitutions (outside the euro 8 area) branches of euro-area based credit institutions 48 41 branches of non EEA based banks

9 credit institutions legally incorporated in the reporting country

Figure 8: Number of banks established in Belgium Source: ECB

INTERNATIONALISATION OF THE BELGIAN BANKING LANDSCAPE

When we look at Figure 9, we see that almost half of the total assets (held by Belgian financial institutions) are in the hands of foreign-controlled credit institutions.

ESTONIA CZECH REPUBLIC SLOVAKIA LUXEMBOURG ROMANIA MALTA BULGARIA LITHUANIA FINLAND POLAND LATVIA IRELAND Percentage of Total Assets held HUNGARY by domestic Credit Institutions BELGIUM CYPRUS Percentage of Total Assets held UNITED KINGDOM by foreign-controlled Credit SLOVENIA Institutions AUSTRIA PORTUGAL GREECE DENMARK NETHERLANDS GERMANY ITALY SPAIN FRANCE SWEDEN 0% 20% 40% 60% 80% 100%

Figure 9: Percentage of total assets held by domestic or foreign-controlled Credit Institutions Source: Consolidated Balance Sheet data: 2010, ECB

CHAPTER 2: THE VALUE CHAIN

In this chapter, we focus on the value chain of the Belgian banking sector as presented in Figure 10. The value chain is broken down into three components on the income side (net interest income, net commissions and financial transactions) and into two blocks on the expenses side (operating expenses and loan losses). We first present the break-down based on the segments proposed by the ECB3 (i.e. investment banks and universal banks). Second, we present the break-down by the size (i.e. total assets) of the banks as suggested by KPMG and then, we look more specifically at the Big Four players in Belgium. Finally, we present the break-down based on the segments defined by BankScope.

Figure 10: The Value Chain of a Bank

In the first section, we briefly describe our sample as well as the categorisation of the banks in different segments. The next sections are dedicated to different parts of the value chain, and the last section concludes our discussion.

3 To make a segmentation, we have borrowed a classification that is used in an ECB paper: ‘Beyond ROE- how to measure bank performance?; Appendix to the report on EU banking structures’, September 2010.

SAMPLE SELECTION AND SEGMENTS DESCRIPTION

Sample Selection In the pre-processing phase of this study, we had to define our research sample. Based on both the list of the ‘members of the BVB’ (list from Febelfin) and on BankScope (database managed by Bureau van Dijk), we started gathering data for the banks in Belgium. Consolidated data for Dexia group and KBC group were removed from the sample, since a major part of their business originates from foreign activities. However, we were able to extract data from Dexia Bank Belgium and KBC Bank Belgium, and these are included in the final sample. This resulted in an initial sample of 33 banks which had accounting data available for 2010. As this study examines a time span of 10 years (i.e. 2001-2010), banks with too many missing values were removed from the sample. This yielded a final sample of 23 banks. However, we must point out that the percentage of the total assets in 2010 of our final sample was 92.12% of the total assets of the original sample.

When we limit the time span of our study to the period of 2001-2009, we are able to include six more banks. These banks are: Banque Delen NV (private), Citibank Belgium NV/SA (commercial), Delta Lloyd Bank (commercial), Deutsche Bank Sa (commercial), F. van Lanschot Bankiers Belgie (private) and Société Générale Private Banking (private).

Adding these six banks to our sample means that, for the period from 2001 to 2009, we have data on 29 banks. For the full period 2001-2010, 23 banks are in the sample. Although these six banks contribute only 1.90% of the total assets in the sample for the year 2009, we believe that adding these banks to our sample brings added value to this research. We also need to stress that the four major banks (i.e. BNP Paribas Fortis (former Fortis), KBC, ING and Dexia) together have a percentage of 88.98% of the total assets in 2009. Appendix 6 shows the division of our sample based on total assets in 2009 and per segment.

ECB segmentation The classification of banks can be based on the sustainability of bank revenues. As suggested by a report of the ECB4, banks can be classified on the basis of their share of core banking income, such as net interest income and fees and commissions from on-balance sheet activities, compared to the share of so-called non-recurring revenue activities, such as income from fees not related to balance sheet activities as well as trading income and other one-off gains. Based on the sustainability of bank revenues, we have created two sub-groups: universal banks and investment banks (see Table 2). As mentioned previously, three income drivers can be distinguished in banking: (1) net interest income, (2) net commissions and (3) financial transactions. In this exercise, banks are identified as driven by investment activity when their share of commissions and trading income was higher than the share of net interest income and other income in most periods, meaning that more than 50% of their total income originates from net commissions and financial transactions.

4 ECB paper: ‘Beyond ROE- how to measure bank performance?; Appendix to the report on EU banking structures’, September 2010.

In this categorisation, an investment bank is a financial intermediary that generates more income from non-traditional deposit-to-loans transformation, whereas a universal bank is more active in the banking book activities than in the trading activities of a bank, thus more involved in traditional commercial banking5.

Investment Banks Universal Banks Banque Degroof SA-Bank Degroof NV Agricaisse - Caisse Coopérative de Dépôts et de Crédit Agricole Banque Delen NV (2001-2009) Antwerps Beroepskrediet Byblos Bank Europe SA Argenta Spaarbank-ASPA F. van Lanschot Bankiers Belgie (2001-2009) Banca Monte Paschi Belgio SA Keytrade Bank SA/NV Bank J. Van Breda en Co NV Société Générale Private Banking N.V.(2001- Banque de la Poste-Bank van de Post 2009) BKCP CBC Banque S.A. Centea Citibank Belgium N.V./S.A.(2001-2009) Crédit Agricole SA-Landbouwkrediet Delta Lloyd Bank (2001-2009) Deutsche Bank SA-Deutsche Bank NV (2001-2009) Dexia Bank Belgium-Dexia Bank Ethias Bank Europabank Fortis Bank SA/ NV-BNP Paribas Fortis ING Belgium SA/NV-ING KBC Bank NV Onderling Beroepskrediet-OBK-Bank C.V.B.A. Record Bank SA/NV Santander Benelux SA/NV VDK Spaarbank NV Table 2: Two categories in bank sample: investment banks and universal banks Source: BankScope + KPMG’s own calculations

5 Some potential issues associated with this approach are related to the unstable nature of the mix of these sources of income. For example, if trading income is negative and commission income positive, they can cancel each other, leaving a small proportion for these types of revenue. However, when we use squared shares to check our results, we arrive at a similar outcome, which makes us confident in our conclusions.

Size segmentation A third categorisation in this study is based on the size of the banks. This allows us to investigate whether there are significant differences among banks of different sizes as measured by total assets (TA), with regard to the characteristics under study. The 4 categories are as follows:

 The first category contains the banks with total assets over €100 billion. This category contains the Big Four banks: Fortis, KBC, Dexia and ING.  Our second category contains the banks with total assets between €100 billion and €10 billion. This category includes 4 banks. Our third category also consists of 4 banks. This category contains banks with total assets between €10 billion and €7 billion.  Our fourth category consists of all of the other banks with total assets of less than €7 billion. This category contains 17 banks. Table 3 shows the categories:

TA > €100 billion TA ≤ €7 billion Fortis Bank SA/ NV-BNP Paribas Fortis Delta Lloyd Bank KBC Bank NV Banque Degroof SA-Bank Degroof NV Dexia Bank Belgium-Dexia Bank BKCP ING Belgium SA/NV-ING Citibank Belgium N.V./S.A. €100 billion ≥ TA > €10 billion Bank J. Van Breda en Co NV Argenta Spaarbank-ASPA VDK Spaarbank NV Record Bank SA/NV Agricaisse - Caisse Coopérative de Dépôts et de Crédit Agricole Deutsche Bank SA-Deutsche Bank NV F. van Lanschot Bankiers Belgie Centea Ethias Bank €10 billion ≥ TA > €7 billion Keytrade Bank SA/NV Crédit Agricole SA-Landbouwkrediet Banca Monte Paschi Belgio SA CBC Banque S.A. Banque Delen NV Santander Benelux SA/NV Onderling Beroepskrediet-OBK-Bank C.V.B.A. Banque de la Poste-Bank van de Post Europabank Société Générale Private Banking N.V. Byblos Bank Europe SA Antwerps Beroepskrediet Table 3: Sample of the banks divided by size

In the last step, we examine the details of the four big banks.

Bank TA in 2010 Fortis Bank SA/ NV-BNP Paribas Fortis €347.967 billion KBC Bank NV €276.723 billion Dexia Bank Belgium-Dexia Bank €247.902 billion ING Belgium SA/NV-ING €155.639 billion

BankScope segmentation In a last step of this research, we defined different categories of banks based on the BankScope categorisation. These 4 categories are: (1) Cooperative banks, (2) Commercial banks, (3) Savings banks and (4) Private banks. Table 4 lists the banks divided into the four categories. This categorisation is based on the business model of the banks during the 10-year period under study.

Agricaisse - Caisse Coopérative de Dépôts et de Crédit Agricole Cooperative Antwerps Beroepskrediet Cooperative Argenta Spaarbank-ASPA Saving Banca Monte Paschi Belgio SA Commercial Bank J. Van Breda en Co NV Private Banque Degroof SA-Bank Degroof NV Private Banque de la Poste-Bank van de Post Commercial Banque Delen NV (2001-2009) Private Byblos Bank Europe SA Commercial CBC Banque S.A. Commercial Centea Saving Citibank Belgium NV/SA (2001-2009) Commercial Crédit Agricole SA-Landbouwkrediet Commercial Delta Lloyd Bank (2001-2009) Commercial Deutsche Bank Sa (2001-2009) Commercial Dexia Bank Belgium-Dexia Bank Commercial Europabank Commercial Ethias Bank Saving F. van Lanschot Bankiers Belgie (2001-2009) Private Fortis Bank SA/ NV-BNP Paribas Fortis Commercial ING Belgium SA/NV-ING Commercial KBC Bank NV Commercial Keytrade Bank SA/NV Commercial Onderling Beroepskrediet-OBK-Bank C.V.B.A. Commercial Record Bank SA/NV Commercial Santander Benelux SA/NV Commercial Société Générale Private Banking (2001-2009) Private VDK Spaarbank NV Saving Table 4: Bank sample divided per category based on BankScope categorisation Source: BankScope

OPERATIONAL INCOME TO TOTAL ASSETS

Step 1 of this research investigates the gross or total revenue (measured as the sum of net interest income, net commissions and fees, and income from financial transactions) of the banks divided by total assets. We see the evolution of their ability to generate revenue over the time period under study (2001-2010). In a second step, we investigate the main drivers for the change in operational income (i.e. net interest income, net commissions and fees, or income from financial transactions).

When we use the 2-category categorisation (universal and investment banks), we see that the investment banks generate substantially higher income than the universal banks. In 2001, investment banks reported more than 7.00% for the ratio of operational income to total assets due to the 16.7% figure of Keytrade. The figure of Keytrade dropped in 2002 after the acquisition of RealBank during 2002.

8,00%

7,00%

6,00%

5,00%

4,00% Investment Banks Universal Banks 3,00%

2,00%

1,00%

0,00% 2001200220032004200520062007200820092010

Figure 11: Operational income to total assets between investment and universal banks Source: BankScope

Figure 12 shows the evolution of operational income to total assets split into the 4 size-based categories. It is noteworthy that the banks with total assets less than €7 billion – i.e. the 17 smallest banks – are generating by far the best operational income relative to total assets. While the overall trend during the period under study is a decrease in the operational income to total assets, after 2005 the banks between €7 and €10 billion in assets were able to outperform the bigger banks, except for a small recovery of the biggest banks in 2009 and 2010.

5,00% 4,50% 4,00% 3,50% 3,00% TA > 100 000 2,50% TA > 10 000 2,00% TA > 7 000 1,50% TA < 7 000 1,00% 0,50% 0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 12: Operational income to total assets based on size (million €) Source: BankScope

When investigating the four biggest banks on an individual basis, we see that operational income to total assets for KBC was very volatile over the period under study. In the period 2002-2007, KBC generated the highest operational income to total assets. However, this ratio plummeted in 2008, from 1.68% to 0.56%. The reason for this drop is the devaluation of certain trading products (annual report KBC Bank, 2008). In 2008, Fortis showed a drop in operational income to total assets as well, also associated with write-downs related to the financial crisis. Only ING and Dexia generated extra income per euro of assets in 2008. Dexia’s results were not harmed in 2008. However, after the financial crisis hit the markets in 2008, operational income for Dexia decreased in 2009. As a conclusion, we can say that ING is the only bank of the Big Four that was not affected in their reported results by the financial crisis. In both 2008 and 2009, their operational income even showed an increase. The main drivers for these remarkable results will be discussed in the next sections. At the end of 2010, KBC was again leading the Big Four with the highest operational income to total assets ratio.

