Loan supply and demand in Germany's three‐pillar banking system during the financial crisis

• Published date: 01 March 2018
• Date: 01 March 2018
• Author: Lina Zwick,Torsten Schmidt
• DOI: http://doi.org/10.1111/infi.12125

DOI: 10.1111/infi.12125

ORIGINAL ARTICLE

Loan supply and demand in Germany's three-pillar

banking system during the financial crisis

Torsten Schmidt

|

Lina Zwick

RWI - Leibniz Institute for Economic

Research, Essen, Germany

Correspondence

Lina Zwick, RWI - Leibniz Institute for

Economic Research, Hohenzollernstr.1-3

Essen 45128, Germany.

Email: lina.zwick@rwi-essen.de

Abstract

Internationally operating German banks had to depreciate

huge amounts of assets during the financial crisis while the

loan volume to the private sector declined simultaneously.

In this paper, we empirically analyse, for the whole banking

sector as well as for selected categories of banks, whether

the reduction in loan volume was supply-side-driven. The

results suggest that a credit crunch did not occur in Germany

during the recent economic crisis. Apart from substantial

policy interventions, our results offer two explanations:

First, excess demand for loans dropped, as firms reverted to

other external financing instruments, particularly industrial

bonds. Second, those banks operating at the regional level

only were not affected by the crisis. In this regard, the three-

pillar structure of the German banking system contributed to

stable credit financing, as banks differ in their operating

behaviours and in their customers.

1

|

INTRODUCTION

In the wake of the recent financial crisis, banks that had been highly active in the U.S. financial markets

were forced to depreciate large amounts of their assets. Especially large and systemically relevant

banks were affected, leading to credit crunches in many countries.

1

Germany was no exception: In

particular, big private banks and Landesbanken (big public banks) reported large losses due to such

impairments (Deutsche Bundesbank, 2009, p. 52). At the same time, bank lending weakened. The loan

volume from banks to the private sector even decreased beginning in the third quarter of 2009. As bank

credit still makes up the most important part of external financing of firms (Deutsche Bundesbank,

2012, p. 28), these developments evoked concerns that firms' funding activities could be constrained in

Germany as well.

International Finance. 2018;21:23–38. wileyonlinelibrary.com/journal/infi © 2018 John Wiley & Sons Ltd

|

23

The purpose of this paper is to investigate whether the reduction in loan volume was supply-side

driven, that is, whether a so-called credit crunch occurred in Germany during the financial crisis.

Recent studies show that loan supply differs between banks or banking groups. Sette and Gobbi (2015)

distinguish between transactional lenders and relationship lenders. While relationship lenders have

long-standing relationships with their clients, transactional lenders collaborate with their clients in the

respective business only. During the recent financial crisis, transactional lenders in Italy reduced their

loan supply more than did relationship lenders (Bolton, Freixas, Gambacorta, & Mistrulli, 2016; Sette

& Gobbi, 2015). In Germany, relationship banking is an important feature of all categories of banks,

but there might be differences as well. Therefore, we analyse demand and supply conditions in the

German loan market during the financial crisis for the main categories of banks.

The German banking system is co mposed of three pillars. The first pillar consists of c ommercial

banks, including big private b anks. Landesbanken and regionally operating saving sb anks constitute

the second pillar. Credit coop eratives and regional insti tutions of credit cooperati ves form the third

pillar. There is some empiric al evidence that banks in the se different categories rea ct differently to

changes in economic activit y (Behr, Norden, & Noth, 2013). T herefore, differences in the three

pillars might have helped sta bilize the overall banking sys tem during the crisis. This wou ld be an

interesting result from an interna tional perspective because, in co ntrast to Germany, in some

countries, savings banks w ere privatized, or other mea sures were implemented that strengthened

private banks (Behr & Schmidt , 2016).

The analysis requires a thorough distinction between both sides of the loan market, as a reduction in

loan volume can also be traced to demand-side factors. During times of recession, it is likely that firms

reduce their demand for loans, due to reduced investment activities. However, disentangling loan

supply and demand is quite challenging, as both sides of the market are unobservable. Generally, two

data-driven approaches in the literature are used to address this problem: The first is to use

microeconomic data, such as survey or firm-level data. The results of such surveys are typically

available without any substantial time lag. However, in many cases, the history covered by such

surveys is rather short. The other approach is to use macroeconomic time series data, which are

available at least at a quarterly frequency. One approach in particular, using macroeconomic data, that

is well established in the literature is to employ a (dynamic) disequilibrium model.

2

In this paper, we

follow the latter approach and use Bayesian inference to solve the model.

Regarding the German loan market during the financial crisis, some studies find supply-side

restrictions for large banks (Blaes, 2011; Rottmann & Wollmershäuser, 2010) and for retail lending

when savings banks are exposed to the crisis through their ownership of affected ‘Landesbanken’(Puri,

Rocholt, & Steffen, 2011), while an overall credit crunch is not identified (Erdogan, 2010). Using a

dynamic disequilibrium model to address the identification problem allows us to distinguish loan

demand and supply as well as analyse the respective influential factors. Compared to the approaches

followed in the aforementioned studies, the advantage of such a model is that crunch periods are

identified and demand and supply factors are determined simultaneously. Moreover, we use the same

specification for demand and supply, respectively, to estimate models for the main categories of banks

to reveal differences in these market segments.

Our results for the overal l banking sector indicate a hi gh probability of supply- side restrictions

in the credit market during the rec essions in 2001 and 2008. During th e Euro area crisis, the

probability of a supply-sid e restriction was elevated but still lower than during previous recessions.

If we examine the categories o f banks, our results reveal, pa rticularly for savings ba nks, credit

cooperatives, and regiona l institutions of credit coo peratives, no evidence for restrictions on loan

supply at all after 2008. Only for the hi ghly affected big private banks and Landesbanken do we find

some evidence of loan supply restr ictions during the financial cr isis and the Euro area crisis. This

24

|

SCHMIDT AND ZWICK