KfW Research. No. 17, July MakroScope. The German Banking Industry in International Comparison: Low profitability, high productivity

KfW Research. No. 17, July 2005 MakroScope. The German Banking Industry in International Comparison: Low profitability, high productivity

The German Banking Industry in International Comparison - Low profitability, high productivity - Measured against earnings ratios such as interest margin, cost/income ratio or return on equity, in international comparison Germany s banks and savings banks have had to take a back seat for quite some time. Yet, with regard to labour productivity growth (calculated on the basis of national accounts indicators) they beat most other domestic economic sectors, placing them far above the national average. In international comparison only the productivity gains of Japan s banking industry are slightly higher, whereas Germany s banks surpass not only the banks in the other EU-15 countries (including the UK, France, Italy, the Netherlands, Spain and Sweden) but also those in the USA by a wide margin in terms of productivity both per person engaged and per hour worked. Measured against the macroeconomic resources that are used, Germany is not over-banked. As in the other large economies, the banking industry employs around 2% of the workforce. All in all the results support the hypothesis that, in global comparison, German banks and savings banks are efficient despite the rigid threepillar structure of Germany s banking industry. In addition, they show that this is tied to very intense price competition that keeps driving the banks and savings banks to create innovations and to rationalise. A similarly distinct transfer of productivity gains to the demand side in the form of falling prices for bank services as is the case in Germany could not be observed in any of the other countries included in the analysis. It cannot be excluded that the low supplier concentration in Germany plays a role in this. Owing to the key position of the banking industry in an economy that is based on a strict division of labour, the productivity gains have a positive effect on the economy as a whole. In spite of this, consolidation could become necessary in order to provide additional investment and innovation incentives via higher profitability. These incentives, in turn, would enable further productivity gains for the banking industry. Yet, if therefore scale and synergy effects could be realised, the prices for bank services would remain below the competitive prices of the status quo despite lower competitive pressure. 1 Introduction. It is impossible to imagine a modern economy without a stable and efficient financial system that organises payment processing as well as debtor-creditor and capital relationships. After all, in an economic structure marked by a high degree of division of labour, having a broad range of financial services adapted to the needs of businesses, start-ups, financial investors, savers and consumers - is essential for growth and prosperity. Therefore, the banking industry takes on a key role, and its performance not only has a heavy influence on its own economic success but also directly or indirectly affects all other economic sectors. In view of this key role, which is accompanied by a high level of overall economic impact and responsibility on the part of the credit institutions, the heated discussion on the condition and future of Germany s banks and savings banks that has been going on for quite some years is just as justified as it is necessary. The assessment of the banking industry s performance is usually based on earnings ratios such as interest margin, cost/income ratio or return on equity, which are compared between countries either on the level of individual credit institutions or on the aggregate sector level. Nearly all of the assessments of this kind produce the same result: in international comparison, Germany's banking industry is characterised by weak returns and low profitability. In this connection, critics tend to refer to the rigid three-pillar banking system comprising savings banks, cooperative banks and private banks, and they call the German market structurally 'over-banked.' According to their theory, the high share of credit

The German Banking Industry in International Comparison Page 2 institutions whose goal is not solely to generate a maximum profit leads to inefficiencies, as indicated by an excessively high number of bank branches, low margins and related low profitability. Yet, a small profit on the microeconomic level is not necessarily caused by inefficiency. It may also be the result of intense competition. From an overall economic point of view, competition is healthy since it means permanent incentives to rationalise and to produce innovations. Over and above that it also ensures that the products and services are not overpriced, as is the case in an oligopoly market, but are instead offered at near-cost prices. Aggregate factor productivity is one way to measure how efficiently an industry makes use of its production factors. The development of aggregate factor productivity is looked at in detail in Section 3 on the basis of national account indicators. Section 2 first takes a brief look at the development of the earnings situation of Germany s banks and savings banks compared to foreign banks. 2 Earnings Situation of the German Banking Industry. An important indicator of the earnings situation of the banking industry is the interest margin. It compares the interest surplus (difference between interest income and interest expense) to the average balance-sheet total. As shown in Chart 1, Germany s and France s credit institutions take a back seat in terms of their interest margin. Yet, the interest margin did not become tighter over time in Germany alone. In the 1990s this trend could be observed everywhere. However, it must be said that the interest margin of German credit institutions declined more than in the other countries included in the comparison. In the period between 1993 and 2001, it contracted in Germany by a total of almost 40%, yet in the UK and in France only by almost 28%, in Italy by 15% and in the USA by a mere 12%. Chart 1. Margin of interest of the banking industry* in selected countries 4 3.5 Net interest income in percent of average total assets 3 2.5 2 1.5 1 0.5 0 USA Italy UK Germany France 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 * USA, UK: Commercial banks only Source: Deutsche Bundesbank, OECD The interest margin is not suited as the sole success criterion, however. Interest surplus as the numerator is, on its own, an expression of the intermediation earnings generated in lending business. Yet, a bank's business success is based on other earnings as well, such as commissions in