3,00%

2,50%

2,00% Fortis

1,50% KBC Dexia 1,00% ING

0,50%

0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 13: Operational income to total assets for Fortis, KBC, Dexia and ING Source: BankScope

When comparing the 4 bank categories (cooperative banks, commercial banks, savings banks and private banks), we clearly see that more income to total assets is generated by private banks than by cooperative, commercial or savings banks. This result might be related to the nature of the activities of a private bank, which requires lower assets than the three other categories. Another noteworthy fact is the overall downward trend in this ratio over the last decade.

It is also noteworthy that private banks are creating twice as much operational income to total assets as savings banks. In 2008, the cooperative banks’ income ratio decreased by approximately 25%: in 2007, their operational income ratio was 2.31%, declining to 1.74% in 2008. Recovery is observable in 2009 and ends with the second highest operating income in 2010. For the other types of banks, operational income to total assets decreased during the financial crisis but not as much as the trend observed in the cooperative banks. When drawing conclusions based on the BankScope segmentation, one has to realise that there are only 2 cooperative banks in the sample and only 4 savings banks.

6,00%

5,00%

4,00% Coop. Bank

3,00% Comm. Bank Savings Banks 2,00% Private Banks

1,00%

0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 14: Operational income to total assets based on the BankScope segmentation Source: BankScope

NET INTEREST INCOME TO TOTAL ASSETS

All the banks in our sample act as financial intermediaries: the bank generates profits by transforming deposits into bank loans. This is the first component of the value chain: net interest income. Net interest income (NII) is measured as the difference between revenues generated by interest-bearing assets and the cost of paying interests on deposits. To assess the importance of financial intermediation relative to the other components of the value chain on the income-side, we compare the different components of the value chain by dividing each of them by total assets.

When explaining the change in net interest income, the maturity structure of the assets and liabilities must be considered as well. Generally speaking, the average maturity on the asset side is larger than the maturities measured on the liability side. This implies that changes in market interest rates have an

almost immediate influence on interest expenses. This could imply that, when interest rates increase, NII decreases relatively6. In addition, recent literature (Bolt et al, 2010) suggests that bank’s lending history should be taken into account when discussing NII in times of financial distress. Long-term interest rates from previous years, characterised by economic growth (resulting in higher lending activity), are found to be important determinants of Interest Income. However, the effect of long-term interest rates on net interest income declines in time, due to repayments of loans as well as the growth of the loan portfolio over time.

The division between universal and investment banks shows that universal banks are generating most of their profits from financial intermediation, which could be expected given the definition of universal banks. The NII to TA for universal banks decreases for most of the period under study, mainly due to an increasing trend in total assets. Looking at the investment banks, no general trend can be found. Note that the gap between the two types of banks is decreasing.

2,50%

2,00%

1,50% Investment Banks

1,00% Universal Banks

0,50%

0,00% 2001200220032004200520062007200820092010

Figure 15: Net interest income to total assets between investment and universal banks Source: BankScope

Figure 16 shows that the four biggest banks generated less net interest income to total assets than the three other categories of banks, except in 2009-2010, when they were slightly higher than the banks between €10 and €100 billion in assets.

Note that the 17 smallest banks generated the highest net interest income to total assets during the entire period. However, the gap between the smallest banks and the third category (between €7 and €10 billion in assets) was very small in 2007. As we go from 2001 to 2010, the four categories of banks move closer to each other. This might reflect the greater interest of the big banks in financial intermediation as the back-to-basics strategy in the aftermath of the financial crisis.

6 This relationship has been well-known for a long time (e.g. Préfontaine and Thibeault, 1993; English, 2002). 28

3,00%

2,50%

2,00% TA > 100 000

1,50% TA > 10 000 TA > 7 000 1,00% TA < 7 000

0,50%

0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 16: Net interest income to total assets based on size (million €) Source: BankScope

When looking at the four biggest banks individually, one can observe a counter-cyclical trend compared to economic growth (we used the Bel-20 as our business cycle indicator). This trend is especially observable for KBC as can be seen in Figure 17. This result contradicts the findings of Albertazzi and Gambacorta (2008). They found that there is a positive relationship between net interest income and the business cycle, represented by GDP.

2,00% 5000 4500 4000

1,50%

3500 3000

1,00% 2500 20 20 (value)

2000 - Bel KBC NII/TA KBC NII/TA (%) 1500 0,50% 1000 500 0,00% 0

Figure 17: Evolution of the NII to total assets from KBC compared to the BEL-20 stock index Source: DataStream (Bel-20) and BankScope (KBC NII/TA)

In 2009, KBC’s Net Interest Income to total assets increased most in comparison with the other years. According to KBC, this was mainly attributable to a stronger performance in deposit and loan margins7 (following the exceptionally high rates of interest paid on deposits since the summer of 2008), and the shift on the deposit side towards products that generate a higher margin. These factors caused the net interest margin to widen from 1.47% in 2008 to 1.58% in 2009. The group’s total volume of loans and deposits in Belgium ended 3% higher and 6% lower, respectively, than in 2008 (KBC annual report 2009, page 29).

When looking at the NII/TA ratio for Fortis, we see an increase of 24.29% in 2009, and an increase of 21.83% in 2010. This is again a consequence of the favourable funding conditions. In 2010, the NII to TA increased further, indicating the same trend as observed for KBC.

During the period 2001-2004, ING’s NII/TA remained stable. In 2005-2007, their income generated by transforming deposits into loans decreased. This decrease was mainly due to a significant increase in their total assets, as they were generating more net interest income. In 2008, NII to TA increased again. Although the TA of the bank decreased (-2.28%), the bank was able to generate more net interest income (+11.98%). Despite difficult Belgian market conditions, ING was able to attract more deposits (+7.1%) and sell more loans (+11%) (Press Release ING Belgium 2008), resulting in a better ratio for 2008. In 2009, ING was able to publish even better NII/TA results. The better market conditions (i.e. more convenient spread between deposits and loans) and an increase in the credits that were granted (more than €4 billion new credits, press release ING Belgium 2009) were the main drivers for the increase in NII to TA.

Dexia shows the same trend as can be observed for ING. In the period 2004-2007, their TA increased more than their NII, resulting in a decreasing trend. In both 2009 and 2010, their NII decreased more than the decline in total assets – therefore, a negative trend in the ratio was observed.

2,50%

2,00%

Fortis 1,50% KBC 1,00% Dexia ING 0,50%

0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 18: Net interest income for Fortis, KBC, Dexia and ING Source: BankScope

7 In 2009, the base rate paid for traditional savings deposits fell to 1%, down from its record level of 4% in the summer of 2008. (KBC annual report 2009, page 20). 30

Figure 19 shows that NII to TA for the cooperative banks is the highest of the whole sample, except in 2008 when this category of banks fell just below the commercial banks after a drop of 16.53% in their net interest income to total assets. The main driver for this decrease is the NII to TA ratio of ‘Antwerps Beroepskrediet’. As there are only 2 cooperative banks in the sample, a big change in the ratio of one of these banks has a substantial influence on the group. The net interest income to total assets for the commercial banks followed the same trend as the trend observed for the cooperative banks. However, their NII to TA was more stable during the entire period under study. Looking at the savings banks, we see that their NII to TA did not change that much. In both 2007 and 2008, a slight decrease can be observed. Starting in 2001 up until 2007, the NII to TA for private banks decreased. In 2008, a swing in the trend can be observed, ending with a big increase in 2010.

3,00%

2,50%

2,00% Coop. Banks

1,50% Comm. Banks Savings Banks 1,00% Private Banks

0,50%

0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 19: Net interest income to total assets based on the BankScope segmentation Source: BankScope

NET COMMISSIONS AND FEES TO TOTAL ASSETS

Financial mediation8 is another way banks generate income. Dexia defines net commissions and fees as commissions arising from negotiating, or participating in the negotiation of, a transaction for a third party, such as the arrangement for selling loans, equity securities or other securities, or the purchase or sale of businesses (Dexia Bank Belgium Annual Report, 2009, page 62). Asset management operations generate fee-business as well. The income from asset management principally originates from unit trusts, mutual fund management, administration fees and loan commitment fees.

When dividing the banks into the two segments – investment banks and universal banks – we can see the clear dominance of investment banks with regard to net fees and commissions to total assets, which is in line with the definition of these investment banks. In 2003 and 2008, their income was at its lowest. For investment banks, the main part of their income originates from asset management. Lower consumer confidence resulted in a lower transaction volume, which is reflected in a decrease in the level of the securities brokerage fees.

3,50%

3,00%

2,50%

2,00% Investment Banks 1,50% Universal Banks

1,00%

0,50%

0,00% 2001200220032004200520062007200820092010

Figure 20: Net commissions and fees to total assets between investment and universal banks Source: BankScope

8 Financial mediation refers mainly to off-balance sheet activities: i.e. activities that do not require the transformation of deposits into loans. 32

Figure 21 shows the evolution of net commissions and fees to total assets based on the size of the banks. This graph shows us immediately that both the Big Four and the 17 smallest banks are the only two groups who never had negative net commissions income. It is also noteworthy that the smallest banks generate the best net commissions and fees to total assets. Their income originating from fee business is even more than twice as high as the Big Four banks. The negative net commissions and fees to total assets for the second category of banks is due to the fact that, in this group, only Deutsche Bank recorded positive net commissions and fees during these years.

1,60% 1,40% 1,20% 1,00% TA > 100 000 0,80% TA > 10 000 0,60% TA > 7 000 0,40% TA < 7 000 0,20% 0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 -0,20% -0,40%

Figure 21: Net commissions and fees to total assets based on size (million €) Source: BankScope

When investigating the four biggest banks individually, we see that the fees and commissions income for ING is relatively stable over time, even at the time when the financial crisis fully impacts the economy. However, in 2008 there was a slight decrease in this source of income for ING. According to ING, this decrease was driven by lower asset management fees and lower income from their securities business (annual report ING Bank 2008, page 5). In 2009, their commissions income increased, thanks to fees earned on waivers and restructuring (annual report ING Bank 2009, page 8).

In 2008, KBC and Fortis reported a sharp decrease in fees and commissions income. According to KBC, this decrease was an effect of the crisis: reduced sales of investment funds and unit-linked life insurance products, which were reflected in lower net fees and commissions income. Due to the financial and stock market crisis, investors turned their backs on investment funds and certain investment-type insurance, opting instead for traditional deposit products. As a result, fee commission income declined. Conversely, fees and commissions paid (primarily to insurance agents) went up, due to increased sales of insurance and new acquisitions. In 2008, total assets under management by the group decreased by approximately 10% to €207 billion, which was attributable entirely to the decline in value of the assets under discussion (-12 percentage points) and a limited inflow of assets (+1 percentage point) (Annual report KBC 2008, pages 18 and 21). In 2010, net commissions were back to the pre-crisis level for most of the banks.

Dexia’s net commissions income is noticeably lower compared to the other big banks (except for the year 2001). The substantial decrease in 2002 is mainly due to off-balance sheet products, such as mutual funds, which generated less income in the context of unattractive capital markets. In addition, the stock exchange orders and commissions linked to debit cards decreased (Dexia annual report 2002, page 63). In 2008, their net commissions started to decrease slowly. This is due mainly to the drop in commissions received on unit trust funds (Dexia Bank Belgium annual report 2008, page 157). In 2009, their net commissions decreased further as a result of the drop in commissions received on unit trust funds and insurances, which was partly offset by the positive impact of the non-renewal in 2009 of commissions paid on credit derivatives in 2008 (Dexia Bank Belgium annual report 2009, page 181).

0,70%

0,60%

0,50% Fortis 0,40% KBC 0,30% Dexia

0,20% ING

0,10%

0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 22: Net commissions and fees to total assets for Fortis, KBC, Dexia and ING Source: BankScope

When dividing our sample per sector, the private banks clearly dominate the industry with regard to net fees and commissions to total assets. In general, they generate higher income to total assets than those involved in pure financial intermediation9. Since financial intermediation is far more important for commercial banks, it is normal that their income generated by mediation amounts to only 10% of their total income. Both the cooperative and the savings banks generated negative income. This is probably due to the commissions paid for the distribution of some products.

4,00% 3,50% 3,00% 2,50% Coop. Banks 2,00% Comm. Banks 1,50% Savings Banks 1,00% Private Banks 0,50% 0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 -0,50% -1,00%

Figure 23: Net fees and commissions to total assets based on the BankScope segmentation Source: BankScope

9 Financial intermediation refers to the transformation on the balance sheet of deposits into loans. 35

FINANCIAL TRANSACTIONS TO TOTAL ASSETS

Another non-interest income for banks is the income generated by financial transactions (trading activities).

Comparing the investment banks and the universal banks, we see more or less what we expected. The investment banks generated more income from financial transactions to total assets than the universal banks. However, in 2009, with the full impact of the financial crisis, both types of banks reduced their exposure substantially, which resulted in small losses for both.

