Besides the introduction and conclusion, this thesis is comprised of six independentchapters. In this thesis we provide an in-sample and out-of-sample assessment of howthe South African Reserve Bank (SARB) sets its policy rate, post 2000 inflation targetingregime, in the context of both linear and nonlinear Taylor-type rule models of monetarypolicy.<p.Chapter 2 provides the theoretical foundations and the case study discussion. Theliterature has shown that the Taylor (1993) rule has gone through many modificationssince the last decade of the 20th century. The modifications of the Taylor rule includeinterest rate smoothing, backward and forward looking versions, and nonlinearapproximations. Furthermore, there has been increasing debate on whether centralbanks should respond to asset prices and financial variables. Despite somedisagreements, economists seem to agree on the role of the financial market indetermining inflation and economic performance. As far as South Africa is concerned, astable financial system is one of the mandates of the central bank.
Chapter 3 discusses the research methods used in the thesis. First, the chapter provides an overview on the Hodrick-Prescott Filter used to detrend some series. Second, more focus is oriented on a class of estimators, used in this thesis, called Generalized Method of Moments (GMM) estimators. GMM is important in that it can be applied to several estimation contexts besides the linear model. In fact, GMM can provide a simple alternative to other estimators, especially when it is difficult to write down the maximum likelihood estimator.
Chapter 4 is aimed to provide the source of data, to show the transformation made tosome of them and to explore the data for preliminary results. The Augmented Dickey-Fuller (ADF), Phillips-Perron (PP), GLS transformed Dickey-Fuller (DFGLS) andKwiatkowski, et. Al. (KPSS) tests suggest that all the series follow a stationary process. The chapter also reveals that the financial conditions index measured as an equal weight average of its components yields a smallest AIC than other alternative suggested herein.
Furthermore, the chapter shows that the models that consider coincident business cycleindicator, rather than industrial production, perform better in terms of goodness of fit.Given the controversial debate on whether central banks should target asset prices foreconomic stability, chapter 5 investigates whether the SARB pays close attention toasset and financial markets in their policy decisions. The main findings are that theSARB policy-makers pay close attention to the financial conditions index when settinginterest rate. In the same chapter, it is also found that nonlinear Taylor rule improves itsperformance with the advent of the financial crisis, providing the best description of insample SARB interest rate setting behaviour. The 2007-2009 financial crisis witnesses an overall increased reaction to inflation and financial conditions. In addition, the financial crisis saw a shift from output stabilisation to inflation targeting and a shift, from a symmetric policy response to financial conditions, to a more asymmetric responsedepending on the state of the economy. Although one could have expected that theSARBs response of monetary policy to output during the crisis to increase, theresponse has dropped significantly. These results show the concern over the high levelof inflation observed during the second semester of 2008.<p.In chapter 6, we test the concept of Opportunistic Approach to monetary policy. Thefindings support the two features of the opportunistic approach. First, we find that themodels that include an intermediate target that reflects the recent history of inflationrather than a simple inflation target improve the fit of the models. Second, the datasupports the view that the South African Reserve Bank (SARB) behaves with somedegree of non-responsiveness when inflation is within the zone of discretion but reactaggressively otherwise. Recursive estimates from the preferred model reveal that overallthere has been a subdued reaction to inflation, output and financial conditions amidstthe increased economic uncertainty of the 2007-2009 financial crisis.
Chapter 7 compares forecast performance of linear and nonlinear monetary policy rulesestimated in the two previous chapters but rewritten in their backward looking versions.Recursive forecasts values are computed for 1- to 12-step ahead for the out-of-sampleperiod 2006:01 to 2010:12. For the nonlinear models we use bootstrap method formulti-step ahead forecasts as opposed to point forecasts approach used for linearmodels. The aim is to evaluate the performance of three competing models in an outof-sample forecasting exercise. Overall ranking reveals the superiority of the nonlinearmodel that distinguishes between downward and upward movements in the businesscycles in closely matching the historical record. As such, forecasting performance testsreveal that the SARB pays particular attention to business cycles movements whensetting its policy rate.
© 2011 Author. All rights reserved. The copyright in this work vests in the author. No part of this work may be reproduced or transmitted in any form or by any means, without the prior written permission of the author.
Please cite as follows:
Kasai, N 2011, Analysis of monetary policy rules for South Africa, PhD thesis, University of Pretoria, Pretoria, viewed yymmdd < http://upetd.up.ac.za/thesis/available/etd-10132011-155515/ >
D11/9/74/ag
