Discussion Paper. Which banks are more risky? The impact of loan growth and business model on bank risk-taking Matthias Köhler - PDF

Description
Discussion Paper Deutsche Bundesbank No 33/2012 Which banks are more risky? The impact of loan growth and business model on bank risk-taking Matthias Köhler Discussion Papers represent the authors personal

Please download to get full document.

View again

of 20
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information
Category:

Public Notices

Publish on:

Views: 20 | Pages: 20

Extension: PDF | Download: 0

Share
Transcript
Discussion Paper Deutsche Bundesbank No 33/2012 Which banks are more risky? The impact of loan growth and business model on bank risk-taking Matthias Köhler Discussion Papers represent the authors personal opinions and do not necessarily reflect the views of the Deutsche Bundesbank or its staff. Editorial Board: Klaus Düllmann Heinz Herrmann Christoph Memmel Deutsche Bundesbank, Wilhelm-Epstein-Straße 14, Frankfurt am Main, Postfach , Frankfurt am Main Tel Please address all orders in writing to: Deutsche Bundesbank, Press and Public Relations Division, at the above address or via fax Internet Reproduction permitted only if source is stated. ISBN (Printversion) ISBN (Internetversion) Abstract In this paper, we analyze the impact of loan growth and business model on bank risk in 15 EU countries. In contrast to the literature, we include a large number of unlisted banks in our sample which represent the majority of banks in the EU. We show that banks with high rates of loan growth are more risky. Moreover, we find that banks will become more stable if they increase their non-interest income share due to a better diversification of income sources. The effect, however, decreases with bank size possibly because large banks are more active in volatile trading and off-balance sheet activities such as securitization that allow them to increase their leverage. Our results further indicate that banks become more risky if aggregate credit growth is excessive. This even affects those banks that do not exhibit high rates of individual loan growth compared to their competitors. Overall, our results indicate that differences in the lending activities and business models of banks help to identify risks, which would only materialize in the long-term or in the event of a shock. JEL-Classification: G20, G21, G 28 Keywords: Banks, risk-taking, business model, loan growth Non-technical Summary In this paper, we analyze the impact of lending growth and business model on bank risk in 15 EU countries. In contrast to the literature on this issue which mainly focuses on large and listed banks, we include a large number of smaller unlisted banks in our sample which represent the majority of banks in the EU. We think that this is important for the broader applicability of the results. We also think that our sample should better allow us to identify the effects of loan growth and banks business models on bank risk, since unlisted differ markedly in their lending behavior and business model from listed banks. Controlling for endogeneity, bank-, year- and country-specific effects we find that it is important to enlarge the number of banks and bank types to come to general conclusions about the effect of banks business model on risk-taking in the EU banking sector. While the previous studies suggest that it may be beneficial for banks to reduce their share of non-interest income, our results indicate the opposite. This finding is consistent with the common view that European banks are better able to exploit the diversification potential of fee-based activities due to their experience with universal banking models than US banks. The diversification effect of a higher share on non-interest income, however, decreases with bank size possibly because larger banks are more likely to be active in volatile and risky trading and off-balance sheet activities such as securitization that allows them to employ a higher financial leverage than small banks. Finally, our paper indicates that supervisors should carefully monitor loan growth on the individual level, since high rates of loan growth are associated with of bank risk-taking. Moreover, they should be aware of the development of aggregate credit growth, since our results show that banks reduce their lending standards and become more risky during periods of excessive lending growth at the country level. This even affects those banks that do not exhibit high rates of individual loan growth compared to their competitors. With respect to aggregate credit growth our paper, therefore, provides support for the introduction of countercyclical capital buffers which should reduce credit growth and the build-up of systemic risk during booms. Nichttechnische Zusammenfassung In der vorliegenden Arbeit untersuchen wir den Einfluss von Kreditwachstum und Geschäftsmodell auf das Risiko von Banken in 15 EU-Staaten. Im Gegensatz zur bestehenden Literatur zu diesem Thema konzentrieren wir uns nicht nur auf große, börsennotierte Banken, sondern beziehen auch eine Vielzahl kleinerer