0,50%

0,40%

0,30%

0,20% Investment Banks

0,10% Universal Banks

0,00% 2001200220032004200520062007200820092010 -0,10%

-0,20%

Figure 24: Financial transactions to total assets between investment and universal banks Source: BankScope

Figure 25 shows that the income from financial transactions for all banks is very low in Belgium compared with the other types of income. Until 2005, the smallest banks in the sample generated more income from financial transactions compared with the other banks. In 2006, the Big Four banks gained ground on the smallest ones, generating more income on the financial markets due mainly to KBC’s speculative portfolio. In 2007, they were able to generate the most income compared with the others. However, in 2008 and 2009, they suffered important losses in their trading portfolios. In 2008, the others managed to generate more income with financial transactions than the year before. But in 2009, their trading income reduced compared with level of the year before, and they even made a loss. So the smaller banks were also influenced by the financial market turmoil that started in 2007 in the United States and reached Belgium in 2008.

0,30%

0,20%

0,10%

0,00% TA > 100 000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 TA > 10 000 -0,10% TA > 7 000 -0,20% TA < 7 000

-0,30%

-0,40%

-0,50%

Figure 25: Financial transactions to total assets based on size (million €) Source: BankScope

In the first years under study, both KBC and ING generated more income with financial transactions than Fortis and Dexia. In the following years, we can see KBC’s dominant position, first positive and then negative in 2008 and 2009, before exiting their speculative position. In 2005, Dexia was the first one to report losses on their financial assets. In 2006, the difference between the other banks and KBC became bigger; KBC generated 0.67% of its income by financial transactions to total assets. The discrepancy between the other banks and KBC was also high in 2008 and 2009. This time, however, KBC generated losses in their financial assets.

Only ING succeeded in generating a positive value during the whole period under study. Their income on financial assets remained stable over time. In the years before the financial crisis, this type of income for ING was rather low compared with the other big banks, but the financial income in 2008 - 2009 did not decline in contrast to the other banks. This indicates that ING invests in less risky products, as their income from financial transactions did not fluctuate very much.

1,00%

0,50%

Fortis 0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 KBC Dexia -0,50% ING

-1,00%

-1,50%

Figure 26: Financial transactions to total assets for Fortis, KBC, Dexia and ING Source: BankScope

Using BankScope criteria to segment the banks, we conclude that savings banks can be seen as less risky, because their income from financial transactions is nearly zero (with their biggest loss reported in 2006). With regard to the cooperative banks, their income was very volatile, and they generated losses in 2002, 2008 and 2009. In contrast, commercial and private banks show a more or less stable evolution. Starting in 2008, this evolution changed and the commercial banks started to generate losses. In 2009, they were all making losses – an indication of the bad economic environment. In 2010, cooperative banks and commercial banks were back with positive results.

1,20%

1,00%

0,80%

0,60% Coop. Banks Comm. Banks 0,40% Savings Banks 0,20% Private Banks

0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 -0,20%

-0,40%

Figure 27: Financial transactions to total assets based on the BankScope segmentation Source: BankScope

OPERATIONAL EXPENSES TO TOTAL ASSETS

When dividing our sample into investment banks and universal banks, the operational expenses10 to total assets are higher for investment banks than for universal banks during the entire period. For universal banks, the ratio is stable during the period 2001-2007. Starting in 2008, the operational expenses to total assets decreased slightly. For investment banks, the ratio’s downward trend is more accentuated over the period under study, with only a slight increase in 2008 and 2009. This increase in the ratio for the investment banks is due to a small decrease of about 2% in total assets, while the assets of the universal banks grew by 6% from 2008 to 2009.

6,00%

5,00%

4,00%

3,00% Investment Banks Universal Banks 2,00%

1,00%

0,00% 2001200220032004200520062007200820092010

Figure 28: Operational expenses to total assets between investment and universal banks Source: BankScope

Looking at the operational expenses relative to the size of the banks, we see that, in general, the operational expenses to total assets have decreased over time. However, in 2008 there was a slight increase for every category except the second one, total assets between €10 and €100 billion. As the economic environment has been difficult since the end of 2007, for banks it has been more difficult to keep their operational expenses at the same level. The growth of the Belgian economy slowed down significantly in 2008. GDP fell in the fourth quarter, accompanied by an erosion in investor confidence. Additionally, the inflation rate, which peaked in July 2008 (+5.9%) and was sustained by the rise in oil prices, came down to a more reasonable, though still high, level of 2.9% in December 2008. This contributed substantially to an increase in operating expenses (Financial report 2008 ING Bank Belgium, page 5).

10 In “BankScope”, operational expenses are labelled “overheads”. 40

Another noteworthy fact is that the smallest banks have by far the highest operational expenses compared to their total assets. Also noteworthy is that the biggest banks do not have the lowest operational expenses to total assets ratio, which could have been expected due to the benefits of economies of scale. However, compared to the others, their operational expenses to total assets are low, but the second category (TA > €10 billion) succeed in having the lowest costs – especially in the years 2008-2010, when their operational expenses to total assets decreased substantially, and the operational expenses to total assets for the Big Four increased slightly.

Also worth highlighting is the fact that, in 2005, the operational expenses for the third category (TA < €7 billion) increased significantly, while for the others the ratio decreased. The main reason for this increase was the change in operational expenses for Santander Benelux. From 0.73% in 2004, the ratio increased to 2.13% in 2005. Their financial statements show that operational expenses increased, but not as much as the decrease in total assets.

4,00%

3,50%

3,00%

2,50% TA > 100 000

2,00% TA > 10 000 TA > 7 000 1,50% TA < 7 000 1,00%

0,50%

0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 29: Operational expenses to total assets based on size (million €) Source: BankScope

Figure 30 exhibits the high volatility in the operational expenses to total assets ratio for the four big Belgian banks, except for Dexia, which shows a steady decrease in the ratio from 2001 to 2007, to stabilise after that. KBC’s operational expenses to total assets ratio increased in both 2008 and 2009. According to KBC, the increase in 2008 was not only due to the economic conditions but also to one-off items such as new early retirement provisions and increased pension fund provisions (Annual Report KBC 2008, page 27). In 2009, the operational expenses for KBC actually shrink but less than the decline in total assets which was -11.6%; therefore, an increasing trend can be found in the ratio. This decline in operational expenses is a consequence of the lower number of employees due to the downturn of investment banking activities (KBC annual report 2009, page 39). ING’s operational expenses to total assets also increased from 2007 to 2009. ING divides their banking activities into two segments: retail banking and wholesale banking. They state that the retail part is responsible for substantially higher investments in marketing and customer services, whereas the wholesale segment needed investments to strengthen the front office in 2008. (Financial Report 2008 ING bank Belgium, page 6). In 2009, their operational expenses to total assets increased; however, their operational expenses decreased. The

ratio increased because the total assets decreased more than the operational expenses did. This decrease in operational expenses was due to the positive impact of cost containment initiatives, with all business lines contributing to the decline (Financial Report ING Bank 2009, page 5).

Fortis exhibits a high volatility in the operational expenses to total assets ratio over the time period under study. Here also, the changes in total assets need to be taken into account to interpret the situation correctly. In 2008, Fortis’ total assets decreased by 23.5%, but at the same time, their operational expenses decreased by 34.9%, pushing down the ratio of operational expenses to total assets. In 2010, Fortis’ total assets increased again.

Looking at the evolution of Dexia, we see that their operational expenses to total assets remained remarkably stable over the last few years. During the period 2001-2006, their operational expenses decreased from 1.35% to 0.7%. From then on, the ratio remained stable.

2,00% 1,80% 1,60% 1,40% 1,20% Fortis 1,00% KBC 0,80% Dexia 0,60% ING 0,40% 0,20% 0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 30: Operational expenses to total assets for Fortis, KBC, Dexia and ING Source: BankScope

Looking at the sub-categories based on BankScope, we can also see a generally decreasing trend in operational expenses to total assets except for the cooperative banks. One can see that the private banks have a much higher operational expenses to total assets ratio compared to the other categories. As stated earlier, this might be due to the fact that private banking activities require fewer assets than other banking activities. An interesting finding is that savings banks achieved the lowest operational expenses to total assets in the industry.

4,50%

4,00%

3,50%

3,00% Coop. Banks 2,50% Comm. Banks 2,00% Savings Banks 1,50% Private Banks 1,00%

0,50%

0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 31: Operational expenses to total assets based on the BankScope segmentation Source: BankScope

As the operational expenses to total assets ratio is the focus of many banks, Table 5 shows this ratio averaged over the period 2008-2010 for all the banks. Santander Benelux outperforms the other banks by far.

Overhead to Bank total assets Agricaisse - Caisse Coopérative de Dépôts et de Crédit Agricole 0,41% Antwerps Beroepskrediet 1,45% Argenta Spaarbank-ASPA 0,36% Banca Monte Paschi Belgio SA 1,12% Bank J. Van Breda en Co NV 1,65% Banque de la Poste-Bank van de Post 0,80% Banque Degroof SA-Bank Degroof NV 3,41% Banque Delen NV 3,07% BKCP 1,96% Byblos Bank Europe SA 1,61% CBC Banque S.A. 1,56% Centea 0,79% Citibank Belgium N.V./S.A. 6,43% Crédit Agricole SA-Landbouwkrediet 1,86% Delta Lloyd Bank 2,42% Deutsche Bank SA-Deutsche Bank NV 0,96% Dexia Bank Belgium-Dexia Bank 0,67% Ethias Bank 0,77% Europabank 4,48% F. van Lanschot Bankiers Belgie 1,31% Fortis Bank SA/ NV-BNP Paribas Fortis 1,08% ING Belgium SA/NV-ING 1,20% KBC Bank NV 1,43% Keytrade Bank SA/NV 1,26% Onderling Beroepskrediet-OBK-Bank C.V.B.A. 1,74% Record Bank SA/NV 0,65% Santander Benelux SA/NV 0,07% Société Générale Private Banking N.V. 5,78% VDK Spaarbank NV 1,11% Table 5: Three-year average of the operational expenses to total assets ratio for the period 2008-2010

LOAN LOSS PROVISIONS TO TOTAL ASSETS

Another cost which can be categorised is provisions for loan losses. These provisions are used as a buffer against expected losses in the bank’s loan portfolio.

A lot of research has already been conducted on the evolution of loan loss provisions and the business cycle. All of the studies (e.g. Laeven and Majnoni, 2003; Bikker and Metzemakers, 2005; Dinamona, 2008) have come to the conclusion that, when cyclical downturns set in, banks tend to delay provisioning for bad loans, possibly magnifying the impact of the economic cycle on the banks’ income and capital. This implies that banks fund the losses on loans in periods of economic downturn more than they do in periods of economic upswing – i.e. loan loss provisions display counter-cyclical behaviour. This is consistent with the shape of Figure 32: in the period 2001-2003, loan loss provisions to total assets are much higher than in the period 2004-2007. In 2008, the loan loss provisions are increasing, and they peak in 2009.

Looking at the universal / investment banks sample, we can see the counter-cyclical trend. As expected from their business segment, the universal banks have more provisions for loans than investment banks. In 2009, the investment banks were starting to recover from the financial crisis, while the universal banks started to be affected by the global economic crisis.

0,35%

0,30%

0,25%

0,20%

0,15% Investment Banks 0,10% Universal Banks

0,05%

0,00% 2001200220032004200520062007200820092010 -0,05%

-0,10%

Figure 32: Loan loss provisions to total assets for investment banks and universal banks Source: BankScope

Figure 33 shows that the loan loss provisions to total assets ratio displays counter-cyclical behaviour in every category. This counter-cyclical behaviour is more accentuated for both the Big Four and the smallest banks in our sample.

0,40% 0,35% 0,30% 0,25% TA > 100 000 0,20% TA > 10 000 0,15% TA > 7 000 0,10% TA < 7 000 0,05% 0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 -0,05% -0,10%

Figure 33: Loan loss provisions to total assets ratio based on size (million €) Source: BankScope

Our four big banks individually show the same counter-cyclical behaviour to different extents. KBC and, to a lesser extent, Fortis exhibit this counter-cyclical behaviour. An interesting point is that the LLP to TA ratio for ING in 2005 is even negative, which means that they are taking back money from their loan loss provisions. In 2008-2010, the LLP to TA ratio for KBC is the highest of the Big Four. In 2010, the buffers for expected losses for both Dexia and Fortis were very low. This is somewhat understandable in the case of Fortis, as in 2009 their provisions were very high. For both Dexia and ING, the pattern of LLP to TA is more stable than for Fortis and KBC.

0,80%

0,70%

0,60%

0,50% Fortis 0,40% KBC 0,30% Dexia ING 0,20%

0,10%

0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 -0,10%

Figure 34: Loan loss provisions to total assets ratio for Fortis, KBC, Dexia and ING Source: BankScope 46

The BankScope segmentation (see Figure 35) shows that, in general, commercial banks have a higher LLP to TA ratio, which makes sense given the nature of their business. Interesting to note: in the period 2003-2006, the LLP to TA ratio for the cooperative banks was negative, which was turned around substantially in 2009 and 2010. In 2008, the LLP to TA ratio increased for all types of banks, except for the savings banks. In 2008, the savings banks even decreased their loan loss reserves, with negative LLP. In 2009, the LLP are increasing further, except for the private banks. In 2010, the LLP to TA ratio decreases again, except for the private banks, whose ratio is now comparable to commercial banks and savings banks.