The German Banking Industry in International Comparison Page 3 investment business or the result of own trade in securities. However, the interest surplus is based on the entire balance-sheet total, which reflects the bank s entire banking business. Consequently, the interest margin can send distorted signals: banks and savings banks with smaller interest margins do not necessarily have weak returns. It should also be pointed out that the interest earnings and expenses may differ depending on the accounting standards used as a basis (German Commercial Code/HGB, IFRS, US-GAAP). A broader range for describing the earnings power of credit institutions is the ratio of general administrative expenditure i.e. personnel and material expenses to the operating result, the socalled cost/income ratio (CIR; see Chart 2). From 1993 until 1998 the CIR of Germany s banks and savings banks was about 63%, thus below the European average of 65%. During the following stock market boom the CIR in Germany increased steeply, since German banks obviously did not succeed in matching the higher costs for new information technologies and for the expansion of their investment banking business with corresponding growth in earnings. 1 In the years 2002 and 2003 the CIR decreased year-on-year, yet a global downwards trend can be noted since 1998, too. So, in terms of their cost/income structure, it can be said that German banks still have some catching up to do. In international comparison, a CIR of around 60% is generally considered to be competitive. Chart 2. Cost/income ratio of the banking industry* in selected countries 75 Personnel and material expenses in precent of interest and non-interest income 70 65 60 55 50 USA Italy UK Germany France 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 * USA, UK: commercial banks only Source: OECD, Heise/Holzhausen (2004), Die Bank The below-average earnings situation of Germany s banks compared to banks in other countries is also clearly demonstrated by the development of profitability indicators (see Chart 3). Measured against the return on equity (RoE) of the banking industry as a whole - in which global investors are keenly interested on average, German banks and savings banks clearly lag behind the RoE of international banks. The contrary development of profitability in Germany, on the one hand, and in France and Italy, on the other hand, becomes quite obvious. Whereas in France and Italy the reforms in the banking sector were followed by a rise in yield from a very low level, the return on equity in 1 See also Heise, M. and A. Holzhausen (2004), Economic Research Allianz Group Dresdner Bank, Working Paper No. 19, p. 4.

The German Banking Industry in International Comparison Page 4 Germany was on a steady decline between the mid-1990s and 2003 with the exception of the year 1998. Chart 3. Return on equity of the banking industry* after taxes in selected countries 25 Profit after taxes in percent of equity 20 15 10 5 USA Italy UK Germany France 0-5 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004** * USA, UK: commercial banks only; USA: no figure available for 2002 ** Estimated figures Source: OECD, DB Research, DekaBank, Die Zeit; own calculations Apart from the various types of cost and income as those compared to calculate the CIR, the return on equity contains the entire net income for the year and, thus, also the risk provisions that have been made. The weak profitability of the German banking industry in the past years is due not least to this item, which is naturally influenced by the amount of defaults and therefore also by the cyclical situation. In 2003 the net expenses for the valuation of loans, receivables and securities still absorbed almost 20% of the operating profit, after absorbing nearly 27% in 2002 (see Chart 4). In addition, one structural cause of the weak earnings situation of the German banking industry that is often cited is the low degree of concentration within the industry. Measured as the share of the five largest banks of the aggregate balance sheet total of the entire banking industry, the figure for Germany is approx. 20% and thus lower than in the UK and Italy (both 30%), France (50%) and Spain (55%). However, such a concentration indicator albeit widespread is informative to only a limited extent since balance-sheet totals do not necessarily reflect the market shares of domestic business. In addition, economically it is possible to although we consider this neither appropriate nor fitting evaluate savings banks and cooperative banks (including their central institutions) on their own as separate suppliers. Looked at from this perspective, Germany's banking industry would hardly appear fragmented at all, with an extremely high degree of concentration of around 75%.

The German Banking Industry in International Comparison Page 5 Chart 4. Net provisions of the German banking industry and real GDP growth 30 3.6 Net provisions in percent of operating result 25 20 15 10 5 Net provisions (LHS) Real GDP growth (RHS) 3 2.4 1.8 1.2 0.6 Real GDP recentage change y-o-y, adjusted for calendar effects 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 0 Source: Deutsche Bundesbank, Statistisches Bundesamt; own calculations The - on average low earnings and lack of profitability pose a problem mainly for those banks whose shareholders expect a yield that can compete internationally. Yet, not only global investors but also domestic savers and borrowers are interested in a banking industry that performs well. In the end an efficient financial system has a positive effect on the entire economy. This justifies adding a macroeconomic perspective to the microeconomic analysis of the banking industry, which would focus on using factor productivity as a gauge for efficiency. 3. Productivity in the Banking Industry A Comparison based on National Accounts Indicators. Up until now the aspect of productivity has, for the most part, been neglected in the discussion on the structural change affecting Germany's banking industry. Specifically, productivity analyses usually focus on labour productivity, defined as real gross value added per person engaged ('productivity per person engaged') or per hour worked ('productivity per hour worked'). On the basis of the 60-Industry Database of the Dutch Groningen Growth and Development Centre (GGDC), this section analyses the development of labour productivity in the German banking industry in a national and international context. The scientifically recognized GGDC database contains internationally comparable indicators for determining the labour productivity of 56 economic sectors in 28 globally important economic areas (26 OECD countries, EU-15 and Taiwan) that are derived from the national accounts data. 2 Since it is highly relevant for a well-founded assessment of the results to understand the national accounts instruments used to calculate the gross value added of the banking industry, the basic elements of the method of approach are briefly described in Box 1. This analysis was based on the assumption that the modern information and communication technologies (ICT) in conjunction with complementary organisational measures have brought about sustainable changes in the processes within the banking industry worldwide and, consequently, have 2 See http://www.ggdc.net/dseries/60-industry.html.

