This paper examines how Basel III capital reforms affected bank lending in Ger- many. We focus on the increase of minimum risk-based capital requirements and the introduction of the leverage ratio. The announcement of stricter risk-based capital regulation significantly affected low capitalized banks. The impact depends on a bank's credit risk model, i.e. whether a bank applies the standardized approach (SA) or an internal ratings-based approach (IRBA) to determine risk weights. Low capitalized SA banks significantly cut lending whereas IRBA banks did not ad- just lending volumes. By contrast, low capitalized IRBA banks significantly in- creased collateralization while low capitalized SA banks adjusted collateralization only marginally. Moreover, the impact on SMEs and large companies also differs. In terms of lending, SMEs were affected more strongly, whilst in terms of collateralization the impact on large companies was bigger. The announcement of the leverage ratio had, however, a rather limited impact. We find some evidence that low capitalized banks reduced lending. Furthermore, low capitalized banks somewhat tightened collateral requirements, especially for large companies.

How does bank distress impact their customers' probability of default and trade credit availability? We address this question by looking at a unique sample of German firms from 2000 to 2011. We follow their firm-bank relationships through times of distress and crisis, featuring the different transmission of bank distress shocks into already weakened firm balance sheets. We find that a distressed bank bailout, which is subject to restructuring and deleveraging conditions, leads to a bank-induced increase of firms' probabilities of default. Moreover, bailouts tend to reduce trade credit availability and ultimately firms' sales. We further find that the direction and magnitude of the effects depends on firm quality and the relationship orientation of banks.

This paper contributes to the literature on early warning indicators by applying a Bayesian model averaging approach. Our analysis, based on Austrian data, is carried out in two steps: First, we construct a quarterly financial stress index (AFSI) quantifying the level of stress in the Austrian financial system. Second, we examine the predictive power of various indicators, as measured by their ability to forecast the AFSI. Our approach allows us to investigate a large number of indicators. The results show that excessive credit growth and high returns of banks' stocks are the best early warning indicators. Unstable funding (as measured by the loan to deposit ratio) also has a high predictive power.

This study analyzes the impact of bank relationships on a firm's borrowing costs. We find that a firm's borrowing costs decrease with relationship strength, proxied by the share of bank debt provided by the lender. Borrowing costs, however, rise with relationship length. While the increase over time is weak on average, bank-dependent borrowers face a substantial premium after several relationship years. Switching the lender initially leads to only a small price discount on average. However, the discount is considerable for borrowers that switch and had a strong relationship with their previous lender. Our results suggest that close lending relationships lead to benefits for the firm, but may also imply hold-up costs in the long term.

This paper’s objective is to contribute to the evolving field of macroprudential supervision in Austria in a twofold way: First, we construct an Austrian financial stress index (AFSI) that quantifies the level of stress in the Austrian financial market. Second, drawing on supervisory, market-based and macroeconomic data for the period from 2000 to 2012, we examine various indicators regarding their predictive power for this stress index. These indicators are categorized to cover the following six risk channels that affect financial stability: risk-bearing capacity, mispricing of risk, excessive growth, interconnectedness, concentration, and the macroeconomic environment and its outlook. In our empirical analysis we apply state-of-the-art econometrics including best subset selection, Kalman filters and model averaging. Our results indicate that, as risk channels, excessive growth, interconnectedness and mispricing of risk have the greatest influence on the AFSI. Furthermore, our findings lead to the conclusion that the complexities of risk channel interactions render univariate analysis and/or stand-alone models ineffective for financial stability-oriented policymaking. Instead, an integrated analysis of different indicators turns out to be the more promising approach when trying to identify the buildup of systemic risk.

This paper investigates contagion in the German interbank market under the assumption of a stochastic loss given default (LGD). We combine a unique data set about the LGD of interbank loans with detailed data about interbank exposures. We find that the frequency distribution of the LGD is markedly U-shaped. Our simulations show that contagion in the German interbank market may happen. For the point in time under consideration, the assumption of a stochastic LGD leads on average to a more fragile banking system than under the assumption of a constant LGD.

This paper provides new evidence that taxes affect capital structure choice, using a unique and comprehensive panel data set which covers 86,173 German non-financial firms over the years 1973-2008. Following the Graham methodology to simulate marginal tax rates, we find a statistically and economically significant positive relationship between the marginal tax benefit of debt (net and gross of investor taxes) and the debt ratio. A 10% increase in the net (gross) marginal tax benefit of debt causes a 1.5% (1.6%) increase in the debt ratio, ceteris paribus. The results are robust to various specifications like using changes in debt or debt to capital ratios. A significantly positive effect of taxes on the debt ratio can also be identified in a partial adjustment model.

This paper investigates contagion at the German interbank market under the assumption of a stochastic loss given default (LGD). We combine a unique data set about the LGD of interbank loans with data about interbank exposures. We find that the frequency distribution of the LGD is u-shaped. Under the assumption of a stochastic LGD, simulation results show a more fragile banking system than under the assumption of a constant LGD. There are three types of banks concerning their tendency to trigger contagion: banks with strongly varying impact, banks whose impact is relatively constant, and banks with no direct impact.

