Monetary Policy in a Downturn: Are Financial Crises Special?

• Published date: 01 March 2014
• Author: Leonardo Gambacorta,Morten L. Bech,Enisse Kharroubi
• DOI: http://doi.org/10.1111/infi.12040
• Date: 01 March 2014

Monetary Policy in a Downturn:

Are Financial Crises Special?

Morten L. Bech, Leonardo Gambacorta

and Enisse Kharroubi

Bank for International Settlements

Abstract

This paper analyses the effectiveness of monetary policy during downturns

associated with financial crises. Based on a sample of 24 developed countries,

our empirical analysis suggests that monetary policy is l ess effective follow-

ing a financial crisis as the monetary transmission mechanism is part ially

impaired. In particular, our results suggest that the benefits of accommoda-

tive monetary policy during a downturn are elusive when the downturn is

associated with a financial crisis. In addition, we find that private sector

deleveraging during a downturn helps to induce a stronger recovery.

I. Introduction

Six years after the global financial crisis, significant parts of the world

economy have yet to recover despite the very accommodative policies of

many central banks. In this paper, we ask whether this outcome is surprising.

The authors wish to thank Philippe Aghion, Claudio Borio, Steve Cecchetti, Ramon

Moreno, Christian Upp er, Elo}d Takáts, Fabrizio Zampoll i and, in particu lar, the editor

Benn Steil and two anonymous ref erees for comments and suggest ions. Gabriele Gasp erini

provided excellent rese arch assistance. Th e opinions expresse d in this paper are th ose of the

authors only and do n ot necessarily refle ct those of the Bank for I nternational Sett lements.

International Finance 17:1, 2014: pp. 99–119

DOI: 10.1111/infi.12040

© 2014 John Wiley & Sons Ltd

In particular, we try to assess whether the impact of monetary policy in

periods that follow a financial crisis differs from its impact in periods that do

not. Monetary policy is often likened to ‘pushing on a string’,andifthe

string –the financial system –is broken, it seems plausible that the

transmission mechanism of monetary policy also becomes impaired to

some degree.

The impact of monetary policy following financial crises is a tr icky issue

to tackle empirically for a number of reasons. First, financial crises are

(fortunately) rare events, and, hence, the sample size is bound to be small

and the statistical inference weak. There are basically two approaches to

increasing the degrees of freedom. One can either take a long look back in

history at a limited number of countries, or one can look across a wider

range of countrie s over a more limited per iod of time.

1

Second, regardless of

approach, this type of analysis inherently involves comparing different

economies and episodes in some way. Here, we try to strike a balance by

focusing on a set of 24 developed countries with data going back to the mid‐

1960s.

Another issue is that monetary policy is not conducted in a vacuum but

instead responds to macroeconomic and financial developments. Hence,

endogeneity is an issue, and the identification strategy is important. We use

separation in time with the aim of distinguishing between cause and effect.

That is, we do not look at the effects of monetary policy over the entire

period following the financial crisis. Rather, we look at how the monetary

policy stance during the downturn affects the strength of the subsequent

recovery. In doing so, we diff erentiate betw een downturns th at occurred in

connection with a financial crisis and those that did not. We find that

accommodative monetary policy during a ‘normal’downturn leads to a

stronger recover y afterwards. How ever, this relationshi p is not evident f or

accommodative monetary policy during downturns associated with financial

crises. The result holds after controlling for the fiscal policy stance, real

exchange rate movements, and economic conditions in other countries.

As debt and leverage have been found to play a crucial part in the booms

prior to financial busts (Jorda et al. 2011), we also analyse whether delever-

aging during downturn episodes produces significant benefits during the

subsequent recover y. In the case of a financial crisis, we find that deleverag-

ing is beneficial for the su bsequent recover y.

1

Schularick and Taylor (2012) provi de an impressive exampl e of the former approach . The

authors study the b ehaviour of money, credit an d macroeconomic ind icators in a histor ical

dataset for 12 devel oped countries over the period 1870–2008. In particular, they document

that monetary policy responses t o financial crises h ave been more aggressi ve since 1945 but

that, despite these po licies, the outpu t costs of crises have remain ed large.

100 Morten L. Bech et al.

© 2014 John Wiley & Sons Ltd