The German Banking Industry in International Comparison Page 6 had considerable impact on the development of productivity, not only of German but also of international banks. Some examples are the pricing of financial products, rating, risk management, management of asset portfolios, new ways to reach customers such as Internet banking, and many more. The indicators presented at the end shed some light on and even help to answer the question of whether the traditional structure of Germany's banking industry prevented the growth of productivity of our banks and savings banks from keeping step with global growth. Box 1. Calculating the nominal and real gross value added of the banking industry in the national accounts. The calculation of the nominal and real gross value added of the banking industry in the national accounts has several inherent problems. In order to solve them, procedures are applied that require explanation to enable a well-founded assessment of the results. To achieve the greatest possible transparency, the fundamentals of these approaches are explained below. The explanations are oriented towards the approach applied in the national accounts prior to the major revision on April 28, 2005, since the GGDC indicators used in this analysis correspond to that set of data. In principle, under the output method of the national accounts gross value added is defined as a production value less intermediate inputs. 3 Determining the production value of units that do not report any turnover (within the meaning of a sold quantity of goods multiplied by their price) is problematic. Apart from the state and the insurance branch, this mainly applies to the banking industry. In the national accounts, its primary function is defined as the collection of financial resources for the purpose of investing them or lending them out in exchange for interest. In most cases, an explicit fee is not collected for this service; instead, the credit institutions generate results by charging higher interest for the loans they grant than the interest they pay on deposits. This difference between interest earned (incl. dividends but excluding earnings for own funds), on the one hand, and paid interest on the other hand is interpreted as a service fee and, in the form of so-called FISIM (Financial Intermediation Services Indirectly Measured), accounts for the majority of the production value. 4 Added to this are what are known as revenues from secondary activities (i.e. sales of goods and services or self-made goods) to arrive at the overall nominal production value. These secondary sales activities (such as renting out the vault) do not pose any conceptual difficulty since they require the payment of a set fee (e.g. commission income, securities account charge). If the nominal production value calculated as described above is deducted from nominal intermediate inputs (without paid interest), the result is the nominal gross value added. As is the case with the secondary activities or bank services performed in exchange for a fee, the intermediate inputs (e.g. commissions and similar charges payable, other administrative expenses) are determined in a conventional manner. A key source of data for these calculations is the Deutsche Bundesbank s statistics on the profit and loss accounts of the credit institutions. Since especially the FISIM cannot be split up into a volume component and a price component, when determining the real gross value added it is not possible to deflate the nominal gross value added using an explicit price deflator. Instead, the national accounts take an indirect path by using a volume indicator. 5 The sum of the average deposit and lending volumes of the credit institutions is used as a nominal output indicator. In a second step, it is deflated with the price index for the final domestic expenditure. The volume indicator generated in this way is introduced into a row of indices via which the FISIM of the year of reference is extrapolated. In addition, the bank services performed in exchange for a fee are split up into a value-dependent component (70%) and a quantity-dependent component (30%). As was the case with the output indicator for the FISIM, the value-dependent component is deflated via the deflator for the most recent domestic expenditure. On the other hand, the quantity-dependent component is deflated using the price index for the services provided by the banks, which is part of the consumer price index. 3 See also Statistisches Bundesamt, Fachserie 18, Reihe p. 22, esp. pp. 93-99. 4 A major change introduced by the latest revision of the national accounts is the breakdown of the FISIM into the aggregates of GDP expenditure instead of deducting them in a lump sum as intermediate inputs from the unadjusted gross value added for all industries taken together, as used to be the case. This new breakdown did not have any serious effects on the calculation of the value added for the banking industry alone, which is a key component of the productivity analysis. 5 See also Statistisches Bundesamt, Fachserie 18, Reihe p. 24, esp. pp. 25 ff.