0,40%

0,30%

0,20% Coop. Banks 0,10% Comm. Banks

0,00% Savings Banks 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Private Banks -0,10%

-0,20%

-0,30%

Figure 35: Loan loss provisions to total assets ratio based on the BankScope segmentation Source: BankScope

CONCLUSION

Studying the value chain of the Belgian banks during the period 2001-2010 in detail yields some interesting findings. First, on the income side: when comparing investment banks and universal banks – except for 2001, when investment banks attained a ratio of more than 7.0%, and in 2010 when they reached only 3.0% – the investment banks are performing twice as well as the universal banks. When comparing banks based on their size, the smallest banks generate an operating income to total assets ratio twice the one of the three other categories, except for 2010. When comparing the four largest banks, KBC stands as the leader except for 2008 and 2009. KBC also shows the largest variation in this ratio, from just above 0.5% in 2008 to 2.5% in 2010. For the period under study, the private banks clearly out-perform the other banks with a ratio of about 4.5% for the period under study. The cooperative banks and the commercial banks stand just above 2%, while the savings banks reach just a little more than 1.0%.

Breaking down the operating income into its three components – net interest income, net fees plus commissions and financial transactions – we find that in the case of net interest income reflecting financial intermediation activities, most of the time the universal banks generate twice as much net interest income than the investment banks do. When looking at the size of the bank and the capacity to generate net interest income compare to total assets, it is interesting to notice the negative correlation between the size in assets and the ratio of net interest income to total assets. This low net interest margin for the big banks in Belgium is reflected in an average ratio between 1.0% and 1.5 % for the period under study, with a peak above 1.5% in 2009 and 2010 for KBC and ING. These findings are fully consistent with the paper by Bos et al (2006) from the Dutch Central Bank. From this financial intermediation focus, in most of the years, cooperative banks and commercial banks out-performed savings banks and private banks.

The second value driver on the income side – net fees and commissions – shows the net dominance of the private banks in the fees and commissions business. The private banks in our sample were able to generate fees and commissions to total assets between 3.0% and 3.5%, while this figure fluctuates between about plus 30 basis points to minus 30 basis points for the other banks. In our comparison of investment banks with universal banks, the dominance of investment banks with a ratio between 3.0% and 3.5% is very clear, compared to universal banks with a ratio close to 0.0%. When looking at the size of the banks, only the very big ones and the very small ones generate significant income from fees and commissions. The banks in between do not make money at all, or they lose a little. However, the net fees and commissions to total assets ratio of the very small banks is about 3 times that of the big banks. Among these big banks, we can observe except for 2001 a high volatility in their results, with KBC clearly leading from 2005 to 2010, followed by Fortis and ING and, far behind by Dexia.

The last value driver on the income side exhibits a very high volatility. Surprisingly enough, the most volatile banks are the cooperative banks, with two peaks in financial transactions to total assets in 2005 and 2007. However, the results of this category is driven by only 2 players. As expected, positive results, mainly from the private banks, are followed by negative results in 2009 and 2010 in the aftermath of the financial crisis. We can observe the same pattern with the investment banks and the universal banks, with the investment banks’ share of this market being more than 3 times that of the universal banks. When categorising banks relative to their total assets, we can observe a very high 48

volatility in the financial transactions to total assets ratio. This volatility is mainly due to the results of the very small banks and of the largest banks from 2001 to 2007. The big banks’ large losses in 2008 and 2009 impact significantly the volatility of financial transactions to total assets. For the big banks, these results were mainly driven by the results of KBC during those years.

Second, on the cost side: we have looked at the evolution of operational expenses and loan loss provisions compared to total assets. For the operational expenses to total assets ratio, in the case of the investment banks, we can observe a significant decrease from more than 5.0% in 2001 to less than 2.5% in 2007. After 2007, their ratio fluctuates between 3.0% and 2.0%. For the universal banks, the ratio decreases steadily from almost 2.0% in 2001 to about 1.0% in 2010. For the period under study, all the banks have been able to reduce their operational expenses relative to their total assets. The smallest banks show the biggest improvement, but they are still 50% above the other banks in 2010. Except for the banks between €10 billion and €100 billion in assets, it seems that, once a bank achieves a ratio of 1.0%, the cost reduction has reached its floor. However, among the big banks, we can observe large fluctuations in the ratio of operational expenses to total assets except for Dexia that stabilises after 2005. The private banks have the highest costs over the whole time period, with a ratio fluctuating between 4.0% and 2.5% reflecting relatively smaller balance sheets than the other categories. The commercial banks follow, with a ratio decreasing almost steadily from above 2.5% to less than 1.5%. For the cooperative banks as well as the savings banks, the ratio is quite stable around 1.0%

Our last value driver, loan loss provisions to total assets shows large fluctuations everywhere. This can be attributed to the banks’ different provisioning strategies and/or to different exposures.

CHAPTER 3: THE FINANCIAL PERFORMANCE

In this chapter, we focus on the profitability of the Belgian banking industry during the past decade. We look at the return on equity ROE), which is the most used benchmark to measure profitability from a shareholder point of view. Then, in a first level of analysis, we break down the ROE into two components: namely, the return on assets (ROA) and the leverage (L). In a second level of analysis, we split the return on assets (ROA) into the asset yield (AY) and the profit margin (PM). This breakdown is called the Dupont analysis (see Figure 36). Finally, we supplement the profit margin with an efficiency measure (i.e. cost to income ratio), and we conclude the analysis of the financial performance of the Belgian banking industry.

Figure 36: The Dupont Analysis

Figure 36 gives us the mathematical relationship between the components of the Dupont system. For example: multiplying the ROA by the leverage, the asset component disappears, leaving us with the ROE. While this is interesting from a mathematical point of view, we can derive much more information by looking at the Dupont system from a managerial point of view.

The first level of analysis tells us that the return to the shareholders is composed of the ROA, which reflects the profit derived from the bank’s overall operations, times a factor called leverage, which reflects the structure of the bank’s balance sheet. While the ROA can be defined as the operational profitability of the bank, the leverage, Asset / Equity, reflects the willingness to take solvency risk. Thus, a bank can improve the return to its shareholders by leveraging itself, which means taking more risk.

The second level of analysis links the asset yield (AY) and the profit margin (PM) to the operational profitability, the ROA. Looking at Figure 36, one can see that multiplying AY by PM cancels out total revenue to yield the ROA.

While looking at asset yield (AY), total revenue / total asset, we are measuring the proportion of gross revenue (as presented in a slightly different way in Chapter 2) to total assets. Notice that, in this current definition of total revenue, we do not subtract the interest expenses and the fees and commissions paid. Thus, AY reflects the capacity of the bank to generate income. Among other things, this capacity to generate income is related to the type of balance sheet and off-balance sheet activities carried out by the bank, as well as its pricing policy.

The profit margin (PM), net profit / total revenue, relates to the way net profit compares to total revenue. The difference between these two figures is a series of various costs. Thus, PM may be associated with the way the bank manages its costs, given a certain level of gross revenue.

Thus, in this chapter, we are addressing the profitability of the Belgian banking industry from an ROE perspective, which encompasses: operational profitability, ROA, the bank’s willingness to take risk, leverage (solvency risk), the capacity of the bank to generate income, and the bank’s costs management.

RETURN ON EQUITY

Return on equity – measured as net income to equity – gives an indication of how much profit the bank is able to generate to remunerate its stockholders. In other words, the ROE shows the return for the shareholders before the decision to give dividends or to retain earnings. The ROE can be seen as one of the most important indicators of a bank’s profitability and potential growth.

Figure 37 looks at the evolution of the ROE for the investment banks and universal banks categories. The graph shows the same evolution for the two types of banks; however, the ROE for the investment banks is remarkably higher than that for the universal banks. This graph shows a higher ROE in good economic conditions and a lower ROE during economic slowdowns (as shown by the BEL-20 in Figure 17). However, the most surprising results are the ones for the Belgian investment banks during the financial crisis of 2008 – 2009. While their ROEs declined during these two years, they were still able to achieve a healthy 15% and 10% in 2008 and 2009, respectively.

25,00%

20,00%

15,00%

10,00% Investment Banks

5,00% Universal Banks

0,00% 2001200220032004200520062007200820092010 -5,00%

-10,00%

Figure 37: Return on equity (ROE) for investment banks and universal banks Source: BankScope

Looking at Figure 38, the four biggest banks are the most profitable under good economic conditions. However, in both 2008 and 2009, they generated a negative ROE. In 2008, they even show a loss of 38.4% – which implies that, for every €100 of equity, the stockholders lost €38.40.

The ROE of the second category in the sample did not fluctuate very much. In both 2001 and 2002, their ROE was pretty low, but in the period 2003-2010 their ROE was more or less stable. However, in 2008 their ROE declined due to a decrease in net income, but this decrease was less than that of the other categories.

The two categories grouping the smallest banks barely reached a 10% ROE under favourable economic conditions. While the banks with assets between €7 and €10 billion have a quite stable ROE after 2003, the average ROE for the smallest banks in 2008 has been severely impacted by two important write-off for Onderling Beroepskrediet-OBK-Bank C.V.B.A. and Ethias Bank.

30,00%

20,00%

10,00%

0,00% TA > 100 000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 TA > 10 000 -10,00% TA > 7 000 -20,00% TA < 7 000

-30,00%

-40,00%

-50,00%

Figure 38: Return on equity (ROE) based on size (million €) Source: BankScope

The profitability of our four big banks taken separately shows some interesting results. In 2001, ING’s ROE was very high (i.e. their ROE is 2 times higher compared to the others). Also in 2002, their ROE is the highest among the Big Four. In 2003, Dexia shows the best ROE performance. In 2004, the four banks achieved more or less the same results in terms of ROE. In 2006, Fortis’ ROE increases substantially, followed by a remarkable decrease in both 2007 and 2008. In 2008, Fortis’ ROE was more than 100% negative, which implies a negative equity. In 2008, both KBC and Dexia had negative ROEs, but not as negative as Fortis’ ROE. In 2009, Fortis’ ROE increased, but the bank was not yet profitable. Furthermore, KBC was still showing a negative ROE. In 2010, the ROEs for our four banks had more or less the same value. (The sections that follow will elucidate the main driver – a high leverage or a higher return to total assets – of the change in profitability.)

60,00% 40,00% 20,00% 0,00% -20,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Fortis -40,00% KBC -60,00% Dexia -80,00% ING -100,00% -120,00% -140,00% -160,00%

Figure 39: Return on equity (ROE) for Fortis, KBC, Dexia and ING Source: BankScope

To understand the evolution of the ROE, we will investigate the main drivers for it: i.e. leverage and return on assets. The next section describes the financial leverage, and the section thereafter details the evolution of the ROA.

Figure 40 shows that the private banks, on average, generate the highest ROE. Even in times of financial distress, when the ROE of all the other categories is below zero (i.e. in 2008), the ROE of the private banks is even above 10%. The ROE of the commercial banks fluctuates the most, which implies that they are the most risky banks of the sample. However, the ROE of both the cooperative banks and the savings banks is just about as volatile.

25,00%

20,00%

15,00% Coop. Banks 10,00% Comm. Banks

5,00% Savings Banks Private Banks 0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 -5,00%

-10,00%

Figure 40: Return on equity (ROE) based on the BankScope segmentation Source: BankScope

LEVERAGE

In this section, we measure leverage as total assets divided by equity. Whenever a bank’s assets exceed its equity base, that bank is considered to be ‘leveraged’. The more assets a bank finances with debt with the view to enhancing its returns, the higher is that bank’s leverage. By leveraging, a bank bets that the interest paid on the borrowed money will be smaller than the return generated, thus improving the bank’s profitability. If leverage is employed successfully, the positive difference between the return on the capital employed and the cost of capital will generate an economic profit.

A lot of research has already been conducted on leverage. Academic literature reports that leverage in the banking sector is pro-cyclical (e.g. Adrian and Shin, 2008). This pro-cyclical nature of leverage (i.e. harvesting higher returns in good times and causing a collapse of financial institutions’ balance sheets in bad times) renders the financial system particularly fragile to shocks. Although leverage may boost ROE in good times, in a recession it can bring the collapse of the bank.

With the onset of the financial crisis, banks were accused of having taken on excessive leverage and are therefore subject to scrutiny. In response, and as a measure to limit systemic risk, the Basel Committee of Banking Supervisors, following the G20 request to enhance financial sector resilience, proposed that banks should be legally obliged to comply with a pre-defined leverage ratio standard.

Figure 41 shows a similar pattern for the leverage of investment banks and universal banks. However, the leverage for the investment banks is always lower than that for the universal banks. In 2010, though, while the universal banks were de-leveraging slightly, the investment banks increased substantially their leverage to its pre-crisis level of 2006.