Banken, die nicht an der Börse notiert sind, in unsere Analyse ein. Da diese Institute die Mehrheit der Banken in Europa repräsentieren, ermöglicht unsere Arbeit allgemeinere Aussagen zum Einfluss des Geschäftsmodells auf das Risiko, das eine Bank eingeht. Während die bisherigen Studien für börsennotierte Banken darauf hindeuten, dass Banken weniger Risiken eingehen, wenn sie ihren Anteil des Nichtzinseinkommens am gesamten operativen Einkommen reduzieren, deuten unsere Ergebnisse auf das Gegenteil hin. Sie stimmen mit der allgemeinen Einschätzung überein, dass europäische Banken die Diversifikationsvorteile, die die Expansion ins Nichtzinsgeschäft bieten, auf Grund ihrer Erfahrungen mit Universalbankenmodellen besser ausnutzen können als US-amerikanische Banken. Der Diversifikationseffekt nimmt jedoch mit zunehmender Größe des Kreditinstituts ab. Ein Grund hierfür könnte sein, dass große Banken stärker im volatilen und riskanten Eigenhandel tätig sind und außerbilanzielle Geschäfte wie Verbriefungen durchführen, die es ihnen erlauben, ihren finanziellen Hebel zu erhöhen. Darüber hinaus zeigen unserer Ergebnisse, dass Bankaufseher das Kreditwachstum von Banken intensiv beobachten sollten, da Banken mit einem hohen Kreditwachstum riskanter sind. Außerdem sollten Aufseher auch die Entwicklung des aggregierten Kreditwachstums im Auge behalten, da die Ergebnisse unserer Studie darauf hindeuten, dass Banken während Phasen exzessiven Kreditwachstums riskanter werden. Das betrifft auch die Banken, die niedrige Kreditwachstumsraten im Vergleich zu ihren Wettbewerbern aufweisen. Insgesamt stützen unsere Ergebnisse somit die Einführung antizyklischer Kapitalpuffer, die das Kreditwachstum und den Aufbau systemischer Risiken in Aufschwungphasen reduzieren sollen. Contents 1 Introduction 1 2 Data 5 3 Bank Risk Taking 6 3 a Bank Characteristics 8 (1) Lending Activity 8 (2) Business Mix 8 (3) Funding Structure 9 3 b Comparison of Bank Characteristics for Different Bank Types 9 3 c Country Characteristics 10 4 Empirical Models 12 5 Results 13 5 a Bank Characteristics and Risk-Taking 14 5 b Country Characteristics and Risk-Taking 16 5 c Does the Effect of Banks Business Mix Differ According to Bank Size? 17 6 Alternative Indicators of Bank Risk 19 7 Conclusions 20 Which banks are more risky? The impact of loan growth and business model on bank risk-taking 1 1. Introduction The financial crisis of 2007/2008 has led to significant losses of banks. However, not all banks were affected equally. In particular, large-complex banking groups with a focus on investment banking recorded large losses (ECB, 2010). Due to their systemic importance their risk-taking behavior has been analyzed frequently in the literature (e.g. Altunbas et al., 2011, Beltratti and Stulz, 2012, Demirgüc-Kunt and Huizinga, 2010 and Laeven and Levine, 2009). Banks with a more traditional banking model, however, suffered large losses as well. In particular, banks with high rates of loan growth reported a significant drop in their performance during the crisis as indicated Figure 1. For example, while the return-on-equity (ROE) of EU banks with the highest average rate of loan growth between 2003 and 2006 decreased, on average, from 13.34% in 2006 to 6.77% in 2008, the ROE of banks with the lowest loan growth rates declined less steeply from 10.46% to 5.65%. Interestingly, while the profitability of the first group of banks dropped further in 2009, the ROE of banks with the lowest rates of loan growth increased. Furthermore, for the first time since 2002 the ROE of banks with the lowest rates of loan growth was higher than the ROE of banks with the highest rates of lending growth. If we consider the drop-off in performance during the crisis as indicator of risk-taking, banks with high loan growth rates seem to have incurred greater risks than banks with low rates of loan growth. In the pre-crisis period, this resulted in a higher profitability of these banks, but in a large decrease in profits in The further decline in bank profitability in 2009 suggests that not all of these risks materialized in 2008, but also in 2009 due to the economic downturn that followed. Since economic growth is still weak and unemployment high, many banks with previously high rates of loan growth continue to report low profitability up to day. Due to pressure from investors and regulators these banks are among those that have to deleverage and change their business model most. Banks with high non-interest income also have to rethink their business model, since non-interest income is highly volatile and led to large losses during the crisis (Liikanen, 2012). This may particularly concern large banks with 1 Deutsche Bundesbank, Wilhelm-Epstein-Straße 14, Frankfurt am Main, Germany. The author would like to thank Tobias Michalak, Nora Srzentic, Christoph Memmel, Heinz Herrmann and the participants at the Bundesbank seminar and the Conference on the Stability of the European Financial System and the Real Economy in the Shadow of the Crisis in Dresden for helpful comments and suggestions. The paper represents the author s personal opinions and does not necessarily reflect the views of the Deutsche Bundesbank or its staff. 