The German Banking Industry in International Comparison Page 7 These three components that are deflated separately (FISIM, value-dependent and quantitydependent share of the secondary activities) are added and, taken together, represent the real production value of the banking industry. If the conventionally deflated intermediate inputs are deducted, the result is the real gross value added. By dividing the nominal by the real gross value added, it is possible to generate an implicit price deflator for the value added of the banking industry. In international comparison the methods used, especially for deflation, differ quite substantially in some cases. The differences are noticeable primarily in the levels of real value added and therefore also in the labour productivity. For this reason, such figures were left out of the analyses in this paper. In addition, small differences in the development of productivity should not be overestimated, since it is difficult to decide beforehand the extent to which various methods ought to play a detailed role in this. Instead, major differences such as those described below in this paper are probably a sign of real material differences. 3.1 National Perspective. First the national dimension of this issue needs to be examined more closely. How does the productivity growth of the banking industry compare to that of other industries within Germany? Table 1 shows the quantitative importance of the banking industry; the qualitative role of banks and savings banks for the aggregate economy has already been discussed in the introduction. Measured in terms of nominal value added of almost EUR 50 billion in the year 2002 more current data on the breakdown of value added calculations by economic sector was not available at the time the banking industry had a structural share of 2.5% of the total economy, or 4.2% of services (excluding trade). As a result, it ranks 13 th of the 56 sectors included and 8 th among the service sectors. Measured in terms of the total number of persons engaged or against the total hours worked, these figures are slightly lower: 2.0% of the total economy and approx. 3 ¾% of services. Table 1. The quantitative importance of the banking industry a in Germany Share in Levels 2002 all industries (56) b service industries (20) c Unit Value [%] Rank [%] Rank Nominal value added [EUR million] 49,232 2.5 13 4.2 8 Total persons engaged [thousands] 785 2.0 18 3.7 9 Total hours worked [thousands] 1,081,640 2.0 18 3.8 9 Contribution to total growth Share in total growth rate of 1991-2002 all industries (56) b service industries (20) c Unit Value [%] Rank [%] Rank Nominal value added [percentage p.] -0.1-0.3 50-0.3 20 Real value added d [percentage p.] 1.4 7.6 6 10.8 4 Total persons engaged [percentage p.] 0.1 15.8 15 0.9 13 Total hours worked [percentage p.] -0.1-1.4 27-1.1 17 a ISIC rev 3 code 65 (Financial intermediation, except insurance and pension funding) b ISIC rev 3 codes 01 (Agriculture) to 95 (Private households with employed persons) c ISIC rev 3 codes 55 (Hotels and catering) to 95 (Private households with employed persons), i.e. excluding trade sector d Chained 1995 prices If, instead of the nominal levels in 2002, the contribution to the growth of real value added in the period being looked at 1991 until 2002 or, in other words, from the first year for which aggregate data for unified Germany is available until the data cut-off the banking industry moves up two notches to 6 th place among all industries: With a contribution to total growth of 1.4 percentage points, it has a 7.6%

The German Banking Industry in International Comparison Page 8 share of real growth over the entire period of 18.7%. 6 Compared to the banking industry s structural share, its contribution to growth is above average in the face of an almost stagnating number of persons engaged and an even slightly decreasing labour volume. This is already a first indication of significant labour productivity growth in the banking industry. Table 2 shows the calculation of the growth of labour productivity. In addition to the rates of change in productivity per person engaged and per hour worked expressed as both a total figure for the entire period and as an annual average, the corresponding rates of change of the individual components (i.e. nominal gross value added, price deflator, total persons engaged and total hours worked) are also shown. These indicators were determined not only for the banking industry but also for the aggregate economy and for the service industries for the purpose of comparison. As illustrated by the two indices of year-on-year development of productivity per person engaged and per hour worked, for example, in national comparison the productivity growth in Germany s banking industry (blue line) outperforms that of other industries. Before taking an in-depth look at this development, however, another important aspect related to the calculation of the productivity growth requires mentioning. As shown by Table 2, the strong growth in the volume of real value added in the banking industry (almost 50% between 1991 and 2002) which was boosted by substantial productivity growth was accompanied by a steep drop in prices. During the same period the price deflator decreased by slightly more than half, or nearly 5% annually. As a result, the value added in current prices stagnated. In other words, the productivity gains were transferred quickly to bank and savings bank customers in the form of lower prices. In the national context this can be considered an achievement, since the development of the price deflators for all economic sectors on average or solely for the service industries was +1.3% annually each between 1991 and 2002. Table 2. Calculation of labour productivity growth in Germany Rate of change a [%] 1991-2002 Annual average Index [1991=100] 1. Value added and prices Nominal value added Labour productivity per person engaged Banking industry -3.0-0.3 160 All industries 32.6 3.0 150 Banking industry Service industries 46.8 4.3 All industries Price deflator 140 Service industries Banking industry -52.6-4.8 All industries 13.9 1.3 130 Service industries 14.2 1.3 120 Real value added Banking industry 49.6 4.5 110 All industries 18.7 1.7 100 Service industries 32.6 3.0 90 2. Employment and labour volume Total persons engaged Banking industry 4.5 0.4 All industries 0.6 0.1 Service industries 20.2 1.8 Labour productivity per hour worked Total hours worked 160 Banking industry -3.3-0.3 150 Banking industry All industries -4.7-0.4 All industries 140 Service industries 12.9 1.2 Service industries 3. Labour productivity 130 Labour productivity per person engaged 120 Banking industry 45.1 4.1 All industries 18.2 1.7 110 Service industries 12.3 1.1 Labour productivity per hour worked 100 Banking industry 52.9 4.8 90 All industries 23.4 2.1 Service industries 19.7 1.8 a Exponential growth rates 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 145.1 118.2 112.3 152.9 123.4 119.7 6 Exponential growth rates were calculated.