30,00

25,00

20,00

15,00 Investment Banks Universal Banks 10,00

5,00

0,00 2001200220032004200520062007200820092010

Figure 41: Leverage for investment banks and universal banks Source: BankScope

Figure 42 shows that the smallest banks were less leveraged during the entire period. Also, the pro- cyclical trend is not as obvious in this figure as would be expected from the discussion above. One can observe that, except for the years 2001 and 2002, the 8 biggest banks present a significantly higher leverage than the smaller ones. One point to highlight is the very high leverage for the Big Four banks in 2008.

45,00

40,00

35,00

30,00 TA > 100 000 25,00 TA > 10 000 20,00 TA > 7 000 15,00 TA < 7 000 10,00

5,00

0,00 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 42: Leverage based on size (million €) Source: BankScope

Looking at the four big banks separately shows that KBC has maintained a very stable leverage during the period under study. KBC’s leverage was also the lowest of the big four. It is interesting to note that ING started to de-leverage in 2006, before the crisis, followed by KBC and Fortis in 2007. However, in the case of Fortis, leverage was back in play in 2008. Thus, the high leverage of 2008 for the Big Four is clearly attributable to the leverage of Fortis and Dexia. With an extremely high leverage of 74.91, Dexia was obliged to restructure the liability side of its balance sheet. Therefore, in 2009 Dexia started to de- leverage: in that year, Dexia stated that it wanted to reduce the group’s risk profile by reducing the balance sheet by 35% by 2014. Dexia Bank Belgium’s balance sheet decreased by 3.54% in 2009 and then decreased another 2.31% in 2010. The reduction in leverage was mainly due to an increase in equity. The leverage for our four big banks in both 2009 and 2010 declined, reflecting the regulatory pressures following the financial crisis. In 2010, the de-leveraging in the Belgian banking sector continued, with total assets falling from some 430% of Belgium’s GDP in mid-2009 to around 360% by mid-2010 (IMF 2010).

80,00

70,00

60,00

50,00 Fortis

40,00 KBC Dexia 30,00 ING 20,00

10,00

0,00 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 43: Leverage for Fortis, KBC, Dexia and ING Source: BankScope

Using BankScope’s categorisation, Figure 44 clearly shows that the commercial banks and the savings banks have a much higher leverage than cooperative banks and private banks. However, the leverage of the private banks in 2005 and 2006 is higher than that for the cooperative banks. Thus, the commercial banks and the savings banks will certainly suffer the most from the new regulations to de-leverage the banking industry.

30,00

25,00

20,00 Coop. Banks

15,00 Comm. Banks Savings Banks 10,00 Private Banks

5,00

0,00 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 44: Leverage based on BankScope segmentation Source: BankScope

RETURN ON ASSETS

To complete the first level of the analysis of the ROE, we examine the return on assets (ROA). The return on assets tells an investor how much profit the bank generated for each €1 in assets, or how well management is using a bank’s assets. Thus, the ROA can be used to assess the operational profitability of a bank.

Figure 45 shows that the ROAs of the investment banks is much higher than that of the universal banks. While the universal banks have a higher leverage and take more solvency risk than the investment banks, the latter have a higher ROA as the main driver to achieving a higher ROE. The large drop in ROA for the investment banks from 2001 to 2002 is due to Keytrade Bank which, as we will see later, shows a drop in asset yield and a negative profit margin in 2002. During the crisis, the universal banks were affected mainly in terms of ROA, i.e. negative value in 2008; the ROA of the investment banks decreased, but was still twice its level of 2002, its lowest level.

2,00%

1,50%

1,00% Investment Banks Universal Banks 0,50%

0,00% 2001200220032004200520062007200820092010

-0,50%

Figure 45: ROA for investment banks and universal banks Source: BankScope

When comparing ROA on the basis of total assets (Figure 46), we can see that, except for the smallest banks, the ROA is in line with the scorecard of the DNB (Bos et al., 2006). Most of the time, the smallest banks have the highest ROA – in combination with the lowest leverage, this makes them less risky. However, in 2008 they were impacted by the financial crisis, and their ROA turned into a slightly negative value. The ROA of the four biggest banks together fluctuates around 0.50%; however, due to the financial crisis, their ROA turned into a significantly negative value in 2008 and remained negative in 2009. In 2010, the ROA for the biggest banks was back at its pre-crisis level. The second category (i.e. TA between €10 and €100 billion) shows the most stable ROA evolution. While they suffered from the economic conditions in both 2001 and 2002, they fully recovered in 2003 and generated their highest ROA. Even in both 2008 and 2009, their ROA remained stable compared to the high fluctuations in the ROA of the very big or very small banks.

2,00%

1,50%

1,00% TA > 100 000 0,50% TA > 10 000

0,00% TA > 7 000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 TA < 7 000 -0,50%

-1,00%

-1,50%

Figure 46: ROA based on size (million €) Source: BankScope

When we compare the ROA of our four big banks (Figure 47), ING can be considered to be the most stable bank based on its ROA during the time period under study. In 2009, ING’s ROA is even at its highest over the period, and in 2008 ING is the only bank in this group to achieve a positive ROA. In 2008, Fortis suffered the largest loss in the Belgian banking sector for the decade under study. In 2009, impairments at KBC resulted in a substantial loss. While Fortis and KBC recovered in 2010, Dexia’s performance was about half that of the other three banks.

1,50% 1,00% 0,50% 0,00% -0,50% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Fortis -1,00% KBC -1,50% Dexia -2,00% ING -2,50% -3,00% -3,50% -4,00%

Figure 47: ROA for Fortis, KBC, Dexia and ING Source: BankScope

Obviously, no specific trends can be found in the evolution of the ROA. As one would expect, the ROA for all types of banks decreased in 2008. For some of them (i.e. the cooperative banks and the savings banks), the ROA became negative. It is also noteworthy that, most of the time, the private banks produce the highest and the most consistent ROA. Even in 2008, their ROA was still 1%. In the case of the cooperative banks in 2005, the small sample size (2 banks) explains the high volatility of their ROA. In 2005, the ROA of 11.23% was driven by the results of the Antwerps Beroepskrediet. This exceptional result is related to an accounting procedure that reintroduced a positive reserve of €40 million (compared to a negative reserve of €6.5 million in 2004) into the income statement. In 2008, the big negative ROA for the cooperative banks is also due to the Antwerps Beroepskrediet. They took a loss of more than €33 million on their bond portfolio. Overall, looking at the ROA of the private banks, and based on the scorecard of the DNB (Bos et al., 2006), we can conclude that these banks are out- performing the industry.

5,00%

4,00%

3,00% Coop. Banks 2,00% Comm. Banks

1,00% Savings Banks Private Banks 0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 -1,00%

-2,00%

Figure 48: ROA based on BankScope segmentation Source: BankScope

ASSET YIELD

The second level of analysis of the Dupont system decomposes the return on assets into asset yield and profit margin.

The asset yield – measured as total revenue to total assets – can be seen as the capacity of a bank to generate income which can be related to the composition of the balance sheet, to the off-balance sheet activities and to pricing.

Figure 49 shows that, when looking at the capacity to generate income, investment banks always do better than universal banks, except in 2010. The big gap between investment and universal banks in 2001 is due to the asset yield of Keytrade bank (i.e. 18.64%). In 2002, their asset yield dropped to 5.88%, which was mainly due to the acquisition of RealBank during 2002.

The years 2001 and 2006 show a higher asset yield for investment banks compared to universal banks. Looking at Figure 17, these results support the well-accepted assumption that the capacity of investment banks to generate income is correlated with the performance of the financial markets.

10,00% 9,00% 8,00% 7,00% 6,00% 5,00% Investment Banks 4,00% Universal Banks 3,00% 2,00% 1,00% 0,00% 2001200220032004200520062007200820092010

Figure 49: Asset yield for investment banks and universal banks Source: BankScope

Figure 50 below shows the evolution of the asset yield based on size. Of course, the category that contains the four biggest banks deviates significantly from the three other categories. In both 2007 and 2008, in the middle of the financial crisis, they were able to achieve record levels for their asset yield. Interesting enough, the smallest banks were most able to generate income before 2007. The categories in between the smallest ones and the biggest ones are quite comparable.

10,00% 9,00% 8,00% 7,00% 6,00% TA > 100 000 5,00% TA > 10 000 4,00% TA > 7 000 3,00% TA < 7 000 2,00% 1,00% 0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 50: Asset yield based on size (million €) Source: BankScope

The details of the evolution of the four biggest banks individually show that the main driver for the exceptional performance of the Big Four in 2007 and 2008 was Fortis, which was able to generate extremely high gross income in both years. ING can also be considered an extreme case, presenting a very volatile asset yield with peaks in 2005 and in 2010. This volatility is due to the large increases of the gross interest and dividend income component of their total income in these two years. In the case of Fortis, the volatility of its asset yield can also be linked to the large volatility of the gross interest and dividend income. In 2007, there was both an increase in total assets and an increase in gross interest and dividend income. As gross income increased from €36 million to €98 million, the increase in AY in 2007 is due mainly to this factor. However, in 2008 the increase in AY for Fortis is due mainly to a decrease in total assets (i.e. a drop of 23% in TA). The decrease in AY in 2009 for Fortis is due to a more significant decrease in gross income (i.e. almost 50%) than in TA (i.e. 25%). In 2010, there was a drop of more than 80% in gross income, which yielded a very low AY for Fortis.

20,00% 18,00% 16,00% 14,00% 12,00% Fortis 10,00% KBC 8,00% Dexia 6,00% ING 4,00% 2,00% 0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 51: Asset yield for Fortis, KBC, Dexia and ING Source: BankScope

Figure 52 shows that the cooperative banks were able to generate a steady income over time, except for 2008. This figure also shows the dominance of the private banks in capacity to generate income. They are closely followed by the commercial banks – at least in 2001, 2007 and 2008.

9,00%

8,00%

7,00%

6,00% Coop. Banks 5,00% Comm. Banks 4,00% Savings Banks 3,00% Private Banks 2,00%

1,00%

0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 52: Asset yield based on BankScope segmentation Source: BankScope

PROFIT MARGIN

Based on the Dupont model, the second component of the ROA is the profit margin, which can be calculated as net income to total revenues. This ratio can be an indicator of a bank's pricing strategies and how well it controls costs. So the higher the ratio, the better the bank is in managing its costs base.

Figure 53 shows the PM for investment banks and universal banks. Both types of banks have a similar pattern associated with pro-cyclical behaviour, except for the outstanding results of the investment banks in 2003. The main driver for this result was Keytrade Bank, which generated a PM of 18.83%. In 2008, the universal banks were much more affected by the financial crisis than the investment banks. Investment banks show a much higher PM than the universal banks over the entire period, which can be largely attributed to the capacity of investment banks to generate higher income as shown by their asset yield in Figure 49.

30,00%

25,00%

20,00%

15,00% Investment Banks Universal Banks 10,00%

5,00%

0,00% 2001200220032004200520062007200820092010

Figure 53: Profit margin for investment banks and universal banks Source: BankScope

Looking at the PM based on size, we can see that the smallest banks perform better than the largest ones, which leads us to conclude that the smallest banks are better at cost management. Furthermore, in 2008 the largest banks experienced a big decrease in their PM: their PM becomes negative in 2008 and stays negative in 2009, which is clearly the result of the financial crisis on the largest banks.

Looking at the banks with TA between €10 and €100 billion, their PM was negative in 2002. This was due to RecordBank, with a PM of -11%. However, in 2003 their PM went up again, due to the strong performance of Centea, with a PM of 31%. During the crisis, these banks were not affected very much in terms of PM. Their PM decreased, but only a little compared to the other categories.

Looking at our third category, in 2006 a big increase in PM can be observed, and it even increased further in 2007. Like the other three categories, the PM went down in 2008 and increased again in both 2009 and 2010.

The smallest banks in the sample also show a pro-cyclical behaviour. In 2006, the PM for both the smallest and biggest banks was at the same level; however, the smallest banks were better at managing their costs during the financial crisis.

25,00%

20,00%

15,00% TA > 100 000 10,00% TA > 10 000

5,00% TA > 7 000 TA < 7 000 0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 -5,00%

-10,00%

Figure 54: Profit margin based on size (million €) Source: BankScope

Figure 55 shows the evolution of the profit margin of the four biggest banks. In both 2008 and 2009, there was a negative profit margin for some of them. Fortis yielded the lowest PM in 2008. In 2009, they started to recover and, in 2010, they were even generating the highest profit margin. KBC and Dexia also generated a negative PM in 2008. The negative result of Dexia was of short duration: in 2009, Dexia joined ING with a positive PM. For KBC on the other hand, 2009 was the worst year, with a negative PM of almost 23% reflecting the write-offs decided by the bank. In 2010, the impact of the write-offs was absorbed, and KBC was back with a positive PM. ING is the only bank of the Big Four to show a positive PM over the whole decade. However, compared to the other banks, ING experienced a substantial drop in its PM in 2005, but it was able to come back to the level of the three other banks in 2006. During the financial crisis (i.e. 2007-2008), ING’s PM was relatively stable. However, its PM decreased substantially in 2010, moving in the opposite direction of the three other banks.