1 substantial trading activities (Liikanen, 2012). Smaller banks with a large share of interest income, in contrast, may benefit from higher non-interest income as it may help them to diversify their income sources which should make them less dependent on overall business conditions and more stable. In this paper, we analyze the impact of loan growth and business model on bank risk in 15 EU countries. In contrast to the previous literature that analyzes the impact of bank s business model on risk-taking based on a sample of listed banks, we include a large number of unlisted institutions which represent the majority of banks in the EU. Our results indicate considerable heterogeneity in risk-taking across banks and countries. We show that banks with high rates of loan growth are more risky. We further find evidence that banks will become more stable if they increase their non-interest income share due to a better diversification of income sources. The effect, however, decreases with bank size. This indicates that it is important to enlarge the sample of banks to come to general conclusions about the effect of banks business model on risk. Our results further show that banks become more risky if aggregate credit growth is excessive. This even affects those banks that do not exhibit high rates of individual loan growth compared to their competitors. Overall, our results indicate that differences in the lending activities and the business models of banks help to identify risks, which would only materialize in the long-term or in the event of a shock. While the literature consistently finds that excessive rates of loan growth lead to greater risktaking (see e.g. Foos et al., 2010 and Jimenez and Saurina, 2007), there is less consensus among academics about the impact of a bank s business mix on bank risk. For example, while some argue that an increase in non-interest activities such as investment banking provides banks with additional sources of revenue and can therefore provide a diversification in their overall income which should make them more stable, others argue that banks may also become less stable if they diversify into non-lending activities due to the higher volatility of non-interest income. 2 Evidence from the recent crisis provides support for the latter hypothesis. Altunbas et al. (2011), for example, show that banks with high non-interest income are more risky. Larger banks and those with more aggressive loan growth are less stable as well, while banks with less risk-taking are characterized by a strong deposit base. Demirgüc-Kunt and Huizinga (2010) obtain similar results. They show that banks with a high level of fee and trading income are more risky. Banks that heavily rely on wholesale funding are more risky as well, while Demirgüc-Kunt and Huizinga (2010) find no evidence that high rates of asset growth result into greater risk-taking. Common to both studies is that the impact of a bank s business model on risk is analyzed for a sample of listed banks which are usually larger and more active in non-lending activities than banks not listed such as savings and cooperative banks. 2 Saunders and Walter, (1994), De Young and Roland (2001) and Stiroh (2004) provide detailed literature reviews. For Germany, Busch and Kick (2009) show that the volatility of commercial banks returns significantly increases if they are involved in fee business. There is, however, no evidence that the returns of German savings and cooperative banks become more volatile. 2 We contribute to these papers in three important ways. First of all, in addition to listed banks our dataset includes a large number of unlisted banks. This should give a more representative picture about the European banking sector as unlisted banks represent the majority of banks in the EU. We think this is important for the broader applicability of the results. We also think that our sample should better allow us to identify the effects of loan growth and banks business models on the level of risk-taking, since unlisted banks are usually smaller and have a more traditional business model with a greater focus on lending activities than listed banks. Including unlisted banks also enlarges the number of bank types, since savings and cooperative banks are usually not listed. For example, among the unlisted banks in our sample more than 70% are savings and cooperative banks. The latter do not only have different business models, but also differ in terms of their business objective and ownership structure from commercial banks (Hesse and Cihak, 2007 and Beck et al., 2009). Second, estimations on the determinants of bank risk are impeded by the problem of endogeneity between the variables used to describe a bank s business model and bank risk. We solve this problem by choosing an econometric approach that instruments endogenous variables with their own lags. Furthermore, we allow the risk-taking behavior of banks to be dynamic as bank risk may be persistent over time due to inter-temporal risk smoothing, competition, banking