The German Banking Industry in International Comparison Page 9 Table 3 contains more detailed information about how the productivity of Germany s banks and savings banks developed compared to other German industries. With annual labour productivity growth of 4.1% per person engaged, the banking industry clearly beats the average for all industries, which is 1.7% (mean) and 1.9 % (median). Compared to the service industries, the banks lead is even larger overall. Among the service industries, for instance, the median branch saw annual growth in productivity per person engaged of 0.3%; the arithmetic mean of 1.1% is slightly higher, yet it is far below the corresponding figure for the banking industry. Table 3. Productivity of the German banking industry in the national perspective Rate of change a [%] 1991-2002 Annual average 1. Labour productivity growth per person engaged Banking industry compared to 45.1 4.1 all industries (56) Mean 18.2 1.7 Median 20.5 1.9 Maximum 518.2 47.1 Minimum -52.5-4.8 Rank banking industry 18 service industries (20) Mean 12.3 1.1 Median 3.8 0.3 Maximum 131.6 12.0 Minimum -52.5-4.8 Rank banking industry 5 2. Labour productivity growth per hour worked Banking industry compared to 52.9 4.8 all industries (56) Mean 23.4 2.1 Median 21.3 1.9 Maximum 510.3 46.4 Minimum -42.8-3.9 Rank banking industry 13 service industries (20) Mean 19.7 1.8 Median 11.5 1.0 Maximum 141.4 12.9 Minimum -42.8-3.9 Rank banking industry 5 a Exponential growth rates b Harmonised US deflators were used in the calculations Industry / Rubber & plastics Electronic valves and tubes b Hotels and catering / Renting of machinery and equipment Water transport Hotels and catering / Railroad equipment and transport equipment nec Electronic valves and tubes b Hotels and catering / Renting of machinery and equipment Water transport Hotels and catering This picture does not change substantially if the productivity per hour worked is used as a basis instead. However, from the banks' perspective, the result is even slightly more favourable. With an annual increase in productivity per hour worked of 4.8%, the banking industry ranks 13 th among the 56 industries included in this comparison, which achieved 2.1% growth on average. Far ahead of the banking industry are the ICT branches as well as those branches of the transport sector that also benefited considerably from the ICT revolution (a list of all branches broken down according to growth in productivity per person engaged and per hour worked is enclosed as an annex). Therefore, at this stage the productivity growth of the banks and savings banks can be said to be more than presentable in national comparison. The productivity growth of the banking industry is higher than for most other economic sectors in Germany. It exceeds the averages of both the service industries and the aggregate economy by far. Contrary to the average of all industries and of the service industries, these productivity gains are passed on to the banks customers in the form of lower prices. But what if we look beyond national borders? Is it possible that German banks and savings banks beat most national industries in terms of productivity growth, but lag behind the credit institution in other major global economies? This question is answered in Section 3.2 below.

The German Banking Industry in International Comparison Page 10 3.2 International Perspective. Before delving into a closer examination of the productivity indicators, Table 4 presents a few structural indicators of the quantitative importance of the banking industry in some of the world's dominant economies selected for the purpose of comparison - the USA, Japan and EU-15 (excluding Germany). Contrary to the common opinion that Germany has too many banks, the structural shares illustrate that there are no more macroeconomic resources tied up in the German banking industry than in other major economies. Around 2% of the persons engaged in the USA, Japan and in EU-15 work in the banking industry. In the other large EU member states this figure ranges from 1.6% (Spain) to 2.4% (UK). Only in Sweden is this figure somewhat lower (1.3%). In terms of nominal value added, the banks and savings banks in Germany and in Sweden even have the lowest share of all the countries included. Thus, in macroeconomic terms Germany measured against international standards is not at all over-banked. Table 4. The quantitative importance of the banking industry a in major world economies Nominal value added Share in nominal value added of b levels 2002 all industries service industries c Unit Value [%] Rank [%] Rank Germany [EUR million] 49,232 2.5 13 (of 56) 4.2 8 (of 20) EU-15 (excluding Germany) [EUR million] 283,858 3.9 9 (of 56) 6.5 6 (of 20) United Kingdom [GBP million] 29,577 3.0 13 (of 55) 5.0 10 (of 19) France [EUR million] 45,435 3.2 10 (of 56) 5.1 7 (of 20) Italy [EUR million] 51,426 4.4 9 (of 56) 7.6 6 (of 20) Spain [EUR million] 29,380 4.5 8 (of 56) 7.8 6 (of 20) Netherlands [EUR million] 15,726 3.8 9 (of 56) 6.4 7 (of 20) Sweden [SEK million] 50,596 2.4 15 (of 56) 4.0 10 (of 20) USA [USD million] 500,422 5.1 6 (of 56) 8.1 4 (of 20) Japan [JPY billion] 25,850 5.4 5 (of 54) 9.4 3 (of 18) Total persons engaged Share in total persons engaged by b levels 2002 all industries service industries c Unit Value [%] Rank [%] Rank Germany [thousands] 785 2.0 18 (of 56) 3.7 9 (of 20) EU-15 (excluding Germany) [thousands] 2,498 1.9 16 (of 56) 3.4 10 (of 20) United Kingdom [thousands] 695 2.4 11 (of 55) 3.9 8 (of 19) France [thousands] 426 1.7 18 (of 56) 2.8 11 (of 20) Italy [thousands] 406 1.7 18 (of 56) 3.4 10 (of 20) Spain [thousands] 257 1.6 16 (of 56) 3.2 9 (of 20) Netherlands [thousands] 176 2.1 14 (of 56) 3.5 10 (of 20) Sweden [thousands] 59 1.3 22 (of 56) 2.2 13 (of 20) USA [thousands] 3,134 2.1 13 (of 56) 3.4 8 (of 20) Japan [thousands] 1,274 2.0 14 (of 54) 4.1 8 (of 18) a ISIC rev 3 code 65 (Financial intermediation, except insurance and pension funding) b ISIC rev 3 codes 01 (Agriculture) to 95 (Private households with employed persons) c ISIC rev 3 codes 55 (Hotels and catering) to 95 (Private households with employed persons), i.e. excluding trade sector In the international comparison of productivity performance portrayed in Table 5 Germany s banks and savings banks fare well. Regardless of whether it is based on persons engaged or hours worked, the productivity growth is significantly higher than in the USA and in the "old" EU-15 excluding Germany. Whereas the German banking industry achieved annual growth in productivity per hour worked of 4.8% in the period from 1991 to 2002, the banks in the rest of the European Union together achieved growth of only 3.4%, with US banks even lower at 2.7%. In the selected EU countries examined individually, the UK stands out in a positive sense. With regard to growth in productivity per person engaged its average annual rate of 3.9% almost matches Germany's 4.1%, but in terms of growth in productivity per hour worked (3.6%) the UK s rate is a good one-fourth lower than in Germany. The banks in France and Spain are particularly disappointing, with

