Introduction

For the Netherlands there is some evidence on the bank-lending channel from both VAR and panel data analysis. VAR analysis generally indicates that the lending channel is not very relevant in the Netherlands from a macroeconomic viewpoint. According to Garretsen and Swank (1998), van Ees, Garretsen and Sterken (1999) and Kakes (2000) the lending channel is partly offset because banks use their holdings of securities as a buffer to shield their loan portfolios from negative monetary shocks. However, the analysis by De Bondt (1999) does not confirm this buffer function. The evidence from the relatively scarce panel data analysis of the cross-sectional differences in bank-lending behaviour is somewhat mixed for the Netherlands. De Bondt (2000) finds some evidence for a lending channel in the Netherlands, while Schuller (1998) does not.

The present chapter contributes to the latter type of empirical evidence of the bank-lending channel for the Netherlands, using individual bank data. Following Ehrmann et al. (chapter 14 in this volume), the investigation concerns the response of bank lending to monetary shocks, together with the influence on this response of bank size, liquidity and capitalisation. The contributions of the present chapter are the following. First, consolidated data representing the Dutch banking population are used on a quarterly basis. Previous studies for the Netherlands have analysed BankScope data, which is biased towards large-sized banks, is available only at an annual frequency, and covers unconsolidated data. Second, the present study extends the analysis to several segments of the bank credit market.