25,00% 20,00% 15,00% 10,00% 5,00% Fortis 0,00% KBC -5,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Dexia -10,00% ING -15,00% -20,00% -25,00% -30,00%

Figure 55: Profit margin for Fortis, KBC, Dexia and ING Source: BankScope

As for the asset yield, Figure 56 shows the clear dominance of the private banks – except for 2009 and 2010, which are dominated by the cooperative banks. Here again, the dominance of the cooperative banks is due to the exceptional result of Antwerps Beroepskrediet. Except for 2009 and 2010, when the cooperative banks out-perform the industry, the private banks are clearly the best from a profit margin point of view (particularly in the good years 2003 – 2007).

30,00%

25,00%

20,00% Coop. Banks 15,00% Comm. Banks

10,00% Savings Banks Private Banks 5,00%

0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 -5,00%

Figure 56: Profit margin based on BankScope segmentation Source: BankScope

COST TO INCOME

As discussed earlier in Chapter 2, the commercial, cooperative and savings banks generate most of their income through financial intermediation. On the other hand, the private banks’ income is mainly derived from the services they provide. The analysis of the cost to income ratio looks at how efficient the banks were in generating their revenues. This cost to income ratio – the operational costs divided by the operational income – gives a clear view of how efficient the bank is in its operations. The lower the ratio, the more profitable the bank is. Changes in the ratio can also highlight potential problems: if the ratio rises from one period to the next, costs are rising at a higher rate than income, which could suggest that the bank lost track of the evolution of its costs perhaps in a drive to attract more business.

Figure 57 shows that, for investment banks, the operational costs to operational income was highest in 2002; whereas, for universal banks, the ratio was highest in 2008. For universal banks, the ratio was still increasing in 2003, as the efficiency measure for investment banks was starting to come down, bringing both types of banks very close to each other in term of efficiency. From 2005 to 2008, the universal banks show more costs relative to their operational income compared to the investment banks. In 2010, both types of banks ended with a cost to income ratio that was more or less the same.

90,00%

80,00%

70,00%

60,00%

50,00% Investment Banks 40,00% Universal Banks 30,00%

20,00%

10,00%

0,00% 2001200220032004200520062007200820092010

Figure 57: Cost to income ratio for investment banks and universal banks Source: BankScope

With regard to the efficiency of the banks based on their size, we clearly see a peak in 2008 and 2009 with the four biggest banks in the sample. The reduction in income is the reason for this increase.

100,00% 90,00% 80,00% 70,00% 60,00% TA > 100 000 50,00% TA > 10 000 40,00% TA > 7 000 30,00% TA < 7 000 20,00% 10,00% 0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 58: Cost to income ratio based on size (million €) Source: BankScope

Figure 59 shows a very stable and comparable ratio for the period 2001-2006. So, from an efficiency point of view, the four big banks were quite similar in this period.

With the crisis occurring in 2007, we start to see significant differences among the Big Four. Fortis was most noticeable in terms of the cost to income ratio. In 2007, its cost to income ratio increased due to a big increase in operational expenses. This trend continued for Fortis in 2008; however, as their operational expenses increased, their operational income decreased even further. In 2008, the cost to income ratio was above 100%, which implies that the costs were higher than their income, yielding a loss that year. KBC was in the same situation that year. In 2009, both KBC and Fortis performed better, but now it was Dexia’s turn to exceed the 100% mark. In 2010, income increased again, and so the cost to income ratio decreased. For some banks, the ratio was even lower than before the financial crisis.

It is noteworthy that the cost to income ratio for ING was steadily low and stable during this period.

140,00%

120,00%

100,00% Fortis 80,00% KBC 60,00% Dexia

40,00% ING

20,00%

0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 59: Cost to income ratio for Fortis, KBC, Dexia and ING Source: BankScope

Looking at the BankScope segmentation in Figure 60, we find that the commercial banks and the private banks achieved the highest ratios in 2001 and 2002. In the following years, savings banks and commercial banks show a very high ratio most of the time, with the savings banks approaching 90% in 2006.

100,00% 90,00% 80,00% 70,00% 60,00% Coop. Banks 50,00% Comm. Banks 40,00% Savings Banks 30,00% Private Banks 20,00% 10,00% 0,00% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 60: Cost to income based on BankScope segmentation Source: BankScope

Given the banks’ focus on this ratio, Table 6 shows each bank’s average cost to income ratio for the last three years (2008, 2009 and 2010). Santander Benelux out-performs the other banks by far.

Cost to income Bank ratio Agricaisse - Caisse Coopérative de Dépôts et de Crédit Agricole 38,59% Antwerps Beroepskrediet 38,15% Argenta Spaarbank-ASPA 50,93% Banca Monte Paschi Belgio SA 67,80% Bank J. Van Breda en Co NV 58,72% Banque de la Poste-Bank van de Post 78,55% Banque Degroof SA-Bank Degroof NV 62,62% Banque Delen NV 46,26% BKCP 113,02% Byblos Bank Europe SA 60,71% CBC Banque S.A. 60,47% Centea 47,48% Citibank Belgium N.V./S.A. 71,47% Crédit Agricole SA-Landbouwkrediet 76,18% Delta Lloyd Bank 96,00% Deutsche Bank SA-Deutsche Bank NV 85,58% Dexia Bank Belgium-Dexia Bank 79,03% Ethias Bank 100,60% Europabank 64,16% F. van Lanschot Bankiers Belgie 90,23% Fortis Bank SA/ NV-BNP Paribas Fortis 93,54% ING Belgium SA/NV-ING 57,69% KBC Bank NV 84,97% Keytrade Bank SA/NV 46,47% Onderling Beroepskrediet-OBK-Bank C.V.B.A. 120,41% Record Bank SA/NV 60,64% Santander Benelux SA/NV 7,62% Société Générale Private Banking N.V. 87,95% VDK Spaarbank NV 65,23% Table 6: Three-year average of the cost to income ratio for the period 2008-2010

CONCLUSION

We have used the Dupont Framework to analyse the financial performance of the Belgian banking industry. When using the categories proposed by the European Central Bank – the Investment banks and the Universal banks – the Investment banks clearly out-perform the Universal banks in terms of ROE, except in 2002. Even more interesting, in the crucial year 2008, the Investment banks were able to achieve an average ROE of about 15%. Since the Universal banks dominate the Investment banks in terms of Leverage, it is the Investment banks’ ROA that gives them their leading position in ROE. The dominance of the Investment banks in terms of ROA is due to their dominance in terms of profit margin as well as asset yield, except for the asset yield of 2010.

When we look at the 4 categories of banks based on their total assets, we find that, most of the time, the Big Four achieved a better ROE than the other categories, except for catastrophic results in 2008 and a slightly negative ROE in 2009. Although the four big banks are characterised by a high leverage which is close to that of the next category (banks with assets between €10 and €100 billion), their ROA is never in first place. Thus, they were able to achieve the best ROEs in some years by taking more risks in terms of solvency. Looking at the capacity to generate income (AY), one can see that the smallest banks clearly dominate in the period 2001 to 2006, while the four big banks lead the industry in the following years. There is no consistent leadership in terms of profit margin. An interesting conclusion for this grouping of banks is that the Big Four’s leadership in terms of ROE is mainly due to their larger risk appetite.

When looking at the Big Four separately, we cannot select a leading bank: first place changes constantly from one player to another. However, ING had the highest leverage in the first year of the study, followed closely by Dexia and Fortis, and Fortis dominated the period from 2002 to 2006. After 2006, Dexia was by far the most leveraged bank. A ranking based on ROE is very volatile. Such a result can be explained by the high volatility of both asset yield and profit margin.

When using the categories defined by BankScope, we found that Private banks clearly out-performed the other categories, followed by Commercial banks and Savings banks, with Cooperative banks far behind. In terms of ROE, the leading position of the Private banks is not related to a high Leverage – like the Cooperative banks, they have a leverage of about half of that of the Commercial banks and the Savings banks – but rather to an almost consistent and steady ROA. This is due to the combination of their capacity to generate income (AY) and to a low costs structure (PM). Concerning the AY, they out- perform the Commercial banks slightly and leave Savings banks and Cooperative banks far behind most of the time. Their leading position in term of ROE is clearly derived from their PM, except for 2009 and 2010.

CHAPTER 4: THE RISK-RETURN TRADE-OFF

In this chapter, we discuss the risk-return trade-off of our 29 banks over the period under study. As it is impossible to obtain the stock market returns for all of the banks, we use the ROE as a measure of return. In Figure 61 and 62, the average annual risk (i.e. the standard deviation of the ROE for the period under study)11 is plotted along the x-axis, and the average annual ROE is on the y-axis.

Looking at Figure 61, we see that the banks are clustered in an area of low volatility with positive returns. While the normal relationship expected in finance is ‘the higher the expected risk, the higher the expected return’, the Big Four players on the Belgian market exhibit an inverse relationship. Fortis yields the lowest ROE, averaged on a yearly basis over the 10 years, and this is in combination with the highest risk. Of our four big banks, ING performs the best. ING has a higher ROE in comparison to both Dexia and KBC and has the lowest risk. The bank that has the highest average annual ROE over the 10-year study is Keytrade Bank; however, this high ROE is associated with a high risk. On the other hand, the bank with the lowest risk is Agricaisse, which also had an average annual ROE of 5%. Europabank and Banque Delen perform best in terms of their risk-return trade-off.

It is interesting to note that, over this 10-year period, Dexia outperformed KBC in terms of its risk- return trade-off.

To obtain a better view on the evolution of the risk-return trade-off for the Big Four players, we examine the evolution of the risk-return trade-off over a rolling window of 5 years. Figure 62 presents a rolling window estimation for the risk-return trade-off of Dexia, Fortis, ING and KBC, starting with the period 2002 – 2006 (labelled as 2006) until the period 2006 – 2010 (labelled as 2010). The impact of the financial crisis is clearly evident in a decreasing risk-return trade-off, primarily for Fortis. However, the five points for ING are all located in the same area, which implies a stable risk-return trade-off even during the financial crisis. Looking at the other three banks, we clearly see that, for the periods 2002-2006 and 2003-2007, they were comparable to ING. However, due to a poorer performance in 2008, their average risk-return trade-off for 2004-2008 deteriorated substantially. For KBC, the decline was small for that period, but in both 2009 and 2010 the decline became more significant. For Dexia, the risk-return trade-offs of 2008, 2009 and 2010 cluster together at a better level than those for KBC for 2009 and 2010. Fortis is the big loser of this comparison, with negative returns in 2008, 2009 and 2010.

11 In finance and in risk management, risk is associated with the volatility of results. Assuming a normal distribution of these results, standard deviation is the method for measuring this volatility. 79

Figure 61: Risk-return trade-off for the 29 banks over a 10-year period

Figure 62: 5 year rolling window risk-return trade-off for Dexia, Fortis, ING and KBC

CHAPTER 5: WHAT DID WE LEARN IN THIS STUDY?

The first point of interest in this study is the relative importance of the financial sector in Belgium’s economy. Its importance is substantial: more than 6% of the Gross Value Added to the Belgian economy in 2000 as well as in 2010. As one would expect, the year 2008 shows a substantial decrease to about 5%. The financial sector’s contribution in Belgian employees’ employment compensation went from more than 6% in 2000 to 5% in 2010. The same trend is observable in the number of people employed in the financial sector: from 148,000 in 2001 to about 137,500 in 2010. However, while the number of people employed decreased during this period, the average salary steadily increased from €58,000 in 2001 to €70,000 in 2008. After 2008, the impact of the financial crisis freezes wages at about €70,000. Another striking feature of the Belgian economy is the savings rate: one of the highest in the world, it fluctuated between 15% and 20% between 2001 and 2010.

When looking more specifically to the Belgian banking sector, one clearly notices its international standing. In Belgium, almost half of the total assets held by Belgian financial institutions are in the hands of foreign-controlled credit institutions.

To study the value chain of the Belgian banking sector, and to break down its profitability, we segmented the banking industry in several different ways. First, by dividing the banks into Investment banks and Universal banks; second, by segmenting the banks according to their size; and finally, by using the “BankScope” segmentation. One main conclusion of the study is that the banking sector in Belgium is highly concentrated with regard to total assets: in 2009, the four big banks accounted for almost 90% of the total assets of the Belgian banking sector.

When breaking down the value chain, it is interesting to note that private banks consistently exhibit the highest operating income to total assets ratio – which makes sense, as their business model relies on off-balance sheet activities like assets under management. In the investment / universal banks segment, the investment banks have a better ratio than the universal banks. However, the pool of investment banks is very small compared to that of universal banks, which includes the Big Four banks. When examining bank size, the smallest banks perform best – but by virtue of the selected sample, they represent a very small portion of the banking sector.