The German Banking Industry in International Comparison Page 11 their stagnating or even declining productivity indicators. Only Japan was able to achieve slightly higher annual growth in real value added per hour (5.3%) and per person engaged (5.2%). Additionally, German and also Japanese banks have outperformed average national productivity growth rates of both the aggregate economy and of the service sector better than banks in the USA or in the other EU countries, as indicated by the middle column in Table 5. In view of the wide variety of methods applied in the national accounts calculations as mentioned in Box 1 which makes an international comparison difficult, it is of some comfort that in global terms, the trend of strong productivity growth of the German banking industry is also confirmed by other studies using alternative approaches. For instance, the McKinsey Global Institute lists growth of productivity per hour worked in retail banking as 7.5% annually for Germany s banking industry in the years 1994 to 2000, compared to only 5.5% for French banks and 4.9% for US banks. For the output a volume measure is used that comprises the real lending volume and the number of transactions. 7 At a KfW workshop on June 27, 2005 the Dutch economist R. Inklaar presented a study that has not yet been completed in which he draws the following interim conclusion on the basis of various indicators: (The) weight of evidence still suggests productivity performance of German banking is relatively strong. 8 Table 5. Productivity of the German banking industry in the international perspective Rate of change a [%] 1991-2002 Annual average of all industries of service ind. 1. Labour productivity per person engaged Germany 45.1 4.1 2.5 3.0 EU-15 (excluding Germany) 35.6 3.2 1.4 2.1 United Kingdom 42.9 3.9 1.6 2.4 France -9.8-0.9-1.9-1.0 Italy 28.6 2.6 1.2 2.0 Spain 2.0 0.2-0.7 0.0 Netherlands 19.9 1.8 1.1 1.4 Sweden 23.7 2.2 0.0 1.5 USA 30.2 2.7 0.8 1.8 Japan 57.0 5.2 3.5 4.3 2. Labour productivity per hour worked Germany 52.9 4.8 2.7 3.0 EU-15 (excluding Germany) 37.8 3.4 1.2 2.0 United Kingdom 39.1 3.6 1.0 1.9 France 0.6 0.1-1.9-0.9 Italy 32.0 2.9 1.3 2.1 Spain 3.4 0.3-0.7 0.0 Netherlands 20.6 1.9 0.5 1.1 Sweden 25.5 2.3 0.3 1.9 USA 30.2 2.7 0.8 2.0 Japan 58.8 5.3 2.8 2.8 a Exponential growth rates Deviation of annual rate of change from mean [Percentage p.] Index [1991=100] 160 150 140 130 120 110 100 Germany EU-15 (excluding Germany) USA Japan 90 1991 1993 1995 1997 1999 2001 160 150 140 130 120 110 100 Germany EU-15 (excluding Germany) USA Japan 157.0 145.1 135.6 130.2 158.8 90 1991 1993 1995 1997 1999 2001 152.9 137.8 130.2 The right-hand column of Table 5 shows the progression of productivity in the form of an index, which makes the development profile easy to see. It is easy to discern that in all of the economies included but above all in Germany - productivity grew particularly fast in the second half of the 1990s, prior to the end of the so-called new economy boom in the year 2000, when it came to a halt. The productivity curve dropped by 2002 in all of the countries included, except for Japan. The reason for this may be that the irrational exuberance at the beginning of this millennium was not as prevalent in 7 See McKinsey Global Institute (2002), Reaching higher productivity growth in France and Germany, Sector case: Retail banking, pp. 7 and 13. 8 See www.rug.nl/economie/faculteit/medewerkers/inklaar/pdf/kfwpresentation.pdf, p. 15.

