The Transmission of Monetary Policy in South Africa Before and After the Global Financial Crisis

• Date: 01 December 2019
• DOI: http://doi.org/10.1111/saje.12238
• Author: Mpho Rapapali,Alain Kabundi
• Published date: 01 December 2019

© 2019 Economic Society of South Africa

South African Journal of Economics Vol. 87:4 December 2019

doi: 10.1111/saje.12238

464

THE TRANSMISSION OF MONETARY POLICY IN

SOUTH AFRICA BEFORE AND AFTER THE GLOBAL

FINANCIAL CRISIS

ALAIN KABUNDI†,* AND MPHO RAPAPALI‡

Abstract

This paper examines whether the transmission mechanism of monetary policy in South Africa has

changed after the global financial crisis (GFC). We use a Bayesian vector autoregressive (BVAR)

model with Minnesota priors and 15 monthly variables, extending the system of Christiano,

Eichenbaum, with Evans (1999). The benefit of the BVAR approach is that it can accommodate a

large cross section of variables without running out of degrees of freedom. To identify the change

in the transmission process, we divide the sample size into two subsamples, namely the pre-GFC

period (March 2001 to August 2008) and the post-GFC period (September 2008 to February

2016). The results indicate that a change in the transmission of monetary policy occurred after the

GFC. The magnitude of the effect of a monetary policy shock on output is considerably greater in

the pre-GFC period compared to the post-GFC period. Moreover, the impact of a policy shock on

inflation is not statistically significant in the post-GFC period. The variance decomposition shows

that the interest-rate channel has possibly weakened in the post-GFC period.

JEL Classification: C32, C54, E52

Keywords: monetary policy, BVAR, forecast error variance, Minnesota priors

1. INTRODUCTION

In the aftermath of the most recent global financial crisis (GFC), real short-term interest

rates were very low compared to historical levels. However, at the same time, the credit

extended to the private sector was declining gradually. The real ex ante repurchase rate

(repo rate) measured −0.2% on average over the period 2009 to 20161 while growth in

the credit extension to the private sector remained below 10% since 2010. This is in con-

trast with the period 2001 to 2008, when the low interest-rate environment saw credit

growth rise by 17% on average. Furthermore, private-sector investment and consump-

tion expenditure are still weak. The question then arises: has the transmission mechanism

of monetary policy changed since the GFC?

1 The ex ante real interest rate is estimated as the difference between the policy rate, i.e. the re-

purchase rate, and the two-year-ahead inflation expectations obtained from the Bureau for

Economic Research.

* Corresponding author: World Bank Group, Washington, DC 20433, USA. Tel.: +1 (202) 458

5663, Fax: +1 (202) 522 3563. E-mail: akabundi@worldbank.org

† World Bank Group

‡ South African Reserve Bank

South African Journal

of Economics

465South African Journal of Economics Vol. 87:4 December 2019

© 2019 Economic Society of South Africa

There are many plausible explanations as to why the low interest-rate environment has

not translated into higher credit growth, including structural bottlenecks to growth, weak

economic conditions and deleveraging by both households and firms. Another plausi-

ble explanation is the effect of the legislative and regulatory frameworks adopted before

and after the GFC, specifically the National Credit Act 34 of 2005 (NCA), which was

implemented in June 2007, and the Basel II regulations, announced worldwide in July

2006 and implemented in South Africa in January 2008. The NCA was introduced to

promote a fairer and more transparent credit market, protect consumers and their rights

in the credit market, and limit the cost of credit. Basel II was intended to amend the in-

ternational standards that controlled how much capital banks needed to hold in order to

guard against the financial and operational risks they faced. These rules sought to ensure

that banks with a large risk exposure held a greater amount of capital to ensure their sol-

vency and overall stability. The Basel III regulations were implemented in January 2013

to address excessive risk taking by banks and thus impose more restrictions on credit

lending. Finally, in March 2015, the Affordability Assessment Regulations (AARs) of the

National Credit Regulator (NCR) became effective. The purpose of these regulations was

to prevent reckless and indiscriminate lending by financial intermediaries.

Against this backdrop, this paper empirically investigates the validity of the hypothesis

that a change occurred in the transmission of monetary policy after the GFC. To identify

the change in the transmission, we divide the sample into two subsamples, namely the

pre-GFC period (March 2001 and August 2008) and the post-GFC period (September

2008 and February 2016). We focus soley on the interest rate channel of transmission and

only consider the period under the inflation-targeting regime. The monetary policy shock

is identified using the Bayesian vector autoregressive (BVAR) approach with Minnesota

priors and 15 monthly variables, expanding the system of Christiano et al. (1999). We

prefer the BVAR approach because of its ability to deal with a large panel of time series

without the risk of running out of degrees of freedom. When using a traditional vector

autoregressive (VAR) approach with 15 monthly variables, we need to estimate at least

225 parameters. This is commonly known as “the curse of dimensionality”. Without

imposing prior information, it would be hard to obtain precise estimates of these 225

parameters from a traditional VAR and thus features such as impulse response functions

(IRFs) and forecasts tend to be imprecisely estimated. It is therefore desirable to “shrink”

the parameters and Minnesota priors offer a way of doing so. The priors incorporate the

belief that more recent lag variables should provide more reliable information than dis-

tant ones and that own lag variables should explain more of the variation of a given vari-

able than other lag variables in each equation. This ensures shrinkage of the parameters

and decreases the risk of overfitting. Specifically, the BVAR approach via the imposition

of Minnesota priors turns “the curse of dimensionality” into a blessing and solves the

puzzling results which are common in traditional VAR approaches.

The monthly proxy of real economic activity is obtained from the nowcast of real

gross domestic product (GDP) growth proposed by Kabundi et al. (2016). The monetary

policy shock is approximated by a 100 basis points rise in the nominal repo rate. We then

assess the reaction of the real GDP growth rate and the headline inflation rate from the

entire sample, i.e. March 2001 to February 2016, and from each subsample. We use the

forecast error variance (FEV) to evaluate the effectiveness of the monetary policy shock