When we broke the total operating income into its components, we found that net interest income was negatively correlated to the size of the bank as measured by total assets. As expected, the private banks dominate the other banks when net fees and commissions are used as the benchmark; but no conclusion can be drawn when using financial transactions as the benchmark, given the high volatility of this value driver. On the costs side of the value chain, all of the banks were able to reduce their operating expenses during the period under study. Private banks experienced the highest level of operating expenses to total assets (4%), while universal banks were able to bring this figure to a low of 1%.

Looking at the performance of the banking industry based on the breakdown of the ROE, the four big banks dominate the market, not because of their good operational profitability (ROA), but because of a bigger risk appetite, which is reflected in a higher leverage compared to the other banks. The comparison of investment banks and universal banks shows that the investment banks performed better in terms of ROE. However, based on their total assets for the period under study, the investment banks account for not even 1% of the sample, while the universal banks account for the rest. In the “BankScope” segmentation, the Private banks have the best ROE, due to their better performance in operational profitability (ROA). While the asset yield is dominated by the Big Four after 2007, the investment banks achieved better profit margins over the period under study (but again, their representation in the overall sample is very small). The cost to income ratio of the Big Four shows a clear deterioration in 2007, 2008 and 2009.

The analysis of the risk-return trade-off shows a cluster of the banks in an area from low to medium risk and from a low to a high ROE. The banks presenting the best risk-return trade-off over the period under study are Banque Delen and Europabank. An interesting point of this risk-return analysis is the inverse relationship for the Big Four banks. ING shows a good risk-return profile, followed by Dexia and KBC with a higher risk and a lower return, and finally by Fortis Bank (BNP Paribas Fortis) with the worst risk-return profile for the period under study.

REFERENCES

Albertazzi U. and L. Gambacorta (2009), “ Bank profitability and the business cycle”, Journal of Financial Stability Vol. 5 pages 393-409.

Allen N., N. Berger and J. Wilson (2009), “The Oxford Handbook of Banking”, Oxford University Press.

Altunbas Y., S. Carbo and E.PM. Gardener (2007), “Examining the Relationship between Capital, Risk and Efficiency in European Banking”, European Financial Management, 13, 1, pages 49-70.

Altunbas Y., E.P.M. Gardener and B. Moore (2001), “Efficiency in European Banking”, European Economic Review, 45, pages 1931-1955.

Balla E. and A. McKenna (2009), “Dynamic provisioning: a countercyclical tool for loan loss reserves.” Economic Quarterly, Federal Reserve Bank of Richmond, issue fall, pages 383-418.

Bikker J.A. and P. Metzemakers (2005), “Bank provisioning behavior and procyclicality”, Journal of International Financial Markets, Institutions and Money 15, 141–157.

Bolt W., L. de Haan, M. Hoeberichts, M. van Oord and J. Swank (2010), “Bank profitability during recessions”, DNB Working Paper, No. 251.

Bos J.W.B., J.A.J. Draulans, D. van den Kommer and B.A. Verhoef (2006), “An international scorecard for measuring bank performance: The case of Dutch banks”, DNB Occasional studies, Vol. 4 / Nr. 2.

Bouvatier V., and L. Lepetit (2007), “Banks’ procyclical behavior: Does provisioning matter?”, Journal of International Financial Markets, Institutions and Money.

Casu B. (2003), “A comparative Study of Efficiency in European Banking”, Applied Economics, 35(17), pages 1865-1876.

Casu B. and C. Girardone (2004), “Productivity in European Banking – A Comparison of Parametric and Non-Parametric Approaches”, Journal of banking and Finance, Volume 28, Issue 10, pages 2521- 2540.

Dermine J. (2009), “Bank valuation & value-based management”, McGraw-Hill, 442 p.

Dexia (2002), “Dexia Annual Report”.

Dexia (2008), “Dexia Bank Belgium Annual Report”.

Dexia (2009), “Dexia Bank Belgium Annual Report”.

Dinamona D.D. (2008), “Bank risks, provisioning and the business cycle: a panel analysis on European intermediaries”, Banks and Bank Systems, Volume 3, Issue 2, 2008.

English W.B. (2002), “Interest rate risk and bank net interest margins”, BIS Quarterly Review, December 2002.

European Central Bank (2010). “‘Beyond ROE- how to measure bank performance?; Appendix to the report on EU banking structures”, September 2010

European Central Bank (1996). “Foreword ESA 95”, Retrieved June 17, 2011, from Europa.eu: http://circa.europa.eu/irc/dsis/nfaccount/info/data/esa95/en/een00000.htm

European Central Bank (2011). “Map of euro area 1999 – 2011”, Retrieved June 4, 2011, from European Central Bank: the Euro: http://www.ecb.int/euro/intro/html/map.en.html

European Central Bank (n.d.). “MFIs and Assets”, Retrieved June 9, 2011, from European Central Bank: https://mfi-assets.ecb.int/user_manual.htm

European Central Bank (n.d.). “Statistics glossary”, Retrieved June 17, 2011, from European Central Bank: http://www.ecb.int/home/glossary/html/act2v.en.html

Eurostat. (2011). “Introduction”, Retrieved June 5, 2011, from Eurostat > European Sector Accounts: http://ec.europa.eu/eurostat/sectoraccounts

Febelfin. (2009). “Statistisch vademecum van de banksector”, Brussels: Febelfin.

Febelfin. (2010). “Banking and Finance in Belgium: facts & figures 2009-2010”, Brussels: Febelfin.

Girardone C. and E.P.M. Gardener (2004), “Analysing the Determinants of Bank Efficiency – The Case of Italian banks”, Applied economics, 36, pages 215-227.

Goddard J. and M. Tavakoli (2007), “European Banking: An Overview”, Journal of Banking and Finance, Volume 31, pages 1911-1935.

ING (2008), “Annual Report ING Bank”.

ING (2008), “Press Release ING Belgium”.

ING (2008), “Financial Report ING Bank Belgium”.

ING (2009), “Annual Report ING Bank”.

ING (2009), “Financial Report ING Bank”.

KBC (2008), “KBC Annual Report”.

KBC (2009), “KBC Annual Report”.

Laeven L. and G. Majnoni (2003), “Loan loss provisioning and economic slowdowns: too much, too late?”, Journal of Financial Intermediation 12, 178–197.

National Bank of Belgium. (2011). Statistics. Retrieved 06 08, 2011, from Home > Statistics > Macroeconomic statistics: http://www.nbb.be/pub/stats/stats.htm?l=en&tab=Figures

Saunders A. and M. M. Cornett (2006), “Financial Institutions Management, a risk management approach”, fith edition, Mc Graw-Hill, 856 p.

Préfontaine J. and A. Thibeault (1993), “Introduction to bank financial management”, financial services studies program. The Institute of Canadian Bankers, September 1993, p. 127

Verweire K., P. Roelandt, J. De Grande and L. Van den Berghe (2006), “What drives performance in the Belgian insurance industry?”, Vlerick Leuven, Gent Management School, 124 p.

Wilson J. (2007), “Developments in European Banking”, Journal of Banking and Finance, Volume 31, pages 1907-1910.

Wilson J. and J. Goddard (2001), “European Banking: Efficiency, Technology and Growth”, John Wiley.

Wilson J. and J. Goddard (2004), “The profitability of European Banks – A Cross-sectional and Dynamic Panel Analysis”, The Manchester School, Vol 72, No. 3, pages 363-381.

Zouaoui M., G. Nouyrigat and F. Beer (2011), “How does investor sentiment affect stock market crises? Evidence from panel data”, FARGO – Centre de recherché en Finance, Architecture et Gouvernance des Organisations.

APPENDICES

APPENDIX 1: THE RATIOS

1. Chapter 2: The value chain a. Operational income to total assets b. Net interest income to total assets c. Net commissions and fees to total assets d. Financial transactions to total assets e. Operational expenses (Overheads) to total assets f. Loan loss provisions to total assets 2. Chapter 3: The financial performance a. Return on equity b. Leverage c. Return on assets d. Asset yield e. Profit margin f. Cost to income 3. Chapter 4: The risk return trade-off

APPENDIX 2: THE VARIABLES

1. Total assets 2. Equity 3. Operational income as net interest income plus net commissions and fees plus financial transactions 4. Net interest income 5. Net commissions and fees 6. Financial transactions 7. Operational expenses (overheads) 8. Loan loss provisions 9. Net income 10. Gross income as gross interest and dividend income plus total non-interest operating income

APPENDIX 3: A38 CODE

Code Description

AA Agriculture, forestry and fishing

BB Mining and quarrying

CA Manufacture of food products, beverages and tobacco products

CB Manufacture of textiles, wearing apparel and leather products

CC Manufacture of wood and paper products, and printing

CD Manufacture of coke and refined petroleum products

CE Manufacture of chemicals and chemical products

CF Manufacture of basic pharmaceutical products and pharmaceutical preparations

Manufacture of rubber and plastics products, and other non-metallic mineral CG products

Manufacture of basic metals and fabricated metal products, except machinery CH and equipment

CI Manufacture of computer, electronic and optical products

CJ Manufacture of electrical equipment

CK Manufacture of machinery and equipment n.e.c.

CL Manufacture of transport equipment

Manufacture of furniture; other manufacturing; repair and installation of CM machinery and equipment

DD Elektricity, gas, steam and air-conditioning supply

EE Water supply: sewerage, waste management and remediation activities

FF Construction

GG Wholesale and retail trade, repair of motor vehicles and motorcycles

HH Transportation and storage

II Accommodation and food service activities

JA Publishing, audiovisual and broadcasting activities

JB Telecommunications

Computer programming, consultancy and related activities; information service JC activities

KK Financial and insurance activities

LL Real estate activities

Legal and accounting activities; activities of head offices; management MA consultancy activities; architecture and engineering activities; technical testing and analysis

MB Scientific research and development

Advertising and market research; other professional, scientific and technical MC activities; veterinary activities

NN Administrative and support service activities

OO Public administration and defence; compulsory social security

PP Education

QA Human health activities

QB Social work activities

RR Arts, entertainment and recreation

SS Other service activities

Activities of households as employers of domestic personnel and undifferentiated TT goods and services production of households for own use

APPENDIX 4: NACE BEL CODE

NACE Bel Code 2008 Description

0 Agricultural, commodities production

1 Food production

2 Chemicals, steel, electronics production

3 Transport, small equipment, electricity production

4 Construction, wholesale, retail

5 Transport, catering industry, publishing

6 Communication, consulting, financial services, real estate

7 Holding, leasing, research, advertising

8 Security, cleaning, environment, government, medical

9 Entertainment, art, profession association

APPENDIX 5: LIST OF ABBREVIATIONS

CIS Credit Institutions

EEA European Economic Area: an agreement establishing a zone of 4 freedoms: free movement of (1) goods, (2) capital, (3) services and (4) persons. The agreement establishes a system ensuring equal conditions of competition.

EMU Economic and Monetary Union (European Central Bank, 2011):

- EMU-11: Austria, Belgium, Finland, France, Germany, Ireland, Italy, Luxembourg, the Netherlands, Portugal, and Spain - EMU-12: Greece joins (1 January 2001) - EMU-13: Slovenia joins (1 January 2007) - EMU-15: Cyprus and Malta join (1 January 2008) - EMU-16: Slovakia Joins (1 January 2009) - EMU-17: Estonia Joins (1 January 2011)

ESA 95 The new European System of National and Regional Accounts (ESA 95) is a high- quality statistical instrument which provides the European Community institutions, governments and economic and social operators with a set of harmonised and reliable statistics on which to base their decisions.

ESA95 is a major improvement on the previous version, which dates from 1979. Progress has been achieved in the harmonisation of methodology and in the precision and accuracy of the concepts, definitions, classifications and accounting rules which have to be applied in order to arrive at a consistent, reliable and comparable quantitative description of the economies of the Member States (European Central Bank, 1996).

GVA Gross Value Added (GVA) measures the contribution to the economy of each individual producer, industry or sector.

The gross value added at basic prices is defined in the ESA 95 as the output valued at basic prices less intermediate consumption valued at purchaser prices, before deduction of consumption of fixed capital (European Central Bank).

MFI Monetary Financial Institutions (MFIs) comprise resident Credit Institutions as defined in Community Law, and other resident Financial Institutions whose business is to receive deposits and/or close substitutes for deposits from entities other than MFIs, and, for their own account (at least in economic terms), to grant credits and/or make investments in securities (European Central Bank).

APPENDIX 6: CONTRIBUTION IN TERMS OF PERCENTAGE OF BOTH THE BANKS INDIVIDUALLY AND PER SECTOR.