This study analyzes the circumstances under which firms choose to have a relationship lender and under which firms switch their relationship lender. Relationship lending is measured by the largest lender’s share of debt. Our study is based on a unique dataset for Germany with more than 13,000 observations. We find that young firms and well capitalized firms are more likely to concentrate their borrowing on one bank. These two groups of firms are also more likely to switch their relationship lender. Moreover, firms concentrate their borrowing more heavily on banks with a high share of core deposits. Small firms are also more likely to stay with relationship lenders with a high share of core deposits. Finally, the proportion of debt borrowed from the relationship lender is reduced if the relationship between the lender and the borrower has been close for several years. Our findings suggest that both the decision in favor of relationship lending and the decision to switch the relationship lender are made in such a way as to balance the potential benefits and costs of a relationship lender’s access to private information.

This is the first study which analyzes the effect of relationship lending on accounting conservatism for private firms, and how this affects corporate finance. Based on a database containing the financial accounting data of approx. 8,600 private German firms, we first show that relationship lending is positively associated with the level of accounting conservatism, even when controlling for leverage and other factors. Relationship lending is measured by the proportion of loans provided by the main lender. Accounting conservatism is proxied by cumulative non-operating accruals and a novel conservatism score, both of which tend to considerably reflect unconditional conservatism. Second, we find that the cost of debt increases with the level of accounting conservatism. Third, we observe that relationship lending positively affects the amount of trade credit. Suppliers seem to interpret the existence of a lending relationship as an indication of a firm's creditworthiness. The results suggest that unconditional conservatism goes hand in hand with lending relationships even controlling for taxation issues. Our findings match the theoretical literature on implicit lending contracts. This literature claims that relationship lenders attempt to retain superior information to realize information rents that compensate for losses when lending to young firms and to firms in financial distress. Unconditional accounting conservatism may be used as a tool to retain superior information and to enhance implicit contracting.

Kapitalstruktur erfolgreicher Venture-Capital-Investitionen: Empirische Evidenz für Deutschland Der vorliegende Beitrag untersucht, welche Finanzierungsinstrumente Venture-Capital-Gesellschaften in Deutschland einsetzen und welche Einflussfaktoren die Wahl der Finanzierungsinstrumente steuern. Die hierfür verwendete Stichprobe besteht aus 92 Finanzierungsfällen, die in einer eigenen Erhebung gesammelt wurden. Im Unterschied zu den USA werden wandelbare Finanzierungsinstrumente in Deutschland nur selten eingesetzt. Für die Einräumung einer Wandlungsmöglichkeit werden zudem in der Regel nicht, wie in den USA üblich, Wandelschuldverschreibungen oder wandelbare Vorzugsaktien gewählt, sondern stille Beteiligungen, die vertraglich erweitert werden. Die häufigsten Finanzierungsformen sind direkte Beteiligungen und eigenkapitalnahe Finanzierungsformen. Die Verwendung direkter Beteiligungen nimmt mit der Größe des finanzierten Unternehmens ab. Ein Einfluss der F&E-Intensität des Unternehmens auf die Verwendung direkter Beteiligungen kann nicht festgestellt werden. (JEL G24)

Relationship lending is a common practice in credit financing all over the world, notably also in the European Union, which has been assumed to be particularly beneficial for Small and Medium-Sized Enterprises (SMEs). During recent years, there has been the impression that relationship lending loses ground due to a change of the banks’ business models, which could ultimately yield to a worsening of the business environment for corporates and SMEs. In this study, we investigate the determinants of relationship lending for Germany, where relationship lending traditionally plays an important role. Compared to previous studies, we refer to much more comprehensive data with information on more than 16,000 firm-bank relationships. Our findings confirm the assumption that relationship lending seems to be an important pillar for economic growth and employment: We find that the firms that are most likely to contribute to (future) economic growth, namely small and R&D-intensive firms, tend to choose a relationship lender. The same is observed for firms of high credit quality, independent of their size or R&D intensity. Furthermore, we also observe that the importance of relationship lending did not decrease since the mid 1990s.

Analyses and empirical studies of the banking sector and financial stability are becoming ever more important. Research using microdata, in particular, has gained in international significance in the past few years. The Bundesbank collects microdata on German banks which are also used for research purposes. Since the early 1990s, the Bundesbank has been electronically storing these data in its prudential information system (Bankaufsichtliches Informationssystem, hereafter BAKIS). These data are of interest for microand macroprudential research of the Bundesbank for the following reasons.

This study analyzes the determinants of relationship banking in Germany and their impact on the firms' borrowing costs. The analysis is based on unique micro data comprising firms' balance sheet data and their lending relationships with banks. We find that the number of lending relationships decreases with the firms' size and their R&D-intensity. Moreover, we find that borrowing costs increase with the number of lending relationships, but that the advantage of concentrating the borrowing on one bank vanishes if the relationship lasts for a longer time.

Relationship lending is a common practice in credit financing all over the world, particularly in Germany. On the basis of a comprehensive data set comprising information on firm-bank relationships for more than 16,000 observations, this study analyses the determinants of relationship lending in Germany. We find that small, young and R&D-intensive firms tend to choose relationship lending. Furthermore, we find that firms with a higher creditworthiness are more likely to choose a relationship lender. We find that the importance of relationship lending stayed roughly constant since the mid 90s.