The German Banking Industry in International Comparison Page 12 Japan, and that the Japanese banking industry has faced strong permanent pressure to rationalise since the property and stock market bubble burst in the early 1990s. As to what extent the customers benefit from the productivity gains of the credit institutions in the form of lower prices, there are huge differences from one country to the next, as shown by the indicators of the GGDC (Table 6). Whereas the increase in the price deflator for the value added of the US banking industry was comparatively strong in the years 1991-2002, namely 4.4% annually, the increase in the average for all other EU countries was - in the face of annual growth of 0.6% - restrained, and the price deflator for Japan even declined (-0.6%). Yet, Germany experienced a price decrease that, at - 4.8% p.a., was stronger than in all other economies included in this comparison. The only other appreciable price decrease occurred in Sweden (-1.8%). In contrast, the price indicators for the larger EU members, primarily Spain (5.5%) and the Netherlands (4.9%) rose considerably, yet nearly stagnated in the UK (0.4%) and Italy (0.2%). Compared to the deflators for all industries and for the service industries, only Sweden kept its prices below the national average to about the same extent as Germany (see middle column in Table 6). Table 6. Price deflators of the banking industry's value added in the international perspective Deviation of annual rate of change Rate of change a [%] from mean [Percentage p.] Index [1991=100] 1991-2002 Annual average of all industries of service ind. Germany -52.6-4.8-6.0-6.4 EU-15 (excluding Germany) 7.0 0.6-1.6-2.3 160 140 148.8 United Kingdom 3.9 0.4-2.1-3.0 France 29.9 2.7 1.3 0.4 Italy 1.9 0.2-2.7-3.5 Spain 60.0 5.5 1.7 0.8 Netherlands 54.4 4.9 2.7 1.8 Sweden -19.9-1.8-3.9-5.6 USA 48.8 4.4 2.5 1.3 Japan -6.2-0.6 0.5-0.9 a Exponential growth rates 120 100 80 60 Germany 40 EU-15 (excluding Germany) USA 20 Japan 0 1991 1993 1995 1997 1999 2001 107.0 93.8 47.4 In the USA, the Netherlands, Spain and France the prices for bank services even increased more than the mean for the other national economic sectors. Therefore, the intensity of competition within the banking market at least with regard to price competition and pressure to rationalise, which increases productivity in Germany in particular is probably far higher than in the other countries mentioned, yet also higher than in Japan, the UK, Italy or Sweden. Above all, Germany s demand side reaps the benefits of this. 4 Conclusion. Measured against earnings ratios such as interest margin, cost/income ratio or return on equity, in international comparison Germany s banks and savings banks have had to take a back seat for quite some time. In contrast, based on national accounts indicators the results of productivity per person

The German Banking Industry in International Comparison Page 13 engaged and per hour worked that are presented in this paper do not suggest that the German banking industry has a productivity problem: The banks and savings banks are ahead of most other German economic sectors, and their productivity growth exceeds the national average by a wide margin. In international comparison only the productivity gains of Japan s banking industry are slightly higher, whereas Germany s banks surpass not only the banks in the other EU-15 countries (including the UK, France, Italy, the Netherlands, Spain and also Sweden) but also those in the USA by a wide margin. Measured against the macroeconomic resources that are used, Germany is not over-banked. As in the other large economies, the banking industry employs around 2% of the workforce. In Germany, it is mainly the customers who benefit from productivity gains in the banking industry in the form of lower prices - not the banks' shareholders. For instance, the GGDC figures prove that the strong rise in real gross value added made possible by the productivity gains goes hand in hand with declining prices for bank services. This price slide is so steep that the nominal gross value added, which is a determining factor for return on equity, has stagnated. 9 Such a distinct transfer of the productivity gains to the demand side could not be found in any of the other countries included in the calculations. All in all the results support the hypothesis that, in global comparison, German banks and savings banks are efficient despite the rigid three-pillar structure of Germany s banking industry. They also illustrate that there is highly intense price competition, which encourages the banks and savings banks to develop innovations and undertake efforts to rationalise on an ongoing basis. The fruits of these efforts benefit their customers much more so than in the other countries included in the calculations. 10 It cannot be excluded that the low supplier concentration in Germany plays a role in this. Owing to the key position of the banking industry in an economy that is based on a strict division of labour, the productivity gains have a positive effect on the economy as a whole. In this way, Germany s credit institutions fulfil their economic task and need not shun an international comparison in this area. Yet, the political recommendations implied by this result are not quite as clear. A consolidation could become necessary in spite of the positive productivity figures in order to provide additional investment and innovation incentives via higher profitability. These incentives, in turn, would enable further productivity gains for the banking industry. Greater concentration on the supply side would then gradually lead to oligopoly prices. Yet, if at the same time scale and synergy effects could be realised, the prices for bank services could remain below the competitive prices of the status quo despite lower competitive pressure. The deciding factor in this scenario is, however, open market access. Edited: July 1, 2005 Whom to contact: Dr. Klaus Borger, +49 (0)69 / 7431-2455, klaus.borger@kfw.de Dr. Tobias Rehbock, +49 (0)69 / 7431-2686, tobias.rehbock@kfw.de 9 Apart from the concepts used in microeconomic and macroeconomic accounting that differ in their detail, the following definition applies: gross value added corresponds to the sum of the operating result, employee salaries, charges and depreciation each one nominal (current prices) or real (adjusted for price changes). The return on equity, on the other hand, compares the nominal operating result to the capital used. 10 The strong productivity growth of the Japanese banks, which have been under heavy pressure to rationalise since the beginning of the 1990s, can also be understood in this sense.