regulations or relationship banking with risky customers (Delis and Kourtas, 2011). Third, even though there are several theoretical papers that show that banks lower their lending standards and collateral requirements during booms (Ruckes, 2004 and Dell Ariccia and Marquez, 2006), the empirical evidence on the impact of lending booms on individual bank risk-taking is limited. Using two different indicators to characterize periods of excessive lending growth, we analyze whether high rates of aggregate credit growth led to an increase in individual bank risk. We follow Altunbas et al. (2011) and Demirgüc-Kunt and Huizinga (2010) and measure bank risk-taking using the Z-Score, defined as the number of standard deviations that a bank s return on asset has to fall for the bank to become insolvent. 3 Our sample shows considerable heterogeneity in risk-taking across banks and countries. We explain this by differences in loan growth and the business model of banks as well as the development of aggregate credit growth. Our results show that loan growth is an important determinant of bank risk. We find evidence that banks with high rates of loan growth are more risky. This may indicate that banks lower their lending standards and collateral requirements to increase loan growth. Furthermore, banks that exhibit significantly higher rates of loan growth than their competitors may attract customers which have not been given a loan by other banks because they asked for too low loan rates or provided not sufficient collateral relative to their credit quality (Foos et al., 2010). 3 Other studies that use the Z-Score to measure bank risk-taking are Laeven and Levine (2009) and Foos et al. (2010). 3 Banks business mix also matters. In contrast to Altunbas et al. (2011) and Demirgüc-Kunt and Huizinga (2010), we show that banks become more stable if they generate a larger fraction of their income from non-interest activities. 4 This effect depends on bank size, however. While smaller banks benefit from the income diversifying effects of a higher non-interest income share, we find the opposite for large banks. We think that this reflects the different sets of non-interest income activities of small and large banks. While large banks are more active in volatile trading activities, smaller banks usually derive a higher share of their income from provisions which are more stable and often linked to interest income due to cross-selling activities (see also Stiroh, 2004). Larger banks may also be more likely to engage in more risky off-balance sheet activities such as securitization which allows them to employ a higher financial leverage than small banks. This is also reflected by our sample which shows a strong negative relationship between bank size, non-interest income share and a bank s capital ratio. Together this may offset the positive effect of a higher non-interest income share and a larger size on bank stability and may ultimately result in greater risk-taking by large credit institutions. Furthermore and in contrast to Altunbas et al. (2011) and Demirgüc-Kunt and Huizinga (2010), we find no evidence that banks that rely on wholesale funding are more risky than banks that primarily fund their activities by customer deposits. The latter are usually considered as a more stable source of funding (Song and Thakor, 2007 and Shleifer and Vishny, 2010). We think that our results are driven by the large number of unlisted banks included in our sample which primarily fund their loans by customer deposits as indicated by relatively low average loan-to-deposit ratios. Moreover, we have no investment banks included. Hence, the risks stemming from the excessive reliance on wholesale funding as, for example, described by Huang and Ratnovksi (2008) should be significantly lower for most of the banks included in our sample. Altunbas et al. (2011) and Demirgüc-Kunt and Huizinga (2010), in contrast, focus on large and listed banks that primarily depend on wholesale funding. Bank risk also significantly differs across countries. We find evidence that aggregate credit growth is an important determinant of bank risk at the country level. This is consistent with the literature that shows that banks reduce their lending standards and collateral requirements during booms due to improved borrows income prospects, rising collateral values (Ruckes, 2004) and a reduction in information asymmetries (Dell Arrica and Marquez, 2006). In contrast to idiosyncratic risk that arises if single institutions reduce their lending standards, a general reduction of lending standards leads to the build-up of systemic risk in the banking sector that once it materializes affects all banks. Taking into account GDP growth, the level of interest rates as well as the size and the level of competition in the banking sector do not change our results. 4 Non-interest income includes activities such as income from trading and securitization, investment banking and advisory fees, brokerage commissions, venture capital, and fiduciary income, and gains on non-
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks