Interest Rate Pass-Through in Poland: Evidence from Individual Bank Data

Abstract

This paper employs individual bank data to analyze interest rate pass-through from money market rates to bank deposit and lending rates. First the speed and completeness of interest rate adjustment is assessed. As the sample covers the period prior to and after the outbreak of the financial crisis, some comparisons of the interest rate transmission process in these periods are made. The influence of individual bank characteristics (e.g., size, liquidity, strength of deposit base, quality of credit portfolio) on features of the interest rate transmission is also examined.

KEYWORDS:

• global financial crisis, interest rate channel
• interest rate pass-through
• monetary policy transmission mechanism

JEL Classification:

• E52
• E43
• G21

Opinions expressed in this paper are those of the author and should not be interpreted as the views of Narodowy Bank Polski. The author would like to thank A. Sznajderska, T. Łyziak, and participants of the Sixth Conference on Economic Challenges in Enlarged Europe, held on June 15–17, 2014, in Tallinn, Estonia, for their comments and discussions, and to N. Cieśla for explaining issues related to the dataset. The author is responsible for all mistakes and omissions.

Notes

¹ For a synthetic description of the structural characteristics of the Polish economy (and their development over time) that affect the monetary transmission mechanism, see Kapuściński et al. (2014).

² International comparisons are presented by Cottarelli and Kourelis (1994), Borio and Fritz (1995), Gigineishvili (2011), Saborowski and Weber (2013); results for the euro area as a whole or for its core economies might be found in Mojon (2000), de Bondt (2002), Sander and Kleimeier (2004), Sørensen and Werner (2006); for evidence for Central and Eastern European countries, see Sander and Kleimeier (2006), Egert et al. (2004, 2006); individual economies were analyzed by De Graeve et al. (2007)—Belgium; Frappa et al. (2008)—France; Hofmann and Mizen (2004)—UK; Schluter et al. (2012)—Germany; Montagnoli, Napolitano, and Siliverstovs (2012)—Italy; Wróbel and Pawłowska (2002), Chmielewski (2003)—Poland.

³ For discussion, see Mojon (2000) or de Bondt (2002).

⁴ Trade credit, corporate bonds, etc.

⁵ E.g., Jobst and Kwapil (2008), Hristov et al. (2012), Harbo Hansen and Weltz (2011), Illes and Lombardi (2013), ECB (2013).

⁶ The few months of the most severe distortions in the interbank market (September–December 2008) are excluded from the analysis.

⁷ For a detailed discussion of disturbances in the interest rate transmission process, see Demchuk et al. 2012.

⁸ This period is shorter than the one employed in the analysis of interest rate pass-through (January 2005–June 2013). However, as these characteristics are structural in nature, and only their mean values over time are needed, it is expected that this discrepancy does not matter.

⁹ The long-run multiplier for household credit rates is very small in the postcrisis sample. It is due to the fact that almost half of total credit to households is attributed to consumer credit, for which the cointegrating relation was broken after September 2008.

¹⁰ As a robustness check, a more general approach was employed in which both groups were allowed to differ in both the scale of the long-run pass-through and in the short-run adjustment dynamics. However, the results (available from the author on request) indicated that the impact of individual bank characteristics on the long-run relationship between retail and money market rates was very limited. This finding is in line with Weth (2002) and Gambacorta (2008).

¹¹ Employing a model with interaction terms would restrict the number of observations available, because some financial indicators investigated are reported only on a quarterly basis, while interest rate statistics are revealed each month. On the other hand, a small number of banks do not allow estimations of the heterogeneous panel model, as they are recommended for panels with large time and cross-section dimensions.

¹² There is also weak evidence (at a 10 percent significant level) of faster adjustment of the rate on total credit to firms in banks with a higher capital adequacy ratio.

¹³ Adjustment of this interest rate to the market rate changes also turns out to be dissimilar to other deposit rates when other bank characteristics are considered. However, their importance for transmission is negligible, as the share of deposit of this maturity in total household deposit is very low (about 10 percent).

¹⁴ Gambacorta (2004) found faster adjustment in bigger, more liquid, and better-capitalized banks in Italy if the influence of each characteristic was investigated separately, as in this paper. However, if all characteristics were considered jointly, more liquid and better-capitalized banks turned out to react less to market conditions and bank size became insignificant.

¹⁵ However, smaller long-run pass-through of lending rates might be attributed not only to disturbances on the money market, but also—at least partially—to asymmetry of interest rate adjustment with respect to interest rate increases and decreases.

Additional information

Notes on contributors

Ewa Stanisławska

Ewa Stanisławska is an economic expert at the Economic Institute, Narodowy Bank Polski, Warsaw, Poland, e-mail: [email protected].