VDK Spaarbank NV 0,23% Société Générale Private Banking N.V. 0,05% Santander Benelux SA/NV 0,72% Record Bank SA/NV 1,13% Onderling Beroepskrediet-OBK-Bank C.V.B.A. 0,08% Keytrade Bank SA/NV 0,13% KBC Bank NV 22,29% ING Belgium SA/NV-ING 12,16% Fortis Bank SA/ NV-BNP Paribas Fortis 34,44% F. van Lanschot Bankiers Belgie 0,14% Europabank 0,07% Ethias Bank 0,13% Dexia Bank Belgium-Dexia Bank 20,09% Deutsche Bank SA-Deutsche Bank NV 0,89% Delta Lloyd Bank 0,43% Crédit Agricole SA-Landbouwkrediet 0,78% Citibank Belgium N.V./S.A. 0,26% Centea 0,82% CBC Banque S.A. 0,76% Byblos Bank Europe SA 0,05% BKCP 0,28% Banque Delen NV 0,10% Banque Degroof SA-Bank Degroof NV 0,37% Banque de la Poste-Bank van de Post 0,58% Bank J. Van Breda en Co NV 0,24% Banca Monte Paschi Belgio SA 0,10% Argenta Spaarbank-ASPA 2,49% Antwerps Beroepskrediet 0,04% Agricaisse - Caisse Coopérative de Dépôts et de… 0,14%

Figure 63: The contribution of the banks (based on TA 2009) to the sample source: BankScope

3,68% 0,90% 0,47%

Commercial Banks Savings Banks Private Banks 94,96% Cooperative Banks

Figure 64: BankScope segmentation Source: BankScope

0,84%

Investment Bank Universal Bank

99,16%

Figure 65: ECB segmentation Source: BankScope

APPENDIX 7: THE SEGMENTATION BETWEEN UNIVERSAL AND INVESTMENT BANKS

Bank Name 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Agricaisse - Caisse Coopérative de Dépôts et de Crédit Agricole UB UB UB UB UB UB UB UB UB UB Antwerps Beroepskrediet UB UB UB UB UB UB UB UB UB UB Argenta Spaarbank-ASPA UB UB UB UB UB UB UB UB UB UB Banca Monte Paschi Belgio SA UB UB UB UB UB UB UB UB UB UB Bank J. Van Breda en Co NV UB UB UB UB UB UB UB UB UB UB Banque de la Poste-Bank van de Post UB UB UB UB UB UB UB UB UB UB Banque Degroof SA-Bank Degroof NV InvB InvB InvB InvB InvB InvB InvB InvB InvB InvB Banque Delen NV InvB InvB InvB InvB InvB InvB InvB InvB InvB BKCP UB UB UB UB UB UB UB UB UB UB Byblos Bank Europe SA InvB InvB UB UB UB InvB InvB InvB InvB UB CBC Banque S.A. UB UB UB UB UB UB UB UB UB UB Centea UB UB UB UB UB UB UB UB UB UB Citibank Belgium N.V./S.A. UB UB UB UB UB UB UB UB UB Crédit Agricole SA-Landbouwkrediet UB UB UB UB UB UB UB UB UB UB Delta Lloyd Bank UB UB UB UB UB UB UB UB UB Deutsche Bank SA-Deutsche Bank NV UB UB UB UB UB UB UB UB UB Dexia Bank Belgium-Dexia Bank UB UB UB UB UB UB UB UB UB UB Ethias Bank UB UB UB UB UB UB UB UB UB UB Europabank UB UB UB UB UB UB UB UB UB UB F. van Lanschot Bankiers Belgie InvB InvB InvB InvB InvB UB InvB UB UB Fortis Bank SA/ NV-BNP Paribas Fortis UB UB UB UB UB UB UB UB UB UB ING Belgium SA/NV-ING UB UB UB UB UB UB UB UB UB UB KBC Bank NV UB UB UB UB UB UB UB UB UB UB Keytrade Bank SA/NV InvB InvB InvB InvB InvB InvB InvB UB InvB UB Onderling Beroepskrediet-OBK-Bank C.V.B.A. UB UB UB UB UB UB UB UB UB UB Record Bank SA/NV UB UB UB UB UB UB UB UB UB UB Santander Benelux SA/NV UB UB UB UB InvB UB UB UB UB UB Société Générale Private Banking N.V. InvB InvB InvB InvB InvB InvB InvB InvB InvB VDK Spaarbank NV UB UB UB UB UB UB UB UB UB UB

Figure 66: Segmentation between Universal banks and Investment banks Source: BankScope

In this table, UB stands for Universal bank (if the main income comes from net interest income), and InvB stands for Investment bank. When a bank is categorised more often as an investment bank than as a universal bank during the entire period, then this bank is designated as an investment bank in our sample. The 6 banks that are categorised as investment banks are: Bank Degroof, Banque Delen, Byblos Bank, F. Van Lanschot Bankiers, Keytrade Bank and Société Générale.

APPENDIX 8: THE EVOLUTION OF THE NATURAL LOGARITHM OF THE TOTAL ASSETS

18,00

16,00

14,00

12,00 Coop. Banks 10,00 Comm. Banks 8,00 Savings Banks 6,00 Private Banks 4,00

2,00

0,00 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 67: Size based on the BankScope segmentation Source: BankScope

18,00

16,00

14,00

12,00

10,00 Investment Banks 8,00 Universal Banks 6,00

4,00

2,00

0,00 2001200220032004200520062007200820092010

Figure 68: Size for investment banks and universal banks Source: BankScope

25,00

20,00

15,00 TA > 100 000 TA > 10 000

10,00 TA > 7 000 TA < 7 000 5,00

0,00 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 69: Size based on size of the banks Source: BankScope

21,00

20,50

20,00 Fortis 19,50 KBC 19,00 Dexia

18,50 ING

18,00

17,50 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 70: Size of Fortis, KBC, Dexia and ING Source: BankScope

APPENDIX 9: RATIOS FOR EACH BANK INDIVIDUALLY

This appendix shows the three year average ratio for each bank separately during the period 2008- 2009 and 2010.

The table below explains the abbreviations used in the tables on the next two pages:

3yaOpIncometoTA10 3 year average operational income to total assets 3yaNIItoTA10 3 year average net interest income to total assets 3yaNCFtoTA10 3 year average net commissions and fees to total assets 3yaFTtoTA10 3 year average financial transactions to total assets 3yaLLPtoTA10 3 year average loan loss provisions to total assets 3yaOpExptoTA10 3 year average operational expenses to total assets 3yaROE10 3 year average return on equity 3yaROA10 3 year average return on assets 3yaLev10 3 year average leverage 3yaPM10 3 year average profit margin 3yaAY10 3 year average asset yield 3yaCostToIncome10 3 year average cost to income

Bank Name 3yaOpIncometoTA10 3yaNIItoTA10 3yaNCFtoTA10 3yaFTtoTA10 3yaLLPtoTA10 3yaOpExptoTA10

Agricaisse - Caisse Coopérative de Dépôts et de Crédit Agricole 0,92% 1,48% -0,56% 0,29% 0,41% Antwerps Beroepskrediet 3,16% 3,26% -0,02% -0,23% 0,26% 1,45% Argenta Spaarbank-ASPA 0,55% 0,80% -0,21% -0,03% 0,02% 0,36%

Banca Monte Paschi Belgio SA 1,59% 1,36% 0,19% 0,04% 0,51% 1,12%

Bank J. Van Breda en Co NV 2,69% 2,06% 0,74% -0,11% 0,11% 1,65%

Banque de la Poste-Bank van de Post 1,17% 2,10% -0,93% 0,02% 0,80% Banque Degroof SA-Bank Degroof NV 4,85% 1,35% 3,67% -0,18% 0,01% 3,41% Banque Delen NV 6,65% 0,64% 6,36% 3,07% BKCP 1,59% 1,49% 0,11% 0,00% 0,12% 1,96% Byblos Bank Europe SA 2,57% 1,15% 1,42% 0,03% 1,61% CBC Banque S.A. 2,57% 1,89% 0,59% 0,10% 0,04% 1,56% Centea 1,66% 1,90% -0,25% 0,01% 0,09% 0,79% Citibank Belgium N.V./S.A. 7,68% 7,67% -0,04% 0,05% 1,78% 6,43% Crédit Agricole SA-Landbouwkrediet 2,19% 2,09% 0,07% 0,03% 0,04% 1,86% Delta Lloyd Bank 2,19% 1,53% 0,44% 0,21% 0,05% 2,42%

Deutsche Bank SA-Deutsche Bank NV 1,19% 0,77% 0,35% 0,06% 0,00% 0,96% Dexia Bank Belgium-Dexia Bank 1,06% 1,00% 0,15% -0,09% 0,08% 0,67% Ethias Bank 0,75% 0,79% -0,03% -0,01% 0,15% 0,77% Europabank 6,24% 5,70% 0,81% 0,04% 4,48%

F. van Lanschot Bankiers Belgie 1,46% 0,78% 0,72% 0,02% 0,08% 1,31%

Fortis Bank SA/ NV-BNP Paribas Fortis 1,09% 0,88% 0,30% -0,09% 0,23% 1,08%

ING Belgium SA/NV-ING 2,00% 1,43% 0,39% 0,18% 0,10% 1,20% KBC Bank NV 1,55% 1,67% 0,57% -0,68% 0,48% 1,43% Keytrade Bank SA/NV 2,83% 1,40% 1,24% 0,20% 0,15% 1,26%

Onderling Beroepskrediet-OBK-Bank C.V.B.A. 1,38% 1,48% -0,10% 0,00% 0,04% 1,74% Record Bank SA/NV 0,99% 1,33% -0,35% 0,01% 0,04% 0,65% Santander Benelux SA/NV 1,03% 0,85% 0,17% 0,01% 0,00% 0,07%

Société Générale Private Banking N.V. 6,58% 0,86% 5,48% 0,49% 0,00% 5,78% VDK Spaarbank NV 1,73% 1,74% -0,01% 0,04% 1,11%

100

Bank Name 3yaROE10 3yaROA10 3yaLev10 3yaPM10 3yaAY10 3yaCostToIncome10

Agricaisse - Caisse Coopérative de Dépôts et de Crédit Agricole 5,97% 0,37% 15,75 9,40% 3,85% 38,59% Antwerps Beroepskrediet 1,41% 0,70% 2,49 59,12% 3,87% 38,15% Argenta Spaarbank-ASPA 9,19% 0,25% 37,00 6,43% 3,86% 50,93%

Banca Monte Paschi Belgio SA 2,02% 0,14% 15,56 4,47% 3,98% 67,80%

Bank J. Van Breda en Co NV 9,61% 0,76% 12,67 13,97% 5,55% 58,72%

Banque de la Poste-Bank van de Post 7,25% 0,20% 36,11 8,08% 2,61% 78,55% Banque Degroof SA-Bank Degroof NV 10,34% 1,13% 8,98 16,21% 7,20% 62,62% Banque Delen NV 20,50% 2,55% 8,33 38,05% 6,66% 46,26% BKCP -1,39% -0,09% 12,53 -1,39% 4,53% 113,02% Byblos Bank Europe SA 10,23% 0,80% 12,68 19,42% 4,25% 60,71% CBC Banque S.A. 14,05% 0,74% 19,04 15,21% 4,92% 60,47% Centea 13,03% 0,70% 18,44 17,85% 3,96% 47,48% Citibank Belgium N.V./S.A. -9,85% -0,93% 11,40 -8,12% 10,96% 71,47% Crédit Agricole SA-Landbouwkrediet 6,59% 0,42% 15,59 17,41% 2,44% 76,18% Delta Lloyd Bank 2,77% 0,16% 17,29 3,16% 5,06% 96,00%

Deutsche Bank SA-Deutsche Bank NV 6,95% 0,22% 32,06 6,05% 3,58% 85,58% Dexia Bank Belgium-Dexia Bank 0,84% 0,07% 55,03 4,61% 2,97% 79,03% Ethias Bank -9,91% -0,49% 19,32 -10,24% 3,57% 100,60% Europabank 15,58% 1,70% 9,20 24,30% 6,98% 64,16%

101

F. van Lanschot Bankiers Belgie 0,29% 0,04% 7,17 1,64% 4,25% 90,23%

Fortis Bank SA/ NV-BNP Paribas Fortis -42,85% -1,02% 26,69 0,21% 11,03% 93,54%

ING Belgium SA/NV-ING 10,54% 0,66% 16,17 7,52% 9,89% 57,69% KBC Bank NV -6,28% -0,25% 22,34 -5,83% 4,31% 84,97% Keytrade Bank SA/NV 27,59% 0,89% 30,93 33,26% 2,74% 46,47% Onderling Beroepskrediet-OBK-Bank C.V.B.A. -10,07% -0,19% 29,37 -1,99% 4,24% 120,41% Record Bank SA/NV 11,18% 0,34% 32,34 8,93% 3,84% 60,64% Santander Benelux SA/NV 6,26% 0,89% 8,22 31,14% 2,77% 7,62% Société Générale Private Banking N.V. 6,62% 0,71% 9,36 8,20% 8,89% 87,95% VDK Spaarbank NV 3,71% 0,30% 12,45 7,92% 3,89% 65,23%

102

© 2012 KPMG Holding (Belgium), a Belgian civil CVBA/SCRL and a member firm of the KPMG network of independent member firms affiliated with KPMG International Cooperative (“KPMG International”), a Swiss entity. All rights reserved.