The German Banking Industry in International Comparison Page 14 Annex Table 7. German industries ranked by average annual rate of growth of labour productivity per person engaged 1991-2002 Rank Industry ISIC rev 3 Growth [%] Rank Industry ISIC rev 3 Growth [%] 1 Electronic valves and tubes 321 47.1 31 Printing & publishing 22 1.7 2 Office machinery 30 36.6 32 Inland transport 60 1.7 3 Water transport 61 12.0 33 Public administration and defence; compulsory social security 75 1.6 4 Communications 64 11.5 34 Research and development 73 1.4 5 Air transport 62 9.1 35 Other electrical machinery and apparatus nec 31-313 1.1 6 Mineral oil refining, coke & nuclear fuel 23 8.5 36 Fabricated metal products 28 1.0 7 Telecommunication equipment 322 8.5 37 Food, drink & tobacco 15-16 0.9 8 Agriculture 01 6.3 38 Health and social work 85 0.8 9 Forestry 02 6.3 39 Fishing 05 0.5 10 Basic metals 27 5.7 40 Renting of machinery and equipment 71 0.3 11 Electricity, gas and water supply 40-41 5.0 41 Retail trade, except of motor vehicles and motorcycles; repair of personal and household goods 52 0.3 12 Aircraft and spacecraft 353 5.0 42 Motor vehicles 34 0.1 13 Chemicals 24 4.9 43 Construction 45 0.0 14 Supporting and auxiliary transport activities; activities of travel agencies 63 4.8 44 Education 80-0.1 15 Building and repairing of ships and boats 351 4.5 45 Private households with employed persons 95-0.3 16 Leather and footwear 19 4.4 46 Mining and quarrying 10-14 -0.3 17 Clothing 18 4.3 47 Activities auxiliary to financial intermediation 67-0.4 18 Financial intermediation, except insurance and pension funding 65 4.1 48 Furniture, miscellaneous manufacturing; recycling 36-37 -0.5 19 Computer and related activities 72 3.2 49 Other community, social and personal services 90-93 -1.0 20 Radio and television receivers 323 3.2 50 Other business activities, nec 749-1.4 21 Wood & products of wood and cork 20 3.1 51 Legal, technical and advertising 741-3 -2.0 22 Scientific instruments 331 2.9 52 Real estate activities 70-2.1 23 Non-metallic mineral products 26 2.9 53 Insurance and pension funding, except compulsory social security 66-2.2 24 Textiles 17 2.5 54 Sale, maintenance and repair of motor vehicles and motorcycles; retail sale of automotive fuel 50-2.3 25 Pulp, paper & paper products 21 2.4 55 Insulated wire 313-3.5 26 Mechanical engineering 29 2.3 56 Hotels & catering 55-4.8 27 Railroad equipment and transport equipment nec 352+359 2.2 28 Other instruments 33-331 2.0 29 Rubber & plastics 25 1.9 30 Wholesale trade and commission trade, except of motor vehicles and motorcycles 51 1.7 Table 8. German industries ranked by average annual rate of growth of labour productivity per hour worked 1991-2002 Rank Industry ISIC rev 3 Growth [%] Rank Industry ISIC rev 3 Growth [%] 1 Electronic valves and tubes 321 46.4 31 Printing & publishing 22 1.9 2 Office machinery 30 34.7 32 Rubber & plastics 25 1.7 3 Water transport 61 12.9 33 Textiles 17 1.5 4 Communications 64 12.4 34 Health and social work 85 1.4 5 Air transport 62 10.0 35 Other instruments 33-331 1.3 6 Mineral oil refining, coke & nuclear fuel 23 8.4 36 Retail trade, except of motor vehicles and motorcycles; repair of personal and household goods 52 1.2 7 Telecommunication equipment 322 7.8 37 Food, drink & tobacco 15-16 1.1 8 Supporting and auxiliary transport activities; activities of travel agencies 63 5.7 38 Renting of machinery and equipment 71 1.0 9 Basic metals 27 5.5 39 Fabricated metal products 28 0.9 10 Agriculture 01 5.5 40 Other electrical machinery and apparatus nec 31-313 0.8 11 Forestry 02 5.5 41 Education 80 0.5 12 Electricity, gas and water supply 40-41 5.2 42 Motor vehicles 34 0.5 13 Financial intermediation, except insurance and pension funding 65 4.8 43 Construction 45 0.4 14 Aircraft and spacecraft 353 4.7 44 Private households with employed persons 95 0.3 15 Chemicals 24 4.6 45 Activities auxiliary to financial intermediation 67 0.3 16 Building and repairing of ships and boats 351 4.3 46 Mining and quarrying 10-14 -0.1 17 Clothing 18 4.1 47 Fishing 05-0.2 18 Computer and related activities 72 3.9 48 Other community, social and personal services 90-93 -0.4 19 Wood & products of wood and cork 20 3.3 49 Furniture, miscellaneous manufacturing; recycling 36-37 -0.4 20 Leather and footwear 19 2.8 50 Other business activities, nec 749-0.7 21 Non-metallic mineral products 26 2.7 51 Legal, technical and advertising 741-3 -1.3 22 Wholesale trade and commission trade, except of motor vehicles and motorcycles 51 2.6 52 Real estate activities 70-1.4 23 Inland transport 60 2.6 53 Sale, maintenance and repair of motor vehicles and motorcycles; retail sale of automotive fuel 50-1.4 24 Pulp, paper & paper products 21 2.5 54 Insurance and pension funding, except compulsory social security 66-1.5 25 Radio and television receivers 323 2.5 55 Insulated wire 313-3.8 26 Scientific instruments 331 2.3 56 Hotels & catering 55-3.9 27 Public administration and defence; compulsory social security 75 2.2 28 Research and development 73 2.1 29 Railroad equipment and transport equipment nec 352+359 1.9 30 Mechanical engineering 29 1.9